ann. behav. med.

(2017) 51 (Suppl 1):S1S2867

DOI 10.1007/s12160-017-9903-3

Seminar 1 9:00 AM-3:00 PM

THE NUTS AND BOLTS OF BEHAVIORAL INTERVENTION DEVELOPMENT: STUDY DESIGNS,

METHODS AND FUNDING OPPORTUNITIES

Susan M. Czajkowski, Ph.D.1, Sylvie Naar, PhD2, Deborah Ellis, PhD2, Elizabeth Towner, PhD2,
Lynda Powell, PhD3, Walter Dempsey, Ph.D. (Statistics); B.S (Math w/ Honors); B.A.
(Economics, Statistics) all at the University of Chicago4, Jesse Dallery, Ph.D.5, Kenneth E.
Freedland, PhD6, Christine M. Hunter, PhD7, Frank Perna, EdD, PhD8

1
National Cancer Institute, NIH, Bethesda, MD; 2Wayne State University, Detroit, MI; 3Rush
University Medical Center, Chicago, IL; 4University of Michigan, Ann Arbor, MI; 5University of
Florida, Gainesville, FL; 6Washington University School of Medicine, St Louis, MO; 7Division of
Diabetes, Endocrinology, and Metabolic Diseases, Bethesda, MD; 8National Cancer Institute,
Rockville, MD

This seminar will provide investigators who are interested in the design and preliminary
testing of health-related behavioral interventions an opportunity to (1) learn about a new
framework for behavioral treatment development -- the ORBIT model (see
http://psycnet.apa.org/psycinfo/2015-03938-001/); (2) learn about appropriate study designs
and methods for early-phase behavioral intervention research, including which methods are
appropriate at each phase of the behavioral intervention development process; (3) apply the
ORBIT model and knowledge about relevant methodologies to their own behavioral
treatment development projects; and (4) identify early-phase translational research funding
opportunities and develop grant applications to support intervention development
research. The format will include didactic presentations, question and answer sessions, and
small group discussion in which participants will be provided with advice to help them design
their own behavioral intervention development project. NIH scientists will provide
information on suitable funding opportunity announcements and advice on how to develop
applications for NIH support. Extramural behavioral scientists will describe their own
experiences in designing behavioral intervention development studies, bringing these lessons
learned to bear in advising seminar attendees on their individual projects.

Didactic presentations will provide detailed information about methodologies and study
designs most applicable to the early phases of behavioral intervention development (e.g.,
qualitative research, single-case designs, dose-finding studies, adaptive and fractional factorial
designs, and pilot and feasibility studies), with ample time allotted for questions and
discussion. In addition, experiential and small-group activities will deepen participants
knowledge of and skills needed for designing a behavioral intervention development program,
allowing time for tailored advice and feedback. Participants will also be asked to submit in

# The Society of Behavioral Medicine 2017

S2 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

advance a 1-2 page synopsis (e.g., abstract, research aims, hypotheses, methods) of a
behavioral intervention development project, which can be one they are considering
submitting for funding, along with specific questions they may have regarding the process of
behavioral intervention development. These will be discussed in small groups led by NIH staff
and senior investigators, allowing participants to receive detailed feedback and advice to
enhance the quality of their grant applications for designing and optimizing behavioral
interventions.

CORRESPONDING AUTHOR: Susan M. Czajkowski, Ph.D., National Cancer Institute, NIH,

Bethesda, MD, 20892; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S3

Seminar 2 9:00 AM-3:00 PM

DETERMINING THE APPROPRIATE MIND-BODY MODALITY FOR GREATER CLINICAL OUTCOMES

Brad S. Lichtenstein, Doctorate of Naturopathic Medicine

Bastyr University, Lake Forest Park, WA

Mindfulness and mindfulness-based approaches have increased in popularity over the past
decade as useful modalities for stress reduction. However, it is important to remember that
mindfulness is only one of many self-soothing mind-body medicine strategies. Unfortunately,
researchers and clinicians typically focus on one particular method, failing to address
variations in individual stress response patterns. Acknowledging the necessity for a person-
centered approach to care, determining which of the numerous mind-body approaches that is
best suited for the individual is paramount. To address this issue, Smith proposed a self-stress
theory recognizing that individuals maintain heightened nervous system arousal by
responding to stressors in one of six particular psycho-physiological patterns. Each of these
patterns, interestingly, corresponds to a group of mind-body techniques. To obtain optimal
improvement, selecting the mind-body strategy that best matches the self-stressed target is
imperative. This does not suggest that only one mind-body approach alone should be
reserved for one particular stressed response target. Instead, patterns can be considered
guidepost to navigate the field of mind-body practices.

This workshop will introduce 6 self-stress psycho-physiological patterns (posture, muscles,

breathing, body focus, emotions, and attention) and how to identify them. Next the
corresponding mind-body medicine techniques will be examined and experienced. The mind-
body modalities covered will include: 1) movement/therapeutic exercise/yoga, 2) progressive
muscle relaxation, 3) breathwork, 4) autogenic training, 5) guided imagery/hypnosis/self-talk,
and 6) mindfulness.

CORRESPONDING AUTHOR: Brad S. Lichtenstein, Doctorate of Naturopathic Medicine , Bastyr

University, Lake Forest Park, WA, 98155; [email protected]
S4 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Seminar 3 9:00 AM-3:00 PM

EVALUATING MEASURES USING ITEM RESPONSE THEORY

Ruben Castaneda, NA

University of California, Merced, Merced, CA

Eager to roll out a research project, researchers often overlook the importance of
proper measurement. Taking a measure at face value may be tempting, but improper
measures may result in a low powered study at best, and reaching a completely wrong
conclusion at worst. This course will introduce you to topics in IRT. I will go over the basics of
measurement starting from selecting a proper model, evaluating the impact of each item and
evaluating the measure as a whole. We will go over strategies for getting the most out of your
measure using data in the behavioral sciences.

CORRESPONDING AUTHOR: Ruben Castaneda, NA, University of California, Merced, Merced,

CA, 95340; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S5

Seminar 5 9:00 AM-3:00 PM

THE NUTS AND BOLTS OF DEVELOPING, TESTING, & IMPLEMENTING DYADIC INTERVENTIONS
FOR CANCER PATIENTS & THEIR INFORMAL CAREGIVERS

Hoda Badr, PHD1, Erin Kent, PhD2, Wendy Demark-Wahnefried, PhD RD3, Betty Ferrell, PhD,
MA, FAAN, FPCN4, Karen Kayser, PhD, MSW5, Jean-Philippe Laurenceau, Ph.D.6, Matthew
Loscalzo, LCSW7, Julia H. Rowland, PhD8, Tracey A. Revenson, PhD9, Deborah A. Kashy, Ph.D.10

1
Baylor College of Medicine, Houston, TX; 2National Cancer Institute, Rockville, MD; 3UAB,
Birmingham, AL; 4City of Hope, Duarte, CA; 5University of Louisville, Louisville, KY; 6University
of Delaware, Newark, DE; 7City Of Hope-National Medical Center, duarte, CA; 8National
Cancer Institute, Bethesda, MD; 9Hunter College, New York, NY; 10Michigan State University,
East Lansing, MI

Traditional approaches for addressing health outcomes and health behaviors after the
diagnosis of cancer have focused on either the patient or his/her primary caregiver (often the
patients partner or another close family member). However, cancer often exacts a heavy toll
on both patients and their caregivers, as well as their relationships by challenging established
communication patterns, roles, and responsibilities. Interest in the development of
psychosocial and behavioral interventions that address the patient-caregiver dyad as a unit is
growing. These dyadic interventions often seek to leverage the relationship between patient
and caregiver to affect behavioral change or to enhance the quality of life of one or both
partners. Despite growing interest in dyads as a promising focus of intervention, little
guidance regarding the development and design of dyadic interventions or how to implement
and disseminate them in oncology clinical care settings is currently available.

This seminar will provide attendees an opportunity to:

1. Learn about the growing field of dyadic interventions in cancer, including

accomplishments and remaining research/practice knowledge gaps, theoretical
models, and therapeutic approaches;
2. Examine appropriate study designs for different phases of the intervention
development and testing process (e.g., qualitative research, dose-finding studies, pilot
studies, RCTs, and effectiveness-implementation hybrid designs) as well as approaches
for analyzing dyadic data;
3. Identify methodological challenges involved in dyadic research (e.g.,
recruitment/retention issues, choice of control group, tailoring for the individual vs
dyad);
4. Discuss challenges related to clinical uptake/implementation;
S6 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

5. Apply knowledge gained about relevant approaches and methodologies to their own
research questions.

The format will include didactic presentations from NIH and extramural researchers, question
and answer sessions, and small group activities/discussions in which participants will be
provided with advice to help them in designing their own dyadic intervention project.
Throughout the workshop, the emphasis will be on designing programs with an eye toward
implementation. Toward that end, presenters will describe their own experiences in
developing, testing, and implementing different types of dyadic interventions, bringing these
lessons learned to bear in advising seminar attendees on their individual projects. In
addition, experiential and small-group activities will deepen participants knowledge and skills
needed for designing dyadic interventions, allowing time for tailored advice and feedback
from both the presenters and the group. Funding opportunities will also be discussed.

CORRESPONDING AUTHOR: Hoda Badr, PHD, Baylor College of Medicine, Houston, TX, 77021;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S7

Seminar 6 9:00 AM-11:45 AM

THE FOUNDATIONS OF IMPLEMENTATION SCIENCE: TOPICS, OUTCOMES, AND METHODS

Paul A. Estabrooks, PhD1, Samantha Harden, PhD2, Mark Stoutenberg, PhD, MSPH3

1
University of Nebraska Medical Center, Omaha, NE; 2Virginia Tech, Blacksburg, VA;
3
University of Miami, Miami Beach, FL

Implementation science is generally defined as the study of methods to promote the

integration of research findings and evidence-based interventions within healthcare policy
and practicethough the methods and scientific contributions for implementation science
can also be generated in community and other organizational contexts outside of healthcare
systems. The National Institutes of Health Dissemination and Implementation Research in
Health program announcement describes research objectives that include supporting
innovative approaches to identify, understand, and overcome barriers to the adoption,
adaptation, integration, scale-up and sustainability of evidence-based interventions, tools,
policies, and guidelines. Unfortunately, scientists that focus on implementation research are
confronted by the challenge of identifying epistemological approaches, developing research
methods that balance internal and external validity in their work, and attempting to speed the
translation of evidence into clinical and community practice settings. The purpose of this
workshop will be to review implementation science theories, research methods, and
outcomes and provide opportunities to apply these concepts to an ongoing program of
research. Specifically, Dr. Mark Stoutenberg will provide an overview of implementation
science definitions and concepts from the perspective of addressing the health of
underserved populations in Miami. He will also review the Consolidated Framework for
Implementation Research and highlight advances in measurement and tools to assess
implementation initiation and sustainability. Dr. Samantha Harden will present on
implementation science outcomes, including a focus on the RE-AIM framework and
applications for physical activity promotion. In addition, she will focus on pragmatic
measurement of implementation outcomes as they relate to adaptation over time. Dr. Paul
Estabrooks will provide examples of hybrid effectiveness implementation designs developed
to reduce the delay between the identification of efficacious clinical approaches and
translation into typical practice. He will provide examples of HEI designs that blend
effectiveness and implementation research into three distinct, but overlapping trial types.
This blending may lead to confusion in that often there may be 2 distinct levels of scientific
manipulationone, at the level of the patient or participant that is intended to change a
behavioral or health outcome and one at the level of the practitioner that is intended to
improve the implementation quality of the clinical practice. Finally, Drs. Stoutenberg,
Estabrooks, and Harden will lead workshop participants through key considerations in
S8 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

developing an implementation science grant proposal using sound theory, methods, and
outcomes.

CORRESPONDING AUTHOR: Paul A. Estabrooks, PhD, University of Nebraska Medical Center,

Omaha, NE, 68198-4365; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S9

Seminar 7 9:00 AM-11:45 AM

WHAT YOU NEED TO KNOW ABOUT WRITING GRANTS: TIPS AND TECHNIQUES FROM EXPERTS

Lara Traeger, PhD1, Felicity Harper, PhD2, Jamilia Sly, Ph.D.3, Qian Lu, Ph.D.4

1
Massachusetts General Hospital/Harvard Medical School, Boston, MA; 2Karmanos Cancer
Institute/Wayne State University, Detroit, MI; 3Icahn School of Medicine at Mount Sinai, New
York, NY; 4University of Houston, Houston, TX

This session focuses on how to integrate successful grantsmanship strategies into your career.
First, a panel of experts in cancer and ethnic minority & multicultural health research will
address key tips for developing grant proposals, with specific attention to NIH R03s, R21s,
R01s and similar mechanisms. Then, we will demonstrate a study section review for two
previously-selected proposals. Finally, we will host three roundtables on 1) balancing grant
writing with other work demands, 2) interpreting grant critiques: when to keep trying and
when start fresh, and 3) lessons learned: rookie mistakes and strategies for success.
Attendees will be able to discuss their questions with diverse clinical research experts;
observe the study section process; gain insight into research career considerations; and
participate in the roundtables.

CORRESPONDING AUTHOR: Lara Traeger, PhD, Massachusetts General Hospital/Harvard

Medical School, Boston, MA, 02114; [email protected]
S10 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Seminar 9 9:00 AM-11:45 AM

THE NEW ACCEPTANCE AND COMMITMENT THERAPY APPROACH FOR CHRONIC PAIN AND
HEADACHE

Maria Karekla, Ph.D.

University of Cyprus, Nicosia, Nicosia, Cyprus

The application of Acceptance and Commitment Therapy (ACT) for the treatment of chronic
pain is well established and indeed ACT is considered an empirically supported treatment for
such chronic and debilitating problems. This workshop will present the latest developments in
ACT manuals for the treatment of chronic pain and headache for increasing well-being of
sufferers. The latest application of these new protocols will be discussed based on findings
from the European Union funded research project ALGEA (the Greek word for suffering).
The main aim of the project was to examine the efficacy of ACT for the treatment of chronic
pain and examine the mechanisms and processes of change in this approach. Concepts will be
illustrated using live demonstrations, experiential exercises, metaphors, and worksheets. This
workshop is designed to teach skills needed to explore ACT as an assessment model and
intervention method for chronic pain. It will be mostly experiential and will balance an
understanding of the model with a personal connection with the issues raised in ACT, and
with skill development.

CORRESPONDING AUTHOR: Maria Karekla, Ph.D., University of Cyprus, Nicosia, Nicosia, 1678;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S11

Seminar 10 9:00 AM-11:45 AM

HARD TO MAINTAIN BEHAVIOR CHANGE: AN INTRODUCTION TO CEOS THEORY

Ron Borland, PhD

Cancer Council Victoria, Melbourne, Victoria, Australia

This seminar provides a brief introduction to CEOS theory, with a focus on how it
reconceptualises a range of challenging issues within behavioral medicine. The theory is
elaborated in Borland, Understanding Hard to Maintain Behaviour Change: A Dual Process
Approach. Oxford, Wiley-Blackwell. 2014). Behaviour is theorised to be the result of the
moment by moment interaction between internal needs (operational processes) in relation to
environmental conditions, and for humans this is augmented by goal-directed, executive
action which can overcome immediate contingencies to create net affective force for action
which is the end pathway to action. This conceptualisation postulates a key role for emotions
as signals to executive processes about internal states and relationships with the context.
Affective force can come from memories of direct experience, vicarious experience,
emotionally-charged communications mediated through stories the person generates and are
available to them at the point of action. Executive behaviour, ie, goal seeking, can only occur
when the operational processes have been activated to support it, and this requires
overcoming any competing action tendencies. This seminar outlines the key postulates of the
theory, then shows via examples taken from the authors 30 years of experience in tobacco
control, how the theory helps make sense and provide new insights around how to
communicate most effectively, and how to facilitate self-regulatory processes, which it
conceives as much more than self-control. It also will show how the theory provides a
framework for understanding the contributions and limitations of virtually all existing
behaviour change techniques. It is hoped attendees will come away with a heightened
understanding of the potential of dual process approaches to behaviour and behaviour
change and new insights into their areas of practice and research.

CORRESPONDING AUTHOR: Ron Borland, PhD, Cancer Council Victoria, Melbourne, Victoria,
3004; [email protected]
S12 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Seminar 11 9:00 AM-11:45 AM

DYNAMIC SYSTEMS MODELING FOR HEALTH IN A SOCIAL CONTEXT

Jonathan E. Butner, Ph.D., Cynthia Berg, Ph.D.

University of Utah, Salt Lake City, UT

Real-time, daily diary, longitudinal methodologies to health research have generated a boon
of data through time. For example, intensive measurement has been applied to health
behaviors broadly (e.g., healthy eating, exercise, smoking) and for managing chronic illnesses
(adherence, coping). Such data are complicated to analyze in a coherent, theory-driven
manner. This seminar examines dynamical systems modeling with the goal of preparing
participants to apply these techniques in their own research.

Attractor dynamics depict health data over time as a function of temporal patterning within a
larger system. For example, many individuals with a chronic illness coordinate multiple
behaviors numerous times in a day to promote adherence. Such behaviors occur within
multiple social relationships (parents, friends, spouses) that can support or interfere with
adherence behaviors. Dynamic systems models capture how these elements may be
coordinated to either maintain homeostasis or create changes in how health behaviors are
regulated.

Dynamic systems modeling captures the patterning among these elements of the system. The
simplest of these patterns is a homeostatic one, where stability is examined as to how quickly
the system returns to homeostasis after being perturbed. This simple representation can be
captured within regression for a single time series, and multilevel modeling for multiple time
series, by treating change in a variable as the dependent variable and the current value of the
variable as the predictor.

Moving from a single outcome in time to two simultaneous outcomes expands homeostasis to
include cyclical patternings. Such is the case when we examine the relationship of two
behaviors (e.g., adherence and support of a close relationship). Analytically, this takes the
regression and multilevel models and expands them to a form of the actor-partner model.

Some discussion will be provided of further expansions including continuous time model
equivalents, models where there are more than one underlying temporal pattern (e.g. two
different homoeostatic points, moving between them), and more complex patterning.
Examples and code will be illustrated in SPSS and R.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S13

The advantage of these models is that they can identify factors that may drive change toward
better or worse health behaviors and capture the dynamic nature of these behaviors within
the social context.

CORRESPONDING AUTHOR: Jonathan E. Butner, Ph.D., University of Utah, Salt Lake City, UT,
84112; [email protected]
S14 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Seminar 13 12:15 PM-3:00 PM

ADVERSE CHILDHOOD EXPERIENCES AND ADULT HEALTH OUTCOMES: EPIGENETICS,
PSYCHONEUROIMMUNOLOGY, AND BRAIN-BASED THERAPY

John B. Arden, PhD, ABPP

Institute of Brain Potential, Sebastopol, CA

The Adverse Childhood Experiences (ACEs) seminar presents a wide variety of factors having
multidirectional causal relationships between ACEs: resulting in stress, depression, anxiety,
the immune system, and gene expression. The interaction between all these factors has been
by illuminated by studies examining ACEs and the effects of life style factors on the incidence
of health and psychological problems. There are significant relationships between immune
system function, stress, insecure attachment, anxiety, depression, poor nutrition, bad quality
sleep, physical inactivity, and neurophysiological dysregulation. For example, insecure
attachment, deprivation, and child abuse contribute to anxiety and depression in far more
extensive ways then was believed. Some of these dysregulating effects includes the
epigenetic down regulation of the cortisol receptors, the stress induced activation of the
sympathetic nervous system, and the breakdown of the neuroendocrine system. The
excessive release of cortisol can result in the eventual development of hypocortisolism with
the simultaneous excessive levels of epinephrine and norepinephrine which in turn, stimulate
the release of proteins called pro-inflammatory cytokines. Chronically high levels of these
cytokines adversely affect the central nervous system, resulting in autoimmune diseases, as
well as physiological symptoms including lethargy, achiness, and disturbances in mood,
cognition, and promoting withdrawal behaviors that contribute to major depression and
anxiety. This seminar will address recovery from all these problems.

CORRESPONDING AUTHOR: John B. Arden, PhD, ABPP, Institute of Brain Potential, Sebastopol,
CA, 95472; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S15

Seminar 14 3:15 PM-6:00 PM

APPLICATION OF BEHAVIOR CHANGE PRINCIPLES IN APP-BASED, MOBILE HEALTH

INTERVENTIONS

Kirstin Aschbacher, Ph.D. & Licensed Psychologist

UCSF/ Jawbone, San Francisco, CA

This seminar will introduce you to the skills needed to translate empirically supported
behavior change techniques to a mobile health, or App-based setting. You will learn through
concrete examples of how lifestyle & weight loss interventions have been designed, deployed,
and tested in the context of a major Silicon Valley companys App and wearables.

The facilitator holds experience as a licensed Clinical Psychologist, an Assistant Professor at

University of California, San Francisco (UCSF), and a Behavioral Data Scientist employed with a
major Silicon Valley company. This seminar will discuss the challenges of deploying
interventions in an integrated system, which includes an App and a tracking band, united by
machine learning algorithms. Key objectives of the system are to provide users with
personalized behavior change strategies.

The seminar will be geared so as to benefit both trainees and established investigators. The
facilitator will discuss specific examples of how he/she developed and implemented
longitudinal weight loss interventions in an industry setting. These examples will provide you
with content knowledge, including: 1) Tips for translating behavior change principles to App
settings, 2) Best practices in experimental design for A/B tests (or App-based randomized
controlled trials (RCTs)), and 3) Key considerations in the analysis of your results: The pros
and cons of machine learning versus traditional statistical techniques. You will also gain
process knowledge, including: 1) What to focus on when pitching collaborations with
industry, 2) How scale changes everything, and 3) Using a Minimum Viable Product (MVP)
orientation to rapidly scale mini-interventions to larger interventions. Finally, attendees will
obtain an overview of the terminology and concepts needed to bridge the gap and develop
productive collaborations between academia and industry.

CORRESPONDING AUTHOR: Kirstin Aschbacher, Ph.D. & Licensed Psychologist, UCSF/

Jawbone, San Francisco, CA, 94117; [email protected]
S16 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Seminar 15 3:15 PM-6:00 PM

PUTTING PATIENT DATA INTO ACTION: THE USE OF PATIENT-GENERATED HEALTH DATA IN
CLINICAL CARE AND RESEARCH

Robin R. Austin, DNP, DC, RN-BC1, Katherine Kim, PhD, MPH, MBA 2, Ellen Beckjord, PhD,
MPH3, Bradford W. Hesse, PhD4, Ashley Wilder Smith, PHD, MPH5, Sherri Sheinfeld Gorin,
PhD, FSBM6, Julie A. Wright, PhD7, Jessica Y. Breland, PhD8, Miho Tanaka, PhD, MPH9, Pei-Yun
Sabrina Hsueh, PhD10

1
University of Minnesota, School of Nursing, Minneapolis, MN; 2UC Davis, Sacramento, CA;
3
UPMC Health Plan, Pittsburgh, PA; 4National Cancer Institute, Rockville, MD; 5National
Cancer Institute, Bethesda, MD; 6New York Physicians Against Cancer, http://nypac.center/,
New York, NY; 7UMass Boston, Boston, MA; 8VA Palo Alto & Stanford University School of
Medicine, Menlo Park, CA; 9Department of Veterans Affairs, Falls Church, VA; 10IBM - Thomas
J. Watson Research Center, Yorktown Heights, NY

Patient generated health data (PGHD), produced either directly through patient reported
outcome tools or indirectly through passive sensing devices, has become an important
behavioral consideration in healthcare. According to the Department of Health and Human
Services, PGHD can provide information about how patients are doing between clinical visits;
it can be gathered on an ongoing, longitudinal basis; it can provide relevant information for
preventive and chronic care management; and it can improve patient safety through ongoing
monitoring of adverse events during treatment. PGHDs are increasing in recognition and
usage across health systems. The focus of this preconference is designed to inform, educate,
and update colleagues on PGHD used for behavior medicine research. We will present the
current state-of-the-science and related evidence for PGHD. This preconference will explore
three areas: (1) presentations from individuals currently working with PGHD for research; (2)
presentations from individuals currently working to implement PGHD within clinical settings
and systems for use by providers and clinicians; and (3) discussions (in small groups, led by the
presenters) about the potential barriers to implementing the use of PGHD and about
innovative solutions to these PGHD barriers. Participants will also be able to work in small
groups to develop strategies for implementing and using PGHD in their own healthcare
settings. Through this workshop, participants will be able to engage in deeper discussion and
dialogue with leaders in the field as well as fellow participants related to best practices and
barriers to the use of PGHD in research, clinical care, and policy.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S17

Goals/Objectives for the Pre-conference:

Understand the use and current state of patient generated health data
Describe current trends and opportunities for research using patient generated health
data
Identify best practices in and potential barriers to the use of patient generated health
data.

CORRESPONDING AUTHOR: Robin R. Austin, DNP, DC, RN-BC, University of Minnesota, School
of Nursing, Minneapolis, MN, 55455; [email protected]
S18 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Seminar 16 3:15 PM-6:00 PM

HOW CAN SOCIAL NETWORK ANALYSIS HELP ME UNDERSTAND THE INFLUENCE THAT THE
FAMILY ENVIRONMENT HAS ON INDIVIDUAL HEALTH BEHAVIORS?

Enrique Ortega, PhD / MPH1, Emily Coviello, B.S.2, Karen Carranza, BS3

1
CSU Dominguez Hills, Alhambra, CA; 2California State University Dominguez Hills, Alhambra,
CA; 3California State University Dominguez Hills, Lawndale, CA

Research in the health sciences has traditionally focused on assessing individual attributes to
understand and predict individual health outcomes. Social Network Analysis (SNA) methods
allow investigators to inquire how such individual attributes develop and become normative
by considering the individual within a social context. This seminar will discuss and
demonstrate how to apply SNA assessment methods to the field of family dynamics in order
to investigate the potential effects that family social relationships may have on individual
health decision making processes, and the evolvement of individual health practices.

SNA involves the study of how individuals monitor their social networks and make use of the
information presented within these networks to modify, among other aspects, their health
attitudes and behaviors. SNA research conceptualizes the individuals environment as a
network to understand the reference points that an individual faces in their daily lives. SNA
research allows researchers to describe the patterns of interactions that an individual
perceives within their social network and to investigate how these interactions may affect
their health choices. These research domains are important for investigators interested in
assessing group health factors and outcomes such as those possibly tied to community and
public health efforts. Social network analyses investigations are currently benefiting from
advances in computer science and graphing technologies that allow researchers to visually
depict the social networks of individuals and how networks influence individual conditions.
While such visual depictions are at the forefront of other fields such as geography and
biology, applications to the field of behavioral health can allow social health scientists to
further understand the links between individual health and its link with group health
patterns.

The present seminar will discuss SNA methods of assessment to understand how individuals
perceive and cognitively represent the family networks that surround them; and how these
perceptions inform their individual health knowledge and behaviors. The application of SNA
methods will be discussed and demonstrated through 3 case studies which have utilized such
methods to inquire how family network dynamics inform the development of health
promotion practices, emotional regulation, and healthy lifestyle choices.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S19

CORRESPONDING AUTHOR: Enrique Ortega, PhD / MPH, CSU Dominguez Hills, Alhambra, CA,
91803; [email protected]
S20 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Seminar 17 3:15 PM-6:00 PM

VARIANCE MODELING OF ECOLOGICAL MOMENTARY ASSESSMENT (EMA) DATA

Donald Hedeker, Ph.D.

University of Chicago, Chicago, IL

For longitudinal data, mixed models include random subject effects to indicate how subjects
influence their responses over the repeated assessments. The error variance and the variance
of the random effects are usually considered to be homogeneous. These variance terms
characterize the within-subjects (error variance) and between-subjects (random-effects
variance) variation in the data. In studies using Ecological Momentary Assessment (EMA), up
to thirty or forty observations are often obtained for each subject, and interest frequently
centers around changes in the variances, both within- and between-subjects. Also, such EMA
studies often include several waves of data collection. In this seminar, we focus on an
adolescent smoking study using EMA at both one and several measurement waves, where
interest is on characterizing changes in mood variation associated with smoking. We describe
how covariates can influence the mood variances, and also describe an extension of the
standard mixed model by adding a subject-level random effect to the within-subject variance
specification. This permits subjects to have influence on the mean, or location, and variability,
or (square of the) scale, of their mood responses. Additionally, we allow the location and scale
random effects to be correlated. These mixed-effects location scale models have useful
applications in many research areas where interest centers on the joint modeling of the mean
and variance structure. Computer application using SAS NLMIXED and the freeware
MIXREGLS program will be described and illustrated.

CORRESPONDING AUTHOR: Donald Hedeker, Ph.D. , University of Chicago, Chicago, IL, 60637;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S21

Wednesday

March 29, 2017

6:00 PM-7:00 PM
S22 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A001 6:00 PM-7:00 PM

60 MINUTES FOR HEALTH: PILOTING AN INNOVATIVE BEHAVIORAL INTERVENTION

PROMOTING RETENTION IN HIV CARE

Laramie R. Smith, PhD1, K. Rivet Amico, PhD2, Chinazo Cunningham, MD, MS3

1
Division of Global Public Health/UCSD, La Jolla, CA; 2School of Public Health/ University of
Michigan, Ann Arbor, Ann Arbor, MI; 3Albert Einstein College of Medicine, Bronx, NY

Sustained retention in HIV medical care is a key health behavior for the long-term health of
people living with HIV (PLWH) and reduced HIV transmission via viral suppression. In the US,
the CDC estimates 60% of PLWH are not retained in HIV care, yet to date, the few evidence-
based retention-promoting interventions available are resource and time intensive to
implement. The current study piloted 60 Minutes for Health, a theory-based, low-resource
single-session intervention designed to be implemented in a busy clinic setting by a health
educator, to PLWH who have had a recent (6-months) gap in care. Intervention content was
informed by the Information Motivation Behavioral Skills model and delivered via
Motivational Interviewing. A series of guided activities aimed to identify and reduce faulty
heuristics guiding HIV care attendance, enhance motivation to maintain care via personal
health goals, and build skills for coping with negative affect towards living with HIV and
increase self-efficacy for navigating the logistics of maintaining care amidst competing
priorities. Participants were randomized to 60 Minutes for Health (n=8) or a theory-based, diet
and nutrition time-and-attention control intervention (n=8). Medical records were abstracted
to measure number of quarters (3-month intervals) with an HIV care visit, and group
differences in visit attendance over time were characterized and evaluated with independent
group t-tests. Groups did not differ in the number of quarters with a visit 12-months prior to
baseline (t(14)=0.894, p(2-tail)=0.386). A substantial trend identified 75% of the retention arm
vs. 50% of the control arm had at least one visit in the first two quarters (t (14)=0.149, p(1-
tail)=0.167). Over the entire 12-month follow-up, retention intervention participants had
dramatically more quarters with documented care visits (M=2.75, SD=1.28) than those in the
comparison condition (M=1.38, SD=1.60), which did reach significance (t (14)=1.898, p(1-
tail)=0.039; Cohens d=0.946). Findings suggest the
60 Minutes for Health intervention may
promote better adherence to HIV care among PLWH with a history of poor retention.

CORRESPONDING AUTHOR: Laramie R. Smith, PhD, Division of Global Public Health/UCSD, La

Jolla, CA, 92093-0507; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S23

A002 6:00 PM-7:00 PM

A COACH BY ANY OTHER NAME: HEALTH COACHING TERMINOLOGY, ASSOCIATIONS,

PREFERENCES, AND EXPECTATIONS

Tess Thompson, PhD, MPH1, Amy McQueen, PhD2, Matthew Kreuter, PhD3, Molly Loughran,
BS4, Katie Childs, BA3, Nicole Caito, MPH, MS, RD5, Timothy Poor, BA3, John H. Wingfield,
PhD6, Jennifer A. Funaro, B.S., Communications7, Karyn Quinn, BSN7, Montserrat Urnek,
MBA8, Lisa Gibson, N/A9

1
Washington University in St. Louis, Webster Groves, MO; 2Washington University, St Louis,
MO; 3Washington University in St. Louis, St. Louis, MO; 4Health Communication Research
Laboratory, Washington University in St. Louis, Saint Louis, MO; 5Washington University,
Fenton, MO; 6Washington University in St. Louis, St Louis, MO; 7Envolve PeopleCare,
Farmington, CT; 8Envolve PeopleCare, Derry, NH; 9Envolve PeopleCare, Farmington, CT

Health coaching can promote health behavior change in contexts including health insurance
plans, workplace wellness programs, and primary care clinics. Online coaching programs may
offer additional flexibility for patients who need it. Little is known, however, about peoples
preferences for health coach terminology or the characteristics and services they expect from
coaching. This formative study was conducted to inform the design of an online health
coaching program. We interviewed 25 adults enrolled in employer-sponsored health
insurance plans (mean age = 42, age range 28-63; 68% women; 76% white). A semi-structured
interview was conducted to explore associations with, and preferences for, health coaching
terminology. Card sort exercises elicited reactions to 31 coaching terms (e.g., navigator,
guide). Participants also answered open-ended questions about coaching. Interviews were
transcribed and coded by two trained coders. Preferred terms for people assisting with
behavior change included coach, advisor, counselor, specialist, and consultant. One dimension
that differentiated terms was education/training (expert) vs. life experience (supporter). Most
participants expected coaches to have extensive experience, but none believed a coach had to
have lived through the same experience as the person being coached. Many participants
believed coaches are successful in motivating health behavior change when they hold people
accountable to their goals. Participants valued coaches empathy, listening skills, and ability to
motivate. Most participants did not anticipate negative consequences of working with
coaches, but a few expressed reservations about data security, poor advice, or too-frequent
contact. Results suggest that some coaching terms are more strongly endorsed than others,
and that people generally have positive expectations of coaching. Future research will
compare this sample with Medicaid-insured individuals to explore whether and how
S24 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

preferences differ. Online coaching programs should reflect participants preferences for
coaching terms and attributes of coaches, and find ways to reassure participants about
perceived negative effects of coaching.

CORRESPONDING AUTHOR: Tess Thompson, PhD, MPH, Washington University in St. Louis,
Webster Groves, MO, 63119; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S25

A003 6:00 PM-7:00 PM

A NOVEL MODEL TO SUPPORT LIFESTYLE INTERVENTIONS IN CANCER SURVIVORS:

UNDERGRADUATE STUDENTS AND MULTIMODAL SOFTWARE PLATFORM

Angela Yung, RDN1, Tracy Crane, BS: MS; PhD2, Amanda Younger, MS, CE-P, CHWC, EIM2,
Loescher J. Lois, PhD, FAAN3, Lindsay Bingham, B.S.4, Cynthia Thomson, PhD, RD2

1
Behavioral Measurement and Interventions Shared Resourc, Tucson, AZ; 2University of
Arizona, Tucson, AZ; 3The University of Arizona, Tucson, AZ; 4Arizona Smokers' Helpline,
Tucson, AZ

Background: There are an estimated 15.5 million cancer survivors in the U.S. Lifestyle
interventions that promote healthy eating and physical activity hold promise to improve
survival and quality of life. Efforts to test and develop scalable, cost-efficient models for the
delivery of lifestyle behavior interventions is an active area of research.

Purpose: To describe the integration of a health coaching training program developed for
undergraduate (UG) students and its effect on cancer survivors.

Methods: The University of Arizona Behavioral Measurements and Interventions Shared

Resource (BMISR) at the University of Arizona (UA) Cancer Center in collaboration with
Arizona Research Labs developed a cloud-based software platform for the delivery of eHealth
coaching (telephone, SMS, email) for lifestyle intervention research. A 120 hour standardized
health coaching training program was developed for UG students in nutrition science and
public health for the delivery of lifestyle interventions. Training for health coaches is grounded
in cognitive behavior theory with integration of the health belief model utilizing motivational
interviewing to guide behavior change. In addition, health coaches are trained in nutrition
and physical activity areas specific to cancer survivors including GI health, neuropathy, and
fatigue.

Results: Since 2012, over 900 patients (predominantly cancer survivors) have received health
coaching from 20 trained UG health coaches at BMISR. Evidence for behavior change within
one cohort of survivors, when comparing baseline to 12 month data suggests average
increase in 5.2 MET-hour /week, reduction in dietary fat of 14 grams/day as well as increases
of 3 servings/day in vegetable/fruit and 8.5 grams fiber. UG students trained utilizing this
model have been highly successful in achieving professional goals with 5 securing dietetic
internships and 4 entering graduate training programs. Importantly, this model for
intervention delivery is cost-effective, with each coaching session costing less than $12.00.
S26 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusion: Novel, student-engaged approaches that integrate resources and expertise across
academic campuses can result in the provision of lifestyle interventions that are mutually
beneficial for patient populations, researchers and students.

CORRESPONDING AUTHOR: Angela Yung, RDN, Behavioral Measurement and Interventions

Shared Resourc, Tucson, AZ, 85719; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S27

A004 6:00 PM-7:00 PM

A PILOT STUDY TO EVALUATE A HEALTHY LIFESTYLE BEHAVIOR CHANGE PROGRAM

Linda Gotthelf, Ph.D.1, Carol Addy, M.D., MMSc2, Steven May, PHD Licensed Psychologist
California3

1
HMR, Boston, MA; 2HMR Weight Management Services Corp., Boston, MA; 3Health
Management Resources, Palm Springs, CA

Data consistently show that few Americans are physically active and follow a healthy diet
even though these behaviors have been shown to be associated with maintaining normal
body weight, reduced risk of chronic diseases, and overall health. It is important to identify
programs to help people be engaged in and reinforced for making healthy lifestyle changes.

HMR, with over 30 years of experience in weight management, conducted a 12-week pilot
study to evaluate the effect of a novel healthy lifestyle behavior change program on health
and quality of life. A lifestyle survey was developed to create a Healthy Eating & Physical
Activity Profile. Individual profiles had activity and healthy diet scores as well as a total
Healthy Lifestyle Score (HLS). The lifestyle survey was completed at baseline, weeks 4, 8 and
12. Quality of life (QoL) was evaluated at baseline and week 12 using the RAND 36-item
Health Survey. The primary endpoint for the study was change from baseline in HLS, an
indication of change in measured healthy lifestyle behaviors.

Participants kept simple records on physical activity (PA) and vegetable/fruit servings (V/F)
each day using an HMR mobile app. Participants who expressed interest in weight loss (98%)
were sent a wireless scale and were instructed to weigh weekly. Hour-long weekly group
coaching sessions were conducted over the phone by trained health coaches.

62 participants enrolled (56 women, 6 men) with a mean starting weight of 218.3 lbs. (range:
149-364). 93.5% completed the 12 week pilot. Preliminary data (n=50) show a mean of 1054
and 2636 PA calories/wk and 9.9 and 34.3 V/F servings (full-cup) at baseline and at week 12,
respectively. The total HLS score was 50.1 at baseline and 92.0 on the post-survey (range of
possible scores is 0-120). The mean weight change was 8.2 lbs. On the post-study survey,
96.5% felt somewhat or very much more optimistic about their overall health and QoL and
94.6% were more optimistic about managing their weight. 91% said they would recommend
the program to family, friends or colleagues. On the RAND Survey, all 8 scales improved from
baseline with the greatest improvement in role capabilities due to physical health and energy.
S28 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Overall, participants were able to use the HLS as a tool to identify specific steps in making
lifestyle changes. Despite modest weight loss, participants made dietary and PA changes and
the improvement in QoL was substantial for most.

CORRESPONDING AUTHOR: Linda Gotthelf, Ph.D., HMR, Boston, MA, 02110;

[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S29

A005 6:00 PM-7:00 PM

A THEORY-BASED SMART PHONE APP TO PROMOTE SAFETY BEHAVIORS

Andrea Gielen, ScD ScM

Johns Hopkins Bloomberg School of Public Health, Baltimore, MD

BackgroundThe leading cause of death to children after the age of 1 is motor vehicle crashes,
and child safety seats (CSS) are known to reduce this risk by between 50-80%. Parents use of
CSS with their older children (> 4years) is low, and misuse is widespread. Health care settings
offer an ideal time to communicate with parents, yet there are many missed opportunities to
do so, especially in emergency departments (EDs) where clinicians time is very limited.
Therefore we developed a theory-based, stand-alone, smart phone app to promote correct
CSS use that does not rely on clinicians to communicate potentially life-saving information.
We tested the efficacy of the app in an RCT conducted in a pediatric ED in Maryland and one
in Arkansas

Methods Parents with children 4-7 years old were recruited in the ED waiting rooms where
they downloaded the Safety in Seconds app to their smart phone, and completed a baseline
assessment of behavior and beliefs. Based on their responses, a personalized feedback report
with tailored education was immediately delivered in the app. Those randomized to the
intervention group received CSS messages; those in the control group received fire safety
messages. The assessment and messages were developed following the Precaution Adoption
Process Model. Development of the app and the application of behavioral theory to
messaging will be further described in the presentation. A follow-up survey was conducted 3
months later.ResultsA total of 707 parents completed follow-up, and there were no
demographic differences between the two study groups. Compared to the control group, the
intervention group was significantly more likely to report using the correct type of car seat for
their childs age and size (77% vs. 68%; p < 0.01), more likely to report their child rides in the
back seat (96% vs. 91%; p < 0.01), and more likely to report having had their seat inspected by
a car seat technician (41% vs. 20%; p < 0.01). Significant results were maintained when
adjusted for behaviors reported at baseline. Conclusions A theory-based, tailored educational
program delivered via smartphone app can improve parents use of car safety seats to protect
their children.

CORRESPONDING AUTHOR: Andrea Gielen, ScD ScM, Johns Hopkins Bloomberg School of
Public Health, Baltimore, MD, 21205; [email protected]
S30 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A006 6:00 PM-7:00 PM

AN ONLINE DAILY FEEDBACK INTERVENTION IMPROVES SUN PROTECTION AMONG PATIENTS

WITH AN ELEVATED RISK OF SKIN CANCER

Tammy Stump, PhD1, Lisa G. Aspinwall, PhD2

1
Feinberg School of Medicine, Northwestern University, Chicago, IL; 2University of Utah, Salt
Lake City, UT

Every year, over two million people are diagnosed with skin cancer. The primary method
recommended for skin cancer prevention is reducing ultraviolet radiation (UVR) exposure.
However, sun protection is often inadequate, even among higher risk patients. To assist
patients efforts to reduce their skin cancer risk, the present study provided daily personalized
feedback on UVR exposure.

This study tested 1) the effectiveness of a 14-day online intervention that provided daily color-
coded feedback illustrating duration of UVR exposure on specific body sites, and 2)
theoretically derived predictions on how feedback impacts the process of reducing UVR
exposure. Participants (N=47; 53.3% women, mean age=49.87) were recruited from
dermatology clinics and had an elevated risk of skin cancer; most (63.8%) had a skin cancer
history, including 44.7% with melanoma. Participants were randomly assigned to a feedback,
self-monitoring, or control condition. For feedback participants, the 14-day intervention
included daily sun-protection reminders, self-reports of sun exposure, color-coded feedback
diagrams, and survey items on health-relevant cognitions and emotions. To control for the
potential benefit of reporting ones behavior, self-monitoring participants completed these
assessments but did not receive feedback. Control participants only received daily reminders.

At baseline, then 1 and 2 months later (and during the intervention, if applicable), sun
exposure was measured by the Minutes of Unprotected Sun Exposure (MUSE) Inventory,
which assesses the duration of sun exposure after accounting for the use of any sun
protection methods (e.g., sunscreen, hat). During the intervention, feedback participants
reported less sun exposure than self-monitoring participants. In both of these conditions,
participants reported higher perceptions of goal fulfillment for sun exposure on intervention
days when their sun exposure was lower, and these perceptions predicted higher self-efficacy
for sun protection. Only feedback participants reported a decrease in sun exposure from
baseline to the 2-month follow-up; significant decreases were reported for the lower face,
arms, and lower legs, which are common sites for melanoma. Tanning was also assessed by
reflectance spectroscopy, an objective measure of skin color. These measurements did not
change over time or by condition, potentially because they were taken on a limited number of
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S31

body sites (wrist, upper face) for which reported exposure did not decrease substantially.

This study supports the feasibility and effectiveness of an online, daily feedback intervention
for sun exposure among higher risk patients. Future directions include testing it among other
populations and investigating additional mechanisms (e.g., changes in goals) through which
feedback operates.

CORRESPONDING AUTHOR: Tammy Stump, PhD, Feinberg School of Medicine, Northwestern

University, Chicago, IL, 60613; [email protected]
S32 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A007 6:00 PM-7:00 PM

ASSESSING THE THEORETICAL FIDELITY OF BEHAVIOR CHANGE INTERVENTIONS USING

TAXONOMY CODING

Shirley M. Moore, PhD1, Clara Adegbite-Adeniyi, BA1, Oluwatomisin Olayinka, MPH, RN2

1
Case Western Reserve University, Cleveland, OH; 2School of Nursing, Case Western Reserve
University, Cleveland, OH

Background: In studies of head-to-head comparisons of theoretically-different behavior

change interventions it is important to have good theoretical fidelity of the interventions
designed and tested. The development of a taxonomy coding system for 93 behavior change
techniques (Michie, 2013) provides a standardized approach to determining the types and
dose of behavior change techniques used in interventions.
Purpose: The purpose of this report is to describe the use of taxonomy coding to assess the
theoretical fidelity of behavior change interventions currently being tested in a randomized
trial for weight management in adolescents. One intervention was based on cognitive-
behavioral theory, the other intervention was based on system process improvement theory.
Methods: Following training in the coding process and the establishment of acceptable
interrater reliability, the Behavior Change Technique Taxonomy (BCTTv1) was used by two
coders to independently code intervention materials (teaching plans, handouts) of two 12-
month family-focused interventions for adolescent weight management (25 90-minute face-
to-face sessions each).
Results: Findings showed that there were distinct differences in the types of behavior change
techniques used in the two interventions. The cognitive behavioral intervention
predominately used behavior change techniques associated with goal setting and problem-
solving, whereas these techniques were negligibly used in the system process improvement
intervention. Alternatively, the use of behavioral experiment, behavioral instruction,
behavioral practice, behavioral contract, and commitment were used to a much greater
extent in the system process improvement intervention. Thus, the behavior change
techniques used in the two interventions did reflect the different theoretical foundations of
the interventions.
Conclusions: In conclusion, the use of a behavior change technique taxonomy was useful to
assess the extent to which intervention components reflected the underlying theories of
different interventions. Greater application of intervention taxonomy coding has the potential
to assist behavioral scientists to design more efficient and effective behavior change
interventions.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S33

CORRESPONDING AUTHOR: Shirley M. Moore, PhD, Case Western Reserve University,

Cleveland, OH, 44106; [email protected]
S34 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A008 6:00 PM-7:00 PM

ASSOCIATION BETWEEN SPIRITUALITY AND INTERVENTION ADHERENCE IN A MIND-BODY

PILOT STUDY AMONG AFRICAN AMERICANS

Nishat Bhuiyan, B.S.1, Scherezade K. Mama, DrPH2, Jornae Rideaux, MPH3, Lorna McNeill,
PhD, MPH4

1
The Pennsylvania State University, State College, PA; 2The Pennsylvania State University,
University Park, PA; 3MD Anderson Cancer Center, Pearland, TX; 4University of Texas MD
Anderson Cancer Center, Houston, TX

Mind-body (MB) practices have been shown to be effective for improving various health
outcomes. Spirituality is an important cultural component in the African American
community, and culturally-adapted MB interventions have potential for reducing health
disparities. However, little is known about the role of spirituality in MB intervention
adherence. A mixed-method approach was used to examine the association between
spirituality and intervention adherence in an MB pilot study among African American
churchgoers. Participants were recruited through an existing church partnership, and eligible
participants (N=50) were randomized to an MB intervention (n=26) or wait-list control (n=24)
group. Participants in the MB intervention (M age=50.1 years, M BMI=33.9 kg/m2) attended
16 MB sessions over 8 weeks. Spirituality was assessed at baseline (T1) and post-intervention
(T2) with the Functional Assessment of Chronic Illness Therapy Spiritual Wellbeing Scale
(FACIT-Sp). A subsample of MB participants (n=6) completed an optional in-depth exit
interview consisting of questions assessing motivators and barriers to intervention
attendance. Among those in the MB intervention, increases in spirituality were associated
with increased session attendance (=.697, p=.033). A thematic content analysis of the
interview transcripts identified spirituality as a key motivator for participants intervention
attendance. Results suggest that spirituality may play an important role in intervention
adherence in an African American church-based population. Church-based MB interventions
should aim to incorporate and foster spirituality as a strategy to overcome barriers to
intervention adherence among African Americans.

CORRESPONDING AUTHOR: Nishat Bhuiyan, B.S., The Pennsylvania State University, State
College, PA, 16801; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S35

A009 6:00 PM-7:00 PM

BIDIRECTIONAL LONGITUDINAL ASSOCIATIONS BETWEEN SMOKING AND AFFECT IN CANCER

PATIENTS USING VARENICLINE TO QUIT SMOKING

Allison J. Carroll, M.S.1, Anna K. Veluz-Wilkins, MA1, Nancy C. Jao, MS1, Frank Leone, MD MS2,
Robert A. Schnoll, PhD2, Brian Hitsman, Ph.D.1

1
Northwestern University Feinberg School of Medicine, Chicago, IL; 2University of
Pennsylvania, Philadelphia, PA

Background: During a quit attempt, low positive affect (PA) and high negative affect (NA)
predict relapse to smoking. In this study we evaluated bidirectional longitudinal associations
between smoking and PA/NA in cancer patients treated with varenicline. Methods: Smokers
(5 cigarettes/week) who were diagnosed with cancer within 5 years were recruited for an
ongoing 24-week trial of extended duration varenicline plus behavioral counseling. Data were
drawn from the 12-week open-label phase. Smoking was assessed via self-reported number of
cigarettes in the past 24 hours (CPD). PA/NA were assessed using the Positive and Negative
Affect Scale. Data were collected at pre-quit (PQ), target quit day (TQD), week 4 (W4), and
week 12 (W12). We evaluated cross-lagged panel models for NA and PA using PROC CALIS in
SAS. We assessed autocorrelations (e.g. PQ CPD to TQD CPD) and cross-lagged associations
(e.g. PQ CPD to TQD NA). Error covariances (e.g. PQ CPD and PQ PA) were included in each
model. Results: Participants (N=93 with complete data) were 48% female, 33% racial minority,
and 59 years old on average. Goodness-of-fit statistics were adequate for both models (NA:
SRMR=0.09, AGFI=0.71, RMSEA=0.18, BCFI=0.87; PA: SRMR=0.07, AGFI=0.81, RMSEA=0.12,
BCFI=0.96). Autocorrelations were strong within PA (path coefficients: 0.81-0.85) and NA
(0.52-0.66), pspspspsConclusion: Using novel statistical methodology we found that self-
reported smoking within the past 24 hours accounted for up to 22% of variance in later NA,
but not PA, through 12 weeks of treatment. Neither NA nor PA predicted later smoking.
Among cancer patients treated for nicotine dependence, the lack of association between PA
and smoking suggests that varenicline may especially modulate PA, independent of smoking
status.

CORRESPONDING AUTHOR: Allison J. Carroll, M.S., Northwestern University Feinberg School

of Medicine, Chicago, IL, 60611; [email protected]
S36 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A010 6:00 PM-7:00 PM

BRIEF COUNSELING AND EXERCISE REFERRAL SCHEME: A PRAGMATIC TRIAL IN MEXICO.

Katia Gallegos Carrillo, PhD MSc1, Carmen Garcia Pea, PhD MSc MD2, Jorge Salmeron, PhD
MSc MD3, Nelly Salgado-de-Snyder, PhD, MPH4, Felipe Lobelo, PhD MD5

1
University of Southern California, Glendale, CA; 2Instituto Nacional de Geriatria Mexico,
Ciudad de Mexico, N/A, Mexico; 3Instituto Mexicano del Seguro Social, Cuernavaca, N/A,
Mexico; 4National Institute of Public Health. Mexico., Cuernavaca, Morelos, Mexico; 5Emory
University, Atlanta, GA

Introduction: The effectiveness of clinical-community linkages for the promotion of physical

activity (PA) has not been explored in low and middle-income countries (LMICs). It was
assessed the effectiveness of a primary-care based, 16-week intervention rooted in behavioral
theory approaches to increase compliance with aerobic PA recommendations.

Methods: Pragmatic cluster randomized trial.Patients with diagnosis (< 5 years) of

hypertension, 35-70 years of age, self-reported as physically inactive, with a stated intention
to engage in PA, and attending primary health care centers (PHCC) in the Social Security
health system in Cuernavaca, Mexico. Out of a total of 23 PHCCs, four were selected based on
proximity (5 km radius) to a center with PA programs and trained personnel and randomized
to a brief PA counseling (BC) (n=2) or an exercise-referral (ER) (n=2) intervention. The study
was conducted between 2011 and 2012. Change in objectively-measured PA levels (ActiGraph
GT3X accelerometers) at baseline, 16 and 24 weeks. Intention-to-treat analyses were used to
assess the overall effectiveness of the intervention overall and according to ER intervention
attendance. Longitudinal multilevel mixed-effects analysis considering the interaction (time by
intervention) were conducted, each model was also adjusted by baseline value of the
outcome measure, demographic and health variables, social support, PA self-efficacy and
barriers.

Results: Minutes/week of moderate-to-vigorous physical activity (MVPA) objectively-

measured increased by, 40 and 53 minutes in the ER and BC groups, respectively (p=0.59).
Participants attending> 50% of ER program sessions increased their MVPA by 104
minutes/week and their compliance with aerobic PA recommendations by 23.8%, compared
to the BC group (both p < 0.05).

Conclusions: Both BC and ER led to modest improvements in PA levels, with no significant

differences between groups. Adequate adherence with the ER program sessions led to
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S37

significant improvements in the compliance with aerobic PA recommendations compared to

BC. These results provide evidence to help guide the development and implementation of
programs integrating standardized PA assessment, counseling and referrals via clinical-
community linkages in Mexico and other LMICs in the region.

CORRESPONDING AUTHOR: Katia Gallegos Carrillo, PhD MSc, University of Southern

California, Glendale, CA, 91204; [email protected]
S38 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A011 6:00 PM-7:00 PM

CELL PHONE USE WHILE DRIVING: DOES PEER-REPORTED USE PREDICT EMERGING ADULT
BEHAVIOR

Neha Trivedi, MPH1, Denise Haynie, PhD, MPH2, Joseph Bible, PhD2, Danping Liu, PhD2, Bruce
Simons-Morton, EdD, MPH2

1
University of Maryland, Adelphi, MD; 2National Institute of Child Health and Human
Development, Bethesda, MD

Secondary task engagement such as cellphone use while driving is a common public health
concern among adolescents and emerging adults. Cell phone use among this population
contributes to the high rates of fatal car crashes. Peer engagement in similar risky driving
behaviors, such as cell phone use, could affect adolescent driving behavior. The present study
investigates the association between peer and emerging adult cell phone use while driving.
Binomial logistic regression analyses, adjusting for gender, race/ethnicity, parental education,
and family affluence, showed that participants with peers who reported the first year after
high school frequently texting while driving, were significantly more likely to text while driving
the following year (odds ratio, 3.01; 95% CI, 1.19 to 7.59; P=0.05) . The findings are consistent
with peer influence on the prevalence of texting while driving among emerging adult.

CORRESPONDING AUTHOR: Neha Trivedi, MPH, University of Maryland, Adelphi, MD, 20783;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S39

A012 6:00 PM-7:00 PM

CHANGES IN SEDENTARY BEHAVIORS AMONG LATINAS IN A PHYSICAL ACTIVITY

INTERVENTION

Sheri J. Hartman, PhD1, Shira Dunsiger, PhD2, Dori Pekmezi, PhD3, Britta Larsen, PhD4, Bess H.
Marcus, Ph.D.5

1
Department of Family Medicine and Public Health, UC San Diego, La Jolla, CA, USA, La Jolla,
CA; 2The Miriam Hospital and Brown University, Portsmouth, RI; 3University of Alabama at
Birmingham, Birmingham, AL; 4University of California, San Diego, San Diego, CA; 5University
of California, San Diego, La Jolla, CA

Greater time spent in sedentary behaviors is associated with increased risk for poor health
outcomes. Emerging research has demonstrated that sedentary behaviors are distinct from
physical activity and that existing interventions to increased physical activity do not
meaningfully impact sedentary behaviors. The purpose of this study was to examine changes
to sedentary time among Spanish-speaking Latinas in a randomized controlled trial of a
physical activity intervention.

Sedentary time and moderate to vigorous physical activity (MVPA) were measured via
accelerometer at baseline, 6 and 12 months. Using a series of generalized linear models, we
examined intervention effects on mean sedentary time at follow-up, controlling for baseline
and wear time.

Participants (n=205) were a mean age of 39 years (SD=10.5) with the majority self-identified
as Mexican (84.4%), White (51.7%) and First Generation (81%). Overall, participants were
sedentary an average of 62.03 hours per week (SD=18.18). Results showed significant
intervention effects such that those randomized to the Intervention arm had significantly
higher mean minutes of objectively measured sedentary time at 6 and 12 months, ps <
.05. Compared to the control group, participants in the Intervention arm were sedentary for
146 more minutes/week at 6 months and 254 minutes/week at 12 months (p=.02). As a
subsequent step, we added MVPA into the model as the Intervention group had significantly
greater increases in min/week of MVPA than the control group (mean differences=31.0,
SE=10.7, p < .01). The Intervention group spent significantly more minutes being sedentary
than the control group even after controlling for time in MVPA. Furthermore, more time
spent in MVPA was positively associated with more sedentary time (p=.04).

These results emphasize that sedentary behavior and physical activity are distinct behaviors,
S40 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

and suggest that, as participants increase MVPA, they may compensate by also increasing
sedentary time. Physical activity interventions may need to simultaneously address decreasing
sedentary time in order to maximize health benefits.

CORRESPONDING AUTHOR: Sheri J. Hartman, PhD, Department of Family Medicine and Public
Health, UC San Diego, La Jolla, CA, USA, La Jolla, CA, 92093; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S41

A013 6:00 PM-7:00 PM

COPING STYLE AND DIABETES DISTRESS AS PREDICTORS OF HABIT STRENGTH FOR INSULIN
USE AMONG ADULTS WITH TYPE 2 DIABETES

Rachel Burns, PhD1, Sonya Deschnes, PhD2, Brbel Knuper, PhD1, Norbert Schmitz, PhD2

1
McGill University, Montreal, PQ, Canada; 2McGill University/Douglas Mental Health
University Institute, Montreal, PQ, Canada

Habits are important processes for health behavior maintenance. Strong habits are automatic
and are formed by repeating a behavior in a stable context. Coping styles can influence how
often a behavior that is perceived to be stressful is performed and thus may influence the
formation of strong habits. In particular, an avoidance-oriented coping style may hamper the
formation of strong habits because behaviors that are avoided are repeated less frequently. In
contrast, task-oriented styles involve problem-solving and planning and may facilitate the
development of strong habits by promoting behavioral repetition. This study tested
prospective associations between dispositional coping styles (avoidance-, task-, and emotion-
oriented) and habit strength for insulin use. Coping strategies are activated under conditions
of stress, so diabetes distress was tested as a moderator. Diabetes distress was hypothesized
to moderate the association between avoidance-oriented coping and habit strength for insulin
use, and the association between task-oriented coping and habit strength. The association
between emotion-oriented coping and habit strength was explored. Participants were 77
adults with type 2 diabetes who were insulin-nave at baseline and initiated insulin therapy
during the 3- or 4-year follow-up waves. Coping styles and diabetes distress were measured at
baseline. Habit strength for insulin use was measured at the follow-up wave that insulin use
was first reported. Diabetes distress moderated the association between avoidance-oriented
coping and habit strength (p = .02, R2interaction = .07); avoidance-oriented coping was
inversely associated with habit strength if diabetes distress was relatively high, but not if
diabetes distress was average or low. Diabetes distress also moderated the association
between task-oriented coping and habit strength (p = .0497, R2interaction = .05). Contrary to
expectations, task-oriented coping was inversely associated with habit strength if diabetes
distress was relatively high or average, but not when diabetes distress was low. This may be
attributable to the vigilance and deliberative decision-making that accompanies task-oriented
coping. Emotion-oriented coping was not associated with habit strength. Results indicated
that specific coping styles were prospectively associated with habit strength for insulin use if
diabetes distress was relatively high. Previously, studies have generally examined if properties
of the target behavior influence the development of habits. This study highlights the utility of
exploring dispositional characteristics as predictors of habit formation. Future research could
S42 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

explore the effectiveness of delivering simple habit formation interventions to individuals with
task- or avoidance-oriented coping styles who are experiencing high levels of diabetes
distress.

CORRESPONDING AUTHOR: Rachel Burns, PhD, McGill University, Montreal, PQ, H4A3L3;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S43

A014 6:00 PM-7:00 PM

CORRELATES OF FLU VACCINE BEHAVIOR AND VACCINE INTENTIONS AMONG COLLEGE

STUDENTS

Kara M. Magane, MS1, Gregory M. Dominick, PhD2

1
Division of Adolescent/Young Adult Medicine, Boston Children's Hospital, Portsmouth, NH;
2
University of Delaware, Newark, DE

Background. Influenza (flu) vaccination rates are lowest among college students, yet research
in this population is limited. This study examined correlates of flu vaccine behaviors and
intentions, and reasons for not getting vaccinated.

Methods. An online survey was emailed to a random sample of 3,201 full-time undergraduate
students between 18 and 25 years of age. Multivariate logistic regression was used to
determine associations between demographic (n=5), psychosocial (n=3), and environmental
(n=10) influences on past flu vaccine behaviors and flu vaccine intentions.

Results. Of the 450 respondents, 36% received a flu vaccine in the past year, and nearly 45%
intended to get vaccinated for the coming season. Being Hispanic (AOR=0.36, 95%CI 0.15-
0.90), having very good health (AOR=0.47; 95%CI 0.26-0.86), and using pharmacy/health
department (AOR=0.55, 95%CI 0.36-0.85) or family/friends (AOR=0.53, 95%CI 0.34-0.82) as
sources of flu vaccine information were associated with lower odds of receiving the flu
vaccine in the past year. Belief that advertising increased ones likelihood of getting
vaccinated was associated with higher odds of receiving the flu vaccine in the past year
(AOR=2.18, 95%CI 1.41-3.35). Flu vaccine intentions were lower for African Americans
(AOR=0.37, 95%CI 0.15-0.91) and those using Internet/social media as a source of flu vaccine
information (AOR=0.38; 95%CI 0.38-0.93). Flu vaccine intentions were higher among those
concerned about flu (AOR=1.97, 95%CI 1.10-3.53) and those aware of the cost and location of
campus flu vaccine services (AOR=1.67, 95%CI 1.11-2.52). The most frequently reported
reasons for not getting vaccinated were never getting the flu (33.73%), unlikely to get sick
from the flu (32.13%), and concerns that the vaccine causes the flu (31.33%).

Discussion. Flu vaccination behaviors and vaccine intentions in this sample are below the
national target of 70%. Results revealed differences in correlates of past flu vaccine behaviors
and intentions that could inform campus-wide programs to promote flu vaccination rates,
particularly among racial/ethnic minorities. Given that most college students are in good
health, behavioral intervention messages should target perceived flu susceptibility and
S44 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

severity, while accurately addressing misconceptions about flu vaccine safety, as well as
vaccine cost and location to reduce barriers and increase accessibility.

CORRESPONDING AUTHOR: Kara M. Magane, MS, Division of Adolescent/Young Adult

Medicine, Boston Children's Hospital, Portsmouth, NH, 03801;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S45

A015 6:00 PM-7:00 PM

DEPRESSION AND ANXIETY PREDICTING RUMINATIONS IMPACT ON HEALTH BEHAVIORS: A

DAILY DIARY STUDY

Kristen E. Riley, Ph.D. 1, Crystal L. Park, PhD2, Dean Cruess, PhD3, Jean-Philippe Laurenceau,
Ph.D.4

1
Memorial Sloan Kettering Cancer Center, New York, NY; 2University of Connecticut, Storrs
Mansfield, CT; 3University of Connecticut, Storrs, CT; 4University of Delaware, Newark, DE

Rumination, or thinking passively, negatively, and repetitively, is a common cognitive process,

and is associated with poor health behaviors. Rumination may impact health behaviors
through two distinct behavioral pathways: acting too quickly (impulsivity) or not acting at all
(amotivation). Anxiety and depression may be differential predictors of the behavioral
manifestations of rumination, and may be able to predict its effect on health behavior
patterns. Our purpose was to test whether individuals reporting anxiety who ruminate will
then act impulsively and those reporting depression who ruminate will then not act
(amotivation) in the context of daily health behaviors. We recruited 285 college students
(mean age=19.3; 76.8% female; 79.4% Caucasian) and had them complete a baseline survey
and an 11-day online daily diary survey. Variables assessed included baseline trait depression
and anxiety (DASS-21), and daily rumination (RRS), impulsivity (BIS-11), and amotivation
(GMS-28), all specifically in the context of daily health behaviors, such as fruit, vegetable, and
alcohol intake, exercise, and sexual risk taking behavior. Participants completed an average of
9.72 of 11 daily diary days, or 88.3% compliance. Results showed that rumination predicted
health behaviors through both impulsivity and amotivation. Mediation models revealed that
baseline anxiety predicted daily rumination ( = .060, p = p = .051), and that baseline
depression predicted rumination ( =.057, p = p = .055). We found that 17% of the association
between rumination and exercise was explained by impulsivity, and 29% of the association
between depression and motivation was explained by rumination. In additional post-hoc
analyses, rumination was shown to predict health behaviors above and beyond depression
and anxiety, demonstrating that rumination is an important, and perhaps more proximal,
predictor of health behaviors. These results showed that those who report depression or
anxiety are likely to experience more daily rumination and the deleterious behavioral
manifestations of that rumination, which the affects health behavior engagement or
avoidance. Future behavioral health interventions can perhaps specifically target impulsivity
in those reporting anxiety and amotivation in those reporting depression in order to promote
healthy behaviors.
S46 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Kristen E. Riley, Ph.D. , Memorial Sloan Kettering Cancer Center,
New York, NY, 10022; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S47

A016 6:00 PM-7:00 PM

DESCRIPTIVE ASSESSMENT OF IMPLEMENTATION INTENTIONS IN A MOBILE ECOLOGICAL

MOMENTARY INTERVENTION FOR SMOKING CESSATION

Vincent Puga-Aragon, MPH1, Katherine L. Sarina, Psychology, B.A.2, Sheila Yu, MPH3, Christian
Cerrada, BS3, Kathleen Colao, BS, MPH4, Eldin Dzubur, MS3, Jimi Huh, PhD3

1
MyQuit USC, Duarte, CA; 2University of Southern California, Huntington Beach, CA;
3
University of Southern California, Los Angeles, CA; 4University of Southern California Keck
School of Medicine, Venice, CA

Introduction: The present study explores the types of implementation intentions (IIs) that
young Korean smokers enact to avoid smoking lapse during a cessation attempt. Though IIs
have demonstrated success in previous smoking cessation studies, assessment of the types of
IIs smokers select and whether or not they have been enacted in situ has been rare. Mobile
health applications delivering ecological momentary assessments (EMA) have enabled
researchers to observe this process in real-time, allowing for exploration of II enactment,
adjustment, and failure to enact during high-risk smoking situations (HRSS) associated with
smoking.

Methods: Korean American emerging adult (KAEA) participants (N=16) responded to EMAs
(n=1,351) following at least five self-scheduled HRSS each day over 4 weeks using MyQuit
USC, a mobile smoking cessation app. Following each HRSS, participants indicated whether or
not they completed the specific II they scheduled with each HRSS at baseline. IIs were
categorized based on the primary method for avoiding lapse and HRSS were categorized by
types of smoking contexts: social, emotional, and habitual. We report descriptive statistics on
the most commonly enacted IIs overall and by HRSS type. Patterns of II completion over time
are also summarized.

Results: II enactment was assessed using EMA during 1,351 HRSS. Among those, 712 IIs were
enacted, with 71.5% completed during the initially scheduled, i.e. congruent, HRSS. The
distraction/substitution methods (e.g., chew gum) were the most commonly enacted IIs,
accounting for 41.5% of the total IIs completed across all congruent HRSS types. Among the
original IIs that were not followed, 67.3% were replaced with an alternate method, with 37%
of them being performed during the congruent HRSS. The averaged proportion of overall II
completion across participants increased over the 4 weeks from 48.3% to 59.4%. IIs in
congruent HRSS had an increase in completion proportion from 60.1% to 71.2% over the 4
weeks.
S48 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions: These preliminary results provide insight on commonly enacted IIs for each HRSS
type during a cessation attempt among KAEA smokers. Future versions of MyQuit USC will
incorporate these findings to develop a more context-relevant cessation aid. Specifically,
information about an individuals II enactment patterns may be leveraged to adapt the
provision of real-time individualized support, e.g. alternative IIs, according to personally
relevant HRSS.

CORRESPONDING AUTHOR: Vincent Puga-Aragon, MPH, MyQuit USC, Duarte, CA, 91010;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S49

A017 6:00 PM-7:00 PM

A LONGITUDINAL ANALYSIS OF SOCIAL SUPPORT RECEIVED FROM AN ONLINE COMMUNITY

FOR CANCER SURVIVORS AND CAREGIVERS

Tenbroeck Smith, M.A.1, Bennett R. McDonald, MPH1, Elizabeth Fallon, Ph.D.1, J. Lee
Westmaas, PhD1, Robert E. Kraut, PhD2, Kenneth M. Portier, PhD Biostatistics3

1
American Cancer Society, Atlanta, GA; 2Carnegie Mellon University, Pittsburgh, PA;
3
American Cancer Society, Global Headquarters, Atlanta, GA

Introduction: Cancer survivors and caregivers increasingly use online communities, but little
research has shown the efficacy of these communities for improving social support. The
American Cancer Societys Cancer Survivors Network (CSN) is an online social networking
community where survivors and caregivers can exchange information and emotional support
regarding the cancer experience. The purpose of this study was to examine whether CSN use
increases social support among cancer survivors and caregivers.

Methods: CSN members completed web surveys at baseline (T1) and 3-months (T2; N = 821).
Perceived Emotional/Informational Social Support from CSN (SSCSN) was measured using
items adapted from previously validated instruments. Member-reported use of CSN features
was summarized in two scores: 1) Interactive Use involved contributing to the community
(e.g., posting to discussion boards, chatting, writing blogs); 2) Reading was limited to
information consumption (e.g., reading discussion boards or blogs). Dichotomized Interactive
scores from T1 and T2 were used to create four groups reflecting patterns of use: Low-Low
(low use at T1 and T2), Low-High (low use at T1 and high use at T2), High-Low or High-High.
Similar groups were created for Reading. Two separate generalized linear models were used
to assess whether patterns of Reading or Interactive Use predicted SSCSN at T2, controlling
for SSCSN at T1 and covariates (age, gender, race, marital status, education, perceived health,
cancer status).

Results: The sample was mostly female (72%), white (87%), survivors (90%) with a mean age
of 54.9 (sd = 10.4) years. Using LS-means with Low-Low Interactive Use as the referent group,
the Low-High group was 0.58 points higher on SSCSN (95%CI: 0.15-1.00), and the High-High
group was 0.42 points higher on SSCSN (95%CI: 0.14-0.70). For Reading, only the High-High
group had a higher LS-means SSCSN (.31; 95%CI: 0.10-0.53) than the Low-Low group.

Discussion: Our results show that CSN use leads to higher levels of social support from that
community by approximately half a point on the 5-point SSCSN scale. Interactive use (e.g.,
S50 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

posting to discussion boards) seems particularly important, whereas passive reading, alone,
was associated with greater perceived social support only when reading levels were
consistently high. The literature suggests that the perceived social support provided by CSN
helps survivors and caregivers adapt to and cope with cancer.

CORRESPONDING AUTHOR: Tenbroeck Smith, M.A., American Cancer Society, Atlanta, GA,
30303; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S51

A018 6:00 PM-7:00 PM

A MODEL OF LONELINESS, SLEEP, AND DEPRESSION AMONG LUNG CANCER PATIENTS ON

ACTIVE ONCOLOGIC TREATMENT: A DAILY DIARY STUDY

Timothy J. Williamson, M.A., MPH1, Alyssa K. Choi, B.A. (in progress)1, Julie C. Kim, B.A.1,
Annette L. Stanton, Ph.D.2

1
University of California, Los Angeles, Los Angeles, CA; 2UCLA, Los Angeles, CA

Background: Loneliness predicts important health outcomes such as depression and mortality
in both healthy and medical populations. Hawkley and Cacioppo (2010) suggest that sleep
quality and daytime functioning (i.e., alertness, fatigue) may be mechanisms through which
loneliness predicts health morbidity. Loneliness has been associated with poor sleep quality
and subsequent impaired daytime functioning among older adults. Yet, no study has
investigated associations between loneliness, sleep quality, and daytime functioning within a
medical population. Loneliness is a prevalent concern among lung cancer (LC) patients, who
also evidence high rates of sleep disturbance and depression. As such, it is important to 1)
understand the extent to which loneliness predicts depression among LC patients and 2)
investigate whether sleep quality and daytime dysfunction mediate the posited relationship
between loneliness and depression. Method: Participants (n=33) were men and women
receiving treatment for LC. Participants completed a sleep diary at home for seven days. They
answered questions about sleep quality at the start of every morning and about loneliness
and daytime functioning at the end of each day. Participants also completed the Center for
Epidemiologic Studies-Depression scale (CES-D) at the end of the week. Serial mediation was
conducted to test the hypothesis that loneliness predicts poor sleep quality, which in turn
predicts daytime dysfunction, which in turn predicts higher depressive symptom severity.
Results: Loneliness and CES-D scores were significantly related (r = .53, p = .001). The
relationship between loneliness and depression symptom severity was significantly mediated
by daytime dysfunction (b = 4.05, SE = 2.47, 95% CI [0.57, 11.66]) but not by sleep quality
(b = 0.40, SE = 2.40, 95% CI [-0.99, 3.75]). Loneliness was unrelated to sleep quality (b = -4.85,
SE = 9.20, 95% CI [-23.56, 13.87]), and sleep quality was the only variable in the serial
mediation model that was not a significant predictor of depressive symptom severity (b = -
0.08, SE = 0.06, 95% CI [-0.20, 36.52]). Conclusions: Findings indicate that loneliness is a
significant predictor of depressive symptom severity among LC patients and that daytime
dysfunction, but not sleep quality, partially mediates this relationship. Although many LC
patients evidence poor sleep quality, our results suggest that daytime dysfunction is a more
potent predictor of depression in this sample.
S52 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Timothy J. Williamson, M.A., MPH, University of California, Los

Angeles, Los Angeles, CA, 90095; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S53

A019 6:00 PM-7:00 PM

A SYSTEMATIC REVIEW OF THE EFFECT OF PATIENT NAVIGATION ON SATISFACTION WITH

CANCER CARE

Kevin W. Campbell, Bachelor of Arts 1, Kristen Wells, Ph.D., MPH1, Ambuj Kumar, M.D., MPH2,
Tatianna Clark, NA3, Pascal Jean-Pierre, Ph.D.4

1
San Diego State University, San Diego, CA; 2Morsani College of Medicine, University of South
Florida, Tampa, FL; 3San Diego State University, Chula Vista, CA; 4University of Notre Dame,
Notre Dame, IN

Cancer is a leading cause of death in the United States, with disparate effects on racial-ethnic
minorities and lower socioeconomic groups. Patient navigation (PN) is a supportive patient-
centered model in which an individual guides a patient through the medical system and helps
him or her overcome barriers to healthcare. It is not known whether PN is associated with
increasing satisfaction with cancer care. A systematic review was conducted to assess studies
between the years 1995 and 2016 to integrate findings pertaining to the effectiveness of
patient navigation in the context of satisfaction with cancer care. To be included, studies were
required to evaluate patient navigation related to cancer care, include a comparison group,
and measure patient satisfaction with care after the PN intervention. Nine studies met
inclusion criteria, but only five (one randomized controlled trial [RCT] and four observational
studies) had adequate data for meta-analysis. Sample sizes for each individual study at
baseline ranged from 44 to 1,788 participants, with 4,210 patient satisfaction surveys
completed. Additionally, the type of cancer care (i.e., head and neck cancer, breast cancer)
among studies was greatly varied, as was the methodological quality of studies. Results from
the pooled observational studies data showed a statistically non-significant association
between PN and satisfaction with cancer care [standardized mean difference (SMD)=0.39;
95% Confidence Interval (CI):-0.02, 0.80, p=.06]. Additionally, results from the only RCT
included in the final analysis showed a significant increase in satisfaction with cancer care
when PN was implemented (SMD: 2.30; 95% CI:1.79, 2.80, p < .001). The contrast in findings,
in addition to the fact that half of the studies were rated as being weak methodologically,
indicates a need to conduct studies of higher quality to further investigate whether patients
who receive PN are more satisfied than those who do not.

CORRESPONDING AUTHOR: Kevin W. Campbell, Bachelor of Arts , San Diego State University,
San Diego, CA, 92114; [email protected]
S54 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A020 6:00 PM-7:00 PM

I DIDNT KNOW THERE WERE OTHER OPTIONS: USING BRECONDA TO ASSIST WOMEN AT
HIGH BREAST CANCER RISK WITH SURGICAL DECISION MAKING

Kerry Sherman, PhD1, Christopher J. Kilby, MRes (Psy)2, Laura-Kate Shaw, BA Psych Hons1,
Kathy Tucker, MBBS3, Judy Kirk, MBBS4, Elisabeth Elder, PhD5

1
Macquarie University, Sydney, New South Wales, Australia; 2Macquarie University, North
Ryde, New South Wales, Australia; 3Prince of Wales, Sydney, New South Wales, Australia;
4
Westmead Hospital, Westmead, New South Wales, Australia; 5Westmead Breast Cancer
Institute, Sydney, New South Wales, Australia

Introduction: Women at risk for hereditary breast cancer may opt for risk reducing
mastectomy. Part of this decision-making entails preference-based decisions regarding
whether, and how, to restore breast shape after mastectomy. We developed an online
decision aid, BRECONDA, to assist with this decision-making. In a randomised controlled trial
we compared the efficacy of BRECONDA with a general educational booklet.
Methods: Women at risk (N=87), who were recruited into the multicentre web-based trial
from hereditary cancer clinics, completed baseline questionnaires including measures of
decisional conflict, satisfaction with information and knowledge. Participants were then
randomly assigned to either: 1) Intervention (INT) - received unlimited access to BRECONDA
and an information booklet about breast reconstruction; or, 2) Control (CONT) - received the
information booklet alone. At 2-months participants completed a follow-up questionnaire re-
assessing satisfaction with information, knowledge, and decisional conflict regarding the
choice to have risk reducing mastectomy. INT participants also completed user acceptability
ratings for BRECONDA.
Results: Linear mixed model analyses demonstrated significant time x condition effects so that
by 2-months INT participants reported greater increases in satisfaction with information
(F=9.98, p < .005) and knowledge (F=12.23, p=.001) than controls. By 2-months there was a
significant effect for time, and condition, but not time x condition, with both INT and CONT
experiencing similar reductions in decisional conflict. However, by 2-months mean scores of
decisional conflict for INT participants (M=17.68) indicated that they had implemented their
decision, whereas the mean score of CONT participants (M=27.39) indicated that they were
still indecisive. The majority (73%) of INT participants reported high to very high satisfaction
with the intervention.
Conclusions: These findings indicate that participants receiving BRECONDA benefitted through
having greater satisfaction with information and knowledge about breast reconstruction
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S55

options, and reduced decisional conflict. Along with high user acceptability, these findings
support the feasibility of implementing BRECONDA into hereditary breast cancer clinics.

CORRESPONDING AUTHOR: Kerry Sherman, PhD, Macquarie University, Sydney, New South
Wales, 2109; [email protected]
S56 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A021 6:00 PM-7:00 PM

I DONT WANT TO LIVE WITH THE REGRET: UNDERSTANDING SURGICAL DECISIONS MADE
BY YOUNG WOMEN WITH BREAST CANCER

Shoshana Rosenberg, ScD, MPH1, Mary L. Greaney, PhD2, Andrea Patenaude, PhD1, Meghan
Meyer, N/A3, Ann Partridge, MD, MPH1

1
Dana-Farber Cancer Institute, Boston, MA; 2University of Rhode Island, Kingstom, RI; 3Dana
Farber Cancer Institute, Boston, MA

Background: Due to an increasing number of young women with unilateral breast cancer (BC)
choosing contralateral prophylactic mastectomy (CPM), there is a need to better understand
this choice of a procedure which is of limited medical benefit for most women. Using a series
of focus groups (FG), we sought to explore surgical decision-making in young BC survivors.

Methods: Women diagnosed with stage 0-3 BC in the past 1-3 yrs at age 40 who had
undergone BC surgery were recruited. Four FG were conducted: 2 with women after bilateral
mastectomy, 2 with women who kept their contralateral breast (lumpectomy or unilateral
mastectomy). FG were recorded and transcribed with identifiers removed. Emergent themes
were identified by thematic content analysis using NVivo 11. Participants also completed a
demographic/clinical survey.

Results: Of 20 participants (4-6 per FG) median age at diagnosis was 37; 10% (2/20) identified
as Hispanic; 70% had Stage 1 (4/20) or 2 (10/20) disease. Few were carriers of a BC pre-
disposing mutation, with a higher proportion of carriers in the CPM (3/11) vs other surgical
group (1/9). Concern about contralateral BC and recurrence were common among women
who chose CPM. Women stated they sleep better at night due to their decision, did not
want to take chances, and that if BC recurred, they had tried at least everything. Other
contributors to choice of CPM included cosmetic symmetry and anxiety about surveillance.
For women who kept their contralateral breast, identified themes associated with their choice
included a desire to breastfeed in the future and to look normal and look my age.

Conclusions: Young women who choose CPM often are concerned about future BC events,
despite the fact that risk of future breast events is low and CPM does not improve survival,
suggesting some have difficulty accepting even a small level of risk and gain peace of mind as
a result of choosing more extensive surgery. There is a need to develop novel, targeted
interventions to help women to manage their concerns in the short and long-term, ensuring
decisions are patient-centered and evidence-based.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S57

CORRESPONDING AUTHOR: Shoshana Rosenberg, ScD, MPH, Dana-Farber Cancer Institute,

Boston, MA, 02215; [email protected]
S58 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A022 6:00 PM-7:00 PM

THE EYE OF THE HURRICANE: AN INTEGRATED MIND-BODY MEDICINE PROGRAM FOR

INDIVIDUALS TREATED IN A UNIVERSITY CANCER CLINIC (UCC)

Brian Arizmendi, MA1, Mark Gilbert, MD2, Catherine Shisslak, Ph.D.3

1
Department of Psychology, University of Arizona, Tucson, AZ; 2Department of Psychiatry,
Banner University of Arizona Medical Center, Tucson, AZ; 3University of Arizona, Tucson, AZ

Considerable evidence has emerged supporting the benefits of mind-body medicine

techniques adjuvant to traditional medical interventions. Many studies have examined various
aspects of this practice including meditation, relaxation, and psychoeducation. However, few
programs utilize a truly integrated approach that interweaves psychoeducation, experiential
practice, and group process elements simultaneously leaving participants to synthesize
separate components on their own. Extant studies are conducted in structured research
settings which lack the veridicality of a multidisciplinary UCC. We examined the effectiveness
of an integrated mind-body medicine program in cancer survivors treated at a UCC. Methods:
Utilizing a quasi-experimental pre-post design, this study examined the effect of a 10-week
mind-body medicine program combining aspects of the traditional skill-building model, with
psychoeducation and group process elements. Changes in psychological symptomatology and
well-being were measured in 45 individuals at various stages of cancer treatment. Results:
Paired-sample T-tests comparing patients pre- and post-intervention yielded consistent
program effects for decreased generalized anxiety (GAD-7), t(44) = 5.24, pt(43) = 3.87, p <
.001, and increased well-being (WHO-5) t(44) = -4.27, p < .001. Average change on these
measures are all considered to be clinically meaningful changes based on extant normative
data. Discussion: This study supports the effectiveness of a comprehensive mind-body
medicine group that simultaneously incorporates skill-building, psychoeducational, and group
process components towards significant positive shifts in patient well-being. Results reflect
statistically reliable and clinically meaningful improvements in anxiety, perceived stress
symptomatology and quality of life, adding to the literature supporting mind-body medicine
practices. Methods for implementing such a group in a multidisciplinary community
environment will be discussed.

CORRESPONDING AUTHOR: Brian Arizmendi, MA, Department of Psychology, University of

Arizona, Tucson, AZ, 85721; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S59

A023 6:00 PM-7:00 PM

ACUTE EXERCISE EFFECTS ON ANXIETY IN BREAST CANCER SURVIVORS

Elizabeth Awick, MS1, Arthur Kramer, PhD2, Kendrith Rowland, MD3, Edward McAuley, PhD4

1
University of Illinois at Urbana-Champaign, Urbana, IL; 2Northeastern University, Boston, MA;
3
Mills Breast Cancer Institute, urbana, IL; 4University of Illinois at Urbana Champaign, Urbana,
IL

Many women report increased levels of anxiety and depression after a breast cancer
diagnosis. While past research suggests that acute bouts (i.e., single sessions) of exercise may
be beneficial in reducing such emotional distress, much of this work is grounded in high
intensity exercises. However, negative side effects of the disease such as fatigue and reduced
physical function make it difficult for many survivors to engage in such activities. The purpose
of the present study was to examine the effects of moderate-intensity walking on anxiety and
depression in breast cancer survivors. Breast cancer survivors (N=25; M age=49.81) engaged
in 30 minutes of supervised, moderate-intensity aerobic exercise that consisted of walking on
a treadmill at 40-60% of their maximal heart rate, as determined by a graded maximal
exercise test. Immediately before and after the exercise session, participants completed the
Hospital Anxiety and Depression Scale (HADS). Repeated measures analyses of variance
revealed a significant time effect for the anxiety subscale [F(1,23)=8.78, p=.007, 2=0.28], such
that the women demonstrated significantly reduced anxiety post-exercise. There were no
significant changes in the depression subscale (p=.21). Additional between-subjects analyses
revealed a significant time by age interaction [F(1,23)=3.36, p=.039, 2=0.17], suggesting that
women over the age of 50 demonstrated greater reductions in anxiety compared to women
aged 50 and younger. The present findings are encouraging in light of the host of negative
psychosocial consequences that accompany a cancer diagnosis. This study suggests that
breast cancer survivors who suffer from increased anxiety levels may engage in a moderate
intensity walk for 30 minutes for positive mood changes. Additionally, survivors over the age
of 50 may benefit more from such an exercise bout. Future studies should determine how
long these effects may last for prolonged anxiety relief.

CORRESPONDING AUTHOR: Elizabeth Awick, MS, University of Illinois at Urbana-Champaign,

Urbana, IL, 61801; [email protected]
S60 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A024 6:00 PM-7:00 PM

AN EXAMINATION OF MEDIATORS IN THE RELATIONSHIPS BETWEEN PERCEIVED CANCER RISK,

HEALTHY EATING, AND PHYSICAL ACTIVITY

Tya M. Arthur, PhD, MPH, CPH, Cassie Maxwell, BS

Ball State University, Muncie, IN

Background: Cancer continues to be a major health problem in the U.S. There were an
estimated 1.69 million new cases of cancer in 2016. Approximately one-third of the most
common cancers could be prevented through maintaining a healthy weight and engaging in
health promoting behaviors (e.g., healthy eating, physical activity). There is some evidence
that perceived cancer risk influences engagement in health promoting behaviors; however,
the mechanisms behind this relationship need further examination. The purpose of this study
was to (a) examine the associations between perceived cancer risk (PCR), fruit and vegetable
(F&V) consumption, and physical activity (PA), and (b) separately test the indirect effects of
PCR on F&V consumption and PA as mediated by cancer worry (CW) and the belief that health
promoting behaviors (HPBs) matter in cancer prevention. Methods: The study sample
included 3,677 participants (Mean Age=51.95, SD=21.17) from the fourth cycle of the 2014
Health Information National Trends Survey. More than half of the sample self-identified as
non-Hispanic White (53.3%) and had a household income of < $75,000 per year (62.4%). Path
analysis was used to test the direct and indirect effects of PCR on F&V consumption and PA
while controlling for race/ethnicity, age, and income. Results: PCR was inversely associated
with F&V consumption (r=-0.24, p < .001) and PA (r=-0.23, p < .001). The indirect effects of
PCR on F&V consumption and PA through CW were both statistically significant as well as the
indirect effects of PCR on F&V consumption and PA through the belief that HPBs matter in
cancer prevention. Conclusions: The results have implications for health messages and
intervention strategies to increase health promoting behaviors and prevent cancer. Cancer
prevention researchers and practitioners should not only educate people about the threat of
cancer but also consider strategies that build affect with regard to cancer (e.g., cancer worry
emotions) and promote the belief that health promoting behaviors have an influence on
preventing cancer.

CORRESPONDING AUTHOR: Tya M. Arthur, PhD, MPH, CPH, Ball State University, Muncie, IN,
47306; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S61

A025 6:00 PM-7:00 PM

AN INTERVENTION TO IMPROVE SYMPTOMS OF DISTRESS IN CAREGIVERS OF PATIENTS WITH

STEM CELL TRANSPLANT: FINAL RESULTS

Alaina Carr, BA1, Tanisha Josi, PhD2, Kristin Kilbourn, PhD, MPH3, Jean Kutner, MD, MSPH4,
Mark L. Laudenslager, PhD5, Nicole Amoyal Pensak, PhD2

1
University of Colorado at Denver, Denvwr, CO; 2University of Colorado Anschutz Medical
Campus, Aurora, CO; 3University of Colorado at Denver, Denver, CO; 4University of Colorado
School of Medicine, Auora, CO; 5University of Colorado Denver Anschutz Medical Campus,
Aurora, CO

Purpose: Caregivers of patients with Autologous Hematopoietic Stem Cell Transplant (auto-
HSCT) experience significant burden and distress including anxiety and depression. This
qualitative study explored valuable feedback of caregivers regarding a psychoeducation and
skills-based (Pep-Pal) program. We aimed to identify caregiver perspectives in terms of the
look and feel, acceptability, usability, and feasibility of Pep-Pal as a web-based intervention
for reducing caregiver distress.

Methods: We conducted individual qualitative interviews with 9 caregivers of patients with

auto-HSCT. Each caregiver participated in a semi-structured interview that consisted of
watching an introduction video and three full-sessions. The interview guide encompassed five
primary domains 1) look and feel 2) content 3) usability 4) acceptability and 5) feasibility of
the program. Applied thematic analysis was conducted on each interview to draw out
themes. The authors (NAP and TJ) independently reviewed the transcripts and discussed
themes emerging from the data. They applied open coding to the transcripts and developed
an initial codebook. After discussing the codebook and agreeing upon the codes, author TJ
coded all transcripts independently and identified broad themes within each domain. These
themes used further developed the Pep-Pal program.

Results: Many themes emerged from the interviews. In terms of look and feel, an
overwhelming response felt the animation was off-putting or cheapened the message the
video had to provide. Many caregivers felt different aspects of several videos took away from
the messages being communicated, for example, the background music being too loud during
a body can exercise or intimacy video. In terms of content, several caregivers reported feeling
validated by the intervention because they felt it accurately reflected their experiences and
the challenges associated with their role. Many caregivers requested more specific examples
about the caregiver experience related to cancer and that the program be used in addition to
S62 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

one on one support. All caregivers felt they could navigate the sessions on their own with little
to no difficulty and some even went further and said it was convenient and easy to use.
Themes emerged regarding improvements for the program and included adding pauses or
instructions on how and when to complete exercises. Participants felt using the Pep-Pal
mobilized program would be an acceptable way to get support. Caregivers noted the brevity
of the sessions, especially the mini-peps, and the semi structured nature of the sessions with
flexibility to move around from topic to topic make it more acceptable to use.

Conclusion: The qualitative results from the caregiver individual interviews were integrated
into the final development of Pep-Pal with the goal of enhancing dissemination, engagement,
acceptability, and usability.

CORRESPONDING AUTHOR: Alaina Carr, BA, University of Colorado at Denver, Denvwr, CO,
80202; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S63

A026 6:00 PM-7:00 PM

ANALYSIS OF NATURAL LANGUAGE: A NOVEL APPROACH TO UNDERSTANDING THE THEMES

OF BREAST CANCER SURVIVORSHIP

Amelia Stanton, BA, Jennifer Currin-McCulloch, LMSC, OSW-C, Barbara Jones, PhD

The University of Texas at Austin, Austin, TX

When individuals are diagnosed with breast cancer, they receive an onslaught of information
about medical treatments, potential changes to their body image, future side effects, and
shifts in their familial and occupational roles. Post-diagnosis, some may turn to social media
for resource information and peer support. Social media offers unique advantages for health
professionals seeking to answer complex research questions. Namely, the quantity of natural
language data that can be gleaned from social media is enormous, and the anonymity of the
Internet may help facilitate the discussion of difficult subjects. The purpose of this study was
to assess themes that emerge when individuals talk about breast cancer in a real-world
settingthat is, outside of the laboratoryin order to inform practitioners who are providing
treatment to this population. Natural language (n = 1051 posts) was extracted from a breast-
cancer related subsection of a popular social media website and analyzed via the meaning
extraction method (MEM), an advanced computerized text analysis technique that identifies
words that relate to specific implicit themes across texts. The procedure extracted seven
unique themes germane to breast cancer: diagnosis, social support, risk, existentialism,
treatment process, information seeking, and surgery. Each of these themes was comprised of
frequently used words across posts. This approach validated other themes that have been
noted in previous qualitative studies and uncovered specific nuances of the breast cancer
experience that have not yet been explored in laboratory-based research. These results also
offer health care providers with insight on relevant topics to address with this population
during treatment or therapy.

CORRESPONDING AUTHOR: Amelia Stanton, BA, The University of Texas at Austin, Austin, TX,
78751; [email protected]
S64 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A028 6:00 PM-7:00 PM

BELIEFS ABOUT CANCER DIAGNOSIS AND THEIR IMPACT ON COMORBID DIABETES

MANAGEMENT

Kimberly Muellers, MPH, Juan Wisnivesky, MD, DrPH, Jenny J. Lin, MD, MPH

Icahn School of Medicine at Mount Sinai, New York, NY

Background

With improved survival and prognosis for many cancers, it is increasingly important to
understand how a cancer diagnosis may impact patients long-term health and how beliefs
about cancer impact management of comorbidities. We undertook this study to assess the
effect of beliefs about cancer on comorbid diabetes management.

Methods

We recruited patients with type II diabetes who were newly diagnosed with early-stage
breast, prostate, lung or colorectal cancer and enrolled age-, gender- and hemoglobin A1c-
matched diabetes patients. All patients completed the Beliefs and Illness Perceptions
Questionnaire (BIPQ) for diabetes, and cancer patients also answered the BIPQ for cancer.
Diabetes management behavior was based on the Summary of Diabetes Self-Care Activities
(SDSCA) questionnaire (On how many of the last seven days did you test your blood sugar?
and On how many of the last seven days did you test your blood sugar the number of times
recommended by your health care provider?). Non-parametric analyses were used to assess
the relationship between beliefs about cancer (or diabetes) and diabetes self-management
behaviors.

Results

To date, we have recruited 141 patients, of which 55 were newly diagnosed with cancer. The
average age was 62.3 years and 57% were male. Of the cancer patients, 53% had prostate
cancer, 34% had breast cancer, 3% lung cancer and 9% colon cancer. Compared to non-cancer
patients, cancer patients tested their blood glucose levels (2.9 vs. 4.1, p=0.018) or followed
recommended glucose monitoring (3.6 days vs. 4.9 days, p=0.017) on fewer days in the past
week. However, there was no significant difference between cancer and non-cancer patients
diabetes beliefs, including perceived control and effectiveness of treatment. Beliefs about
diabetes were also not associated with blood glucose monitoring frequency or adherence.
Among cancer patients, perceived control over cancer was associated with blood glucose
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S65

monitoring adherence. Those with low perceived control over their cancer reported blood
glucose testing 1.6 days per week (SD=2.29) compared to 3.9 days among patients with high
perceived control (p=0.003). Patients with low perceived control over cancer also monitored
as recommended fewer days per week (2.3 vs. 4.8, p=0.007).

Conclusion

Our results indicate that patients with newly diagnosed cancer and comorbid diabetes had
poorer adherence to glucose monitoring recommendations than patients with no cancer
history. While cancer and non-cancer patients did not differ significantly in their beliefs about
diabetes, cancer patients who felt they had less control over their cancer were less adherent
to glucose monitoring. This finding suggests that beliefs about cancer, rather than beliefs
about diabetes, may have a stronger impact on cancer survivors management of their
comorbid diabetes.

CORRESPONDING AUTHOR: Kimberly Muellers, MPH, Icahn School of Medicine at Mount

Sinai, New York, NY, 10029; [email protected]
S66 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A029 6:00 PM-7:00 PM

BEYOND DISTRESS ASSESSMENT: LEVERAGING TECHNOLOGY FOR PSYCHOSOCIAL SCREENING

AND DYNAMIC REFERRAL

Cori McMahon, Psy.D.1, Tina Harralson, PhD2, Laura Dietzen, M.S.3, Rebecca Watson, BA3,
Philip Fizur, MA4, Kelly Lynn. Gilrain, PhD5, Mark Redlus, BA6

1
Cooper University Hospital, Department of Medicine, Woodbury, NJ; 2Polaris Health
Directions, Wayne, PA; 3Polaris Health Directions, Inc, Wayne, PA; 4MD Anderson Cancer
Center at Cooper, Cooper University Hospital, Camden, NJ; 5Cooper University Hospital,
Camden, NJ; 6Polaris Health DIrections, Inc, Wayne, PA

Based on the American College of Surgeons Commission on Cancer (CoC) standards, all
accredited cancer centers were required to implement procedures for distress screening by
the year 2015. As distress is often unrecognized and greatly impacts quality of life and cancer
outcomes, effective screening for and management of distress can improve overall health
functioning. Nationally, accredited cancer centers have implemented distress screening
procedures in a variety of ways based on resources and efficiency of work flow, including
paper-and-pencil questionnaires or face-to-face report with medical providers, for example.
Given the opportunity to achieve regular distress screening procedures, MD Anderson Cancer
Center at Cooper University Hospital established a more comprehensive psychosocial
screener system by leveraging technology for patient report and dynamic referral to
appropriate providers. The purpose of this study is to determine the utility of a computerized
psychosocial screener that employs iPads, interfaces with the electronic medical record, and
assesses distress as well as other important issues. Using the Polestar platform, 379
subjects were screened with mean age 61.3 (range: 21-90); with top 3 disease sites: 35%
breast cancer, 13% lung cancer, and 13% unknown. Seventy-seven % were Caucasian, 19%
African American; and 55% married. 59% reported moderate to severe distress, 23%
screened positive for both anxiety and depression (PHQ2 & GAD2); 53% reported at least a
little functional disability due to physical issues and 31% due to emotional issues. The
majority of subjects reported good social support (82% always get help, 78% someone listens)
and few problems with personal relationships (86% no problems). Regarding substance use,
21% acknowledged risky drinking, 21% reported smoking and 11% of those are interested in
quitting smoking. Finally, other problems reported include financial (27%) and transportation
(7%). Based on specific criteria, referrals were triggered to appropriate services, including:
Behavioral Medicine (62%), Social Work (50%), Dietician (31%), and Spiritual Counseling (6%).
Implementation of this screening system significantly increased the number of referrals sent
to various supportive care services (Behavioral Medicine increased by 41-623% monthly in a
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S67

two-year period), thus demonstrating the ability to identify patients in need and to do so
efficiently via dynamic electronic referral and report placed in EMR for ease of use for
providers.

CORRESPONDING AUTHOR: Cori McMahon, Psy.D., Cooper University Hospital, Department of

Medicine, Woodbury, NJ, 08096; [email protected]
S68 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A030 6:00 PM-7:00 PM

BODY IMAGE CHANGES IN BREAST CANCER SURVIVORS DURING A WEIGHT MAINTENANCE

INTERVENTION

Rebecca L. Hunter, Ph.D.1, Christie Befort, PhD2

1
Rush University Medical Center, Downers Grove, IL; 2University of Kansas Medical Center,
Kansas City, KS

Breast cancer treatments can have significant effects on a womans body image due to the
treatment itself (e.g., surgery) or due to treatment-related side effects (e.g., weight gain).
While previous work suggests improvements across dimensions of body image during a
weight loss intervention, little is known about subsequent changes in body image during
weight maintenance or weight regain. This study evaluated changes in body image dimensions
in breast cancer survivors during a weight maintenance intervention and predictors of those
changes, including demographic characteristics, breast cancer treatment history, baseline
depressive symptoms, percent weight regain, and physical activity change. The Body Image
and Relationships Scale (BIRS) assessed 6 dimensions of body image relevant for breast cancer
survivors: changes in social activities due to treatment-related symptoms, energy and
strength, discomfort/embarrassment due to treatment-related appearance changes, body
integrity (feeling natural and whole), sense of control over health, and sexual attractiveness.
While enrolled in a weight control trial, participants lost an average of 13.65.7% of baseline
weight during the 6-month group phone-based weight loss intervention. All 6 body image
dimensions and total score improved upon completion of the weight loss intervention (effect
sizes ranged from d=0.5 to 1.4), although percent weight loss and physical activity were not
significant predictors. Participants (n=75, age=58.68.1 years, time since treatment=3.62.4
years, baseline BMI=34.44.1 kg/m2) regained an average of 36.444.8% of lost weight during
a 12-month randomized phone- and newsletter-based weight maintenance intervention. Five
body image dimensions with the exception of changes in social activities worsened upon
completion of the weight maintenance intervention (effect sizes ranged from partial 2=0.06
to 0.18). Multiple linear regressions revealed that married women experienced less worsening
of discomfort due to appearance (t=2.0, p < .05) and sense of control (t=2.1, p < .05) than their
single counterparts. Higher baseline depressive symptoms predicted greater loss of control
(t=-2.1, p < .05). Notably, women who regained more weight experienced greater worsening
of energy and strength (t=-2.5, p < .05), discomfort due to appearance (t=-4.0, p < .001), and
body integrity (t=-2.0, p < .05). Treatment history and changes in physical activity were not
significant predictors. In contrast to the impact of weight loss, weight regain may directly
impact dimensions of body image in breast cancer survivors. However, the magnitude of
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S69

change was less profound during the weight maintenance intervention, suggesting that
survivors may still benefit from improved body image when compared to their body image
prior to engaging in the weight loss intervention.

CORRESPONDING AUTHOR: Rebecca L. Hunter, Ph.D., Rush University Medical Center,

Downers Grove, IL, 60515; [email protected]
S70 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A031 6:00 PM-7:00 PM

BREAST AND OVARIAN CANCER RISK MANAGEMENT FOLLOWING TELEPHONE VS. IN-PERSON
GENETIC COUNSELING AND TESTING

Hannah Segal1, Beth N. Peshkin, MS, CGC2, Marc D. Schwartz, PhD3, Heiddis Valdimarsdottir,
PhD4

1
Georgetown University Medical Center, Alexandria, VA; 2Georgetown University,
Washington, DC; 3Georgetown University/Lombardi Cancer Center, Washington, DC;
4
Reykjavik University, Reykjavik, N/A, Iceland

Genetic counseling and testing is central to the clinical care of women at risk for hereditary
breast and ovarian cancer (HBOC). Women at high risk following testing can reduce their risk
through surgery or increased screening. In a multi-site randomized noninferiority trial, we
compared telephone genetic counseling/testing (TC) to standard in-person counseling/testing
(SC) for HBOC. TC was noninferior to SC on all genetic counseling/psychosocial outcomes.
Here, we compare TC to SC on cancer risk management outcomes 12-months post-
randomization.

Women 21-85, referred for genetic counseling at one of the study sites, without a new or
metastatic cancer diagnosis were randomized to TC vs. SC. Of 669 women randomized, 514
completed the 12-month follow up. We found no differences in attrition between TC and SC.

Although not powered to test noninferiority on risk management outcomes, we compared TC

vs. SC on use of risk reducing mastectomy (RRM) and oophorectomy (RRO), and breast and
ovarian cancer screening (mammography; MRI; CA125; ultrasound). In bivariate analyses, TC
and SC did not differ on any risk management outcome (p>.10). In separate multivariate
logistic models, we included significant bivariate predictors of the outcome along with test
result and group assignment. RRM was predicted by: positive test result (OR=2.5, 95% CI =
1.04, 6.2) and cancer distress (IES; OR=1.3, 95% CI = 1.04, 1.5); mammogram was predicted
by: positive test result (OR=4.1, 95% CI = 1.8, 9.4), being married (OR=2.7, 95% CI = 1.4, 5.4),
being above age 50 (OR=5.2, 95% CI = 2.2, 12.5), and having a lower a priori mutation risk
(OR=.74, 95% CI = .58, .94); MRI was predicted by: positive test result (OR=7.4, 95% CI = 3.8,
14.9) and being affected with BrCa (OR=3.7, 95% CI = 2.1, 6.7); RRO was predicted by: positive
test result (OR=16.7, 95% CI = 7.2, 38.5) and being married (OR=6.2, 95% CI = 2.2, 17);
ultrasound was predicted by: positive test result (OR=2.6, 95% CI = 1.3, 5.4), being affected
with BrCa (OR=2.4, 95% CI = 1.4, 4.2) and higher cancer distress (OR=1.2, 95% CI 1.1., 1.4);
CA125 was predicted by positive test result (OR=2.8, 95% CI = 1.5, 5.5). We also evaluated the
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S71

group by test result interaction effect in each model. This effect was significant for CA125. SC
participants with a positive BRCA test result were more likely to obtain a CA125 than TC
participants who tested positive.

Despite being underpowered for noninferiority analyses on risk management outcomes, these
traditional analyses are consistent with our prior findings indicating that TC is a safe and cost
effective approach to increase access to genetic testing. The finding that mutation carriers in
the TC arm were less likely than those in the SC arm to undergo ovarian cancer screening
should be interpreted in light of the fact that ovarian cancer screening is not recommended
for BRCA mutation carriers given evidence that it does not reduce ovarian cancer mortality.

CORRESPONDING AUTHOR: Hannah Segal, Georgetown University Medical Center, Alexandria,

VA, 22314; [email protected]
S72 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A032 6:00 PM-7:00 PM

BREAST DENSITY KNOWLEDGE AND INTENTIONS FOR BREAST CANCER SCREENING IN A

DIVERSE SAMPLE OF WOMEN AGE ELIGIBLE FOR MAMMOGRAPHY

Marimer Santiago-Rivas, PhD1, Shayna Benjamin, BA1, Janna Andrews, MD2, Lina Jandorf, MA1

1
Icahn School of Medicine at Mount Sinai, New York, NY; 2Hofstra School of Medicine,
Hempstead, NY

While there is considerable evidence showing the association between breast density and
breast cancer risk and detection, little is known about womens awareness and knowledge of
their own breast density, and its effect on the likelihood of completing breast cancer
screening. The objective of this study was to assess breast density knowledge, breast density
awareness, and intention to complete screening for breast cancer in a diverse sample of
women age eligible for mammography. We quantitatively assessed breast density awareness
and knowledge in black, Latina, and white women recruited online (N=164) and in the
community (N=100). The majority of the participants reported having heard about breast
density (69.2%), but less than one third reported awareness regarding the New York State
Breast Density Notification Law (30.9%) and less than one third knew their own breast density
status (30.4%). Language preference for data collection (English), site of data collection
(online), breast density awareness, knowledge of own breast density, and responding yes to
Would knowing your breast density make you feel informed to make decisions regarding
your breast health? predicted higher likelihood of intending mammogram completion in the
future. An important implication of these findings is that many women may not have access
to important health information that may influence their decision making regarding their
breast health care.

CORRESPONDING AUTHOR: Marimer Santiago-Rivas, PhD, Icahn School of Medicine at Mount

Sinai, New York, NY, 10029; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S73

A033 6:00 PM-7:00 PM

CANCER DISTRESS SCREENING AT AN URBAN SAFETY NET HOSPITAL: PATIENT

CONSIDERATIONS AND ITS IMPACT ON SUPPORT SERVICES

Stephen Lo, MA1, Leanne Ianniello, BS/MPH2, Malti Sharma, RN3, Diane Sarnacki, RN, MSN,
AOCN4, Kathleen T. Finn, MSN, ANP, AOCN5

1
Boston Medical Center, Columbus, OH; 2Boston Medical Center, Brighton, MA; 3Boston
Medical Center, somerville, MA; 4Boston Medical Center, Framingham, MA; 5City of Hope,
Duarte, CA

Background:

Distress screening and providing support services to cancer patients were points of emphasis
made by the US Institute of Medicine. This study explored the relationship between Distress
Thermometer (DT) screening, support service usage, and other patient characteristics at an
urban safety-net hospital.

Methods:

From January 2009 to December 2013, medical records of 3485 patients with complete data,
a cancer diagnosis and pursued care with radiation and/or medical oncology at an urban
safety-net hospital were retrospectively reviewed; the DT was implemented in 2011. x2 tests
were used to evaluate differences in DT usage, patient characteristics (race/ethnicity, gender,
language, marital status, education, insurance and number of chronic illness) and support
service utilization (patient navigation, social work, and behavioral health) between patients
starting care before and after the implementation of the DT. In post-DT implementation
patients, multiple regression analyses were conducted to evaluate whether DT screening and
other patient characteristics predicted the use of support services and an ANOVA with post-
hoc Tukeys tests were used to find differences in support service and DT usage on based on
insurance type. Opinions on the experience using the DT were evaluated 27 oncology staff
from June to August 2014.

Results:

DT usage significantly increased in patients who started care after DT implementation, but
only one-third of patients were successfully screened for distress in a span of 3 years since the
implementation. Patients screened for distress had greater odds of utilizing patient navigation
S74 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

(OR=2.3; 95% CI=1.7-3.0) and social work (OR=1.6; 95% CI=1.2-2.0) after controlling for
patient characteristics. More patients on Medicare and Medicaid utilized patient navigation
and social work than those with private insurance. No difference in the DT screening based on
insurance type was found. The majority of staff reported the DT to be helpful in deciding
referral to support services; however, patient interviews were thought to be more effective
than using the DT. Patient language and literacy were reported to be the largest barriers to
screening.

Conclusion:

The impact on DT usage on support service usage is unclear given a potential selection bias in
the study and additional support service staff is needed to address cancer distress. Future
research to efficiently operationalize distress screening when caring for vulnerable
populations is warranted.

CORRESPONDING AUTHOR: Stephen Lo, MA, Boston Medical Center, Columbus, OH, 43201;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S75

A034 6:00 PM-7:00 PM

CANCER PATIENT PERSPECTIVES REGARDING PREPAREDNESS FOR END-OF-LIFE CARE: A

QUALITATIVE STUDY

Allen C. Sherman, PhD, Stephanie Simonton-Atchley, PhD, Cindy W. Mikeal, MS, Kendra
Anderson, MS, Konstantinos Arnaoutakis, MD, Laura F. Hutchins, MD, Issam Makhoul, MD,
Fade Mahmoud, MD, Natasha Milojkovic, MD, Sarah E. Harrington, MD, James Y. Suen, MD

Winthrop P. Rockefeller Cancer Institute, University of Arkansas for Medical Sciences, Little
Rock, AR

The extent to which patients feel prepared for end-of-life care may be associated with
important clinical outcomes, such as decisions about aggressive treatment, receipt of care
consistent with preferences, healthcare costs, and quality-of-life. Despite growing interest in
the concept of "preparedness," however, there is insufficient information about what cancer
patients actually need to feel prepared. This qualitative study examined patient perspectives
re: preparedness for end-of-life care, in a cohort with terminal disease.

In-depth interviews were conducted with cancer patients who had advanced-stage disease
and limited life-expectancies. Enrollment continued until data saturation. Mean age of
participants was 63.7 years, 39.8% were non-white, and the sample was diverse with respect
to type of malignancy and socioeconomic status. Interviews were audiotaped and
transcribed. Thematic analysis was undertaken by 3 investigators, each of whom coded all
transcripts. Derivation of codes and themes was facilitated by use of NVivo 11 software.

Six overarching themes emerged. These included concerns about: (1) ) end-of-life decisions
(e.g., aggressive treatment vs. comfort care, location of care); (2) interactions with the
medical team (e.g., pursuit vs. avoidance of prognostic information, symptom control,
availability of resources); (3) interactions with family/friends (e.g., burdening the family,
support vs. conflict, communication); (4) emotional well-being (e.g., distress, life completion,
existential suffering); (5) spiritual well-being (e.g., comfort vs. alienation); and (6) financial
well-being (e.g., estate planning, medical expenses).

Findings highlight areas that patients themselves regard as necessary for a sense of
preparedness. The ability to manage these challenges was seen as critical at this phase of
care. Results offer an important foundation for further research regarding instrument
development, prediction of clinical outcomes (e.g., care concordant with preferences, ICU
admissions, quality-of-life), and development of targeted interventions.
S76 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Allen C. Sherman, PhD, Winthrop P. Rockefeller Cancer Institute,

University of Arkansas for Medical Sciences, Little Rock, AR, 72205; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S77

A035 6:00 PM-7:00 PM

CHILDHOOD ADVERSITY PREDICTS TRAJECTORIES OF INFLAMMATION, FATIGUE, AND

DEPRESSION DURING BREAST CANCER SURVIVORSHIP

Avelina C. Padin, MA1, Stephanie J. Wilson, PhD2, Janice K. Kiecolt-Glaser, PhD2

1
The Ohio State University, Columbus, OH; 2The Ohio State University Wexner Medical Center,
Columbus, OH

Breast cancer survivors commonly experience symptoms such as depression, fatigue, and
inflammation during and after treatment. For many women, symptoms improve in the
months following treatment, but a subset of survivors experience persistent symptoms years
after treatment ends. Prior research indicates that childhood adversity predicts elevated
symptoms before, during and in the year following cancer treatment; however, studies have
not assessed whether adversity produces differential trajectories of symptoms over longer
periods of time. The current study assessed childhood maltreatment as a risk factor for poorer
short-term and long-term psychological and physical recovery during survivorship. In a parent
study, breast cancer survivors stages 0-IV (N = 112) provided data prior to primary cancer
treatment (i.e., surgery, radiation, or chemotherapy) and 6 and 18 months post-treatment.
The current study utilized scores from the Childhood Trauma Questionnaire, Center for
Epidemiological Studies-Depression (CES-D) scale, and Multidimensional Fatigue Symptom
Inventory (MFSI-SF), as well as inflammatory marker plasma C-reactive protein (CRP). Women
who reported more physical abuse (b = 0.04, SE = 0.02, p = .047), emotional abuse (b = 0.04,
SE = 0.02, p = .014), and emotional neglect (b = 0.04, SE = 0.02, p = .023) evidenced greater
increases in CRP from baseline to 6 months post-treatment compared to women who
reported fewer adversities. Women who reported more childhood physical abuse also had
greater increases in fatigue (b = 0.29, SE = 0.12, p = .014) and marginally persistent depressive
symptoms (b = 0.37, SE = 0.22, p = .091) from baseline to 18 months post-treatment
compared to those who reported less abuse. Sexual abuse and physical neglect did not predict
symptom recovery. Given these results, it appears that childhood adversity, particularly
physical abuse, has lasting effects on depression, inflammation, and fatigue following breast
cancer treatment. For women with a history of adversity, these symptoms may persist up to
18 months after treatment, long after the physical and psychological repercussions of cancer
treatment are expected to dissipate. Identifying childhood adversity as a risk factor for
persistent inflammation, fatigue, and depression has important implications for survivorship,
as these symptoms have been associated with increased risk for metastasis, recurrence, and
morbidity among cancer survivors.
S78 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Avelina C. Padin, MA, The Ohio State University, Columbus, OH,
43210; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S79

A036 6:00 PM-7:00 PM

CLINICAL AND PSYCHOSOCIAL RISK FACTORS FOR RESIDUAL PAIN IN ADULTS ON A LONG-
ACTING OPIOID FOR CANCER-RELATED PAIN

Emily Wright, BA1, Areej El-Jwahri, MD2, Jennifer S. Temel, MD3, Lara Traeger, PhD2

1
Massachusetts General Hospital, Boston, MA; 2Massachusetts General Hospital/Harvard
Medical School, Boston, MA; 3Massachusetts General Hospital Cancer Center/Harvard
Medical School, Boston, MA

Two thirds of U.S. adults with advanced cancer experience cancer-related pain, despite the
clinical use of opioids in this population. Given multiple opioid formulations and recent
widespread media coverage of opioids, patients medication taking behaviors and their
perceptions of medications may be related to pain outcomes. This study examined clinical and
psychosocial risk factors for residual pain among patients on a long-acting (LA) opioid for
cancer-related pain.

We analyzed cross-sectional data from ambulatory adult patients (n=100) with locally
advanced or metastatic cancer who had a LA opioid prescription for nociceptive cancer pain
and were responsible for administering their medications. Participants reported demographic
information, perceptions of pain medications (e.g., concerns about efficacy, addiction, side
effects), LA opioid adherence in the past week (0-100%), and pain severity (Brief Pain
Inventory). Cancer type and LA opioid regimen were obtained from health records. Clinical
and psychosocial risk factors for pain were explored using multiple linear regression at p <
.05.

Participants (53% female, 89.7% non-Hispanic white, M age=57.5 years [SD=12.3]) were
diagnosed with diverse cancers (thoracic 27%, genitourinary 22%, sarcoma 20%, breast 17%,
gastrointestinal 13%, gynecologic 1%). LA opioid regimens included oral medication (8, 12 or
24 hrs; 74%) or duragesic patch (48 or 72 hrs; 26%), with morphine equivalent dose M=109.3
mg (SD=127.8). 46.4% of participants reported at least moderate pain (4/10). In a linear
regression model that controlled for patient age, gender, and racial/ethnic minority status,
more negative perceptions (B=.32[SE=.12], p=.009) and use of an oral LA opioid (vs. duragesic
patch; B=1.18[SE=.53], p=.027) were associated with higher pain. Opioid dose
(B=.001[SE=.002], p=.49) and self-reported LA opioid adherence (B=.24[SE=.82], p=.77) were
not associated with pain.

Almost half of patients reported at least moderate pain. Patients who had more concerns
S80 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

about pain medications and who were on daily oral doses reported greater pain, regardless of
opioid dose or self-reported adherence. These findings support that patients attitudes and
medication taking patterns may be related to pain outcomes. Longitudinal studies are needed
to examine pain medication taking behaviors and identify modifiable targets for improving
pain management in patients with chronic cancer pain.

CORRESPONDING AUTHOR: Emily Wright, BA, Massachusetts General Hospital, Boston, MA,
02141; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S81

A037 6:00 PM-7:00 PM

CLINICAL EVALUATION OF EMOTIONAL RESPONSES IN PATIENTS WITH CANCER: DIAGNOSTIC

ACCURACY COMPARED TO TWO REFERENCE STANDARDS

Wendy Aerts, MSc1, Joost Dekker, PhD2

1
VU University Medical Center, Amsterdam, Noord-Holland, Netherlands; 2VU University
Medical Centre, Amsterdam, Noord-Holland, Netherlands

Objective. Previous research has suggested that the clinical evaluation by doctors and nurses
of emotional responses in patients with cancer is not always adequate. However, we
hypothesize that doctors and nurses may be differentiating between emotional responses
that do or do not necessitate professional care. The aim of this study was to explore the
diagnostic accuracy (sensitivity and specificity) of clinical evaluation of emotional responses
compared to two reference standards: (i) emotional distress as reference standard, and (ii)
care needs related to emotional problems as reference standard.

Methods. This study was designed as a retrospective cohort study. The clinical evaluation of
emotional responses was derived from the medical and nursing file. The Distress
Thermometer and Problem List were used to assess emotional distress and care needs related
emotional problems. Sensitivity, specificity and the diagnostic odds ratio (DOR) of clinical
evaluation were calculated, compared to two reference standards, i.e. (i) emotional distress
and (ii) care needs related to emotional problems.

Results. The sample included 120 patients: in 35.7%, the doctor or nurse made notes
describing fear, anxiety, stress, low mood or depression, or referred the patient to a mental
health professional. According to the Distress Thermometer, 36.8% of the patients
experienced emotional distress; and 11.3% expressed a need for care related to emotional
problems. The sensitivity of clinical evaluation was 51%, specificity 65% and DOR 1.94 with
emotional distress as reference standard. With care needs as reference standard, the
sensitivity of clinical evaluation was 77%, specificity 63% and DOR 5.77.

Conclusions. As expected, we found indications that when using care needs as reference,
sensitivity of clinical evaluation of emotional responses is higher compared to emotional
distress as reference. This provides preliminary support for our hypothesis that doctors and
nurses distinguish between emotional responses that do or do not necessitate professional
care.
S82 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Wendy Aerts, MSc, VU University Medical Center, Amsterdam,

Noord-Holland, 1081 HL; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S83

A038 6:00 PM-7:00 PM

MERITORIOUS AWARD WINNER

COMMUNITY IMPLEMENTATION OF THE BIOBEHAVIORAL INTERVENTION FOR CANCER

PATIENTS: A COLLABORATIVE STUDY OF PATIENT OUTCOMES

Claire C. Conley, MA, Barbara L. Andersen, PhD

The Ohio State University, Columbus, OH

Objective: Evidence-based treatments (EBTs) for cancer patients are available, but have not
been disseminated for wide use among community mental health practitioners. In a
dissemination effort, mental health professionals were trained to deliver the Biobehavioral
Intervention (BBI) at their home institutions. A collaborative group reports the effects of BBI
on patient outcomes.

Methods: Therapists from 11 sites (82% community settings; hospitals [n = 5, 46%], Cancer
Support Community facilities [n = 3, 27%], oncology practice [n = 1, 9%]) delivered the BBI as
part of standard care. To maximize external validity, therapists in each setting determined the
BBI format (i.e., individual therapy, open group, or closed group), duration, and components
of treatment delivered. Therapists reported treatment integrity data at each session. Cancer
patients (N=110) completed self-report measures (including the Profile of Mood States
[POMS]) pre- and post-BBI. Intervention effectiveness was tested using hierarchical linear
multiple regression analyses. Exploratory analyses examined the relationship between
significant patient outcomes and treatment integrity variables.

Results: As in BBI efficacy studies, patients POMS scores significantly decreased from pre-to-
post (R2=0.1, =-0.3, pr=-0.3, p < .01). Using the Reliable Change Index, 26% of patients
demonstrated clinically significant improvement on the POMS. POMS scores correlated with
BBI group type (r=-0.3, p < .05). M = -22 v. -8). Open group participants evidenced larger
decreases than closed group participants (M=-22 v. -8). POMS scores also correlated with
receipt of BBI in individual format (r=0.2, p < .05). Participants in individual treatment
evidenced smaller decreases than participants in group treatment (M=-2 v. -17).
S84 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions: This is the first study to report patient-level data in the context of multi-site
therapist implementation of an EBT for cancer patients. Data from this collaborative group
provides evidence of the feasibility and effectiveness of the BBI on patient outcomes when
delivered by newly trained mental health providers in the community.

CORRESPONDING AUTHOR: Claire C. Conley, MA, The Ohio State University, Columbus, OH,
43210; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S85

A039 6:00 PM-7:00 PM

COMPARISON OF CANCER BELIEFS BETWEEN APPALACHIANS AND NON-APPALACHIANS AND

RELATED SOCIODEMOGRAPHIC CHARATERISTICS

Katrina J. Serrano, PhD1, Elise L. Rice, PhD2, Minal Patel, PhD, MPH3, Chan Thai, PhD, MPH4,
Kelly D. Blake, ScD5, Robin Vanderpool, BS, MPH, DrPH6

1
National Cancer Institute/Behavioral Research Program, Bethesda, MD; 2National Cancer
Institute, Bethesda, MD; 3National Cancer Institute, Rockville, MD; 4Santa Clara University,
Santa Clara, CA; 5Behavioral Research Program, Division of Cancer Control and Population
Sciences, National Cancer Institute, National Institutes of Health, Bethesda, MD; 6University of
Kentucky College of Public Health, Lexington, KY

Background: Residents of the 13-state Appalachian region experience poorer health

outcomes compared to other geographic populations, including higher rates of cancer
incidence and mortality. Studies suggest that cultural beliefs and practices may influence
these health disparities. Previous research shows that Appalachians have higher cancer risk
perceptions and negative beliefs about cancer but few studies have examined the associations
between Appalachian residence, socio-demographics, and cancer belief variables.

Objective: This study examines cancer-related beliefs between Appalachians and non-
Appalachians, as well as correlates (i.e., Appalachian residence, socio-demographics, and
health status) of those beliefs.

Methods: Data from four administrations (2011-2014) of the National Cancer Institutes
Health Information National Trends Survey were combined and analyzed (n=14,451).
Descriptive analyses were conducted to examine four items related to cancer beliefs between
Appalachians and non-Appalachians. Regression analysis with the entire sample was
conducted to examine the correlates of a composite variable representing cancer beliefs.

Results: About 8% (n=1,015) of the survey respondents were from Appalachia. A higher
percentage of Appalachians vs. non-Appalachians agreed with statements such as everything
causes cancer (69.8% vs. 64.4%) and there were too many recommendations about cancer
prevention (79.4% vs. 72.7%). A higher percentage also believed that they were likely to get
cancer in their lifetime in general (44.0% vs. 41.6%) as well as compared with people their age
(23.1% vs. 18.7%). Regression analyses with the entire sample showed that Appalachians had
higher overall negative cancer beliefs than non-Appalachians (=0.37, p < 0.05). Older vs.
younger respondents had lower negative cancer beliefs (=-0.03, p < 0.05). Additionally, all
S86 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

other racial/ethnic groups had lower negative cancer beliefs compared with non-Hispanic
Whites (p < 0.05). Those with a lower education (e.g., high school: =0.25, p < 0.001) and
health status (e.g., poor health: =1.32, p < 0.001) had higher negative cancer beliefs
compared with those with a bachelors degree and excellent health, respectively.

Conclusions: Appalachians held more negative beliefs about cancer than non-Appalachians,
and residing in Appalachia was significantly associated with greater negative cancer beliefs.
Tailored communication strategies and interventions for certain socio-demographic groups
are needed to provide individuals with accurate information about cancer risk, and to dispel
negative cancer beliefs that may be associated with and contribute to Appalachian cancer
disparities.

CORRESPONDING AUTHOR: Katrina J. Serrano, PhD, National Cancer Institute/Behavioral

Research Program, Bethesda, MD, 20892; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S87

A040 6:00 PM-7:00 PM

CONFIRMING RESEARCH RESULTS AND OBTAINING CANCER SCREENING 6 MONTHS AFTER

LEARNING CDKN2A VARIANT MUTATION STATUS

Carmen Radecki Breitkopf, PhD1, Emma Leof, BA2, Jennifer B. McCormick, PhD, MPP3, Kari G.
Chaffee, M.S.1, Gloria M. Petersen, PhD1

1
Mayo Clinic, Rochester, MN; 2University of Minnesota Twin Cities, Minneapolis, MN;
3
Pennsylvania State University College of Medicine, Hershey, PA

This study examined whether participants who agreed to learn genetic research results would
pursue confirmatory testing in a clinical laboratory of a germline CDKN2A mutation (as
recommended), which predisposes to pancreatic cancer and melanoma. Post-disclosure
behavioral intent to pursue pancreas and skin exams was assessed via self-report immediately
after disclosure and at 6 months. Participants were enrollees in a Mayo Clinic Pancreas
Research registry and who gave a blood sample. Samples were tested in a research laboratory
for the CDKN2A mutation. 133 at-risk individuals without pancreatic cancer were invited by
mail into the disclosure study and to complete a baseline survey, followed by a telephone call
in which they would learn their result. Eighty individuals (60%) returned a baseline survey, and
73 (91.3%) learned their mutation status and participated in semi-structured interviews
immediately post-disclosure; 72 completed the interview, and one withdrew prior to
completion. Six months following disclosure, a survey was mailed which asked about
confirmatory genetic testing and uptake of cancer screening (response rate of 87.3% (n=62)).
Of 62 participants with complete (interview and survey) data, 14 (22.5%) were mutation
carriers and 48 (77.5%) were non-carriers. In the post-disclosure interviews, 29 (46.8%)
participants (71.4% (n=10) carriers and 39.6% (n=19) non-carriers) indicated intent to pursue
confirmatory testing. At the 6-month follow-up, 21.4% (n=3) of carriers and 2.1% (n=1) of non-
carriers reported obtaining confirmatory testing. With regard to cancer screening, 84.6%
(11/13) of carriers stated they were definitely planning on getting a pancreas check in the
next 6 months, as compared to 12.5% (n=6) of non-carriers (p < 0.0001). All carriers indicated
that they definitely planned to get a skin exam in the next 6 months, compared to 59.6%
(28/47) of non-carriers (p=0.003). Six-month follow-up data revealed that 29.4% (n=5; all were
carriers) of those with a strong intention to get a pancreas check did, while 56.1% (n=23; 9
carriers and 14 non-carriers) of those with a strong intention to get a skin exam did. Eleven
(78.6%) carriers stated that they were definitely planning to get a pancreas check in the next
year, compared to 4.2% (n=2) of non-carriers (p < 0.0001), and 84.6% of carriers compared to
31.9% of non-carriers intended to get a skin exam in the next six months (p=0.001). Recipients
of genetic research results relating to cancer predisposition may not follow-through with their
S88 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

intentions to pursue recommended confirmatory testing, yet carriers at potentially increased

risk of pancreatic cancer and melanoma appear to use the genetic information to motivate
participation in cancer detection/screening behaviors.

CORRESPONDING AUTHOR: Carmen Radecki Breitkopf, PhD, Mayo Clinic, Rochester, MN,
55905; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S89

A041 6:00 PM-7:00 PM

COPING AND ENDOCRINE THERAPY ADHERENCE AMONG WOMEN WITH BREAST CANCER

Emma E. Bright, M.A.1, Annette L. Stanton, Ph.D.2

1
University of California, Los Angeles, Los Angeles, CA; 2UCLA, Los Angeles, CA

Rates of non-adherence to endocrine therapies (i.e., tamoxifen, aromatase inhibitors) among

women with breast cancer remain high despite their demonstrated effects on survival. In this
longitudinal study, we examined contributors to adherence among survivors of breast cancer
(N = 130). Participants with breast cancer receiving their initial endocrine therapy prescription
were recruited from a large community oncology breast clinic. Participants took part in three
interview and questionnaire sessions at the clinic at prescription initiation, one month and
four months later. Participants completed a final interview and questionnaire session by
phone 12 months later. Questionnaires addressed coping processes used to manage the
cancer experience. At the initial appointment, participants were provided a Medication Event
Monitoring System (MEMS) cap to assess adherence; the MEMS cap was collected at the four-
month follow-up appointment. The MEMS cap measurement, the percent of days a dose was
taken, was regressed on hypothesized predictors of adherence. The final regression model
accounted for 22.1% of the variance in adherence. Changes in coping accounted for 18.5% of
the variance (p < .05), and relevant medical/demographic factors accounted for 3.6% of the
variance (p > .05). An increase in coping through emotional support seeking from study entry
to one month was significantly associated with lower adherence at 4 months (r = -0.26, p <
.05), while an increase in coping through positive reinterpretation and growth from study
entry to one month was significantly associated with higher adherence at 4 months (r = .19, p
< .05). Increases in coping through seeking emotional support may reflect a deficit in
emotional support provided. Taken collectively, these findings addressing womens cognitive
appraisals and support needs regarding the cancer experience might promote endocrine
therapy adherence.

CORRESPONDING AUTHOR: Emma E. Bright, M.A., University of California, Los Angeles, Los
Angeles, CA, 90025; [email protected]
S90 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A042 6:00 PM-7:00 PM

DESCRIPTION OF PHYSICAL ACTIVITY LEVELS AMONG MYELOPROLIFERATIVE NEOPLASM

PATIENTS ENGAGING IN AN ONLINE YOGA INTERVENTION

Ryan Eckert, B.S.1, Jennifer Huberty, PhD 2, Krisstina Gowin, D.O3, Linda Larkey, Ph.D.1,
Matthew Buman, Ph.D.1, Ruben Mesa, M.D.3

1
Arizona State University, Phoenix, AZ; 2Arizona State University School of Nutrition and
Health Promotion, Phoenix, AZ; 3Mayo Clinic Cancer Center, Phoenix, AZ

Introduction: Cancer patients report lower physical activity (PA) levels than those that are not
diagnosed with cancer, particularly during treatment and often times persisting into cancer
survivorship. Myeloproliferative neoplasm (MPN) cancer patients are unique in the sense that
these individuals often spend the rest of their life undergoing treatment to manage disease
progression and symptom burden. MPN patients report a high symptom burden (i.e., fatigue,
depression, anxiety, sleep disturbance, etc.) with inactivity reported as a symptom in upwards
of 60% of patients. It has been demonstrated that PA can improve a variety of physical and
psychosocial side effects of cancer and its treatment. To date there is no information
regarding PA levels in MPN patients. Therefore, the purpose of this abstract is to summarize
the objective PA levels of MPN patients participating in a 12-week online-streamed yoga
intervention.

Methods: Participants were MPN patients (n=38) enrolled in a 12 week online-streamed yoga
intervention. Participants were instructed to wear a Fitbit Flex on their non-dominant wrist
throughout the study. The Fitbit Flex provides an estimate of daily step counts as well as time
spent in sedentary, light, moderate, and vigorous activity. Trajectories of objectively-
measured PA were tested using linear mixed models with fixed and random effects for time
included in the model.

Results: At baseline, MPN patients averaged 871.841.5 min/week of sedentary time,

204.813.9 min/week of light activity, 19.33.5 min/week of moderate-vigorous physical
activity (MVPA), and 6466.9652.3 steps/day. Throughout the course of the 12-week study,
there were no changes in sedentary time, light activity, or steps. However, there was a
significant downward trend in daily MVPA (-0.050.02; p=.034) over the 12 weeks.

Conclusion: MPN patients engaged in only 19.33.5 min/week of MVPA at baseline (as
measured by a Fitbit) of the 12-week yoga intervention, and the time spent in MVPA
decreased significantly over the course of the study. However, time spent in sedentary and
light activities did not change over the course of the study, nor did daily step counts. Future
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S91

research is warranted to objectively measure PA in a larger sample of MPN patients, examine

unique methods to improve PA behavior, and to identify if increased PA is efficacious for
improving symptom burden in MPN patients.

CORRESPONDING AUTHOR: Ryan Eckert, B.S., Arizona State University, Phoenix, AZ, 85004;
[email protected]
S92 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A043 6:00 PM-7:00 PM

DETERMINANTS OF MULTIPLE HEALTH BEHAVIOR CHANGE IN CANCER SURVIVORS WHO

SMOKE

Emily Cox-Martin, PhD1, Karen Basen-Engquist, PhD, MPH2, Janice A. Blalock, Ph.D.3

1
Department of Medicine, University of Colorado Anschutz Medical Campus, Denver, CO; 2The
University of Texas MD Anderson Cancer Center, Houston, TX; 3University of Texas MD
Anderson Cancer Center, Houston, TX

Objective: This project investigated associations between determinants of health behavior

change for smoking cessation and physical activity (PA) in cancer survivors. Improving our
understanding of the relationships between common psychological determinants for different
health behaviors may help improve intervention efficiency and effectiveness through multiple
health behavior change. Methods: Participants included cancer survivors who were current
smokers and had an initial assessment at the Tobacco Treatment Program at MD Anderson
Cancer Center. Patients who had been abstinent for over six months (i.e., in maintenance)
were excluded. Measures of self-efficacy and stage of change were administered for smoking
cessation and PA. Results: Participants (n = 76) identified mostly as white (79%) and male
(58%) with a variety of cancer diagnoses. For smoking cessation stage of change participants
were mostly in contemplation (20%) or preparation (70%); none were in precontemplation
and few were in action. For PA, participants showed a wide range of stage of change:
precontemplation (9%), contemplation (21%), preparation (38%), action (15%) and
maintenance (16%). A significant correlation was demonstrated between self-efficacy for
smoking cessation and self-efficacy for PA (r = .445, p = .002). Stages of change for smoking
cessation and PA were not related. The relationship between self-efficacy and stage of
change for PA was significant [F(4,55) = 4.058, p =.006], but not for smoking cessation [F(2,52)
= 1.399, p=.256]. Relationships for stage of change and self-efficacy across behaviors were
non-significant. Conclusion: A significant relationship was found between self-efficacy for
these two behaviors but not for stage of change. Furthermore, a relationship between these
two determinants was evidenced for PA but not across behaviors. Survivors approaching
smoking cessation may have generalized efficacy for health behavior change, but low
motivation for changing more than one behavior at a time. Motivational interventions to
capitalize on high levels of self-efficacy for both behaviors and influence stage of change
might be applicable.

CORRESPONDING AUTHOR: Emily Cox-Martin, PhD, Department of Medicine, University of

Colorado Anschutz Medical Campus, Denver, CO, 80230; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S93

A044 6:00 PM-7:00 PM

EXAMINING POSTTRAUMATIC GROWTH AS A PREDICTOR OF HEALTH-RELATED OUTCOMES IN

PROSTATE CANCER SURVIVORS

Elizabeth L. Addington, PhD

Northwestern University Feinberg School of Medicine, Chicago, IL

Background: Due to high incidence and survival rates of prostate cancer, survivors outcomes
are important public health issues. Some men with prostate cancer experience positive
psychological changes known as posttraumatic growth (PTG), which is posited to operate
through psychophysiologic and behavioral pathways to ameliorate survivors health and well-
being. However, scant research has tested outcomes of PTG, particularly those related to
physical health. This study therefore examined PTG as a predictor of health and health
behaviors in a sample of 168 prostate cancer survivors. Methods: On average, participants
completed baseline measures at 6 months post-diagnosis and follow-up surveys at 5 years
post-diagnosis. Demographic and medical covariates were obtained from self-report and
medical records. Using the PTG Inventory-Short Form, participants indicated the extent of
positive changes in 5 domains (relationships, spirituality, new possibilities, personal strength,
appreciation for life) as a result of their prostate cancer. Physical activity was measured with 6
items that assess aerobic activity and strength training, allowing for calculation of weekly MET
minutes. Fruit and vegetable consumption was measured with the NCI Food Frequency
Questionnaire. Participants rated their general health from poor to excellent. Results: The
sample included 58% White and 42% Black men; 99% were diagnosed with locally or
regionally confined prostate cancer. Average levels of PTG were moderate (M=2.75). Health
behaviors tended to decline over time, such that less than half the sample at 5 years post-
diagnosis met activity and dietary guidelines for cancer survivors. In multiple regression
analysis controlling for socioeconomic status and medical/health indicators (e.g., education,
income, baseline health, Charlson comorbidity index), PTG (b=.12, p=.005) independently
predicted prostate cancer survivors self-rated health at approximately 5 years post-diagnosis
(total R2=.41, change in R2=.03, p=.005). PTG was not significantly related to physical activity
or to fruit and vegetable consumption, nor did those health behaviors mediate the
relationship between PTG and health. Conclusions: Among 5-year survivors of prostate
cancer, PTG positively predicted self-rated health, above and beyond socioeconomic and
medical indicators. Future studies should continue to examine outcomes related to PTG in
diverse samples of cancer survivors, using objective measures when possible. In addition,
interventions are needed to improve and sustain health behaviors among prostate cancer
survivors, many of whom failed to meet recommended levels of physical activity and fruit and
vegetable consumption.
S94 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Elizabeth L. Addington, PhD, Northwestern University Feinberg

School of Medicine, Chicago, IL, 60647; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S95

A045 6:00 PM-7:00 PM

A FORMATIVE EVALUATION STUDY OF AGENCY LEADERS WHO SERVE VULNERABLE

POPULATIONS AS PART OF THE CALIFORNIA FIRST FIVE INITIATIVE

Kristen T. Ruckstuhl, MS, BA1, Stephanie A. Brito, BS1, Mengyu Liu, MS1, Chih-Ping Chou, PhD1,
Lourdes Baezconde-Garbanati, PhD, MPH, MA1, Susana Ramirez, PhD, MA2, Mary Ann Pentz,
PhD, MA1

1
University of Southern California, Department of Preventive Medicine, Institute for Health
Promotion and Disease Prevention Research, Los Angeles, CA; 2University of California,
Merced, Public Health Communication, Merced, CA

Background: This pilot study aims to address health disparities research on chronic disease
prevention in low-income Hispanic/Latino families with children under the age of five. Agency
leaders were randomly selected from a larger list of health care agency directors and
providers who are charged with delivering evidence-based prevention programs as part of the
California First Five Initiative in counties that include a focus on vulnerable populations in
urban and rural communities in Southern and Central California, respectively. The main
objective was to learn about their ideas and experiences in an effort to determine how to best
engage and keep such families enrolled in health programs. Methods: Two members of the
research team conducted phone interviews from JulySeptember 2016. One member
facilitated the phone interviews, while another recorded the responses in separate, de-
identified Word documents for qualitative analysis in ATLAS.ti. Interview questions sought to
assess the perceived usefulness of 12 different health programs that are either already part of
the California First Five Initiative or are being considered on a pilot basis. The response rate
was 100% for all organizations interviewed. Both team members then analyzed key words and
phrases in all interviews for each of the nine questions. In order to assess inter-rater
reliability, kappa coefficients were computed across all participants (N=24) for each interview
question. A score of 0.8 was used as a cutoff to demonstrate high agreement. The kappa
coefficients for questions 1-9 ranged from 0.8042-1.0 with a total kappa coefficient of
0.9356. Results: 13/24 (54.2%) agency leaders were from rural communities, while 11/24
(45.8%) were associated with urban environments. Conclusions: Of the 12 health programs,
these 5 garnered the most support: 14/24 (58.3%) endorsed working together in groups to
improve access to health programs; 12/24 (50%) endorsed discussing family health needs at
home with a promotora (i.e., a trained person from the community); 11/24 (45.8%) endorsed
taking 10 minutes a day to do easy physical activity at home; 12/24 (50%) endorsed learning
about ways to relax and reduce stress for mothers; and 13/24 (54.2%) endorsed assisting
mothers with family weight management and infant eating. Increased implementation of such
S96 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

health programs is likely to reduce health risk behaviors (e.g., sedentary behavior, low
physical activity, and poor nutrition).

CORRESPONDING AUTHOR: Kristen T. Ruckstuhl, MS, BA, University of Southern California,

Department of Preventive Medicine, Institute for Health Promotion and Disease Prevention
Research, Los Angeles, CA, 90032; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S97

A046 6:00 PM-7:00 PM

THE IMPACT OF CUMULATIVE RISK ON ASTHMA OUTCOMES IN CHILDREN AND ADOLESCENTS

LIVING IN AN URBAN ENVIRONMENT

Samantha A. Miadich, M.A. 1, Jeremy Barsell, B.S.1, Lariel V. Turner, Student2, Robin S.
Everhart, PhD1

1
Virginia Commonwealth University, Richmond, VA; 2Viriginia Commonwealth University,
Hopewell, VA

Objective: Children and adolescents with asthma from low-income, urban families are more
likely to experience high levels of stress, which can further compound their risk for poor
asthma outcomes. A cumulative risk model may be a more accurate depiction of how risk
factors impact health outcomes in that risk exposures do not occur in isolation. This ongoing
study investigated the association between a cumulative risk model (comprised of poverty,
caregiver stress, family functioning, child discrimination, neighborhood stress) and child
asthma outcomes (quality of life [QOL], asthma control, functional severity, quick-relief
medication use). This study also assessed whether the cumulative risk model was a stronger
predictor of asthma outcomes than a single risk exposure.

Methods: Participants included 94 children/adolescents (64% male) with asthma 7-17 years of
age (M=11.85 years, SD=2.86) and their primary caregivers (92% African American). Children
and caregivers completed questionnaires at an initial research session.

Results: The cumulative risk model significantly predicted child QOL (R=.17, F(1,93)=18.54, p
< .001), ), asthma control (R=.09, F(1,93)=9.35, p=.003), functional severity (R=.06,
F(1,93)=6.09, p=.015), and quick-relief medication use (R=.08, F(1,93)=7.95, p=.006). Higher
cumulative risk indices were associated with worse asthma outcomes. Additionally, the
cumulative risk model predicted QOL above and beyond any single risk factor. However, the
cumulative risk index was not a stronger predictor than neighborhood stress when assessing
functional severity, quick-relief medication, and asthma control. In addition, it was not a
stronger predictor than poverty when assessing functional severity and quick-relief
medication use.

Conclusions: Findings suggest that an accumulation of stress related to urban living may be
associated with poorer QOL, worse asthma control, worse functional severity, and more
quick-relief medication use in children and adolescents with asthma. Healthcare providers and
clinicians may need to focus on family-based strategies that reduce the accumulation of stress
S98 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

to improve child asthma outcomes. Future research is also needed to further examine the
independent impact of neighborhood stress and poverty on asthma outcomes, as these
factors were found to be most robust in associations with several asthma outcomes (e.g.,
functional severity, quick-relief medication use).

CORRESPONDING AUTHOR: Samantha A. Miadich, M.A. , Virginia Commonwealth University,

Richmond, VA, 23221; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S99

A047 6:00 PM-7:00 PM

TRAJECTORIES OF WEIGHT FOR LENGTH GROWTH FOR INFANTS DURING THE FIRST YEAR OF
LIFE

Andrea Lopez-Cepero, BS, MHSN, Lisa Nobel, MSc, Tiffany A. Moore Simas, MD, MPH, MEd,
FACOG, Milagros C. Rosal, PhD

University of Massachusetts Medical School, Worcester, MA

Background: Childhood obesity is a major public health problem. Studies of patterns of child
growth contributing to the development of obesity are scarce, particularly in infancy. Group-
based trajectory analyses among infants are a novel procedure that may help characterize
subgroups of infants with similar longitudinal growth profiles.

Objective: To identify trajectories of weight for length growth during the first year of life.

Methods: Subjects were singleton infants and their mothers (N=90 mother-infant pairs) who
participated in the Pregnancy and Postpartum Observational Dietary Study. Women
completed assessments throughout their infant's first year of life and included socio-
demographic characteristics and feeding behaviors. Infant weight for length measures from
birth to 12 months were abstracted from pediatric office records. Weight for length
percentiles were calculated according to the World Health Organization guidelines for infants.
Group-based trajectory analysis was done to identify subgroups of infants with similar growth
profiles.

Results: Infants were from mothers with average of 28 years (SD=5.2), 70.0% White, 60.0%
high-school educated and 63.2% had two or more children. Over half of mothers introduced
solid foods to their infants by 6 months of age (63.2%) and about one third self-reported
breast feeding at 12 months post-partum (31.9%). Three growth trajectories were identified:
a low and stable growth group (38.3%), a rapid growth group (35.0%) and a moderate growth
group (26.7%). Maternal and feeding variables were all similar across the three infant growth
trajectory groups (p>0.05).

Conclusion: Trajectory models suggested three patterns of infant growth. If replicated, future
studies can help identify and subsequently target modifiable risk factors associated with rapid
infant growth trajectories.

CORRESPONDING AUTHOR: Andrea Lopez-Cepero, BS, MHSN, University of Massachusetts

Medical School, Worcester, MA, 01655; [email protected]
S100 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A048 6:00 PM-7:00 PM

A QUANTITATIVE STUDY OF THE INTEGRATION OF YOGA IN NATURE

Zachary J. Bertone, Ph.D.

Naval Medical Center San Diego, Coronado, CA

A Quantitative Study of the Integration of Yoga in Nature

Objective: Current trends of ubiquitous technology use and significant immersion in artificial
urban environments may be significantly impacting the health of individuals through a
disconnection from both nature and the human body. As alternatives, this quantitative study
examined the practice of yoga and immersion in nature settings as a means to facilitate
greater physical, psychological, spiritual health, and embodiment.

Method: A pre and posttest design was conducted in a sample of adults recruited from the
Silicon Valley, CA. The following measures were collected: Body Awareness Questionnaire
(BAQ), Spiritual Well-Being Scale (SWBS), Quality of Life Questionnaire (QLQ), Exercise
Induced Feeling Inventory (EFI), and blood pressure and salivary cortisol (both obtained once
pre- and post-test). Participants were assigned to one of three groups that engaged in weekly
90 minute activities for six sessions over an 8-10 week period. The experimental group
practiced yoga in a nature setting; the primary control group hiked in a designated nature
area; and the secondary control group practiced yoga in an indoor facility. Repeated measures
ANOVA and t-tests were used for analyses.

Results: A well-educated, racially diverse sample (2.3% African American; 27.3% Asian; 45.5%
Caucasian; 11.4% Hispanic; 4.5% Middle Eastern; 2.2% Native American; 6.8% Other) of 44
participants (15 men; 29 women) enrolled. Within subject effects for EFI 1 scores for the
sample as a whole emerged significant. EFI 6 emerged as significant within groups with higher
mean scores for the Outdoor and Studio yoga group. Repeated measures analysis of variance
(ANOVA; p < .05) of all self-report measures did not find significant effects for BAQ, SWBS,
QLQ, cortisol, and blood pressure assessments. A paired t-test found only significant
decreases in diastolic blood pressure effects for the full sample.

Conclusions: This study is the first to quantitatively explore potential differences between
indoor and outdoor yoga and the embodied effects of yoga in an outdoor environment.
Results support findings in previous research on overall physiological health benefits of
exercise and exercise differentiation literature that shows the benefits of yoga over other
exercise modalities.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S101

CORRESPONDING AUTHOR: Zachary J. Bertone, Ph.D., Naval Medical Center San Diego,
Coronado, CA, 92118; [email protected]
S102 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A049 6:00 PM-7:00 PM

AN EXPLORATION OF SHAME, SPIRITUAL PRACTICES, AND COMPASSION IN A COMMUNITY-

BASED GROUP

Gillian D. Grannum, MA, MDiv1, Alexis D. Abernethy, PhD2, David Allen, Director3

1
Graduate School of Psychology, Fuller Theological Seminary, Pasadena, CA; 2Travis Research
Institute/ Fuller Theological Seminary, Pasadena, CA; 3Renascence Institute International, Ft .
Lauderdale, FL

Psychological research has explored the maladaptive role of shame in contributing to poor
self-image, impaired relationships, and psychopathology. Compassion-focused individual and
group therapy has demonstrated effectiveness in repairing shame through increasing self-
acceptance and empathy. Culturally-tailored group interventions, which may include spiritual
practices, may help to mitigate the effects of maladaptive responses to shame, particularly
within communities that have experienced societal fragmentation. The present study
explored how engaging in spiritual practices within a community-based group may be related
to transformation and the alleviation of shame. Semi-structured interviews were conducted
with 16 members of The Family: People Helping People, a Bahamas-based group
resocialization program. Thematic analysis of data was conducted using Interpretive
Phenomenological Analysis and themes were coded with NVivo 11 software. A total of 54
themes were identified and cluster analyses were conducted to assess relationships between
themes based on word similarity. Participants noted that prayer enhanced their group
experience and helped to foster group cohesion and an environment of increased compassion
for others. In addition, group-related factors (particularly interpersonal elements such as
connectedness to group and learning from or teaching other members) were prominent in
participants description of transformation and decreased shame. Compassion for others was
significantly more prevalent than self-compassion; this finding may reflect the unique
characteristics of the cultural context. Research findings indicate that culturally congruent
spiritual practices, including prayer and musical expression, may contribute to restoring
healthy intrapersonal and interpersonal functioning. This research offers support for a holistic
approach to addressing shame that incorporates psychological, social, and spiritual
dimensions.

CORRESPONDING AUTHOR: Gillian D. Grannum, MA, MDiv, Graduate School of Psychology,

Fuller Theological Seminary, Pasadena, CA, 91101; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S103

A050 6:00 PM-7:00 PM

ASSESSING THE KENTUCKY INVENTORY MINDFULNESS SCALE: A COST EFFECTIVE TOOL FOR
HOSPITALS?

Rubi Gonzales, B.A1, Juliana Cardoso, M.A2, Craig Field, PhD, MPH3

1
University of Texas at El Paso, El paso, TX; 2University of Texas at El Paso, El Paso, TX; 3Latino
Alcohol and Health Disparities Research (LAHDR) Center, Department of Psychology, The
University of Texas at El Paso, El Paso, TX

According to the Centers for Disease Control and Prevention (2012), 117 million people
suffer from chronic health conditions and one of four individuals has two or more conditions.
These conditions are common and most importantly preventable; they include heart disease,
diabetes, obesity, stroke, and arthritis. Several treatments are available for theses illness but
such treatments can be a financial burden for the patients. In 2010, approximately 90% of
health care spending was allocated to individuals with more than one chronic illness.

Mindfulness may be defined as a psychological state of awareness of experience

without judgement. Research has demonstrated several benefits to mindfulness, such as
reducing rumination (Chambers et al., 2008), reducing stress (Hoffman et al., 2010), improving
working memory (Jha et al., 2010), and reducing emotional reactivity (Orthner et al., 2007).
The literature has established several benefits; however, limited research has been conducted
on patients in a hospital setting. According to Johansson et al. (2012), a mindfulness base
stress reduction treatment significantly improved self-report for mental fatigue on
neuropsychological tests in patients who suffer from mental fatigue after a stroke or
traumatic brain injury. Furthermore, Tsafou et al. (2016) reported satisfaction as a mediator
on the effect of mindfulness on physical activity.

The present study investigated the psychometric properties of the Kentucky Inventory
Mindfulness Skills (KIMS) scale by Baer, Smith, & Allen, 2004 from an online sample (N=601). A
Principal Component factor analysis was conducted and indicated multidimensionality with 7
factors. Cronbachs Alpha coefficient was .789 and Spearman-Brown coefficient of .794,
indicating good reliability. A chi-square test of goodness-of-fit was performed to determine
whether there is a difference between the distributions (458, N=601)=785.091, p=.000. The
average age of participants was 38.64 (SD= 15.329).

Given the lack of research of mindfulness in clinical settings, one crucial future
direction is to assess the psychometric properties of the scale for hospital setting patients.
The KIMS may serve as a potential cost effective tool for health care providers as a measure of
S104 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

mindfulness strategies in patients; more specifically, such identification may facilitate the
development of specific mindfulness techniques to reduce negative patient health outcomes.

CORRESPONDING AUTHOR: Rubi Gonzales, B.A, University of Texas at El Paso, El paso, TX,
79912; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S105

A051 6:00 PM-7:00 PM

BODY-MIND-SPIRIT GROUP THERAPY FOR CHINESE MEDICINE STAGNATION SYNDROME - RCT

WITH SELF-REPORT AND PHYSIOLOGICAL MEASURES

Siu-man Ng, PhD

The University of Hong Kong, Pokfulam, N/A, Hong Kong

Chinese medicine diagnosis stagnation syndrome is characterized by mind-body obstruction-

like symptoms such as repressed emotions, feeling stomach clogged, something obstructing
throat and heavy head. Its emphasis on somatic symptoms makes it a socially legitimate
health condition for adults in distress to seek help. Stagnation syndrome is common, with a
point-prevalence at 6.2% revealed in our epidemiological study. Our previous studies have
also operationalized stagnation as a construct useful to all mental health practitioners, and
piloted a 6-session group therapy for stagnation syndrome grounded on a body-mind-spirit
intervention model. The current randomized controlled trial aims to further rigorously
evaluate the efficacy of the manualized intervention. Totally 126 participants were recruited
and randomly assigned to either treatment or control group. Outcomes were measured at
pre, post and 2 months after intervention, and evaluated by self-report scales and salivary
cortisol level. Repeated measures ANOVA revealed significant superiority of treatment over
control groups in self-report measures and mean cortisol level, mostly at moderate effect size.
Regarding diurnal slope of cortisol level, changes were nonsignificant. The findings provide
evidence supporting the efficacy of the body-mind-spirit group therapy, which actualizes the
holistic tradition of Chinese medicine in modern health practice. Evaluating the stagnation
concept and group intervention in other cultures is worth pursuing.
This research is funded by Research Grant Council of the Hong Kong SAR Government (Ref. no.
GRF 748013)

CORRESPONDING AUTHOR: Siu-man Ng, PhD, The University of Hong Kong, Pokfulam, N/A, -;
[email protected]
S106 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A052 6:00 PM-7:00 PM

DIVERSE WOMENS PERSPECTIVES ON MIND-BODY PRACTICES THROUGHOUT PREGNANCY

AND BEYOND

Patricia Kinser, PhD, WHNP-BC, RN, FNAP1, Nancy Jallo, PhD, FNP, WHNP, CNS2

1
Virginia Commonwealth University School of Nursing, Richmond, VA; 2Virginia
Commonwealth University, Richmond, VA

Background: Unmanaged perinatal stress poses public health concerns for maternal, child,
and family health. Pregnant and postpartum women who experience high levels of
psychosocial stress are at risk of substance use, poor spousal relationships, suicidality, poor
maternal-fetal/child attachment, and poor maternal self-efficacy. The prevalence and long-
term negative effects of stress on the well-being of the mother, fetus, and child during the
perinatal continuum underscores its importance as a public health problem. Because maternal
stress and the associated negative symptoms of depression and anxiety are potentially
modifiable, addressing maternal stress during this critical time may signify one of the most
feasible strategies for improving the health and well-being of women, infants, and children.
Mind-body interventions are often used for managing stress and depressive symptoms in
pregnant women.

Purpose/Methods: The purpose of this presentation is to report recent research about diverse
womens interest and participation in mind-body interventions for stress and depression in
the perinatal period. Focus groups and interviews with pregnant and postpartum women
were conducted to evaluate the feasibility and acceptability of mind-body practices, such as
guided imagery (GI) and yoga, for stress and depressive symptom management during
pregnancy and in the postpartum period.

Results: Black, White, and Hispanic women who participated in the interviews and focus
groups (n=25) reported that stress and depressive symptoms are pervasive in their lives and in
their community and consistently expressed a strong interest in mind-body practices. Women
who participated in GI (n=15) reported less stress and anxiety with increased feelings of
calmness, peace, and sense of coping. Women who participated in yoga (n=12) as an aspect of
group prenatal care expressed great satisfaction with the ease of access and the community-
support involved in the prenatal yoga sessions.

Conclusions: This study suggests that diverse women experience significant levels of stress,
anxiety, and depressive symptoms during their pregnancies and that a mind-body
intervention may be helpful for coping with stress and minimizing its negative effects. These
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S107

findings are relevant to healthcare providers and researchers who should continue to focus on
the unique needs of pregnant women, particularly when designing prevention and
intervention strategies regarding stress and depressive symptoms.

CORRESPONDING AUTHOR: Patricia Kinser, PhD, WHNP-BC, RN, FNAP, Virginia

Commonwealth University School of Nursing, Richmond, VA, 23298; [email protected]
S108 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A053 6:00 PM-7:00 PM

EXPLORING THE RELATIONSHIP BETWEEN SELF-ESTEEM AND SPIRITUALITY IN PATIENTS WITH

EATING DISORDERS

Andrene Spencer, MA1, Alexis D. Abernethy, PhD2, Kallie Kobold, MA3

1
Fuller Theological Seminary, South Pasadena, CA; 2Travis Research Institute/ Fuller
Theological Seminary, Pasadena, CA; 3Fuller Theological Seminary, Pasadena, CA

Studies have shown that there is a relationship between spirituality and positive health
outcomes for people with eating disorders (EDs). Low self-esteem is one of several risk factors
for EDs. For ED patients, the relationship between self-esteem and spirituality is not well
understood. The present study sought to examine the relationships among self-esteem,
spirituality, and disordered eating. Drawing from a parent study of 390 participants,
qualitative responses from a subsample of 150 adult and adolescent patients who were
treated for an ED at various treatment centers in either residential, partial, or intensive
outpatient programs were examined. Patients responded to two open-ended qualitative
questions about identity and future orientation. Interpretive phenomenological analysis (IPA)
was utilized to identify themes from both questions taken at admission and discharge.
NVivo11, a qualitative analysis software, was used to conduct cluster analyses. A dendrogram
is a type of cluster analysis where themes that are similar based on word usage are clustered
together on the same branch and those that are different are further apart. Self-esteem
themes included low self-worth and a strong sense of self. Spirituality themes included
religious comfort and religious strain. Eating disorder themes included recovery/health and
symptom/problem focused. Themes associated with religious strain were related to low self-
worth and symptom/problem focused while themes associated with religious comfort were
related to recovery/health. These findings suggest that patients who described their
relationship with God in a negative way seemed to focus more on their eating disorder
diagnosis and had a negative view of themselves. Those who described their relationship with
God or a higher being in a positive way tended to have a more positive outlook on life. An
unexpected finding was the shared similarity between strong sense of self and low self-worth.
Responses that were coded as indicating a strong sense of self captured the more positive
aspects of self-esteem as well as a strong sense of identity. It is possible that the themes were
similar since they both pertain to how patients see themselves (e.g., positive, negative, or
neutral). These findings shed light on how ones relationship with a higher being may be
associated with self-esteem and disordered eating.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S109

CORRESPONDING AUTHOR: Andrene Spencer, MA, Fuller Theological Seminary, South

Pasadena, CA, 91030; [email protected]
S110 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A054 6:00 PM-7:00 PM

GENDER REPRESENTATION OF MINDFULNESS BASED CLINICAL TRIALS

Jamie Bodenlos, Ph.D.1, Kathryn Strang, BA2, Juliana M. Heffern, B.S. Psychology3, Kathryn
Lyle, N/A2, Brien Ashdown, Ph.D.2, Rosalind Gray-Bauer, Bachelor of Arts 4

1
Hobart and William Smith Colleges, Baldwinsville, NY; 2Hobart and William Smith Colleges,
geneva, NY; 3Hobart and William Smith Colleges, Geneva, NY; 4Hobart and William Smith
Colleges, Braintree, MA

Mindfulness-based treatments for psychological disorders and health issues have become
mainstream. Studies have confirmed the efficacy of treatments for a wide array of disorders;
however, its unclear if these treatments work as well in males as they do in females.
According to the Evidence Based Practice in Psychology model, its important to have support
for the use of a treatment in the population you plan to treat. In order to understand how
well males have been represented in the mindfulness based literature, we conducted a
systematic review of the published randomized clinical trials testing either Mindfulness-Based
Stress Reduction (MBSR) or Mindfulness-Based Cognitive Therapy (MBCT). We reviewed 120
articles (M number of participants per article = 87.72, S.D. = 75.08). MBCT was used in 68
(56.6%), MBSR was utilized in 51 of the articles (42.5%), and one article utilized both types of
interventions (0.9%). Of the 10,526 total participants, males accounted for less than a third
and only 35% of the articles made it clear that they included ethnic minority participants.
These findings suggest that studies which focus on mindfulness based interventions among
males is needed to ascertain the efficacy of these interventions in this group.

CORRESPONDING AUTHOR: Jamie Bodenlos, Ph.D., Hobart and William Smith Colleges,
Baldwinsville, NY, 13027; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S111

A055 6:00 PM-7:00 PM

HIGH FAITH VS. NO FAITH: MORE EVIDENCE OF A CURVILINEAR RELATIONSHIP BETWEEN

RELIGIOSITY AND HEALTH

R. Brian Giesler, PhD, Natalie Nusinow, BA, Shelby Eaton, BA

Butler University, Indianapoils, IN

Although positive linear associations between religiosity and physical health have frequently
been found in previous research, our recent work with younger adults suggests that these two
variables may be better characterized as being related in a curvilinear fashion. To assess
whether curvilinearity is present in older adults, a sample of residents (mean age=85, sd=5.4,
n=113) from a retirement community in a large Midwestern city were recruited to take part in
a cross-sectional survey study. Health was assessed through self-report measures used in
prior work, including: 1) a current symptom count, 2) a comorbidity count, 3) a single, Likert
type item assessing general health, 4) a single item assessing pain, and 5) a single item
assessing fatigue. Religiosity was assessed using a modified version of Hoges (1972) religious
commitment scale, a validated measure of intrinsic religiosity. As predicted, curvilinear
relationships were found across most of the individual measures of health. When individual
measures were combined into a global measure of health (Cronbachs alpha = .81), a
significant curvilinear relationship was again found (p = .005, R2=.09), such that the most
religious and the most secular reported the highest levels of health. To investigate this
relationship further, participants were categorized into five groups by level of religiosity. A
repeated measures ANOVA, with health measure as the repeated factor, indicated marginally
significant differences across the groups (p < .08): highly religiously committed individuals
tended to show the best health (i.e., fewer symptoms, better self-rated health, etc.), and
health declined as commitment decreased. However, this downward trend was broken by
individuals who were not at all committed (i.e., secularly oriented individuals), who reported
relatively high levels of physical health that most closely approximated those of the
moderately religiously committed. This curvilinear relation between health and religiosity
may have been missed in the past because analyses assuming linearity have generally been
employed in this context. However, across all groups, the most highly religious still appear to
enjoy the highest levels of physical health.

CORRESPONDING AUTHOR: R. Brian Giesler, PhD, Butler University, Indianapoils, IN, 46208;
[email protected]
S112 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A056 6:00 PM-7:00 PM

IMPLEMENTATION OF AN INTEGRATIVE CARE CENTER IN AN NCI DESIGNATED ACADEMIC

CANCER CENTER

Sarah Braun, MS1, Masey Ross, MD, MS2, Carrie Cybulski, BA, PMP3, Mary Ann Hager, MS,
CHPN3, Tara A. Albrecht, PhD, ACNP-BC4

1
Virginia Commonwealth University, Richmond, VA; 2Virginia Commonwealth University
Health System, Richmond, VA; 3VCU Health, Richmond, VA; 4Virginia Commonwealth
University and Massey Cancer Center, Richmond, VA

Purpose: Oncology patients and survivors report increased use and interest in integrative
practices, including meditation, diet, yoga, and more. Research supports the use of integrative
practices as complementary to conventional medicine for the improvement of physical and
psychological symptoms for patients with cancer. However, the implementation of an
Integrative Health Program in an academic and NCI designated Cancer Center within a safety
net hospital is not without challenges. Therefore, we will document the processes
implemented and challenges encountered over the past two years, and discuss future
directions as we work to develop an integrative health program in our NCI designated Cancer
Center.

Methods: An Integrative Health committee was created within the NCI designated Cancer
Center comprised of an interdisciplinary team of healthcare providers, administrators, and
professional practitioners trained in integrative practices. Over the past two years, the steps
to meet the aims for program initiation included development of a conceptual framework,
research initiatives, resource management, and the conduction of a SWOT (strengths,
weaknesses, opportunities, threats) analysis and a needs assessment of both
patients/survivors and health care providers.

Results: Results from monthly meetings and subgroup initiatives of the Integrative Health
committee are discussed including the four pillars identified in the mission statement (Cancer
Rehabilitation, Mind/Body, Psychosocial, Nutrition), program development, creation of a
business plan to address challenges specific to our cancer center, and future directions based
on the results of our needs assessments.

Conclusion: Discussion of the early development of an Integrative Health Program will include
resource discovery, creation of a mission statement and conceptual framework, interpretation
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S113

of the SWOT analysis and needs assessments, and marketing development. Important lessons
for future directions include the need for interdisciplinary and interdepartmental
collaboration, patient-centered care, informed providers, and a well-developed business plan.
Successes, challenges, and specific plans for the future are discussed within the context of an
NCI designated Cancer Center in a safety net hospital.

CORRESPONDING AUTHOR: Sarah Braun, MS, Virginia Commonwealth University, Richmond,

VA, 23223; [email protected]
S114 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A057 6:00 PM-7:00 PM

RELIGION, PHYSICAL ACTIVITY, AND FAT INTAKE

Karen Yeary, PhD, Page Moore, MS, PhD

University of Arkansas for Medical Sciences, Little Rock, AR

A substantial body of literature has documented religions association with health, and health
behaviors have been hypothesized as a sizeable mediator in the religion-health relationship.
Few studies have examined the potential role of physical activity and nutrition in religions
association with health over time. Also, a limited number of studies have examined religions
relationship with health behaviors in underserved populations.

Secondary data analyses using a preexisting data set were conducted to explore associations
between religion, physical activity, and fat-intake among African American adults. The data
set (n=26) was from a single-group pre-posttest study that examined the feasibility of an
evidence-based, 16-week behavioral weight loss intervention. Total physical activity
(Paffenbarger), recreational physical activity (Paffenbarger), fat-intake (NCI Quick Food Scan),
religious attendance, private religiosity, daily spiritual experiences, religious social support,
religious coping, perceived sacredness of the body, religious locus of control, and
demographics were assessed at baseline and 16-week follow-up. Linear regression analyses
were used to explore the relationships between religion, physical activity and fat intake after
controlling for demographic variables (gender, age, employment, marital status, education,
and income) and BMI.

At baseline, there was a statistical trend between anticipated religious support and higher
total physical activity (adjusted =7.7, SE=3.5; p < .10), negative religious coping and higher
fat intake (adjusted =2.6, SE=1.3; p < .10), and private religious activity and lower fat intake
(adjusted =-7.7, SE=4.0; p < .10). Positive religious coping at baseline (adjusted =7.2, SE=1.7;
p < .05), God locus of control at baseline (adjusted =2.0, SE=0.50; p < .05), and church
attendance at baseline (adjusted =26.4, SE=1.8; p < .05) were each significantly associated
with greater total physical activity from baseline to 16-week follow-up. Positive religious
coping at baseline (adjusted =4.4, SE=6.6; p < .05), religious social support given at baseline
(adjusted =5.2, SE=0.82; p < .05), and church attendance at baseline (adjusted =13.3,
SE=3.1; p < .05) were each significantly associated with greater recreational physical activity
from baseline to 16-week follow-up.

Religion may be involved in shaping physical activity among African Americans attempting to
lose weight.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S115

CORRESPONDING AUTHOR: Karen Yeary, PhD, University of Arkansas for Medical Sciences,
Little Rock, AR, 72205; [email protected]
S116 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A058 6:00 PM-7:00 PM

THE EFFECTS OF YOGA ON ADULTS WITH TYPE II DIABETES: A SYSTEMATIC REVIEW

Herpreet Thind, PhD, MPH, MBBS1, Lori A. J. Scott-Sheldon, PhD2, Ryan Lantini, MA3, Brittany
L. Balletto, BS4, Marissa Donahue, M.A Candidate3, Melissa M. Feulner, BS5, Elena Salmoirago-
Blotcher, M.D., Ph.D.6, Beth C. Bock, PhD7

1
University of Massachusetts Lowell, Lowell, MA; 2The Miriam Hospital and Brown University,
Providence, RI; 3The Miriam Hospital, Providence, RI; 4The Miriam Hospital, Cranston, RI; 5The
Miriam Hospital, Milford, CT; 6The Miriam Hospital & Alpert Medical School of Brown
University, Providence, RI; 7Brown Medical School, Providence, RI

Objective: The purpose of this systematic review was to examine the effects of yoga for
glycemic control among adults with type II diabetes.

Methods: Comprehensive electronic databases searches located 2,556 unique studies with
relevant key terms; of these, 21 studies met inclusion criteria. Studies were included if they a)
evaluated a yoga intervention to promote type II diabetes management, b) used an objective
measure to assess glycemic control at post-intervention, and c) were available through
December 2015. Studies were excluded if yoga was not the primary intervention focus (e.g., if
yoga was part of a mindfulness-based intervention). Independent raters coded participant,
design and methodological characteristics and intervention content. Within- and between-
group changes in glycemic control were assessed.

Results: Studies were published between 1992 and 2015; samples included 2,402 participants
(M age = 54 years; 43% women). Most studies (17) were conducted in India; 2 were
conducted in England, 1 in Cuba, and 1 in Iran. Treatment as usual was the most common (15)
control condition; 3 studies used a diabetes education and 3 used an exercise comparison
condition. Interventions were typically conducted over a median of 50 sessions; mean
duration of intervention was 11 weeks. Only five studies reported the style of yoga used in the
intervention (3, Hatha Yoga; 2, Iyengar). Within-group glycemic control improved at post-
intervention among participants exposed to the yoga intervention but between-group
findings compared to comparison group were mixed. Overall, studies satisfied an average of
42% of the methodological quality criteria indicating poor to moderate quality.

Conclusion: Yoga improved glycemic control in adults with type II diabetes from pre- to post-
intervention but the findings were mixed when yoga was compared to a control condition.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S117

Future studies using more rigorous designs with detailed reporting of methodology are
needed to determine the potential benefit of yoga for adults with type II diabetes.

Learning Objectives:
1) Describe the efficacy of yoga interventions for improving glycemic control for type II
diabetes.
2) Identify the gaps in research on the use of yoga for adults living with type II diabetes.

CORRESPONDING AUTHOR: Herpreet Thind, PhD, MPH, MBBS, University of Massachusetts

Lowell, Lowell, MA, 01854; [email protected]
S118 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A059 6:00 PM-7:00 PM

USE OF TRADITIONAL MEDICINES TO COPE WITH CLIMATE-SENSITIVE DISEASES IN A

RESOURCE POOR SETTING IN BANGLADESH

Md Aminul Haque, PhD

University of Dhaka, Dhaka, Dhaka, Bangladesh

Background
This study aims to explore the use of traditional medicines to cope with climate sensitive
diseases in areas vulnerable to climate change. We assessed the extent to which traditional or
alternative medicines were used for the treatment of the climate sensitive diseases by
villagers as part of their health-coping strategies.

Methods
The study deployed a mixed-method research design to know the health-coping strategies of
the people in a resource-poor setting.

A cross sectional study was conducted from September 2010 to March 2011 among 450
households selected randomly in the districts of Rajshahi and Khulna, Bangladesh. The elder
males or females of each household were interviewed. For qualitative methods, twelve focus
group discussions (six with females and six with males) and fifteen key informant interviews
were conducted by the research team, using interview guidelines on the use of traditional
medicine.

Results
Univariate analysis showed that the use of traditional medicines has increased among
community members of all socio-economic and demographic backgrounds. Due to the
increased incidence of disease and sickness respondents had to increase the use of their
cultural means to cope with adverse health situations.

Conclusions
A systematic collection of knowledge on the use of traditional medicines to cope with climate-
sensitive diseases can help the adaptation of communities vulnerable to climate change. In
addition it can be instrumental in creating a directory of traditional medicine components
used for specific diseases and highlight the effectiveness and relevance of traditional
medicines as health-coping strategies. This may be useful for policymakers, researchers, and
development partners to adapt existing health care policy in resource-limited contexts. It may
also encourage WHO, national and international institutions, such as pharmaceutical
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S119

companies, to carry out research investigating the effectiveness of these traditional medicines
and integrate them with modern medicine. Overall, it could increase the health coping
capacity of people in a resource-poor setting and contribute to their adaptation capabilities.

CORRESPONDING AUTHOR: Md Aminul Haque, PhD, University of Dhaka, Dhaka, Dhaka, 1000;
[email protected]
S120 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A060 6:00 PM-7:00 PM

ASSESSING COMMUNITY MEMBERS PREFERENCES OF A FAMILY-BASED DIABETES

PREVENTION PROGRAM IN RURAL ALABAMA

Alexandra Khodadadi, B.A. 1, Susan Andreae, PhD, MPH2, Andrea Cherrington, MD MPH3

1
University of Alabama at Birmingham, Auburn, AL; 2University of Alabama at Birmingham,
Birmingham, AL; 3University of Alabama at Birmingham, Division of Preventive Medicine,
birmingham, AL

INTRODUCTION: Developing strategies to engage all family members in a lifestyle

management program may be a promising strategy to improve the health of families living in
rural underserved communities in Alabama.

OBJECTIVES: The aim of this study was to better understand barriers and facilitators in
engaging the entire family in a family-centered diabetes prevention program.

METHODS: Five focus groups were held in 2016 in a rural, southern Alabama community with
community members who had diabetes or cared for a family member with diabetes. Of the 58
focus group participants, 50 were women, 41 were over the age of 50, 53 had graduated from
college or high school, and 16 worked full or part time. Participants were asked to discuss
their current beliefs about diabetes and obesity and ideas on how to involve all family
members in lifestyle changes. Focus group findings were analyzed by two coders using open
coding.

RESULTS: Several themes emerged indicating family centered programs would be feasible and
acceptable including: high levels of family cohesion, health information shared among family
units and neighbors, and the desire for acquiring health knowledge and skills necessary for
healthy lifestyles as a family unit. Potential barriers included the limited availability of
community level resources, busy schedules, and multiple competing demands. Individual level
barriers included stress and chronic pain in some family members.

CONCLUSIONS: With multiple generations living close by and a strong sense of family, a family
based intervention would be relevant and feasible in this community. However, several
barriers will need to be considered in order to achieve a culturally relevant lifestyle
intervention applicable to the whole family.

CORRESPONDING AUTHOR: Alexandra Khodadadi, B.A. , University of Alabama at Birmingham,

Auburn, AL, 36830`; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S121

A061 6:00 PM-7:00 PM

LINKS OF COMMUNAL COPING TO MOOD AND SELF-CARE BEHAVIOR IN TYPE 2 DIABETES: A

DAILY DIARY STUDY

Melissa Zajdel, M.S.1, Vicki Helgeson, PhD1, Howard Seltman, M.D., PhD1, Mary Korytkowski,
M.D.2, Leslie Hausmann, PhD3

1
Carnegie Mellon University, Pittsburgh, PA; 2University of Pittsburgh, Pittsburgh, PA; 3VA
Pittsburgh Healthcare System, Pittsburgh, PA

Communal coping in the context of couples coping with chronic illness reflects the extent to
which the illness is perceived as a shared problem and couples collaborate to manage the
illness. We sought to determine how communal coping is related to adjustment to chronic
illness in terms of daily mood and daily self-care behaviors. We conducted a daily diary study
of 125 couples in which one person was recently diagnosed with type 2 diabetes ( < 3 years).
Patients were 59% white, 55% male, with a mean age of 54 years. Patients and partners
completed questionnaires separately at the end of each day for 14 days, with an average
completion rate of 10 days. Days were nested within individuals, and within-person
communal coping and between-person communal coping values were created so that the
between subject variance could be statistically controlled in all analyses. Therefore, within-
person communal coping on any given day reflects the amount that an individual deviates
from his or her average level of communal coping. Patient daily communal coping was related
to higher levels of happiness ( = .12), lower levels of anger ( = -.16), higher levels of dietary
adherence ( = .27), and higher levels of medication adherence ( = .41; all ps < .05) for
patients. Partner communal coping was also related to patient outcomes, specifically higher
patient happiness ( = .10), dietary adherence ( = .10), and medication adherence ( = .38; all
ps < .05). When patient and partner communal coping were entered into the same equation
to predict patient outcomes, patient communal coping emerged as the stronger and more
robust predictor. Partner communal coping was related to their own mood higher happiness
( = .21), lower anger ( = -.15), and lower depressed mood ( = -.13; all ps < .05). These
results suggest that communal coping on a daily basis is linked to enhanced mood among both
patients and partners as well as better patient self-care behaviors.

CORRESPONDING AUTHOR: Melissa Zajdel, M.S., Carnegie Mellon University, Pittsburgh, PA,
15206; [email protected]
S122 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A062 6:00 PM-7:00 PM

A COMPREHENSIVE PATIENT NAVIGATOR PROGRAM FOR DIABETES PATIENTS LIVING IN

RURAL APPALACHIA

Liz A. Beverly, PhD

Ohio University Heritage College of Osteopathic Medicine, Athens, OH

In the Appalachian region of rural southeastern Ohio, diabetes rates are 30%-40% higher than
the national averages. Thus, strategies that complement standard medical interventions are
critically important to mitigate the risk of complications and reduce the economic burden of
the disease. The Comprehensive Patient Navigation Program for Rural Appalachia is designed
to improve health outcomes (e.g., glycemic and blood pressure control; reduced depression,
distress) and lower health care expenditures (e.g., reduced admissions, readmissions, and
emergency department utilization) for individuals with type 2 diabetes through the
development and coordinated implementation of a Diabetes Patient Navigator and Medical
Legal Partnership (MLP). To date, 35 patients (mean age=6013 years, A1C=8.92.4%, 61%
female, 100% white) have received navigation and/or MLP services. Qualitative interviews and
baseline surveys revealed the following diabetes management barriers: mental health issues,
food insecurity, lack of insurance coverage, unemployment and/or disability, limited or no
transportation, housing issues, limited finances, and other legal issues (e.g., custody rights,
divorce). Navigation and MLP services have resulted in increases in insurance coverage (n=13),
Social Security Extra Help (n=9), food stamps (n=7), reduced hospital bills (n=6), permanent or
temporary housing (n=5), diabetes supplies at no cost (n=5), and utility repairs (n=5). Only two
patients were hospitalized at follow-up and no patients utilized the emergency department.
Eleven patients demonstrated improvements in diabetes self-care at 6-month follow-up; A1C
values are currently being collected. Preliminary findings suggest that diabetes patients living
in rural southeastern Ohio benefit from a combination of patient navigation and legal
services.

CORRESPONDING AUTHOR: Liz A. Beverly, PhD, Ohio University Heritage College of

Osteopathic Medicine, Athens, OH, 45701; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S123

A063 6:00 PM-7:00 PM

LISTOS CON LISTAS? EVALUATING WHETHER SPANISH-SPEAKING PATIENTS WHO BRING A

LIST A QUESTIONS FOR THE DOCTOR GET THEM ANSWERED.

Iridian Duran, N/A1, Adriana Orellana, N/A1, Brett Cervantes, N/A1, George Heredia, A.S.1,
Karina Galvan-Martinez, N/A2, Karla Garcia, B.S.1, Herlinda Guzman, B.A.1, John Billimek,
Ph.D.1

1
University of California, Irvine, Irvine, CA; 2Health Policy Research Institute at UCI, Irvine, CA

Type 2 diabetes negatively impacts the underserved low-income Hispanic community, giving
rise to inadequate guidance for the proper plan of care due to ineffective communication.
The disconnect between patients and their physicians may lead to medication non-adherence
and poor health outcomes. A proposed solution to improve physician-patient communication
is to prepare patients with a set of questions to ask during their visit. Bringing a list of
questions to medical visits could facilitate communication between the physician and patient.
For this study, community health workers (CHW) assisted patients in determining three
questions to bring for their medical visit. The patients then wrote a list of important questions
they wanted to discuss with their physician. The medical visit was audio recorded from the
entrance of the physician until the encounter was completed. The recordings were then
transcribed and coded by selected audio-coders. Data from the audio recordings were
analyzed and compared to determine if the patients asked their questions and if the physician
answered them. The data indicated that in 49.1% of visits physicians made reference to or
read questions from the patient's question list. Overall, 57 recordings were audio coded with
an average of 2.7+0.6 questions written down, out of those, an average of 2.3+0.8 questions
were asked by patients and an average of 2.2+0.9 questions answered by the physician. A
total of 64.9% (37 of 57) of patients asked all of the questions from their list, and all questions
asked by the patient were answered by the physician in 82.5% of (47 of 57) visits. The results
from this study suggest that preparing lists of questions prior to a doctor visit may be helpful
to promote discussion of important health topics, but that many questions from the lists are
never discussed with the doctor. Further research must be conducted to identify strategies to
improve communication about these issues.

CORRESPONDING AUTHOR: Iridian Duran, N/A, University of California, Irvine, Irvine, CA,
92697-5800; [email protected]
S124 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A064 6:00 PM-7:00 PM

ASSOCIATION BETWEEN WEIGHT LOSS OUTCOMES AND PERCEIVED RISK OF TYPE 2 DIABETES
AMONG PERSONS WITH PREDIABETES

Mona AuYoung, PhD, MS, MPH1, Tannaz Moin, MD, MBA, MSHS2, Laura Damschroder, MS,
MPH1, Caroline Richardson, MD3

1
VA CCMR, Ann Arbor, MI; 2VA Greater Los Angeles Healthcare System, David Geffen UCLA
School of Medicine, Los Angeles, CA; 3University of Michigan, Ann Arbor, MI

Theories such as the Health Behavior Model posit that perceived risk and perceived severity of
health conditions are related to actual health behaviors. Previous studies have shown mixed
results regarding perceived risk and health behaviors. Among chronic conditions, a previous
study indicated diabetes is perceived to be less severe than conditions such as heart disease,
breast cancer, and stroke. Another study found that those with higher perceived risk of
diabetes were not more likely to intend to make healthy lifestyle changes. This study
examines whether perceived diabetes risk is associated with weight loss after participating in
a lifestyle management program.

Eligible Veterans with prediabetes were randomly selected into one of three arms of a
pragmatic trial: in-person diabetes prevention program (DPP), online DPP, and a usual care
weight management program (MOVE!). All three arms encouraged patients to lose weight,
since this can reduce diabetes risk. The analytic sample included consented patients who
completed baseline questionnaires and had 12-month weight outcomes available (either
through clinical data sources or wireless scales). Participants rated perceived risk (chances of
developing type 2 diabetes in their lifetime) and perceived severity (how much diabetes
would affect their life if they were diagnosed) on Likert scales. Multiple linear regression
analyses assessed the association between 12-month weight change and perceived risk and
perceived severity, separately, controlling for sex, age, study arm, race/ethnicity, and baseline
body mass index.

Survey respondents (n=376) had been diagnosed with prediabetes; the perceived risk of
developing type 2 diabetes was fairly evenly distributed from low to high although perceived
severity skewed towards high. There was no association between perceived risk or perceived
severity and 12-month weight change, after controlling for covariates. This lack of association
suggests that perceived risk and perceived severity of diabetes may not be sufficient to
motivate weight loss to prevent diabetes.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S125

CORRESPONDING AUTHOR: Mona AuYoung, PhD, MS, MPH, VA CCMR, Ann Arbor, MI, 48113-
0170; [email protected]
S126 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A065 6:00 PM-7:00 PM

CHALLENGES OF RECRUITING PARTICIPANTS WITH A HEALED DIABETIC FOOT ULCER: THE STEP
UP TO AVERT AMPUTATION IN DIABETES TRIAL

Yvonne Ye, B.S1, Madeline Russell, BA2, Megan MacDougall, BA3, Kimberly Stone, MPH4, Xiao
Xiao, M.S.5, Tova Bergsten, B.A.6, Trina N. Wijangco, B.A.7, Samantha Hill, B.S.8, Sundar
Natarajan, MD9

1
VA Harbor Healthcare System, New York, NY; 2Veteran Affairs Medical Center, New York, NY;
3
Veterans Affairs New York Harbor Healthcare System, La Canada, CA; 4New York Harbor
Healthcare System, New York, NY; 5VA NY Harbor Healthcare System Manhattan Campus,
New York, NY; 6Veteran's Affairs Hospital, New York, NY; 7Veterans Affairs Medical Center,
Manhattan, NJ; 8Veteran Affairs Medical Center - Manhattan, Manhattan, NY; 9Department of
Veterans Affairs, NY, NY

Objectives: Lower extremity amputation (LEA) is a devastating complication of diabetes with

a mortality rate exceeding that of most cancers. LEA is preceded by foot ulcers in >80% of
cases. We examined the challenges to enrolling participants in the context of a clinical trial to
avert ulcer recurrence.

Methods: We identified patients with diabetes and a healed diabetic foot ulcer (DFU) through
an electronic medical record search for patients with ICD-9 codes for diabetes and DFU. The
previously healed DFU was confirmed through medical record review. Interested patients
provided informed consent and underwent a formal screening. Descriptive statistics (numbers
and proportions) illustrate our success in enrolling patients and provide reasons why patients
were ineligible or not enrolled.

Results: From August 2015 to September 2016, we identified 334 patients from an electronic
search and medical record review. Of these patients, 133 (41%) met ineligibility criteria such
as: a current or recent DFU (n=42), an inability to ambulate without assistance (n=21), not
currently taking medication for diabetes mellitus (n=10), lower extremity amputation (n=4),
recent hospitalization or surgery (n=12), and other comorbidities such as cancer (n=6) and
end-stage renal disease (n=2). Among the 201 eligible patients, 55 (27%) were
enrolled. Common reasons for non-enrollment among the remaining patients (n=146)
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S127

include: temporarily not having the time to be screened (n=18, 12%), ineligibility after
providing consent (n=17, 12%), hesitancy in committing to a long-term study (n=14, 10%),
commitment to managing serious coexisting health conditions (n=14, 10%), not being
interested in education (n=7, 5%), denial of DFU history (n=3, 2%), hesitancy due to
unfavorable experiences with the hospital in the past (n=2, 1%), and travel concerns (n=2,
1%).

Conclusions: There are several challenges to recruiting patients with a previous foot ulcer.
Most significantly, the electronic search identifies only a small number of patients and a large
fraction of these patients have either a non-healed DFU or other serious coexisting conditions
that lead to ineligibility. To improve recruitment and generalizability, recruitment strategies
should consider developing a better electronic search, dedicating staff to reaching patients,
enrolling participants with limited mobility, incorporating patients who are not taking
medications, and motivating patients to participate in clinical trials.

CORRESPONDING AUTHOR: Yvonne Ye, B.S, VA Harbor Healthcare System, New York, NY,
10010; [email protected]
S128 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A066 6:00 PM-7:00 PM

COGNITIVE AND BEHAVIORAL PREDICTORS OF SOCIAL RESILIENCE AMONG ADOLESCENTS

WITH TYPE 1 DIABETES (T1D)

Rebecca N. Adams, Ph.D.1, Esti Iturralde, Ph.D.1, Anthony T. Vesco, Ph.D.2, Jill Weissberg-
Benchell, Ph.D.3, Korey K. Hood, Ph.D.1

1
Stanford University School of Medicine, Palo Alto, CA; 2Ann & Robert H. Lurie Childrens
Hospital of Chicago, Chicago, IL; 3Northwestern Universitys Feinberg School of Medicine,
Chicago, IL

Social resilience, including a sense of relatedness and the ability to use relationships to buffer
against stress, is associated with T1D management behaviors and health outcomes. However,
little is known about the core components of social resilience in adolescents with T1D and
what may be modifiable cognitive and behavioral intervention targets. Addressing this gap will
inform intervention strategies to improve health outcomes given nearly 75% of adolescents
with T1D do not meet treatment targets for glycemic control. Guided by cognitive theory, the
current study examined these associations longitudinally and tested a hypothesized mediator.
It was hypothesized that cognitive and behavioral skills (i.e., greater positive problem-solving
(PS) and lower negative PS skills) would predict fewer maladaptive cognitions, which in turn
would predict greater social resilience. Adolescents with T1D (n=264, mean age=15.7, 59.8%
female, 65.5% non-Hispanic white, 70.1% on insulin pump, mean T1D duration=6.9 years)
participating in a clinical trial were assessed 4 times across 1 year. Participants completed
questionnaires to assess positive and negative PS skills, maladaptive cognitions, social
resilience, diabetes self-care, and glycemic control. Mediation effects were tested using 2
time-lagged autoregressive path models, one for each PS skill controlling for intervention
assignment, age, diabetes duration, gender, race, and family income. For negative PS, model
fit was adequate (RMSEA=0.07; CFI=0.92) and the hypothesized mediation effect was found
(CI did not contain 0): over time, greater negative PS skills predicted more maladaptive
cognitions, which in turn led to reduced social resilience. Social resilience in this sample was
significantly correlated with diabetes self-care and glycemic control (ps < 0.05). Effects did not
differ between intervention groups. In contrast, the positive PS model showed poor fit and no
mediation effect. Findings suggest that factors targeted in cognitive-behavioral treatments,
especially negative PS and maladaptive cognitions, impact social resilience and may be
important to consider in interventions that ultimately impact diabetes self-care and glycemic
control.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S129

CORRESPONDING AUTHOR: Rebecca N. Adams, Ph.D., Stanford University School of Medicine,

Palo Alto, CA, 94304; [email protected]
S130 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A067 6:00 PM-7:00 PM

CULTURAL ADAPTATION OF DIABETES SELF-MANAGEMENT EDUCATION FOR U.S. RESIDING

MARSHALLESE

Karen Yeary, PhD1, Nia Aitaoto, BA, MPH, MS and PhD2, Karra Sparks, BS3, Mandy Ritok-
Lakien, BA3, Jonell Hudson, PharmD3, Peter Goulden, MD1, Williamina Bing, BS in
Management3, Sheldon Riklon, MD3, Jellesen Rubon-Chutaro, N/A3, Pearl Mcelfish, PhD,
MBA3

1
University of Arkansas for Medical Sciences, Little Rock, AR; 2University of Arkansas
Northwest, Fayetteville, AR; 3University of Arkansas for Medical Sciences Northwest,
Fayetteville, AR

Type 2 diabetes is a significant public health problem, with U.S. Pacific Islander communities
bearing a disproportionate burden. In one Pacific Islander communitythe Marshallesethe
stark disparities in type 2 diabetes outcomes are due in part to the US Nuclear Testing
Program that tested a payload equivalent to 7200 Hiroshima-sized bombs in the Republic of
Marshall Islands (RMI), which displaced over 12,000 RMI residents to Arkansas, the home of
the largest population of Marshallese in the U.S. Among Arkansas residing Marshallese, rates
of diabetes are ~32% compared to 8% of the general US population.

Despite the appalling health inequities between Marshallese and the general U.S.
population, research with Marshallese has been limited because of data aggregation of
Marshallese and other Pacific Islander groups with Asian Americans, resulting in Marshallese
and other Pacific Islanders groups an invisible population within an already understudied
minority.

Given the weak body of evidence to guide service planning in managing type 2 diabetes
among Marshallese, we used a community-based participatory research approach to adapt an
evidence-based diabetes self- management education (DSME) intervention for Marshallese.
We began the DSME curriculum adaptation by reviewing the results of previously conducted
focus groups (n=41) and pilot studies regarding Marshalleses diabetes beliefs and behaviors.
We combined this previously collected formative data with a literature review to create a
curriculum, which was reviewed using an iterative process with community and academic
experts.

Specific cultural components were recommended for adapting the seven key
components of the DSME curriculum. Cultural considerations included the dichotomous vs.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S131

gradient conceptualization of ideas, the importance of engaging the entire family, the use of
nature analogies, the role of spirituality, the value of pacific natural medicine, and a
collectivistic orientation. These identified key cultural considerations can help inform others
work in type 2 diabetes with Marshallese communities.

CORRESPONDING AUTHOR: Karen Yeary, PhD, University of Arkansas for Medical Sciences,
Little Rock, AR, 72205; [email protected]
S132 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A068 6:00 PM-7:00 PM

DECONSTRUCTING ANXIETY AND POSTTRAUMATIC STRESS AMONG ADULTS WITH TYPE1

DIABETES

Bret Boyer, Ph.D.1, Chelsea Renna, PsyD2

1
Widener University, Philadelphia, PA; 2CHaC, Philadelphia, PA

Posttraumatic stress symptoms (PTS) have been reported among adults with type1 diabetes
(T1D), both related to experiences with hypoglycemia and with hyperglycemia, but studies
have not simultaneously assessed these two phenomena within the same populations. This
study investigated PTS related to 1) hypoglycemia, 2) hyperglycemia, and 3) T1D diagnosis
among 241 adults with T1D (Age 18-79, Mean=36.9, SD=12.5; 79% women; 84% Caucasian).
Exploratory Factor Analyses identified a 2-factor structure on the posttraumatic diagnostic
scale (PDS) for hyperglycemia experiences (Triggered T1D-specific Anxiety/Active Avoidance
[T-sAAA], Numbing/General Reactivity), and a 3-factor structure for hypoglycemia experiences
(T-sAAA, Numbing/General Reactivity, hypervigilance). Using only the scores for the T-sAAA
for hypoglycemia and hyperglycemia, and for general T1D, a cluster analysis identified a three-
cluster solution (High PTS scores on all 3 [HiAll], Low PTS scores on all 3 [LoAll], High Pts
regarding hypoglycemia, but low for hyperglycemia and T1D [PTShypo], and high PTS for
hyperglycemia and T1D, low for hypoglycemia [PTShyper]). MANOVA were run to test for
differences between these cluster groups on self-management and T1D outcomes, finding
significant differences for HbA1c (F(3)=9.57,p < .001), total T1D complications (F(3)=3.63, p <
.014), Severity of hypoglycemic episodes during previous month (F(3)=11.51, p < .001), and
Self Care Inventory scores (F(3)=3.28, p < .022). The LoAll cluster showed best scores on these
variables, with the HiAll group showing worst glycemic control and poorest outcomes. While
the PTShypo cluster showed significantly lower HbA1c, total T1D Complications, less severe
hypoglycemia, and better self-management behavior than the HiAll cluster, the PTShyper
cluster was not significantly lower than the HiAll cluster for total complications or higher on
SCI score. The PTShypo cluster did not differ significantly from the LoAll cluster. A Cluster
Analysis to examine sub-samples of cases with different patterns of HbA1c, #hypoglycemia,
Severity of hypoglycemia, and T1D Complications identified 3 clusters with different patterns
of these outcomes, and PTShyper versus PTShypo clusters showed a significant chi-squared
for their representation in these outcome clusters (Chi-Square=7.83, p < .02). Clinical
implications of identifying cases with these distinct patterns of diabetes-specific anxiety are
interpreted from the patterns in their diabetes self-management and glycemic control.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S133

CORRESPONDING AUTHOR: Bret Boyer, Ph.D., Widener University, Philadelphia, PA, 19119;
[email protected]
S134 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A069 6:00 PM-7:00 PM

ILLNESS BELIEFS OF SPECIALTY CLINIC PATIENTS WITH TYPE 2 DIABETES PRESCRIBED INSULIN

Jennifer F. Ward, PhD1, Barbara A. Stetson, Ph.D.2, Stephanie R. McDonough, M.A.2, Christen
McKiel, BA 3, Sriprakash Mokshagundam, MD4

1
University of Louisville, Hermitage, TN; 2University of Louisville, Louisville, KY; 3University of
Louisville, Jeffersonville, IN; 4University of Louisville, Louisville, KY, Louisville, KY

Background: Exogenous insulin use in the management of Type 2 (T2) Diabetes Mellitus (DM)
has grown in recent years. Patient beliefs related to treatments in this population will help to
inform clinical work and research related to both insulin treatment adherence and decision-
making efforts.

Aim: 1) Illness beliefs (IPQ-r) related to T2DM, such as beliefs about DM chronicity, symptoms
and severity, and insulin treatment effectiveness are examined and (2) associations between
these beliefs and time since DM diagnosis, insulin prescription duration and medical
outcomes (HbA1c%) are explored.

Method: Percent of participants endorsing agree or strongly agree beliefs for items and
subscales related to personal and treatment management of DM are reported. Linear
Regression and ANOVA used for continuous and group associations.

Result: Participants were 167 adults with T2DM (49% male) prescribed exogenous insulin at a
specialty diabetes clinic. Self-report missing data was high (30%). Participants were
moderately controlled [HbA1c%M=8.13 (1.4)]. The majority experienced recent changes in
insulin dose or frequency (59.1%) per chart review. Self-reported adherence to insulin was
high (SDSCA, 76.7% daily). Insulin duration [M=66.4 (49.1), range 9-257 months] and DM
diagnosis duration [M=190.0 (100.1), range 9 558 months] varied, though there was
significant (43.4% and 17.9%) missing chart data.

Participants attributed several symptoms, including pain (58.3%), fatigue (69.5%) and
sleeplessness (49.0%) to their T2DM. Most participants reported beliefs that T2DM is
permanent (85.4%) and serious (94.3%). Beliefs about personal (90.3%-96.0%) and treatment
management of DM (68.6%, 81.0%) were generally positive with some reporting cure
potential with T2DM treatment (31.7%). Tested relationships between DM beliefs, HbA1c%
and DM diagnosis duration were not significant. Longer insulin duration was associated with
negative beliefs about treatment, but not personal, management (=.28, p=.02).
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S135

Conclusions/Implications: High compliance with insulin treatment was reported in clinical

participants with T2DM prescribed exogenous insulin. Self-reported insulin treatment beliefs
suggest that most view treatment and personal management of T2DM as achievable and
likely to improve outcomes. Results indicate that longer insulin treatment duration was
related to more negative beliefs regarding insulin treatment. Missing responses may reflect
uncertainty about these beliefs. Reported beliefs were not associated with HbA1c%. Longer
insulin treatment in this population may lead to less positive beliefs in treatment effectiveness
and potentially reduced compliance. Understanding beliefs about DM and insulin treatment in
this growing population will help inform treatment adherence and decision-making efforts
both clinically and in future research.

CORRESPONDING AUTHOR: Jennifer F. Ward, PhD, University of Louisville, Hermitage, TN,

37076; [email protected]
S136 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A070 6:00 PM-7:00 PM

MHEALTH-BASED MULTICOMPONENT LIFESTYLE INTERVENTION IN OBESE PATIENTS WITH

DIABETES AND CHRONIC KIDNEY DISEASE

Lu Hu, PhD1, Kathleen Woolf, PhD, RD2, David St-Jules, RD, PhD1, Ram Jagannathan, PhD3,
David S. Goldfarb, MD4, Stuart Katz, MD4, Aditya Mattoo, MD5, Alex Sierra, MS6, Mary Lou
Pompeii, RDN, CSR, CDN, CDE3, Nicole Skursky, BA7, Lisa Ganguzza, MS, RD, CDN8, Mary Ann
Sevick, ScD, RN7

1
Center for Healthful Behavior Change, NYU School of Medicine, New York, NY; 2NYU
Steinhardt, New York, NY; 3Center for Healthful Behavior Change Department of Population
Health New York University School of Medicine, New York, NY; 4NYU Langone Medical Center,
New York, NY; 5NYU Langone Medical Center, NEW YORK, NY; 6Center for Healthful Behavior
Change Department of Population Health New York University School of Medicine, NEW
YORK, NY; 7Center for Healthful Behavior Change Department of Population Health New York
University School of Medicine, new york, NY; 8NYU School of Medicine, New York, NY

Purpose

In-person lifestyle interventions are effective for chronic disease management; however,
several fundamental barriers (e.g., transportation, schedule conflicts) limit real-world
implementation and scalability. In this pilot study, we evaluated the use of mobile technology
to deliver a multi-component lifestyle intervention for engaging patients with type 2 diabetes
(T2D) and concurrent stage 2-4 chronic kidney disease (CKD) in multiple behavior changes
(restriction of dietary energy, sodium and phosphate additives, and increased physical
activity). In this report, we describe feasibility, acceptability, and participant satisfaction.

Methods

Participants were randomized to either a usual care (UC) or a multicomponent, behavioral

intervention group. The intervention components included: (1) Social Cognitive Theory (SCT)-
based group counseling, and (2) technology-based self-monitoring of diet, weight and physical
activity. All intervention sessions were delivered remotely using videoconferencing weekly for
4 weeks and then biweekly for the next 20 weeks. Participants satisfaction was assessed by
an investigator-developed, 5-point, Likert-scaled survey (1=strongly agree, 5=strongly
disagree) at 24 weeks.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S137

Results

A total of 14 participants were randomized to the UC (n=7) or SCT-based intervention (n=7).

The study sample was predominantly white (71.4%), male (71.4%), and currently or living as
married (57.1%) with a mean age of 65.9 (9.7SD) years, and BMI of 34.4 (7.0SD) kg/m2. The
median time devoted to technology training for this aging sample was 45 minutes (IQR=30-45
minutes). Of the 1,134 person-days of possible self-monitoring, the median proportion of days
in which at least 3 meals was logged was 95.8% (IQR=72.0-99.4%), and the median number of
meals entered was 3.7 meals/day (IQR=2.8-3.8). Attendance for the 14 online group sessions
averaged 94% (.07SD). Participants reported: ease in using the touchscreen to operate
(MeanSD: 1.2.4) and access group sessions on the iPad (2.01.0); a desire to use the iPad to
manage their health once the study was over (1.4.5); technology-based self-monitoring
helped them understand how to change their diet (1.6.9), keep track of nutrition (1.81.3),
stay motivated (1.6.5), and limit sodium (1.6.9) and calorie intake (1.4.5).

Conclusions

This pilot study demonstrated that a multicomponent intervention involving remotely

delivered group counseling and technology-based self-monitoring is feasible and acceptable
to an aging population with T2D and concurrent CKD. Additional research is needed to
evaluate the incremental contribution of the individual intervention components and their
interaction effects for engaging this patient population in multiple lifestyle behavior changes.

CORRESPONDING AUTHOR: Lu Hu, PhD, Center for Healthful Behavior Change, NYU School of
Medicine, New York, NY, 10016; [email protected]
S138 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A071 6:00 PM-7:00 PM

MINDFULNESS AMONG ADULTS WITH DIABETES: RELATIONSHIPS WITH SELF-CARE AND

GLYCEMIC CONTROL

Andrea E. Lagotte, MA1, Chandra Osborn, PhD, MPH2, Lindsay S. Mayberry, MS, PhD1

1
Vanderbilt University Medical Center, Nashville, TN; 2One Drop, Nashville, TN

Several recent diabetes interventions target mindfulness to improve adherence to self-care

and outcomes. However, few have examined mindfulness as a determinant of diabetes self-
care behaviors or glycemic control (A1c).

We recruited English-speaking adults with type 2 diabetes (T2D) from Vanderbilt University
Medical Center to participate in a randomized control trial of the Diabetes MAP web-based
medication adherence intervention. We collected self-reported demographics (age, gender,
race, ethnicity, education, income), responses to the Mindfulness Attention Awareness Scale
(MAAS), Adherence to Refills and Medications Scale for Diabetes (ARMS-D; dichotomized as
perfect vs. suboptimal adherence) and Summary of Diabetes Self-Care Activities (SDSCA)
general diet, specific diet, exercise and self-monitoring of blood glucose (SMBG) at baseline.
Participants also completed a baseline lab A1c test. Spearmans rho and Mann-Whitney U
tests examined bivariate associations between demographics and mindfulness. Linear or
logistic regression models examined relationships between mindfulness and adherence to
each self-care behavior and A1c at baseline. All models were adjusted for demographics.

Participants (N=151) were 55.310.9 years old, 60% female, and 24% non-White, with
15.21.8 years of education, and 25% annual incomes < $40K. A1c was 8.01.5%. MAAS scores
were 4.90.8 (possible range 1-6), with higher scores indicating greater
mindfulness. Demographics were not associated with mindfulness. Greater mindfulness was
marginally associated with a higher odds of perfect adherence (AOR=1.77, p=.11) and
significantly associated with better adherence to general diet (=.30, p < .01) and specific diet
(=.31, p < .001).

Among adults with T2D, greater mindfulness was independently associated with better
adherence to general diet and specific diet, and marginally associated with greater odds of
having perfect medication adherence. Greater mindfulness was not associated with
adherence to exercise or SMBG. Mindfulness skills may be an effective target for interventions
to improve adherence to diet and medications, but may be less effective for improving
exercise or SMBG. Although mindfulness was not associated with A1c, increasing mindfulness
may affect A1c through improvements in adherence to diet and medications.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S139

CORRESPONDING AUTHOR: Andrea E. Lagotte, MA, Vanderbilt University Medical Center,

Nashville, TN, 37203; [email protected]
S140 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A072 6:00 PM-7:00 PM

OUTCOMES OF A DIGITAL HEALTH PROGRAM WITH HUMAN COACHING FOR DIABETES RISK
REDUCTION IN A MEDICARE POPULATION

Cynthia M. Castro Sweet, PhD1, Vinay Chiguluri, MPH2, Rajiv Gumpina, M.S.3, Paul Abbott,
MBA4, Erica Madero, BS, MS5, Mike Payne, MBA, MSci6, Laura Happe, PharmD, MPH3, Andrew
M. Renda, MD MPH4, Todd G. Prewitt, MD, FAAFP3

1
Omada Health, Inc., San Francisco, CA; 2Humana, Inc, Harbor City, CA; 3Humana Inc.,
Louisville, KY; 4Humana, Louisville, KY; 5University of California Berkeley, San Francisco, CA;
6
Virta Health, Burlingame, CA

Older adults have increasing risk for obesity and related chronic diseases, particularly Type 2
diabetes and cardiovascular disease (CVD). Current clinical guidelines recommend offering
behavioral counseling programs for lifestyle management for disease risk reduction. Digital
programs that combine health technology with live human coaching have recently emerged as
scalable alternatives to the more traditional in-person and face-to-face formats of lifestyle
management programs. Little research has examined the outcomes of these newer,
technology-driven programs specifically with older adults. The goal of this observational study
was to examine the outcomes of a Medicare population who participated in a digital health
program with human coaching for diabetes risk reduction. Health plan members ages 65 and
older who were at risk for diabetes enrolled in the program. They received weekly online
curriculum, remote access to a peer support group of other participants, regular
communication with a trained lifestyle coach, a wireless scale, wearable activity monitor, and
food and activity tracking tools. Program participation and health outcomes were examined at
4, 6 and 12 months. A total of 501 participants enrolled in the program from 37 states. 92%
completed at least 9 of 16 core lessons. Participants averaged 20 of 31 possible opportunities
for weekly engagement with the program. At 12 months, participants lost 7.4% (SD=7.8%) of
initial body weight, and among participants with clinical data, glucose control improved
(HbA1c change = -0.14%, p=0.001) and total cholesterol decreased (-7.08, p =.008). Self-
reported well-being, depression and self-care also improved (p < 0.0001). This Medicare
population demonstrated sustained program engagement and improvements in weight,
disease risk factors and well-being. The findings support use of digital programs leveraging
human coaching for reducing chronic disease risk among older adults.

CORRESPONDING AUTHOR: Cynthia M. Castro Sweet, PhD, Omada Health, Inc., San Francisco,
CA, 94111; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S141

A073 6:00 PM-7:00 PM

THE IMPORTANCE OF HAVING FUN: SHARED PLAY IN COUPLES COPING WITH TYPE 1 DIABETES

Meredith Van Vleet, Ph.D.1, Vicki Helgeson, PhD1, Mary Korytkowski, M.D.2, Cynthia Berg,
Ph.D.3

1
Carnegie Mellon University, Pittsburgh, PA; 2University of Pittsburgh, Pittsbugh, PA;
3
University of Utah, Salt Lake City, UT

Chronic illness research has traditionally focused on stressful contexts associated with the
illness and their impact on health and well-being, while overlooking the role of positive events
on health and well-being. We examined shared play as a positive relationship context that is
likely to affect well-being and self-care among couples coping with type 1 diabetes: Shared
play is defined as activities that are pursued for fun/amusement, that are carried out with an
enthusiastic and in-the-moment approach, and that are highly interactive. Shared play is likely
to be an enjoyable experience that provides a momentary break from strains and
responsibilities, and serves a bonding function between partners. When partners are in better
moods, less stressed, and feel closer to one another, they are likely to engage in better
problem-solving together in diabetes contexts, discuss diabetes more, and have greater
communal coping (i.e., appraise diabetes problems as shared and collaborate in resolving
problems), and engage in better self-care. We examined links between average daily shared
play to positive affect, general stress, diabetes stress, diabetes disclosure, communal coping,
and self-care behavior in a study of 143 patients with type 1 diabetes and their spouses (95%
white, 61% female, average age= 46, range= 25-75). Couples completed online questionnaires
for 14 days assessing shared play and proposed outcomes. Aggregates of shared play and
outcomes across the 14 days were used in analyses. Results indicated that greater shared play
across the 14 days was linked with greater positive affect, less general stress, more diabetes
disclosure, more communal coping, and more exercise (all ps < .05), but was unrelated to
diabetes stress or self-care behavior. These results suggest that shared play is associated with
a more intimate, positive relationship atmosphere, which is more conducive to diabetes
communication and a shared approach to diabetes self-care.

CORRESPONDING AUTHOR: Meredith Van Vleet, Ph.D., Carnegie Mellon University,

Pittsburgh, PA, 15213; [email protected]
S142 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A074 6:00 PM-7:00 PM

DO AS I SAY NOT AS I DO: THE RELATIONSHIP BETWEEN PARENTS DIETS AND THEIR
ATTITUDES TOWARD THEIR ADOLESCENTS DIETS

Sasha A. Fleary, PhD

Tufts University, Medford, MA

Background: The parent-child diet relationship has been widely investigated with parent
modeling and availability of foods in the home being implicated. Further, studies have found
that parenting practices around diet determine the quantity and types of foods children
consume. Few studies have focused on adolescents and little is known about how attitudes
toward their adolescents diet are related to parents diet. Thus, this study aimed to
determine whether parents practice what they preach by exploring the relationship between
parents diet and their attitudes towards their adolescents fruits and vegetables (FV), and
sugary foods and drinks intake. Given the role of diet in obesity and the likely internal conflict
for obese parents trying to regulate their adolescents diet, the second aim of this study
explored if parents obesity status affected the relationship between their diet and attitudes
toward their adolescents diet.

Methods: Parent data were obtained from the National Cancer Institutes Family Life, Activity,
Sun, Health, and Eating Study. Parents (n=1859, 72% female, 69% White, 62%
overweight/obese) completed measures of their FV, and sugary foods and drinks intake,
supportive and controlling attitudes for adolescents diet, and availability of foods at home.
Regressions were computed to address the aims of the study.

Results: For fruits, availability of (B=.32, p < .001) and supportive attitudes (B=.17, p < .001)
for FV were related to intake. Availability of FV (B=.31, p < .001) and supportive (B=.21, p <
.001) and controlling attitudes for FV (B=.15, p < .001) were also related to vegetable intake.
FV regressions did not differ by obesity status. For sugary foods, availability of (B=.33, p <
.001) and supportive attitudes (B=-.07, p=.001) for a non-sugary foods diet were related to
sugary foods consumption. Additionally, regressions coefficients for controlling attitudes
differed by obesity status (F=4.52, p=.03). Availability of sugary drinks (B=.34, p < .001) was
related to sugary drinks intake and regression coefficients for controlling attitudes differed by
obesity status (F=4.40, p=.03).
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S143

Discussion: This study confirms that parents adopt a do as I say not as I do approach
regarding healthy eating and this is especially true for sugary foods and drinks intake.
Implications for future research include exploring the cognitive processes behind parents'
misaligned attitudes and behaviors and investigating how these inconsistencies are related to
adolescents behaviors to better develop parent-adolescent diet interventions.

CORRESPONDING AUTHOR: Sasha A. Fleary, PhD, Tufts University, Medford, MA, 02155;
[email protected]
S144 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A075 6:00 PM-7:00 PM

FACTORS ASSOCIATED WITH LOW-INCOME PRESCHOOLERS HAVING A TV ON AND CHILD

WATCHING TV DURING DINNER

Teresia O'Connor, MD, MPH1, Sheryl Hughes, PhD1, Ann Chen, PhD1, Thomas Power, PhD2

1
Baylor College of Medicine, Houston, TX; 2Washington State University, Pullman, TX

Background: Turning the TV off during meals is a common obesity prevention strategy
promoted by interventions. However few directly observed data are available of the
prevalence of having the TV on during meals or the amount of time a child spends watching
TV during meals to inform who would most benefit from these messages. The aim of this
study was to assess factors associated with a) having a TV on during dinner and b) duration of
child watching during meals, among low-income ethnic minority families with preschool aged
children.

Methods: Analysis of a cross-sectional home dinner observation study of Hispanic and black
families whose 3-5 year old child attended Head Start. Three home dinners were video
recorded among 144 families. Parents self-reported demographics and Caregivers Feeding
Style Questionnaire. Videos were coded for whether TV was on and number of people eating
together. Of those with TV on, two coders assessed and reached consensus if they could
determine whether and when child was watching TV (depending on position of TV and video-
recording equipment). Those videos were coded for the duration of the meal the child
watched TV. 20% of videos were double coded to assess intercoder reliability. Multilevel
models with video observation nested within family assessed factors associated with having
TV on and duration (as % of meal time) child watched TV.

Results: The sample consisted of 59.7% Hispanic and 40.3% black families. The average
number of people eating dinner, including the child, was 2.1 (SD 1.3) and 43.8% of parents
were married. Of the 428 videos, 39.3% had TV off, 59.8% had TV on (44.2% were codable,
15.7% TV-on, but whether child was watching was uncertain/not codable), and in 0.9% of
videos we could not tell if TV was on (removed from analysis). The inter-coder ICCs were 0.85-
0.99 for the duration codes. For codable videos with TV on, children spent 17.1% (SD 19.1) of
the meal watching TV. Only ethnicity was significantly associated with having the TV on
during dinner in the full model, with odds ratio of 0.18 (95% CI 0.08, 0.41) of Hispanic having
TV on as compared to black families. Having unmarried parent (p=0.05) and fewer number of
people eating together at dinner (p
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S145

Conclusion: A majority (60%) of the observed dinner meals had a TV on, but having the TV on
during dinner was more common among black than Hispanic families. Targeting black families
and families with unmarried parents may be particularly important. Future studies will need
to see if increasing the number of people eating with the child at dinner decreases their
tendency to watching TV during the meal.

CORRESPONDING AUTHOR: Teresia O'Connor, MD, MPH, Baylor College of Medicine,

Houston, TX, 77030; [email protected]
S146 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A076 6:00 PM-7:00 PM

USING THE INTEGRATIVE MODEL OF BEHAVIORAL PREDICTION TO PREDICT SUGARY DRINK

CONSUMPTION OF OVERWEIGHT AND OBESE ADULTS

Maria Collado-Rivera, MS, CHES, Paul Branscum, PhD, RD

The University of Oklahoma, Norman, OK

Researchers and practitioners draw upon a number of theoretical paradigms within social and
behavioral health. In an attempt to integrate these paradigms, the Integrative Model of
Behavioral Prediction (IMB) was developed. The development of the IMB began at a
workshop sponsored by the National Institute of Mental Health, and was attended by a group
of behavioral theorists who worked together to identify similarities and differences among
prominent theories of behavior change and prediction. The IMB posits that individual
behavior is predicated upon having sufficient intentions to behave in certain ways, barring any
deficiency in skills and abilities to perform the behavior, and environmental constraints.
Behavioral intentions are further influenced by three factors: attitudes towards a behavior,
perceived norms (PN), and perceived behavioral control (PBC). Since its creation, few
researchers have fully operationalized its constructs for investigation. Therefore, the purpose
of this study was to explore the utility of the IMB with obese and overweight adults
attempting to lose weight, for predicting the behavior To stop drinking regular soda and
other sugary drinks for the next 6 months. This study was conducted over the course of 2
phases. During the first phase, an elicitation of beliefs from the target population was
accomplished, to elicit salient beliefs for survey development. Next, weight loss patients
(n=410) completed a valid and reliable survey, evaluating the essential constructs of the IMB.
For this phase, a hypothesized path model was tested (AMOS 17.0) for which attitudes, PN
and PBC predicted behavioral intentions, and intentions, PBC, skills, and environmental
constructs predicted sugary drink consumption. Overall, fit indices of the model were good
(CFI=.970, TLI=.921, RMSEA=.062), and a significant amount of the variance of intentions
(R2=.407) and sugary drink consumption (R2=.162) were predicted. Results suggest that the
IM is an advantageous and practical theory for explaining health behaviors related to obesity
prevention and treatment. Further work is needed to better operationalize constructs of the
IM that have not been fully explored.

CORRESPONDING AUTHOR: Maria Collado-Rivera, MS, CHES, The University of Oklahoma,

Norman, OK, 73071; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S147

A077 6:00 PM-7:00 PM

VEGANS: JUST LIKE US, ONLY HEALTHIER

Sydney Heiss, B.A.1, Jaime Coffino, M.P.H.2, Julia M. Hormes, Ph.D.2

1
University at Albany, SUNY, Delmar, NY; 2University at Albany, SUNY, Albany, NY

Introduction. It has been posited that vegans (i.e. individuals who refrain from consuming all
animal products) engage in chronic unhealthy behaviors, especially in regard to diet and
weight. Unfortunately, the large majority of studies interested in the behaviors of meat-
refrainers are unable to obtain a large enough sample of vegans to make meaningful
comparisons. This study examines indicators of health in one of the largest samples of vegans
recruited from the general population, in comparison to omnivores (i.e. individuals who do
not refrain from any animal products). Methods. Respondents were categorized based on
self-reported vegetarian status. For the purposes of the present study vegans (N = 287; female
= 84.0%; mean age = 31.7 12.5) were compared to omnivores (N = 165; female = 76.4%
mean age = 30.9 13.7). Relevant measures include reports on demographics, height, actual
and ideal weight, smoking behaviors, the Alcohol Use Disorders Identification Test, Food
Frequency Questionnaire, Orthorexia Scale, Binge Eating Scale, Drive for Thinness Scale, Drive
for Muscularity Scale, and Yale Food Addiction Scale. Groups were compared in a series of
chi-square, independent samples t-tests, and univariate and multivariate analyses of variance
(for measures containing multiple subscales). Gender was included as a covariate when
significant. Results. Vegans were more likely than omnivores to be female (p = .05). Vegans
more frequently ate fruits, vegetables, nuts, beans, and grains, but less frequently consumed
caffeinated soft drinks (all p < .001). Vegans were more likely to prepare food at home and
more likely to consider themselves healthy (both p = < .001), but scored lower on the
Orthorexia Scale (p < .001), were less likely to exhibit moderate binge eating (p = .01), and
endorsed a lower drive for thinness (p = .01). There were no differences between groups in
age, BMI, proportion of respondents categorized as underweight, normal weight, overweight,
or obese, reported discrepancy between current and ideal weight, frequency and length of
exercise, alcohol use, smoking behaviors, drive for muscularity, or scores on the Yale Food
Addiction Scale. Discussion. This study begins to paint a more accurate picture of health
behaviors in vegans. It seems as though vegans exhibit markers of healthier behaviors (e.g.,
increased fruit and vegetable consumption, increased home-cooking, self-reported
healthfulness) but simultaneously do not exhibit the negative effects often thought to result
from continuous self-monitoring of the diet (e.g., increased drive for thinness, low BMI,
increased orthorexia score). Ultimately, vegans do not appear to differ from omnivores to a
large degree, and when they do, it seems to be in meaningful healthful ways.
S148 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Sydney Heiss, B.A., University at Albany, SUNY, Delmar, NY,
12054; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S149

A078 6:00 PM-7:00 PM

YEEGO GARDENING!: A COMMUNITY GARDEN INTERVENTION TO PROMOTE HEALTH ON THE

NAVAJO NATION

India Ornelas, PhD1, Desiree Deschenie, Bachelors2, Sonia Bishop, Kevin Lombard, Shirley A.A.
Beresford, PhD1

1
University of Washington, Seattle, WA; 2New Mexico State University, Farmington, NM

Background: American Indians and Alaskan Natives (AI/AN) are at increased risk for obesity
and diabetes in part due to their diet, including low consumption of fruits and vegetables.
Several factors contribute to the low consumption of fruits and vegetables among Navajo,
including lack of access to, barriers to growing, and high cost of fruits and vegetables. The goal
of our study was to develop a theory-based and culturally relevant community garden
intervention to increase gardening practice and fruit and vegetable consumption in two
Navajo communities. In this presentation, we describe our process of designing the
intervention by both drawing on behavioral theory and solicited input from Navajo
community members. Methods: Using Social Cognitive theory, focus group findings and input
from our community partners, we developed an intervention consisting of three integrated
components: 1) a community/demonstration garden plot, 2) technical assistance workshops
to increase gardening and healthy eating and 3) community outreach. These three
intervention components aimed to increase gardening and healthy eating by increased access
to fresh produce grown in the community gardens, as well as by increasing participants self-
efficacy, self-monitoring and behavioral capability to garden, prepare and consume fruits and
vegetables. Gardens were constructed and maintained in collaboration with community-
based organizations in 2 Navajo communities. Monthly workshops were held throughout the
growing season, which incorporated aspects of Navajo culture and opportunities to build
confidence and skills in gardening and healthy eating behaviors. In addition, program staff
attended community events to promote gardening and healthy eating. Conclusions:
Community input was essential throughout the planning and implementation of the
intervention. If effective, community garden interventions may improve the food
environment in areas with limited access to healthy foods. Increasing access and knowledge
about healthy eating behaviors may ultimately reduce risk for obesity and diabetes in Navajo
communities.

CORRESPONDING AUTHOR: India Ornelas, PhD, University of Washington, Seattle, WA, 98195;
[email protected]
S150 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A079 6:00 PM-7:00 PM

EXPLORING THE RELATIONSHIPS BETWEEN SOCIAL MEDIA TECHNOLOGY, COMPETITION, AND

FITNESS MOTIVATION AND ENJOYMENT

Lauren Galantai, none, Joanne Hash, PhD

Whittier College, Whittier, CA

Fewer than half of American adults meet the recommended guidelines for aerobic physical
activity (CDC, 2016), which may contribute to the high rate of obesity in the US (NIH, 2012).
This trend may be due to a lack of motivation, and/or lack of social support, to exercise.
Mobile fitness applications, like Fitbit and Strava, have been introduced to help increase
physical activity through the use of technology (Bauer & Kriglstein, 2015). The current study
explores how fitness technology and friendly competition may interact with self-efficacy,
motivation, effort, and enjoyment in fitness. Forty healthy adults (30 F, age 18-61, M=22.6,
SD=7.4) were assigned to 1 of 2 experimental groups, Fitbit (n=12) or Strava (n=13), or to the
social context-free pedometer control group (n=15). The Fitbit group tracked their steps,
while the Strava group timed their run through a specified 0.2-mile course (the pedometer
participants either tracked steps or ran the course without social context). Each experimental
group competed in counterbalanced 1 week of app-based competition and 1 week of non-
competition tracking. The social settings of competition differed in each experimental group;
Fitbit participants competed against friends, while Strava participants competed against
strangers. During the non-competition and competition phases, each of the groups showed
significant positive correlations between competitiveness, enjoyment, effort, self-efficacy, and
motivational styles (p < 0.05). Interestingly, the Fitbit group had a significant relationship
between enjoyment and intrinsic motivation during the non-competition phase, r(12)=0.630,
p < 0.05, but lost it during competition, where a significant relationship was found between
enjoyment and extrinsic motivation, r(12)=0.578, p < 0.05. Similarly, during the competition
phase, the relationship between extrinsic motivation and competitiveness also became
significant for the Fitbit group, r(12)=0.631. For the Strava group, during competition there
was a significant negative relationship between effort and intrinsic motivation, r(13)=-0.580, p
< 0.05. It appears that the type of tech-based social context that an individual engages in has
an influence on motivational styles and enjoyment in fitness activities. Competing against
friends appeared to increase extrinsic motivation and enjoyment, while competing against
strangers resulted in a negative relationship between effort and intrinsic motivation. Having a
more nuanced understanding of the complex relationships between motivational style,
mobile fitness applications, and social context may not only have implications for altering
personal fitness behaviors, but also may impact adherence to fitness-based treatment
regimens.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S151

CORRESPONDING AUTHOR: Lauren Galantai, none, Whittier College, Whittier, CA, 90601;
[email protected]
S152 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A080 6:00 PM-7:00 PM

HEALTH PRIVACY: PERCEPTIONS AND PREFERENCES AMONG ADULTS AND TEENS

Matthew J. Bietz, Ph.D.1, Cynthia Cheung, MPH/MA (c)2, Cinnamon S. Bloss, Ph.D.2

1
University of California, Irvine, Irvine, CA; 2University of California - San Diego, La Jolla, CA

This paper reports on a qualitative study of individual preferences and preferences around the
privacy of personal health information. We conducted 7 focus groups with 51 people to
understand the expectations that individuals have for the privacy of their health information
and how privacy preferences for health information may be different from preferences for
other kinds of personal information. Focus groups were conducted with a variety of cohorts,
including: middle school students (2 groups); members of a breast cancer research cohort (2
groups); and members of community-serving organizations (3 groups). Focus groups included
between 3 and 12 participants (median = 6). Three of the focus groups (both middle-school
and one community group) were conducted with teenagers (age 14-19 years, n=24). Focus
groups were conducted using a semi-structured protocol that covered issues relating to
general privacy definitions and preferences, electronic health records, genomic data, tissue
donation, and personal health and lifestyle data (e.g. wearable health devices). Focus groups
were transcribed and analyzed using a grounded theory approach to understand emergent
themes in the discussions. After the focus groups, participants filled out surveys that included
demographic information as well as information about their interactions with health providers
and attitudes toward privacy.

The results of this study suggest that both adults and teenagers are concerned with privacy,
but that their specific concerns differ. For example, teenagers were more likely to see privacy
through a commodity lens in which personal information had inherent value. Teens were
more willing to exchange their personal data for the use of a service, and more often cited a
belief that they should receive some form of compensation (monetary or otherwise) from
those who wanted to mine their data. Teenager concern for the privacy of health data was
also lower than for adults. Post-focus-group surveys showed that while adults were more
likely to distrust corporations and legal protections for privacy, teenagers were more likely to
be privacy pragmatists who weighed the benefits of services against the potential privacy
intrusions. Participants of all ages wanted to own their personal data but were unsure of who
actually did own it. This paper discusses these findings and their potential impact on
acceptance and adoption of new health technologies.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S153

CORRESPONDING AUTHOR: Matthew J. Bietz, Ph.D., University of California, Irvine, Irvine, CA,
92697-3440; [email protected]
S154 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A081 6:00 PM-7:00 PM

HOW TO STAY HEALTHY WHILE STUDYING ABROAD: AN ELECTRONIC MAGAZINE FOR COLLEGE
STUDENTS

Hannah Lightcap, Bachelor of Science in Dietetics1, Marie Kuczmarski, PhD2, Lisa Chieffo, EdD2

1
University of Delaware, Elkton, MD; 2University of Delaware, newark, DE

Title: How to Stay Healthy While Studying Abroad: An Electronic Magazine for College
Students

Objectives: List the top three areas which students utilized from the electronic [e] magazine.
List the three lifestyle behaviors students attributed to their physical wellness whilst on their
travel study.

Introduction: Increasing numbers of U.S. students participate in international and domestic

travel study programs. While most institutions include material on health-related topics such
as immunization, psychological distress, personal assault, and water quality, comprehensive
coverage of strategies to ensure optimal physical wellness during travel study has been
overlooked. To fill that void How to Stay Healthy While Studying Abroad, a comprehensive
wellness guide, was designed as an e magazine for undergraduate students.

Research Objectives: The primary objective was to evaluate the usefulness of the e magazine.
The secondary objective was to enhance the content of the guide based on an assessment of
students' attitudes, perceptions, and behaviors related to wellness pre- and post-participation
in travel study programs.

Methods: Of the 1,137 students enrolled in travel study programs in the winter 2016 term,
344 students completed the pre-travel questionnaire and 172 the post-travel questionnaire.
Students were recruited via email invite. Study protocol was approved by the University of
Delawares Institutional Review Board. Questionnaires were pilot tested and focused on
evaluating aspects of the e magazine, as well as attitudes, perceptions, behaviors, and barriers
to optimizing physical wellness.

Results: Hydration (66%), health during travel (57%), and getting enough sleep (54%) were
topics most utilized. Self-motivation, social support, and tips within How to Stay Healthy While
Studying Abroad were the top three factors contributing to students' achievement of physical
wellness on their program. The majority of respondents (87%), indicated they would
recommend How to Stay Healthy While Studying Abroad to others going on a travel study
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S155

program.

Conclusions: These findings support the need for inclusion of physical wellness guides in pre-
departure materials to help students maintain their health while navigating new
environments during travel study programs.

CORRESPONDING AUTHOR: Hannah Lightcap, Bachelor of Science in Dietetics, University of

Delaware, Elkton, MD, 21921; [email protected]
S156 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A082 6:00 PM-7:00 PM

LESSONS LEARNED FROM A PROCESS EVALUATION OF AN ONLINE ORGAN DONATION

EDUCATIONAL INTERVENTION

Nakeva Redmond, MPH1, Laura Harker, MPH2, Yvan Bamps, PhD3, Shauna Flemming, MPH,
MDiv4, Kimberly J. Arriola, PhD/MPH5, Jennie P. Perryman, RN, PhD6, Nancy J. Thompson,
Ph.D., M.P.H.7, Rachel E. Patzer, PhD, MPH8

1
Emory University, Atlanta, GA; 2Georgia Budget and Policy Institute, Atlanta, GA; 3School of
Medicine, Emory University, Decatur, GA; 4Emory University, Rollins School of Public Health,
Atlanta, GA; 5RSPH of Emory University, Atlanta, GA; 6Emory Transplant Center/Emory
Healthcare/Emory University, Atlanta, GA; 7Department of Behavioral Sciences and Health
education/Rollins School of Public Health/Emory University, Atlanta, GA; 8Emory University
School of Medicine, Atlanta, GA

Lessons Learned from a Process Evaluation of an Online Organ Donation Educational

Intervention
The number one problem in transplantation today is the lack of available organs. Internet use
is at an all-time high1, creating an opportunity to increase public commitment to organ
donation through the broad reach of web-based, behavioral interventions. Implementing
Internet interventions, however, presents some challenges, including preventing fraudulent
respondents and ensuring intervention uptake. While web-based organ donation
interventions have increased in recent years, process evaluation models fit for web-based
interventions are lacking. This study outlines a refined process evaluation model adapted for
online settings and used to assess the implementation of a web-based intervention aimed to
increase organ donation among African Americans (AA).

We used a randomized, pretest-posttest control design to assess the effectiveness of the

intervention website, which addressed barriers to organ donation through corresponding
videos. Eligible participants were AA adult residents of Georgia (GA) who were not registered
on the state donor registry. Drawing from previously developed process constructs 2, we
adapted reach (proportion of eligible participants), recruitment (online mechanism used to
recruit participants), dose received (intervention uptake) and context (factors in the cyber
world affecting implementation) for use in Internet-settings.

Reach: A total of 1,415 individuals completed the eligibility screener; 948 (67%) were
determined eligible, of which 97% (n=918) completed the study. Recruitment: A total of 67%
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S157

of participants were recruited from an online recruiter. Context: Approximately 75% of

responders resided outside of GA (n=577) and 22% of resided outside of the U.S. (n=172); 27%
of all participants were repeat visitors (n=206). After eliminating these data, only 17
participant surveys remained. Dose received: No videos were watched in their entirety and
the average website session duration was just 17 seconds over the minimum. Our adapted
process evaluation model confirmed fraud and minimal uptake to be serious threats to
Internet studies, and underscore the need for using a process evaluation model fit for web
settings to identify such threats.

References

1. Pew Research Center. Americans Internet Access: 2000-2015. In: Andrew Perrin,
Maeve Duggan, eds2015.

2. Steckler A, Linnan L. Process evaluation for public health interventions and research.
San Francisco, CA: John Wiley & Sons, Inc.; 2002.

CORRESPONDING AUTHOR: Nakeva Redmond, MPH, Emory University, Atlanta, GA, 30322;
[email protected]
S158 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A083 6:00 PM-7:00 PM

OLDER ADULTS' INTERNET USAGE PATTERNS AND RELATIONSHIP WITH SOCIAL ENGAGEMENT

Ching-Ju Chiu, Ph.D.1, Cheng-Sian Chang, Ph.D.2, Dai-Chan Lin, M.S.2

1
Institute of Gerontology, College of Medicine, National Cheng Kung University, Tainan,
Tainan, Taiwan (Republic of China); 2National Cheng Kung University, Tainan, Tainan, Taiwan
(Republic of China)

Objectives: Literature has offered little guidance on whether diverse types of Internet use
among older adults exist and whether they have significant reciprocal associations with older
adults social engagement in their daily lives.

Methods: This study randomly selected older adults aged 50 and older representative of silver
surfer in Taiwan from two urban and two rural areas and interviewed the participants through
telephone interviews (n=248). We employed cluster analysis to identify their Internet use
patterns, the multinomial logistic regression analyses to ascertain characteristics associated
with the Internet usage patterns, and the multiple regression to examine if the Internet usage
pattern associated with social engagement in real life.

Results: Four user types were identified: Leisure (32%), Sporadic (26%), Instrumental (21%),
and Eager (21%). Education level, Internet use history, perspectives, and social engagement in
daily life were the authentic predictors of Internet user types. Eager and Leisure Users had
significantly higher scores on social engagement in their daily lives compared with Sporadic
Users after controlling for significant sociodemographic and behavioral covariates.

Conclusions: Findings from this study underline the importance of the older population
segment regarding their different characteristics and needs. It also suggests the possibility of
Internet use for increasing social engagement in old age.

CORRESPONDING AUTHOR: Ching-Ju Chiu, Ph.D., Institute of Gerontology, College of

Medicine, National Cheng Kung University, Tainan, Tainan, 70101; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S159

A084 6:00 PM-7:00 PM

QUALITATIVE STUDY OF TEXT MESSAGING FOR SMOKING CESSATION

Claire A. Spears, PhD1, Natalie K. Anderson, BA2, Sadaf Lotfalian, MA2, Lorien Abroms, ScD,
MA3, David W. Wetter, Ph.D.4

1
Georgia State University, Atlanta, GA; 2Catholic University, Washington, DC; 3George
Washington University, Washington, DC; 4University of Utah, Salt Lake City, UT

Mobile health technologies offer unique opportunities for cost-effective dissemination of

health promotion interventions. A growing body of research supports text messaging
programs for smoking cessation. However, qualitative research could be critical for informing
the development of more engaging, acceptable, and effective text-messaging interventions.
This is particularly important among underserved populations (e.g., low-income and
racial/ethnic minority smokers) at risk for tobacco-related health disparities. This qualitative
study assessed views of text messaging for promoting smoking cessation and suggestions for
improving these programs among low-income, primarily African American smokers. Four
focus groups were conducted with a total of 15 adult smokers interested in quitting. Most
participants were women (80%), African American (86.7%), and reported annual income <
$18,000 (60%). The facilitator asked participants about their level of interest in text messaging
to help them quit smoking; preferences for structure and timing of messages; and suggestions
for content of messages (including reactions to sample text messages), as well as other
suggestions for making the program more helpful and user-friendly. Groups were audio-
recorded, transcribed, and coded using QSR NVivo 10. Overall, participants were open to text
messaging for smoking cessation, and most were enthusiastic. Some were skeptical that texts
would help them quit in the absence of other support, but believed that text messages in
addition to in-person treatment and nicotine patches would be helpful. Primary themes
regarding suggestions for improving text-based interventions included: incorporating as much
personalization as possible (e.g., sending reminders of personal reasons for quitting); making
text messages as short as possible; sending visual messages (e.g., motivating pictures);
providing specific strategies for coping with cravings; and, in the context of smoking lapses,
encouraging people to forgive themselves for the slip and get back on track. Results provide
suggestions that could be useful for improving engagement with and effectiveness of text
messaging smoking cessation programs among low-income African American smokers.

CORRESPONDING AUTHOR: Claire A. Spears, PhD, Georgia State University, Atlanta, GA,
30303; [email protected]
S160 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A085 6:00 PM-7:00 PM

READABILITY OF PRIVACY POLICIES FOR APPS TARGETED TO YOUTH

Gitanjali Das, BS, Cynthia Cheung, MPH/MA (c), Camille Nebeker, Ph.D., Cinnamon S. Bloss,
Ph.D.

University of California - San Diego, La Jolla, CA

Privacy policies are pervasive in the digital age. In practice, users, including youth, must
read and agree to the contents of these policies in order to access and use nearly all
computer
and mobile device apps. The extent to which youth, in particular, are able to comprehend
these
contracts, however, remains questionable. We assessed reading grade level, reading ease, and
word count of privacy policies for apps that are either available or targeted to youth. Privacy
policies for 64 popular apps were collected and analyzed. Readability metrics were compared
to
the average US adult reading level, as well as to the Patient Privacy Rights (PPR TF) Trust
Framework. Policies were further compared as a function of app category (free vs. paid;
entertainment vs. social networking vs. utility). We found that the average reading grade level
for the privacy policies we analyzed was higher (12.78) than the average reading grade level of
adults in the United States (8.0) and higher than the PPR TF recommended level (12.0). In
general, app policies did not differ in their readability as a function of category. Although
users
must agree to app privacy policies in order to access digital tools and products, readability
analyses suggest these agreements are not comprehensible by the average adult, let alone
youth.
Because companies often collect, use and sell users personal information, it is concerning
that
agreements describing and governing these activities are not accessible to most users. Given
that
youth are considered a vulnerable population, we propose that pediatricians and healthcare
professionals could play a role in educating youth and their guardians about the use of online
services and potential privacy risks, including the unintentional sharing of personal data.

CORRESPONDING AUTHOR: Gitanjali Das, BS, University of California - San Diego, La Jolla, CA,
92093; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S161

A086 6:00 PM-7:00 PM

UPTAKE OF AN INCENTIVE-BASED MHEALTH APPLICATION FOR CHRONIC DISEASE

PREVENTION: PROCESS EVALUATION OF THE CARROT REWARDS APP

Marc Mitchell, Post-Doctoral Fellow1, Lauren White, MSc2, Paul Oh, MD MSc FRCPC FACP1,
David A. Alter, MD; Ph.D.,; F.R.C.P.(C)3, Tricia M. Leahey, PhD4, Guy Faulkner, PhD5

1
University Health Network, Toronto, ON, Canada; 2CARROT Insights Inc, Toronto, ON,
Canada; 3Toronto Rehabilitation Institute-University health Network, Toronto, ON, Canada;
4
University of Connecticut, Storrs, CT; 5University of British Columbia, Vancouver, BC, Canada

Background: Behavioral economics has stimulated renewed interest in financial health

incentives worldwide. The Carrot Rewards app was developed as part of a novel public-private
partnership to reward Canadians with loyalty points (e.g., movies, groceries) for downloading
the app, referring friends and completing one to two health quizzes per week (micro-
learning).

Objective: The main objective is to evaluate app uptake during the exclusive 3-month launch
period in British Columbia (BC), Canada. The secondary aims are to describe the health and
demographic characteristics of users, as well as engagement levels (i.e. proportion of quizzes
completed).

Methods: The app was promoted via loyalty program email campaigns (1.64 million emails).
Number of downloads and valid registrations (users enter age, gender and valid BC postal
code to register) were collected. Additional demographics were inferred by linking postal
codes with census data at the Local Health Area (LHA) level. A health risk assessment was also
deployed. Engagement level was collected over three months and descriptive data are
presented.

Results: In three months, 67,464 individuals downloaded the app in its first week, the Carrot
Rewards app was the most downloaded health app in Canada. Among active users
(n=57,885; at least one quiz completed), the majority were female (63%) and aged 18-35yrs
(54%). More than half of users (52%) resided in LHAs where the median income is below the
provincial average ($28, 765 Cnd). Sixty-four percent of users lived in metropolitan LHAs,
compared to 56% of the general BC population. The most prevalent risk factors were not
meeting physical activity guidelines (73%) and not getting the flu shot last year (68%).
Regarding engagement, 21%, 7%, 13%, and 59% of users were classified as low (74%)
engagers, respectively.
S162 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusion: Early results suggest that loyalty points may promote mHealth app uptake and
engagement. The impact of loyalty points on behavioral outcomes (e.g., steps/day, flu shot)
are worth exploring.

CORRESPONDING AUTHOR: Marc Mitchell, Post-Doctoral Fellow, University Health Network,

Toronto, ON, M4C 3K6; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S163

A087 6:00 PM-7:00 PM

BEHAVIORAL MEDICINE RESIDENT EDUCATION AND A FORMAT FOR CURRICULUM

MODIFICATION

Jennifer Harsh, PhD, Tyler Lawrence, M.A. Professional Counseling

University of Nebraska Medical Center, Omaha, NE

Psychosocial problems are common in healthcare settings, with greater than 70% of patients
seen in primary care visits presenting with a psychosocial comorbidity. Non-comorbid mental
health and substance use disorders are anticipated to surpass physical disease as a cause of
worldwide disability by 2020. Also, patient-provider interactions are important because they
dictate relevant patient outcomes, affecting, for example, patient engagement in healthcare
decisions and medication adherence.

Because of the prevalence of patient psychosocial difficulties and the importance of patient-
provider interactions, we developed a behavioral medicine curriculum to assist residents in
enhancing behavioral medicine-related knowledge and clinical skills. The two mainstays of the
education program are monthly hour-long noon conference presentations and behavioral
medicine seminars. Noon conferences consist of didactics, audience participation, and small
group discussion. Topics are related to patient care in primary care and hospital settings.
Examples include motivational interviewing techniques, approaches to working with difficult-
to-treat patients, and brief systemically-focused interventions for depression and anxiety.
During behavioral medicine seminar, a behavioral medicine faculty member meets with two
to three residents to view recordings of patient-provider interactions for each resident in
attendance. The faculty member facilitates discussion surrounding the presented patient
cases, including strategies for agenda setting, building rapport, displaying empathy, and asking
open-ended questions. The goal of the behavioral medicine seminar is to provide residents
with feedback that can assist them with enhancing strategies for providing patient-centered
care.

Following each behavioral medicine seminar and noon conference, residents completed a
survey that contained questions on which behavioral medicine topics they plan to implement
in their practice, which they would like to learn, and how useful the training was overall.
Residents reported behavioral medicine education as being very useful. Findings from a
S164 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

qualitative thematic analysis of open-ended survey questions highlighted Motivational

Interviewing skills and agenda setting as topics which were particularly beneficial for their
practice. Results also highlighted residents desire to learn more about pain management
strategies and engaging families in care in the future.

CORRESPONDING AUTHOR: Jennifer Harsh, PhD, University of Nebraska Medical Center,

Omaha, NE, 68108; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S165

A088 6:00 PM-7:00 PM

DELIVERY OF MOTIVATIONAL INTERVIEWING WORKSHOPS TO CLINICS SERVING VULNERABLE

POPULATIONS

Lola Coke, PhD, ACNS-BC, RN-BC1, Beth-Anne Christopher, MS, RN, CNL2, Grantner Mary, MA,
CHCP1

1
Rush University Medical Center, Chicago, IL; 2Rush University College of Nursing, LaGrange, IL

Background: With grant support, a large urban medical center has formed a collaborative
partnership with a group of 13 Federally Qualified Health Centers to develop and implement a
motivational interviewing curriculum for its health care providers, care coordinators and
community health workers.

Purpose: Development, implementation and evaluation of a 4 hour tailored workshop with

follow-up concordant testing to teach principles of motivational interviewing, coaching and
relapse prevention techniques. Data was collected, pre-workshop, immediately post-
workshop and 8 weeks post-workshop.

Results: Workshops were delivered to 13 sites (N=114 health care providers). The majority of
participants were care coordinators (36.6% n = 37). Pre-workshop data showed that 94.3%
were familiar with MI, but only 68.6% thought they could use the technique to encourage a
commitment to change. Post -workshop data showed that compared to pre-workshop,
participants felt better prepared to: 1) build collaborative partnerships (84.7% vs 98.2%;
change increase 13.5%); and, 2) encourage patient commitment to change (68.6% vs 93.5%;
change increase = 29.4%). Post-workshop barriers included: 1) uncomfortable with content
(13.3%); 2) change is too time-consuming (33.7%); 3) patients are resistant to change (54.1%).
At 8 weeks, participants (n = 28) reported they were all (100%) building collaborative
relationships, and 93% were able to encourage patient commitment to change.

Conclusion: Each clinic valued the workshop as evidenced by high attendance, engagement
and participation in long-term follow-up measures. Learner responses indicated that within
this sample, the workshop was effective in moving participants to actual change in practice
S166 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

and in effectively promoting patient change. Additional strategies include reinforcement of

key learning objectives through-mail testlets and other on-going programing.

CORRESPONDING AUTHOR: Lola Coke, PhD, ACNS-BC, RN-BC, Rush University Medical Center,
Chicago, IL, 60654; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S167

A089 6:00 PM-7:00 PM

HPV VACCINATION TRAINING EXPERIENCE AND PREFERENCES AMONG FAMILY MEDICINE

RESIDENTS AND FACULTY: A MIXED-METHODS STUDY

Monica Kasting, PhD1, Paige W. Lake, BA1, Courtney L. Scherr, PhD2, Tracy Johns, PharmD,
MS3, Teri L. Malo, PhD, MPH4, Richard Roetzheim, MD5, Gwendolyn Quinn, Ph.D. 6, Susan
Vadaparampil, PhD MPH1

1
Moffitt Cancer Center, Tampa, FL; 2Northwestern University, Chicago, IL; 3USF-MPM Family
Medicine Residency, Clearwater, FL; 4UNC Lineberger Comprehensive Cancer Center, Chapel
Hill, NC; 5U. of South Florida, Tampa, FL; 6Moffitt Cancer Center - USF, Tampa, FL

Background: Physicians in the U.S. do not consistently recommend human papillomavirus

(HPV) vaccination for adolescents. Residency training is a pivotal time to establish skills for
career-long practices, particularly for challenging skills such as HPV vaccine recommendation.
We examined Family Medicine (FM) residents and facultys training experience and
preferences related to delivering HPV vaccine recommendations.

Methods: Residents (n=28) and faculty (n=19) were identified through a national FM
residency directory and recruited from training programs in Florida. Participants completed a
phone interview assessing key aspects of HPV vaccine recommendation training. Interviews
were audio recorded, transcribed verbatim, and analyzed using content analysis. A brief
follow-up survey was emailed after the interview. Survey data were analyzed using descriptive
statistics and faculty and resident responses were compared using Fishers exact tests.

Results: Qualitative. Residents training experience with HPV vaccination recommendation

varied from none to extensive, and was often self-directed. Variation in training was seen
between and within programs. Faculty often noted HPV vaccination training was not
standardized and residents lacked instruction about effective communication. Quantitative.
Most residents preferred to receive training focused on providing recommendations for HPV
vaccine through formal lectures (86%). All faculty participants (100%) indicated this format
was useful (p=0.14). Only 46% of residents indicated communication training was provided via
formal lectures in their program, but 90% of faculty reported such training was available
(p=0.005). Additionally, 79% of faculty thought videotaping residents recommending the
vaccine would be a useful approach for training, whereas only 14% of residents preferred this
method of instruction (p

Conclusions: Our study identified a lack of consistent and standardized training for delivering
S168 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

HPV vaccine recommendations. A training curriculum that uses multiple modalities and
reflects resident and faculty preferences is needed.

CORRESPONDING AUTHOR: Monica Kasting, PhD, Moffitt Cancer Center, Tampa, FL, 33612;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S169

A090 6:00 PM-7:00 PM

INCREASING BEHAVIORAL HEALTH & INTEGRATED CARE IN K-12 SCHOOLS

Lauren Bigham, Eds1, JungSu Oh, M.Ed.2, Jeffrey D. Okun, M.S., NCC2, Dontavious Ford, M.Ed.,
NCC3, Bernadette Heckman, PhD2, Jolie Daigle, PhD, LPC2

1
University of Georgia, Macon, GA; 2University of Georgia, Athens, GA; 3University of Georgia,
Stockbridge, GA

Background: In Georgia, there is a critical lack of behavioral health professionals and services
that results in children not receiving vital care. An urgent need exists to increase childrens
access to behavioral health services in Georgias school systems via innovative school
counseling training programs that move beyond generalist training. Guided by a
biopsychosocial and multicultural framework, school counseling masters students are
receiving training in behavioral health to place larger numbers of professionals in schools to
mitigate the gaps to such services. Objectives: To evaluate this innovative training approach,
the current study examines school counseling masters students perceived self-efficacy and
competency in the delivery of behavioral heath and integrated primary care services in
school-based settings. Research Design: Funded by HRSA, this study reports on preliminary
cross-sectional data collected from school counseling master's students (n=54) receiving
training in a large southeastern public university. In addition to completing the School
Counselor Self-Efficacy Scale and the Revised Multicultural Counseling Competence and
Training Survey, students completed the Readiness for Interprofessional Learning Scale and
the Interprofessional Education Collaborative Competency Survey. Results: Analyses of
baseline data suggests that masters level school counseling students maintained moderate
levels of self-efficacy (Mean=3.03, Possible Range=1-5), valued providing multidisciplinary
services in schools (Mean=4.30, Possible Range=1-5), were highly amenable to working with
interdisciplinary professionals (Mean=3.47, Possible Range=1-5), and held a more positive
view of interprofessional teamwork and collaboration (Mean=4.68, Possible Range=1-5).
Conclusions: School counselors in training appear to be open to and value innovative training
in the delivery of interprofessional and multicultural behavioral health services in K-12
schools. Future research will evaluate the long-term effectiveness and sustainability of this
innovative training approach to increase access to behavioral health and integrated primary
care services in K-12 schools.

CORRESPONDING AUTHOR: Lauren Bigham, Eds, University of Georgia, Macon, GA, 31210;
[email protected]
S170 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A091 6:00 PM-7:00 PM

INTEGRATING SIMULATION INTO TEAMSTEPPS TRAINING TO IMPROVE INTERPROFESSIONAL

COLLABORATIVE PRACTICE

Patricia Francis-Johnson, DNP, RN, Sharon Decker, PhD, RN, ANEF, FAAN, Charles F. Seifert,
Pharm.D., FCCP, BCPS, Brian Irons, PharmD, FCCP, BCACP, BC-ADM, Christina Esperat, RN,
PhD, FAAN, Huaxin Song, PhD

Texas Tech University Health Sciences Center, Lubbock, TX

Background
Poor communication between health professionals has been identified as a major cause of
care delay and poor team performance. Simulation-based activities have demonstrated the
ability to facilitate collaboration and improve communication among interprofessional teams.
TeamSTEPPS training has demonstrated the ability to improve interprofessional
communication, and teamwork skills improving quality of care and patient safety. Therefore,
the purpose of this project was to investigate how TeamSTEPPS training, which included clinic
specific simulations, improved the communication and team behaviors of the clinics
interprofessional personnel.
Methods
An interprofessional group of identified leaders were designated as champions of the project
and trained in TeamSTEPPS. Focus groups with the champions and clinic administration, along
with findings from observations by the researchers identified challenges affecting patient care
delivery. These challenges were integrated into the case scenarios for the simulation based
activities. The health professionals in our nurse managed community health center were
asked to complete this two-day training. A pre-post survey of the participants and observation
by the researchers in both simulation and at the clinic setting were used to obtain data. Tools
included: TeamSTEPPS Teamwork Attitudes Questionnaire, TeamSTEPPS Teamwork
Perceptions Questionnaire, and the Collaborative Practice Assessment Tool (CPAT).

Results
The training improved teamwork attitudes significantly on overall scores and subscales for
team structure, leadership, situation monitoring and communication. The training also
improved teamwork perception on communication significantly. The CPAT subscale
scores were found highly correlated to TeamSTEPPS Teamwork Attitudes Questionnaire
subscale scores.

Discussion
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S171

The changes in patient care delivery has resulted in unsolicited positive comments by
patients, and actual chronic disease management.

CORRESPONDING AUTHOR: Patricia Francis-Johnson, DNP, RN, Texas Tech University Health
Sciences Center, Lubbock, TX, 79430-6264; [email protected]
S172 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A092 6:00 PM-7:00 PM

THE INTEGRATIVE HEALTH & LIFESTYLE PROGRAM (IHELP): IMPACT OF AN ONLINE

INTERPROFESSIONAL CURRICULUM

Audrey Brooks, PhD1, Molly Burke, N/A2

1
University of Arizona Center for Integrative Medicine, Tucson, AZ; 2University of Arizona,
Tucson, AZ

Integrative health (IH) takes account of the whole person, including all aspects of lifestyle. It
emphasizes the therapeutic relationship between practitioner and patient, is informed by
evidence, and makes use of all appropriate therapies. The Integrative Health & Lifestyle
Program (IHeLp) was developed to meet a training gap in IH approaches for allied health care
professionals. IHeLp is 6-month, 250-hour interprofessional, online training program that aims
to educate health and wellness professionals in an IH approach to patient/client care focused
on addressing 7 core areas of health: Sleep, Movement, Nutrition, Relationships, Resiliency,
Environment, and Spirituality. IHeLp targets licensed, registered, or certified allied health
professionals with the goal of supporting interprofessional collaboration and well-trained IH
care teams. Participants learn to utilize lifestyle modifications that improve health and
wellbeing (nutrition, mindfulness, physical activity, complementary therapies, etc.) and
techniques to motivate change and reduce stress with an emphasis on Self-Care as applied
learning. IHeLp utilizes an online curriculum, weekly faculty mentor meetings, self-care teams,
ongoing assignments, faculty-moderated dialogues, and 4-day experiential retreat.
Knowledge-based exams assess attainment of learning objections. The program has 140
graduates. Mean score on the final exam was 91% (66-100% range). A final program
evaluation survey assess overall experience. Items are rated on a 5-point scale (5 is most
favorable) and the top 2 categories summed. The program evaluation was very positive with
97% finding the experience transformative, personally (100%) and professionally enriching
(98%), and 97% felt able to apply learning to real-world situations. Pre- and post-course
changes in burnout (Maslach Burnout Inventory), lifestyle behaviors (Arizona Lifestyle
Inventory), emotional intelligence (Interpersonal Reactivity Index), and attitudes toward
interprofessional collaboration (Attitudes toward Health Care Teams; ATHC) are examined.
Statistically significant improvements (p < 0.05) were found for burnout, specific lifestyle
behaviors in the areas of diet, mind-body practices, leisure, and sleep), and quality of care
(ATHC). No differences were found for emotional intelligence, time constraints (ATHC) and
some lifestyle behaviors. IHeLp is a unique, accessible option for allied health professionals to
address interprofessional and IH training needs.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S173

CORRESPONDING AUTHOR: Audrey Brooks, PhD, University of Arizona Center for Integrative
Medicine, Tucson, AZ, 85724-5153; [email protected]
S174 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A093 6:00 PM-7:00 PM

THE NATIONAL CENTER FOR INTEGRATIVE PRIMARY HEALTHCARE - ENHANCING

INTERPROFESSIONAL INTEGRATIVE HEALTH EDUCATION

Audrey Brooks, PhD1, Patricia Lebensohn, MD2

1
University of Arizona Center for Integrative Medicine, Tucson, AZ; 2University of Arizona,
Tucson, AZ

Evidence is accruing for the clinical and cost-effectiveness of integrative healthcare (IH);
however, there is a knowledge gap for primary care professionals, which has hindered
widespread adoption of IH into the healthcare system. The University of Arizona Center for
Integrative Medicine received a HRSA grant to establish the National Center for Integrative
Primary Healthcare (NCIPH) to address this need. Based on a coordinated set of IH
competencies across primary care professions and needs assessment, a 45-hour online
interprofesssional IH course, Foundations in Integrative Health (FIH), was developed and pilot-
tested in primary care training programs, e.g., family medicine, internal medicine residencies,
nursing, pharmacy, behavioral health, oriental medicine, chiropractic. Units include:
Introduction to IH in Primary Care; Prevention and Lifestyle Behavior Change; Healthcare
Professional Wellbeing; Addressing Patients through an Integrative Lens in Primary
Care; Integrative Interventions; Community Settings and Systems at Large. There were 2,200
course participants in the pilot, of which 696 (32%) completed the course. An evaluation
survey administered at the end of each unit assesses whether objectives were met,
educational depth, clinical utility, helpfulness of resources and reflections and ease of
technology. Items are rated on a 5-point scale with 5 the most favorable. Met objectives (4.5),
usefulness to patient care (4.2) and ease of technology (4.2) received the highest ratings.
Helpfulness of sharing reflections received the lowest rating (3.2). When asked whether the
unit should be incorporated into mandatory material, a majority (61-74%) rated the units as
mandatory. A final course evaluation survey assesses interest in applying IH principles in
practice, desire to seek additional IH education, whether course enhanced educational
experience, recommend course to others, incorporation of self-care practices learned in the
course, and site leader support for completing and relating course content to training. The FIH
course experience received very high ratings with most interested in applying what they
learned (4.5), continuing to seek IH education and training (4.3), and rated the course as
enhancing their educational experience (4.3). Two-thirds (67%) reported incorporating new
self-care practices into their lives based on what they learned in the course. The goal of NCIPH
is to transform primary care health professional education to include an emphasis on
providing an integrative approach to patient care utilizing an interprofessional collaborative
team. The FIH course can serve as a foundation in this effort. An online course addresses the
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S175

challenges of time, cost, and curriculum consistency and can be widely disseminated to the
entire spectrum of primary care training programs.

CORRESPONDING AUTHOR: Audrey Brooks, PhD, University of Arizona Center for Integrative
Medicine, Tucson, AZ, 85724-5153; [email protected]
S176 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A094 6:00 PM-7:00 PM

AWARENESS AND TRUST OF U.S. PUBLIC HEALTH AGENCIES: RESULTS FROM A NATIONALLY
REPRESENTATIVE SAMPLE OF ADULTS AND ADOLESCENTS

Sarah Kowitt, MPH1, Allison M. Schmidt, MPH2, Anika Hannan, MPH student3, Adam O.
Goldstein, MD, MPH4

1
University of North Carolina at Chapel Hill, Gillings School of Global Public Health, Chapel Hill,
NC; 2Gillings School of Global Public Health, UNC Chapel Hill, Chapel Hill, NC; 3Rollins School of
Public Health at Emory University, Fayetteville, NC; 4University of North Carolina at Chapel
Hill, Chapel Hill, NC

Trust in government agencies plays a key role in advancing these organizations' agendas,
changing health behaviors, and effectively implementing health policies. However, few studies
have examined how individuals trust the leading U.S. agencies devoted to protecting the
publics health. Using two national samples of adults (N=1125) and adolescents (N=5014), we
examined demographic factors, with a focus on vulnerable groups, that predicted awareness
and trust of the Centers for Disease Control and Prevention (CDC), Food and Drug
Administration (FDA), and the federal government. From nine different weighted logistic
regression models, we found high levels of awareness of the FDA and CDC (ranging from
55.7% for adolescents awareness of the CDC to 94.3% for adults awareness of the FDA) and
moderate levels of trust (ranging from a low of 41.8% for adults trust in the federal
government and a high of 78.8% for adolescents trust of the FDA). With respect to
awareness, across both adolescents and adults, being African American, of younger age,
having lower education, and having low numeracy were all associated with lower awareness.
With respect to trust in these agencies and the federal government as a whole, fewer
demographic differences were found. This research provides key novel insights into the
drivers of awareness and trust in the US federal government and its public health agencies.
Our findings suggest groups that these agencies may want to target to enhance trust and thus
facilitate their communication and regulatory agendas.

CORRESPONDING AUTHOR: Sarah Kowitt, MPH, University of North Carolina at Chapel Hill,
Gillings School of Global Public Health, Chapel Hill, NC, 27516; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S177

A095 6:00 PM-7:00 PM

BLACK-WHITE RACIAL DIFFERENCES IN PATIENT-PHYSICIAN COMMUNICATION: A SYSTEMATIC

REVIEW

Megan Shen, PhD1, Emily B. Peterson, PhD2, Rosario Costas-Muiz, PhD3, Carma Bylund, PhD4

1
Weill Cornell Medicine, New York, NY; 2National Cancer Institute, Rockville, MD; 3Memorial
Sloan Kettering Cancer Center, New York, NY; 4Hamad Medical Corporation, Houston, TX

Background: Ample evidence exists for the presence of racial disparities in health care, but it
is less clear what contributes to these disparities. Recent data suggests disparities in patient-
provider communication may contribute to these racial disparities in health care. The goal of
the present study was to systematically review studies examining the effect of race on
patient-physician communication.

Methods: A comprehensive search was conducted across six online databases between 1995
and 2016. Studies were included when the sample consisted of patients in health care
contexts and the communication outcome was observational or patient-reported. The search
resulted in 4,672 records for review and 40 articles for final inclusion in the review. Studies
were divided into eight main domains of patient-physician communication: (1) quality, (2)
satisfaction, (3) information-giving, (4) partnership building, (5) patient participation and
participatory decision-making, (6) positive and negative affect/talk, (7) length of visit/time and
talk-time ratio, and (8) other.

Results: Studies were heterogeneous in health contexts and communication measures, but
most results seemingly pointed to black patients having worse communication outcomes than
white patients, although some results were mixed. These results were most consistent for
quality, information-giving, and patient participation and participatory decision-making.
Results were mixed for satisfaction, partnership building, length of visit, and talk-time ratio.
Mixed results seemed to depend on the specificity of the communication measure as well as
its consistency across studies.

Conclusions: Overall, results of this study indicate that black patients may both receive and
perceive poorer communication than white patients. However, results also highlight the lack
of consistent, valid measured used in the research literature and the need for more
consistency across studies to illuminate these results. Nevertheless, results highlight the
potential need to train physicians on how to engage in higher quality communication with
S178 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

black patients by focusing on improving their patient-centeredness, information-giving,

partnership building, and engagement of black patients in communication processes.

CORRESPONDING AUTHOR: Megan Shen, PhD, Weill Cornell Medicine, New York, NY, 10065;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S179

A096 6:00 PM-7:00 PM

COMMUNITY HEALTH IMPROVEMENT RESEARCH IN SOUTHWEST KANSAS: A QUALITY

IMPROVEMENT PROJECT

Judy A. Johnston, MS, RD, LD1, Lisette Jacobson, PhD, MPA, MA2, Liuqiang (Kelsey) Lu, MS2,
David F. Gamble, Bachelor of Science in Education3, Sarah Weber, Bachelor of Business
Administration, May 20184, Jacob Rha, N/A5, Gillian Jolly, Bachelor of Science6

1
University of Kansas School of Medicine -Wichita, Wichita, KS; 2University of Kansas School of
Medicine-Wichita, Wichita, KS; 3Baylor University, Estes Park, CO; 4Baylor University, Colorado
Springs, CO; 5Baylor University, Carrollton, TX; 6Baylor University, Madison, MS

Background: In the United States, quality improvement (QI) initiatives mainly focus on the
health of urban populations and usually target one specific disease or health concern such as
mental health, cardiovascular disease, childhood obesity, the healthcare system, or a sub-
population. There are no studies that have assessed the health needs of an entire community
by economic sector. The purpose of this study was to assess community members
perceptions, knowledge, attitudes, and desires related to health and wellness resources in an
extremely frontier rural region (referred to as county) of a Midwestern state.

Methods: Households in the two largest cities of this county, with 235 and 781 households
respectively, were targeted to complete the assessment. The county had a total of 1,341
households. The research team trained 45 volunteers who went door-to-door to collect one
survey per household. The survey instrument was available in English and Spanish. Survey
questions focused on availability of health and wellness (h/w) services in the county and h/w
services offered at the worksite, daycare/preschool, public school, church, county extension
office, and at the community-level. Incentives were provided. Data were collected in
July/August, 2016. Descriptive analyses were conducted.

Results: With 865 surveys completed, the cities response rate - 85.1%; total county response
rate - 64.5%. The majority of respondents were female (72.9%), age 25-44 (39.3%), earned
$10,000-$50,000/year (32.3%), and non-Hispanic white (64.7%) followed by Hispanic (27.3%).
The majority of participants (41.1%) defined health and wellness as an active, lifelong
process of becoming aware of choices and making decisions toward a more balanced and
fulfilling life, and 60.9% reported that health/wellness is influenced by the environments
where I live, learn, work, play and pray. About 44.7% reported that they partially agreed
with the statement the healthcare system is responsible for my health/wellness followed by
32.8% who disagreed with this statement. Participants identified after-hours non-emergency
S180 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

care, weight management coaching, and mental health services as the top three most
important community priorities.

Conclusion: This community health assessment is a first step toward enhancing the quality of
h/w resources provided to rural residents. Follow-up studies are needed to identify feasible
interventions thereby improving health and wellness for this population.

CORRESPONDING AUTHOR: Judy A. Johnston, MS, RD, LD, University of Kansas School of
Medicine -Wichita, Wichita, KS, 67214; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S181

A097 6:00 PM-7:00 PM

DEVELOPMENT AND VALIDATION OF AN ENVIRONMENTAL HEALTH LITERACY ASSESSMENT

TOOL

Veronica Irvin, PhD, MPH, Diana Rohlman, PhD, Amelia Vaughn, BA, Rebecca Amantia, MPH,
Molly Kile, ScD

Oregon State University, Corvallis, OR

Introduction: Private well users in the US have access to well water testing assistance through
their county health departments. Though many citizens take advantage of this oppurtunity,
few follow through with treatment if they receive negative test results. A potential barrier to
treatment could be a lack of environmental health literacy. Environmental health literacy
integrates concepts from both environmental literacy and health literacy to develop skills and
competencies needed to seek out, comprehend, evaluate, and use information to make
informed choices, reduce health risks, improve quality of life and protect the environment.
However, measures of environmental health literacy have not been developed.

Methods: The purpose of this presentation is to describe the development of an

environmental health literacy (EHL) assessment tool and to assess the relationship of the EHL
against a validated health literacy scale administered in clinic settings, the Newest Vital Sign
(NVS). We used crowd sourcing techniques to administer the survey online to a sample of US
adults ages 18 and older. The two different measurement tools were compared on 1) absolute
values, 2) ranking ability in tertiles, and 3) Bland-Atlman plots with limits of agreement.

Results: A total of 911 adults started the survey, 95% completed it for a final sample of 865.
Participants were 57% female, 86% college-graduates, 53% home owners and reported a
mean age of 37.8 (SD=11.5). Both surveys averaged between 2 2.5 minutes to complete.
Scores ranged from 0-6 for both tools. Mean score was 5.0 (SD=1.2) for the standardized NVS
and 5.0 (SD=1.0) our EHL measure. Cronbach alpha for our EHL measure was alpha=0.51.
Correlation between both measures was r=0.468, p < 001. The number of participants in the
top tertile (score of 5 or 6) was 78% for both NVS and EHL with a percent agreement of 74.5%
across all tertiles. Histograms and Bland-Altman plots will be included in poster presentation.

Conclusions: The EHL was quick to administer and easy to complete. The EHL measure had
similar distributional properties and ranking abilities to the standardized NVS health literacy
measure. This tool should be tested in populations with lower access to internet or lower
educational attainnment. An assessment tool for environmental health literacy could identify
participants at low environmental literacy and aid with intervention development to improve
home environmental hazards.
S182 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Veronica Irvin, PhD, MPH, Oregon State University, Corvallis, OR,
97331; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S183

A098 6:00 PM-7:00 PM

GOOGLE ANALYTICS USAGE IN THE STUDY OF AN ONLINE PROGRAM DESIGNED FOR

PROSTATE CANCER SURVIVORSHIP

Sara Fleszar, N/A1, Alexander Adia, N/A2, Gianna M. Torre, BA1, Erin K. Tagai, PhD, MPH3,
Simon Hall, MD4, Manish Vira, MD4, Alexander Kutikov, MD3, David Y.T. Chen, MD3, Curtis
Miyamoto, MD5, Adam Reese, MD6, Suzanne M. Miller, PhD7, Shawna Hudson, PhD8, Michael
A. Diefenbach, Ph.D.1

1
Northwell Health, Manhasset, NY; 2Northwell Health, Wantagh, NY; 3Fox Chase Cancer
Center, Philadelphia, PA; 4Northwell Health, New Hyde Park, NY; 5Fox Chase Cancer Center at
Temple University Hospital, Philadelphia, PA; 6Lewis Katz School of Medicine/Temple
University Hospital, Philadelphia, PA; 7Fox Chase Cancer Center/Temple University Health
System, Philadelphia, PA; 8Rutgers Robert Wood Johnson Medical School, New Brunswick, NJ,
New Brunswick, NJ

Introduction: Google Analytics (GA) is a powerful tool to analyze web-based software usage.
While GA is widely used as a business tool, it has only sparsely been used in behavioral
research. GA was used as part of a larger trial to test the efficacy of PROGRESS, an online
multimedia program for prostate cancer survivorship. PROGRESS was designed to boost
coping and adjustment in prostate cancer survivors post treatment. Methods: GA enabled
tracking of individual user behavior within the PROGRESS website, recording accessed pages,
videos and downloaded documents. Usage was then related to common demographic
variables. Results: Users were mostly white (77%), married (83%), and well-educated (76%).
Of the 162 patients who received access to the website, 78 (48%) participants logged on and
viewed multiple pages. There was no association between users and non-users along race,
marital status, and education variables. The most popular feature on the website was the
Health Tracking Tool, which allows users to track their sleep, moods, and recovery of erectile
functioning. Other popular features were informational pdfs to construct a personalized
survivorship care plan (e.g. treatment information, follow-up care, and post-treatment health
changes), accessing external information of prostate cancer-relevant exercises (yoga &
Kegels), and a collection of physician and patient videos. Discussion: We succeeded in our
goal to create a web based support tool that was attractive to users across demographic
groups. Patients preferred interactive components, such as the health tracking tool, and
informational content that allowed them to record personalized information relevant for
future follow-up care (i.e., survivorship care plan). Measuring website usage in real time
through an online tool, such as GA, provides important information about patient needs and
S184 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

behaviors. The use of real-time access data reveals patient needs that are far more accurate
than retrospective reporting of usage.

CORRESPONDING AUTHOR: Sara Fleszar, N/A, Northwell Health, Manhasset, NY, 11030;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S185

A099 6:00 PM-7:00 PM

IMPACT OF AMERICAN SIGN LANGUAGE EDUCATION VIDEOS ON NUTRITION AND EXERCISE

KNOWLEDGE AMONGST THE DEAF COMMUNITY

Luis Mandujano, N/A1, Sandy Bohan, MPH2, Matthew Fager, N/A3, Patricia Branz, N/A3,
Vanessa L. Malcarne, Ph.D.4, Georgia Sadler, Ph.D5

1
University of California, San Diego, La Jolla, CA; 2San Diego State University/University of
California, San Diego Joint Doctoral Program in Public Health, La Jolla, CA; 3University of
California, San Diego Moores Cancer Center, La Jolla, CA; 4San Diego State University, San
Diego, CA; 5University of California, San Diego Moores Cancer Center, San Diego, CA

Reducing obesity and cardiovascular disease are public health goals. Studies support
improving nutrition and exercise knowledge to promote weight loss and physical activity. The
Deaf community faces communication barriers that hinder education efforts, such as limited
English proficiency and limited health terminology available in American Sign Language (ASL).
Intervening with culturally competent education about nutrition and exercise in ASL is needed
to overcome these communication barriers. In this study, Deaf-relevant education videos in
ASL were created to improve nutrition and exercise knowledge amongst the Deaf community.
The aim of this study was to test whether positive changes in nutrition and exercise
knowledge resulted from exposure to the education videos. We enrolled 215 Deaf men and
women. Enrollees were randomly assigned to one of two intervention groups: nutrition or
exercise video. Each group served as controls for the other group. Participants completed a
demographic survey, a nutrition knowledge test, and an exercise knowledge test. Knowledge
tests were completed at pre- and post-intervention and two months afterwards. Descriptive
statistics were presented as means and standard deviations. Repeated measures models
within linear mixed-effects models compared changes in knowledge test scores between
groups. The sample that completed the study consisted of nearly equal numbers of Deaf men
and women, between 18 and 87 years of age (N = 190; M SD: 42 16 years). They were
mostly non-Hispanic White and had completed education beyond high school. Nutrition
knowledge test scores can range from 0 to 17; exercise knowledge test scores, from 0 to 12.
Higher scores indicate higher knowledge. The nutrition group gained nutrition knowledge
after viewing the education video, and retained knowledge after two months (M SD: pre, 8.9
2.7; post, 11.0 2.3; two months, 10.5 2.8). The exercise group also gained and retained
exercise knowledge (M SD: pre, 5.7 1.9; post, 8.0 1.8; two months, 6.8 1.9). Adjusting
for age and education level, nutrition knowledge test scores increased by 0.8 (F(1, 379) = 67.2,
p < 0.0001) and exercise knowledge test scores increased by 0.5 (F(1, 379) = 33.9, p < 0.0001)
from pre-intervention. However, the nutrition group did not score differently on the nutrition
S186 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

knowledge test than the controls. Exercise knowledge scores amongst the exercise group
were generally higher by 1.05 than that of the controls (F(1, 184) = 22.2, p < 0001). ASL
education videos are a promising tool to communicate nutrition and exercise information to
the Deaf community. Exposure to the ASL exercise education video resulted in modest
improvements in exercise knowledge. Further study is needed to explore the impact of the
ASL education videos on promoting healthy dietary and exercise behavioral changes.

CORRESPONDING AUTHOR: Luis Mandujano, N/A, University of California, San Diego, La Jolla,
CA, 92093-0850; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S187

A100 6:00 PM-7:00 PM

PERCEPTIONS OF ADDICTIVENESS OF WATERPIPE TOBACCO AMONG YOUNG ADULT

WATERPIPE TOBACCO USERS

Darren Mays, PhD, MPH1, Kenneth P. Tercyak, PhD1, Isaac Lipkus, PhD2

1
Georgetown University Medical Center, Lombardi Comprehensive Cancer Center,
Washington, DC; 2Duke University School of nursing, Durham, NC

Waterpipe tobacco is a commonly used tobacco product among U.S. young adults, linked to
symptoms of nicotine dependence and acute and long-term negative health effects. Very little
research exists as to how to design and deliver public health messages to prevent and reduce
waterpipe tobacco use among young adults. To this end, this study investigated young adult
waterpipe tobacco users perceptions of the addictiveness of waterpipe tobacco and
preferences for waterpipe tobacco messaging. Young adults (n = 44, M age 25.3 years, 52%
female, 80% white) who had used waterpipe tobacco in the past month completed a brief,
cross-sectional online survey. Closed-ended measures assessed demographics, waterpipe and
other tobacco use, and perceived addictiveness of waterpipe tobacco. Open-ended items
measured additional perceptions of addictiveness of waterpipe tobacco and preferences for
waterpipe tobacco messages. Quantitative data were analyzed descriptively, and open-ended
data were coded using an iterative, constant comparison method to identify emerging
themes. Most participants used waterpipe tobacco monthly (59%); 86% also used one or
more other tobacco product. Participants averaged 4.0 (SD 2.0, range 0-13) on a waterpipe
tobacco dependence measure, but reported low perceived addictiveness (M 2.0, SD 0.9, range
1-4), low perceived chances of becoming addicted (M 3.0, SD 1.6, range 1-7), and low desire
to quit (M 3.0, SD 1.8, range 1-7). Open-ended data indicated participants believe intermittent
social use does not lead to addiction and that they can easily quit. However, some participants
expressed concerns that waterpipe tobacco addiction may lead to health harms, social stigma,
and financial costs. Participants indicated messages emphasizing negative health effects, using
attention-grabbing imagery, and conveying that waterpipe users are at risk of long-term
tobacco use would motivate cessation. Despite reporting moderate levels of waterpipe
tobacco dependence, study participants largely viewed that waterpipe tobacco is not
addictive. Messaging vividly conveying the health harms and risk of long-term addiction may
be optimal to reduce waterpipe tobacco use.

CORRESPONDING AUTHOR: Darren Mays, PhD, MPH, Georgetown University Medical Center,
Lombardi Comprehensive Cancer Center, Washington, DC, 20007; [email protected]
S188 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A101 6:00 PM-7:00 PM

PROMOTING HEATH AMONG MILITARY FAMILIES: 5210 HEALTHY MILITARY CHILDREN

Jennifer M. DiNallo, Ph.D.1, Erica Rauff, PhD2, Daniel F. Perkins, Ph.D.3

1
Pennsylvania State University, Boalsburg, PA; 2Pennsylvania State University, Seattle, WA;
3
Pennsylvania State University, State College, PA

Background: Disseminating evidence-informed health promotion strategies to parents has the

potential to improve health and decrease the prevalence of obesity among children. The 5210
Healthy Military Children campaign (HMC) is designed to spread a consistent health message
throughout Military communities. The campaign emphasizes 5 + fruits and vegetables, 2
hours or less of recreational screen time, 1 + hours of vigorous physical activity, and 0
sweetened drinks.

Methods: An effectiveness-implementation Hybrid Type III design (Curran et al., 2012) study
was conducted among six military installations, where 15 venues (e.g., child development
centers and commissaries) at each site were targeted by the 5210 HMC campaign using
venue-specific toolkits, posters, social media messaging, and flyers. The Point of Contact at
each installation was trained on content and campaign implementation strategies via a 7-
module online training. Opt-in text based surveys were used to evaluate the impact of the
campaign on awareness, knowledge, and health behaviors at 0, 4, 8, and 12 months.

Results: Survey participants were 115 adults, with the majority 25-40 years (61%), female
(71%), and with children (60%). Most people opted in to participate at a medical center (29%)
or via social media (23%). At baseline, 78% of participants were not aware of the 5210
message. Most participants indicated that they sometimes ate 5+ fruits/vegetables daily
(61%), sometimes spent more than 2 hours a day watching television, playing video games, or
surfing the web (42%), always spent 1 + hours engaging in daily physical activity (39%), and
rarely drank sweetened beverages daily (40%). Parental reports of their childrens behaviors
mirrored those of their own behaviors. Although attrition in the study declined over the 12-
month period, for those remaining, awareness and knowledge of the 5210 message increased,
with no change in behavior.

Discussion: Dissemination and implementation research seeks to translate knowledge into

practice. Improving the quality and consistency of health promotion messaging may improve
family and child health outcomes. This pilot study produced a wide array of lessons learned
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S189

and next steps in implementing and evaluating an evidence-informed health promotion

campaign in a military context.

CORRESPONDING AUTHOR: Jennifer M. DiNallo, Ph.D., Pennsylvania State University,

Boalsburg, PA, 16827; [email protected]
S190 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A102 6:00 PM-7:00 PM

PUBLIC UNDERSTANDING OF THE PURPOSE AND LIMITATIONS OF CANCER SCREENING: DATA

FROM HEALTH INFORMATION NATIONAL TRENDS SURVEY 4

Rebekah H. Nagler, PhD1, Sarah Parvanta, PhD, MPH2, Seth K. Goldman, Ph.D.3

1
University of Minnesota, Minneapolis, MN; 2RTI International, Research Triangle Park, NC;
3
University of Massachusetts Amherst, Northampton, MA

In recent years, questions about the appropriate use of cancer screeningincluding whether
and when screenings risks may outweigh its benefitshave risen to the fore. One such risk is
overdiagnosis, which refers to a diagnosis of cancer that would otherwise never have caused
symptoms or death in a persons lifetime. Although debates about the value of cancer
screening find their way into media coverage, we know little about public understanding of
these issues or whether there are significant communication inequalities in social groups
ability to access, understand, and act on this information. Using weighted data from the
National Cancer Institutes Health Information National Trends Survey (HINTS 4, Cycle 4;
N=3,677), we find that public understanding of the purpose and limitations of screening
varies. Most respondents (91.2%) agreed that when a [screening] test finds something
abnormal, more tests are needed to know if it is cancer. However, only one-quarter (28.2%)
correctly identified as false the statement that These tests can definitively tell that a person
has cancer. Even fewer (20.0%) recognized that the harms of these tests and exams
sometimes outweigh the benefits, despite evidence that this is true for certain screening
modalities, such as prostate-specific antigen testing. Importantly, screening perceptions were
socially patterned: Respondents with more education and income were more likely to agree
that screening tests cannot definitively tell that a person has cancer (p < .001), and while
dont know responses were frequent overall, this was particularly true among respondents
of lower socioeconomic position (SEP). Given ongoing debates about the value of screening
and concerns about overdiagnosis, it is important for providers to help patients understand
screenings risks and benefits so they can make informed decisions. This is particularly critical
for lower SEP and other underserved populations, given persistent communication
inequalities. Additional implications of study findings for behavioral interventions will be
discussed.

CORRESPONDING AUTHOR: Rebekah H. Nagler, PhD, University of Minnesota, Minneapolis,

MN, 55455; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S191

A103 6:00 PM-7:00 PM

RATIONALE, PROCEDURES, AND RESPONSE RATES FOR THE 2015 ADMINISTRATION OF NCIS
HEALTH INFORMATION NATIONAL TRENDS SURVEY

Kelly D. Blake, ScD1, David B. Portnoy, PhD, MPH2, Annette Kaufman, PhD, MPH3, Chung-Tung
Jordan Lin, Ph.D.4, Serena Lo, PhD5, Eric L. Backlund, M.S.6, David Cantor, Ph.D.7, Lloyd Hicks,
MS8, Amy Lin, MS7, Andrew R. Caporaso, MS7, Terisa Davis, MPH7, Richard P. Moser, PhD9,
Bradford W. Hesse, PhD10

1
Behavioral Research Program, Division of Cancer Control and Population Sciences, National
Cancer Institute, National Institutes of Health, Bethesda, MD; 2Food and Drug Administration,
Center for Tobacco Products, Silver Spring, MD; 3Tobacco Control Research Branch, Behavioral
Research Program, Division of Cancer Control and Population Sciences, National Cancer
Institute, National Institutes of Health, Rockville, MD; 4FDA Center for Food Safety and
Applied Nutrition, College Park, MD; 5Center for Food Safety and Applied Nutrition, Food and
Drug Administration, Washington, DC; 6U.S. Food and Drug Administration, Silver Spring, MD;
7
Westat, Rockville, MD; 8Westat Inc, Rockville, MD; 9National Cancer Institute, Bethesda, MD;
10
National Cancer Institute, Rockville, MD

Background: The National Cancer Institute (NCI) developed the Health Information National
Trends Survey (HINTS) to monitor population trends in cancer communication practices,
information preferences, health risk behaviors, attitudes, and cancer knowledge. The US Food
and Drug Administration (FDA) recognized HINTS as a unique data resource for informing its
health communication endeavors, and partnered with NCI to field HINTS-FDA 2015. Methods:
HINTS-FDA 2015 was a self-administered paper instrument sent by mail May 29-September 8,
2015 using a random probability-based sample of US postal addresses, stratified by county-
level smoking rates, with an oversample of high and medium-high smoking strata to increase
the yield of current smokers responding to the survey. Results: The response rate for HINTS-
FDA 2015 was 33%, N=3738. The yield of current smokers (N=495) was lower than expected,
but the sampling strategy achieved the goal of obtaining more former smokers (N=1132).
Conclusion: The public-use HINTS-FDA 2015 data and supporting documentation are available
for download and secondary data analyses as of June 2016 at http://hints.cancer.gov. NCI and
FDA encourage the use of HINTS-FDA for health communication research and practice related
to tobacco-related communications, public knowledge, and behaviors, as well as beliefs and
actions related to medical products and dietary supplements.
S192 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Kelly D. Blake, ScD, Behavioral Research Program, Division of

Cancer Control and Population Sciences, National Cancer Institute, National Institutes of
Health, Bethesda, MD, 20892-9671; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S193

A104 6:00 PM-7:00 PM

RESOURCES FOR RESEARCHERS: A NEW PORTAL TO ASSIST PREVENTION INVESTIGATORS

Deborah Langer, MPH, Jody Engel, MA, RD, Elizabeth Neilson, PhD, MPH, MSN, Andrey
Kuzmichev, PhD, Murray David, PhD

National Institutes of Health, Office of Disease Prevention, Rockville, MD

Background

The Office of Disease Prevention (ODP) is the lead office at the National Institutes of Health
(NIH) responsible for assessing, facilitating, and stimulating research in disease prevention and
health promotion, and disseminating the results of this research to improve public health.
Increasing the scope, quality, dissemination, and impact of prevention research supported by
the NIH is central to the ODPs mission.

Methods

To assist the extramural research community to develop quality research projects that
address a wide range of research needs, gaps, and priorities, the ODP has created a new
online information portal: Resources for Researchers. This dissemination effort, which
incorporates domains and strategies from the Framework for Knowledge Translation, aims to
increase investigator capacity and lead to research that informs improved clinical practice,
health policy, and community health programs. The ODP collected information from NIHs
Office of Extramural Research, Center for Scientific Review, and other Institutes and Centers
to develop this portal. The ODP also collaborated with the Agency for Healthcare Research
and Quality and the Centers for Disease Control and Prevention to provide information on
research gaps identified through rigorous systematic reviews by the U.S. Preventive Services
Task Force and the Community Preventive Services Task Force.

Results

Resources for Researchers provides investigators with a variety of tools and resources
including (1) instructions for finding NIH-funded research projects; (2) directions for applying
for NIH funding; (3) a list of prevention-related funding opportunity announcements; (4) a list
of methods-related funding opportunity announcements; (5) a directory of prevention-related
NIH programs and offices; (6) a list of prevention-related study sections at the NIH; (7) an
S194 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

innovative database that is manually curated by experts, keyword-searchable, and categorized

by research topics; (8) a comprehensive list of resources to assist dissemination and
implementation researchers; and (9) information on high-priority evidence gaps identified by
the U.S Preventive Services Task Force and the Community Preventive Services Task Force.
The information about evidence gaps provides investigators seeking new topics of inquiry
with a selection of understudied areas in prevention science.

Conclusions

Resources for Researchers combines prevention-related information across various NIH

websites with newly created tools and materials in one online repository specifically tailored
to the needs of the prevention research community. This dissemination effort is intended to
build investigator capacity and has shown to be a useful resource for prevention researchers,
particularly new investigators.

CORRESPONDING AUTHOR: Deborah Langer, MPH, National Institutes of Health, Office of

Disease Prevention, Rockville, MD, 20892; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S195

A105 6:00 PM-7:00 PM

SOCIAL AND COMMUNICATION NETWORKS, CANCER INFORMATION, AND SELF-COLLECTED

HPV TESTING AMONG WOMEN IN APPALACHIAN KENTUCKY

Katherine S. Eddens, PhD, MPH1, Jesse M. Fagan, MA2, Tom Collins, BS3

1
University of Kentucky College of Public Health, Lexington, KY; 2University of Kentucky
Gatton College of Business and Economics, lexington, KY; 3University of Kentucky Rural Cancer
Prevention Center, Lexington, KY

In Appalachian communities, social networks are powerful community ties. However, there
are few studies that examine how these network ties might be used to communicate
information and norms about cancer screening and prevention behaviors. This study explores
peer word-of-mouth communication networks in rural Appalachian women who have used
self-collected vaginal swab (SCVS) for HPV testing, an innovative screening methodology in
this population. The goal is to investigate how best to activate networks to disseminate cancer
prevention and screening information, specifically innovative screening methodologies, in
rural Appalachia. Network interviews were conducted with 50 women in Appalachian
Kentucky who have not had a Pap test in three or more years, and 34 members of their
networks (alters). Community health workers conducted interviews in the field using tablet-
based softwarefor guided network data collection. They obtained information about
participants social and communication networks, study referral activity, source of health
information and health care services, and characteristics of participants network members
(alters) and their relationships. Participants seek health information from health care
providers (HCP; 51%), the internet (25%), and family and friends (16%), but get cancer
information primarily from the internet (30%), followed by HCP (21%), and friends and family
(11%). Of the 80% of participants that use the internet, most (67%) access it from a
smartphone. Preliminary network descriptives show a mean network size of 12.6 alters
(median 9, range 2-27) and a mean network density of 0.67 (median 0.63, range 0.13 1.00).
However, 43% of participants have networks with a density of 1.00, meaning everyone in their
network knows one another. This type of dense network structure can limit access to new
information and resources, such as innovative cancer screening methodologies. Additional
results will include communication network characteristics, what type of network
characteristics distinguish women who refer others to SCVS testing, and whom they refer.
Determining how network structure and content influences access to cancer information and
cancer prevention norms can allow public health practitioners to tailor dissemination
interventions in peer networks in low-income, marginalized populations such as rural
Appalachian women.
S196 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Katherine S. Eddens, PhD, MPH, University of Kentucky College of

Public Health, Lexington, KY, 40506; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S197

A106 6:00 PM-7:00 PM

TAILORED HEALTH COMMUNICATION FOR PRECLINICAL ALZHEIMER'S DISEASE

Jonathan Gooblar, MA

VA Palo Alto Health Care System, Palo Alto, CA

Educating patients about symptoms, prognosis, and treatment of disease is a cornerstone of

competent patient care. Tailored Health Communication (THC) accounts for an
individual's preferences, characteristics, and needs in the education process. The present
study examines the influence of individual differences on the effectiveness of an educational
protocol about preclinical Alzheimers disease (pAD), a novel at-risk diagnostic category. A
total of 581 adults between the ages of 18 and 65 were recruited on Amazon Mechanical Turk
(50.8% female; 64.9% Caucasian; mean age=37.32 years). Participants were randomly
assigned to one of three education conditions: education-as-usual (EAU; basic disease
information); learning-to-criterion-basic (LTCbasic; basic disease information plus criterion
learning); and learning-to-criterion-extended (LTCextended; basic disease information plus in-
depth risks, benefits, and limitations plus criterion learning). Participants responded to two
knowledge questions, and LTCextended participants responded to two additional in-depth
knowledge questions. Participants also reported on their level of education, health literacy,
experience with AD, and knowledge about AD. N=283 participants (51.3%) responded
correctly to the knowledge questions. On the two basic knowledge questions, performance in
the LTCextended condition was significantly better than in the other conditions. On the two
basic knowledge questions, performance in the LTCextended condition was significantly
better than in the other conditions (2(2, N=581)=40.80, p < .001). Participants in the
LTCextended intervention reported lower levels of interest in obtaining predictive testing for
AD compared to participants in the other interventions (F(2,581)=11.31, p < .001, partial
2=.038). Individual differences affected performance on the knowledge questions
(2(11)=94.75, p < .001), with health literacy (lower confidence in filling out medical forms and
few problems learning about medical conditions) and higher objective AD knowledge
predicting better performance. More experience with AD predicted higher interest in
obtaining testing for AD (F(1,570)=8.09, p=.005, partial 2=.014) with education condition in
the model. Overall, participants knowledge of this complex diagnostic category depended on
individual difference variables and type of education. This study provides preliminary
evidence for the effectiveness of THC in education for pAD.

CORRESPONDING AUTHOR: Jonathan Gooblar, MA, VA Palo Alto Health Care System, Palo
Alto, CA, 94304; [email protected]
S198 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A107 6:00 PM-7:00 PM

WHAT (PREGNANT) WOMEN WANT: RESULTS FROM A WEB-BASED NEEDS ASSESSMENT OF

WEIGHT-RELATED MOBILE APP MOTIVATIONS AND EXPERIENCES

Alicia A. Dahl, MS1, Brie Turner-McGrievy, PhD, MS, RD2, Sara Wilcox, PhD2, Jihong Liu, Sc.D.2,
Rachel Davis, PhD2

1
University of South Carolina, West Columbia, SC; 2University of South Carolina, Columbia, SC

BACKGROUND: Technology has revolutionized the way people communicate. Mobile

applications (apps) and websites hold potential as a convenient, readily available, and
personalized resource for women during pregnancy. Electronic health (e-health) interventions
could be used as a parallel or supplemental tool to deliver pregnancy weight-related
information.

METHODS: Using Facebook and Twitter social media sites, pregnant women living in the
United States who owned a smartphone were recruited to complete a brief one-time online
survey regarding the needs and interests of pregnant women to guide the development of an
e-health intervention. The survey included questions regarding pregnancy history, weight-
related knowledge and behaviors, topics discussed with health care practitioners, information
sought by pregnant women, and interests in electronically delivered health information.
Descriptive statistics were used to summarize the data.

RESULTS: A total of 330 pregnant women completed the survey, and 327 participants self-
reported pre-pregnancy height and weight (mean BMI =26.1+6.5). Half of the participants
(50.6%) reported that they did not receive recommendations for healthy gestational weight
gain (GWG) from their physicians during pregnancy. Ninety-one percent reported that they
did not receive counseling on weight-specific information but 37.3% indicated interest in
receiving such information during their pregnancy from a doctor, nurse, or other health care
worker. Participants reported receiving their nutrition and physical activity information from
websites (59.9%), a doctor (49.5%), friends/family (40.7%), cell phone apps (23.9%), a
nurse/health care worker (23.5%), blog posts (21.1%), social media (18.3%), or
books/magazines (16.8%). When asked about interest in receiving health information via
electronic resources, 73% of participants were interested in podcasts, online group-based
nutrition challenges (75.5%), group-based exercise challenges (76.9%), and mobile apps
(81.5%). Moreover, 83.9% were interested in a mobile app for tracking GWG, and 87.3%
reported use of pregnancy apps.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S199

CONCLUSIONS: Participants from social media sites reported high use of electronic resources
during pregnancy for health-related information. Focusing e-health interventions around
weight management information and GWG tracking may fill a need in prenatal care. Future
development of e-health interventions using websites or mobile apps would be beneficial for
pregnant women.

CORRESPONDING AUTHOR: Alicia A. Dahl, MS, University of South Carolina, West Columbia,
SC, 29169; [email protected]
S200 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A108 6:00 PM-7:00 PM

FEAR OF DEPORTATION AND ITS EFFECTS ON HEALTH SERVICE USE AMONG LATINO
IMMIGRANT MEN WHO HAVE SEX WITH MEN

Thespina Yamanis, Ph.D., MPH1, Maria Cecilia Zea, Ph.D.2, Suyanna Linhales Barker, Dr. Sc.3,
Allison O'Rourke, MPH1, Ana Maria del Rio Gonzalez, Ph.D.2

1
American University, Washington, DC; 2The George Washington University, Washington, DC;
3
La Clnica del Pueblo, Washington, DC

Introduction. In the past fifteen years, there has been a 600% increase in the number of
deportations, yet few studies have explored fear of deportation and how it affects health
disparities. Fear of deportation may be conceptually related to vigilance, the everyday
anticipation of discrimination by African-Americans. Research is needed to understand how
fear of deportation impacts health care use among young Latino men who have sex with men
(MSM), the Latino population at highest risk for HIV, for whom health care avoidance could
result in swifter progression to AIDS and increased HIV transmission.

Methods. We conducted in-depth interviews with 18 non-US citizen immigrant Latino MSM
aged 18-34 who live in the Washington, DC metro area. We asked about places and
circumstances where they feel unsafe, experiences with immigration authorities, and use of
health services. Interviews were conducted in Spanish for up to 90 minutes. Verbal consent
was obtained. Data were protected through a federal Certificate of Confidentiality.

Results. Nearly all participants were from Central America (El Salvador, Honduras, and
Mexico). Their average age was 25. The average length of time the participants were in the
US was 6.4 years (range 0.2 year 18 years). Eleven participants were undocumented
(without legal papers to live/work in the U.S.) and were undergoing a legal process to adjust
their immigration status, six were undocumented with no legal process, and one was a
permanent resident. Five participants were HIV positive. Most participants described being
worried about deportation because of violence, threats of death, and discrimination in their
countries of origin. Some reported anxiety due to recent raids on undocumented immigrants.
Participants fears resulted in avoiding certain activities, including driving, going out late at
night, and traveling to other states. As one participant explained: I dont leave the
apartment. I only go out for important things. This [fear of deportation] is something that you
always have on your mind. It is very different for someone who has papers. They can go out to
the places they want. Thirteen participants had health insurance, mostly through DC Health
Care Alliance, one of the few state-sponsored programs that provides health insurance for the
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S201

undocumented. One undocumented participant became unemployed after lacking medical

care for an injury. A few participants were afraid to access health care for fear they would
encounter law enforcement and be deported, including one participant who was HIV positive.

Conclusion. The persistent threat of deportation generated everyday stress for our
participants and contributed in some cases to their avoidance of health services. We will
discuss how these qualitative themes have been translated into survey items for further
measurement and testing among immigrant Latino MSM.

CORRESPONDING AUTHOR: Thespina Yamanis, Ph.D., MPH, American University, Washington,

DC, 20016; [email protected]
S202 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A109 6:00 PM-7:00 PM

GENERATIONAL DIFFERENCES IN SEXUAL BEHAVIORS AND PERCEPTIONS AMONG GAY,

BISEXUAL, AND QUEER MEN IN RURAL AREAS OF THE USA

Natalia Truszczynski, MPH1, Danielle N. Lambert, MPH, CHES2, Anne Marie Schipani-
McLaughlin, MPH1, Carolyn Lauckner, PhD1, Nathan Hansen, PhD1

1
University of Georgia, Athens, GA; 2University of Georgia, Atlanta, GA

Background: Gay, bisexual, queer and other men who have sex with men (GBQ men) have
been put into one monolithic group, MSM (men who have sex with men), when considering
sexual risk. However, there are many factors that create differences of risk among GBQ men,
including age. For example, older GBQ men may perceive HIV transmission as more likely and
deadly than younger gay man due to growing up during the 1980/90s AIDS crisis.

Methods: This qualitative study explored the health, relationships, and technology use of GBQ
men in rural areas of the United States. Eligibility criteria included identifying as a man who
experiences attraction to other men, residing in a rural area, and being at least 18 years old. In
total, 20 semi-structured interviews were conducted from July 2015 to June 2016. Interviews
were transcribed verbatim and analyzed in NVivo 11 by a diverse coding team using grounded
theory approach.

Results: Themes of generational differences were common throughout the interviews. Most
men who were 35 years of age or older, described feeling as if younger men, usually under 18
years of age, did not perceive sexual risk as severe and did not fear HIV/AIDS. They described
younger men as more willing to engage in unsafe sex, and even ask their partners to forego
protection, either for pleasure or to actually help transmission of the disease. Additionally,
older men perceived younger GBQ men to engage in more casual sex and to be less focused
on pursuing long-term relationships than older GBQ men. Young GBQ men felt pressure to
commit and be in relationships, and often rejected those pressures. Lastly, almost all older
GBQ men shared stories of being asked to be younger mens silver daddies, where the
partners would have unprotected, casual sex. These propositions were universally met with
disdain.

Conclusion: Understanding how young GBQ men perceive HIV/AIDS and transmission are
important in creating effective interventions to reduce transmission of HIV. Interventions can
no longer just focus on reducing barriers and increasing knowledge; instead we have to
additionally focus change perceptions of risk and safety. Interventions should also be tailored
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S203

toward different generations of GBQ men, adolescents, young adults, and older adults, as all
groups view sexual and relationships risk differently.

CORRESPONDING AUTHOR: Natalia Truszczynski, MPH, University of Georgia, Athens, GA,

30602; [email protected]
S204 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A110 6:00 PM-7:00 PM

MERITORIOUS AWARD WINNER

HIV TESTING AVOIDANCE: PSYCHOSOCIAL FACTORS AFFECTING HIV TESTING IN GAY AND
BISEXUAL MEN

Devon Price, MS, Stephanie Finneran, MA, Redd Driver, BA, Seth Kalichman, PhD

University of Connecticut, Storrs, CT

The CDC recommends men who have sex with men (MSM) get tested for HIV every three
months after having unprotected anal sex (CDC, 2015). HIV testing is treated as the first step
towards HIV prevention. However, there are significant social and psychological barriers to
testing uptake. Even with improved HIV testing technology that can deliver accurate results in
a matter of minutes, 1 in 8 people living with HIV are still unaware of their positive status
(CDC, 2015). With the HIV testing gap still such a pressing issue, we sought to pin point which
psychosocial factors might affect testing. This study examined a three psychosocial factor
model of HIV testing decisions among men who have sex with men (MSM). Two of the
psychosocial factors, HIV stigma and fatalistic beliefs regarding an HIV positive diagnosis, have
been well documented in the literature on HIV testing and psychosocial barriers. However,
the third psychosocial factor, the defensive process of avoiding relevant health information,
has received less attention in relation to HIV testing. The present study investigated how
these factors impact both past and contemporaneous HIV testing behaviors in MSM.

Surveys were collected from 400 gay and bisexual men at the 2015 Atlanta Pride Festival. Men
answered questions about their HIV beliefs and attitudes, HIV testing history, sex behavior,
and whether or not they would like to receive an at-home HIV test. A multivariate multinomial
logistic regressions predicting testing history (never tested, tested but not recently, recently
tested) revealed the only significant psychosocial predictor of group membership was HIV
status related information avoidance. Those who reported greater HIV status related
avoidance tendencies were more likely to have never been tested than those who were
tested recently (in 2015), OR 0.62, 95% CI [0.50-0.76], or those who had been tested but not
recently (not in the year 2015) OR 0.72, 95% CI [0.57-0.91]. A logistic regression analyses was
conducted in order to assess predictors of contemporaneous HIV testing uptake. After
controlling for covariates, HIV status related information avoidance was significantly related
to testing uptake above and beyond the other psychosocial factors, OR 0.85, 95% CI [0.73-
0.98]. Those who were higher in HIV status related information avoidance were less likely to
accept the at-home HIV test.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S205

Overall, our results indicate that an important potential mechanism has been missing from
the study of psychosocial barriers to HIV-testing in the literature. HIV status related
information avoidance predicts HIV-testing behavior above and beyond the psychosocial
factors that have previously been considered the primary motivators for avoidance. Further
examination of information avoidance and how it functions in relation to HIV testing decisions
is warranted for furthering the understanding of why at-risk populations choose not to test for
HIV.

CORRESPONDING AUTHOR: Devon Price, MS, University of Connecticut, Storrs, CT, 06269-
1020; [email protected]
S206 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A111 6:00 PM-7:00 PM

SEX TRADE, VIOLENCE, HIV: MULTIPLE TRAUMAS IMPACTING UTILIZATION OF CARE

Lianne Urada, PhD1, Laramie R. Smith, PhD2, Jennifer Yore, MPH3, Daniel P. Triplett, Sanne P.
Moller, Sankari Ayyaluru, John Edmiston, Melissa Ramos, Pamela M. Schwartz, Alexadra X.
Caraballo, Anita Raj, PhD4

1
School of Social Work, San Diego State University and Center on Gender Equity and Health,
Division of Global Public Health, UC San Diego, La Jolla, CA; 2University of California San Diego
School of Medicine, La Jolla, CA; 3Center on Gender Equity and Health (GEH) at University of
California, San Diego, La Jolla, CA; 4University of California San Diego, La Jolla, CA

As part of Kaiser Permanentes Community-Based HIV Test & Treat Initiative, this study
hypothesizes that sex trade involvement and experiencing recent physical or sexual violence
are associated with lower levels of engagement in HIV care, HIV case management, and in
other health services (substance use treatment, mental health services). Methods: Baseline
survey data collected from October 2013-June 2016 as part of the multi-site Community-
Based Linkage to and Retention in HIV Care study were used in the current analyses.
Community recruitment of HIV-positive participants (N=581) was conducted by seven agency
sites across six US cities. Multivariate multilevel logistic regression models were used to assess
associations of sex trade and recent history of victimization from violence (physical and sexual
violence) with outcomes of HIV care utilization in the past 6 months (defined as a medical
appointment in which they received their CD4 count), HIV case management, mental health
services, and substance abuse treatment. Models adjusted for age, race, gender, recent
incarceration and substance use, being insured, running out of cash, and individuals nested
within study sites. Results: Participants (N=582, median age 45 (IQR: 33-52) were majority
heterosexual (53.6%), Gay/Lesbian (28.2%), and Bisexual (15.1%), although 14 (2.4%) did not
specify their sexual orientation. Most identified as male (65.8%); 40.9% were cisgender
females, and 10.7% were transgender females. Most participants were Black/African
American (74.7%); 15.1% were Latino. 11.4% were involved in sex trade (past 90 days); 3.6%
were physically abused (past year), and 2.8% were sexually abused (past year). 36.4% did not
receive an HIV care appointment in the past 6 months, 63.9% missed a medical appointment
in the past 3 months; 30.8% did not receive an HIV case management appointment in the past
3 months. 35.2% received mental health services in the past 12 months, and 18.4%
participated in substance abuse treatment in this same timeframe. Adjusted analyses
indicated no associations between recent sex trade involvement and receipt of HIV medical
care, case management, or mental health services, though sex trade was associated with
likelihood of having missed a healthcare appointment in the past 3 months (OR=2.58, CI:1.37-
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S207

4.88). Incarceration (OR=1.78, CI=1.16-2.83), being uninsured (OR=1.88, CI=1.08-2.35), and

not having enough money was associated (OR=0.86, CI:0.74-0.99) with missing healthcare
appointments. No associations were seen between recent physical or sexual violence and our
outcomes of interest. However, being insured (OR=2.44, CI:1.51-3.95) was associated with HIV
case management and mental health care (OR=1.71, CI: 1.01-2.87). Conclusion: HIV-positive
participants involved in sex trade were more likely to have missed healthcare appointments
and to utilize substance abuse treatment.

CORRESPONDING AUTHOR: Lianne Urada, PhD, School of Social Work, San Diego State
University and Center on Gender Equity and Health, Division of Global Public Health, UC San
Diego, La Jolla, CA, 92037; [email protected]
S208 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A112 6:00 PM-7:00 PM

USING NEW MEASURES TO EXPLORE FACTORS ASSOCIATED WITH PREP INTENTIONS AND USE

Jennifer L. Walsh, PhD

Center for AIDS Intervention Research, Department of Psychiatry and Behavioral Medicine,
Medical College of Wisconsin, Milwaukee, WI

Background. Pre-exposure prophylaxis (PrEP) shows great promise to significantly impact the
HIV epidemic; however, use among those most at risk, including men who have sex with men
(MSM) and transgender (TG) individuals, remains relatively rare. Research is needed to
determine appropriate targets for interventions to increase PrEP uptake. One basis for HIV
prevention interventions is the Information-Motivation-Behavioral Skills (IMB) model.
However, we do not yet know whether this model applies to PrEP intentions and use.

Methods. Participants completed a survey at a community event in a mid-sized city in the

midwestern U.S. in 2016 (N=357, 64% White, 22% Black, 14% Latino, Mage=35). Only HIV- men
(93%) and TG men and women (6%) who engaged in sex with men were included in this
analysis. Items to assess PrEP knowledge, attitudes, stigma, descriptive and subjective norms,
and intentions were developed and reviewed by experts in HIV prevention and HIV providers.
Participants responded to the new measures and reported their current PrEP use. Structural
equation modeling was used to test the IMB model, with separate models for intentions (for
those not on PrEP) and use (for all participants). Models included demographic controls.

Results. All new measures performed well based on evaluations of items and correlations,
confirmatory factor analysis, and tests for differential item functioning; all scales were reliable
(s=.83-.94). Only 12% of participants were taking PrEP, although 69% had heard of PrEP and
37% said they would probably/definitely start taking PrEP in the next 3 months.

A model with acceptable fit (2(267)=417.37, RMSEA=.04, CFI=.95) explained 35% of the
variance in PrEP use intentions. Stigma and descriptive norms predicted self-efficacy, B = -
0.18, p < .05, and B = 0.26, p < .05, respectively. Attitudes, descriptive norms, and self-efficacy
predicted intentions, B = 0.28, p < .05, B = 0.32, p < .01, and B = 0.17, p < .05, respectively.
However, neither knowledge nor subjective norms was associated with intentions, and there
was no mediation.

A second model with good fit (X2(90)=308.57, RMSEA=.03, CFI=.96) explained 57% of the
variance in use. Attitudes, stigma, and descriptive norms predicted self-efficacy, B = 0.28, p <
.05, B = -0.20, p = .001, and B = 0.22, p < .01, respectively. Knowledge, stigma, and self-efficacy
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S209

predicted use, B = 0.28, p < .01, B = -0.18, p < .05, and B = 0.29, p = .001, respectively.
Attitudes, stigma, and descriptive norms had indirect effects on use via self-efficacy, in line
with the IMB model. However, subjective norms were not associated directly or indirectly
with use.

Conclusions. New PrEP-related measures were successfully piloted with a diverse sample of
MSM. Results suggested that the IMB model may be useful when developing PrEP
interventions, with knowledge, attitudes, stigma, descriptive norms, and self-efficacy all
playing roles.

CORRESPONDING AUTHOR: Jennifer L. Walsh, PhD, Center for AIDS Intervention Research,
Department of Psychiatry and Behavioral Medicine, Medical College of Wisconsin, Milwaukee,
WI, 53202; [email protected]
S210 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A113 6:00 PM-7:00 PM

FACTORS CONTRIBUTING TO PRIMARY CARE VISITS AMONG HEMODIALYSIS PATIENTS IN A

PATIENT-CENTERED MEDICAL HOME INTERVENTION

Ifeanyi Chukwudozie, MPH1, Marian Fitzgibbon, PhD2, Linda Schiffer, MS, MPH3, Michael
Berbaum, PhD4, Cheryl Gilmartin, PharmD3, Pyone David, MSW3, Eson Ekpo, RN3, Michael
Fischer, MD, MHS5, Anna Porter, MD6, Alana Aziz-Bradley, MPH3, Denise M. Hynes, MPH, PhD,
RN7

1
Cancer Center/University of Illinois, Chicago, IL; 2Pediatrics/University of Illinois, Chicago, IL;
3
University of Illinois at Chicago, Chicago, IL; 4Institute for Health Research and Policy,
Chicago, IL; 5Nephrology Division/Department of Medicine/University of Illinois, Chicago, IL;
6
Nephrology Division/Department of Medicine/University of Illinois at Chicago, Chicago, IL;
7
ACADEMIC INTERNAL MEDICINE/UNIVERSITY OF ILLINOIS & US DEPT OF VETERANS AFFAIRS,
CHICAGO, IL

Background: Utilization of thepatient centered medical home (PCMH) model reduces

hospitalizations, emergency room visits, and healthcare costs for some complex chronic
diseases. Yet PCMH has not been examined for chronic kidney disease.

Purpose: In a non-randomized quasi-experimental 18-month intervention study, we

implemented an adaptation of the patient centered medical home for kidney disease (PCMH-
KD) at two urban dialysis centers. PCMH-KD integrated a primary care physician (PCP), nurse
coordinator, pharmacist and community health worker (CHW) to the health care delivery
team. In this analysis, we examined factors associated with PCP visits by adult chronic
hemodialysis (CHD) patients.

Methods: We used logistic and Poisson regression analysis to examine factors contributing to
having at least one visit with one of the study PCPs and the number of PCP visits. Patients
were grouped based on their response to having a PCP or regular doctor and for how long
during the baseline survey.

Results: Among 173 patients included in the analysis, 91 (53%) patients had at least one (1)
visit with the PCMH-KD study PCP. Thirty-two of these patients 32 (35%) reported having a
PCP at baseline which they have seen for six or more months (established PCP) while 59 (65%)
reported no PCP or have a PCP for less than 6 months (no established PCP). Having more visits
with the CHW was a significant predictor for having 1 visit(s) with one of the study PCPs in
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S211

both the established PCP and no established PCP groups. CHD patients with no established
PCP and had greater PCMH-KD PCP visits were females, self-reported Diabetes, used a
medically arranged transportation to dialysis, had a stressful life events in the prior 6 months,
and had a family member of friend involved with their dialysis care.

Conclusions: Both social and clinical factors influenced the use of the PCMH-KD PCP by CHD
patients. This study is the first to examine factors affecting use of primary care by CHD
patients. Understanding these relationships is important in care redesign aimed at improving
outcomes for complex chronic diseases.

CORRESPONDING AUTHOR: Ifeanyi Chukwudozie, MPH, Cancer Center/University of Illinois,

Chicago, IL, 60608; [email protected]
S212 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A114 6:00 PM-7:00 PM

UTILIZATION OF PRIMARY CARE-MENTAL HEALTH INTEGRATION: A PERSPECTIVE FROM

PRIMARY CARE PROVIDERS IN THE VA

Chelsea L. Rothschild, Ph.D.1, Ashley Barroquillo, Psy.D.2, Lucille Carriere, Ph.D.3, Christine
Davis-Hayes, LCSW 4, Stephanie Smith, LCSW 2, Nicole Webb, Ph.D. 2

1
VA Tennessee Valley Healthcare System, Nashville, TN; 2VA Tennessee Valley Heathcare
System, Murfreesboro, TN; 3Tennessee Valley Healthcare System, Murfreesboro, TN; 4VA
Tennessee Valley Heathcare System, MURFREESBORO, TN

Co-located, collaborative care is one component of the Department of Veterans Affairs model
of integrated care. Primary Care-Mental Health Integration (PC-MHI) is a multi-disciplinary
program designed to address the mental health needs of Veterans in the primary care
setting. The study aimed to assess primary care providers perspective on two critical
elements of integration: 1) Level of PC-MHI utilization and 2) potential referral barriers.
Qualitative survey data was collected from Primary Care Clinicians (PCC) (N=22) from two VA
medical centers to evaluate utilization rates and obtain information regarding current
strengths and challenges of providing integrated primary care.

Results indicated the following utilization rates: 27.3% 2-3x/mo., 45.5% 1-2x/wk, 27.3%
3+/wk. PCCs indicated roughly similar preferences for face-to-face, same day referral (68.2%,
)vs. requesting a future PC-MHI appointment (63.6%). PCCs most frequently request PC-MHI
assistance to co-manage anxiety (100%), depression (95.5%), stress management (95.5%), and
cognitive assessment (86.4%).

Regarding prescribing psychotropic medication to co-treat psychiatric conditions, it is notable

that 4.5% were not comfortable prescribing for depression/anxiety, while 22.7% were not
comfortable prescribing specifically for PTSD. PCCs noted minimal barriers to PC-MHI
referral, with the most prevalent being patient refusal (often a barrier 36.4%). PCCs overall
impressions of PC-MHI were favorable, reporting significant benefit to both themselves
(M=9.36, SD= .90) and their Veterans (M=9.27, SD=1.07).
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S213

Preliminary findings provide insight into perspectives toward integrated care and suggest
opportunities to further enhance care. Data indicates lower referral rates for behavioral
health conditions such as chronic pain (45.5%), coping with medical illness (22.7%), and
smoking cessation and weight management (4.5%), and suggests a target for increased PCC
education. Findings also indicate an opportunity to decrease stigma associated with PTSD and
increase PCCs comfort in initiating/managing psychiatric medications with PC-MHI support.
Only 50% of PCCs reported awareness of availability of the depression medication monitoring
protocol. Additionally, efforts must be made to decrease requests for future appointments,
as same-day access is a central goal of the PC-MHI program. Focus on these areas is essential
for ongoing improvement of the co-located, collaborative model of care.

CORRESPONDING AUTHOR: Chelsea L. Rothschild, Ph.D., VA Tennessee Valley Healthcare

System, Nashville, TN, 37212; [email protected]
S214 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A115 6:00 PM-7:00 PM

CHANGES IN RESILIENCE PREDICT FUNCTION IN ADULTS WITH PHYSICAL DISABILITIES: A

LONGITUDINAL STUDY

Karlyn A. Edwards, B.A.

University of New Mexico, Albuquerque, NM

Resilience has been identified as an important factor that may contribute to improved
functioning and overall quality of life. Resilience, defined as the ability to restore or maintain
psychological and physical health in the face of life stressors, may be especially important for
those with physical disabilities, who often experience a multitude of stressors. Little is known
about how resilience may relate to overall physical and psychological health of this
population. Even further, if resilience is shown to positively relate to functioning, there is
limited research examining whether resilience may be a viable target for treatment among
this population. The current study sought to first examine the impact of resilience on four
important measures of functioning (physical functioning, fatigue, sleep quality and
depression) over the period of one year among those with physical disabilities. Second, the
study compared how much resilience varied over one year in relation to other well-
established and supported targets of treatment, such as fatigue, sleep quality, and
depression. The study was comprised of 893 adults that had one of four physical disabilities -
Muscular Dystrophy (18%), Multiple Sclerosis (40%), Spinal Cord Injury (18%), or Post-Polio
Syndrome (24%). The participants were mostly White/Caucasian (90%), female (64%), and
highly educated (81% receiving at least some college). The participants were assessed at two
time points one year apart by mail-in survey. Four separate linear regressions examined if
changes in resilience would predict one of four criterion variables (physical functioning,
fatigue, sleep quality, and depression), while controlling for the baseline score of the criterion
variable, age, sex, and diagnosis. Four Pearson R correlations were also done to examine the
stability of resilience in comparison to depression, fatigue, and sleep quality. Our findings
showed that changes in resilience were significantly associated with changes in physical
functioning, fatigue, sleep quality, and depression over one year, with no differences among
age, sex, or diagnosis group. Our findings also indicated that resilience did show similar test-
retest stability over one year as depression, fatigue, and sleep quality. Our findings support
that resilience is important in maintaining physical and psychological health among those with
physical disabilities. Second, our findings provide support that resilience may be a viable
target for treatment, such as depression, fatigue, and sleep quality are.

Learning Objective 1: Define resilience, and its relation to physical functioning, fatigue, sleep
quality, and depression among those with physical disabilities.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S215

Learning Objective 2: Discuss the variability and/or stability of resilience over time, as well as
its viability as a target for treatment.

CORRESPONDING AUTHOR: Karlyn A. Edwards, B.A. , University of New Mexico, Albuquerque,

NM, 87108; [email protected]
S216 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A116 6:00 PM-7:00 PM

CORRELATES OF THE OCI-R SUB-SCALES AND MISOPHONIA

Shannon Cusack, N/A, Rachel Wallace, N/A, Therese Cash, M.S., Ph.D., Christina Sheerin, M.S.,
Ph.D., Scott Vrana, PhD, LCP

Virginia Commonwealth University, Richmond, VA

Correlates of the OCI-R sub-scales and misophonia

Misophonia is characterized by an aversion to specific sounds, usually made by other people.

It has been proposed as potentially falling within the category of obsessive-compulsive
spectrum disorders (OCSD). Evidence for a relation between misophonia and the OC spectrum
is present in case reports of patients with misophonia endorsing co-morbid OCSDs, including
obsessive-compulsive disorder. Data suggest that misophonia is associated with both
obsessive rumination about the sound and compulsive behavior designed to stop the sound or
mitigate its effects. Additional research is needed to better characterize this association and
begin to explore potential mechanisms underlying misophonia. This study examines the
association of misophonia with six subscales of the Obsessive Compulsive Inventory-Revised
(OCI-R) and a brief misophonia screening question, Compared to other people, are you
sensitive to certain sounds made by other people? in a cross-sectional survey of college
students and community adults.

The current study (n=826, Mage=27.5, 63.6% female) investigates misophonia and its OCI-R
sub-scale correlates. 16.2% of the total sample indicated they regularly experience
misophonia, while 44.7% endorsed misophonia some of the time. Misophonia was
significantly correlated with the full OCI-R (r = .22) as well as the obsessing subscale (r = .231)
and the compulsive checking (r = .157), hoarding (r = .15), neutralizing (r = .15), ordering (r =
.17), and washing (r = .20) subscales (all p Further analyses were conducted to explore the
hypothesis that misophonia would be more strongly correlated with the obsessive vs.
compulsive symptoms. The correlation between misophonia and the obsessing subscale was
stronger than with the checking (z = 2.2, p = 0.03), hoarding (z = 2.3, p = 0.02), neutralizing (z =
2.4, p = 0.02), ordering (z = 1.8, p = 0.07), and washing (z = 1.5, p = 0.13) subscales. These
results expand upon the literature regarding the OCD and misophonia relationship, suggesting
that misophonia is more closely associated with obsessive rumination than compulsive
behavior designed to reduce anxiety. Future research should extend these findings by further
examining obsessive rumination as a potential mechanism of misophonia.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S217

CORRESPONDING AUTHOR: Shannon Cusack, N/A, Virginia Commonwealth University,

Richmond, VA, 23220; [email protected]
S218 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A117 6:00 PM-7:00 PM

CRIME PREVENTION THROUGH SCHOOL RISK BEHAVIOR SURVEILLANCE OF ADOLESCENTS IN A

HIGH CRIME RATE COMMUNITY

Imelda G. Alcala-Snchez, Ph.D.1, Carlos Gonzalo Ibez-Alcal, M.Sc.2, Miguel Conchas-

Ramirez, M.A.1, Manuel B. Gonzalez-Gonzalez, J.Dr.1

1
Instituto Felipe Angeles/Universidad Autonoma de Chihuahua, Chihuahua, Chihuahua,
Mexico; 2Instituto Tecnologico de Monterrey, Chihuahua, Chihuahua, Mexico

The public health model is suitable for preventing risky behavior in adolescents living in high
crime communities. Children and adolescent wellbeing and mental health can be
compromised by high crime environments. Regular or recurrent participation in risky and
antisocial behavior is an indication of lower self-control associated with criminal behavior in
adolescent and youth. Assessment and early detection of individuals participating in risky and
antisocial behaviors can lead to improve crime prevention efforts in communities with high
crime rates. The Youth Risky Behavior Surveillance System (YRBSS), form the USA CCD, is an
effective scholar system to monitor troubled behavior in adolescents. Risk behaviors are
grouped in six categories of priority behaviors contributing to unintentional injury and
violence, undesired pregnancy, sexually transmitted diseases including HIV-AIDS, tobacco,
alcohol and illicit drug consumption, as well as inappropriate eating and sedentary life style.
Mexican schools located in high crime communities may benefit from a monitoring system
assessing high risk behaviors among middle school students. Objetive. To assess the
psychometric properties of a Spanish version of the YRBSS in a sample of 581 middle school
adolescents 11-14 years old, attending one of four middle schools located in high crime
neighbors at the city of Chihuahua, Mexico. Methods: The 34 item transadapted Spanish
version of the YRBSS questionnaire was completed anonymously in a 50 minute classroom
session. Male and female students were recruited voluntarily to participate in the study after
previously obtaining their parents authorization. School teachers collaborated with
questionnaire delivery. Results: Overall Cronbach alpha was .746, with 81.1% of complete
cases. In a principal component Exploratory Factor Analysis with a Varimax rotation and Kaiser
normalization analysis two general factors were identified: a) Addiction and Sex, b) Violence
and Leisure Time. Male students showed significantly higher frequencies of Risk Behaviors
than females. Males endorsed significantly more alcohol consumption, sexual activity and
violence items, than females. More females reported involvement in risky behavior related to
affective reactions and monitoring their weight through exercise, diet and other means, than
males. Despite some non-significant differences, gender trend described hold both with and
without significant differences. Conclusion: YRBSS can be useful to monitor risky behaviors in
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S219

adolescent Mexican students differentiating males from females. Grouping of items into two
sub-scales may reflect a link among behaviors, those relationships deserve further exploration
as it may reflect a common dependence on psychosocial factors present in the studied
sample.

CORRESPONDING AUTHOR: Imelda G. Alcala-Snchez, Ph.D., Instituto Felipe

Angeles/Universidad Autonoma de Chihuahua, Chihuahua, Chihuahua, 31000;
[email protected]
S220 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A118 6:00 PM-7:00 PM

EXAMINING THE POTENTIAL IMPORTANCE OF FREE WILL BELIEF FOR ANXIETY AND
DEPRESSION

Courtney Alderson, BA, Hassan Khuram, NA, Ashlee Sawyer, MS

Virginia Commonwealth University, Richmond, VA

Previous findings from social cognitive research indicate that belief in free will is a factor of
important interpersonal consequence. However, a dearth of research exists regarding the
intrapersonal effects of free will belief. The goal of this study was two-fold: 1) to increase the
knowledge base concerning the intrapersonal effects of believing in free will; and 2) to
investigate the potential importance of this psychological construct for mental health. By
employing lay free will theory to the context of mental health, we hypothesized two very
different routes by which the belief in free will would impact mental healthone route
predicting a negative impact, and the other predicting a positive impact. As for the negative
impact on mental health, we predicted free will belief would be positively associated with
perceived stress and anxiety. Regarding the positive impact on mental health, we predicted
that free will belief would be positively associated with coping self-efficacy and negatively
associated with depression. A cross-sectional survey method was implemented to test the
predicted associations. A sample consisting of 1048 undergraduate college students
completed a survey containing several measures aimed at assessing individuals self-reported
mental health and belief in free will. Bivariate correlations partially supported our predictions.
Per our predictions, free will belief was positively associated with perceived stress and coping
self-efficacy (r = .119, p = .01; r = .222, p = .01, respectively). Contrary to our predictions, no
associations for free will belief were found for anxiety or depression (r = -.017, p > .05; r = -
.043, p > .05, respectively). However, a series of mediation analyses using a bootstrapping
technique provided more nuanced support for our predictions. While free will belief was
found to negatively impact (i.e., increase) both anxiety and depression by way of perceived
stress, it was also found to positively impact (i.e., decrease) anxiety and depression by way of
coping self-efficacy. These results provide preliminary evidence that suggests free will beliefs
have important intrapersonal implications for mental health/psychological wellbeing. While
basic research is needed to more accurately understand these associations, future applied
work should identify ways that free will beliefs could be targeted so as to increase self-efficacy
with healthy coping strategies, as well as reframe the free willperceived stress linkage.

CORRESPONDING AUTHOR: Courtney Alderson, BA, Virginia Commonwealth University,

Richmond, VA, 23219; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S221

A119 6:00 PM-7:00 PM

FINANCIAL WELL-BEING AND DEPRESSIVE SYMPTOMS AMONG OLDER ADULTS

Tucker-Seeley D. Reginald, MA, ScM, ScD1, Gillian L. Marshall, MSW, PhD2

1
Harvard T.H. Chan School of Public Health/ Dana-Farber Cancer Institute, Boston, MA;
2
University of Washington, Tacoma, WA

Financial well-being (FWB) has recently been conceptualized in cancer survivorship research
across three domains: material, psychosocial, and behavioral. The material domain refers to
the lack of financial resources (e.g. material hardship), the psychosocial domain refers to how
one feels about the lack of resources (e.g. financial worry), and the behavioral domain
captures what one does with their financial resources (e.g. saving/consumption behaviors).
Research has consistently shown an association between FWB concepts such as financial
hardship and health across a range of specific population groups such as low-income housing
residents, cancer survivors, and smokers. However, the association between multiple
domains of FWB and depressive symptoms among older adults has yet to be explored. Thus,
it remains unclear whether the various domains of FWB are differentially associated with
mental health among older adults. The purpose of this study was to determine the
association between the material, psychosocial, and behavioral domains of financial well-
being and depressive symptoms. Cross sectional analysis was conducted using data from the
2010 Health and Retirement Study Leave Behind Questionnaire (N=7,407). Multiple items
within each FWB domain were summed to create material, psychosocial, and behavioral FWB
scores, with increasing scores indicating worse FWB. Depressive symptoms were
operationalized using the Center for Epidemiologic Studies Depression (CESD) scale (range 0-
8), with a cut-off CESD score greater than four indicating depression. Multivariable logistic
regression was used to obtain the odds of depression (reporting more than four depressive
symptoms). Our results showed that increasing scores on each of the FWB domains (worse
FWB) was positively associated depression, even after controlling for demographic (age, race,
ethnicity, marital status) and socioeconomic (education, household income) characteristics:
material (Odd ratio(OR)=1.41; 95% Confidence Interval (CI): 1.35, 1.47), psychosocial
(OR=1.17, 95% CI: 1.15, 1.20), and behavioral (OR = 2.09; 95% CI: 1.80, 2.41). Though these
results show a similar pattern of worse FWB positively associated with depression for each
domain, their differential association is important for the further explication of the pathway
between socioeconomic circumstances and mental health among older adults, over and
above traditional measures of socioeconomic status such as education and income. In
addition, such explication provides greater specificity for potential intervention targets where
programs can be designed for improving the material, psychosocial, and/or behavioral
S222 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

domains of FWB to reduce depressive symptoms and overall mental health among older
adults..

CORRESPONDING AUTHOR: Tucker-Seeley D. Reginald, MA, ScM, ScD, Harvard T.H. Chan
School of Public Health/ Dana-Farber Cancer Institute, Boston, MA, 02215;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S223

A120 6:00 PM-7:00 PM

FREQUENCY AND PREDICTORS OF DISTRESS IN A DIVERSE CANCER SAMPLE

Laura M. Perry, BS1, Ashley Beggin, N/A2, Michael Hoerger, PhD, MSCR2

1
Tulane University, New Orlenas, LA; 2Tulane University, New Orleans, LA

Title: Frequency and Predictors of Distress in a Diverse Cancer Sample

Objective: Determining the prevalence of distress and its predictors among patients with
cancer is important in developing and implementing tailored interventions to help reduce
these symptoms. Recent studies addressing these issues have reported results from samples
that are majority female and white. The purpose of this study was to examine the frequency
and demographic and health predictors of distress in an ethnically diverse sample of patients
with multiple cancer diagnoses.

Method: A sample of 1,616 individuals with a cancer diagnosis who visited a cancer center in
2015 responded to items on distress, demographics, and health history. Distress was
evaluated using the Distress Thermometer. Logistic regression (odds ratios) was used to
determine the demographic and health variables associated with clinically significant distress
(Distress Thermometer score 4).

Results: Participants were relatively equally distributed across gender (female = 47%; male =
53%), and the majority of participants were either White (49%) or African American (47%).
They had a range of cancer diagnoses, with the most frequent being prostate cancer (27%),
hematological cancer (25%), and breast cancer (24%). Clinically significant distress was
present in 19.5% of the sample. The variables associated with increased risk for distress
included female gender (OR = 1.56, p = .027), younger age ( < 65 years old; OR = 2.38, p <
.001), having lung/bronchus cancer (OR = 1.94, p = .049) or head/neck cancer (OR = 2.36, p =
.023), presence of metastases (OR = 1.81, p = .003), and presence of symptom burden (OR =
1.75, p < .001).
S224 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusion: The prevalence of distress in this sample was 19.5%. Significant predictors of
distress included certain demographic (gender and age) , diagnostic (lung/bronchus and
head/neck cancer), and health (metastases and symptom burden) variables. This study helps
to generalize findings on the prevalence and predictors of distress in cancer to a more diverse
population of patients with cancer.

CORRESPONDING AUTHOR: Laura M. Perry, BS, Tulane University, New Orlenas, LA, 70118;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S225

A121 6:00 PM-7:00 PM

INTERPERSONAL VIOLENCE HISTORY AND CHILDBIRTH-SPECIFIC POSTTRAUMATIC STRESS

SYMPTOMS

Meghan Sharp, M.A., Melissa Decker, M.A., Heather Littleton, Ph.D., Christyn Dolbier, Ph.D.

East Carolina University, Greenville, NC

Introduction: Childbirth can be a traumatic event for some women due to risks of death or
serious injury to the baby and/or mother during birth. Prior research has supported that a
history of childhood sexual abuse (CSA; sexual contact by a relative or authority figure before
age 14) is associated with elevated risk of childbirth-related PTSD symptoms. CSA survivors
may be at greater risk to experience trauma symptoms due in part to deficits in emotion
regulation and adaptive coping, as well as pre-existing elevations in psychological distress,
potentially including PTSD. However, prior work has not examined the extent to which
childbirth-specific PTSD relates to other experiences of interpersonal violence including adult
sexual assault (ASA), childhood physical abuse (CPA), and adult physical intimate partner
violence (IPV). The current study evaluated CSA, CPA, ASA, and IPV as predictors of childbirth-
specific PTSD symptomology in the postpartum period.

Method: Participants were 159 women who gave birth to a living child in the past four months
recruited via social media for an online survey of pregnancy and childbirth experiences (M=29
years; majority White, married, college-educated, primaparous). A total of 30.8% reported
prior sexual or physical abuse or assault.

Results: CPA survivors reported higher PTSD arousal symptoms (n = 15; M = 10.6) than both
IPV survivors (n = 35; M = 5.55) and women with no PA history (n = 39; M = 5.3), F (3,91) =
3.14, p = .03, 2 = .10. SA survivors exhibited no differences in PTSD symptomology according
to timing of abuse or compared with women with no SA history.

Conclusions: Results highlight potential differences in postpartum trauma symptomology with

timing and type of interpersonal trauma. Unexpectedly, CSA survivors did not report greater
severity of PTSD, and CPA survivors reported greater symptoms of arousal only. The high SES
of the sample may have affected results, as women of low SES may be more likely to
experience elevated distress, more life stressors prior to birth, as well as more traumatic
events during birth. Future research should examine effects of trauma duration and severity,
as well as the cumulative effects of multiple traumas on risk for childbirth-related PTSD in
larger samples of mothers. Finally, as trajectories of PTSD symptomology differ, future study
should examine differences in PTSD symptoms over time.
S226 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Meghan Sharp, M.A., East Carolina University, Greenville, NC,
27858; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S227

A122 6:00 PM-7:00 PM

SEDENTARY BEHAVIOR AND POST-TRAUMATIC STRESS DISORDER SYMPTOMS IN ACUTE

CORONARY SYNDROME SURVIVORS

Deanna Zhu, B.A.1, Ian Kronish, MD, MPH2, Donald Edmondson, PhD, MPH3, Joseph Schwartz,
PhD1, Keith M. Diaz, PhD1

1
Center for Behavioral Cardiovascular Health, Columbia University Medical Center, New York,
NY; 2Columbia University Medical Center, Center for Behavioral Cardiovascular Health, New
York, NY; 3Columbia University Medical Center, New York, NY

Introduction: Many survivors of acute coronary syndrome (ACS) develop post-traumatic stress
disorder (PTSD). Patients with ACS-induced PTSD have twice the risk of recurrent cardiac
events; however the contributing mechanisms have not been elucidated. The purpose of this
study was to examine the patterns of sedentary behavior (an emerging and potent
cardiovascular risk factor) among ACS survivors with and without PTSD symptoms in the one
month following hospital discharge.

Methods: Hospitalized patients with confirmed ACS (n=50, 24% female, mean age: 62 10
years) were invited to wear an Actical accelerometer on their non-dominant wrist for 30-days
post discharge, after which they completed the PTSD CheckList (PCL) tailored to those who
incurred a cardiac-related event. Patients were classified as those reporting (54%) and not
reporting (46%) any PTSD symptoms. Total sedentary time was quantified as the number of
min/day spent sedentary for each week (e.g. weeks 1-4) post-hospitalization.

Results: In multivariable adjusted analyses, there was a significant main effect of PTSD
symptoms on levels of sedentary time in weeks 1-4 post-hospitalization (F=4.2, p=0.046).
Total sedentary time was significantly higher among patients reporting PTSD symptoms
compared to those without PTSD symptoms in week 1 post-hospitalization (PTSD symptoms:
824 37 vs. No PTSD symptoms: 700 40 min/day; p=0.028). A trend for higher sedentary
time in patients with PTSD symptoms was also observed in week 2 (PTSD symptoms: 787 42
vs. No PTSD symptoms: 675 45 min/day; p=0.078), week 3 (PTSD symptoms: 766 42 vs. No
PTSD symptoms: 657 45 min/day; p=0.088) and week 4 (PTSD symptoms: 741 42 vs. No
PTSD symptoms: 630 45 min/day; p=0.082) post-hospitalization.

Conclusion: ACS survivors reporting cardiac-related PTSD symptoms had greater sedentary
time in the four weeks following hospitalization than those not reporting any PTSD symptoms.
S228 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Future research should consider examining the role of sedentary behavior as a risk factor for
recurrent cardiac events in ACS survivors who develop PTSD symptoms.

CORRESPONDING AUTHOR: Deanna Zhu, B.A., Center for Behavioral Cardiovascular Health,
Columbia University Medical Center, New York, NY, 10032; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S229

A123 6:00 PM-7:00 PM

THE DEVELOPMENT AND INITIAL EVALUATION OF A CALL CENTER FOR CONCERNED FAMILY
MEMBERS OF MILITARY VETERANS

Jennifer M. Picanso, B.A., MSW candidate

Department of Veterans Affairs, Los Angeles, CA

The Development and Initial Evaluation of a Call Center for Concerned Family Members of
Military Veterans

Military Veterans with mental health issues can often struggle in recognizing that they need
care or are silenced by the stigma associated with reaching out for help. Family and friends
are the first to identify that care is needed but feel powerless in convincing the Veteran to
seek mental health care. Coaching Into Care is a program developed by the Department of
Veterans Affairs to educate and provide support to people surrounding the Veteran to
promote help-seeking behavior. Since the programs inception in 2011, call volume has risen
to over 3000 incoming yearly calls. The most frequent caller was a female spouse or parent of
a Veteran calling about a male Veteran, and about 50% of those Veterans had deployed to
Iraq or Afghanistan. The majority of problems identified in our 9 month examination cohort
(79.9%) concerned a psychosocial issue, such as psychiatric/behavioral problem, medical
problems, and questions about VA services. Fifty-eight percent of calls were referred by first-
line call responders to a professional coach for targeted intervention. Those receiving the
extended telephone-based coaching service were provided approximately 17 calls over 6
months. Among those callers in our examination cohort receiving coaching (N = 165), about
31% of Veteran were engaged in some type of mental health care; 6 months of coaching calls
was associated with an almost 50% increase in Veterans receiving mental health treatment
(up to 48%), which was statistically significant. Overall, 76.2% of the callers reported obtaining
the service they hoped for, 76.9% were more hopeful, and 85.2% felt more prepared to
the help the Veteran. Coaching Into Care provides a unique clinical service to families and
friends of Veterans who: 1) have not yet engaged in care; 2) who need additional mental
health services; or 3) are minimally, or sporadically, engaged in mental health treatment.

CORRESPONDING AUTHOR: Jennifer M. Picanso, B.A., MSW candidate, Department of

Veterans Affairs, Los Angeles, CA, 90073; [email protected]
S230 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A124 6:00 PM-7:00 PM

TRAUMA CORRELATES OF DECREASED SOUND TOLERANCE CONDITIONS

Rachel Wallace, B.S. Psychology1, Shaina Gulin, B.S. Psychology; M.S. Clinical Psychology2,
Therese Cash, M.S., Ph.D.2, Christina Sheerin, M.S., Ph.D.2, Danielle Dick, Ph. D.3, Kenneth
Kendler, M.D.4, Scott Vrana, Ph. D.5

1
Clinical Psychology/Department of Psychology/Virginia Commonwealth University,
Richmond, VA; 2Virginia Commonwealth University, Richmond, VA; 3Devlopmental and
Clinical Psychology; Virginia Commonwealth University, Richmond, VA; 4Virginia Institute for
Psychiatric and Behavioral Genetics, Richmond, VA; 5Virginia Commonwealth University
Departments of Psychology and Psychiatry, Richmond, VA

Decreased sound tolerance (DST) conditions are emerging clinical conditions in behavioral
medicine. Misophonia is an extreme negative emotional response to specific sounds (e.g.,
people chewing, swallowing). Hyperacusis involves high sensitivity to sounds below normal
sound sensitivity thresholds. Both often co-occur with tinnitus (the perception of ringing or
buzzing in the ears). Although research on DSTs has found that they are related to anxiety
sensitivity and obsessive-compulsive traits, little is known about their associations with
traumatic life events and posttraumatic stress symptoms (PTSS). This study examines the
association of DSTs with past year traumatic events in a longitudinal health and genetics
survey study of undergraduates. The present work (n=975, Mage=20.9, 68.6% female)
examines DST conditions and their trauma correlates. 21.6% of students endorsed tinnitus,
31.4% endorsed misophonia, and 6.9% endorsed hyperacusis. Rates of endorsement for
trauma experiences were as follows: 6.4%, natural disaster; 13.4%, transportation accident;
6.4%, physical assault; 3.0%, sexual assault; 13.5%, another unwanted/uncomfortable sexual
experience. Chi-square analyses demonstrated that tinnitus was associated with physical
assault [2(1)=7.0, p < .01], natural disaster [2(1)=4.9, p < .05], transportation accident
[2(1)=4.5, p < .05], and other unwanted/uncomfortable sexual experience [2(1)=5.9, p <
.05]. Hyperacusis was related to other unwanted or uncomfortable sexual experience
[2(1)=9.6, p < .01], while misophonia was not related to any traumatic events. Of those who
experienced a traumatic event in the past year, 41.0% (n=114) reported experiencing
posttrauma-related symptoms. Chi-square analyses revealed a significant relationship
between PTSS and hyperacusis [2(1)=5.0, p < .05], but not with tinnitus or misophonia. In this
large, normative university student sample, DSTs and traumatic life events were prevalent.
Tinnitus and hyperacusis were associated with past traumatic experiences, whereas
misophonia was not. Trauma-related symptoms were prevalent and associated with
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S231

hyperacusis. Future research should extend these findings on trauma and DST conditions by
studying PTSS symptoms more comprehensively and examining the mechanisms by which
PTSS correlates with hyperacusis.

CORRESPONDING AUTHOR: Rachel Wallace, B.S. Psychology, Clinical Psychology/Department

of Psychology/Virginia Commonwealth University, Richmond, VA, 23220;
[email protected]
S232 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A125 6:00 PM-7:00 PM

ASSESSING FOR REACTIVITY TO AN ECOLOGICAL MOMENTARY ASSESSMENT PROTOCOL

USING DATA FROM INDIVIDUALS RECOVERING FROM ADDICTION

Nisha C. Gottfredson, PhD

UNC Gillings School of Global Public Health, Chapel Hill, NC

Ecological momentary assessment (EMA) methodology permits inference about within-person

processes in real life setting. Compared with cross-sectional designs, EMA enables more
objective measurement of temporal associations between variables without relying on
participants to introspect or recall this information. Nevertheless, repeated assessment may
introduce a new challenge: participant reactivity.

EMA designs may draw a participants attention to a nonconscious process, resulting in

behavior change. Alternatively, participants might fall in to a habitual pattern of responding
after several assessments. Using an example of negative affect and substance use,
introspective behavior change might result in lower rates of substance use or a weaker
association between negative affect and substance use over time. Habitual responding might
result in reduced within-person variation over time, which would thereby reduce the
magnitude of predictor-outcome associations.

Previous research on participant reactivity to EMA methods has explored whether mean
levels of the outcome changed during the study; most of this research has found little
evidence of reactivity (Collins et al, 1998; Farchaus et al, 2003; Hoffman, 2007; Hufford et al,
2002; Nishina, 2012; Stone et al, 2002). No previous study considered changes in predictor
prevalence, changes in variation of responding, or bivariate changes in predictor-outcome
associations. Failure to recognize and account for participant reactivity may result in biased
inference. On the other hand, identifying protocols that lead to participant reactivity has the
potential provide valuable information about mechanisms for behavior change.

I outline how to assess for different types of participant reactivity and I discuss options for
modeling data in the presence of reactivity through a combination of data visualization, semi-
parametric modeling (Time-Varying Effect Model, Li et al, 2015), and parametric modeling. N =
126 individuals receiving treatment for addiction received four EMA surveys per day for 7 days
in two bursts spaced 6 weeks apart. Surveys assessed participant negative affect (NA) and
substance use (SU). By using a measurement burst design, I can parse true change over time
from artefactual change resulting from the EMA design. Variation in NA decreases over the 7
days in a burst and the bivariate association peaks mid-week. Looking at stable change across
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S233

bursts, SU increases but the bivariate association between NA and SU does not change.

In addition to presenting methods for assessment, I suggest methods for limiting reactivity in
EMA protocol design.

CORRESPONDING AUTHOR: Nisha C. Gottfredson, PhD, UNC Gillings School of Global Public
Health, Chapel Hill, NC, 27599; [email protected]
S234 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A126 6:00 PM-7:00 PM

DO INJUNCTIVE AND DESCRIPTIVE NORMATIVE BELIEFS NEED AN OUTCOME EVALUATION IN

VALUE EXPECTANCY MODELS?

Paul Branscum, PhD, RD, Maria Collado-Rivera, MS, CHES

The University of Oklahoma, Norman, OK

Value-expectancy models are common place in health behavior research and practice, and
posit that behaviors are determined by antecedents (ex. attitudes), which are formed by an
individuals belief that a behavior is likely to lead to certain outcomes (expectancy), and the
value one places each outcome (value). A popular value-expectancy model in social and
behavioral health is the Theory of Planned Behavior. One construct of this model, perceived
norms, consists of two types of normative pressure: injunctive norms (the perception that
significant referents want you to behave in a certain way) and descriptive norms (the
perception that a behavior is normal for others like yourself). Each construct follows the
value-expectancy approach for measurement. Injunctive norms are formed by a combination
of injunctive normative beliefs and motivation to comply, and descriptive norms are formed
by a combination of descriptive normative beliefs and identification with each referent. Some
evidence has shown that measuring the value construct (motivation to comply) in conjunction
with injunctive normative beliefs does little for improving the utility of the TPB, and some
have called for its removal. Concurrently, no study has investigated whether measuring the
value construct (identification with referents) in conjunction with descriptive normative
beliefs improves the utility of the TPB. Therefore, the purpose of this study was to evaluate
the value-added of measuring injunctive and descriptive norms using a value-expectancy
model, compared to an expectancy only based model. Overweight and obese adults (n=410)
at a weight-loss clinic in a southwestern city completed an TPB survey related to consuming
no sugary beverages in the following 6-months. Direct and indirect measures of injunctive
and descriptive norms were evaluated. Each expectancy-value pair (model 1) and expectancy-
only measure (model 2) was correlated with the composite direct measures of either
construct. Results showed little difference between methods. In model 1, all expectancy-
value pairs were significant (p < 0.001) for injunctive norms [spouse (r=0.50), friends (r=0.45),
children (r=0.45), and parents (r=0.49)], but only spouse (r=0.17; p < 0.01), was significantly
associated with descriptive norms. In model 2, the associations were slightly improved, but
mostly unchanged. For injunctive norms all correlations were significant (p < 0.001) [spouse
(r=0.50), friends (r=0.48), children (r=0.50), and parents (r=0.53)], and for descriptive norms
only spouse (r=0.21; p < 0.001), was significantly associated. Results suggest that there was
little to no value added to having the value based measure in this value-expectancy model.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S235

Future researchers should consider dropping these measures to reduce respondent burden,
or find alternative ways of measuring these constructs.

CORRESPONDING AUTHOR: Paul Branscum, PhD, RD, The University of Oklahoma, Norman,
OK, 73071; [email protected]
S236 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A127 6:00 PM-7:00 PM

IMPROVING DATA QUALITY OF HEALTH SURVEYS THROUGH WEB-BASED EXPERIMENTS

Stephanie Fowler, PhD, MPH, Gordon Willis, PhD, Richard Moser, PhD, Rebecca Ferrer, PhD,
David Berrigan, PhD, MPH

National Cancer Institute, Bethesda, MD

Health surveys are vital for surveillance of health behaviors. Importantly, the quality of data
obtained from these surveys is crucial for drawing accurate conclusions about population
health and informing interventions. Survey experiments and repeated measures designs allow
researchers to assess validity and reliability. However, health surveys are seldom tested
comprehensively because it is costly and time-consuming. Given that large cohort studies and
surveillance efforts often rely on survey data, more efficient approaches to testing are
needed. The present investigation examined the utility and efficiency of using an online,
crowd-sourcing platform as a means for conducting validity and reliability testing of health
surveys. We used Amazon.coms Mechanical Turk, an online labor market, as our testing
platform. The study incorporated both a between-subjects manipulation of question order to
assess order effects, and a within-subjects repeated measure of the survey to assess test-
retest reliability. At Time 1, respondents were randomly assigned to one of two experimental
conditions in which the order of questions on walking behavior, and perceptions of the
walking environment, was varied. Note these questions were used in the 2015 National
Health Interview Survey. At Time 2, four weeks later, respondents were invited to complete
the survey again in the same order to assess test-retest reliability. The 1,446 respondents at
Time 1 represented a wide range of demographic characteristics that roughly matched those
of the U.S. adult population: 56.2% male, 84% Non-White Hispanic, 50% with a Bachelors
degree minimum, Mage = 31.8, SDage = 10.6. The 960 respondents who participated again at
Time 2 allowed us to assess item test-retest reliability for the newly developed questions
pertaining to perceptions of the walking environment. These questions demonstrated
moderate to high test-retest reliability overall (Kappa = .48-.84). Further, consistent with
previous findings on question ordering and walking behaviors, we obtained a significant order
effect such that respondents first answering questions on perceptions of the walking
environment had lower odds of transportation walking than did respondents first asked about
walking behavior: OR = 0.61 (95% CI 0.46-0.75), p = .02. Given that physical activity questions,
like walking, are often embedded among a larger set of health behavior questions, an efficient
and cost-effective platform for conducting survey experiments can help increase the
confidence in findings obtained from health surveys. Success of this rapid (~5 week), low-cost
experiment ($800), and feasiability of conducting repeated measures, supports further use of
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S237

online platforms as vehicles for the conduct of survey experiments and reliability testing to
improve the quality of data obtained from health surveys.

CORRESPONDING AUTHOR: Stephanie Fowler, PhD, MPH, National Cancer Institute, Bethesda,
MD, 20850; [email protected]
S238 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A128 6:00 PM-7:00 PM

PSYCHOMETRIC EVALUATION OF THE RELAPSE SITUATION EFFICACY QUESTIONNAIRE AMONG

ADULTS SEEKING WEIGHT LOSS TREATMENT

Yang Yu, MSN, Jeffrey Rohay, PhD, MSIS, Susan Sereika, PhD, Ran Sun, MSN, Jacob Kariuki,
PhD, Lora Burke, PhD, MPH

University of Pittsburgh, Pittsburgh, PA

Background: Abstinence self-efficacy (ASE) has been identified as a key construct in relapse
prevention of addictive behaviors; however, no scale existed for relapse following intentional
weight loss. We adapted an existing scale measuring ASE among smokers and used the
Relapse Situation Efficacy Questionnaire (RSEQ) to measure individuals confidence in staying
on their weight maintenance plan. Objective: The psychometric properties of the RSEQ were
assessed, including internal consistency reliability, convergent validity, predictive validity and
construct validity. Methods: In a 12-mo weight loss maintenance study, the RSEQ was
administered at 0, 6 and 12 mos. Convergent validity was assessed at baseline by using
Spearman's correlation between RSEQ and Weight Efficacy Lifestyle questionnaire; predictive
validity was examined by the Spearman's correlation between 6-mo RSEQ and 12-mo
measures of adherence to physical activity goal, adherence to calorie/fat goal and body
weight; construct validity was assessed via using principal component analysis (PCA) with
Promax rotation. Results: The sample (n=148) was mostly female (90.5%) and white (81.1%)
with a mean body mass index of 34.1 + 4.6 kg/m2. The RSEQ showed good internal
consistency (Cronbachs = .81 to .95) and convergent validity (r=.71). The 6-mo RSEQ was
positively associated with 12-mo adherence to calorie goal (r=.23) and 12-mo adherence to fat
goal (r=.22); 6-mo RSEQ was negatively associated with 12-mo body weight (r= -.25). PCA
suggested that the 31 items can be factored into 6 components: negative emotion (8 items),
positive emotion (6 items), social-food occasions (5 items), distracting situations (5 items),
work-leisure activities (5 items) and physical negative situation (2 items), accounting for 67.5%
of the total item variance. All items had loaded primarily on one factors, with factor loadings
>.40, and all factors were moderately to highly positively correlated (r= .25 to .58).
Conclusion: This study provides preliminary support for the reliability and validity of the
RSEQ. Future work needs to add more items to the factor physical negative situation and to
examine RSEQ in other studies with larger and more diverse samples.

CORRESPONDING AUTHOR: Yang Yu, MSN, University of Pittsburgh, Pittsburgh, PA, 15213;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S239

A129 6:00 PM-7:00 PM

RRR-DERIVED INDEX OF DIETARY BEHAVIORS IS STRONGER CORRELATE OF REPEATED

MEASURES OF OBESITY

Wendy E. Barrington, PhD, MPH, Shirley A.A. Beresford, PhD

University of Washington, Seattle, WA

Background: Identification of reliable dietary measures of obesity-risk is needed to facilitate

evaluation of prevention efforts. Given that key research supporting specific dietary
recommendations has recently come into question, continued use of data-driven analytic
techniques in nutritional epidemiology is warranted. Traditional derivation of dietary indices
has involved factor analysis of nutrients to maximize exposure variance. Yet, dietary behaviors
like intake of fast food, fruits and vegetables, and soft-drinks have each been associated with
obesity and may be more easily and reliably measured than nutrient intake in larger studies.
Using statistical methods that maximize the variance in obesity as a response may also
produce a more salient dietary factor. The purpose of our study was to identify a dietary
behavior index most related to obesity.

Methods: The Promoting Activity and Changes in Eating (PACE) study is a group-randomized
weight gain prevention trial among 34 worksites in the Seattle Metro Area. Behavioral data
and body mass index (BMI) were collected among all participating employees at baseline
(2005-2007) and follow-up (2007-2009) while measured waist circumference (WC) was
collected on a random subsample (n=622 at baseline). Factor, partial least squares, and
reduced rank regression (RRR) analyses were conducted among a random test half of these
individuals without missing data at baseline. Confirmatory analyses were performed within
the validation half at baseline and among all those with follow-up data. Multilevel linear
models included adjustment for covariates and clustering within worksites.

Results: Fast food, fried potato, and soft-drink intake were the most obesogenic behaviors
using RRR analyses with BMI and WC as response variables. Each quartile increase in dietary
index score was associated with a 5% higher geometric mean BMI (Ratio= 1.053; 95% CI:
1.031, 1.075) and approximate 4% higher geometric mean WC (1.036; 95% CI: 1.019, 1.054) at
baseline adjusted for covariates. Results were similar at follow-up and remained statistically
significant for WC after baseline adjustment.

Conclusions: Our findings, although small in magnitude, may still have significant public health
implications given associations were demonstrated among all men and women in the study
for relatively low frequencies of reported dietary behaviors. These findings may have
S240 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

application for identifying those at risk of central obesity in clinical and community settings
and may inform intervention strategies.

CORRESPONDING AUTHOR: Wendy E. Barrington, PhD, MPH, University of Washington,

Seattle, WA, 98195; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S241

A130 6:00 PM-7:00 PM

STRUCTURAL VALIDITY AND MEASUREMENT INVARIANCE OF A SPANISH LANGUAGE

TRANSLATION OF THE PEARLIN MASTERY SCALE

Janna Gordon, M.S. 1, Vanessa L. Malcarne, Ph.D.2, Scott Roesch, Ph.D2, Richard Roetzheim,
MD3, Kristen Wells, Ph.D., MPH2

1
SDSU/UC San Diego Joint Doctoral Program in Clinical Psychology, San Diego, CA; 2San Diego
State University, San Diego, CA; 3U. of South Florida, Tampa, FL

The Pearlin Mastery Scale is a seven-item scale frequently used in psychological research to
assess individuals personal mastery, or the extent to which they believe they are in control of
their own life and fate. Since its development in the 1970s, the Mastery Scale has been
utilized in diverse samples, including caregivers, psychiatric populations, and individuals with
chronic illnesses. It has also been adapted from English into multiple languages, including
Spanish. However, to date, no published studies have examined the psychometric properties
of Spanish translations of the Mastery Scale. This analysis evaluated the structural validity and
measurement invariance of a Spanish translation of the Mastery Scale in two samples of
Spanish-speaking individuals receiving healthcare at community clinics in Florida. Sample 1 (N
= 63) was comprised of adults enrolled in a study investigating the relationships among stress,
depression, and glycemic control in low-income patients with type 2 diabetes. Participants in
Sample 2 (N = 82) were recruited as part of the Patient Navigation Research Program and had
received an abnormal screening test for breast or colorectal cancer. Confirmatory factor
analysis (CFA) and multiple group CFA were conducted both on the full seven-item scale as
well as a five-item version commonly utilized in the literature. CFA indicated that the seven-
item and five-item scales yielded two-factor and one-factor solutions, respectively, for the
combined sample of Spanish-speaking patients. However, although the specified factor
solutions fit each individual sample separately, multiple group CFA revealed that the Mastery
Scale items did not load equivalently on the specified factors across samples. This suggests
that the Spanish version of the Mastery Scale may not measure mastery consistently across
groups and that researchers should use this version with caution, especially with Spanish-
speaking participants from diverse cultural backgrounds.

CORRESPONDING AUTHOR: Janna Gordon, M.S. , SDSU/UC San Diego Joint Doctoral Program
in Clinical Psychology, San Diego, CA, 92120; [email protected]
S242 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A131 6:00 PM-7:00 PM

THE CROSSOVER EFFECT OF PREDICTING INTENTIONS WITH THEORETICAL ANTECEDENTS OF

VEGETABLE SUBGROUP CONSUMPTION

Paul Branscum, PhD, RD1, Valerie Senkowski, MS in Health Promotion2

1
The University of Oklahoma, Norman, OK; 2University of Oklahoma, Norman, OK

Oftentimes health behaviors are in fact behavioral categories that comprise of a number of
interrelated behaviors that meet a common recommendation. For example, while the USDA
has a daily recommendation for vegetable consumption (2.5-3.0 cups), this recommendation
can be met by consuming a variety of vegetables, which in turn, can be viewed as
independent behaviors. Another set of recommendations the USDA published is that
individuals should vary their intake of vegetables by consuming from five subgroups each
week (beans and peas, starchy, orange and red, green, and other). An advantage to studying
sub-behaviors of a behavioral category, it that the results can provide targeted information,
that can help shape future theory-based interventions. When sub-behaviors are studied
together, a cross-over effect can also be evaluated, through which one sub-behavior (ex.
starchy vegetables) can be examined by theory-based antecedents of the other interrelated
behaviors (i.e. attitudes towards eating beans and peas). The purpose of this study was to
examine the crossover effect for predicting each vegetable subgroup, with theoretical
antecedents from all vegetable subgroup categories. The Integrative Model of Behavioral
Prediction provided the theoretical underpinnings of this study. Undergraduate students
(n=386) completed an online survey evaluating each vegetable subgroup, and its theoretical
antecedents (attitudes, perceived norms, and perceived behavioral control (PBC)). Two
stepwise multiple linear regression models were run for each vegetable subgroup. The first
(base) model used its own behavioral antecedents to predict intentions. The second
(omnibus) model included behavioral antecedents from all vegetable subgroups to predict
intentions. Results from the first set of models indicated that attitudes, perceived norms and
PBC predicted a significant amount of variance of intentions (40.5% to 54.6%) for each
vegetable subgroup. Results from the second set of models showed that adding antecedents
from the interrelated behaviors improved the amount of variance explained for intentions
(41.7% to 56.1%) for each vegetable subgroup, but the improvements were minimal (added
5.8% to 1.2% to each model). Overall, the IM was found to be a useful framework for
predicting the intentions of vegetable subgroup consumption. When testing the crossover
effect of behavioral antecedents, while all models improved to some extent, improvements
were minimal, and may not carry much statistical relevance.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S243

CORRESPONDING AUTHOR: Paul Branscum, PhD, RD, The University of Oklahoma, Norman,
OK, 73071; [email protected]
S244 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A132 6:00 PM-7:00 PM

A GENDER- AND CULTURALLY-SENSITIVE WEIGHT LOSS INTERVENTION FOR HISPANIC MALES

David O. Garcia, PhD1, Luis Valdez, MPH2, Benjamin Aceves, II, MPH, MA 3, David Campas, B.S.
Public Health 4, Julio Loya, BSN, RN, CMSRN5, Melanie D. Hingle, PhD, MPH, RD3, Melanie L.
Bell, PhD3, Steven P. Hooker, Ph.D.6, Marylyn M. McEwen, PhD, RN, FAAN3

1
University of Arizona, Mel and Enid Zuckerman College of Public Health, Tucson, AZ; 2Mel
and Enid Zuckerman College of Public Health: University of Arizona, Tucson, AZ; 3University of
Arizona, Tucson, AZ; 4University of Arizona Health Promotion Research Assistant, Tucson, AZ;
5
University of Missouri - Columbia, Marana, AZ; 6Arizona State University, Phoenix, AZ

Background: Hispanic males have the highest rates of overweight and obesity when
compared to males of other racial/ethnic groups. While weight loss can significantly reduce
obesity related health risks, Hispanic males are significantly underrepresented in weight loss
research. Our work addresses a critical gap by informing how tailored intervention strategies
improve weight management in this health disparate population. Purpose: This study
compares a 12 week gender- and culturally-sensitive weight loss intervention (GCSWL) to a
waist-list control (WLC) in sedentary, overweight/obese Hispanic males. Methods: Fifty
Hispanic males (age: 42.1 12.7 years; BMI: 34.1 5.3 kg/m; 48% Spanish speaking only)
were randomized to one of two groups: GCSWL (n=25) or WLC (n=25). GCSWL participants
attended weekly in-person individual sessions guided by a trained bilingual Hispanic male
lifestyle coach, were prescribed a daily reduced calorie goal with a specific focus on
reducing/modifying the types of food and liquids consumed (e.g., alcohol/sugar sweetened
beverages), and 225 minutes of moderate-intensity physical activity (PA) per week. Additional
GCSWL features included a free gym membership and optional spouse/significant other
attendance at intervention sessions. The WLC were asked to maintain their usual dietary
intake and PA habits during the 12 weeks. Results below represent preliminary data for
participants who have reached their 6-week assessment time point to date (n=31).
Results: At week 6, attendance at individual GCSWL sessions is approximately 80%. The
overall attrition rate is 9.7%, with 28 of 31 participants completing 6-week assessments
(GCSWL: n=1, 6.3% vs. WLC: n=2, 13.3%). The observed pre-post mean weight loss for the
completers analysis in the GCSWL is (n=15; -2.6 2.1 kg) compared to the WLC (n=13; 0.6
1.3 kg) (difference = -3.2, 95% CI -4.6, -1.8, p < 0.001) and intention-to-treat (GCSWL: -2.4
2.2 kg; WLC: 0.5 1.2 kg) (difference = -3.2, 95% CI -4.2, -1.6, p < 0.001). It is anticipated 12-
week data collection for all 50 men will be completed December 2016. Conclusion: GCSWL
participants are engaged in intervention sessions, attrition is low, and short-term weight loss
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S245

has been achieved. The intervention appears to be a feasible strategy to engage Hispanic
males in weight loss/management. However, confirmation of these findings is needed upon
study completion.

CORRESPONDING AUTHOR: David O. Garcia, PhD, University of Arizona, Mel and Enid
Zuckerman College of Public Health, Tucson, AZ, 85750; [email protected]
S246 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A133 6:00 PM-7:00 PM

ASSOCIATIONS BETWEEN PARENTING FACTORS AND WEIGHT-STATUS IN OVERWEIGHT

AFRICAN AMERICAN ADOLESCENTS

Lauren Huffman, MA, Dawn K. Wilson, Ph.D., FSBM

University of South Carolina, Columbia, SC

Close to 40% of African American adolescents are overweight nationally. However, there is
limited research examining the relationship between authoritative parenting style (high
nurturance and monitoring) and body mass index (BMI) in overweight, African American
adolescents. The current study expands on previous literature by examining associations
between parenting style, parenting behaviors specific to diet and physical activity (PA)
including social support, autonomy support, home environmental supports, and modeling on
predicting youth BMI. Analyses utilized baseline data from African American adolescents (N =
148; Mage = 13.5 years; 66% female; MBMI% = 96.5) and their caregivers (Mage = 43.4 years;
94% female; MBMI = 37.3) enrolled in the Families Improving Together (FIT) for Weight Loss
randomized controlled trial. Parenting factors were measured using self-report surveys and
daily minutes of PA were measured using 7-day accelerometry estimates. Child zBMI was
calculated from objective height and weight measurements and standardized using CDC
growth reference curves. Regression analyses indicated that the overall model was significant
(F(15, 132) = 2.38, p < 0.05; R2 = 0.22) with parenting variables accounting for significant
variance in predicting BMI beyond covariates (child age, sex, diet, PA, and family income;
R2=0.17, pB= -0.09, SE= 0.05) and parent BMI (weight-modeling) was positively associated
with higher adolescent zBMI (B= 0.02, SE= 0.01). No other parenting effects specific to health
behaviors were significant. Overall, these results suggest that general authoritative parenting
and healthy weight modeling may be more closely associated with weight status compared to
parenting behaviors specific to diet or PA in overweight, African American youth. This study is
cross-sectional and future research is needed to more specifically test longitudinal and causal
relationships.

CORRESPONDING AUTHOR: Lauren Huffman, MA, University of South Carolina, Columbia, SC,
29201; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S247

A134 6:00 PM-7:00 PM

COMMUNICATION BETWEEN TREATMENT-SEEKING ADULTS WITH OBESITY AND THEIR

HEALTHCARE PROVIDERS

Jocelyn E. Remmert, B.A., Amani D. Piers, BA, Meghan L. Butryn, PhD

Drexel University, Philadelphia, PA

Introduction: Little is known on how obese treatment seeking adults interact with their
primary care providers (PCPs) around weight control. These interactions could be a catalyst to
lifestyle modification, but we do not know if, or how, positive weight control interactions are
occurring in primary care. This study surveyed obese treatment seeking adults recently
enrolled in a lifestyle modification program to gather information on these interactions, in
hopes of identifying future targets for improved treatment engagement. Methods: Eighty-five
adults enrolled in a large scale research obesity treatment trial were administered a survey
upon enrollment to assess provider-participant interactions. The survey included Likert scale
questions, where 1=strongly disagree and 7=strongly agree. Results: Two-thirds (64.7%) of
participants spoke about weight loss to their PCPs prior to enrolling in this study; the other
third (35.3%) never spoke about weight control with their PCPs. Of the participants who did
speak to their PCPs about weight loss, 53.7% spoke about weight loss at 2-5 previous
appointments, and 27.8% only spoke about weight loss at 1 previous appointment. Advice
given by PCPs was: lose weight (n=19), attend structured weight loss program (n=15),
monitor food intake (n=14), calorie counting/set a weight loss goal (n=11), exercise (n=11),
and other options (see nutritionist, bariatric surgery, no white food) (n=10). On average,
participants indicated they agreed to strongly agreed (6.55, SD=0.79) with the statement: I
feel comfortable talking to my provider about my weight. Participants on average indicated
they somewhat agreed (5.13, SD=1.47) with the statement: I feel like my healthcare provider
understands my barriers to weight control. Conclusions: Overall the results suggest that
there is variability in communication between treatment-seeking adults with obesity and their
PCPs. While approximately one-third of participants had multiple conversations about weight
control with their PCPs, one-third never spoke about weight control with their PCP.
Understanding barriers and facilitators of communication will be important for future
research. The results also suggest that providers may need additional support and resources
in order to make evidence-based recommendations when discussing weight with their
patients.

CORRESPONDING AUTHOR: Jocelyn E. Remmert, B.A., Drexel University, Philadelphia, PA,

19104; [email protected]
S248 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A135 6:00 PM-7:00 PM

INCREASING EFFECT OF EXECUTIVE FUNCTIONING ON WEIGHT OUTCOMES THROUGHOUT

BEHAVIORAL WEIGHT LOSS AND FOLLOW-UP

Emily Wyckoff, B.A.1, Stephanie Manasse, MS2, Talea Cornelius, M.A., M.S.W.3, Meghan L.
Butryn, PhD2, Evan M. Forman, PhD2

1
University of Connecticut, Storrs, CT; 2Drexel University, Philadelphia, PA; 3University of
Connecticut, Vernon, CT

Group behavioral weight loss (BWL) is the current gold standard weight loss treatment;
however, weight regain is normative and long-term outcomes are poor. Due to the decline in
clinician contact and support as BWL treatment concludes, executive functioning (cognitive
control processes that allow one to carry out goal directed behavior) may play an increasingly
important role in maintaining adherence to dietary and physical activity recommendations.
The current study tested the hypothesis that inhibitory control (i.e. ones ability to regulate
automatic behavioral responses) and planning (i.e., generating mental representations of
steps to achieve an intention) would positively predict weight loss in BWL treatment and
follow-up. Additionally, we examined whether executive functions were increasingly
predictive of outcome across BWL treatment phases and follow-up. It was hypothesized that
poor executive functioning would increasingly predict poorer weight outcomes as group
contact lessened and at follow-up.

Overweight and obese participants (n=190) received 25 group BWL treatment sessions over a
12-month period. At baseline, participants completed behavioral measures of planning (D-
KEFS Tower Test) and inhibitory control (D-KEFS Color-Word interference Test). Assessment
weights were collected at months 0 (baseline), 6 (mid-treatment), 12 (end-of-treatment), and
24 (follow-up).

Two multilevel models were run examining the effects of 1) inhibitory control and 2) planning
on weight change. Both models revealed a negative, main effect of time on weight, ps < .001,
as well as a quadratic effect of time, ps < .001. Participants had an initial decrease in weight,
but this effect eventually tapered and reversed. Inhibitory control and planning interacted the
quadratic effect of time on weight (inhibitory control: B = -1.38, t(444.88) = 3.34, p < .001;
planning: B = .30, t(444.20) = 3.03, p < .01) such that those with poor inhibitory control and
poor planning had similar weight loss months 0-6, lost more weight between months 6 and
12, and had slower regain months 12-24, compared to participants with better inhibitory
control and planning ability.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S249

Contrary to the hypothesis, results suggest that poor inhibitory control and planning are
predictors of better outcomes in BWL. These findings could indicate that BWL may be most
efficacious in the long term for those whose weight control challenges include poor planning
and self-control, where as BWL may not address factors (e.g. unmeasured variables or even
metabolic/biological factors) contributing to weight regain in those with better executive
functioning. The somewhat counterintuitive results of this study highlight the need for
treatment customization based on individual characteristics (e.g. executive functioning) as
BWL does not consistently produce the enduring behavioral changes needed for weight loss
maintenance.

CORRESPONDING AUTHOR: Emily Wyckoff, B.A., University of Connecticut, Storrs, CT, 06269;
[email protected]
S250 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A136 6:00 PM-7:00 PM

INDIVIDUALS WITH OBESITY DID NOT ENGAGE IN GREATER EMOTIONAL EATING THAN THEIR
LEAN COUNTERPARTS IN A LABORATORY SETTING

Laura Aylward, B.S.1, Kristin Schneider, Ph.D.2, Christine May, PhD3, Sherry Pagoto, PhD4

1
Rosalind Franklin University of Medicine & Science, Chicago, IL; 2Rosalind Franklin University
of Medicine & Science, North Chicago, IL; 3University of Massachusetts Medical School,
Fiskdale, MA; 4University of Massachusetts Medical School, Worcester, MA

Background: The literature on the relationship between emotional eating and body mass
index (BMI) is mixed, where some studies support an association between emotional eating
and BMI and some do not. Many studies have examined emotional eating via self-report
questionnaires, but less have examined emotional eating in a laboratory setting. The present
study compared individuals with obese and lean BMI on difference in caloric intake during
negative and neutral mood induction conditions. Greater caloric intake in negative mood
condition versus neutral mood condition was considered emotional eating.

Methods: Participants (n=61; 74% female, 39% obese) underwent neutral, anxiety, and anger
mood inductions in counterbalanced order and received 2400 kcal of a variety of palatable
foods immediately following. They were left alone for 20 minutes and calories consumed was
measured via weighing leftovers. Participants with BMI of 18.5 to 25 were considered lean
and 30 and above were considered obese. Emotional eating was measured as the difference
in caloric intake between the negative and neutral mood conditions.

Results: Both mood inductions were found to significantly impact mood (anxiety p < .001 and
anger p < .001) while the neutral mood induction had no impact on negative mood (p>.05).
The condition by BMI interactions on food intake were not significant for the anxiety, F(1,
177)=0.032, p=.857, or anger conditions, F(1, 177)=0.326, p=.569.

Discussion: Lean and obese individuals did not differ in emotional eating, as defined by
greater calories consumed following anxiety or anger induction relative to a neutral condition.
Results are consistent with other studies reporting no relationship between emotional eating
and BMI. Vulnerability to emotional eating is likely due to factors other than BMI.

CORRESPONDING AUTHOR: Laura Aylward, B.S., Rosalind Franklin University of Medicine &
Science, Chicago, IL, 60640; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S251

A137 6:00 PM-7:00 PM

PARENTS MISPERCEIVE THE WEIGHT STATUS OF GIRLS WHO ARE OVERWEIGHT OR OBESE

Daniella Carucci, B.A., M.S.1, Sierra Bainter, Ph.D.2, Mary Kate Clennan, B.S.2, Marissa D. Alert,
M.S.2, Shannon Chiles, M.S3, Qingyun Liu, N/A2, Patrice G. Saab, Ph.D.2

1
University of Miami, Miami, FL; 2University of Miami, Coral Gables, FL; 3University of Miami,
Louisville, KY

Introduction:

Research suggests that parental misperception of childs weight status is a predictor of

childhood obesity. Parents underestimation of their childs weight status can inhibit
regulation of health behaviors and lead to less favorable health outcomes for children. The
purpose of this study was to investigate whether parental perception of daughters weight
status differed from actual weight status.

Methods:

The participants were a diverse sample of 267 middle school aged girls (M = 11.8 years).
Approximately 88% of the participants were ethnic minorities and 56% came from
communities with a poverty rate of 20% or higher, where the median family income was less
than 80% of the median family income for the state. The participants' BMI was calculated
after measuring height and weight. Approximately 3% of the girls were slightly or very
underweight, 56% were average weight, 21% were overweight, and 20% were obese. Parental
perception of weight status was determined by their response to the question Describe your
childs weight. Response options were very underweight, slightly underweight, about
the right weight, slightly overweight, or obese. Because of the small number of girls
rated as underweight, the underweight categories were combined in the analyses.
Approximately 9% of the parents reported their child as slightly or very underweight, 66%
reported average weight, 21% reported slightly overweight, and 4% reported obese.

Results:

Chi-square tests of independence were performed to determine whether parent perception

of childs weight status differed from the childs actual weight status. The results indicated
that there were differences in parent perception and the childs actual weight status, c 2 (9, N
= 267) = 206.58, p < .001. In fact, of the 57 girls in the overweight weight status, 64.3% were
S252 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

underestimated as average by their parents. Of the 55 girls in the obese weight status,
60.4% were underestimated as overweight.

Conclusion:

Understanding weight perception and identifying strategies to correct parents misperception

may inform interventions targeting weight management efforts for overweight and obese
children. As weight misperception is often influenced by cultural norms, early identification
and interpretation of actual weight status may improve health outcomes for ethnically diverse
young girls. Further research on the predictors of parental weight misperception is warranted
to better serve these at-risk populations. Moreover, interventionists and clinicians should
promote discussion of appropriate weight status between parents and children in efforts to
adjust their perceptions and motivate weight-related behavior change.

CORRESPONDING AUTHOR: Daniella Carucci, B.A., M.S., University of Miami, Miami, FL,
33186; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S253

A138 6:00 PM-7:00 PM

PREFERRED WEIGHT TERMS AND WEIGHT DISSATISFACTION: A STUDY OF BLACK WOMEN IN

THE SHAPE TRIAL

Christina M. Hopkins, B.S., Dori M. Steinberg, PhD, RD, Gary G. Bennett, PhD

Duke University, Durham, NC

Background: Obesity is not sufficiently treated in primary care. Only 20% of obese patients
receive weight loss counseling, and uncertainty about patients preferred terms for obesity
may thwart treatment. No studies have investigated preferences of black women, a
population that experiences disproportionately high rates of obesity.

Objective: To examine desirability of commonly used weight terms in a previously unexplored

population of low-income Black women, and to identify the impact of weight satisfaction.

Method: Participants were 176 low-income Black women, involved in a weight gain
prevention study (Mage = 35.5 5.5 yr, MBMI = 30.2 2.5 kg/m2). At baseline, commonly used
weight terms were ranked for desirability on a 5-point Likert scale (1 = very desirable, 5 = very
undesirable). Weight satisfaction was assessed on a 4-point Likert scale (1 = very satisfied, 4 =
very unsatisfied). Participants were grouped by baseline weight satisfaction and group
differences were analyzed using t-tests.

Results: More euphemistic terms (e.g., weight) were rated as neutral (M=2.81.2); weight
was rated as very undesirable by 9.7% of participants. Conversely, more blunt terms (e.g.,
fatness) were rated as very undesirable (M=3.81.2) by 32.8% of women. On average, no
weight-related term was rated as desirable.

Most participants (72.3%) reported weight dissatisfaction. Harsher terms were rated
significantly less favorably by women who were unsatisfied with their weight (M=3.91.2),
relative to women who were more weight satisfied (M=3.41.1, p < .005). Additionally,
women with BMI > 34 kg/m2 rated the harsher terms as more palatable than their peers in
lower weight classes (ps < .05).

Conclusion: Euphemisms are preferred by patients, but differences by weight satisfaction and
weight class warrant consideration. Women with weight dissatisfaction gave poorer ratings to
the harsher weight-related terms. Conversely, women in the highest BMI class rated these
terms more favorably, suggesting that women most in need of weight loss intervention are
most able to tolerate terms that are currently seen as stigmatizing. Future research should
S254 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

examine clinical effectiveness of the full range of weight-related terms to maximize impact for
behavior change among a population most in need of treatment.

CORRESPONDING AUTHOR: Christina M. Hopkins, B.S., Duke University, Durham, NC, 27705;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S255

A139 6:00 PM-7:00 PM

RECRUITING LOW INCOME POST-PARTUM WOMEN INTO A WEIGHT LOSS TRIAL: IN-PERSON
VERSUS FACEBOOK DELIVERY

Valerie J. Silfee, PhD, Stephenie Lemon, PhD, Milagros C. Rosal, PhD

University of Massachusetts Medical School, Worcester, MA

Several studies, such as the Diabetes Prevention Program (DPP), have provided foundational
evidence for the effect of lifestyle modification interventions on diabetes prevention and
weight loss. However, translating these programs to the real-world has proven difficult. There
remains a need to increase the feasibility and reach of translational weight loss interventions,
particularly among socioeconomically disadvantaged populations. PURPOSE: To compare the
recruitment rate of overweight low income postpartum women into a DPP-adapted
behavioral weight loss program delivered in-person versus delivered via Facebook. We
hypothesized that the Facebook-based trial may be easier to recruit for as the delivery mode
may reduce participant burden (i.e. transportation, child care) compared to the in-person
trial. METHODS: We compared two 8-week pilot behavioral weight loss trials; one delivered
via weekly in-person group sessions and the other delivered entirely via Facebook. Both trials
used the same recruitment methods: participants were overweight low income postpartum
women within five Women Infants and Children (WIC) clinics in the Worcester, Massachusetts
area recruited by nutritionists during routine WIC visits. Inclusion criteria included, childbirth
in the previous 6 weeks to 6 months, age 18 or older, a body mass index (BMI) of > 27 kg/m2,
and obstetric provider approval for participation in the diet and physical activity components
of the intervention. Additional eligibility criteria for the Facebook intervention included: 1)
ability to use the Internet daily; 2) having a Facebook account; 3) currently using Facebook at
least once per week. RESULTS: 27 and 29 women participated in the in-person and Facebook
pilot trials, respectively. Among eligible women, 62.1% gave permission to be contacted for
the in-person trial, and 23.3% enrolled. 58.7% of women eligible for the Facebook pilot agreed
to be contacted, and 24% enrolled. CONCLUSIONS: Recruitment rates for a Facebook-based
weight loss intervention were similar to that of an in-person intervention, thus both programs
may have the same reach. Future efforts are needed to increase reach of weight loss
interventions among overweight and obese low income diverse women.

CORRESPONDING AUTHOR: Valerie J. Silfee, PhD, University of Massachusetts Medical School,

Worcester, MA, 01655; [email protected]
S256 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A140 6:00 PM-7:00 PM

REST ETHIC: THE INFLUENCE OF BELIEFS ABOUT REST IN OVERWEIGHT BLACK WOMEN

T Jack Brooks, BA, Dori M. Steinberg, PhD, RD, Gary G. Bennett, PhD

Duke University, Durham, NC

Background Few Americans meet national physical activity guidelines. However, an expanding
body of research suggests that sedentary time may confer additional health risks even among
those who meet or exceed MVPA guidelines. We investigate the hypothesis that people who
have positive attitudes about the health-promoting value of rest might prioritize rest over
exercise for good health. Relatively little is known about how such attitudes toward rest affect
patterns of physical activity and sedentarism. Objective This study investigates associations
between rest ethic beliefs about the health-promoting value of sedentary behaviors
and weight and physical activity. We also examined possible interactions with
sociodemographic characteristics. Methods Data were collected during an 18-month weight
gain prevention intervention (the Shape Program) conducted in North Carolina community
health centers. Participants were 194 premenopausal black women diagnosed as overweight
or obese. Rest ethic was measured by survey questions about the importance and benefits of
resting, administered at baseline and 12 months. Weight was collected at baseline, 6, 12 and
18 months. Participants wore accelerometers for 14 days at baseline and 12
months. Results At baseline, participants had a mean BMI of 30.2. 68.5% of the sample
reported that rest was somewhat/very likely to relieve stress. 88% reported that regular rest
was somewhat/very important. There was no significant difference in beliefs about rest
between treatment arms, and beliefs were overall stable from baseline to 12 months for both
intervention and usual care arms. Belief about the importance of rest predicted weight
change at 6, 12 and 18 months, with greater mean weight loss for participants who thought
rest was not very important (x=-2.23, -2.55, -4.19 lbs respectively) than for those who said
rest was somewhat/very important (x=-0.01, .18, .59 lbs respectively), p < .05. Demographic
characteristics and accelerometer-measured physical activity were not significantly associated
with rest beliefs. Conclusions Participants with high rest ethic particularly those who
believed rest was important for health were more likely to to gain weight during the study,
while those with low rest ethic tended to lose weight. Thus, rest ethic may be an important
consideration in both weight loss and weight gain prevention.

CORRESPONDING AUTHOR: T Jack Brooks, BA, Duke University, Durham, NC, 27710;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S257

A141 6:00 PM-7:00 PM

RISK OF OBESITY-ASSOCIATED DISEASES BY BMI AND WAIST CIRCUMFERENCE AMONG

CHILDBEARING RURAL WOMEN

Leigh Ann Simmons, Ph.D., M.F.T.1, Devon Noonan, PhD2, Tiffany Bice-Wigington, LMSW,
PhD3, Jewel Scott, MSN, FNP-C1, Richard Sloane, MPH4

1
Duke University School of Nursing, Durham, NC; 2Duke University, Durham, NC; 3Tarleton
State University, Little Elm, TX; 4Duke University Center for the Study of Aging, Durham, NC

Objective: Compared to their urban counterparts, rural women in the United States continue
to experience significant disparities in rates of preventable chronic diseases and associated
morbidity and mortality. Moreover, weight gain associated with childbearing has been
identified as an independent risk factor for obesity-associated diseases, including type 2
diabetes, hypertension, and cardiovascular disease. The purpose of this study was to examine
obesity-associated disease risk among childbearing rural women, including demographic and
health correlates.

Data and Methods: We analyzed data from Rural Families Speak about Health, a multi-state,
epidemiologic study of rural women and their families living in 11 states (N=444). Eligible
women had at least one child aged 13 years or younger. Data were collected via computer-
assisted, interviewer-administered questionnaires. Participants were classified according to
the National Heart, Lung, and Blood Institutes risk classifications for type 2 diabetes,
hypertension, and cardiovascular disease, which uses combined body mass index (BMI) and
waist circumference as indicators of normal, increased, high, very high, and extremely high
risk. Associations between disease risk status and demographic and health correlates were
explored.

Results: The majority of participants were overweight (31.1%) or obese (45.3%); 19.8% were
class I obese, 14% were class II obese, and 11.5% were class III obese, all exceeding national
obesity rates for women. Despite a mean age of just 32.0 years, the sample had significant risk
for obesity-associated diseases. Based on combined BMI and central adiposity indicators,
27.1% had high, 33.1% had very high, and 11.5% had extremely high risk. Factors associated
with elevated disease risk included less than a high school education, part-time work,
unemployment, and depression. There were no differences in risk status based on having a
regular provider or receiving preventive care in the last year.

Conclusions: Risk for obesity-related diseases is high among childbearing rural women and
S258 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

occurs well before the midlife years. Intervening in early adulthood and outside traditional
clinical venues will be critical to reducing rural health disparities. Given that childbearing is
frequently cited as a cause of weight gain for adult women, interventions during pregnancy as
well as the postpartum and internatal periods, may provide the greatest benefit to improve
the health status of rural female Americans.

CORRESPONDING AUTHOR: Leigh Ann Simmons, Ph.D., M.F.T., Duke University School of
Nursing, Durham, NC, 27710-3322; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S259

A142 6:00 PM-7:00 PM

SECONDARY BENEFITS OF FAMILIES IMPROVING TOGETHER WEIGHT LOSS TRIAL ON MENTAL

AND SOCIAL WELLBEING IN AFRICAN AMERICAN YOUTH

Haylee Gause, BS1, Dawn K. Wilson, Ph.D., FSBM1, Heather Kitzman-Ulrich, PhD2, Brittany
Haney, BA3, Kayla Ford, N/A1

1
University of South Carolina, Columbia, SC; 2Baylor Scott & White Health, Dallas, TX;
3
University of South Carolina, West Columbia, SC

The Families Improving Together (FIT) for weight loss trial is a randomized controlled trial to
test a positive parenting and motivational intervention on promoting weight loss in
overweight African American adolescents. While the intervention focuses on decreasing
weight there may be cascading effects on other secondary benefits for adolescents and their
families. These secondary benefits could improve both mental and social wellbeing beyond
the health benefits of losing weight. The purpose of this study was to examine the potential
improvements beyond weight loss to increase understanding of the programs impact on
participants overall wellbeing. This was accomplished through the development of an
evaluation tool using qualitative methods. Data was obtained from African American
adolescents (N = 148; Mage = 13.5 years; 66% female; MBMI% = 96.5) and their caregivers
(Mage = 43.4 years; 94% female; MBMI = 37.3) who took part in the FIT trial. An evaluation tool
was created to quantitatively evaluate familys qualitative responses during the final week of
the program in which they provided testimonials of their successes and challenges. A top-
down approach was used to identify themes based on Social Cognitive Theory in a total of 9
different intervention groups of Project FIT that have been completed to date. Transcriptions
from each intervention session were developed and coded for themes related to health,
social, and cognitive benefits including but not limited to family support, positive
communication, autonomy, self-efficacy, self-regulation, and monitoring strategies for weight,
physical activity, and diet behaviors. The inter-rater reliability was 0.80. The majority of
themes addressed cognitive benefits related to improving self-regulation and skills for relapse
prevention. Families also reported increases in social support and positive communications
between youth and their parents. This evaluation tool may be used in future research to
understand how weight loss programs lead to secondary benefits for individuals and their
families and provides preliminary support for the ripple effects of weight loss programs on
improving both cognitive and social wellbeing in underserved African American youth.
S260 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Haylee Gause, BS, University of South Carolina, Columbia, SC,
29204; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S261

A143 6:00 PM-7:00 PM

TITLE: AN APPETITE AWARENESS INTERVENTION FOR AT RISK YOUNG WOMEN: A

RANDOMIZED CONTROLLED TRIAL

Lara LaCaille, PhD1, Claire L. Guidinger, BA1, Lauren Vieaux, MA2, Rick LaCaille, PhD1, Brooke
Collins, BS3, Courtney Clark, N/A1, Maggie Mason, NA4

1
University of Minnesota Duluth, Duluth, MN; 2University of Minnesota Duluth, Little Canada,
MN; 3University of Minnesota Duluth, Superior, WI; 4University of Minnesota- Duluth, Buffalo,
MN

Weight gain among college freshmen women is a common concern, and research suggests
that this is a time of risk for the development of unhealthy eating patterns and weight change.
The current randomized clinical study evaluates the efficacy of a brief intervention aimed at
increasing an individuals ability to eat intuitively based on bodily hunger and satiety cues,
thereby potentially preventing weight gain and disordered eating. A pilot study was
conducted with 34 college freshman women in Fall of 2015 and was replicated with a larger
sample in Fall 2016. In total, 129 women were randomly assigned to one of three conditions:
Appetite Awareness Training (AAT; n=45), Nutrition Education (NE; n=42), or a no-treatment
control (NTC; n=42). Treatment groups received four group sessions over 6 weeks. Outcomes
assessed at baseline, post-intervention, and 4-month follow up included: weight, BMI, % body
fat, waist circumference, measures of self-efficacy to manage weight and to resist eating in a
variety of contexts, intuitive and restrictive eating patterns, and overeating frequency.
Repeated measures ANOVAs, with follow up contrast tests, were conducted to explore group
differences over time. Data from the pilot study indicated that both the AAT and NE groups
lost weight, whereas the control group gained weight [p=.05, 2=.17 (large effect)]. Both
treatment groups also reported a greater increase in weight maintenance self-efficacy than
the NTC [p=.04, 2=.18 (large effect)]. There was a smaller effect for ability to resist eating
across a variety of contexts, though the AAT group showed the greatest increase in control in
situations where food was highly available. Preliminary evidence suggests that brief group
interventions were effective compared to no intervention. It is anticipated that the AAT group
may be more likely to sustain weight gain prevention in the long term. Post-treatment and 4-
month follow up data from the larger replication study will be presented, along with 1-year
follow up data from subjects in the pilot study. Implications and evaluation of interventions
will also be discussed.
S262 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Lara LaCaille, PhD, University of Minnesota Duluth, Duluth, MN,
55812; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S263

A144 6:00 PM-7:00 PM

NATURALLY-OCCURRING PAIN CONVERSATIONS WITHIN THREE CLINICAL SAMPLES AS AN

ALTERNATIVE MEASURE TO SELF-REPORTED PAIN

Robert C. Wright, MA, Megan Robbins, PhD

University of California, Riverside, Riverside, CA

Pain is a complex, subjective experience which is difficult to fully measure with a simple self-
report. Because pain is not confined to a medical visit and unfolds through patients daily
experiences, it is paramount to access the pain experience with methods beyond traditional
measures. To explore how patients naturally discuss pain, three clinical groups, rheumatoid
arthritis (RA; n=13), breast cancer (BC; n=52), and major depressive disorder (MDD; n=29),
wore the Electronically Activated Recorder (EAR) over a weekend. The EAR captured snippets
of ambient sound, totaling in roughly 10% of the patients waking day. The patients audio
recordings were transcribed. A pain-specific Linguistic Inquiry and Word Count dictionary was
developed and validated, and then used to identify patients pain-related conversations. Self-
reported pain measures were completed just prior to the recording weekend. As predicted,
the proportion of pain conversations to total conversations was highest in the RA group (11%;
95% CI [-5.8, 28.64]), then BC (9%; 95% CI %, [1.41, 17.19]), and MDD (5%; 95% CI [-0.02,
10.48]) groups. Correlational analysis showed a trend between the proportion of pain
conversations in the RA group and their reported pain in the past week (r = .53, p=.09) and a
marginal relationship to pain in the past month (r = .43, p=.14). The other two groups showed
small, nonsignificant positive relationships to self-reported pain (rs < .20), suggesting little
overlap between self-reported pain and mentions of pain in daily conversation. Interestingly,
in an analysis of the types of words used, patients used more cognitive processing words (e.g.
cause, know, ought), in pain versus non-pain conversations, t(81) = 3.60, p < 0.01, d = .398.
The linguistic differences in the types of language used in pain and non-pain conversations
may indicate an active form of coping such as reappraisal. Notably, this study was one of the
first to show patients naturally discuss pain in their everyday lives. Ultimately, this study adds
insights into how patients discuss pain without being prompted by a researcher or clinician.
The insights gained from this research illuminate important paths to improving the pain
experience.

CORRESPONDING AUTHOR: Robert C. Wright, MA, University of California, Riverside,

Riverside, CA, 92506; [email protected]
S264 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A145 6:00 PM-7:00 PM

PAIN CATASTROPHIZING AND COGNITIVE FUSION IN RELATION TO PAIN AND UPPER

EXTREMITY FUNCTION AMONG HAND SURGERY PATIENTS

Sezai Ozkan, MD1, Emily Zale, MS1, David Ring, MD2, Ana-Maria Vranceanu, PhD1

1
Massachusetts General Hospital/Harvard Medical School, Boston, MA; 2The University of
Texas at Austin, Dell Medical School, Austin, TX

Objective: Patients who present to hand surgery practices are at increased risk of
psychological distress, pain, and disability. Catastrophic thinking about pain is associated with
greater pain intensity, and initial evidences suggest that, together, catastrophizing and
cognitive fusion (i.e., the tendency to interpret thoughts as true or representative of reality)
are associated with poorer pain outcomes. We tested whether cognitive fusion and
catastrophizing interact in relation to pain and upper extremity limitations among patients
seeking care from a hand surgeon.

Methods: Patients (N = 110; Mage = 47.51; 59% women) presenting to an outpatient hand
surgery practice completed computerized measures of sociodemographics, pain intensity,
cognitive fusion, pain catastrophizing, and upper extremity limitations.

Results: ANCOVA revealed an interaction between cognitive fusion and catastrophic thinking
with respect to pain intensity (p < .01). Participants who scored high on both cognitive fusion
and catastrophic thinking reported the greatest levels of pain relative to those who scored
high on a single measure. The interaction between cognitive fusion and catastrophizing was
also associated with upper extremity limitations (p < .01) with the greatest levels of upper
extremity limitations among those who scored high on both catastrophic thinking and
cognitive fusion. A similar pattern of results was observed when we tested each
catastrophizing subscale individually.

Conclusion: Pain intensity and upper extremity limitations were poorest among hand surgery
patients who reported high levels of both catastrophic thinking and cognitive fusion.
Maladaptive cognitions about pain (i.e., catastrophic thinking) may be particularly problematic
when interpreted as representative of reality (i.e., cognitive fusion). Psychosocial
interventions that address catastrophic thinking and cognitive fusion concurrently merit
investigation for alleviation of pain and limitations among people with arm illness.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S265

CORRESPONDING AUTHOR: Sezai Ozkan, MD, Massachusetts General Hospital/Harvard

Medical School, Boston, MA, 02114; [email protected]
S266 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A146 6:00 PM-7:00 PM

SEX AND GENDER DIFFERENCES IN CHRONIC PAIN-RELATED DISABILITY

Mindy L. McEntee, M.S., Kevin E. Vowles, PhD

University of New Mexico, Albuquerque, NM

Sex and gender differences in chronic pain are nuanced and contextual. However, the extent
of these differences and the degree to which they impact assessment remains unclear. In
2007, the International Association for the Study of Pain put forth recommendations
suggesting that while the same general measures should be used to assess chronic pain in
men and women, additional research was needed to determine whether these assessments
need to be further tailored by sex or gender. To date, however, it remains unclear whether
current measures of functioning perform adequately for both men and women. The present
study: 1) evaluated a newly abbreviated measure of functioning developed for chronic pain
patients, the SIP-CP in a sample of 808 men and women seeking treatment for their chronic
pain and 2) examined sex differences in chronic pain disability, pain-related acceptance, and
pain-related anxiety. Items of the SIP-CP were entered as dependent variables into a
confirmatory factor analysis (CFA) to verify the proposed two factor structure (Physical and
Psychosocial disability). Measurement invariance of the SIP-CP was then tested at configural,
metric, and scalar levels for men and women using WLSMV estimation with a probit link and
THETA parameterization to estimate all models, with model fit statistics describing the fit of
the polychoric correlation matrix among items for each group. Sex differences on disability
subscales, pain-related acceptance, and pain-related anxiety were examined using one-way
ANOVAs.The CFA supported the two factor structure of the SIP-CP with adequate fit across
indices (RMSEA= 0.35, 90% CI [0.32, 0.38], CFI=0.92, and TLI=0.92). Individual items accounted
for 26%-68% of the variance in physical disability and 16%-64% of the variance in psychosocial
disability. Measurement invariance was supported at the configural and metric levels,
suggesting that the physical and psychosocial disability scores of the SIP-CP adequately assess
the same construct in both men and women. Sex differences were found for the Physical
disability domain score and the social interaction subscale score (both p < .001) of the SIP-CP,
with men reporting greater disability. While there is evidence that sex and gender influence
chronic pain, additional research is needed to clarify the extent and overall impact of these
differences on assessment and treatment. At present, the SIP-CP appears to be a valid
measure assessing the same construct of chronic pain for both men and women.

CORRESPONDING AUTHOR: Mindy L. McEntee, M.S., University of New Mexico, Albuquerque,

NM, 87120; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S267

A147 6:00 PM-7:00 PM

THE EFFECTS OF DURATION OF SYMPTOMS ON HEALTH STATUS DEPENDS ON SELF-EFFICACY

LEVELS

Mathew M. Mansoor, Bachelor of Arts in Psychology1, Symone McKinnon, Bachelors of Arts in

Psychology2, Charles Van Liew, MA3, Lauren E. McKinley, Student1, Terry A. Cronan, Ph.D.1

1
San Diego State University, San Diego, CA; 2San Diego State University, La Mesa, CA; 3Grand
Canyon University, San Diego, CA

Fibromyalgia Syndrome (FMS) is a chronic musculoskeletal pain disorder characterized by

symptoms of fatigue, stiffness, depression, and diminished physical functioning. Age and
duration of symptoms predict the health status of FMS patients. However, researchers have
not examined whether the effects of duration of symptoms on health status depends on the
patientsself-efficacy levels. The purpose of the present study was to examine these
relationships. Participants were 573 women diagnosed with FMS, recruited from a health
maintenance organization who were part of a larger intervention study. Data were analyzed
using the baseline measures. Age and symptom length were collected using a demographic
questionnaire. Self-efficacy (SE) and health status were measured via the Arthritis Self-efficacy
Scale, modified for FMS, and the Fibromyalgia Impact Questionnaire (FIQ). A linear regression
was performed to assess the effects of age, symptom length, and SE on FIQ scores. All
predictors were centered. An interaction term was included to test whether the effect of
symptom length on FIQ depended on SE levels. The overall model was statistically significant,
F(4, 566) = 89.33, p < .001, and accounted for 38.7% of the variance in FIQ scores. The main
effects of age, B = -.2895, p < .001, and SE, B = -.5319, p < .001, were significant, as was the
interaction of symptom length with SE, B = -.0046, p = .04. The main effect of symptom length
was not significant when the interaction was included in the model, because its moderation
by SE was qualitative (i.e., the direction of the effect of symptom length changed across SE
levels). The main effects of age and SE demonstrated that, as individuals aged and had higher
SE scores, they had lower FIQ scores (i.e., less FMS impact).The interaction of symptom length
by SE showed that longer duration of symptoms predicted greater FMS impact for those who
have low self-efficacy, but for those who have high self-efficacy, longer symptoms predicted
less FMS impact. These findings indicate that interventions designed to increase self-efficacy
may reduce the effect of FMS as patients age.

CORRESPONDING AUTHOR: Mathew M. Mansoor, Bachelor of Arts in Psychology, San Diego

State University, San Diego, CA, 92182; [email protected]
S268 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A148 6:00 PM-7:00 PM

THE MEDIATING ROLE OF SOCIAL SUPPORT IN THE RELATIONSHIP BETWEEN MINDFULNESS

AND CHRONIC PAIN : A LONGITUDINAL EVIDENCE

Shin Ye Kim, Ph.D., Monika Gaspar, MA

Texas Tech University, Lubbock, TX

More than 100 million Americans suffer from Chronic Pain (CP). This is four times more than
people with Diabetes and 10 times more than Cancer patients (Institute of Medicine Report,
2010). As CP is a widespread and stressful condition, it also has a significant impact on
patients intrapersonal (e.g., depression, anxiety) and interpersonal (e.g., social isolation,
relational satisfaction) functioning (Breivik et al. 2006).

In understanding CP from a biopsychosocial framework, prior research has identified benefits

of both dispositional mindfulness (Turner et al., 2016) and the role of social support (e.g.,
Evers, et al., 2003) across younger, middle age, and, to a lesser extent, older adults with CP.
Curiously, few studies have examined the postulated links between CP and mindfulness or
between the role of social support and adjusting to a CP condition, and no study to date has
directly tested the possible mediating effect of social support in the relationship between
ones engagement in mindful living and CP conditions among older adults; therefore, the
extent and nature of these theorized relations is unclear. One purpose of this study was to
extend the current understanding of the relation between mindfulness living and CP
conditions among older adults by examining antecedent-consequent relations of mindfulness,
social support, and CP.

The present study utilized two waves of data collected 9 to 10 years apart from 3294
participants of the Midlife Development in the United States Survey (MIDUS) (M age = 56.27
years, SD=12.4; 55.4% female; 88.7% White). We applied (multigroup) change score to two
waves of data. Dispositional mindfulness was measured by the MIDUS-II Mindfulness scale,
which is based on Langer and Moldoveanus (2000) conception of mindfulness. Social support
was assessed with 24 items, 8 of which asked about the social support and social strain
associated with each of three sources, spouses/partner, family, and friends. CP conditions
were assessed with participants self-report of whether pain disorder was diagnosed by
physician, pain medication use frequency, duration of CP, and degree to which pain interfered
with activity and mood.

Analyses revealed evidence for direct and independent multidirectional accounts. Greater
mindfulness predicted lower CP scores and greater social support. In turn, increases in
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S269

mindfulness were predicted by greater social support and lower CP scores. Additionally, using
50,000 bootstrapped samples, mindfulness was found to have a significant indirect effect on
CP conditions as mediated by social support. This is the first study to demonstrate a possible
psychosocial mediator of mindfulness in a national sample of midlife and older adults.
Implications are discussed in terms of possible intervention studies that target enhancement
of social support in mindfulness-based CP management.

CORRESPONDING AUTHOR: Shin Ye Kim, Ph.D., Texas Tech University, Lubbock, TX, 79410;
[email protected]
S270 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A149 6:00 PM-7:00 PM

A SOCIAL MEDIA APPROACH TO INFLUENCE LIFESTYLE PHYSICAL ACTIVITY IN NAVY

RESERVISTS: THE ACTIVE RESERVES (ACRE) TRIAL

Neil Goldstein, Bachelor of Arts - Exercise Science, Thomas Wojcicki, PhD

Bellarmine University, Louisville, KY

In an effort to create a culture of year-round fitness, the US Navy is seeking innovative programs to assist in the promotion of health-related behaviors, such as physical
activity (PA). A recent examination of Physical Readiness Test scores of Reservists (N=189) from the Naval Operational Support Center in Louisville, KY, further highlights this
need, as 13.2% failed this assessment due to poor fitness, and an additional 19.6% barely met the minimum criteria for passing. The purpose of the AcRe Trial was to examine
the feasibility and efficacy of delivering an 8-week lifestyle PA intervention to a sample of low-active Navy Reservists (N=15; Mage=33.80 years) entirely over Facebook.
Primary outcomes included changes in health-enhancing PA (assessed at baseline and 8-week follow-up), as well as post-intervention evaluations regarding program delivery
and usefulness. Participants were randomized to either a social cognitive-based condition (i.e., Behavioral group) or an attentional control (i.e., Informational group). Both
conditions received access to a study-specific Facebook Group, where they received and engaged with PA-related information and resources specifically aimed at military
personnel, as well as adult civilians. Participants in the Behavioral group also received weekly video-based behavioral modules designed to promote PA. Effect sizes (Cohens
d) were calculated within groups to determine differential treatment effects on PA. Results revealed that involvement in the Behavioral group produced greater effects on PA
compared to the Informational group. Specifically, the Behavioral group experienced large and positive changes in moderate-vigorous PA (d=1.40), whereas the Informational
group experienced small to moderate declines (d=-0.44). Post-intervention evaluations revealed that participants found the AcRe Trial to be very useful and that they would
consider participating in a similar social media-delivered program again (agree=60%, strongly agree=40%, respectively). These results suggest that a theoretically-based
intervention delivered via social media has the potential to positively influence PA levels in low-active Reservists. This practical and well-received approach to promoting PA
could be easily modeled and adapted to meet the dynamic needs and health-related goals of the reserve component of the US Navy a critically understudied population,
which plays a paramount role in the continuity of national defense and security.

Neil B. Goldstein, Department of Exercise Science, Bellarmine University, 2001 Newburg Rd, Louisville, KY 40205, (502) 272-7613; [email protected]

CORRESPONDING AUTHOR: Neil Goldstein, Bachelor of Arts - Exercise Science, Bellarmine

University, Louisville, KY, 40220; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S271

A150 6:00 PM-7:00 PM

ITS HARD TO MANAGE CONFLICTING LEISURE GOALS: DO CONCURRENT GOAL-RELATED

COGNITIONS PREDICT EXERCISE VOLUME?

Jocelyn Blouin, BSc. Kin. (Hon.)1, Nancy C. Gyurcsik, PhD2, Larry R. Brawley, PhD2, Kevin Spink,
Professor2

1
University of Saskatchewan, North Battleford, SA, Canada; 2University of Saskatchewan,
Saskatoon, SA, Canada

According to social cognitive theory (SCT), concurrent management of exercise and other
highly valued leisure goals may compete and challenge successful exercise self-regulation.
Concurrent goal-related cognitions may be self-regulatory mechanisms important for planned
exercise engagement. Concurrent self-regulatory efficacy (SRE), or confidence to concurrently
manage multiple goals including exercise, has been predictive of adults' exercise. Intergoal
conflict, or the extent that concurrent goals conflict, is suggested to predict exercise.
However, studies have not supported the prediction. The lack of support may be because a
truly challenging conflict between concurrent goals was never examined. Specifically,
participants exercise and non-exercise goals were not highly valued and thus did not conflict
with each other. Our purpose was to examine whether adults perceived intergoal conflict
enhanced exercise prediction beyond concurrent SRE when concurrent goals were conflicting
and highly valued. Healthy adult exercisers (N = 112; Mage = 30.9 10.5 years) reported a
highly valued ( 6 on a 1 to 9 value response scale) exercise and concurrent non-exercise
leisure goal. Concurrent SRE, intergoal conflict, and moderate-vigorous exercise volume over
the prior month were also assessed. Hierarchical multiple regression was used to examine the
relationship between goal-related cognitions and exercise. Step 1 included concurrent SRE
and step 2 included intergoal conflict. The overall model was significant, F(2, 109) = 8.95, p =
.001, R2adjusted= .13, with concurrent SRE as the sole significant predictor, = .32, p = .001.
The concurrent SRE finding supports SCT contentions that SRE beliefs are predictive of goal-
directed behavior when goals compete. But this is specific to exercisers who have mastery
experiences. In contrast, exercise initiates may be more likely to have goal conflicts that they
need to learn how to master. Future research should prospectively examine intergoal conflict
and concurrent SRE among adults who initiate exercise and follow the predictors of their
adherence trajectory over time. Such a study might offer insight on when intergoal conflict
does and does not matter.
S272 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Jocelyn Blouin, BSc. Kin. (Hon.), University of Saskatchewan,

North Battleford, SA, S9A 3A3; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S273

A151 6:00 PM-7:00 PM

ADAPTIVE GOAL SETTING AND FINANCIAL INCENTIVES: A 2 X 2 FACTORIAL RANDOMIZED

TRIAL TO INCREASE PHYSICAL ACTIVITY AMONG ADULTS

Marc A. Adams, PhD, MPH1, Jane C. Hurley, MS1, Michael Todd, PhD1, Nishat Bhuiyan, B.S.2,
Catherine Jarrett, MS1, Wesley J. Tucker, PhD RD3, Kevin Hollingshead, BSEET4, Siddhartha S.
Angadi, PhD5

1
Arizona State University, Phoenix, AZ; 2The Pennsylvania State University, State College, PA;
3
University of Texas at Arlington, FORT WORTH, TX; 4Arizona State University, Scottsdale, AZ;
5
Arizona State University - Downtown Phoenix Campus, Phoenix, AZ

Purpose: Recent perspectives in health behavior theory propose that behavior, such as
physical activity, has temporal dynamics that are unique to each individual. Emerging
interventions that rely on and harness such variability and adapt intensively to individuals
over time may outperform static interventions that prescribe static goals (e.g., 10,000
steps/day). The purpose of this study was to compare adaptive goal setting and immediate
financial rewards to static goal setting and delayed, non-contingent rewards for promoting
free-living physical activity (PA).

Methods: A 4-month 2x2 factorial randomized trial tested main effects for goal setting
(adaptive vs. static goals) and reward (immediate vs. delayed) and interactions between
factors to increase physical activity as measured by a Fitbit Zip. This analysis reports
the secondary outcome of changes to minutes of moderate-to-vigorous PA (MVPA) using
previously validated minute-level cadence thresholds. Static goals (10,000 steps/day) did not
change while adaptive goals adjusted up or down daily over the course of the 4 months using
a percentile-rank algorithm based on sliding window of a participants performance.
Immediate financial incentives were made available for goal attainment while delayed, non-
contingent incentives were provided monthly for participation.

Results: Participants (N=96) were mainly female (77%), aged 41 9.5 years, and all were
overweight/obese (mean BMI = 34.1 6.2). Participants across all groups increased by 12.2
min/day from baseline to intervention phase on average (p < .001). Participants receiving
static goals increased by 2.5 min/day more on average compared to adaptive goals regardless
of reward group (p = .11). Participants receiving immediate rewards increased by 4.7 min/day
more on average compared to the delayed rewards regardless of goal-setting group (p = .006).
Average rate of change during the intervention phase differed statistically between groups,
and showed that MVPA for participants receiving adaptive goals decreased at a slower rate
S274 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

(2.5 seconds less per day, p = .005) on average, compared to participants receiving static
goals, such that by the end of the 4-month intervention phase, participants receiving adaptive
goals differed by 4.6 min/day compared with static goals, on average. Rate of change did not
differ between reward groups during the intervention phase. Significant interactions between
goal setting and reward factors were observed.

Conclusions: This study evaluated independent and joint contributions of daily goal setting
and financial incentives for increasing MVPA min/day. The factorial design further allowed for
exploring the optimized version of these interventions. These results contribute to
advancement of the science of optimizing behavioral interventions for promoting physical
activity among adults.

CORRESPONDING AUTHOR: Marc A. Adams, PhD, MPH, Arizona State University, Phoenix, AZ,
85004; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S275

A152 6:00 PM-7:00 PM

ARE CARDIAC PATIENTS MEETING PHYSICAL ACTIVITY GUIDELINES?

Bernice Yates, PhD, RN1, Bunny Pozehl, PhD, APRN2, Kevin A. Kupzyk, PhD1, Crystal Epstein,
MSN, APRN1, Pallav Deka, BSN, MS, PhD (c), RN2

1
University of Nebraska Medical Center, Omaha, NE; 2University of Nebraska Medical Center,
Lincoln, NE

Title: Are Cardiac Patients Meeting Physical Activity Guidelines?

Purpose: Physical inactivity is an important independent risk factor for coronary heart
disease. In primary and secondary prevention of coronary heart disease, it is recommended
that individuals achieve > 150 minutes/week of > moderate intensity physical activity (PA)
accumulated in at least 10 minute bouts. Although this amount is recommended, few studies
have examined how many community-dwelling cardiovascular patients achieve this
recommendation. The purpose of this study was to describe subjective and objective physical
activity (PA) levels of two groups of cardiovascular patients who were either post-coronary
artery bypass graft (CABG) surgery or diagnosed with heart failure (HF).

Methods: A descriptive comparative design was used for this secondary analysis of data from
two prior studies. A convenience sample of 62 outpatients was used to examine PA
objectively (Actiheart accelerometer) and subjectively (PA interview). HF and CABG patients
were compared using were examined using Mann-Whitney U statistics.

Results: Based on accelerometer data, HF patients engaged in < 1 minute of moderate-to-

vigorous physical activity (MVPA) per day. They engaged in more light activities (1.6 2.9
METs) than moderate-vigorous activities (18.2 min/d). Thirty eight percent of HF patients met
the national guidelines of > 150 min/wk by self-report; however, none met the guidelines
based on accelerometer data. CABG patients engaged in 18.5 minutes of MVPA per day
(129.5 min/wk) based on accelerometer data. They also engaged in more light than MVPA
(146.8 min/d). Fifty-six percent of CABG patients met the national guidelines of > 150 min/wk
based on self-report; however, only 33% met the guidelines based on accelerometer data.
CABG patients were significantly more active than HF patients as evidenced by significantly
greater minutes spent in light and moderate-to-vigorous activity.

Conclusions: Few cardiac patients in the current study met PA recommendations. Innovative
intervention strategies are needed (e.g., use of activity trackers, making PA a vital sign) to
S276 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

assist patients in gaining the knowledge and skills to be more active and adhere to PA
recommendations.

CORRESPONDING AUTHOR: Bernice Yates, PhD, RN, University of Nebraska Medical Center,
Omaha, NE, 68198-5330; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S277

A153 6:00 PM-7:00 PM

ASSOCIATION BETWEEN CHANGES IN PHYSICAL ACTIVITY & DEPRESSIVE SYMPTOMS &

PERCEIVED STRESS IN ADULTS WITH DIABETES & CHRONIC PAIN

Susan Andreae, PhD, MPH1, Susan L. Davies, PhD1, Joshua Richman, MD PhD2, Andrea
Cherrington, MD MPH3, Monika Safford, MD4

1
University of Alabama at Birmingham, Birmingham, AL; 2UAB, Birmingham, AL; 3University of
Alabama at Birmingham, Division of Preventive Medicine, birmingham, AL; 4Weill Cornell
Medicine, New York, NY

Background: Individuals with diabetes often experience higher than average symptoms of
stress and depression. Barriers to addressing stress and depression include negative attitudes
and beliefs regarding mental health care, distrust of the mental health care system, and
stigma related to mental illness. A promising approach to alleviating such symptoms may be
through the use of physical activity (PA) and exercise. We examined the effects of the change
in PA levels on depressive symptoms and perceived stress levels in a sample of primarily
African American adults in rural Alabama.

Methods: The study included adults with diabetes and chronic pain enrolled in the
intervention arm of a community-based, cluster- randomized controlled trial testing a 12-
week program that included cognitive reframing to increase physical activity. Depressive
symptoms were assessed using the Iowa short form of the Centers for Epidemiological Data
Scale. Perceived stress levels were measured using the 10-item Perceived Stress Scale. PA
levels were assessed using questions that measured intensity and frequency of PA, perception
of ones own PA compared to others of the same age, and usual daily PA levels. Linear
regression with generalized estimating equations were used to account for the clustering and
longitudinal nature of the data.

Results: The 93 participants in the study had a mean age of 59.9 ( 9.8 SD) years, 98% were
African Americans, 84% were female, 73% reported annual income of less than $20,000, and
65% had a high school education or lower. Change in usual daily activity levels were
associated with changes in perceived stress levels (=-2.7 [-5.02, -0.35] p < 0.05), but not with
changes in depressive symptoms.

Discussion: Increases in reports of usual daily PA levels were associated with decreased levels
of perceived stress but not depressive symptoms in this cohort. Future studies are needed in
order to better understand the relationship between depression, stress, and physical activity
S278 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

in other communities, to assess the directionality of the relationships, and to understand the
mechanisms through which PA may affect depression and stress

CORRESPONDING AUTHOR: Susan Andreae, PhD, MPH, University of Alabama at Birmingham,

Birmingham, AL, 35205; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S279

A154 6:00 PM-7:00 PM

ASSOCIATION BETWEEN PA AND SELF-EFFICACY FOR PAIN MANAGEMENT AND DIABETES

MANAGEMENT IN ADULTS WITH DIABETES AND CHRONIC PAIN

Susan Andreae, PhD, MPH1, Susan L. Davies, PhD1, Joshua Richman, MD PhD2, Monika
Safford, MD3, Andrea Cherrington, MD MPH4

1
University of Alabama at Birmingham, Birmingham, AL; 2UAB, Birmingham, AL; 3Weill Cornell
Medicine, New York, NY; 4University of Alabama at Birmingham, Division of Preventive
Medicine, birmingham, AL

Introduction: In addition to preventing chronic disease and premature death, physical activity
(PA) may have immediate and positive effects on self-efficacy and confidence. Emphasizing
these quickly noticeable effects of PA may be a more effective method to promote PA
compared to focusing on long-term benefits (e.g., prevention or control of disease). We
examined the relationship between PA with diabetes management self-efficacy and pain
management self-efficacy in a sample of minority adults with diabetes and chronic pain.

Methods: Our sample was the intervention arm of a community-based, cluster-randomized

controlled trial testing the effectiveness of a 12-week community health worker delivered
program that emphasized cognitive reframing on diabetes outcomes in individuals with
chronic pain and diabetes. Diabetes management self-efficacy was assessed using the
Perceived Diabetes Self-Management Scale. Pain management self-efficacy was assessed
using the pain management self-efficacy subscale of the Arthritis Self-Efficacy Scale. PA was
measured using questions assessing intensity and frequency of PA, perception of PA
compared to others, and usual daily PA levels.

Results: Of the 93 study participants, 98% were African American, 84% were women, 73%
reported an annual income of less than $20,000, and 63% had a high school education or
lower. The mean age of participants was 609.9 SD years. There were positive and significant
associations between the change in perception of PA compared to others and change in pain
management self-efficacy levels (=3.89 [0.53, 7.26], p < 0.05). There were no significant
associations between measures of PA and diabetes management self-efficacy.

Conclusion: Increases in the perception of ones own PA levels compared to others were
associated with improvements in pain management self-efficacy. Reframing the benefits of PA
S280 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

to include affective benefits such as improvements in self-efficacy may be a potential strategy

to promote PA.

CORRESPONDING AUTHOR: Susan Andreae, PhD, MPH, University of Alabama at Birmingham,

Birmingham, AL, 35205; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S281

A155 6:00 PM-7:00 PM

DECREASES IN PHYSICAL ACTIVITY AND PUBERTAL MATURATION: TESTING EVOLUTIONARY

HYPOTHESES AMONG THE TSIMANE OF LOWLAND BOLIVIA

Annie Caldwell, PhD1, Paul Hooper, PhD2

1
University of Colorado, School off Medicine, Aurora, CO; 2Emory University, Atlanta, GA

Understanding factors that affect physical activity can have tremendous public health
implications. Epidemiological studies have consistently demonstrated decreases in physical
activity during adolescence in modern, Western contexts. This trend is typically interpreted as
adolescents adopting unhealthy behaviors in response to the psychosocial stress of puberty,
similar to smoking, drinking, or risky sexual behaviors. Much less is known about physical
activity patterns across adolescence in pre-Industrial settings. In the current study, we utilized
an evolutionary, life history framework to examine physical activity across adolescence among
a cross-section of young Tsimane individuals (N=110; age 8-22). The Tsimane are forager
horticulturalists that live in small villages in the Amazon region of lowland Bolivia. They have a
subsistence based economy that requires much higher levels of obligatory physical activity to
produce food, water, and shelter. We tested the hypothesis that the Tsimane would exhibit
similar decreases in physical activity and increases in sedentary behavior across adolescence,
despite having higher levels of obligatory physical activity. We specifically tested the
hypotheses that physical activity trades-off against energy devoted to reproductive
maturation, and is sensitive to individual energetic condition. Physical activity was measured
objectively using Actigraph tri-axial accelerometers; investment in pubertal maturation was
measured with Tanner stage, urinary DHEA and testosterone; and energetic condition was
measured with urinary C-peptide and serum HbA1c. Minutes of the day spent being sedentary
significantly increased with Tanner stage ( = -28.66, t = 3.47, p < 0.001) controlling for age
and sex. Minutes of the day spent in moderate to vigorous physical activity significantly
decreased with Tanner stage ( = -9.00, t = -2.00, p < 0.05), controlling for the age and sex.
Further, the relationship between DHEA and physical activity was moderated by C-peptide,
such that those who were in poorer energetic condition, exhibited steeper decreases in
activity with increasing levels of DHEA. Results will be interpreted from a life history
framework and considered in light of implications for public health and health disparities in
modern contexts.

CORRESPONDING AUTHOR: Annie Caldwell, PhD, University of Colorado, School off Medicine,
Aurora, CO, 80045; [email protected]
S282 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A156 6:00 PM-7:00 PM

EASING THE TRANSITION TO COLLEGE: COMPARING PHYSICAL ACTIVITY AND MINDFULNESS

INTERVENTIONS AMONG FIRST-YEAR STUDENTS

Pamela J. Lundeberg, M.A.1, Dan J. Graham, PhD2

1
Colorado State, Fort Collins, CO; 2Colorado State University, Fort Collins, CO

Background: Engaging in physical activity (PA) is associated with an array of physical benefits
(e.g., increased longevity, enhanced cardiovascular functioning, and a reduced risk for chronic
diseases). In addition, PA offers numerous benefits for psychological health (e.g., decreased
risk for depression and anxiety, reduced stress, and enhanced mood). Many of the
psychological benefits of PA are also associated with mindfulness training, which like PA
has demonstrated efficacy in improving attention, increasing self-control, and reducing stress.
As PA and mindfulness offer similar psychological benefits, the present study sought to
compare the two among first-year college students, a group of individuals experiencing
significant stress as they transition to the university setting.

Method: First-year college students (N = 52) voluntarily elected to participate in a 4-week PA

or mindfulness workshop (with weekly 75-minute sessions led by a certified instructor) during
which students learned methods of and benefits of practicing PA or mindfulness. Additionally,
participants practiced newly learned techniques in the workshop sessions (i.e., students
engaged in different types of PA or mindfulness). The PA and mindfulness workshops were
built to parallel each other in terms of relative content and structure, and both workshops
emphasized college adjustment outcomes (e.g., stress reduction). Outcome variables were
measured prior to the start of the workshop (T1) and at the final workshop (T2).

Results: Preliminary analyses indicate greater efficacy for the PA intervention compared to
the mindfulness program. Students in the PA group reported lower perceived stress, higher
positive affect, lower homesickness, and higher self-efficacy at T2 compared to T1.
Additionally, participants in the PA group reported higher global health and an increase in
perceived benefits of PA at T2 compared to T1. Participants in the mindfulness group did not
report changes from T1 to T2 in any of the measured outcome variables related to college
adjustment.

Conclusions: First-year college students completing a 4-week PA intervention reported

acquiring useful knowledge and gained competency in multiple PA modalities. Furthermore,
participants in the PA group reported an increase in college adjustment outcomes after
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S283

participating in the 4-week PA workshop, suggesting a benefit of utilizing PA as a method of

easing the transition to college.

CORRESPONDING AUTHOR: Pamela J. Lundeberg, M.A., Colorado State, Fort Collins, CO,
80525; [email protected]
S284 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A157 6:00 PM-7:00 PM

ENHANCING ACCESSIBILITY OF PHYSICAL ACTIVITY DURING PREGNANCY

Patricia Kinser, PhD, WHNP-BC, RN, FNAP1, Christine Aubry, student2, Saba Masho, MD, DrPH2

1
Virginia Commonwealth University School of Nursing, Richmond, VA; 2Virginia
Commonwealth University, Richmond, VA

Background: Although the CDC recommends that pregnant women participate in physical
activity at least five times each week for 30 minutes, not even 16% of pregnant women are
engaging at that level. Pregnant women have expressed interest in yoga as a form of low-
intensity physical activity and desire a convenient method for accessing yoga. Yoga is safe for
pregnancy and is associated with improvements in quality of life, stress and related symptoms
(e.g., depression, anxiety), and may be a strategy for increasing physical activity in pregnancy.

Purpose: The purpose of this study is to evaluate the feasibility and acceptability of integrating
prenatal yoga into a group prenatal care setting, as informed by patients and providers who
participated in combined group prenatal care and prenatal yoga classes (CenteringPregnancy
plus Yoga, or CPC+Y classes).

Methods: This community-based qualitative study utilized focus groups with pregnant women
(n=16) who participated in CPC+Y and with healthcare providers who were involved in CPC+Y
(n = 5). Open-ended questions were used to elicit thoughts, beliefs, attitudes, and concerns
regarding the integration of low-intensity physical activity, in the form of prenatal yoga, with
group prenatal care. Qualitative data was analyzed through content analysis based on
descriptive qualitative methodology with phenomenological overtones in an iterative process.

Results: The following key themes arose from the data: (1) the integration of yoga into group
prenatal care provided a convenient way for participants to engage in gentle physical activity;
(2) participants experienced decreased stress and enhanced confidence with low-intensity
physical activity; and (3) time, space, and partner issues were key barriers to full engagement
with the yoga aspect of the group prenatal care meetings.

Conclusions/Implications: Given the imperative by the Institute of Medicine to encourage

pregnant women to engage in physical activity and control weight gain, this study is timely
and relevant for evaluating whether the incorporation of gentle physical activity, in the form
of yoga, into group prenatal care could enhance womens engagement in physical activity.
Standard prenatal care does not currently fully address pregnant womens low engagement in
physical activity, hence the findings from this study may have the potential to impact the way
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S285

prenatal care is delivered. The findings from this study are encouraging and will be used to
inform the development of future large-scale randomized controlled trials.

CORRESPONDING AUTHOR: Patricia Kinser, PhD, WHNP-BC, RN, FNAP, Virginia

Commonwealth University School of Nursing, Richmond, VA, 23298; [email protected]
S286 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A158 6:00 PM-7:00 PM

EXERCISING FOR TWO: ASSOCIATIONS OF MATERNAL EXERCISE, FETAL GROWTH, & INFANT
BIRTH WEIGHT IN WOMEN WITH GESTATIONAL DIABETES

Krista S. Leonard, B.A.1, Danielle Symons Downs, Ph.D.2

1
The Pennsylvania State University, State College, PA; 2The Pennsylvania State University,
University Park, PA

Exercise (EX) may help women with gestational diabetes mellitus (GDM) to control blood
glucose and possibly prevent or delay the onset of insulin. However, little is known about the
association between maternal EX and infant growth (i.e., fetal abdominal circumference [AC];
birth weight) despite that EX has the potential to positively impact fetal growth and
subsequently lower the risk for macrosomia (i.e., baby weighing more than 4,000 g at
birth). Purpose: Examine associations between maternal 2nd trimester EX and insulin use,
fetal growth (AC [mm] and percentile [%tile]), and infant birth weight among GDM women in
a randomized EX intervention. Methods: GDM women (N = 65) were randomized to either the
face-to-face (F2F) EX intervention, home EX intervention, or control group at 20-weeks
gestation through delivery. Average steps/day and minutes spent in moderate-vigorous
EX/day were measured at 20-weeks gestation with an NL1000 pedometer. AC mm and %tile
were measured by ultrasound at 32- and 36-weeks gestation. Insulin use and infant birth
weight were obtained from medical records. Of the 65 women, 24 had sufficient data on all of
these measures to conduct the analyses. Results: Although group differences were not
significant, theF2F group was less likely to need insulin compared to the home and control
groups. The F2F group initiated insulin use later in gestation (32.83 weeks) compared to the
home group (27.75 weeks; p < .10); however, there was no difference compared to the
control group (32.67 weeks; p > .10). Pearsons correlations showed that higher NL1000
steps/day was associated with lower AC %tile at 32- and 36- weeks gestation and AC mm at
36-weeks gestation among the entire sample. Also, higher NL1000 min moderate-vigorous EX
was associated with lower AC %tile and mm at 32-weeks gestation and lower infant birth
weight. There were no significant individual group correlations. Lastly, while a small
proportion (14%) of the overall sample gave birth to infants with macrosomia, twice as many
were in the home group compared to the F2F and control groups. Conclusion: These
preliminary findings suggest that higher maternal EX is associated with lower AC (%tile and
mm) and infant birth weight. However, further research is needed to replicate these findings
given the small sample size. A F2F EX intervention may help to regulate fetal growth among
GDM women and delay the onset of insulin without adverse effects to fetal growth or infant
birth weight.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S287

CORRESPONDING AUTHOR: Krista S. Leonard, B.A., The Pennsylvania State University, State
College, PA, 16801; [email protected]
S288 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A159 6:00 PM-7:00 PM

INCREASING PHYSICAL ACTIVITY IN MOTHERS USING MOBILE APPS AND VIDEOCONFERENCE:

MOVE TRIAL RESULTS

Maya Mascarenhas, MPH

University of California, San Francisco, San Francisco, CA

Intro: Mothers have lower levels of activity compared to their own levels before children and
compared to fathers and women of the same age with no children. They experience unique
barriers towards activity including a lack of leisure time, fatigue, childcare needs, and a sense
of isolation.

Methods: MOVE (Mothers Online Video Exercise) is an online program designed for mothers
with at least one child under the age of 13. Videoconferencing tools (Google Hangouts) were
used to create virtual spaces where mothers exercised together using popular mobile apps
that guided short intensive workouts.

We enrolled 64 women, 30 of whom were randomized to an 8-week online exercise program

that started upon enrollment (intervention) and 34 were randomized to start a group after 8-
weeks (waitlist control). Our primary outcome was minutes of moderate and vigorous physical
activity (MVPA) measured using the Godin physical activity instrument. Secondary outcomes
included weight loss (self-report), physical activity self-efficacy and social support. We
compared changes from baseline to 8 weeks between the intervention and control arm using
t-tests. Feasibility and acceptability were assessed for intervention participants.

Results: In the intervention arm (n=27 with completed data), women had a baseline MVPA of
65 minutes, which increased by 30 minutes at 8 weeks. In the waitlist arm (n=27 with
completed data), women had a baseline MVPA of 68 minutes, which decreased by 4 minutes
at 8 weeks. The difference of 39 MVPA (95% CI [-20-99]) in the change in minutes between
arms was not significant. Women reported an average baseline weight of 147 pounds in the
intervention arm, and 145 in the control arm. After 8 weeks, women in the intervention
reported a loss of 2 pounds (95% CI [-4.1-0.27]) compared to a loss of 0.19 pounds in the
control arm, a difference that was almost statistically significant (p=0.08).

All participants reported that they would recommend this program to a friend. Women in the
intervention arm reported feeling good or great after exercise group sessions, and all reported
feeling an increase in their fitness levels. Half the women chose to continue with their group
at the conclusion of their 8-week participation. On average, women participated in 3 sessions
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S289

per week with routines that lasted 15 minutes. The most popular workout apps included
Nike+, Sworkit and Johnson & Johnson 7 Minute App.

Discussion: An online exercise intervention proved feasible, and acceptable to mothers. A

larger trial with continuous measures of physical activity and fitness measures is warranted to
understand if the intervention if effective at increasing activity levels and fitness and whether
weight loss can be reproduced and sustained in a new sample.

CORRESPONDING AUTHOR: Maya Mascarenhas, MPH, University of California, San Francisco,

San Francisco, CA, 94110; [email protected]
S290 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A160 6:00 PM-7:00 PM

MERITORIOUS AWARD WINNER

PHYSICAL ACTIVITY AND EXECUTIVE FUNCTION TRAJECTORIES FROM 4TH THROUGH 6TH
GRADE: DIFFERENCES BY PHYSICAL ACTIVITY TYPE

Christopher Warren, BA1, Mary Ann Pentz, PhD, MA2

1
University of Southern California Keck School of Medicine, Los Angeles, CA; 2University of
Southern California, Department of Preventive Medicine, Institute for Health Promotion and
Disease Prevention Research, Los Angeles, CA

Background: Multiple studies in the pediatric literature have linked physical activity (PA) to
enhanced executive function (EF), which in turn has been linked to improved health, and well-
being across the lifespan. While evidence suggests that various forms of PA may differentially
impact EF, such work has been limited to acute studies and others limited in their ability to
evaluate longitudinal relationships between multiple PA types and EF trajectories across early
adolescence. Moreover, extant work has largely studied aerobic exercise paradigms which
are not cognitively demanding (e.g. treadmill running, stationary bicycling) and thus may not
tap or promote executive functioning to the same degree as other, more cognitively taxing
forms of physical activity (e.g. certain sports, martial arts).

Methods: A school-based sample of 709 youth were administered 4 subdomains of the

Behavioral Rating Inventory for Executive Function-Self Report scale (inhibitory control,
emotional control, working memory, organization of materials), and an abridged Self-
Administered Physical Activity Checklist (SAPAC) at four time points from 4th through 6th
grade. Growth curve models (GCMs) were fit examining relationships between EF trajectories
and each of 12 PA types assessed by the SAPAC adjusting for participant SES, gender,
ethnicity, group assignment, school, and total physical activity. Multiple group conditional
GCMs were fit to test whether baseline EF or EF slope growth factors differed significantly
between students who reported participating in each activity at 4th, 5th, and 6th grade
assessments, vs. students who did not. All models demonstrated excellent fit (CFI >.95;
RMSEA < .05).

Results: Regular participation in martial arts (N=35) and skating (including ice, roller, and
skateboarding) (N=148) were each independently associated with significantly increased EF
growth from 4th through 6th grade [B (martial arts participants) =.58; B (non-participants)
=.09; chi-squared difference test p < .05] [B (skating participants) =.36 ; B (non-participants)
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S291

=.05; chi-squared difference test p < .05]. No significant group differences in EF were observed
at 4th grade (baseline) between participants and non-participants in either activity. No
significant differences in baseline EF or EF trajectories across time were found for the other
activities assessed by the SAPAC (i.e. ball playing, bicycling, swimming, gymnastics, exercises
like pushups/sit-ups/jumping jacks, team sports, dancing, racket sports, active outdoor games,
and active chores.

Conclusions: This study found that only martial arts and skating were associated with EF
growth trajectories, whereas other types of PA were not. While previous work has linked
martial arts to EF promotion, future research should confirm and further investigate
mechanisms through which skating may lead to improved EF from 4th through 6th grade.

CORRESPONDING AUTHOR: Christopher Warren, BA, University of Southern California Keck

School of Medicine, Los Angeles, CA, 90032; [email protected]
S292 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A161 6:00 PM-7:00 PM

POSITIVE PARENTS' PERCEPTIONS OF IMPLEMENTING A PHYSICAL ACTIVITY PROGRAM AT

HOME IN YOUNG CHILDREN WITH PRADER-WILLI SYNDROME

Kasen Bras, Undergraduate student1, Kathleen Wilson, Ph.D.2, Daniela A. Rubin, Ph.D.3

1
California State University Fullerton/ Cal Poly Pomona, Fullerton, CA; 2California State
University Fullerton, Fullerton, CA; 3California State University, Fullerton, Fullerton, CA

Prader-Willi Syndrome (PWS) is a rare (1 in 15,000) genetic neurodevelopmental disorder

leading to hyperphagia, hypotonia, high adiposity and developmental delay. PWS affects
motor skills, communication and behavior. Thus, characteristics of the syndrome pose
challenges for children and adults with PWS to being physically active. This study used a
multiple case study design to explore parent perceptions of their experience with leading a
physical activity intervention with their child at home. Home-based Active Play for Parents
and Youth (HAPPY) Early Discoveries is a 12-week active play program facilitated by parents
that focuses on improving motor skills in children ages 4 to 7 years old. This multiple-case
study included four different parents of children with PWS who completed the HAPPY Early
Discoveries program. Each parent participated in an exiting semi-structured interview
exploring their experiences with the program and the changes they perceived in their child.
The transcriptions of these interviews were analyzed using a hierarchical content analysis to
extract common themes. The themes were separated into implementation and outcomes of
the program. When implementing the program, major themes described by all parents were
the challenges of the program execution and finding the right strategies to motivate their
childs participation. The challenges included balancing the program implementation with a
busy life and scheduling the program in. The main motivation to do the program, for both
parents and children, was the use and appeal of the equipment and activities provided by the
program. With regards to the outcomes, major aspects observed by parents were
improvements in their childs motor skills (i.e. coordination and muscular strength) and the
motivation for physical activity. Parents mentioned their child was motivated by having
choices in the program and would initiate some activities on their own incorporating their
imagination. Another outcome of the program was the enjoyment experienced by the parents
and children as the parents enjoyed the family togetherness and children viewed physical
activity as fun. Despite challenges with implementing the program, parents of young children
with PWS gave positive feedback about implementing a game-based physical activity program
at home.
Funded by the US Army Medical Research and Materiel Command Award W81XWH-11-1-
0765.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S293

CORRESPONDING AUTHOR: Kasen Bras, Undergraduate student, California State University

Fullerton/ Cal Poly Pomona, Fullerton, CA, 92831; [email protected]
S294 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A162 6:00 PM-7:00 PM

PREDICTORS OF DAILY PATTERNS OF PHYSICAL ACTIVITY AMONG LATINA PARTICIPANTS IN A

PHYSICAL ACTIVITY INTERVENTION

Alexander W. Sokolovsky, MA1, Shira Dunsiger, PhD2, Lucas Carr, Ph.D.3, Bess H. Marcus,
Ph.D.4

1
Brown University, Providence, RI; 2The Miriam Hospital and Brown University, Portsmouth,
RI; 3University of Iowa, Iowa City, IA; 4University of California, San Diego, La Jolla, CA

Physical activity (PA) interventions based on the Transtheoretical model frequently target
behavioral, cognitive, and other psychological factors to move people through a series of
stages associated with adoption and maintenance of PA. This study examined long-term daily
patterns of PA among adult Latinas enrolled in a 12-month PA intervention.Objective daily PA
data was collected using pedometers over 12 months among 122 Spanish speaking Latinas
(age M = 41.1 years, SD = 10.0; baseline BMI M = 29.5, SD = 4.3; 50.0% married; 45.1%
unemployed). We fit a linear mixed effects regression model, regressing PA (i.e. total daily
steps) onto demographic variables (partner and employment status), number of young
children, day of week, season, baseline depression, days with pedometer, and time-varying
predictors including BMI, self-efficacy, stage of change, behavioral and cognitive processes,
and social support. Participants averaged 6,603 steps/day. PA was highest during weekdays
and summer. PA was significantly negatively associated with days with pedometer, number of
young children, and baseline depression; and positively associated with enjoyment,
behavioral processes, and stage of change. Neither self-efficacy nor social support
significantly predicted PA. We further examined the interaction between self-efficacy and
stage of change and results suggest individuals in early stages of change demonstrated a
significant negative relationship between self-efficacy and PA relative to the non-significant
effects observed among those in more advanced stages of change.Progression through stages
of change in this intervention was associated with increased daily PA. The significant
interaction between self-efficacy and stage of change indicated that individuals in early stages
of change may be more confident in their ability to increase their daily PA relative to those
who have begun implementing such changes. Seasonality, weekday, and number of young
children all impeded daily PA, the latter being consistent with subjects self-reports of barriers
to engaging in PA. Results suggest that the addition of a daily activity monitors to
interventions can provide some insight into how and when specific intervention approaches
impact physical activity.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S295

CORRESPONDING AUTHOR: Alexander W. Sokolovsky, MA, Brown University, Providence, RI,

02912; [email protected]
S296 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A163 6:00 PM-7:00 PM

SOCIAL CUES INFLUENCE ON THE RELATIONSHIP BETWEEN GOAL-SETTING AND EXERCISE

PERFORMANCE

Charles Heidrick, MS, Dan J. Graham, PhD

Colorado State University, Fort Collins, CO

In two previous studies, we identified social factors that influence goal-setting and
performance in an exercise context. In study 1, participants were more likely to achieve
exercise goals when their goals were affirmed by a researcher than when their goals were
kept private or were presented to, but unacknowledged by, a researcher, suggesting that
social affirmation can increase task motivation and persistence. In study 2, participants were
more likely to set more challenging exercise goals when provided social comparison
information; participants in study 2 were randomly assigned to view normative exercise
performance data that was accurate, overestimated, or underestimated, and in all three
conditions, participants set more challenging exercise performance goals than when no social
comparison data were provided. Additionally, being provided with overestimated normative
data led participants to set even more challenging goals, leading to improved exercise
performance vs. receiving accurate or underestimated normative data. In these studies, social
influences exerted prior to beginning a task impacted goal-setting, task motivation, and
performance. In the present study, we sought to determine whether the content of these
influential communications had the same positive impacts if presented electronically, rather
than directly from another person. College students (n = 171) participated in a controlled
experiment. A researcher demonstrated four exercises (pushups, planks, jumping jacks, and
single-leg balancing), after which participants set personal goals regarding their own imminent
performance of these exercises. Participants were randomly assigned to one of two
conditions: 1) Goals affirmed: after entering their goals into a computer program, participants
received a message of These goals look great! or 2) Goals acknowledged: after entering
their goals into a computer program, participants received a message of These goals have
been received. Performance averages on the four exercises were as follows for the
Affirmation group: 24.6 consecutive pushups, 78.5 second plank, 70.4 jumping jacks in one
minute, and 20.8 second balance. None of these values differed significantly (p-values .21-.70)
from the performances by the Acknowledgment group (24.9, 76.1, 68.8, 23.0, respectively). In
contrast to the positive social benefits seen in studies 1 and 2, affirmation from a computer
had no effect on exercise performance in study 3, suggesting that the social aspect of human
affirmation is critical. Implications of these results include the importance of incorporating
positive social cues as useful components in promoting motivation and engagement in health
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S297

behavior change interventions, and highlight a potential shortcoming in eHealth and mHealth
behavioral interventions that are built around online and mobile messaging.

CORRESPONDING AUTHOR: Charles Heidrick, MS, Colorado State University, Fort Collins, CO,
80521; [email protected]
S298 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A164 6:00 PM-7:00 PM

THE ASSOCIATION OF MUNICIPAL OFFICIALS PARTICIPATION IN PHYSICAL ACTIVITY POLICY

AND CITY-LEVEL PHYSICAL ACTIVITY

Kristin Schneider, Ph.D.1, Daniel Kern, M.S.2, Jacey Keeney, M.S.3, Karin Valentine Goins,
MPH4, Stephenie Lemon, PhD4

1
Rosalind Franklin University of Medicine & Science, North Chicago, IL; 2Rosalind Franklin
University of Medicine & Science, Chicago, IL; 3Rosalind Franklin University of Medicine &
Science, Marseilles, IL; 4University of Massachusetts Medical School, Worcester, MA

Municipal officials support of physical activity promotion policies is critical to the

development of communities that encourage physical activity. We examined whether
municipal officials reported participation in policies to increase physical activity was
associated with city-level physical activity. We also examined whether officials endorsement
of greater consideration of physical activity in their work, and their perception of resident
support for addressing physical activity in their position, were associated with physical
activity. Municipal officials from municipalities with > 50,000 residents in 8 states completed a
2012 survey where they indicated their participation (yes/no) in the development, adoption
or implementation of 1) municipal land use policy to increase mixed use, density, street
connectivity, or pedestrian or bicycle access; 2) municipal public works or transportation
policy to increase pedestrian or bicycle safety; and 3) municipal parks and recreation policy to
increase access. They rated their consideration of physical activity in their position and their
perception of resident support for physical activity on 5-pt scales. City-level physical activity
(% engaging in any leisure-time physical activity) from the 2012 Behavioral Risk Factor
Surveillance System (BRFSS) was matched to the officials city. Analyses controlled for officials
job (e.g., mayor, public health official), race/ethnicity, education, city residence, state, use of
active transportation, and degree of social and fiscal conservatism/liberalism. Of 453 officials
with survey data, 161 (68% male) had BRFSS data. Participation in land use policy was the only
policy associated with physical activity [F(1, 144)=5.05, p=.03, 2=.03]. Officials perceptions of
residents support for addressing physical activity was associated with physical activity
(t=3.65,p2= .05); officials reported importance of physical activity to their position was not.
Officials perceptions of residents support for addressing physical activity was associated with
their participation in land use design policy (r=.17, p=.04). Municipal officials focus on land
use policies that encourage physical activity may be key to increasing physical activity rates.
Though counties with greater physical activity participation may appear to be more
supportive of physical activity, encouraging residents to voice their support of environmental
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S299

changes that increase physical activity potential, might persuade officials to consider physical
activity more, and thereby improve public health.

CORRESPONDING AUTHOR: Kristin Schneider, Ph.D., Rosalind Franklin University of Medicine

& Science, North Chicago, IL, 60064; [email protected]
S300 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A165 6:00 PM-7:00 PM

THE ROLE OF ENJOYMENT MODERATES REASONS FOR EXERCISE AND PHYSICAL ACTIVITY

Emily Brooks, B.A.1, Paige Hildreth, B.A.2, Afiah Hasnie, BA2, Heather Fellmeth, MA 2, Caitlin
LaGrotte, PsyD, MEd3, Nataliya Zelikovsky, Ph.D. 4

1
LaSalle University, Philadelphia, PA; 2La Salle University, Philadelphia, PA; 3Temple University,
Philadelphia, PA; 4Psychology, La Salle University, Philadelphia, PA

Various factors have been found to encourage physical activity (PA) in adults, including
exercise enjoyment (EE) and health reasons. Identifying reasons to exercise (RTE) is crucial
since only 20.8% of adults in the United States engage in the CDC recommended PA
guidelines. Research examining enjoyment as a motivation to exercise have provided mixed
results. Individuals underestimate enjoyment before PA and report higher rates of EE after
participating in PA. Additionally, previous research has found gender differences in RTE:
females are more likely to exercise for physical appearance and males are more likely to
exercise for competition and social recognition. Both genders, however, have endorsed
health management as an important RTE. Based on the established relationship between RTE
and PA, our aim was to determine whether enjoyment moderated this relationship above and
beyond gender differences.

Undergraduates (N= 289, 78% female, 61% non-hispanic white, age = 19.7) were surveyed
about weekly exercise engagement, personally motivating reasons for exercise and how much
they enjoy the reported PA. Regression analyses were conducted for the 14 constructs
representing RTE in the Exercise Motivation Inventory. Gender was controlled for based on
existing RTE research. Exercise enjoyment was regressed on amount of PA (days per week)
and RTE revealing a significant moderation for health related RTE: weight management R2 =
.100 (F(2, 286) = .105, p < .05), health pressure R2 = ..233 (F(2, 286) = .10, p < .05), and
avoidance of negative health outcomes R2 = .531 (F(2, 286) = .114, p < .05). The results of this
study suggest that undergraduates who are motivated to exercise for health related reasons
are more likely to exercise when participating in PA that are perceived to be more enjoyable.
Since enjoyment moderates PA for health-motivated individuals, future studies should
identify types of activities that individuals with health conditions find both feasible and
enjoyable to increase exercise compliance.

CORRESPONDING AUTHOR: Emily Brooks, B.A., LaSalle University, Philadelphia, PA, 19104;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S301

A166 6:00 PM-7:00 PM

BARRIERS TO HEALTHCARE ACCESS AND UTILIZATION AMONG US BORN AND IMMIGRANT

AFRICAN AMERICANS, AND LATINAS

Uwaila Omokaro, High School Student1, Mayra Serrano, MPH2, Kimlin T. Ashing, PhD2

1
City of Hope Summer Intern, Duarte, CA; 2City of Hope, Duarte, CA

Background:Health disparities exist and persist for preventive health care and screening
including cancer. African American (AA) women and Latinas have higher later stage cancer
diagnosis and mortality than white women. There is a lack of studies comparing US born and
immigrant African Americans and Latinas on this population healthcare access and
utilization. This study examines healthcare access and utilization barriers among US and
foreign born AA and Latinas.

Methods: We were guided by a community engaged approach. The participants were 395 AA
and Latinas from 18-65 years. Participants were recruited passively via fliers and actively at
health-fairs and through elementary school mailers. Demographic factors including country of
birth, and barriers to Healthcare Access and utilization were assessed

Results: 395 female participants: US born AA (n = 129), foreign born AA (n = 58), US born
Latinas (n = 57), and foreign born (n = 151). Less than half (47.1%) reported that they were
born in the US. Ages ranged from 18-65 years, and the mean age was 42.8 years (SD = 12.93).
Most participants were married or partnered (65%), did not have a college degree (78.2%),
and had an annual income below $25K (50%). AA were significantly less likely to be partnered
and born outside the US, and more likely to have higher educational attainment than Latinas.
AA despite having higher education did not have significantly greater household income.

In general, all groups endorsed moderate to severe barriers to health care access and
utilization.Significant group differences in barriers to healthcare access and utilization were
observed among the four groups. There were significant group differences in barriers to
access to care with respect to no regular place to go to for care (medical home) (p < .001)
and health insurance coverage (p < .001) with 90.1% and 27.2% of immigrant Latinas
reporting no private insurance coverage and no medical home, respectively. Also, there were
significant group differences among the four groups in terms of barriers to health care
utilization: (1) wait too long (p < .05); (2) do not know where to go for medical care (p <
.01); (3) cost (p < .05); (4) language (p < .01); (5) navigator (p < .05); and (6) stress using
healthcare (p < .01). Immigrant Latinas reported the greatest barriers, followed by immigrant
AA, US born Latinas and US born AA.
S302 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusion: Results suggest that barriers to healthcare access and utilization especially among
the Latina immigrant population exist as they reported the highest percentage of people who
experienced each barrier. Multilevel factors like ethnicity and immigrant status among a
group of lower income US born and immigrant population seem to be associated with
healthcare access and utilization barriers among the groups studied.

CORRESPONDING AUTHOR: Uwaila Omokaro, High School Student, City of Hope Summer
Intern, Duarte, CA, 91010; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S303

A167 6:00 PM-7:00 PM

EXAMINING BIOPSYCHOSOCIAL PREDICTORS OF WEIGHT GAIN IN TREATMENT-SEEKING

AFRICAN AMERICAN SMOKERS

Marcia McNutt Tan, PhD, MPH1, Kola Okuyemi, MD, MPH2, Ken Resnicow, Ph.D.3, Noella A.
Dietz, PhD4, Michael Antoni, Ph.D.5, Monica Webb Hooper, PhD6

1
Northwestern University Feinberg School of Medicine, Chicago, IL; 2University of Minnesota,
Minneapolis, MN; 3University of Michigan, Ann Arbor, MI; 4Broward Health Systems, Ft.
Lauderdale, FL; 5University of Miami, Coral Gables, FL; 6Case Comprehensive Cancer Center,
Case Western Reserve University, Cleveland, OH

Research has shown that African Americans gain excessive weight after smoking cessation.
Yet, African Americans have been underrepresented in post-cessation weight gain research.
Few studies have examined 1) the pattern of weight gain and 2) factors associated with
weight gain among African American smokers. The current study aimed to examine
biopsychosocial predictors of weight gain in a sample of treatment-seekers.

Data were drawn from a randomized controlled trial testing the efficacy of a culturally specific
cognitive behavioral therapy (CBT) smoking cessation intervention among African Americans.
Participants (N=342) completed assessments at baseline, the end of counseling, 3-, 6-, and 12-
month follow ups. Baseline measures were salivary cortisol, depressive symptoms, and weight
concern. Weight and self-reported smoking status were measured at all assessments. Random
effects multilevel modeling was used to examine the pattern and predictors of weight gain
over twelve months post CBT. Smoking status was included as a time-varying variable on level
1, and effects of the biopsychosocial variables were examined on level 2. Cross-level
interactions of the biopsychosocial variables and smoking status on weight were explored.

Results revealed that weight significantly increased among those who remained abstinent
over twelve months post CBT. Controlling for intervention condition and baseline obesity,
smoking status positively predicted weight gain in the full sample; abstinence was associated
with increased weight. In a multivariate analysis, male sex and weight concern were positively
associated with baseline weight. Furthermore, exploratory analysis revealed that depressive
symptoms moderated the association between smoking status and weight.

In this sample of African American smokers, weight gain was comparable to previous post-
cessation weight gain research. We also found that psychosocial factors emerged as
important in predicting weight gain. Future research examining predictors of weight gain will
S304 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

inform smoking cessation interventions and help elucidate factors that contribute to tobacco-
and obesity-related health disparities.

CORRESPONDING AUTHOR: Marcia McNutt Tan, PhD, MPH, Northwestern University Feinberg
School of Medicine, Chicago, IL, 60615; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S305

A168 6:00 PM-7:00 PM

EXPLORING MUSLIM WOMENS USE OF CONTRACEPTIONS: PRELIMINARY FINDINGS OF A

SAMPLE SURVEY CONDUCTED IN THE UNITED STATES

Henna Budhwani, PhD, MPH, Kristine Ria. Hearld, PhD

University of Alabama at Birmingham, Birmingham, AL

Introduction: Studies on American, Muslim women are limited, and little is known about their
reproductive health. However, studies on contraception utilization in the general public have
found that socioeconomic and cultural barriers are associated with limited contraceptive
utilization, and the method of contraception used varies significantly when considering
demographics, ethnicity, and marital status. Specifically, 62-75% of American women aged 15-
44 used some form of contraception with oral contraceptive pills being the most
popular. Racial and ethnic minorities areless likely to use contraception and tend to use lower-
efficacy methods more frequently than non-minority, American women. Consequently, racial
and ethnic minorities are twice as likely to experience an unintended pregnancy compared to
non-minority women in the United States. Considering this knowledge gap, the aim of this
study was to explore American, Muslim womens utilization of contraception in general, oral
contraceptive pills, and condoms.

Methods: Self-reported data were collected through an online portal in late 2015 and were
analyzed in 2016. Women who identified as Muslim, were at least eighteen years old, were
sexually active, and were current residents of the United States (N=274) met the inclusion
criteria. Chi-square tests and t-tests were used to assess associations between demographic
characteristics, socioeconomic status, religious characteristics, use of contraception, and type
of contraception.

Results: American, Muslim women utilized contraception at a similar rate found in American
women, in general. Respondents identifying as Sunni Muslims had lower utilization. Ethnicity
was associated with contraception utilization; Middle Eastern or North African Muslims had
lower utilization of oral contraceptive pills and condoms.

Discussion: Findings suggest American, Muslim womens contraception utilization patterns are
complex. Although general contraceptive utilization was high, Muslim sub-groups utilization
of oral contraceptive pills and condoms varied. These disparities may reflect cultural
preferences, knowledge of contraceptive options, or even differences in health care access.
These preliminary findings serve as a foundation for health disparities researchers and public
health practitioners seeking to better understand health behaviors of American, Muslim
S306 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

women. Additionally, although online data collection tools are viewed with some trepidation,
this tool did enable us to reach this particularly stigmatized and hard-to-reach population.

CORRESPONDING AUTHOR: Henna Budhwani, PhD, MPH, University of Alabama at

Birmingham, Birmingham, AL, 35294; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S307

A169 6:00 PM-7:00 PM

RACIAL DISPARITIES IN HEALTHCARE-PROVIDER ADVICE TO QUIT SMOKING

Hope Landrine, PhD1, Irma Corral, PhD2, Jukelia Bess, BS1, Marla Hall, PhD3, Kelly Reburn,
B.A.2, Gavin Locklear, Bachelors 4, Elizabeth Milton, Bachelors 5, Jimmy T. Efird, PhD1

1
Center for Health Disparities ECU, Greenville, NC; 2East Carolina University, Greenville, NC;
3
Brody School of Medicine, East Carolina University, Greenville, NC; 4East Carolina University
Center for Health Disparities, Greenville, NC; 5Center For Health Disparities, Greenville, NC

African-Americans (AAs) have a higher prevalence of smoking-related diseases than Whites

even though they smoke fewer cigarettes. This is in part because AAs are significantly less
likely than Whites to quit smoking. Being advised to quit by a healthcare provider (HCP) is a
robust predictor of smoking cessation, but AAs are significantly less likely than Whites to
receive such advice. Health insurance, regular doctor, age, gender and education are the five
major predictors of HCP cessation advice (HCP-CA) but do not explain racial disparities in it.
We examined other variables that also may contribute to racial disparities in HCP-CA.
Specifically, we included the major correlates of race and examined their effect on the
contribution of race to HCP-CA. Our random, southern sample of 1670 AA and White adults
included 512 smokers, and no racial difference in smoking prevalence (Whites = 32.7%, AAs =
31.7%; X2 = 0.130, p = .718). A significantly larger percentage of White (74%) than AA (55%)
smokers had ever-received HCP-CA (X2 = 13.869, p = .0005). The first hierarchical logistic
regression included the 5 major predictors (above) and race entered on the last step. Health
insurance and education were not significant, but age (older), regular doctor (yes), and gender
(woman) increased the odds of HCP-CA. After controlling for those 5 variables, Whites
remained (Odds Ratio [OR]) 2.6 times more likely than AAs to receive HCP-CA. An additional
logistic regression included co-morbidities (hypertension, diabetes, obesity) and religiosity
(both strong correlates of being AA), with race again entered on the last step. Adding these
variables improved model fit (Cox & Snell R2) but increased (rather than decreased) racial
disparities in HCP-CA: After controlling for all of these variables, Whites were (OR) 9.3 times
more likely than AAs to receive HCP-CA. These data suggest that AAs who have comorbidities
receive HCP-CA less often than Whites but do indeed receive it, whereas AAs who lack
comorbidities are extremely unlikely to ever receive HCP-CA. Chi-square analyses of the AA
smokers supported this interpretation, and thereby highlight the need to improve HCP
compliance with federal policies requiring them to assess and treat tobacco use.
S308 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Hope Landrine, PhD, Center for Health Disparities ECU, Greenville,
NC, 27858; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S309

A170 6:00 PM-7:00 PM

SHOP-BASED HEALTH PROGRAM PLANNING: PERCEPTIONS AMONG RURAL OWNERS USING A

FORMATIVE RESEARCH APPROACH

Marla B. Hall, PhD1, Jukelia Bess, BS2, Hope Landrine, PhD2, Irma Corral, PhD1, Jimmy T. Efird,
PhD2

1
East Carolina University, Greenville, NC; 2Center for Health Disparities ECU, Greenville, NC

African American barbershops and beauty salons are settings that have been identified as a
significant and culturally relevant venue to reach minority populations for health promotion
activities. By being located in almost every town in the United States, this setting is a viable
means to promote healthy lifestyles among African Americans. The purpose of this formative
research project was to assess African American barbershop and beauty salon owners
perceptions of providing health promotion programming in their shops, as well as to obtain
information on health topics of interest and strategies for implementation. Interviewees were
recruited using snowballing among clientele and owner referrals, between November 2014
and August 2015. A total of 20 barbershop and salon owners, across 11 counties in eastern
North Carolina, completed face-to-face interviews. Responses were stratified by barbershops
and beauty salons. Across both groups, all owners stated it would be a good idea to have
health programs/interventions within the shop setting. Most noted topics of interest included
diet and nutrition, hypertension and (wo)mens reproductive health. When asked further
about these desired topics, both benefits and relevance to customers and the African
American community were the reasons for their selections. In addition, across barbershops
and salons, 90% of owners stated interest in having a program implemented in their shop.
This information will be used to guide the development of shop-based interventions, with the
aid of a community advisory board composed of shop owners, individual barbers and stylists;
and customers.

CORRESPONDING AUTHOR: Marla B. Hall, PhD, East Carolina University, Greenville, NC,
27834; [email protected]
S310 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A171 6:00 PM-7:00 PM

STRIKING NATIONAL-LEVEL DISPARITIES IN LGB TOBACCO USE

Kristen Emory, PhD1, Dennis Trinidad, PhD2

1
Moores Cancer Center at University of California, San Diego & San Diego State University, San
Diego, CA; 2Moores Cancer Center at University of California, San Diego & San Diego State
University, La Jolla, CA

Purpose: Despite recent increases in attention to lesbian, gay, bisexual and transgender
(LGBT) issues, tobacco use among racial/ethnic LGBT populations remains understudied,
particularly in large nationally representative samples. This study seeks to address this issue
by exploring and identifying patterns in tobacco use across diverse LGB populations.

Methods: We utilized public use data from the Population Assessment of Tobacco Health
(PATH), a nationally representative US sample of adults ages 18+ (N=31,081), exploring for
differences in tobacco use across self-reported sexual orientation, including 29,369
heterosexual/straight and 1,712 lesbian, gay or bisexual (LGB) respondents. We explore past
30 day use of cigarettes and e-cigarettes. Race/ethnicity was dichotomized to Non-Hispanic
white (NHW) and racial/ethnic minorities (People of Color POC) due to sample size
limitations. We report basic descriptive data and adjusted logistic regression.

Results: Past 30-day use of both cigarettes and e-cigarettes was higher among LGB
participants (15.8% and 52.9%, respectively) compared to non-LGB participants (11.8% and
34.8%, respectively). Racial/ethnic differences were observed in past 30-day use of cigarettes
and e-cigarettes among LGB and heterosexual/straight populations in both the unadjusted
and adjusted analysis. White LGB (OR=1.2, 95%CI:1.0-1.5), Heterosexual POC (OR=1.8,
95%CI:1.7-2.0), and LGB POC (OR=2.5, 95%CI:2.1-3.0) reported higher odds of past 30-day
cigarette use, compared to NHW heterosexual participants. Similarly NHW (OR=2.0,
95%CI:1.7-2.2) and POC (OR=1.2, 95%CI:1.0-1.4). LGB had higher odds of past 30-day e-
cigarette use, when compared to NHW heterosexuals; while heterosexual POC had lower odds
(OR=0.7, 95%CI:0.6-0.7).
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S311

Conclusions: These striking differences in tobacco use across race/ethnicity and sexual
orientation groups suggest there may be within population disparities in the risk for tobacco
among diverse LGB populations. This preliminary evidence supports the idea that being a
member of a multiple minority group (e.g. both a racial/ethnic minority and sexual minority)
may play an important role in tobacco use among diverse LGB. Further research on the role of
multiple minority stress is needed to examine this relationship with diverse LGB tobacco use.

CORRESPONDING AUTHOR: Kristen Emory, PhD, Moores Cancer Center at University of

California, San Diego & San Diego State University, San Diego, CA, 92101;
[email protected]
S312 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A172 6:00 PM-7:00 PM

THE LIVED EXPERIENCE OF INFORMAL AFRICAN AMERICAN CAREGIVERS OF FAMILY MEMBERS

WITH ALZHEIMERS DISEASE AND RELATED DEMENTIA

Dell G. Mars, PhD, RN

Southeastern Louisiana University, Baton Rouge, LA

The aim of this study was to explore the lived experience of informal African American (AA)
caregivers of family members with Alzheimers Disease and Related Dementias (ADRD). A
descriptive phenomenological study was used to conduct semi-structured interviews with 16
AA informal caregivers in Southeastern United States. Data was collected during the months
of September 2014 to November 2014. The descriptive phenomenology ofHusserl and the
methodological interpretations of Colaizzis method of analysis underpin this study. Four
themes were identified in this analysis: (a) sense of obligation, (b) arduous journey, (c)
sentinel events, and (d) faith in God. Results of this study suggest that informal caregivers lack
support, knowledge, and guidance.

CORRESPONDING AUTHOR: Dell G. Mars, PhD, RN, Southeastern Louisiana University, Baton
Rouge, LA, 70815; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S313

A173 6:00 PM-7:00 PM

CONDOM CARNIVAL EVALUATION: COMPARISON OF A NOVEL SEXUAL RISK REDUCTION

INTERVENTION TO "BUSINESS AS USUAL" AND A GOLD STANDARD

Mollie B. Anderson, PhD1, Idia B. Thurston, PhD2, Theresa Montgomery. Okwumabua, Ph.D.1

1
The University of Memphis, Memphis, TN; 2University of Memphis, Memphis, TN

College students frequently report not using condoms, placing them at risk for unplanned
pregnancy and sexually transmitted infections, including HIV. This study aimed to investigate
the preliminary efficacy, acceptability, and feasibility of The Condom Carnival, a novel, brief,
interactive, culturally-tailored, and peer-led sexual risk reduction group intervention for
college students. A longitudinal, randomized controlled trial was utilized to compare the
efficacy of the Condom Carnival to an education-only control condition (HIV/STI 101) and a
treatment control condition (VOICES/VOCES, a CDC effective behavioral intervention). To
encourage college students to increase their condom use and lower their sexual risk, the
Condom Carnival has three specific aims: 1) address knowledge deficits in sexual health
information, 2) improve condom-related self-efficacy, and 3) increase awareness of risky
sexual behaviors. Due to the interactive, skills-based, and peer-led nature of the Condom
Carnival, we hypothesized that participants would report greater efficacy and acceptability of
the Condom Carnival compared to the other interventions. Undergraduate and graduate
students were trained as Condom Carnival peer-facilitators.119 undergraduates, aged 18-57
years (M=21.8), were recruited for this study. Students were 77% Female, 52% Black, 42%
White, 6% Latino, and 1% Asian. All questionnaires (pre-, post-, and follow-up) were
administered online. Two-way Mixed ANOVAs, McNemars tests, and a One-way ANOVA were
used to examine the interventions comparative efficacy and acceptability; frequencies were
examined to determine the feasibility of Condom Carnival activities. The Condom Carnival had
equivalent acceptability and efficacy as VOICES/VOCES in teaching sexual health information
(HIV and lubricant safety knowledge), improving facets of condom-related self-efficacy
(condom negotiation strategies and correct condom use skills), and increasing awareness of
risky sexual behaviors (lowering number of sexual partners, decreasing general sexual risk,
and increasing safe sex behaviors). All Condom Carnival participants engaged in every activity,
thus displaying excellent feasibility. The Condom Carnival, with its scalability, has utility for
teaching college students sexual risk reduction and condom use promotion. This study is
promising for intervention researchers, community preventionists, and campus service
providers.
S314 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Mollie B. Anderson, PhD, The University of Memphis, Memphis,

TN, 38152; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S315

A174 6:00 PM-7:00 PM

HIGH PREVALENCE OF OVER THE COUNTER SLEEP AID USE AMONG INDIVIDUALS WHO HAVE
EXPERIENCED INVOLUNTARY JOB LOSS

Patricia Haynes, PhD1, Graciela E. Silva, PhD2, Devan Gengler, B.S.1, Darlynn Rojo-Wissar, BA1,
Uma Nair, PhD1, Rachel Oliver, N/A1, David Glickenstein, PhD1, Duane Sherrill, PhD1, Stuart
Quan, MD3

1
University of Arizona, Tucson, AZ; 2College of Nursing / University of Arizona, Tucson, AZ;
3
Harvard Medical School, University of Arizona, Tucson, AZ

Sleep is a barometer of mental and physical health, highly influenced by environmental

demands. Stressful life events, such as job loss, are thought to precipitate insomnia. Few
studies have directly examined sleep or treatment patterns for disrupted sleep after stressful
events. As part of recruitment for an ongoing, large-scale, prospective study examining sleep
and involuntary job loss, we sent flyers and conducted phone screens with individuals
applying for unemployment insurance through the Arizona Department of Economic Security.
A total of 399 individuals reporting involuntary job loss called the study line and participated
in the initial phone screen (M age = 41.69 years, SD = 11.68 years). On average, callers lost
their job approximately 39.60 days previously (SD = 53.86 days). A significant number of
callers (29%, n = 116) reported having recently taken drugs or medications to help them sleep
or relax. Of these callers, the majority reported using only one drug/medication for sleep
(90%, n = 104). The most frequently used drug or medication class was over the counter (OTC)
sleep or anti-inflammatory medications (8%), followed by marijuana (6%), and melatonin
(3%). Twelve percent of the sample (n = 48) reported taking prescription medications for sleep
(3% tricyclic antidepressants, 3% pain medications, 3% benzodiazepines, 2% hypnotics, 1%
antipsychotics/other). Only one individual reported recently engaging in Cognitive Behavioral
Therapy for Insomnia, the gold standard behavioral treatment for insomnia. Overall, these
findings suggest that a significant portion of individuals experiencing job loss independently
seek treatment for sleep symptoms and often via OTC sleep medication and marijuana use.
Future studies are recommended to investigate barriers to accessing evidence based sleep
care and potential side effects of OTC sleep aids among people who have lost their jobs.

CORRESPONDING AUTHOR: Patricia Haynes, PhD, University of Arizona, Tucson, AZ, 85724-
5209; [email protected]
S316 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A175 6:00 PM-7:00 PM

THE SCENT OF A GOOD NIGHTS SLEEP: OLFACTORY CUES OF A ROMANTIC PARTNER

INCREASE SLEEP EFFICIENCY

Marlise Hofer, Masters in Psychology, Frances Chen, Psychology PhD

University of British Columbia, Vancouver, BC, Canada

Sleep deficits are widespread and lead to a variety of negative health outcomes. We present
novel evidence for the effectiveness of a practical and easy-to-implement strategy for
increasing sleep efficiency. Forty women slept with a shirt which had been previously worn by
their romantic partner for two nights, and an unworn shirt for two nights. Participants sleep
efficiency was measured each night using an actigraphy watch, and their perceived sleep
quality was measured each morning via self-report. Exposure to a romantic partners scent
increased womens sleep efficiency by an average of 6.9%, equating to over 24 minutes per
night, or almost 3 hours a week. This effect size is larger than the effect size reported for
melatonin, and similar to the effect size reported for benzodiazepines. The effect did not
appear in perceptions of sleep quality, and emerged among people who were not aware they
were sleeping with their partners scent, indicating that this phenomenon occurs outside of
conscious awareness. These findings suggest that human scent may be an under-examined
but powerful means through which cues to a loved one can exert positive effects even during
periods of physical separation.

CORRESPONDING AUTHOR: Marlise Hofer, Masters in Psychology, University of British

Columbia, Vancouver, BC, V6K 2A6; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S317

A176 6:00 PM-7:00 PM

A QUALITATIVE ASSESSMENT OF BARRIERS TO ADEQUATE HEALTHCARE AMONG MEN WHO

HAVE SEX WITH MEN RESIDING IN RURAL AREAS

Danielle N. Lambert, MPH, CHES1, Natalia Truszczynski, MPH2, Anne Marie Schipani-
McLaughlin, MPH2, Carolyn Lauckner, PhD2, Nathan Hansen, PhD2

1
University of Georgia, Atlanta, GA; 2University of Georgia, Athens, GA

Background: Men who have sex with men (MSM) are most affected by the HIV epidemic in
the United States, accounting for 63% of all new infections. Although research has historically
demonstrated that rural areas have significantly more barriers to HIV testing and treatment,
little is known about the impact of rural residence on MSMs sexual health and access to care.
The current qualitative study aimed to explore barriers to adequate healthcare among MSM
in rural areas of the United States.

Methods: A total of 20 semi-structured qualitative interviews were conducted from July 2015
to June 2016 with respondents ranging in age from 18 to 60 years. Eligibility criteria included
identifying as a man who dates or has sex with other men, residing in a non-metropolitan area
of the United States, and being at least 18 years of age. The interviews were transcribed
verbatim and coded by a diverse coding team for emerging themes based on Grounded
Theory using NVivo 11.

Results: Respondents repeatedly discussed limited resources, distance, provider knowledge,

and stigma as barriers to healthcare in the rural areas where they reside. For many
respondents, barriers to healthcare impeded access to needed sexual education, testing, and
PrEP. Often, they turned to the Internet, peers, or sexual partners for health information,
which were not always accurate sources of information. Not only did respondents speak to
the inadequate number of clinics and providers within driving distance, but they also
described how their insurance failed to cover all the local providers in-network. Respondents
living in small towns with only one provider, who they described as closely connected with
residents in the area, feared seeking care or openly communicating with this provider due to
stigma and a perceived lack of confidentiality. For respondents who were able to access care
and disclose their sexual orientation, many recounted negative experiences in which providers
said they were not able to answer their questions, that LGBTQ individuals was not their
specialty, or as one respondent described, it got weird so [they] referred me to someone
else.
S318 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions: The findings from this study can aid in creating tailored programming to increase
the accessibility of care, improve provider knowledge, and decrease discrimination.
Respondents provided recommendations for creating more inclusive clinical environments
and ways to improve provider communication to increase the quality of healthcare.

CORRESPONDING AUTHOR: Danielle N. Lambert, MPH, CHES, University of Georgia, Atlanta,

GA, 30322; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S319

A177 6:00 PM-7:00 PM

A GOOD RESULT: COLORECTAL CANCER SCREENING AND NARRATIVES OF RISK IN A

PHOTOGRAPHIC RESEARCH PROJECT

Robert J. Gallo, MD Candidate1, Jean Hunleth, PhD, MPH1, Emily Steinmetz, PhD2, Aimee
James, PhD, MPH3

1
Washington University School of Medicine, Saint Louis, MO; 2Antioch College, Yellow Springs,
OH; 3Division of Public Health Sciences, Washington University School of Medicine, Saint
Louis, MO

Cancer screening rates have been increasing, with positive impacts on mortality for several
types of cancer. While screening is generally thought of as a healthy behavior, there has been
some study of the overall effects of breast cancer screening on womens lives and
understandings of their identities. In comparison to breast cancer, there are fewer analyses of
how people make meaning from their colon cancer screening experiences. There are also few
studies of how people who face resource shortages and healthcare disparities experience and
interpret cancer screening when cancer is not found. Of particular interest is consideration of
how people who have had a colonoscopy and who do not have colon cancer conceptualize
risk and how that influences their health beliefs.

For this study, 18 participants across three groups were recruited for a Photovoice study that
involved a combination of photo-elicitation, individual interviews, and group sessions. Eligible
participants had completed CRC screening and had not been diagnosed with colon cancer. The
sessions followed a format in which people presented a photo of their choosing, frequently
offering a personal story that they related to the photo. All group and individual sessions were
audio-recorded and professionally transcribed and uploaded into NVivo 10. Transcripts from
the group sessions were analyzed using narrative analysis techniques (Reissman 1993, 2008).

Most of the participants described colon cancer screening as a life-saving intervention, and
they were eager to learn more about colon cancer and its prevention both for themselves and
to share with their communities. This optimistic picture was complicated though by many of
the narratives that they shared. One participant was reluctant to become reliant on medicine
due to concerns about financial stability. The colonoscopy could come back with a cancer
diagnosis that she may not be able to afford in the future. Others recounted ambiguity over
the findings of the screening despite not receiving a cancer diagnosis. Several participants
shared narratives in which their concern about their cancer risk was clearly evident, a concern
that extended to their friends and family as well.
S320 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

While the participants where all healthy in the sense that their screening colonoscopy did
not show colorectal cancer, this analysis demonstrates that colonoscopy took on meaning and
had consequences outside the clinic and within other health experiences. These
consequences varied among individuals, but the participants highlighted the multitude of
ways that even an effective, life-saving intervention can be experienced by a patient.
Ultimately, this study speaks to the need to understand the contextual factors and
experiences that shape peoples decisions to receive cancer screening and the meaning that
such screening has in their lives, even when cancer is not found.

CORRESPONDING AUTHOR: Robert J. Gallo, MD Candidate, Washington University School of

Medicine, Saint Louis, MO, 63108; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S321

A178 6:00 PM-7:00 PM

DEVELOPING NEED INDICES TO UNDERSTAND AND ADDRESS SOCIAL DETERMINANTS OF

HEALTH

Rachel Barth, MSW, Nikole Lobb Dougherty, MA, Emily Kryzer, MSW, MPH, Jasmine L.
Williams, MPH, Jason Purnell, PhD, MPH

Washington University in St. Louis, St. Louis, MO

Background: Social determinants of healthincluding such factors as poverty, education, and

neighborhood qualityhave the greatest impact on the health and well-being of individuals.
Evidence suggests that investing in strategies that improve equity in social conditions is
essential to reducing community health disparities and supporting individual health behaviors.
For example, there is evidence that financial assets like child development accounts is
associated with better development for children and lower levels of depressive symptoms for
mothers. Similarly, protecting fair housing policy ensures equitable access to safe
neighborhoods, thus promoting healthy lifestyles.

For the Sake of All, a report and multi-disciplinary project on the health and well-being of
African Americans in St. Louis, highlights how various social and economic factors impact
regional disparities and supports cross-sector strategies to address disparities.

Methods: Eight cross-sector work groups convened to address the six areas of
recommendation highlighted in the report. Over 100 stakeholders and community members
identified a need for using community-level data to better understand and prioritize where
policy and programmatic efforts should focus. Four of eight individual work groups pursued
the development of recommendation-specific need indices.

Results: Following national methodological examples, three different need indices have been
developed and used to answer prioritization questions in the following areas related to social
determinants of health: 1) A ranking based on birth rates, child poverty rates, and college
completion by zip code was developed for child development accounts; 2) a ranking based on
area socioeconomic status, health, and social support indicators was developed to determine
need for school-based clinics; 3) and an analysis of zoning and land use among St. Louis metro
municipalities revealed areas of opportunity for improving access to inclusive, affordable
housing. These are being used by work groups to prioritize programmatic and advocacy
efforts to improve community health.

Discussion/Conclusion: The use of work group expertise to guide the development and
S322 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

application of community-level need indices offers a framework by which other communities

may work collaboratively to prioritize regional efforts to reduce disparities. This framework is
particularly useful within the resource-limited and politically-complex environment of public
health research.

CORRESPONDING AUTHOR: Rachel Barth, MSW, Washington University in St. Louis, St. Louis,
MO, 63130; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S323

A179 6:00 PM-7:00 PM

ETHNIC AND GENDER DIFFERENCES IN PERCEIVED SOCIAL SUPPORT DURING EMERGENCY

ROOM EVALUATION FOR ACUTE CORONARY SYNDROME

Gabriel Sanchez, B.A.1, Deanna Zhu, B.A.1, David Lopez Veneros, M.A.1, Redeana Umland,
M.A.1, Donald Edmondson, PhD, MPH2

1
Center for Behavioral Cardiovascular Health, Columbia University Medical Center, New York,
NY; 2Columbia University Medical Center, New York, NY

Background: Emergency department (ED) evaluation for acute coronary syndrome (ACS;
myocardial infarction, unstable angina) is common and highly stressful. Social support is
important for ACS outcomes. Hispanic or Latino (HL) patients report more perceived social
support than other ethnicities, but no study has assessed ethnicity and gender effects on
perceived social support during life-threatening medical situations.
Methods: Within a 3 day median of ED ACS evaluation, 508 participants (age 611; 33.3% HL
women, 17.7% non-HL women, 24.8% HL men, 24.2% non-HL men) reported the degree to
which the social support provider present with them during ED evaluation was responsive to
your needs and able to comfort you [scored from 1 (none) to 5 (all of the time)].
Analysis of covariance tested the influence of gender, ethnicity and gender*ethnicity on
responses (covariates:age, race, Charlson Comorbidity Index, GRACE Risk Score).
Results: There was a significant interaction between gender and ethnicity on perceived
responsiveness, F(1,499)=4.12, p=.043, p2 =.008. There were no main effects for gender (p
=.962) or ethnicity (p =.146). HL men reported their support provider was more responsive
(M=4.90, SE=.050) than HL women (M=4.80, SE=.045), non-HL women (M =4.82, SE=.063) and
non-HL men (M=4.72, SE=.051). There was a significant main effect of ethnicity on providers
ability to comfort patients F(1,497) = 4.03, p=.014, p2=.012 [gender (p=.773)
gender*ethnicity (p=.091)]. HL patients reported more received comfort (M=4.74, SE=.05)
than non-HL patients (M=4.53, SE=.062). Though not statistically significant, HL men reported
more received comfort (M=4.81, SE=.073) than non-HL men (M=4.48, SE=.076). HL women
reported feeling comforted slightly more (M=4.66, SE=.066) than non-HL women (M=4.58,
SE=.094).
Conclusion: HL men reported the most social support received while in the ER, followed by
both HL and non-HL women, then non-HL men. Prior studies suggest more social support may
improve health outcomes. These are the first data to address gender and ethnic differences in
received support during life-threatening medical events.
S324 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Gabriel Sanchez, B.A., Center for Behavioral Cardiovascular

Health, Columbia University Medical Center, New York, NY, 10032;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S325

A180 6:00 PM-7:00 PM

FIRST GENERATION COLLEGE STUDENT STATUS, ACCULTURATION STRESS, AND

PSYCHOSOCIAL RESOURCES PREDICT ANXIETY AND DEPRESSION

Katherine Saxton, MPH, PhD, Mariam Khan, BS, Morgan Vien, N/A

Santa Clara University, Santa Clara, CA

Background: Mental health of college students is an important public health concern. College
students in the US report high levels of stress, anxiety, and depression, which may lead to
decreased academic performance and quality of life. First generation college students may
experience increased risk, due to lack of academic and social support. We examined the
prevalence of depression and anxiety in college students, risks associated with first generation
college student status, and potential psychosocial mediators.

Methods: Students at a Northern California university (n=967) completed a survey, including:

a demographic questionnaire, measures of self-efficacy, social support, acculturation stress,
subjective social status, and parent/guardian relationship quality, as well as depression (PHQ-
9) and anxiety (GAD-7) screening measures. The sample included 28% first generation, 65%
female, and 40% non-white college students. Data were analyzed using T-tests and
hierarchical linear regression. Depression and anxiety scores were square-root transformed to
meet regression assumptions.

Results: 25% of students reported moderate or more severe depression, and 27% of students
reported at least moderate levels of anxiety. Women reported more severe symptoms of both
anxiety (T=3.42, df=912, p < 0.001) and depression (T=2.23, df=922, p < 0.05) than men. After
adjusting for race and sex, first generation college students reported increased symptoms of
depression and anxiety (ps < 0.01) relative to continuing generation students. However, this
effect was fully mediated by acculturation stress (ps < 0.001) for both outcomes. In the fully
adjusted model, self-efficacy, social support, and parental caring were inversely associated
with depression and anxiety symptoms (ps < 0.05). Subjective social status was inversely
associated with anxiety (p < 0.05), while overprotective parenting was positively correlated
with depression (p < 0.05).

Conclusions: First generation college students are at increased risk of anxiety and depression,
due to acculturation stress. Interventions should increase support for first generation college
students, and promote coping strategies, enhance social support, and increase access to
mental health resources for all students.
S326 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Katherine Saxton, MPH, PhD, Santa Clara University, Santa Clara,
CA, 95053; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S327

A181 6:00 PM-7:00 PM

IN THEIR OWN WORDS: THE IMPACT OF STRESS ON CHRONIC DISEASE RISK FACTORS AMONG
AFRICAN-AMERICAN CHILDCARE WORKERS

Kia Davis, ScD, MPH1, Laura Linnan, ScD2

1
UNC-Chapel Hill, Durham, NC; 2UNC-CH, Chapel Hill, NC

Background: Stress, a potential mechanism explaining racial/ethnic and socioeconomic

disparities in health, is socially patterned in the U.S. such that people of color, particularly
African-Americans, and those of low SES are likely to experience more stressors. Yet, little
attention is given to what stressors these groups experience and how these experiences
impact their health risk factors like smoking status or physical activity. The purpose of this
study is to quantitatively assess different sources of stress and health risk among African-
American women who work in childcare, a low-wage industry and qualitatively include their
perceptions of stress and its health impact.

Methods: This was amixed methodspilot studyconducted across 15 childcare centers in North
Carolina during June 2016. Women (n=120) were surveyed about stress (job, financial strain,
perceived stress, etc.) and mental and physical health. Descriptive, bivariate, and regression
analyses were conducted in SAS. A subset of the women (n=38) also participated in semi-
structured interviews that covered: identifying various sources of stress they experience; their
greatest source of stress; how stress impacts their health, and recommendations for
policymakers to reduce the amount of stress they experience. Transcribed interviews were
analyzed using a deductive and inductive approach.

Results: Results suggest that reported perceived stress are low. Also, while women discuss
financial strain as a major source of stress in interviews; only 15% report not having enough
to make ends meet yet 43% reported having to borrow money from friends/family. Only 8%
report being current smokers. Similarly from interviews, women discuss how experiencing
multiple stressors leads to unhealthy eating, headaches, disrupted sleep, and high blood
pressure, but also walking with friends to reduce stress. They also articulate that making more
money will help reduce the stress they experience in all domains.
S328 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions/Lessons learned: Stress could trigger participation in health enhancing or

damaging behaviors as a form of coping. Overall, addressing fundamental causes of stressors
could reduce the disproportionate burden of health risk factors among African-American
women. However, researchers should consider the meaning of the stress measures they use
in order to adequately capture the reality of the communities they survey.

CORRESPONDING AUTHOR: Kia Davis, ScD, MPH, UNC-Chapel Hill, Durham, NC, 27713;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S329

A182 6:00 PM-7:00 PM

PATHWAYS TO HARM: ALCOHOL USE PATTERNS BY SEXUAL ORIENTATION

Alyssa Norris, M.A.1, David Marcus, Ph.D.2

1
Brown University, Providence, RI; 2Washington State University, Pullman, WA

Although lesbian, bisexual, and questioning individuals are more likely to experience negative
consequences from alcohol use than heterosexual individuals, they also tend to use alcohol in
greater amounts. Researchers have not examined whether substance use is associated with
more harmful outcomes controlling for level of use. Using the American College Health
Associations National College Health Assessment (N = 43,252), the present study examined
whether sexual identity moderated the relationship between frequency of alcohol use and
alcohol-related negative consequences. Consistent with existing literature, lesbian and
bisexual women were more likely to use alcohol than heterosexual women, although the
pattern of alcohol use was more variable for men. Sexual identity moderated the pathway
from alcohol use frequency to alcohol-related harm. Bisexual women and individuals who
were unsure of their sexual identity had stronger associations between use and harm
compared to their heterosexual peers. A lesbian identity appeared to protect students from
harm.

CORRESPONDING AUTHOR: Alyssa Norris, M.A., Brown University, Providence, RI, 02906;
[email protected]
S330 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A183 6:00 PM-7:00 PM

PERSONALITY TRAIT INTERACTIONS AND SUBJECTIVE WELL-BEING THROUGHOUT THE

LIFESPAN

Katherine A. Duggan, PhD1, Chandra A. Reynolds, PhD2, Howard S. Friedman, PhD2

1
University of Pittsburgh, Pittsburgh, PA; 2University of California, Riverside, Riverside, CA

High conscientiousness is related to decreased mortality risk, in part because of salubrious

health behaviors and healthy social support networks. High neuroticism is associated with
decreased subjective well-being and poorer self-rated health, but there is dispute as to
whether this reflects actual health problems or heightened sensitivity to somatic symptoms
without a true increase in measurable disease. Combinations of conscientiousness and
neuroticism, however, have received relatively little attention, and no study has examined
associations between these combinations and well-being. Here, we tested the concept of
healthy neuroticism (Friedman, 2000), which proposes that the vigilance associated with
neuroticism is health protective when combined with the responsibility and persistence
associated with conscientiousness. Participants were from the Terman study, following 1528
gifted children since 1921. Multiple regressions examined whether healthy neuroticism
(modeled via simple linear interactions and multiplicative linear spline terms) in adulthood
(1940) was related to multifactorial well-being (1940-1950), including measures of self-rated
physical health, self- and observer-rated mental health, living up to potential, extent enjoying
social relationships, and satisfaction with life.

For men, linear interactions were found for extent enjoying social relationships (B=-0.50,
p=.01), such that men high in conscientiousness and neuroticism reported less enjoyment of
social relationships. Spline interactions were found for self-rated health for men (B=-0.62,
p=.01), observer-rated mental-adjustment (B=-0.44, p=.009), and satisfaction with life (B=-
1.28, p=.05), such that men high in conscientiousness and neuroticism scored lower on these
variables. For females, the only significant interaction was the spline for living up to potential
(B=-1.69, p=.006), such that women high in conscientiousness and neuroticism reported
feeling less likely to have lived up to their potential. Consistent with distinctions between
subjective and objective health, the results suggest that healthy neuroticism is not
predictive of a sense of well-being, particularly for men. These results have implications for
understanding the associations between personality and well-being, as contrasted with
mortality risk.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S331

CORRESPONDING AUTHOR: Katherine A. Duggan, PhD, University of Pittsburgh, Pittsburgh,

PA, 15213; [email protected]
S332 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A184 6:00 PM-7:00 PM

SELF-EFFICACY AND AFFECTIVE EVALUATIONS: PARALLEL MEDIATING EFFECTS ON WEIGHT

LOSS SUCCESS

Sean Rice, BS1, Jordan Vossen, BS2, Joyce Ehrlinger, PhD2, Renee E. Magnan, PhD3

1
Washington State University, Vancouver, WA; 2Washington State University, Pullman, WA;
3
Washington State University Vancouver, Vancouver, WA

Both self-efficacy and affect consistently influence weight loss success. Less is known about
how weight loss may affect self-efficacy and affect, or how these variables may have
reciprocal deterministic relationships with each other. We expected successfully losing weight
would promote feelings of self-efficacy and positive affect and, together, these factors would
lead to continued progress towards weight loss goals. In each of 8 weekly meetings within a
structured weight loss program, participants (N=51) were weighed, and they reported their
confidence in achieving their goal weight and their affective evaluations of their progress
towards that goal. Statistical analyses were conducted using structural equation modeling
with maximum likelihood estimation. Cumulative weight loss by the program midpoint
significantly predicted midpoint positive affective evaluations (b =.26, SE=.13, p=.047). Affect
then predicted weight loss at the end of the program (b =.59, SE = .14, p < .001). The results
revealed an indirect effect of midpoint weight loss on weight loss at the end of the program,
through positive affective evaluations of goal progress, b =.15, 95% CI [.01, .43]. This suggests
that greater weight loss predicted greater positive affect at the midpoint, which, in turn,
predicted increased weight loss by the end of the program. Contrary to our hypothesis, self-
efficacy at the beginning of the program did not predict weight loss or affective evaluations in
later weeks, nor did it function as a mediator in parallel with affect. However, positive affect
did emerge as a significant mediator between earlier weight loss and later weight loss. It may
be that affective processes are particularly important in the continued success of weight loss,
as it was both influenced by and predictive of weight loss throughout the program. Overall we
found preliminary evidence of a possible reciprocal relationship between weight loss and
affect, but no such association was found with self-efficacy.

CORRESPONDING AUTHOR: Sean Rice, BS, Washington State University, Vancouver, WA,
98661; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S333

A185 6:00 PM-7:00 PM

SOCIAL INFLUENCES ON SMOKING IN THE SOCIAL NETWORKS OF ADULTS WITH MENTAL

ILLNESS ENROLLED IN CESSATION TREATMENT

Kelly Aschbrenner, PhD1, John Naslund, PhD Candidate2, Emily Schneider, Undergraduate
Student3, A. James O'Malley, PhD4

1
Geisel School of Medicine at Dartmouth, Nashua, NH; 2The Dartmouth Institute for Health
Policy and Clinical Practice, Lebanon, NH; 3Dartmouth College, Hanover, NH; 4The Dartmouth
Institute for Health Policy and Clincial Practice, Lebanon, NH

Introduction: Despite significant decreases smoking rates in the US over the past 50 years, an
estimated 50-85% of adults with serious mental illness (SMI), such as schizophrenia, bipolar
disorder, and major depression, smoke cigarettes compared to 16.8% of adults in the general
population. Individuals with SMI generally have social networks of people similarly affected by
mental illness and high rates of addiction. Whether and how these networks encourage or
limit smoking cessation in this group has received little attention . The purpose of this study
was to: (a) describe smoking behaviors in the social networks of individuals with SMI; and (b)
explore participants perceptions of how these networks influence their efforts to quit
smoking.

Methods: Participants were 41 individuals with SMI who participated in a state Medicaid
demonstration project of smoking cessation at public mental health centers. The study sample
consisted of quitters (n=11) and non-quitters (n=30) following treatment. Data collected
during cross-sectional egocentric social network interviews was used to estimate proportions
of current smokers, former smokers, and never smokers in participants networks. In addition,
a series of open-ended qualitative questions explored participants perceptions of the
influence of individuals in their network (i.e., alters) on their attempts to quit smoking. Alters
were defined as family members, friends, peers, significant others, and non-professionals with
whom participants spent the most time with in the past 12 months.

Results: Most alters identified by participants were friends (45%) or family members (44%). In
the study sample, 53% percent of alters were current smokers; 19% former smokers; and 28%
had never smoked. These proportions did not differ significantly by participants smoking
status following cessation treatment. Participants reported that pro-smoking social norms
and attitudes, and alters smoking habits were barriers to quitting smoking. Social facilitators
to quitting included having cessation role models, quitting together with alters, and receiving
emotional and practical support from alters for quitting smoking.
S334 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions: Similar to the general population, social factors appear to influence efforts to
quit smoking among individuals with SMI enrolled in cessation treatment. A better
understanding of the social context of smoking may help to enhance tobacco control research
and practice for this high-risk population.

CORRESPONDING AUTHOR: Kelly Aschbrenner, PhD, Geisel School of Medicine at Dartmouth,

Nashua, NH, 03064; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S335

A186 6:00 PM-7:00 PM

THE EFFECT OF FAMILY COMPOSITION ON RISK COMMUNICATION AND SUBSEQUENT RISK

TAKING BEHAVIORS

Danielle E. Reece, MPH Candidate1, Erin R. Johnson, MPH, CHES2, Danielle N. Lambert, MPH,
CHES3

1
Emory University, Atlanta, GA; 2American Institutes for Research, Washington, DC;
3
University of Georgia, Atlanta, GA

The effect of family composition on risk communication and subsequent risk taking
behaviors

Background: Upbringing and family dynamic in the home play a vital role in shaping behavior
later in life. To date, there appears to be limited research on how varying differences in family
composition affects risk communication in the home and the likelihood that children residing
in the home participate in high-risk behaviors later in life. This study aims to fill that gap.

Methods: The sample was drawn from an online survey that was disseminated through
academic institutions, community venues, and social media sites. Responses from 1,011
respondents were collected. Of these, 230 (22.7%) were classified as growing up with non-
traditional families and 781 (77.3%) as growing up with traditional families. For the purposes
of this study, a traditional family was defined as being raised by two opposite sex parents.

Results: Chi Square tests of independence were performed and results show that individuals
who grew up with a non-traditional family were less likely to have received communication
about alcohol use than were individuals who grew up with a traditional family (X 2=24.47,
df=1, p < .001). Individuals who specifically grew up with a non-relative, grandparent, or
extended relative in the home were less likely to have received risk communication about safe
sex, substance use, and alcohol (X2=5.91, df=1, p=.017). Additionally, individuals who grew up
with a non-traditional family were more likely to report history of an STI (X2=4.28, df=1,
p=.041). Two-sided t-tests showed that individuals who grew up in a non-traditional family
reported being spoken to about fewer risk related topics (mean=1.49, sd=1.24) than
individuals who grew up with traditional families (mean=1.72, sd=1.13) (t=2.66, df=1009,
p=.008). They also reported a younger age at first sex (mean=17.25, sd=3.74) than those who
S336 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

grew up with traditional families (mean=17.99, sd=3.80) (t=2.46, df=335.88, p=.014).

Conclusions: These findings suggest that family composition during upbringing is related to
both risk communication within the home and subsequent risk behaviors later in life. These
results illustrate the importance of considering the social environment of the family in
developing effective risk communication and reduction interventions.

CORRESPONDING AUTHOR: Danielle E. Reece, MPH Candidate, Emory University, Atlanta, GA,
30306; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S337

A187 6:00 PM-7:00 PM

ASSESSING COPING IN PREGNANCY: PSYCHOMETRIC PROPERTIES OF THE REVISED PRENATAL

COPING INVENTORY

Sirena M. Ibrahim, B.A.1, Marci Lobel, Ph.D.2, Jennifer Nicoloro SantaBarbara, MSW, MA3,
Melissa Auerbach, PhD in Social and Health Psychology4, Lisa Rosenthal, PhD5, Cheyanne
Busso, B.A.6, Christina Kocis, C.N.M., D.N.P7

1
Stony Brook University, Port Jefferson, NY; 2Stony Brook University, Stony Brook, NY; 3Stony
Brook University, Long Island CIty, NY; 4Fox Chase Cancer Center/Temple University Health
System, Williamstown, NJ; 5Pace University, New York, NY; 6Stony Brook University, deer park,
NY; 7Stony Brook School of Medicine, Stony Brook, NY

It is well established that emotional distress during pregnancy increases womens risk of
delivering a low birthweight or preterm infant. Coping research and theory suggest that the
ways women manage stress during pregnancy can alleviate or exacerbate distress (Guardino
& Dunkel-Schetter, 2013). However, few appropriate tools exist to measure coping in
pregnancy. The purpose of the present study was to examine the psychometric properties of
an instrument first developed almost 20 years ago to assess coping in pregnancy (Yali & Lobel,
1999).

Data were derived from a larger investigation of women receiving prenatal care from a
university hospital midwifery health practice (N = 185). On average, participants were 30 years
of age; married or partnered, with at least some college education. They were heterogeneous
with respect to parity and ethnicity/race. The Revised Prenatal Coping Inventory (NuPCI;
Hamilton & Lobel, 2008) was administered on average at 30 weeks of pregnancy (SD = 4.18).
The 42-item NuPCI includes ways of coping that women use to try to manage the strains and
challenges of being pregnant in the past month on a scale from 0 (never) to 4 (very often).

Data were analyzed with exploratory factor analysis using principal components analysis and
orthogonal rotation. Items were eliminated if they failed to load at least 0.30 on any factor (11
items) or failed to contribute to the reliability of any factor (3 items). Five distinct,
conceptually interpretable, and internally consistent coping factors were identified from the
remaining 28 items: Prayer/Spirituality; Seeking Social Support; Planning/Preparation; Positive
Appraisal; and Avoidance. Scores on the avoidance subscale were correlated with pregnancy-
specific distress (rs .30 to .37, p < .001) and with state anxiety (rs .20 to .31, p < .01) at four
time points across pregnancy; positive appraisal was inversely associated with state anxiety
(rs -.20 to -.29, p < .05) in early and mid-pregnancy.
S338 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Results indicate that the NuPCI is a psychometrically sound instrument to assess coping in
pregnancy. The instrument reliably differentiates ways of coping; frequency of use of two of
these ways of coping is associated with emotional distress in predicted directions. The NuPCI
can be used to examine how coping may alleviate or aggravate womens distress during
pregnancy and thereby alter the impact of prenatal stress on birth outcomes.

CORRESPONDING AUTHOR: Sirena M. Ibrahim, B.A., Stony Brook University, Port Jefferson,
NY, 11777; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S339

A188 6:00 PM-7:00 PM

NEUROTICISM, STRESS, AND ERYTHROCYTE SEDIMENTATION: LONGITUDINAL FINDINGS

FROM THE VA NORMATIVE AGING STUDY

Carolyn M. Aldwin, PhD1, Soyoung Choun, PhD2, Lewina O. Lee, Ph.D.3, Avron Spiro, PhD4

1
Oregon State Univeristy, Corvallis, OR; 2Oregon State University, Corvallis, OR; 3BOSTON
UNIVERSITY SCHOOL OF MEDICINE, Arlington, MA; 4VA Boston Healthcare System, Boston,
MA

Objectives. Both neuroticism and stress have been shown to affect inflammatory processes,
which in turn underlie many of the chronic illnesses of aging, including cancer, heart disease,
diabetes, and neurodegenerative disorders. However, most studies of the association
between psychosocial factors and inflammation are either cross-sectional or short-term
longitudinal studies, and often use patient samples, in part because of the relative newness of
many inflammatory measures and their expense. Erythrocyte sedimentation rate (ESR) is an
assessment of general inflammation widely used in clinical settings. It is inexpensive, and has
been in use since the 1970s, but is rarely used in behavioral medicine research. The purpose
of the current study is to examine the relative contribution of neuroticism and stress to
changes in ESR assessments of inflammatory process over 13 years.

Method. We used longitudinal data from the VA Normative Aging Study, selecting 1,229 men
(Mage = 65.6 years, SD = 6.8) who completed a measure of neuroticism in 1988 and who
subsequently completed stressful life events (SLEs) and hassles measures during one or more
triennial physical examinations between 1989 and 2002 (observations = 3,134; M = 2.6, range
= 1-5). ESR was assessed using the standard Wasserman test from a fasting blood draw (M =
13.99, SD = 7.93). We used multilevel models (MLM) to examine the between- and within-
person associations among neuroticism, stress, and ESR levels and slope across time.

Results. Unconditional analyses showed that ESR increased over time. Between-person
analyses showed that men with higher baseline neuroticism had steeper ESR increase over
time. but SLE was only linked to ESR in within-person analyses. Adjusting for between-person
differences in SLE, ESR levels were lower on occasions which men reported more SLEs.
Controlling for health behavior habits and smoking attenuated this effect slightly. Hassles
were not associated with ESR.

Discussion. Neuroticism was associated with increasing levels of inflammation in later life.
However, major stressors were associated with lower ESR levels, supporting a two-stage
S340 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

model; acute stress may decrease inflammation, but chronic stress (presumably experienced
by those higher in neuroticism) may show increases (Franceschi & Campisi, 2014). Further
long-term longitudinal research into the role of personality and stress in aging and
inflammation is warranted.

CORRESPONDING AUTHOR: Carolyn M. Aldwin, PhD, Oregon State Univeristy, Corvallis, OR,
97331; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S341

A189 6:00 PM-7:00 PM

STAFF PERCEPTIONS OF ALTERED DECISION MAKING IN NICU PARENTS

Carissa Ballew, MPH

Human Lacatation Center, UC Davis, Vacaville, CA

BACKGROUND:

Extreme stress may result in acute stress disorder (ASD). Symptoms of ASD include emotional
numbing, avoidance of feelings, sleep disturbances, concentration deficits, intense reactivity
to reminders of the stressful event, flashbacks, or irritable behavior. Decisions must be made
in all medical environments, including the neonatal intensive care unit (NICU), and some of
those decisions influence health outcomes. The prevalence of ASD is 23-55% in NICU parents,
and all NICU parents experience one or more symptoms of ASD. Despite the high prevalence
of ASD in NICU parents, no known studies have directly examined parents impaired decision-
making in the NICU.

OBJECTIVES:

Characterize how NICU staff members perceive manifestations of NICU parents impaired
decision-making and explore how the staff reacts to this impairment.

METHODS:

A series of 25 semi-structured interviews were conducted among registered nurses,

neonatologists, registered dieticians, lactation consultants, and social workers who worked in
NICU environments. The NICU professionals were asked to share their in-hospital experiences
related to parents stress and impaired decision making.

RESULTS:

The staff perceptions of causes for parents decision-making impairment were broadly
categorized into circumstantial and personal reasons and then further subcategorized.
Circumstantial reasons for impairment included the NICU environment, socio-economic status
and limited access to enabling services, infant acuity, culture and religion, and information
sources such as the internet or family members. Staff perceptions of personal reasons why
the parents were having decision-making impairment included age and gender, family history,
education and intelligence, preferred languages, the parents personal capacity for stress,
physical or emotional condition, and the parents desire for control. The staff reported that
they employed many strategies to help the parents with their decision-making including
S342 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

assessing parents circumstances first and then deciding on a course of action such as taking
over, focusing on empowerment, providing emotional and practical support, and tailoring the
information delivery to meet the needs of the parents.

CONCLUSION:

All of the staff members recognized that the NICU parents were stressed. Staff also recognized
and described manifestations of this decision-making impairment but, rather than attributing
these behaviors to stress-related causes, the staff attributed the behaviors to the parents
circumstances or personal characteristics.

CORRESPONDING AUTHOR: Carissa Ballew, MPH, Human Lacatation Center, UC Davis,

Vacaville, CA, 95687; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S343

A190 6:00 PM-7:00 PM

THE EFFECTS OF A STRESS MANAGEMENT PROGRAM ON SELF-EFFICACY, STRESS, AND MOOD

AMONG LOW INCOME AFRICAN-AMERICAN MOTHERS

Grace M. Thornburgh, B.A.1, Lauren Dunne, B.A.2, Colette Brown, B.A.1, Guido Urizar, Jr., PhD3

1
CSULB PRO-Health Lab, Long Beach, CA; 2PRO-Health Lab/Build/CSULB, Huntington Beach,
CA; 3California State University Long Beach, Long Beach, CA

African-American women experience a number of chronic stressors (e.g. poverty, racial

oppression; Lindblad-Goldberg et. al., 1985) and are twice as likely to die from stress-induced
health issues when compared to White women (Heisler et al., 2004). Despite being
disproportionately affected by stress, there are few effective interventions that are tailored to
address specific risk-factors for African American women (El-Bassel et al., 2009). The present
study evaluated the effectiveness of a community-based stress management program for low
income African-American mothers. Classes met for two hours, once a week, for six weeks. The
class topics included stress awareness, thought replacement, coping skills, social support, and
communication. The majority of participants (n = 26) were single with an annual income of
$15,000 or less. Mothers were evaluated at baseline and post-intervention using self-report
scales to measure self-efficacy (Measure of Current Status), perceived stress (Perceived Stress
Scale), depression (Edinburgh Postnatal Depression Scale), and anxiety (State Trait Anxiety
Inventory). Pairwise samples t-test analyses revealed significant improvements in womens
self-efficacy with reducing their tension by using relaxation skills[t (25) = -5.41, p < .001], being
able to recognize that there body is physically tense during times of stress [t (25) = -5.81, p <
.001], being assertive in communicating their needs to others [t (25) = -3.48, p = .002], and
coping with their stress during difficult situations [t (25) = -3.67, p = .001]. Results also showed
a significant decrease in perceived stress [t (25) = 3.20, p < .05], depression [t (25) = 5.68, p <
.001], and anxiety [t (25) = 3.80, p = .001]. As an underserved population considered high-risk
for stress-related health issues, it is imperative that similar interventions be developed and
implemented. Given the effectiveness of the current study, the intervention could serve as a
model for future clinicians and researchers.

CORRESPONDING AUTHOR: Grace M. Thornburgh, B.A., CSULB PRO-Health Lab, Long Beach,
CA, 90803; [email protected]
S344 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A191 6:00 PM-7:00 PM

AN EXAMINATION OF THE DECISIONAL BALANCE FOR DRUG USE AMONG TRANSGENDER

WOMEN IN NEW YORK CITY

Raymond L. Moody, MA1, Sitaji Gurung, MPH, MD2, Demetria Cain, MPH, BS2, Jeffrey T.
Parsons, PhD2

1
Center for HIV Educational Studies and Training, Hunter College, City University of New York
(CUNY); Graduate Center, CUNY, Brooklyn, NY; 2Center for HIV Educational Studies and
Training, Hunter College, City University of New York (CUNY); Graduate Center, CUNY, New
York, NY

Background: Greater rates of drug use and associated problems have been observed among
transgender women compared to the general population. Efforts to curb drug use in high risk
populations have often focused on the potential costs of use. Yet some evidence suggests that
the perceived benefits of drug use may be a significant target for risk reduction interventions.
The present study sought to examine the association between the perceived costs and
benefits of drug use and rates of illicit drug use and associated problems in a sample of
transgender women.

Methods: A sample of 184 transgender women living in NYC completed baseline assessments
as part of the screening process for a transwomen-focused intervention. Baseline assessments
included measures of demographic covariates, the Decisional Balance for Substance Use scale,
the Short Inventory of Problems for Drug and Alcohol, and a timeline follow-back interview of
drug use (previous 60 days).

Results: In total, 52% of the sample reported using one or more illicit drugs in the previous 60
days. After adjusting for demographic covariates, perceived consequences of drug use
distinguished low and heavy marijuana users from non-users (B=-1.01, p=03; B=-1.55, p=.002,
respectively) and was associated with drug/alcohol related problems (B=6.57, p < .001).
Perceived benefits distinguished heavy marijuana users from non-users
(B=1.48, p=.001), heavy club drug users from non-users (B=.94, p=.02), polydrug users from
non-users (B=1.02, p=.005), and was associated with drug/alcohol related
problems (B=2.16, p=.04).

Conclusions: Our findings are consistent with previous research and extend findings from
previous research to transgender women. The results suggest that the perceived benefits of
drug use may be a more meaningful target for reducing rates of drug use compared to
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S345

interventions that emphasize the potential consequences of use. Interventions that aim to
identify and challenge the perceived benefits or assist in generating alternative strategies for
obtaining the desired benefits will likely benefit transwomen more than the traditional risk
focused approaches.

CORRESPONDING AUTHOR: Raymond L. Moody, MA, Center for HIV Educational Studies and
Training, Hunter College, City University of New York (CUNY); Graduate Center, CUNY,
Brooklyn, NY, 11201; [email protected]
S346 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A192 6:00 PM-7:00 PM

EXPERIENCES OF DRUG EXPOSURE AND USE: A QUALITATIVE STUDY OF BLACK FORMERLY

INCARCERATED DRUG OFFENDERS.

Rahma S. Mkuu, MPH, CPH1, Tawandra Rowell-Cunsolo, PhD, MA, BA2, Idethia Harvey, DrPH3,
Ehikowoicho Idoko, Ms.1

1
Texas A&M University, College Station, TX; 2Columbia University, School of Nursing, New
York, NY; 3Texas A&M University , Department of Health and Kinesiology, Transdisciplinary
Center for Health Equity Research, College Station, TX

Background: Black Americans are vastly overrepresented among those incarcerated for drug
related offenses in the United States. Evidence suggests that post-incarceration, formerly
incarcerated individuals may be at high risk of drug overdose deaths.
Objective: The purpose of this study is to explore the factors that influence drug use among
formerly incarcerated Black drug offenders during the immediate post-incarceration period.
Methods: Qualitative structured interviews were conducted with thirty Black Americans who
had been released from prisons in New York State within the past year. One-on-one in-depth
interviews explored participants drug behavior, institutional, environmental and social factors
influencing drug use. Transcripts were transcribed, and analyzed using line-by-line coding that
resulted in identifying themes by the research team.
Results: The following themes related to facilitating drug use emerged. Participants reported
re-entering drug-enticing physical environments (i.e., neighborhoods or shelters) and social
environment or social networks were at a greater risk of relapse. The following themes were
reported sources for abating drug use: being on parole or other forms of community
supervision, access to drug treatment facilities, and social support from family and friends.
Conclusion: Formerly incarcerated individuals continue to face challenges when attempting to
reduce or avoid substance use post-incarceration. Interventions that incorporate a social
determinants approach and consider the environment and social networks as significant
leverage points for behavior change may be warranted.

CORRESPONDING AUTHOR: Rahma S. Mkuu, MPH, CPH, Texas A&M University, College
Station, TX, 77840; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S347

A193 6:00 PM-7:00 PM

EXPOSURE TO SECONDHAND MARIJUANA SMOKE: FINDINGS FROM A NATIONAL SURVEY

Jessica Y. Sun, BA, Yue-Lin Zhuang, PhD, Sharon E. Cummins, PhD, Shu-Hong Zhu, PhD

University of California San Diego, La Jolla, CA

As marijuana use, both medicinal and recreational, is legalized in a growing number of states,
there is concern about the public health ramifications. There is evidence that secondhand
marijuana smoke may be harmful. This study presents the first national data on exposure to
marijuana secondhand smoke, and it compares the prevalence of exposure to that of tobacco
smoke, which is a well-known public health hazard.

A national survey was conducted with a representative sample of the U.S. adult population in
July-August 2016 (N=11,491).

Overall, 8.4% of U.S. adults used marijuana and 15.3% smoked cigarettes in the last 30 days.
Of the population, 4.7% currently used marijuana only, 11.6% used cigarettes only, and 3.7%
used both, while 80.0% of the population currently used neither.

Overall, 9.5% of the population was exposed to secondhand marijuana smoke in the last 30
days, while 21.4% of the population was exposed to secondhand cigarette smoke. Among
those currently using neither marijuana nor cigarettes, 4.1% were exposed to secondhand
marijuana smoke, while 12.6% were exposed to secondhand cigarette smoke. Among those
exposed to secondhand marijuana smoke, 54.2% were exposed 1-3 days out of the last 30
days, 23.5% were exposed 4-14 days, and 22.3% were exposed 15+ out of the last 30 days.
Among those exposed to secondhand cigarette smoke, 42.5% were exposed 1-3 days out of
the last 30 days, 29.7% were exposed 4-14 days, and 27.8% were exposed 15+ out of the last
30 days.

When asked about rules about smoking marijuana and cigarettes at home, 81.6% of the
population banned indoor marijuana smoking, while 80.3% banned indoor cigarette smoking.
Interestingly, only 26.2% of all marijuana users were banned from smoking marijuana at
home; in comparison, 50.1% of all cigarette smokers were banned from smoking cigarettes at
home.

Overall, 71.2% of the population perceived exposure to secondhand marijuana smoke as

harmful. This included 76.4% of those who used neither marijuana nor cigarettes, 58.2% of
cigarette-only users, 40.2% of marijuana-only users, and 37.8% of dual users.
S348 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Nationally, secondhand exposure to marijuana smoke has reached approximately half that of
the exposure to secondhand cigarette smoke. Given the spread of legalization, which is likely
to increase prevalence (and in turn, exposure), and the preliminary evidence on risks of
exposure to marijuana smoke, there is a need for continued monitoring of the extent of
secondhand exposure to marijuana smoke.

CORRESPONDING AUTHOR: Jessica Y. Sun, BA, University of California San Diego, La Jolla, CA,
92093-0905; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S349

A194 6:00 PM-7:00 PM

INTEGRATED DATA ANALYSIS OF TWELVE NIDA CTN SUBSTANCE USE TREATMENT TRIALS:
BASELINE HIV RISK BEHAVIORS AND SUBSTANCE USE

Jennifer L. Brown, Ph.D.1, Michael D. Eriksen, PhD1, Nicole Gause, M.A.1, Gene Brody, PhD2,
Jessica Sales, PhD3

1
University of Cincinnati, Cincinnati, OH; 2Emory University, Atlanta, GA; 3Rollins School of
Public Health, Emory University, Atlanta, GA

Background: Integrative data analysis (IDA) allows for secondary analyses of multiple original
data sets that have been pooled into one using harmonized variables. The aim of this study
was to (a) create a pooled data set with harmonized variables combining data from 12
individual behavioral substance use treatment trials conducted within the National Institute
on Drug Abuses (NIDA) Clinical Trials Network (CTN); and (b) characterize participant
characteristics and baseline levels of HIV risk behavioral engagement and substance use
during the past month across trials.
Methods: 12 behavioral drug treatment trials data (CTN Trials: 0004, 0005, 0006, 0007, 0009,
0011, 0013, 0014, 0015, 0018, 0019, 0021) were integrated in the pooled dataset (N = 4,865).
Participants had a M(SD) age of 34.8(11.4). Most participants were male (55.5%) and 58.6%
of the sample were a racial/ethnic minority.
Results: At baseline, the number of participants reporting one or more days using the
following substance(s) in the past month was: cocaine (n = 1,766); heroin (n = 661); other
opiates (n = 609); and amphetamines (n = 418). Of those reporting injecting with previously
used needles in the past month (n = 245), 60.4% endorsed inconsistent or no use of bleach to
clean syringes. 48.9% reported having one sexual partner in the past month with 13.8%
reporting more than one sexual partner; 37.3% had no sexual partners in the past month. Of
those reporting vaginal sex in the past month (n = 2,818), 80.6% reported inconsistent
condom use.
Discussion: Results highlight the utility of an IDA approach for examining HIV risk behavior
patterns among individuals engaged in NIDA CTN behavioral substance use treatment trials.
Data indicate elevated prevalence of HIV risk behavior engagement among individuals
participating in community-based substance use treatment trials.

CORRESPONDING AUTHOR: Jennifer L. Brown, Ph.D., University of Cincinnati, Cincinnati, OH,

45229; [email protected]
S350 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A195 6:00 PM-7:00 PM

PRESCRIPTION OPIOID MEDICATION USE AMONG PEOPLE WITH MIGRAINE: NATIONAL

SURVEY RESULTS

Christopher Malone, M.A.1, Amy Wachholtz, PhD, MDiv, MS2, Amrita Bhowmick, MBA3

1
William James College, Brighton, MA; 2University of Colorado Denver, Denver, CO; 3Health
Union, Philadelphia, PA

The abuse of prescribed and illicit opioid substances has increased among US residents in
recent years; with an estimated 1.9 million abusing prescribed opioid analgesics and 586,000
using heroin. Recent research has suggested that the recent rise in opioid abuse among US
residents has been driven largely by the increased use of prescription opioid medication.
Understanding the contributing factors to the prescription of opioid medication in a
chronically ill population may inform and improve the use of opioid analgesics in this
population. In this study, we report the findings of an online national survey of 4,139 adult US-
based patients diagnosed with migraine. 989 (22.1%) were currently using prescription
painkillers for their migraines. 1860 (41.5%) had been prescribed painkillers in the past but
were no longer using them and 1290 (28.8%) had never used painkiller medication for their
migraines. We assessed negative life events due to migraines and collected information on
mental and physical comorbid diagnoses, a self-report mood measure, and other aspects of
migraine treatment. Results reveal that the number of mental and medical comorbid
disorders (R2=0.029; pfor migraine headache and reported functional impairment due to
migraine. The results of the present study enhance understanding of the use of opioid
medication prescribing for chronic migraine sufferers. While there are limitations to the
present study, the results suggest an underdiagnosed level of psychological distress among
chronic migraine sufferers that may be inappropriately addressed with prescription opioids
since elevated levels of any type of comorbidity (including non-pain related comorbidities
such as seasonal allergies) may lead to increase prescribing of opioids for migraine headaches.
Methods to more accurately identify levels of psychological distress in people with chronic
medical conditions in order to avoid opioid over prescribing are discussed.

CORRESPONDING AUTHOR: Christopher Malone, M.A., William James College, Brighton, MA,
02135; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S351

A196 6:00 PM-7:00 PM

VIDA PURA: FEASIBILITY OF SCREENING AND BRIEF INTERVENTION TO REDUCE UNHEALTHY

ALCOHOL USE AMONG LATINO DAY LABORERS

India Ornelas, PhD1, Bonnie Duran, Dennis M. Donovan, PhD2, Samantha Serrano, MS1,
Vanessa Noel. Torres, PhD3

1
University of Washington, Seattle, WA; 2University of Washington nschool of Medicine,
Seattle, WA; 3University of Washington, seattle, WA

Background: Due to stressors associated with immigration, Latino day laborers are at
increased risk for unhealthy alcohol use. Vida PURA is a culturally adapted intervention that
consists of promotores providing screening and brief intervention to Latino day laborers at a
day labor worker center. Methods: We conducted a pilot randomized control trial to test the
feasibility of the Vida PURA intervention and study procedures. Participants were screened
for eligibility using the AUDIT (N = 181). Those with an AUDIT score 6 (N = 121) completed a
baseline survey and were randomized into an intervention or control group. Follow-up
surveys assessed changes in alcohol use at two weeks and eight weeks following the baseline.
Participants in the intervention group received a brief counseling session including
personalized feedback about the quantity of their daily and weekly drinking displayed on a
tablet screen. Promotores also used motivational interviewing (MI) to assess their readiness
to change and encourage behavior change. We evaluated intervention fidelity through a
review of 44% of audio recorded counseling sessions for the presence of three steps of the
intervention protocol. Additionally, we rated the promotores motivational interviewing
technique with the Motivational Interviewing Treatment Integrity (MITI) 4.2.1 tool and
compared scores to competency and proficiency thresholds. We assessed recruitment,
retention, and reach using logs and tracking forms. Results: Promotores delivered each of the
three intervention steps in 78% of selected recordings. Evaluation of MI indicated that the
promotores achieved basic competency across all domains and proficiency in 50% of the
measures. At 58 weeks of data collection, 69% of screened participants were eligible for the
study. Of those eligible, 50% had AUDIT scores 20, indicating alcohol dependence.
Participant retention rates were 86% for the two-week survey and 85% for the eight-week
follow-up. Conclusions: Initial results demonstrate that promotores can be adequately trained
to deliver SBI and increase access to this intervention among an underserved population.
Future research should assess the efficacy of Vida PURA to reduce unhealthy alcohol use,
especially among those with high AUDIT scores.
S352 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: India Ornelas, PhD, University of Washington, Seattle, WA, 98195;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S353

A197 6:00 PM-7:00 PM

ENGAGEMENT IN SMOKING CESSATION TREATMENT FOLLOWING A BRIEF TELEPHONE

EVALUATION AND REFERRAL SESSION

Sofie L. Champassak, Ph.D., Suraya Jubaiah, Pharm.D., Timothy Chen, Pharm.D., Mark G.
Myers, Ph.D.

Veterans Affairs San Diego Healthcare System, San Diego, CA

Despite evidence that combined treatment with medication and counseling yields the best
outcomes, few smokers employ both modalities when attempting to quit. This indicates the
importance of efforts to enhance engagement in optimal evidence based treatments. The
following reports on various aspects of treatment engagement in a sample of Veterans
referred for outpatient smoking cessation treatment who completed a brief telephone call
intended to assist with treatment engagement. Veterans were offered participation in
outpatient groups or an in-house telephone tobacco cessation program. Of those who
completed the brief telephone call, 78.2% (N = 308) indicated a desire to attend outpatient or
telephone treatment. The present study examined the association of gender, age, ethnicity,
diagnosis of a mental health disorder, importance and confidence to quit, attempt to quit in
the last year, and number of cigarettes smoked per day with whether smokers were
interested in and attended treatment. Veterans who indicated a desire to attend treatment
smoked significantly more cigarettes per day (M = 19.05, SD = 19.73; M = 9.77, SD = 9.47; p <
.001) and endorsed significantly higher levels of importance to quit smoking (M = 7.66, SD =
3.62; M = 4.59, SD = 4.31; p < .001) than those who did not. Of those who expressed a desire
to attend treatment, 37.7% (N=116) attended. Veterans who attended treatment were older
(M = 54.08, SD = 11.93; M = 48.56, SD = 13.47; p = .001), smoked more cigarettes per day (M =
25.63, SD = 27.81; M = 13.01, SD = 9.37; p < .05) and endorsed higher levels of importance to
quit smoking (M = 7.81, SD = 3.39; M = 6.65, SD = 4.16; p = .01) than those who did not attend
treatment. There were no significant differences found between those who selected
individual telephone counseling (38.79%, N=45) and those who selected outpatient group
counseling (61.21%, N=71) suggesting that Veterans are receptive to using various treatment
modalities for smoking cessation. This study represents an initial effort to explore the
treatment engagement process among Veterans who express interest in smoking cessation
treatment and attending treatment. These findings highlight the need to further examine
barriers to attending treatment and the importance of efforts to engage Veterans who are
younger, smoke fewer cigarettes per day, and endorse lower levels of importance to quit
smoking.
S354 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Sofie L. Champassak, Ph.D., Veterans Affairs San Diego Healthcare
System, San Diego, CA, 92126; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S355

A198 6:00 PM-7:00 PM

PERCEPTIONS OF AN AFFORDABLE HOUSING SMOKING BAN DIFFER AMONG SMOKERS AND

NONSMOKERS: IMPLICATIONS FOR POLICY IMPLEMENTATION

Vaughan Rees, PhD1, Jessica Davine, MSW, MPH2, Robyn Keske, MSW, MPH1, Natasha Sokol,
MPH1, Alan Geller, RN, MPH1

1
Harvard T.H. Chan School of Public Health, Boston, MA; 2Harvard TH Chan School of Public
Health, Boston, MA

BACKGROUND: In low incomecommunities, which have high smoking rates combined with a
high proportion of multi-unit dwellings, smoking and exposure to secondhand smoke (SHS)
pose a chronic health burden. A proposed rule to ban indoor smoking by the U.S. Department
of Housing & Urban Development (HUD) may reduce tobacco use and improve protection for
up to 2 million public housing residents. To tailor future smoke-free housing implementation
strategies, we assessed factors associated with residents support for a smoking ban and self-
reported exposure to SHS, pre- and post-adoption of a smoke-free policy in a multi-unit,
affordable housing setting.

METHODS: Adult residents(n=70 smokers; n=168 non-smokers) of 12 affordable housing

developments in 4 eastern US states were surveyed immediately prior to, and 6 months after
the adoption of a smoking ban. Change in support for the policy and self-reported seeing or
smelling SHS in the past month was assessed at baseline and follow-up. Regression analyses
assessed odds of supporting the ban and self-reported exposure to SHS at baseline after
controlling for smoking status, gender, number of children in the household, level of
education and race.

RESULTS: At baseline, only 25% of smokers supported the ban, compared with 83% of non-
smokers. These proportions did not change significantly after adoption (30% smokers; 79%
non-smokers: ps>0.3). Unemployed residents had lower odds of supporting the ban
compared to employed (OR = 0.22; p = 0.021). SHS exposure decreased significantly among
nonsmokers after adoption (66% at baseline vs. 40% at follow-up; p

CONCLUSIONS: Smokers and nonsmokers differed in their attitudes and perceived benefits
(reduction in SHS exposure) of a smoking ban in multi-unit, low income housing. Smokers
were less likely to support the ban compared with non-smokers, and attitudes for each group
did not change after adoption. Self-reported SHS exposure decreased after policy adoption
among nonsmokers, but not smokers. The findings suggest that smoke-free implementation
S356 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

and communication strategies must be tailored to the needs of different residents, including
smokers and non-smokers, as well as unemployed and younger residents, who may
experience unique challenges in complying with a ban.

CORRESPONDING AUTHOR: Vaughan Rees, PhD, Harvard T.H. Chan School of Public Health,
Boston, MA, 02115; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S357

A199 6:00 PM-7:00 PM

SLEEP QUALITY IS A SIGNIFICANT PREDICTOR OF QUIT-DAY SMOKING SELF-EFFICACY AMONG

LOW-INCOME TREATMENT-SEEKING SMOKERS

Uma Nair, PhD1, Patricia Haynes, PhD1, Bradley Collins, PhD2

1
University of Arizona, Tucson, AZ; 2Temple University, Philadelphia, PA

Introduction: Sleep disturbances are prevalent among smokers. Compared to non-smokers,

smokers have an increased risk for poor sleep quality. Extant evidence points to smoking self-
efficacy as predictor of smoking behavior change outcomes, but few studies have examined
whether sleep quality influences smoking self-efficacy. Purpose: To examine if baseline sleep
quality is associated with quit day smoking self-efficacy among treatment-seeking smokers
enrolled in a smoking cessation trial. Methods: After completing baseline assessments, N=101
low-income smokers were randomized to receive 4-weeks of (a) an integrated low-intensity
physical activity programming with evidence-based standard smoking cessation intervention
or (b) standard care cessation only. On their quit day (week4), participants attended an in-
person session to complete quit day assessments. Baseline sleep quality was assessed using
the Pittsburgh Sleep Quality Index (PSQI). Results: Sample characteristics included: 51% male,
77% African American, 59% high school education. Participants mean age was 42.1(+ 10.9),
BMI =29.9 (+ 7.4) and nicotine dependence = 5.2 (+ 1.2). Baseline sleep quality score was 6.8
(+3.9) and 55% scored >5, the clinical cut-off for poor sleep quality. Multivariate regression
analysis showed that better sleep quality (b=-.63; p=.03) was associated with greater quit day
smoking self-efficacy, after controlling for empirically relevant variables known to be
associated with self-efficacy (e.g., nicotine dependence, depressive symptoms). Conclusion:
Results suggest that sleep quality during the pre-quit period may undermine smoking self-
efficacy. Given the overall poor levels of baseline sleep quality, these data suggest that sleep
may be an important consideration in smoking cessation programs.

CORRESPONDING AUTHOR: Uma Nair, PhD, University of Arizona, Tucson, AZ, 85714;
[email protected]
S358 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A200 6:00 PM-7:00 PM

SMOKING AND URGENCY IN EARLY ADOLESCENCE: A 5 YEAR LONGITUDINAL STUDY

Gabriella E. Puleo, M.S., Elizabeth Riley, MS, Gregory T. Smith, Ph.D., Jessica L. Burris, PhD

University of Kentucky, Lexington, KY

Background: Among US early adolescents (i.e., youth ages 12-14 or middle school students),
the prevalence of cigarette smoking is at its lowest level in recent decades. However, since the
overwhelming majority of adult smokers began in early adolescence (and once established,
smoking is a very hard behavior to curtail), it remains critically important to study early
trajectories of smoking and identify new targets for prevention and intervention. Objective:
This longitudinal study examines the prevalence of early adolescent smoking and tests a
model of reciprocal determinism between smoking and urgency (i.e., the tendency to act
rashly in response to intense positive or negative mood). Method: 1906 US early adolescents
(50% female, 39% racial minorities) completed questionnaires 1-2 times per year starting in
5th grade and ending in 9th grade; retention was 75% across all eight waves. Structural
equation modeling allowed tests of longitudinal, reciprocal relationships between smoking
and urgency controlling for pubertal status and negative affect at each wave. Results: Ever
smoking increased from 6% to 27% over time, with current smoking near 5% at the final wave.
Urgency at each wave predicted smoking at the next wave, above and beyond the predictive
utility of covariates and prior smoking (all p < .001). Likewise, in all but one case, smoking
predicted an increase in urgency at subsequent waves above and beyond the predictive utility
of covariates and prior urgency (all p < .05). Conclusion: With a large sample of US
adolescents, this longitudinal study is unique in its examination of a bidirectional relationship
between behavior and personality. Results show that while the overall prevalence of smoking
may be low among early adolescents, risk for smoking increases with higher levels of urgency.
Results also show that urgency increases secondary to participation in smoking. Moving
forward, treatment studies should explore urgency as a point of prevention/intervention for
smoking as well as smoking cessation as a means to lessen mood-based impulsivity.

CORRESPONDING AUTHOR: Gabriella E. Puleo, M.S., University of Kentucky, Lexington, KY,

40506; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S359

A201 6:00 PM-7:00 PM

TOBACCO PRODUCT USE AMONG PERSONS WITH MENTAL HEALTH SYMPTOMS

Sandra Japuntich, Ph.D.1, Rochelle K. Rosen, PhD2, Lori A. J. Scott-Sheldon, PhD3

1
The Miriam Hospital, Providence, RI; 2Brown University School of Public Health, Providence,
RI; 3The Miriam Hospital and Brown University, Providence, RI

Background: Individuals with current mental illness are twice as likely to use tobacco, smoke
more cigarettes per day, and experience more difficulty quitting than those without mental
illness. Challenges to quitting may lead smokers with mental illness try alternative tobacco
products at higher rates than smokers without mental illness. The purpose of this study was to
determine the use of cigarettes and other tobacco products among smokers with poor mental
health. Method: Data from the adult sample of the Population Assessment of Tobacco and
Health (PATH) study, a nationally representative survey of 32,320 tobacco users and non-
users, were used. Participants rated their current mental health on a 5-point scale (excellent-
poor) and self-reported awareness, ever and current use of tobacco products. Results: Poorer
mental health was associated with higher rates of ever (OR=1.33, p < .001) and current
smoking (OR=1.48, p < .001); e-cigarette awareness (OR=1.16, p < .001), ever (OR=1.25, p <
.001) and current (OR=1.23, p < .001) e-cigarette use; cigar/cigarillo awareness (OR=1.11, p <
.001); ever (OR=1.14, p < .001) and current (OR=1.19, p < .001) cigarillo use; pipe awareness
(OR=1.07, p < .001), ever pipe smoking (OR=1.08, p < .001); and smokeless tobacco awareness
(OR=1.08, p < .001), higher ever (OR=1.04, p=.004) but lower current (OR=0.94, p=.006)
smokeless tobacco use. There was no association between self-reported mental health status
and ever or current cigar use; current pipe smoking; hookah awareness, ever or current
hookah use. Poorer mental health was associated with a greater number of ever and current
tobacco products used (ps < .001). Conclusion: Consistent with cigarette smoking, awareness
and use of other tobacco products was more prevalent among those with poorer mental
health. Notable exceptions include tobacco products typically used in social situations (e.g.,
hookah). Multiple tobacco use places individuals with mental illness at a higher risk for
tobacco-related disease. Targeted prevention and intervention strategies are needed for
individuals with poor mental health.

CORRESPONDING AUTHOR: Sandra Japuntich, Ph.D., The Miriam Hospital, Providence, RI,
02906; [email protected]
S360 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A202 6:00 PM-7:00 PM

USING ECOLOGICAL MOMENTARY ASSESSMENT TO INVESTIGATE SUBSTANCE CO-USE

AMONG YOUNG ADULT SMOKERS

Johannes Thrul, PhD1, Louisa Holmes, PhD2, Pamela Ling, MD, MPH1

1
University of California, San Francisco, San Francisco, CA; 2SUNY Binghamton, Binghamton,
NY

Introduction: The use of alcohol and marijuana plays an important role for tobacco smoking
behavior, especially among young people. Young adults smoke a large proportion of their
cigarettes under the influence of these substances. However, previous findings about co-use
of cigarettes, alcohol, and marijuana may be unreliable as they are based predominantly on
retrospective survey studies covering recall windows of 30 days or more, which are limited by
recall bias. Studies are needed to investigate patterns of co-use using assessments more
closely timed around use occasions.

Methods: We used a smartphone-based assessment system to collect Ecological Momentary

Assessment (EMA) data from young adult smokers, ages 18-26, and completed a pilot
feasibility study with 9 participants (age M=23, 56% female, 56% daily smokers). Participants
living in the San Francisco Bay Area were recruited on Facebook and completed screening,
informed consent, and baseline questionnaires online. For a total of 30 days, participants
were prompted once each morning to complete an assessment of tobacco, alcohol, and
marijuana use, including detailed questions about substance co-use (e.g., smoking tobacco
while using another substance) covering the entire previous day. Participants completed 78%
of these previous-day assessments, contributing a total of 210 participant-days (23.3 days per
participant).

Results: Cigarette use was reported on 96.2% of days, alcohol use on 32.9%, and marijuana
use on 29.5%. Co-use of cigarettes and alcohol was reported on 19.5% of days and co-use of
cigarettes and marijuana on 22.4%. Participants reported to have smoked the majority of
cigarettes (all or most of them) under the influence of alcohol on 48.8% of co-use days and
under the influence of marijuana on 25.5% of co-use days. On co-use days, the number of
cigarettes was significantly associated with number of alcoholic drinks (r=.23; p=.05), but not
with number of times of marijuana use (r=.10; p=.45).

Conclusion: Young adult smokers participating in our study reported a similar number of days
co-using cigarettes and alcohol or marijuana. However, on these co-use days, a greater
proportion of cigarettes were co-used with alcohol compared to marijuana. Smartphone
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S361

based EMA studies can help to elucidate patterns of substance co-use in great detail, improve
our understanding of co-use, and inform development of mobile interventions.

CORRESPONDING AUTHOR: Johannes Thrul, PhD, University of California, San Francisco, San
Francisco, CA, 94143; [email protected]
S362 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Thursday

March 30, 2017

8:00 AM-9:15 AM
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S363

Symposium 1 8:00 AM-9:15 AM

CONNECTING THE DOTS: TRANSLATING EVIDENCE TO ADVANCE HEALTH EQUITY IN PRECISION

MEDICINE

Kassandra I. Alcaraz, PhD, MPH1, Clement K. Gwede, PhD, MPH, RN, FAAN2, Karen Yeary,
PhD3, Usha Menon, PhD, RN, FAAN4, Kimlin Tam. Ashing, Ph.D5

1
American Cancer Society, Atlanta, GA; 2Moffitt Cancer Center, Tampa, FL, FL; 3University of
Arkansas for Medical Sciences, Little Rock, AR; 4University of Arizona, Tucson, AZ; 5City of
Hope National Medical Center, Duarte, CA

Precision medicine (PM) holds great promise for advancing disease prevention and treatment
by considering genetic, environmental, and behavioral influences on health. However,
underserved populations must be appropriately represented and engaged to eliminate health
disparities. The ConNECT Framework, focusing on five broad and synergistic principles
(Integrating Context, Fostering a Norm of Inclusion, Ensuring Equitable Diffusion of
Innovations, Harnessing Communication Technology, and Prioritizing Specialized Training), is a
new and actionable model designed to foster health equity in behavioral medicine. A
multidisciplinary panel of experts will present the application of the ConNECT Framework to
evidence-based health equity approaches pertinent to PM. The first presenter will discuss
strategies for proactive and consistent inclusion of medically vulnerable and socially
marginalized populations. The second presenter will discuss ways to elucidate context, using
the social context of religion and spirituality as an example. The third presenter will discuss
methods for equitable dissemination of new research knowledge with respect for unique
cultural and social factors (e.g., religious beliefs) that influence individual receipt and co-
production of health information. The discussant will synthesize and critically examine
strategies presented including identifying potential challenges and barriers to
implementation. The symposium will provide a plan of action for advancing health equity in
the era of PM.

Symposium 1A

FOSTERING A NORM OF EQUITABLE INCLUSION IN PRECISION MEDICINE

Dr. Clement K. Gwede, PhD, MPH, RN, FAAN

S364 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

The emphasis on precision medicine (PM), with its advancements in the biological bases on
health, brings exciting new opportunities for the integration of behavioral, contextual, and
place-based determinants of health. PM holds great promise for reducing health disparities,
and stimulating equity efforts in behavioral medicine science and practice. Community-based
participatory research (CBPR) provides an important and salient framework and approach to
enhance inclusion and representation of underserved and vulnerable populations in
behavioral medicine research and practice. CBPR-informed initiatives have effectively helped
to increase participation of diverse groups in population health efforts including community-
based interventions and clinical trials. The ConNECT framework, comprised of five broad and
synergistic health equity-centric principles, emphasizes the essential role of CBPR to improve
diversity and inclusion with the ultimate goal of maximizing equity in all areas of health. This
presentation addresses the second ConNECT principle, fostering a Norm of Inclusion
(consistently engage and examine diverse groups) to further integrate CBPR, PM, and
emerging communication technologies to achieve health equity. It offers exemplars of how
ConNECT improves inclusion in four areas: 1) stepwise integration of CBPR in the context of
prevention, clinical, and community trials to increase representation and participation of
historically underserved racial/ethnic minorities and other marginalized groups (e.g., sexual
minorities); 2) promoting biobanking and participation in biospecimens research among
historically vulnerable and at-risk groups; 3) leveraging communication technologies (e.g.,
smartphones, patient portals) for broader reach to diverse and underserved populations, and
4) disseminating cultural competence training for equitable care delivery across diverse levels
of health care providers, public health practitioners, and policy makers. The ConNECT
Framework helps generate and translate emerging PM knowledge into tangible and
meaningful benefit for all patients, communities, and society so that no one is left behind.

CORRESPONDING AUTHORS: Linda Fleisher, PhD, Childrens Hospital of Philadelphia;

[email protected]; Shannon M. Christy, PhD, Moffitt Cancer Center;
[email protected]; Jamilia Sly, PhD, Icahn School of Medicine at Mount Sinai;
[email protected]; Kimlin Tam Ashing, PhD, City of Hope Medical Ctr; KAshing-
[email protected]; Virginia Gil-Rivas, PhD, UNC Charlotte; [email protected]; Karen Yeary, PhD,
University of Arkansas for Medical Sciences; [email protected]; Jean C Yi, PhD, Fred Hutchinson
Cancer Research Center; [email protected]; Qian Lu, PhD, University of Houston;
[email protected]; Usha Menon, PhD, RN, FAAN, University of Arizona School of Nursing;
[email protected]; Kassandra Alcaraz, PhD, American Cancer Society;
[email protected]; Cathy D. Meade, PhD, RN, FAAN, Moffitt Cancer Center;
[email protected]

Symposium 1B

INTEGRATING CONTEXT: CONSIDERING RELIGION AND SPIRITUALITY IN PRECISION MEDICINE

Karen Yeary, PhD

ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S365

Religion and spirituality (R/S) are important components of the social context that have
documented relationships with health behaviors, in particular the acceptability, decision
making, and uptake of health screening, diagnostics, and therapeutics. Precision medicine
(PM) has prompted excitement in the scientific community for its capacity to treat and
prevent disease through individual tailoring treatments and prevention strategies. Specifically,
precision medicine takes into account an individuals unique biological, behavioral, and
environmentalincluding socialcontext. Yet the social context component of R/S has not
been adequately considered. R/S may have particular salience in health equity, as some
groups bearing a disproportionate burden of disease report higher levels of R/S than the
general population (e.g. African Americans). The ConNECT Framework, which emphasizes the
importance of considering social and contextual health factors to foster health equity, is a
model through which R/Ss role in PM can be discussed.

Epidemiological literature has documented R/Ss significant associations with both physical
and mental health outcomes, including biological markers (e.g., CD4 counts/percentages).
Qualitative research has explored the role of R/S beliefs in disease perception, treatment, and
preventative behaviors (e.g., cancer, cancer screening). The fields of sociology and psychology
have investigated the powerful role of religious institutions as social forces that influence
health behaviors (e.g., alcohol use) through social norms. Moreover, intervention science has
reported the important role of R/S in health promotion.

Guided by ConNECT Framework strategies, the proposed presentation will discuss: 1) the
relevance of R/S to health equity in PM; 2) recommended R/S assessments to include in data
collection for individualized health recommendations; 3) the use of R/S data to explicitly tailor
information exchange and the design of studies (including communication technology) for
individuals (e.g., health education material tailored to an individuals R/S beliefs); and 4)
incorporation of R/S into provider and health practitioner training in PM delivery.
Consideration of the R/S domain of care can align with the underlying tenets of individualized
clinical care and treatment, and contribute to health equity in PM.

CORRESPONDING AUTHORS: Usha Menon, PhD, RN, FAAN,, The University of Arizona; Dr.;
Mei-Wei Chang, PhD, Ohio State University; [email protected]; Shannon Kristie, PhD,
Moffitt Cancer Center; [email protected]; Kimlin Ashing, , PhD, City of Hope;
[email protected]; Katarina Felsted,, MS, University of Utah;
[email protected]; Cathy Meade , PhD, RN, FAAN, Moffitt Cancer Center;
[email protected]; Virginia Gil-Rivas, PhD, University of North Carolina;
[email protected]; Kassandra Alcaraz, PhD, MPH, American Cancer Society; Kassandra
[email protected]

Symposium 1C
S366 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

PRECISION MEDICINE AND EQUITABLE DISSEMINATION OF INFORMATION

Dr. Usha Menon, PhD, RN, FAAN

In our recently published ConNECT Framework: A Model for Advancing Behavioral Medicine
Science and Practice to Foster Health Equity,we addressed health equity from a behavioral
medicine lens with one of the focal points being the equitable dissemination of research
findings.Equitable dissemination of information is an important component in the growing
field of precision medicine (PM).

PM offers a tremendous opportunity for revolutionizing how we prevent and treat disease
and improve health. If we achieve the national goals set forth for the one million-strong
Precision Medicine Initiative cohort by the White House, health care providers will be able to
apply cutting edge genomics and genetic testing data to the personalized treatment of risk
and disease. Ideally, these innovative treatments will also take into account individual
differences in peoples genes, environments, and lifestyles.

However, as with any advances in medicine, we should learn from the past, in that as we
write the playbook for PM, considerations of health equity and underserved groups are well
integrated into the planning and not as an afterthought. The dissemination component of the
ConNECT Framework provides a roadmap for PM whereby participants are engaged in
receiving and providing health information for the purposes of enhancing research,
information exchange of study findings, and ensuring ethical practices in PM. Engaging
multiple organizations and institutions within communities (e.g. community health clinics,
faith-based organizations, non-profit organizations) is essential in equitable dissemination of
information and to ensure that the information is responsive to the unique needs of these
groups. To be fully effective, precision treatments of disease and risk reduction must be
accessible to all communities and must not contribute further to health inequities. Further,
factors that influence the participation of individuals in genetic studies as well as their desires
to receive information related to their risk factors and current health conditions (e.g., cultural
norms, spiritual and religious tenets, and medical truth-telling) should also be considered. We
elucidate a framework for robust engagement through information exchange, treatment
access, and diffusion of research findings that promotes health equity.

CORRESPONDING AUTHORS: Karen Yeary, PhD, niversity of Arkansas for Medical Sciences;
[email protected]; Cathy Meade, PhD, FAAN, Moffitt Cancer Center; [email protected];
Kassandra Alcaraz, PhD, American Cancer Socity; [email protected]; Clement
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S367

Gwede, PhD, FAAN, Moffitt Cancer Center; [email protected]; Kimlin Ashing, PhD,
Beckman Research Institute, City of Hope Medical Center; [email protected]; Virgina Gil
Rivas, PhD, University of North Carolina Charlotte; [email protected]; Jamilia Sly, PhD, Icahn
School of Medicine at Mount Sinai; [email protected]; Qian Lu, PhD, University of
Houston; [email protected]
S368 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 2 8:00 AM-9:15 AM

INCORPORATING CANCER SCREENING INNOVATIONS IN INTEGRATED HEALTH SYSTEMS

Sherri Sheinfeld Gorin, PhD, FSBM1, Sara J. Knight, PhD2, Suzanne O'Neill, PhD3, Amy M.
Kilbourne, PhD4

1
New York Physicians Against Cancer, http://nypac.center/, New York, NY; 2University of
Alabama at Birmingham, Birmingham, AL; 3Georgetown University, Washington, DC; 4U.S.
Dept. of Veterans Affairs, Veterans Health Administration, Ann Arbor, MI

Integrated health systems have a vested interest in population health focusing on cancer
prevention and control. Yet, implementing innovations in cancer screening can be complex,
often necessitating engagement of both patients and other stakeholders across multiple levels
of the health system. Beyond dissemination of evidence, planned implementation processes
are often needed to incorporate novel behavioral medicine interventions in integrated health
systems. In this symposium, we examine two instructive cases of the dissemination and
implementation of innovations in cancer screeningthe introduction of genomic services to
screen for Lynch syndrome, a hereditary colon cancer, in the Department of Veterans Affairs
(VA) health system and the incorporation of technology to support breast density notification
as a component of breast cancer screening in an integrated care setting in the Pacific
Northwest. The two cases will illustrate how the characteristics of the innovations, conceptual
models, research designs, stakeholder engagement methods, and organizational settings
influence the adoption of new approaches to cancer screening in clinical practice in health
systems. The first speaker will reflect on the multilevel contexts (policy, provider, and patient)
for the translation of novel cancer screening interventions in large healthcare organizations
and the broader social and political emphasis on efficient coordinated and patient-centered
care. In the second presentation, the speaker will consider findings from a pre-
implementation mixed methods project using the Greenhalgh Diffusion of Innovation model
to examine the adoption of screening for Lynch syndrome in Veterans under age 50 diagnosed
with colorectal cancer and cared for in the VA health system. The third speaker will describe a
hybrid efficacy implementation project using the RE-AIM framework to inform a personalized
web-based breast cancer risk communication and decision support tool for women with
dense breasts and increased cancer risk seen in an integrated care setting. The discussant will
compare and contrast the two cases offering an implementation science perspective on
opportunities to improve the integration of novel cancer interventions in large health care
organizations through the use of deliberate implementation processes.

Symposium 2A
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S369

MULTILEVEL CONTEXTS FOR IMPLEMENTING NOVEL INTERVENTIONS IN CANCER

Dr. Sherri Sheinfeld Gorin, PhD, FSBM

New sources of medical evidence or best practices, such as those provided by these two
studies, and long-standing USPSTF guidelines, are not always put into use by health providers,
or are put to use only inconsistently. This presentation will explore the multilevel (policy,
provider, patient) barriers to successful implementation of novel, evidence-based cancer
interventions that promote informed preference based medical decisions in large, integrated
healthcare systems; Each level influences the other, either implicitly or explicitly. For example,
patients and clinicians are never really isolated from the healthcare organization in which they
are embedded. Patients may not make preference-informed medical decisions if clinicians do
not engage them in discussion about the best evidence. Similarly, healthcare organizations
may not incentivize such patientclinician communication if patient loads and processes leave
clinicians with little time or support to address the complexity of individualized decisions for
patients. Organizational resources and tools to encourage, empower, and motivate clinicians
to adopt the practice of individualized decision making and to ensure sustained awareness of
the decision-making process can range from the simple to the complex, including decision aids
and programs that might model likely patient outcomes. In response to external
accountability demands or internal quality improvement initiatives, organizations may assess
conflicts or inconsistencies within the context of health services delivery, particularly those
arising between clinicians and patients regarding the use of new technologies, such as breast
density measurement or routine Lynch syndrome assessment. In the event of a positive test,
the follow up and monitoring, possible diagnosis and treatment of cancer are costly, especially
among patients with multiple chronic problems. Population-based measures of access, uptake
of preventive services, and mortality rates may be used by organizations to assess
individualized decisions and the quality of the healthcare organization. Existing organizational
performance indicators may be inadequate to assess the impact of routine genetic testing,
however. The multilevel framework has its genesis in the Patient Protection and Affordable
Care Act and the Institute of Medicines emphasis on efficient coordinated and patient-
centered care. The findings from a recent systematic review and meta-analysis suggest that
multilevel interventions have positive outcomes across a number of health behavior
outcomes for racial/ethnic subgroups and the poor. The findings also suggest that multilevel
interventions may improve the quality of health-care system processes for subgroups that
often receive sub-par healthcare.

CORRESPONDING AUTHORS: Erica Breslau, PhD, NCI; [email protected]; Mara

Schonberg, MD, MPH, Harvard Medical School; [email protected]; Heather
Edwards, PhD, PCORI; [email protected]; Nicole Saiontz, MPH, NCI;
[email protected]; Louise Walter , MD, UCSF; [email protected]

Symposium 2B
S370 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

LYNCH SYNDROME SCREENING IN ROUTINE CARE FOR VETERANS WITH COLON CANCER IN
THE VETERANS AFFAIRS HEALTH SYSTEM

Dr. Sara J. Knight, PhD

Precision medicine, including genomics, is dramatically changing the way health risk
assessments inform health decision making, a Department of Veterans Affairs (VA) health
system priority. Screening for Lynch syndrome, a hereditary cancer associated with an 80%
lifetime risk of colon cancer and lesser risk of seven other cancers, was one of the first
applications in genomics to accumulate evidence for its clinical validity and utility. While
national guidelines supported screening for Lynch syndrome, in many health systems,
including the VA, adoption has been lower and more variable than would be expected based
on the recommendations. To address this gap between evidence and practice, we sought to
characterize the adoption of Lynch syndrome screening in the VA health system between
2003 and 2013. The Greenhalgh Diffusion of Innovation Model guided our selection of
methods to examine system readiness for the innovation, stakeholder engagement, and
potential for deliberate implementation processes.

We used mixed methods to: 1) establish a national retrospective cohort of all Veterans under
age 50 and cared for in the VA between 2003 and 2010 using VA administrative data, clinical
registry data, and the electronic health record (n=886) and 2) conduct national key informant
interviews with stakeholders in 2013 to identify barriers and facilitators to the adoption of
Lynch syndrome screening (n=58 clinicians). Quantitative and qualitative analyses were
triangulated to understand system readiness for adoption of genetic consultation and tumor
analysis, key elements used in screening to inform subsequent germline testing.

The analysis revealed a low rate of tumor analysis (8%) and genetic consultation (4.85%) from
2003 to 2008 with a slight rise in utilization during 2009. Of the 108 VA Medical Centers with
colon cancer cases, 62% and 69% lacked documentation of tumor testing or referral to a
clinical genomics expert, respectively. Key informant interviews in 2013 indicated that
clinicians perceived both tumor testing and consultation with a genomics expert as
advantageous and inexpensive, but clinicians reported challenges--lack of expeditious request
and approval routines for referrals, knowledge gaps in genomics among health professionals,
and lack of genomics expertise in VA clinical settings.

Triangulating information from multiple data sources and stakeholder engagement activities
revealed receptiveness to Lynch syndrome screening among VA clinicians, but multiple
barriers occurring at the individual and system level impeded routine use. Subsequent
engagement of VA system level leaders has used these results to inform planning for
deliberate implementation.

Symposium 2C
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S371

PREPARING TO TEST A DECISION SUPPORT TOOL FOR WOMEN WITH DENSE BREASTS IN AN
INTEGRATED CARE SETTING: LESSONS LEARNED

Suzanne O'Neill, PhD

Background: Over half of US states require breast density disclosure to women following
screening mammography. This proliferation has occurred without evidence of clinical benefit,
clinical guidelines, or established best practices for risk communication or clinical integration.
To address this gap, we have developed a personalized, web-based breast cancer risk
communication and decision support tool for women with dense breasts and clinically
elevated cancer risk. As we prepared to test this intervention vs. usual care in an integrated
care setting, we completed a series formative stakeholder engagement activities.

Methods: Engagement activities included focus groups with patients, in-depth pre-testing of
the preliminary tool with patients, interviews with health care providers and delivery system
leaders and engagement of the breast care guideline committee within the health care
system.

Results: Most focus group participants were aware of breast density, but unfamiliar with its
role as a breast cancer risk factor. Women noted that the tools appearance and branding
should convey its credibility and that the intervention methods should engage their primary
care providers, as these were noted as a trusted source of information. Patients in usability
testing (n=8, mean age=60) rated the tool high on acceptability and ease of use, and provided
feedback to improve the communication of cancer risk and risk management approaches.
Health care providers and clinical leaders highlighted key areas for provider engagement prior
to the trials launch. These included clarifying referral pathways and mechanisms, alignment
with the clinical guidelines within the system, and monitoring of clinical impact within the
health system.

Discussion: Incorporating multiple stakeholder perspectives greatly facilitated the

development of this web-based decision support intervention. Iterative feedback across a
group of diverse stakeholders within this integrated health care system additionally supports
the execution of our RCT. This formative work will also allow us to contribute to developing
additional guidance to fill the ongoing clinical vacuum created by legislative mandates.

CORRESPONDING AUTHORS: Karen Wernli, PhD, Group Health Research Institute;

[email protected]; Kelly Ehrlich, MS, Group Health Research Institute; [email protected];
Sarah Knerr, PhD, University of Washington; [email protected]; Amanda Graham, PhD, Truth
Initiative; [email protected]; David Farrell, MPH, People Designs, Inc.;
[email protected]; Chalanda Evans, BS, Georgetown University;
[email protected]; Marc Schwartz, PhD, Georgetown University;
[email protected]; Kathleen Leppig, MD, Group Health Cooperative;
[email protected]
S372 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 3 8:00 AM-9:15 AM

ACADEMIC-BUSINESS PARTNERSHIPSBENEFITS, BARRIERS, AND OPPORTUNITIES, OH MY!

Amy McQueen, PhD1, Gregory Norman, PhD2, Jamie L. Studts, PhD3, Audie Atienza, PhD4

1
Washington University, School of Medicine, St. Louis, MO; 2West Health Institute, La Jolla,
CA; 3University of Kentucky College of Medicine, Lexington, KY; 4ICF, Rockville, MD

Academic researchers develop scientific expertise in niche areas and seek to apply their
theories and results in dynamic real-world environments. For-profit and non-profit
organizations engage with hundreds to millions of diverse individuals, and continually seek to
innovate their services and products to increase their competitive advantage in the
marketplace. Partnerships between academia and business are not new, but are not well
understood. This symposia will present three successful examples of academic-business
collaborations. Each speaker will address common phases of development (e.g., forming,
storming, norming, performing) including challenges and benefits of each. They will discuss
experienced-based approaches to dealing with funding, publication, conflicts of interest,
protecting intellectual property and research credibility, as well as gains in knowledge,
innovation, and the co-creation process. The first presentation will involve an overview
of tobacco cessation research conducted as part of Small Business federal-government
(SBIR/STTR) grants with a health technology company. The second presentation will provide
an overview of a unique industry-funded Center for Health Behavior Change and the
collaborative health communication research being conducted to engage commercially-
insured and Medicaid-insured populations in life-centered health behavior change programs.
The third presentation will describe the role of a unique medical research organization in
facilitating collaborative research with hospitals, healthcare organizations, and foundations.
The discussant will offer some perspective on the strengths, opportunities, and challenges of
different models of academic-business partnerships that involve government, university,
business, non-profit and healthcare stakeholders.

Symposium 3A

INNOVATIVE HEALTH COMMUNICATION RESEARCH INVOLVING MEDICAID BENEFICIARIES AT

THE ENVOLVE CENTER FOR HEALTH BEHAVIOR CHANGE

Dr. Amy McQueen, PhD

ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S373

Health communication campaigns can have great reach and impact attitudes, intentions and
behaviors. However, people are inundated with health information from diverse sources.
People do not pay attention to all messages, nor do they always prioritize health when other
life demands are more salient. The industry-funded Center for Health Behavior Change is a
research collaboration between two universities and a large corporation. The health
communication research team examines new methods for reaching, informing, and engaging
people in health programs. Specifically, we seek to explore life priorities, preferences for life
activities, and life values as novel approaches to tailor messages and programs to
meaningfully engage adults in their healthcare. Study 1 recruitment is ongoing; we completed
qualitative interviews with adults with commercial (n=61) and Medicaid (n=25) health
insurance. Interviews include card sort exercises involving 31 alternative terms for a health
coach (e.g., advisor, helper, guide) or exposure to one of twelve paper prototypes of a future
online health coaching and goal-setting program. Prototypes vary in approach (risk-based
recommendation, choice of health behaviors to address, or choice of life-centric priorities to
address) and formatting (when coaches are introduced, data capture methods). Results
include preferences for coaching terms, expected services and characteristics of coaches;
reactions to the different approaches and design features; and motivations and barriers for
behavior change. Study 2 began recruitment of ~1200 Medicaid beneficiaries in four states to
complete an online survey to validate a new measure of life values and examine how values
influence health attitudes and behaviors. These formative studies will support a future
intervention trial to compare different messaging or engagement approaches with Medicaid
beneficiaries to improve health behaviors and outcomes. This unique academic-industry
partnership has produced a synergistic learning environment for academic researchers and
business professionals to expand their horizons toward impacting millions of Medicaid
beneficiaries through evidence-based strategies tailored on meaningful life values and
priorities.

CORRESPONDING AUTHORS: Karyn Quinn, Envolve PeopleCare; [email protected];

Jen Funaro, Envolve PeopleCare; [email protected]; Matthew Kreuter, PhD,
Washington University; [email protected]

Symposium 3B

WEST HEALTH: AN INNOVATIVE MODEL OF COLLABORATIVE RESEARCH TO ENABLE SENIORS

TO SUCCESSFULLY AGE IN PLACE.

Gregory Norman, PhD

The fastest growing segment of the US population is people over age 80. This group is part
of a demographic shift happening in the US where about 10,000 baby boomers turn 65 each
day, which will result in 20% of the US population over age 65 by 2030. There is a growing
S374 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

recognition that the current health care system is unable to support the unique needs of
seniors including clinical and social supportive services.

The West Health mission is to enable seniors to successfully age in place, with access to
high-quality healthcare and support services that preserve and protect seniors dignity, quality
of life and independence. West Health comprises the non-profit West Health Institute, the
non-partisan West Health Policy Center and the Gary and Mary West Foundation. Through
applied medical research, policy and advocacy and outcomes-based funding, West Health
focuses on studying, developing and advancing scalable, sustainable and affordable
healthcare delivery models that enable seniors to successfully age.

The presentation will discuss ongoing projects and lessons learned on conducting multi-
discipline collaborations that address the patient, provider, and payer perspective of
healthcare. One example of West Healths efforts is the establishment of a Qualified Clinical
Data Registry (QCDR) for home-based medical practices. Currently, no nationally recognized
quality of care measures exist. West Health is partnering with two academic institutions, a
national professional society and a technology vendor to equip home-based medical care to
engage in value-based care. Utilizing the QCDR, practices can engage in payer-required
reporting, performance measurement, and quality improvement. In addition, the QCDR can
fill gaps in payers knowledge of the needs and outcome measures of seniors with multiple
chronic conditions.

Another example is West Healths engagement in initiatives around geriatric emergency

care. West Health is collaborating with an academic health system to establish the first
Geriatric Emergency Department (GED) in San Diego. The GED will test a senior-specific
workflow including screenings for fall risk, delirium, and functional ability. On a national level,
West Health is partnering with health systems to incorporate GED guideline-based care
improves patient outcomes and is cost effective, as well as build the data infrastructure for
performance monitoring and evaluation of GEDs.

West Health works with researchers, clinicians, policy makers, and community organizations
to create inter-disciplinary collaborations with a strong emphasis on adoption and
sustainability. This unique research collaboration and funding model creates opportunities to
address gaps in the science of aging, advance policy changes, and most importantly drive
adoption of senior-specific care models at a more rapid pace than other extramural support
mechanisms.

CORRESPONDING AUTHORS: Kelly Ko, PhD, West Health Institute; [email protected]; Jill
Solaboda, PhD, West Health Institute; [email protected]; Zia Agha, MD, West Health
Institute; [email protected]

Symposium 3C
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S375

CONDUCTING APPLIED TOBACCO CONTROL RESEARCH WITH SBIR/STTR FUNDING: AN

ACADEMIC-INDUSTRY PARTNERSHIP

Dr. Jamie L. Studts, PhD

The Treating Tobacco Use and Dependence Clinical Practice Guideline firmly advocates for the
integration of evidence-based tobacco treatment strategies in all healthcare settings.
However, many clinicians receive suboptimal training in this important area of care and fail to
implement tobacco control strategies with their patients who are current and former tobacco
users. In a collaboration between a university and a health technology company, we have
worked to adapt and evaluate QuitAdvisor, a web-based point-of-care clinical decision
support tool designed to facilitate implementation of evidence-based tobacco treatment
interventions in diverse health care settings and by a range of clinicians. This academic-
industry collaboration has benefitted from a series of SBIR and STTR funded grants that have
facilitated adaptation of the original primary care-based QuitAdvisor tool into QuitAdvisorDDS
and QuitAdvisorOB and subsequent usability testing, pilot testing and clinical trials evaluating
feasibility, acceptability, and efficacy of these tools. While the initial partnership was born of a
desire for an independent evaluation of the QuitAdvisorDDS tool, the team evolved into
sharing responsibility for generating targeted and tailored content while the industry partner
maintained responsibility for technological design and support and the academic partner
oversaw the research design and evaluation. This relationship allowed for independent
responsibility of two separate domains of expertise but shared responsibility for the content
arena. From the beginning of the collaboration, there were shared values pertaining to
evidence-based care and a desire to expand access to high quality tobacco treatment. The
principal challenge has involved the distributed nature of the collaboration with the two
partners working thousands of miles apart and for arranging sufficient in-person and
technology enhanced meetings to facilitate progress. This successful partnership has created
new opportunities and is evolving toward a new line of research dedicated to integrating
QuitAdvisor into a broader package of tools dedicated to improving health outcomes in
alignment with current health policy and healthcare innovations.

CORRESPONDING AUTHORS: Hannah Knudsen, PhD, University of Kentucky College of

Medicine; [email protected]; George Reynolds, BA, Health Decision Technologies,
LLC; [email protected]
S376 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 4 8:00 AM-9:15 AM

MULTIMORBIDITY IN OLDER ADULTS AND CANCER SURVIVORS: BEHAVIORAL, COGNTITIVE-

AFFECTIVE, AND BIOLOGICAL PROCESSES

Jerry Suls, PhD1, Thomas V. Merluzzi, PhD2, Elliot Friedman, PhD3, Melanie Keats, PhD4, M.
Bryant Howren, PhD, MPH5

1
National Cancer Institute, Bethesda, MD; 2University of Notre Dame, Notre Dame, IN;
3
Purdue University, West Lafayette, IN; 4Dalhousie University, Halifax, NS, Canada; 5Veterans
Administration, Iowa City, IA

Multimorbidity--the presence of two or more primary chronic conditions in an individual--is

increasingly common and places considerable burden on patients and the healthcare system
alike, including cost, adverse mental and physical health outcomes, and quality of life.
Although health psychology and behavioral medicine have long focused on the co-occurence
of psychological and medical morbidities, there has been much less attention paid to the
behavioral mechanisms and associated factors which may contribute to the development of
and adjustment to multiple primary chronic conditions. This symposium, sponsored by SBM's
Multimorbidity Special Interest Group, will discuss the role that behavioral, cognitive-
affective, and biological processes may play in understanding the development, progression,
treatment, and recovery in those with multimorbidity. Dr. Merluzzi will examine the
relationship between serious comorbidities and functional impairment and a mediated coping
model testing how coping attenuates or exacerbates the relationship between
comorbidities/functional impairment and distress. Dr. Friedman will outline a guiding model
for, and present results from, studies examining the interactive influences of psychological
well-being and inflammation on the links between multimorbidity and disability. Finally, Dr.
Keats will describe results from a national prospective cohort study, focusing on the
prevalence of physical activity in individuals with a self-reported history of chronic disease,
particularly those with two or more chronic diseases.

Symposium 4A

THE ROLE OF COPING IN THE RELATIONSHIP BETWEEN SERIOUS COMORBIDITIES AND

EMOTIONAL DISTRESS

Dr. Thomas V. Merluzzi, PhD

ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S377

Introduction: Comorbidities add burden by increasing allostatic load (Sterling & Eyer, 1988)
and negatively affecting quality of life (Vissers et al, 2013). Moreover, comorbidities are
correlated with functional impairment, which may intensify distress. Data support the effects
of comorbidities on patient care and mortality (Sogaard et al, 2013). However, comorbidity
models that take into account functional impairment and coping mechanisms may clarify
processes that mitigate or exacerbate outcomes. The current study examined 1) the
relationship between comorbidities and functional impairment and 2) a mediated coping
model testing how coping attenuates or exacerbates the relationship between
comorbidities/functional impairment and distress. Method: 407 patients and survivors (mixed
diagnoses) completed measures of Serious Comorbidities, Coping (BriefCOPE), Functional
Impairment (Sickness Impact Profile-SIP) and Emotional Distress (HADS+QLACS-Negative
Feelings). Regression analysis included age, time since diagnosis, education level, and income
as covariates. The independent variable was a composite of the number of Comorbidities and
Functional Impairment scores. Mediating variables were coping processes (Action Planning;
Support/Advice Seeking; Disengagement/Denial). The outcome was Emotional Distress.
Results: The mediation effect from Comorbidities/Functional Impairment to
Disengagement/Denial Coping to Emotional Distress was significant: Path coefficients:
Comorbities/Functional Impairment to Disengagement/Denial (Est.=1.839; CI: 1.257, 2.426),
and Disengagement/Denial to Distress (Est.=4.604; CI: 3.751, 5.457). Also, there was
significant direct effect of Comorbidities/ Functional Impairment on Distress (Est.=.437; CI:
0.552, 0.831); thus, the mediation was partial. Conclusions/Implications: Comorbidities and
functional impairment can contribute directly to emotional distress; moreover, the only
coping mechanism that was active in the relationship between comorbidities and distress was
disengagement/denial, which exacerbated rather than mitigated distress. These results
support the need to provide supportive care services for survivors to limit destructive coping
in the context of comorbidities in order to improve emotional well-being.

Symposium 4B

THE PATH FROM MULTIMORBIDITY TO DISABILITY: EXAMINING PSYCHOSOCIAL MODERATION

AND BIOLOGICAL MEDIATION

Dr. Elliot Friedman, PhD

Single and multiple chronic medical conditions have become more prevalent as the
population of older adults has grown; by some estimates, most adults over age 60 have more
than one chronic condition (multimorbidity). Multimorbidity not only presents challenges for
medical management and cost of care, but it markedly increases the risk of a range of adverse
health outcomes, including disability, cognitive impairment, and mortality. Importantly,
however, there is variability in these outcomes among those with multimorbidity. Studies on
S378 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

very long-lived individuals, for example, show that multimorbidity need not compromise
quality of life. Our research program seeks to better understand the factors that influence the
extent to which multimorbidity results in adverse health outcomes. We focus specifically on
the potential for psychosocial resources to mitigate the impact of chronic conditions as well as
the potential mediating role of biological processes, notably inflammation, that have been
implicated in a range of disease processes. This presentation will outline our guiding model
and present results from studies examining the interactive influences of psychological well-
being and inflammation on the links between multimorbidity and disability in particular. We
have shown, for example, that higher ratings of eudaimonic well-being (e.g. purpose in life;
fulfilling personal relationships) are linked to lower circulating levels of inflammatory proteins
in general, and in those with multimorbidity specifically, compared to lower levels of well-
being. Moreover, we have shown in longitudinal analyses that while multimorbidity increases
later disability, greater eudaimonic well-being moderates this association. Finally,
inflammation partially mediates the association of multimorbidity and later disability. As well-
being in older adults appears to be modifiable, the results from this work suggest that
interventions targeting well-being may have the potential to improve health outcomes across
a range of chronic medical conditions and improve overall quality of life.

Symposium 4C

MULTIMORBIDITY AND PHYSICAL ACTIVITY IN ATLANTIC CANADIANS

Dr. Melanie Keats, PhD

Physical inactivity has been identified an important contributor to global chronic disease

burden and as the fourth leading risk factor for mortality. Worldwide, physical inactivity is

estimated to contribute to as much as 21-25% of breast and colon cancer, 27% of diabetes,

and approximately 30% of ischemic heart disease burden. Given the shared lifestyle risk

factors and the potential of a shared underlying biology for some diseases (e.g.,

inflammation), the co-occurrence of more than one chronic disease (i.e., multimorbidity) is

becoming an increasing concern. While a wealth of data is available to show that physical

activity is an effective preventive measure for more than 25 chronic diseases, interestingly

although emerging data suggests an important dose-response relationship, much less is

ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S379

known about the potential role of physical activity reducing the burden of multimorbidity.

Similarly, while the relationship between the presence of a chronic illness and physical activity

has been reported, the impact of multiple chronic illnesses on physical activity levels has not

been well described. This retrospective, population-based study will draw data from over

31,000 participants from the Atlantic Partnership for Tomorrows Health (PATH) cohort.

Atlantic PATH is part of Canadian Partnership for Tomorrow Project, a national prospective

cohort study examining the influence of genetic, environmental, and lifestyle factors in the

development of cancer and chronic disease. In brief, we will describe the prevalence of

physical activity in individuals with a self-reported history of a chronic disease (e.g., cancer,

cardiovascular disease, diabetes). We will also report the association between multimorbidity

(i.e., 2 or more chronic illnesses) and physical activity levels. Implications of findings will be

shared and discussed.

CORRESPONDING AUTHOR: Trevor Dummer, PhD, School of Population & Public Health,
University of British Columbia; [email protected]
S380 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 5 8:00 AM-9:15 AM

DESIGNING BEHAVIORAL HEALTH TECHNOLOGIES FOR ADOLESCENTS AND YOUNG ADULTS:

FINDING COMMONALITIES ACROSS PATIENT GROUPS

Emily Lattie, Ph.D.1, Sherif M. Badawy, MD, MS, MBBCh2, Carly M. Goldstein, PhD3, Sarah S.
Jaser, PhD4

1
Northwestern University, Chicago, IL; 2Ann & Robert H. Lurie Childrens Hospital of Chicago,
Northwestern University Feinberg School of Medicine, Chicago, IL; 3Weight Control and
Diabetes Research Center; Brown Medical School, Providence, RI; 4Vanderbilt University
Medical Center, Nashville, TN

Technology-based interventions, such as those delivered via mobile phones and online, offer
the possibility to expand behavioral health treatment options and to reduce barriers to
services. However, sustained long-term engagement with such interventions remains a
challenge, particularly in younger age groups. Adolescents and young adults (AYA) appear to
be ideal candidates for many of these technology-based interventions because they are
typically digital natives and confident in their ability to access and navigate different digital
interfaces. However, counter to older adults, AYA tend to view usability issues as the fault of
the intervention or website and are quick to disengage when their needs and expectations are
not met. For this session, we will present research examining different methods of using
technology to promote healthy behaviors in AYA subgroups. The first presenter will discuss
findings from a study on college student interest in and willingness to use mobile phone apps
for a variety of health behaviors, and will provide tips on collaborating with technologists with
designing for young adults. The second presenter will discuss her work in developing and
testing a positive psychology intervention for teenagers with Type 1 diabetes delivered via
text message and automated voice messages. The third presenter will share insights from his
work in designing a mobile phone app to improve medication adherence and self-
management in AYA with sickle cell disease. Finally, the discussant will provide a summary of
these findings and lessons learned to date in designing behavioral health technologies for
AYA. She will lead a discussion regarding ways to move the intervention field forward, with
attention to sharing knowledge across disease/target behavior subgroups and capitalizing on
the diverse expertise found in the members of the Society for Behavioral Medicine.

Symposium 5A
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S381

HU-GO APP: A NOVEL ENHANCING INTERVENTION FOR HYDROXYUREA IN ADOLESCENTS AND

YOUNG ADULTS WITH SICKLE CELL DISEASE

Dr. Sherif M. Badawy, MD, MS, MBBCh

Background: Sickle cell disease (SCD) is the most common genetic disorder in the US, seen in
80-100,000 Americans. Hydroxyurea (HU) reduces morbidity, and mortality, and lowers
healthcare utilization in SCD. Yet HU adherence is suboptimal. Data on the use of technology
to improve HU adherence in adolescents and young adults (AYA) with SCD are limited.

Aims: (1) to determine access to smartphone technology among AYA with SCD and to identify
preferred technology-based strategies for improving medication adherence; (2) to develop a
smartphone app (HU-Go) to improve HU adherence using iterative user-centered design; and
(3) to conduct a feasibility trial.

Methods: Part I: A cross-sectional survey was administered to AYA (1222 yrs) with SCD. Part
II: A usability phase of HU-Go app (n=5) over 4-weeks and a feasibility phase (n=15) over 12
weeks.

Results: Part I: Eighty participants completed the survey (51% female, 94% Black), and had a
mean age of 15.32.8 years. Most owned or had access to smartphones (85%), tablets (83%),
laptops (72%), or desktops (70%). The top app feature was daily medication reminders,
followed by education about SCD, adherence text prompts, education about SCD medications,
and medication log. Part II: Based on patients input and our conceptual model, we were able
to develop a novel multi-functional HU-Go app with personalized features as a single platform
that provides patients with: a) education (SCD/HU), b) reminders for daily medications, refills
and appointments, c) interactive communication with SCD team, d) community resources, and
e) health tracker (medications, pharmacies, event/symptom log, quality of life, patient portal
and wearable sensors). All patients approached agreed to enroll. The usability phase is
ongoing, and the feasibility trial will begin in October 2016.

Conclusions: Most AYA with SCD owned or had access to smartphones or other electronic
devices. The development of a multi-functional smartphone app to improve HU adherence is
feasible. The feasibility trial of HU-Go will provide a "proof-of-concept" evidence for the
feasibility and acceptability of a novel app-based approach to disease management that we
plan to extend to AYA with other chronic conditions.

CORRESPONDING AUTHORS: Sherif Badawy, MD, MS, MBBCh, Ann & Robert H. Lurie
Childrens Hospital of Chicago; Northwestern University Feinberg School of Medicine;
[email protected]; Alexis Thompson, MD, MPH, Ann & Robert H. Lurie Childrens
Hospital of Chicago; Northwestern University Feinberg School of Medicine; a-
[email protected]; Robert Liem, MD, MS, Ann & Robert H. Lurie Childrens
S382 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Hospital of Chicago; Northwestern University Feinberg School of Medicine;

[email protected]

Symposium 5B

WILLINGNESS TO UTILIZE SMARTPHONE APPS FOR HEALTH INTERVENTIONS: IMPORTANT

CONSIDERATIONS FOR EMERGING ADULTS

Dr. Carly M. Goldstein, PhD

Objective:

Currently, 86% of emerging adults (ages 18-29) own smartphones. This studys purpose was to
measure willingness to use smartphone applications (apps) to monitor or change health
behaviors and common targets of behavioral interventions.

Methods:

Undergraduates (n=249) at a large, Midwestern university were recruited to complete surveys

about engagement in and willingness to use apps to monitor health behaviors through an
online secure-socket layer technology program.

Results:

This sample of emerging adults (19.78 2.98 years; 91.2 % white; 81.9% female) were most
interested in using apps to target exercise patterns, nutrition, calorie intake, stress, and
overall health; they were less willing to target body image, sleep patterns, and anxiety. Higher
BMI was related to willingness to use apps for overall health ( = .15, p < .05). Greater
perceived body fatness in the lower body was related to higher willingness to use apps to
monitor diet ( = .14, p < .05), body image ( = .24, p < .001), caloric intake ( = .22, p < .01),
and overall health ( = .13, p < .05). Higher feelings of fatness predicted higher willingness to
use apps for monitoring body image ( = .29, p < .001), diet ( = .20, p < .01), calorie intake (
= .24, p < .01), and exercise ( = .15, p < .05). More intense body loathing scores were related
to willingness to use apps to monitor body image only ( = .13, p < .05). Worry was related to
increased willingness to use apps to monitor anxiety ( = .17, p < .05).

Discussion:

Emerging adults willingness to use technology was related to numerous measurable factors
within the present study. Researchers are encouraged to design interventions to match
emerging adults willingness and needs; researchers interested in designing app-based or app-
inclusive interventions should allot additional development focus to matching design
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S383

intervention components uniquely tailored for this age group, their preferences, and potential
factors that could improve engagement.

CORRESPONDING AUTHORS: Brooke L. Bennett, MS, University of Hawai'i at Manoa;

[email protected]; Emily C. Gathright, MA, Brown Medical School, Kent State University;
[email protected]; KayLoni L. Olson, MA, Brown Medical School, The Ohio State
University; [email protected]

Symposium 5C

USING TECHNOLOGY TO DELIVER A POSITIVE PSYCHOLOGY INTERVENTION FOR ADOLESCENTS

WITH TYPE 1 DIABETES

Sarah S. Jaser, PhD

The majority of adolescents with type 1 diabetes (T1D) are not meeting treatment goals, and
many struggle with adherence to the intensive treatment regimen. Positive psychology
interventions have been shown to improve adherence in adults with chronic illness (e.g.,
Charlson et al., 2010), but they have not been widely tested in pediatric populations. In our
study, we adapted the protocol successfully used in adults to promote positive affect,
including self-affirmations, small gifts, and gratitude, for adolescents with T1D, by increasing
the frequency of contact from bi-weekly to weekly and including parental affirmations.

Adolescents with T1D (n = 120, mean age 14.8 yrs, 52.5% female, 87.5% White) were
randomized to either an education (n=60) or positive affect (PA) intervention (n=60). We used
automated phone calls (interactive voice technology, n = 30) and text messages (SMS, n=30)
to deliver the PA reminders and to collect data on adolescents mood during the 8-week
intervention. There was a significant difference in the response rate to the weekly PA
messages: only 14.3% of adolescents answered or returned phone calls, as compared to
67.5% of adolescents who responded to text messages (p<.001). The response rate to the
mood assessments were not significantly different between text message (68.5% complete)
vs. interactive voice technology (57.8%) groups (p = .134). Exit interviews revealed that the
greatest barrier to participation was timing/inconvenience. The PA intervention had a positive
effect on quality of life; adolescents in the PA group demonstrated significant improvement in
quality of life at 3 months, compared to the education group (estimated mean difference =
3.9, p = .020), after adjusting for age, sex, race/ethnicity, income, depressive symptoms, pump
use, and baseline measurement for each outcome. In addition, the PA group showed a trend
toward improvement in adolescents adherence (estimated mean difference = 1.2, p=.129),
particularly in the text group (1.5, p=.136). The intervention had no effects on glycemic
control.
S384 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

The next phase of our work will be to administer the intervention solely through text
messages, increasing the frequency of messages (from weekly to daily), and enhancing
involvement of parents (training on parental affirmations).
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S385

Symposium 6 8:00 AM-9:15 AM

PHYSICAL ACTIVITY AND WELLNESS: IMPLICATIONS FOR INTERVENTIONS IN AGING

POPULATIONS

Brenna N. Renn, PhD1, Neha Gothe, MA, MS, PhD2, Susan Aguiaga, PhD3, Elizabeth Awick,
MS4, Kelsey C. Bourbeau, MA5, Sandra Jane. Winter, PhD, MHA6

1
University of Washington School of Medicine, Seattle, WA; 2Wayne State University, Detroit,
MI; 3University of Illinois at Urbana-Champaign, Urbana, IL; 4University of Illinois at Urbana
Champaign, Urbana, IL; 5Wayne State Univeristy, Warren, MI; 6Stanford University School of
Medicine, Palo Alto, CA

The health benefits of physical activity (PA) are extensively documented across the lifespan.
However, many aging individuals engage in low levels of PA and have an elevated risk of
developing chronic diseases associated with inactivity. Understanding the benefits of PA in
this population is necessary to better tailor PA interventions and engage individuals in
behavior change. In this symposium, presenters will describe data regarding PA among aging
adults in the context of wellness and quality of life (QoL), with special attention to
implications for behavioral interventions. The first presenter will discuss perceived benefits
and affective experience of PA in midlife, which is an age group that represents a key
developmental time point for improving health and preventing age-related chronic disease
processes associated with physical inactivity. Her data provide a foundation to further
investigate motives for PA among aging individuals to inform interventions and improve
wellness. The second speaker will present data from an exercise intervention for older adults
in which PA improved global QoL indirectly by reducing distress. The presenter conceptualizes
distress broadly (i.e., anxiety, depression, perceived stress, and disturbed sleep) and will
underscore the importance of these variables in PA interventions for greater well-being in
older adults. The third presenter draws on findings from a Latin dance program for Spanish-
speaking older adults with mild cognitive impairment. While she noted improved physical
function during the course of the intervention, such improvements were also observed in
wait-list participants receiving other services in an enriched adult day center environment.
She will present her findings in the context of implications for future interventions with this
population who have increased risk of chronic disease. Finally, the fourth speaker will discuss
how PA contributes to physical functioning in older adults who have suffered a stroke. Her
findings, based on objectively measured activity levels, suggest that both light and moderate-
to-vigorous activity play an important role in functional ability in this vulnerable population.
The discussant will comment on these findings and initiate a discussion from the perspective
of health promotion interventions among older adults. Together, these presentations
highlight the importance of PA promotion as part of general wellness and QoL in aging.
S386 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 6A

PERCEIVED BENEFITS AND OUTCOME EXPECTATIONS OF PHYSICAL ACTIVITY IN MIDLIFE

Dr. Brenna N. Renn, PhD

Midlife represents a key time point for engaging in behaviors to promote health and forestall
age-related chronic disease. Physical activity (PA) is one such behavior of interest. However,
adults become less active with age, and there is a scarcity of literature examining
psychological factors associated with PA in middle-aged adults. Social cognitive theory (SCT)
highlights the role of outcome expectations in both the adoption and maintenance of PA, but
little attention has been paid to such influence in middle age.

The present study examined cross-sectional outcome expectations of PA in 195 middle-aged

adults (Mage = 58.294.15 years, range 50-64 years, 65.6% female, 9.4% ethnic minority)
recruited from community settings. Primary outcomes were the Exercise Benefits/Barriers
Scale (EBBS) and Physical Activity Enjoyment Scale (PACES). Physical activity was assessed
using the short-form International Physical Activity Questionnaire (IPAQ-SF). Additional SCT
variables of self-efficacy and self-regulation were assessed.

Less than half (43.1%) of the sample reported activity consistent with national guidelines (150
min/week of moderate-to-vigorous PA [MVPA]). Although the majority of the sample was
aware of the benefits of regular PA, individuals who reported 150 min/week of MVPA
perceived significantly more benefits of activity than their less active counterparts, F(1, 182) =
35.61, p < .001, p2 = .16. Those meeting MVPA guidelines also reported higher levels of
enjoyment associated with PA, F(1, 182) = 47.89, p < .001, p2 = .21. Both active and
sedentary respondents perceived physical fitness benefits from PA, including improved
cardiovascular functioning and muscle strength. However, those who reported meeting
MVPA guidelines were more likely to perceive psychological benefits of PA, such as decreased
stress, improved mental health, sense of personal accomplishment, and improved well-being
(ps < .001). After controlling for total activity levels, greater enjoyment of PA (beta = .47, p <
.001) and greater use of goal-setting strategies for PA (beta = .24, p = .002) predicted greater
perceived benefits of PA.

Causal inferences are limited by the cross-sectional methodology, but these data suggest that
appreciating both physical and psychological benefits of PA are associated with higher levels
of MVPA. Given the importance of outcome expectations in predicting engagement in PA,
future research should explore how best to leverage or modify such outcome expectations.
Identifying and addressing how to increase perceptions of benefits and enjoyment of PA may
lead to more effective interventions and sustained behavior change among aging individuals.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S387

CORRESPONDING AUTHOR: Leilani Feliciano, PhD, University of Colorado-Colorado Springs;

[email protected]

Symposium 6B

IMPACT OF A LATIN DANCE PROGRAM ON PHYSICAL ACTIVITY AND PHYSICAL FUNCTION IN

OLDER LATINOS WITH COGNITIVE IMPAIRMENT

Susan Aguiaga, PhD

Older Latinos have a greater risk of developing chronic conditions involving modifiable disease
factors in comparison to non-Latino whites. Presence of these chronic diseases, compounded
with poverty, may increase the risk of mobility disability for this population. Physical activity
(PA) has been shown to improve chronic diseases and may protect against mobility disability.
Unfortunately, older Latinos participate in low levels of leisure-time PA. Thus, the purpose of
this study was to examine the impact of a Latin dance program on PA and physical function
among older Latinos with mild cognitive impairment (MCI) attending an adult day center.
Spanish-speaking older Latinos [N=21, 75.4 6.3 years old, 76.2% female, 22.4 2.8 MMSE
score] were randomized into a 16 week, twice-weekly dance intervention or to a wait-list
control group; the wait-list control group received the dance intervention at week 17.
Participants were given a waterproofed ActivPal inclinometer to wear on their non-dominant
thigh for 7 consecutive days. Physical function was assessed using the Late-life function and
disability instrument (LLFDI), Short Physical Performance Battery (SPPB), and the Timed up
and go (TUG). Data was collected at baseline, month 2, 4, 6, and 8. Repeated measures
ANOVA were used to determine overall time effect and time x group interaction effects. We
calculated Cohens d values as an estimate of the effect sizes. Results revealed no significant
time or interaction effects for light or moderate-vigorous PA as measured by the ActivPal, and
no effects for the LLFDI. There was a time effect for the chair stand component of the SPPB, F
(2, 38) = 10.02, p= .001, d= .83, and the TUG, F (2, 38) = 8.27, p = .001, d =.74. These results
suggest that a Latin dance program for older Latinos with MCI has modest effects on physical
function, but not more so than a wait-list control group. Whether improvement was due to
our intervention or due to the enriched adult day center environment, these changes may be
clinically meaningful for this population. Researchers interested in this population should
consider conducting research with older Latinos with MCI who are not attending adult day
centers in order to determine if PA interventions are efficacious among individuals who do not
have access to these centers.

CORRESPONDING AUTHOR: David X. Marquez, PhD, University of Illinois at Chicago;

[email protected]

Symposium 6C
S388 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

PHYSICAL ACTIVITY, DISTRESS AND QUALITY OF LIFE IN OLDER ADULTS: A PANEL MODEL

Elizabeth Awick, MS

Physical activity has been consistently and significantly associated with improved life
satisfaction in older adults. However, less is known about the mediating effects of changes in
distress on this relationship, specifically within the context of an exercise intervention. The
purpose of the present study was to examine how changes in moderate to vigorous physical
activity (MVPA) and distress influenced changes in quality of life (QOL) in older adults across a
six-month period. Community-dwelling older adults (N=247, M age=65.39) were recruited to
participate in a six-month randomized controlled exercise trial designed to examine cognition
and brain function. Participants wore accelerometers to measure MVPA and completed
measures of psychological health outcomes and QOL at baseline and after the intervention,
six months later. Distress was defined as a latent composite of self-reported measures of
anxiety, depression, sleep disturbance, and perceived stress. In a panel model, parallel
changes in MVPA, distress, and QOL during the six-month intervention were estimated as
latent change scores, including constraints for measurement invariance over time. The
hypothesized model provided excellent fit to the data (=62.099(53), p=.184, RMSEA=.026
[90% CI=.000, .050], CFI=.992, TLI=.990, SRMR=.037). Overall, individuals with greater
increases in average daily MVPA over the intervention period demonstrated significantly
reduced distress from baseline to 6 months ( = -.18, p<.05). In turn, individuals with greater
reductions in distress across the intervention reported significantly greater increases in QOL
from baseline to 6 months ( = -.30, p<.001). Reductions in distress partially explained the
benefits to QOL from increased MVPA [indirect effect = 0.05; bias-corrected bootstrapped
95% CI=.005, .125]. These findings suggest that several domains of psychological health (e.g.,
anxiety, depression, sleep disturbance, stress) may play an important role in the influence of
physical activity behavior change on global life satisfaction. Future research should target such
psychological factors in concurrence with physical activity behavior for greater improvements
in well-being in an aging population.

CORRESPONDING AUTHORS: Jason Fanning, PhD, Wake Forest University; [email protected];

Diane Ehlers, PhD, University of Illinois; [email protected]; Susan Aguinaga, PhD,
University of Illinois; [email protected]; Ana Daugherty, PhD, University of Illinois;
[email protected]; Arthur Kramer, PhD, Northeastern University; [email protected];
Edward McAuley, PhD, University of Illinois at Urbana-Champaign; [email protected]

Symposium 6D

RELATIONSHIP BETWEEN LIGHT PHYSICAL ACTIVITY AND FUNCTIONAL FITNESS IN STROKE

SURVIVORS
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S389

Kelsey C. Bourbeau, MA

Purpose:

Impairment caused by stroke is a major cause of disablement in older adults. Loss of physical
function often results in a decreased ability to perform activities of daily living. Physical
activity has been shown to improve physical functioning, however little research has been
done to explore how activities of different intensities may affect functional fitness outcomes
in stroke survivors. The purpose of this study was to examine the relationship between
physical activity and self-reported and objective functional fitness measures in this
population.

Methods:

Stroke survivors (N=30, Mean age=61.7711.17) completed the Short Physical Performance
Battery (SPPB) and the Late-Life Function and Disability Instrument (LLFDI). Physical activity
levels were measured objectively using a 7-day Actigraph accelerometer wear period and
scored using the Copeland cutoffs for sedentary (counts/minute 50), light (counts/minute
<1,041), and moderate to vigorous (MVPA; counts/minute 1,041) activity.

Results:

Light activity significantly correlated with total SPPB score (r= .390 p=.033). Light activity also
significantly correlated with several subscales of the function component of the LLFDI,
including: basic lower extremity (r= .490, p=.006), advanced lower extremity (r= .568, p=.001),
upper extremity (r= .465, p=.010) and total function (r= .585, p=.001). MVPA significantly
correlated with total SPPB score (r= .491, p=.006), and the LLFDI including: advanced lower
extremity subscale (r= .443, p=.014) and total function (r=.380, p=.038).

Conclusion:

These findings suggest that light activity and MVPA may contribute to better physical
functioning in stroke survivors. While MVPA had a greater association with objective
measures of physical function, light activity displayed greater significant correlations across all
subscales of the self-reported function component of the LLFDI than MVPA. Disabilities
resulting from stroke may limit this population from engaging in MVPA, and these findings
highlight the importance of light physical activity, which may offer similar functional benefits.
Future studies should focus on development of effective exercise interventions for stroke
survivors by incorporating not only MVPA but also light intensity physical activities.
S390 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHORS: Bradley Kendall, MS, Wayne State University;

[email protected]; Neha Gothe, PhD, Wayne State University;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S391

Symposium 7 8:00 AM-9:15 AM

LOCAL AREA HEALTH ASSESSMENT: COLLECTING ACTIONABLE DATA FOR PLANNING,

RESEARCH AND INTERVENTIONS

Richard P. Moser, PhD1, Kate Brett, PhD2, John M. Boyle, PhD3, Kelly D. Blake, ScD4, Bradford
Hesse, PhD5

1
National Cancer Institute, Bethesda, MD; 2National Center for Health Statistics/CDC,
Hyattsville, MD; 3ICF International, Rockville, MD; 4Division of Cancer Control and Population
Sciences, National Cancer Institute, National Institutes of Health, Bethesda, MD; 5National
Cancer Institute, Rockville, MD

Increasingly, health systems such as cancer centers and other hospitals are being held
responsible for assessing and improving the health of people in their local catchment areas,
not just those patients who walk in their doors. These data can be used by providers,
planners and health researchers to improve the health of a local area, however defined, and
ultimately contribute to improved public health. In order to accurately assess these needs at
a local level, however, relevant and precise data are needed. Having these precise and
accurate data are critical as national data will not contain the detailed information to identify
the needs of a local area such as a state, county, city or even a neighborhood. These local data
can then be compared with results from national health-related surveillance systems, such as
the Health Information National Trends Survey (HINTS), to identify and understand the needs
and potential unique health disparities of a local area. These health concerns, for example,
may include understanding rates of behaviors such as use of tobacco products including
combustible cigarettes and e-cigs, HPV vaccination rates, preferred health communication
media channels and other health-related knowledge, perceptions and attitudes. Data from
these local assessments can potentially be disaggregated to understand and target the needs
of important racial/ethnic subgroups or other under-represented groups, and together, be
used to improve local and larger population health outcomes. This symposium will present
several efforts to perform local area health assessments including: 1) a measurement
framework to support community health; 2) research using non-probability survey collection
methods to assess health outcomes for local geographic areas; and 3) a description of health
assessments being done in cancer center catchment areas. These health assessments are
critical and can provide researchers, providers, coalitions and local planners with important
information to stimulate hypothesis-driven research, better characterize local disparities and
inequities and develop appropriate evidence-based interventions.
S392 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 7A

A MEASUREMENT FRAMEWORK FOR COMMUNITY HEALTH AND WELL-BEING: REPORT ON

BEHALF OF THE NATIONAL COMMITTEE ON VITAL AND HEALTH STATISTICS

Dr. Kate Brett, PhD

As healthcare pivots toward greater accountability for population health, professionals in

behavioral medicine are being called upon to design and implement actionable programs for
disease prevention and health promotion. This shift takes place at a time many communities
are driving to achieve health equity. Despite many increasingly favorable conditions,
communities have articulated their ongoing struggle with barriers that make it difficult for
them to work effectively with data to decide where and how to intervene. Although a host of
indicators and metrics exists, there is a shortage of sub-county data, compounded by other
data gaps and by issues with timeliness, standardization, analytic capacity, and data
stewardship. More fundamentally, many communities remain uncertain about precisely what
they should measure, given the multitude of health determinants. On September 27 2016, the
Population Health Subcommittee of the National Committee on Vital and Health Statistics
convened a daylong workshop to promote multi-sector approaches for improving community
health and well-being. In this symposium, the lead staff member of the Subcommittee will
provide SBM conference attendees an in-depth report on the conclusions and follow-up
actions reached by members of the subcommittee. Topics to be covered will include:
(a) status of a multi-sector measurement framework to serve as the basis for a public/private
collaboration at the community-level; (b) opportunities and gaps in sub-county level data
metrics; (c) opportunities to align, leverage, and build multi-sectoral sub-county level metric-
centric efforts to improve health and well-being; and (d) report on collaborative efforts to
continue work in this arena.

Symposium 7B

USING NON-PROBABILITY SURVEY DATA COLLECTION METHODS TO ASSESS HEALTH NEEDS,

BEHAVIORS AND OUTCOMES FOR LOCAL GEOGRAPHIC AREAS

John M. Boyle, PhD

Most public health programs are implemented and evaluated at the community level, but few
public health surveillance systems generate community measures. National health surveys
lack the sample size to provide estimates at the community level. Local area health surveys
could enhance the information needed by researchers, providers and planners to assess
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S393

intervention needs and program outcomes at the local level. However, costs for traditional
probability surveys have been a barrier to adoption of community health surveys.

Non-probability samples provide a low cost and quick turnaround alternative to traditional
probability samples for local area assessments. Non-probability samples are not based on
statistical models that permit variance estimates and confidence intervals about their
estimates. However, they are not necessarily inferior to probability samples in other aspects
of total survey error. They are commonly used in evaluations, including clinical trials.

In this paper, we will describe the methods used in several recent surveys we have conducted
using non-probability web panels and results related to health needs, behaviors and
outcomes. This information includes health status, exercise, height, weight, tobacco use,
doctor visits, screening tests, and other health-related knowledge, attitudes, beliefs and
behaviors. We also tested respondents willingness to use new technology, like smartphones
to take pictures of their health records and their willingness to share their electronic health
records with health agencies.

CORRESPONDING AUTHORS: Lewis Berman, ICF International; [email protected];

Ronaldo Iachan, ICF International; [email protected]

Symposium 7C

POPULATION HEALTH ASSESSMENT IN CANCER CENTER CATCHMENT AREAS

Dr. Kelly D. Blake, ScD

The National Cancer Institute (NCI) recently announced the opportunity for supplemental
funding to conduct research to better characterize the populations and communities within
the NCI-designated cancer center catchment areas. Supplemental funding was provided to 15
cancer centers to enhance their capacity to acquire, aggregate and integrate data from
multiple sources, as well as plan, coordinate, and enhance catchment area analysis activities,
including reviewing and selecting measures and planning a strategy for enhancing data
collection in the catchment area. The short-term goal is to conduct local research to better
define and describe the cancer center catchment area using a multilevel population health
framework. The long-term goal is to facilitate collaborations in which local area providers,
public health practitioners, policymakers, and nonprofit organizations can utilize data to
develop or expand applied cancer control research, planning, intervention, and
implementation efforts, with particular attention to local level health disparities and
communication inequalities. Although there are several national level, public use datasets for
health behavior research, national surveillance vehicles such as HINTS, BRFSS, and NHIS do
not collect a large enough sample at smaller geographic units to be optimally useful at the
local level for targeted outreach, intervention planning, clinical trials recruitment, and patient
S394 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

engagement. A number of methodological questions remain as to the feasibility of using small

area samples to complement data streams embedded within national surveys. How do small
area estimates compare to national estimates? How do nonprobability samples compare to
random samples in terms of bias, error, coverage, and reliability? Can bias in local data be
overcome by calibrating estimates to national anchors? The research funded by NCIs
population health assessment initiative will help to answer these methodological questions,
and by so doing should improve the viability of aggregating local and national data to inform
local needs assessment and planning in cancer prevention and control. We will report on
innovative data collection methods being used across the funded sites, as well as common
constructs and measures that will be used for local and national comparisons.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S395

Symposium 8 8:00 AM-9:15 AM

CONTEXTUALIZING HEALTH BEHAVIORS AND OUTCOMES: THE ROLE OF GEOGRAPHY

Jennifer L. Moss, PhD1, Kevin Cassel, DrPH, MPH2, Natalie Colabianchi, PhD3, S. Bryn Austin,
ScD4, Rebecca E. Lee, PhD5

1
National Cancer Institute, Bethesda, MD; 2University of Hawaii Cancer Center, Honolulu, HI;
3
University of Michigan, Ann Arbor, MI; 4Boston Children's Hospital, Boston, MA; 5Arizona
State University, Phoenix, AZ

Behavioral medicine has traditionally focused on intrapersonal determinants of health risks

and outcomes. Increasingly, however, researchers have placed more emphasis on the role
of geography in contextualizing health behaviors. Geographic disparities in health reflect
variation in factors such as norms, sociodemographics, access, and the built environment,
which can have direct effects on health or can be mediated through individual-level
constructs. Understanding the influence of neighborhoods, communities, and environments
on health behaviors is important for developing messages and interventions that can
effectively promote public health. A variety of data collection and analytic techniques are
available for researchers interested in studying geography and health. In this symposium,
we propose to present three studies that illustrate the breadth of opportunity for
incorporating geographic perspectives into health behavior research. First, Speaker 1 will
discuss research on childhood obesity that examines the role of the built environment
across 130 U.S. communities. Next, Speaker 2 will present results from a geographic
information systems (GIS) analysis of census block groups that have higher-than-expected
numbers of tanning salons across seven large cities in the U.S. Then, Speaker 3 will describe
findings from an in-depth analysis of the role of social and community factors for health
behaviors and well-being among residents of a designated Hawaiian home land. Finally, the
Discussant will integrate findings and themes from these presentations and suggest
directions for future innovations in this field of research. This symposium will highlight
creative analyses of how geography contextualizes a variety of health behaviors and
outcomes and will consider implications for future behavioral medicine research studies.

Symposium 8A

THE PAPAKOLEA HAWAIIAN HOMESTEAD COMMUNITY HEALTH SURVEY

Dr. Kevin Cassel, DrPH, MPH

S396 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

In an effort led by Prince Kalanianaole, Congress passed the "Hawaiian Homes Commission
Act" (HHCA) in 1921, which set aside 200,000 acres in Hawaii as a land trust for Native
Hawaiians. This trust was recompense for the U. S. siege of Crown lands in 1893. The HCCA
was established to help Hawaiians maintain traditional ties to their land and grow culturally
important crops. Today, many homesteads are located in lower socioeconomic areas
associated with poor health outcomes. However, there has not been a comprehensive health
assessment of Hawaiian homesteads and their unique geographical locations.

Papaklea, Kewalo and Kalawahine are neighboring homesteads, jointly referred to as

Papaklea, established in 1934 in Honolulu. It is the only homestead location chosen by
Hawaiians. In 2014, Papakoleas community-based organization, Kula no na Poe Hawaii,
expanded a partnership with the University of Hawaii to assess the health of residents. The
information was intended to inform programs to address the communitys health needs.

In 2015, a survey was developed, pilot-tested, and mailed to 390 lessees and/or residents in
Papakolea. The survey consisted of five sections, including questions on: 1) Demographics -
age, ethnicity, education, income and household size; 2) Health Status - diagnosis of
hypertension, cancer, diabetes, and other medical conditions; 3) Health Behaviors diet and
physical activity; 4) Cancer Screening - mammography, Pap , prostate, and colorectal tests; 5)
Psychosocial Factors - life satisfaction, flourishing

In all 124 surveys (32%) were completed. Most respondents were ages 50 to 70 (55%), female
(72%), and Native Hawaiian (99%). Many residents (35%) reported very good or excellent
health, 35% good health, and 26% poor or fair health. More residents reported poor health
than that reported by the general population (15%) and Hawaiians statewide (19%). Most
residents (56%) reported hypertension. Cancer screening data indicated most residents were
screened for colon cancer (72%), had Pap testing (97%) and mammography (99%).

The Papakolea Hawaiian Homestead Community Health Survey was one of the first health
assessments of a Hawaiian homestead. The survey provided important data on residents
health status, identified strengths, including high rates of cancer screening and areas needing
interventions. The survey provides a model for the planned assessment of 30,855 residents in
49 Hawaiian homestead communities across 7 islands.

CORRESPONDING AUTHORS: Kevin Cassel, Dr PH, University of Hawaii Cancer Center;

[email protected]; Claire Townsen-Ing, DrPH, University of Hawaii John A. Burns School of
Medicine; [email protected]; Mapuana Antonio, DrPH(c), University of Hawaii John A.
Burns School of Medicine; [email protected]; Adrienne Dillard, MSW, LSW, Kula no na
Po'e Hawaii; [email protected]; Puni Kekauoha, Papakolea; [email protected];
Joseph Keawe'aimoku Kaholokula, PhD, University of Hawaii John A. Burns School of
Medicine; [email protected]

Symposium 8B
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S397

THE BUILT ENVIRONMENT AND CHILDHOOD OBESITY: THE HEALTHY COMMUNITIES STUDY

Dr. Natalie Colabianchi, PhD

The built environment has been shown to have an important influence on childhood obesity.
For children, built environment characteristics of the street segment on which they live may
be an important influence on their weight status. This study documented five different street
segment characteristics using an in-person audit tool for 3,600 children in 130 communities.
Specifically, presence of litter, residential quality, presence and quality of sidewalks, street
type (e.g., major/moderate busy thoroughfare) and presence of burned up/abandoned
houses were examined. In addition, an index was created that combined these 5 items. Body
Mass Index (BMI) was calculated from objectively measured height and weight and converted
to BMI z-scores. Two separate linear mixed models were used to examine the association
between 1) the individual street segment attributes and BMI z-score and 2) the overall street
segment index and BMI z-score. Both models controlled for age, gender, race, ethnicity,
family income, parental BMI, and community poverty rate, and adjusted for the clustering of
participants within communities. The sample was between the ages of 4 and 15. Forty-five
percent of the sample identified their ethnicity as Hispanic (via parental report). Two-thirds of
the sample identified their race as White while 22% of the sample identified their race as
African-American. About 50% had a family income under $35,000. The sample was nearly split
between boys and girls. When the five individual street segment components were examined
in the same model, street type was significantly associated with BMI z-score (p=.04) after
controlling for the other covariates. Those who lived on streets that were not
major/moderate busy thoroughfares had lower BMI z-scores relative to those who did live on
those types of streets. The overall index of street characteristics was significantly associated
with BMI z-score in adjusted models (p=.01) with those who had more favorable street
segment characteristics having lower BMI z-scores. These findings highlight the role the built
environment plays on childhood BMI, in particular the area near their home.

CORRESPONDING AUTHORS: Natalie Colabianchi, PhD, University of Michigan;

[email protected]; Kerry McIver, PhD, University of South Carolina;
[email protected]; Marsha Dowda, DrPH, University of South Carolina;
[email protected]; Russell Pate, PhD, University of South Carolina;
[email protected]

Symposium 8C

PREDICTING SPATIAL DISTRIBUTION OF BUSINESSES OFFERING UV INDOOR TANNING ACROSS

SEVEN U.S. METRO AREAS

Dr. S. Bryn Austin, ScD

S398 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Introduction: Societal appearance ideals drive consumer demand for UV indoor tanning,
which now generates approximately $3 billion in revenue annually in the US. The rapid growth
and infiltration of UV tanning into other industries (e.g., gyms) has been followed by a
corresponding rise in melanoma. While proximity to businesses offering UV tanning is likely an
important predictor of use by consumers, little is known about their geographic distribution.
Methods: In fall 2014 and 2015, we gathered phone survey data on whether businesses
offered UV tanning from 21,387 businesses in 7 major U.S. metro areas (Boston, NYC, Chicago,
Seattle, LA, Dallas, Miami). We then linked business survey data to block group characteristics
(sociodemographics, land use, presence of businesses of all categories) to build multivariable
logistic regression models to predict probability of businesses offering UV tanning. We
generated predictive maps of the spatial distribution of these businesses. Results: UV indoor
tanning services were reported in the highest prevalence by tanning salons (90.2%), health
clubs (17.3%), gyms (10.8%), and laser-treatment and tattoo-removal businesses (both 8.0%).
The probability of UV tanning services in a block group was positively associated with percent
of population of white race/ethnicity (p<0.01) and higher density of businesses (p<0.05) and
negatively associated with percent of population with college degree (p<0.05). Sunbelt metro
areas (Miami, Dallas, LA) had lower prevalence of businesses offering UV tanning compared to
northern areas (p<0.01 for all 3 metros). Discussion: We observed important variability in
spatial distribution of UV tanning. Presence of these noxious services may both normalize UV
tanning as a societally valued behavior and increase use by nearby residents. Societal
idealization of tanned skin, particularly in populations of white race/ethnicity with lower
education in northern cities, fuels demand for UV indoor tanning, putting consumers at
elevated risk for melanoma.

CORRESPONDING AUTHORS: S. Bryn Austin, Adolescent Medicine, Boston Children's Hospital;

ScD; Longxiang Li, MS, Harvard Chan School of Public Health; [email protected];
Allegra R. Gordon, ScD, MPH, Boston Children's Hospital;
[email protected]; Jeffrey Blossom, MS, Harvard University;
[email protected]; Alan Geller, PhD, Harvard Chan School of Public Health;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S399

Symposium 9 8:00 AM-9:15 AM

ONE SIZE DOESNT FIT ALL: TAILORING INTERVENTIONS TO ADDRESS INDIVIDUAL VARIABILITY
IN TREATMENT RESPONSE

Christine Pellegrini, PhD1, Allison J. Carroll, M.S.2, Jessica Unick, PhD3, Michelle Drapkin, PhD4,
Angela Fidler Pfammatter, PhD2

1
University of South Carolina, Columbia, SC; 2Northwestern University Feinberg School of
Medicine, Chicago, IL; 3Warren Alpert Medical School at Brown University, Providence, RI;
4
Rutgers, The State University of New Jersey, New Brunswick, NJ

There are many established treatment approaches for improving health behaviors; however,
even the most effective programs are unlikely to improve behaviors across all participants.
While most programs adopt a one size all approach, adapting interventions based on
individual treatment response and adherence may be a more promising and cost-effective
approach. This symposium highlights the need to develop interventions that account for
variability in treatment response across multiple health behaviors. Additionally, speakers will
present potential approaches to target or adapt interventions to address this individual
variability in response. The first presenter will present the results from a systematic review
and meta-analysis evaluating the response to various smoking cessation treatments among
smokers with a history of depression, and discuss potential opportunities to develop targeted
treatment for these high-risk smokers. The second presenter will describe a preliminary study
exploring the effect of providing additional support to individuals who are not responding
early to an internet-delivered weight loss intervention. The results suggest that identifying
early non-responders and providing additional support can enhance adherence and weight
loss outcomes over a 3 month period of time. Finally, the third presenter will present two
sequential multiple assignment randomized trials (SMART) which examine whether offering
patients who are not engaged or drop out of treatment for substance abuse disorders
improve outcomes. Engagement and the results from making additional outreach or providing
choices to patients not engaged will be presented. The discussant will provide an overview on
individual variability in the response to treatment across several health behaviors, collectively
integrate the findings from the three presentations, and highlight evolving challenges and
future directions related to the development and implementation of adaptive interventions.

Symposium 9A
S400 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

DO ANTIDEPRESSANT SMOKING CESSATION TREATMENTS PRODUCE THE BEST OUTCOME FOR

DEPRESSED SMOKERS?

Allison J. Carroll, M.S.

Background: In a prior systematic review and meta-analysis (Hitsman et al. 2013 Addiction),
past major depression (MD) was associated with 17% and 19% lower odds, respectively, of
short- and long-term abstinence. Using the same database, this study examined whether
antidepressant smoking cessation treatments produce the best outcome for past MD
smokers. Methods: For each study in our 2013 meta-analysis (N=42), we coded the highest-
level experimental arm and the lowest-level placebo/alternative control arm for psychological
and pharmacological components. Four factors were coded: 1) standard behavioral treatment
(none, self-help, face-to-face); 2) cognitive behavioral mood management (CBMM; none, self-
help, face-to-face); 3) standard cessation pharmacotherapy (none, placebo, nicotine
replacement therapy [NRT], other); 4) and antidepressant cessation pharmacotherapy (none,
placebo, varenicline, serotonergic, catecholaminergic). We used Generalized Estimating
Equations, with study ID as the cluster identifier, to predict short- (3 months) and long-term
abstinence (6 months) using the coded treatment components. For the analysis of long-term
abstinence, we included separate terms to distinguish standard (Results: Short-term
abstinence (N=30, n=2,604) was significantly increased by CBMM face-to-face (6 arms; odds
ratio [OR]=2.04), NRT (21 arms; OR=2.86), catecholaminergic medication (11 arms; OR=1.36),
and varenicline (2 arms, OR=2.92). For long-term outcomes (N=35, n=2,836), only
catecholaminergic medication (14 arms; OR=1.49) and varenicline (2 arms, each 12-week
duration; OR=2.49) improved abstinence. Neither CBMM (6 arms) nor NRT (22 arms) had an
effect on long-term abstinence. Conclusion: Varenicline was shown to provide the greatest
long-term benefit for smokers with past MD, more than doubling long-term abstinence rates.
However, a novel, targeted behavioral treatment model may be needed to produce sustained
smoking cessation among these high-risk smokers.

CORRESPONDING AUTHORS: Brian Hitsman, PhD, Northwestern University Feinberg School of

Medicine; [email protected]; Amanda Mathew, PhD, Northwestern University
Feinberg School of Medicine; [email protected]; Ray Niaura, PhD, Johns
Hopkins University; [email protected]; Robert Schnoll, PhD, University of
Pennsylvania Perlman School of Medicine; [email protected]; George
Papandonatos, PhD, Brown University; [email protected]

Symposium 9B

EARLY INTERVENTION AS A TREATMENT APPROACH FOR ADDRESSING VARIABILITY WITHIN

BEHAVIORAL WEIGHT LOSS PROGRAMS
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S401

Jessica Unick, PhD

A large degree of individual variability in response to behavioral weight loss (WL) programs is
consistently observed. Intensive lifestyle interventions have been shown to elicit a 10% WL in
approximately one-third of participants, while an additional one-third of participants fail to
achieve a clinically significant WL (i.e., <5% WL). Thus it is important to understand factors
which contribute to this variability and develop tailored interventions targeting those with a
poor likelihood of success. This presentation examines the variability in response to both in-
person and Internet-based WL interventions and discusses early non-response (i.e., 4-week
WL) as a potential target for improving overall treatment outcomes and increasing the
proportion of individuals achieving a clinically significant WL. Preliminary data will be
presented from a randomized trial comparing WL and treatment adherence among those
receiving a 12-week Internet-based WL program (IBWL; n=50) to those receiving the same
Internet program plus the possibility of extra support (IBWL+ES; n=50). The only difference
between the two treatment conditions was that participants in IBWL+ES with a 4-week WL
d=0.84). Further, those early non-responders receiving additional support viewed
significantly more video lessons (p=0.02, Cohens d=0.91) and had better self-monitoring
adherence (p=0.06, Cohens d=0.87) compared to IBWL early non-responders. These
preliminary data suggest that individuals at risk for poor WL success following treatment may
be able to be rescued if intervened upon early. Additional studies are needed to determine
whether this improved adherence and altered WL trajectory observed with early intervention
leads to a greater proportion of individuals achieving clinically significant WL at longer follow-
up periods. Investigations into whether this stepped-care approach is also beneficial from a
cost-effectiveness perspective are warranted.

Symposium 9C

SEQUENTIAL MULTIPLE ASSIGNMENT RANDOMIZED TRIAL (SMART) ADAPTIVE STUDIES FOR

SUBSTANCE USE DISORDERS (SUD)

Dr. Michelle Drapkin, PhD

Individuals with substance use disorders (SUD) have limited options for and mixed
experiences with treatment. High dropout rates are common. In these studies we examined
several research questions using a sequential multiple assignment randomized trial (SMART):
Does offering patients who do not engage in treatment a choice of other interventions
improve outcomes? Does offering patients who engage but then drop out a choice of other
interventions improve outcomes? Does a second attempt to offer treatment choice to non-
engagers improve outcomes? A SMART design enabled us to examine an alternative option
with greatest chance of success for patients who do not engage while at the same time
recognizing that there will still be some patients who remain disengaged after. Sequential
S402 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

randomization is needed to address these goals. In this presentation we will review the study
design and findings and discuss the unique aspects of two SMART trials. Participants recruited
from intensive outpatient (IOPs) in publicly funded and VA programs. One study recruited
individuals who were cocaine dependent (N=300), 80% with past or current alcohol
dependence; the other recruited individuals who were alcohol dependent (N=200), 40% with
past or current cocaine dependence. Participants were randomized to two engagement
strategy interventions, both using Motivational Interviewing (MI) delivered over the phone:
one used MI to help participants re-engaged in the IOP; the other used MI to provided choices
for treatment (IOP, medication, individual cognitive behavioral therapy (CBT), or telephone
monitoring and adaptive counseling). By the second week of the treatment, 37-38% of the
samples were no longer engaged; before the eighth week, an additional 20-23% had
disengaged. Our findings indicated that there was not an advantage to providing outreach and
a choice of interventions to patients who engage initially but then drop out. Providing further
outreach with a choice of interventions to those not engaged at two and eight weeks did not
improve SUD outcomes compared to no further outreach. We will discuss the main effect
analyses and implications in this presentation.

CORRESPONDING AUTHOR: James McKay, Ph.D., University of Pennsylvania;

[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S403

Symposium 10 8:00 AM-9:15 AM

SLEEP INTERVENTIONS IN CHRONIC ILLNESS POPULATIONS

Kelly Glazer. Baron, PhD, MPH, CBSM1, Michelle Drerup, PsyD2, Christina McCrae, PhD3,
Patricia Haynes, PhD4

1
Rush University Medical Center, Chicago, IL; 2Cleveland Clinic, Cleveland, OH; 3University of
Missouri-Columbia, Columbia, MO; 4University of Arizona, Tucson, AZ

There is a large literature demonstrating the relationship between sleep disruption and a
variety of health conditions including neurologic health, chronic pain and cardiometabolic
disorders. Furthermore, there are several well established and highly effective behavioral
sleep interventions for sleep disorders which are known to improve sleep and quality among
patients with chronic illness. However, less is known about the application of sleep
interventions to prevent and treat medical disorders and mechanisms driving those effects.
The focus of this symposium will be to identify sleep interventions tested in three chronic
illness populations (neurologic disorders, chronic pain and cardiometabolic disorders), discuss
how interventions such as cognitive behavioral therapy for insomnia (CBT-I) have or have not
been adapted for use among patients with chronic illness and discuss how sleep interventions
have affected clinical outcomes or disease process. The first presentation will discuss
applications of CBT-I to MS and Parkinsons Disease (Dr. Drerup). The second presentation will
discuss CBT-I in chronic pain populations (Dr. McCrae). The third presentation will discuss the
role of improving sleep duration and quality in patients with cardiometabolic disorders. Dr.
Haynes will serve as the discussant and integrate the presentations, discussing current
challenges and opportunities for future research in sleep interventions among chronic illness
populations.

Symposium 10A

SLEEP INTERVENTIONS IN CARDIOMETABOLIC DISORDERS

Dr. Kelly Glazer. Baron, PhD, MPH, CBSM

Short sleep duration and poor sleep quality have been associatiated with increased risk for
cardiovascular disease, hypertension and diabetes. However, few studies have evaluated
sleep interventions to reduce risk for cardiometabolic disorders. Sleep may represent a
modifiable risk factor for these conditions. Several proposed pathways have been explored
including indirect pathways (role of sleep in influencing diet and physical activity) and direct
pathways (e.g. sympathetic nervous systems activity). The presention will demonstrate the
S404 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

association of sleep duration and timing with diet and physical activity. In addition, a novel
intervention to extend sleep duration among individuals with short sleep duration and
prehypertension will be presented.

Symposium 10B

TREATMENT OF INSOMNIA AMONG PATIENTS WITH NEUROLOGIC DISORDERS

Dr. Michelle Drerup, PsyD

The incidence of sleep problems in individuals with neurological disorders, including multiple
sclerosis and Parkinsons disease, has been estimated to be 3-5 times higher as compared to
the general population. Sleep disorders such as insomnia in individuals with neurological
conditions are often underdiagnosed and undertreated, oftentimes misunderstood as other
symptoms of the disorder such as fatigue. Initial research suggests that Cognitive Behavioral
Therapy for Insomnia (CBT-I) may be an effective clinical intervention for individuals with
neurological disorders but minimal research to date has been done looking at this behavioral
intervention in this specific population. Recent data will be reviewed including a
retrospective case analysis of individuals who underwent CBT-I (individual or group modality)
who demonstrated improvements in insomnia symptoms, increased self reported total sleep
time, as well as decreased fatigue and depression ratings. In addition, pilot data exploring the
feasibility and outcomes of using a web-based CBT-I intervention in individuals with
Parkinsons disease comorbid with insomnia will be reviewed.

Symposium 10C

SLEEP INTERVENTIONS IN PATIENTS WITH FIBROMYALGIA

Christina McCrae, PhD

Fibromyalgia (FM) is a prevalent and costly chronic pain condition. FM sufferers experience
high rates of insomnia and fatigue. Targeting and improving pain, insomnia, and related
symptoms (I.e., mood, fatigue) in these patients can be challenging. Cognitive behavioral
therapy (CBT) holds great promise, because it can be designed to target multiple symptoms.
Recent trail data comparing the impact of CBT for insomnia (CBT-I) to CBT for pain (CBT-P)
on sleep, pain, fatigue, and mood in patients with comorbid fibromyalgia and chronic
insomnia will be reviewed. Pilot data examining the impact of a hybrid CBT-PI on those
same patient centered outcomes will also be presented.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S405

Symposium 11 8:00 AM-9:15 AM

MHEALTH TECHNOLOGY TO IMPROVE MEDICATION ADHERENCE

Janet DeMoor, PhD, MPH1, Michael J. Stirratt, PhD2, Karen B. Farris, PhD3, Frank A. Treiber,
PhD4, William Riley, PhD5, Sara Browne, MD, MPH6

1
National Cancer Institute, National Institutes of Health, Rockville, MD; 2NIMH Division of AIDS
Research, Bethesda, MD; 3University of Michigan College of Pharmacy, Ann Arbor, MI;
4
Medical Univ of South Carolina, charleston, SC; 5National Institutes of Health, Bethesda, MD;
6
Division of Infectious Diseases, University of California, San Diego, San Diego, CA

Recent years have seen exponential growth in the use of mobile health (mHealth)
technologies to address medication adherence. Various examples include electronic
monitoring of prescription refills, wireless electronic drug monitors, pills that wirelessly
detect/report drug ingestion, and cellphone text messaging (SMS) and smartphone/tablet-
based programs to promote medication adherence. These technologies provide novel
capacities for active monitoring of adherence behavior, improved delivery of medication
adherence reminders, patient-centered education on medication and requirements, provision
of feedback to and from patients and providers, and individually-tailored adherence and
monitoring interventions. The symposium will feature three examples of mHealth approaches
for supporting medication adherence and explore how these approaches contribute to
advancing the science of medication adherence. The first presentation will discuss how
ingestible sensors can be used to monitor adherence to diabetes treatment. The second
presentation will explore how artificial intelligence and text messages can improve medication
adherence among adults with uncontrolled hypertension. The third presentation will examine
how mHealth technology can improve medication management among individuals with
chronic illness. Following the three presentations, a discussant will review the major
outstanding knowledge gaps that need to be addressed as well as the barriers and drivers to
incorporating mHealth tools into clinical practice.

Symposium 11A

MHEALTH TECHNOLOGY TO IMPROVE MEDICATION ADHERENCE

Professor Karen B. Farris, PhD

S406 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Mobile health (mHealth) strategies including SMS may improve medication adherence, yet
most are based on deterministic protocols with programmed responses. We propose a model
medication adherence system using Reinforcement Learning that can automatically adapt
SMS communication based on whats working for each individual. In Phase I, we conducted
a single group study among 19 participants with hypertension. Electronic pill bottle openings
served as the outcome that the reinforcement learning agent sought to optimize. Five
categories of messages were developed and up to two messages were sent daily. The
distribution of messages delivered changed over time, indicating that the messages adapted
based on participants experience. The proportion of participants reporting they took their
medications as prescribed on 5+ days in the past week increased from 89.5% at baseline to
94.7% at six months. There were no statistically significant differences in the disease and
medication beliefs from baseline to 6 months, although medication necessity beliefs increased
and medication concern beliefs decreased. In Phase II, 48 participants, with a medication
possession ratio <0.5 and taking at least one anti-hypertensive medication, were enrolled into
a 6-month, randomized, controlled trial. Participants in the intervention group could receive
one of five messages up to once daily. In interim analysis for n=43 with an average of 102 days
of participation, participants had a baseline self-reported Morisky adherence score of 7.02
1.55 in the treatment group and 6.78 1.32 in the control group (p=0.58). We built a
preference for no messages into the system, and the percentage of time that no message was
sent increased from 43% in month one to 65% in month three and to 79% in month six.
Intervention participants were more likely to be adherent at 6 months than controls (94%
versus 80%; p=0.06). Results from two studies suggest that a reinforcement learning SMS
system can adapt the content of SMS messages over time and may improve medication
adherence.

CORRESPONDING AUTHORS: John Piette, PhD, University of Michigan School of Public Health;
[email protected]; Sean Newman, MS, University of Michigan School of Public Health;
[email protected]; Peter Batra, MS, University of Michigan Institute for Social Research;
[email protected]; Teresa Salgado, PhD, University of Michigan College of Pharmacy;
[email protected]; Satinder Singh, PhD, University of Michigan College of Engineering;
[email protected]

Symposium 11B

FACILITATING SUCCESSFUL MHEALTH MEDICATION ADHERENCE: LISTENING TO THE PATIENTS

Frank A. Treiber, PhD

Poor medication adherence (MA) remains a critical problem in management of all chronic
diseases and impacts ~50% of patients. Numerous technology-enabled devices/ systems have
emerged to address the problem. They enable real time capture of use of pill bottles/trays,
actual pill intake, provide various reminder tactics for med intake and often include efforts to
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S407

increase health literacy. Clinical trial data are emerging which suggest some solutions remain
cost prohibitive, do not link MA to biofunctioning relevant to the meds and/or use of
reminder alerts (e.g., visual/audio signals , SMS messages, etc.) dont necessarily result in
robust sustained improvements. Approaches may not have fully capitalized upon expertise of
the intended users and lacked tactics to facilitate sustained motivation to adhere to the
technology enhanced clinical regimens.

This presentation will provide an overview and update of ongoing refinements of a

patient/provider centered, theory guided mHealth medical regimen self-management system.
This iterative design process began with key informant interviews and focus groups with
patients and healthcare providers for prototype development. Usability testing followed
including expert and user based heuristic and cognitive walk through evaluations.
Refinements were made and field based usability tests run. Subsequent large group
survey/demo studies followed to gather patients guidance for final refinement. Three
feasibility trials were run with different patient groups, all with uncontrolled hypertension
(HTN) (i.e., transplant recipients, FQHC HTN patients, post-stroke patients). Findings revealed
high usability and satisfaction, 95-100% MA, large BP reductions, and 65-100% BP control
compared to far lower outcomes in standard care groups. Our most recent examination of
these trials and interim analyses from ongoing research have shown that patients who
maintain MA scores >.90 during first 3 months and are taking < 5 meds desire to revert back
to a standard plastic tray. They have successfully maintained BP control. Subjects with single
morbidity of HTN are often more challenging to establish sustained adherence than those
HTNs with multi-morbidities .Experiences during the trials/ post-trial interviews verified
importance of identifying patients motivational drivers (e.g., family, faith, friends, short /long
term life goals).Linking patients behavioral changes (e.g., increased MA) to these drivers
facilitated regimen adherence . Recent analyses suggest MA is sustained following trial
cessation. We will present how we have refined our system including expansion of tailored
motivational /social reinforcement SMS message library and use of audio recordings, photo
voice, etc. to further enhance relevance of the feedback messages aimed at fostering greater
autonomous motivation for sustained adherence.

CORRESPONDING AUTHORS: John McGillicuddy, MD, Dept. Surgery, MUSC;

[email protected]; Vanessa Diaz, MD, Dept. Family Medicine, MUSC; [email protected];
Jessica Chandler, PhD, TACHL, Col. Nursing, MUSC; [email protected]; John Sieverdes, PhD,
TACHL, Col. Nursing, MUSC; [email protected]; Brenda Brunner-Jackson, MPH, TACHL, Col.
Nursing, MUSC; [email protected]; Sachin Patel, MSc, College of Nursing, MUSC;
[email protected]

Symposium 11C

LET VISUALS TELL THE STORY: MEDICATION ADHERENCE IN PATIENTS WITH TYPE II DIABETES
CAPTURED WITH AN INGESTION SENSOR PLATFORM
S408 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Dr. Sara Browne, MD, MPH

Background: A novel sensing platform, Digital Health Feedback System (DHFS), can detect
ingestion events and physiological measures simultaneously, using an edible sensor, personal
monitor patch, and paired mobile device. The DHFS generates dense data.

Objective: To use modern methods of visual analytics to represent continuous and discrete
DHFS data, plotting data simultaneously to evaluate the potential of the DHFS to capture
longitudinal patterns of medication-taking behavior and self-management.

Methods: Visualizations were generated using time domain methods of data obtained by the
DHFS use in 5 Subjects with type II diabetes over 37-42 days. The DHFS captured at-home
metformin adherence (via sensor co- ingestion) with heart rate, activity, and sleep/rest. A
mobile glucose monitor captured glucose testing and level (mg/dl). Algorithms were
developed to analyze data over: the entire study, daily, and weekly. Following visualization
analysis, correlations between sleep/rest and medication ingestion were calculated.

Results: A total of 197 subject days, and 141,840 data events were analyzed. Individual
continuous patch use varied between 87-98%. On average, the cohort took 78% (SD 12) of
prescribed medication and took 77% (SD 26) within the prescribed 2-hour time window.
Average activity levels per subjects ranged from 4000-12,000 steps per day. The combination
of activity level and heart rate indicated different levels of cardiovascular fitness between
subjects. Visualizations over the entire study captured the longitudinal pattern of missed
doses (the majority took place in the evening), the timing of ingestions in individual subjects,
and the range of medication ingestion timing, which varied from 1.5-2.4 hours (Subject 3) to
11 hours (Subject 2). Individual morning self-management patterns over the study period
were obtained. Visualizations combining multiple data streams over a 24-hour period
captured patterns of broad daily events: when subjects rose in the morning, tested their
blood glucose, took their medications, went to bed, hours of sleep/rest, and level of activity
during the day. Visualizations identified highly consistent and highly erratic daily patterns in
different Subjects. Correlation between sleep /rest and medication ingestion was evaluated.
Subjects 2 and 4 showed correlation between amount of sleep/rest over a 24-hour period and
medication-taking the following day (Subject 2: r = 0.47, P< .02; Subject 4: r = 0.35, P< 0.05).
With Subject 2, sleep/rest disruptions during the night were highly correlated (r = 0.47, P<
0.009) with missing doses the following day.

Conclusions: Visualizations integrating medication ingestion (via sensor co-ingestion) and

physiological data from the DHFS over varying time intervals captured detailed individual
longitudinal patterns of medication adherence and self-management in the natural setting.

CORRESPONDING AUTHORS: Gwen Littlewort, DPhil, Specialists in Global Health;

[email protected]; Yashar Bezhadi, PhD, Proteus Digital Health; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S409

Symposium 12 8:00 AM-9:15 AM

EMERGING PERSPECTIVES ON TREATMENT AND SELF-MANAGEMENT CHALLENGES OF

CHRONIC DISEASE IN EMERGING ADULTHOOD

Amanda S. Phillips, M.S.1, Deborah Wiebe, MPH, PhD2, Susmita Kashikar-Zuck, PhD3, Elizabeth
J. Siembida, Ph.D.4, Cheryl LP. Vigen, PhD5

1
University of North Texas, Denton, TX; 2University of California, Merced, Merced, CA;
3
Cincinnati Children's Hospital Medical Center, Cincinnati, OH; 4University of Connecticut,
Bethesda, MD; 5University of Southern California, Los Angeles, CA

Treatment and self-management considerations for emerging adults managing chronic

disease are a growing priority. Emerging adults experience unique challenges concerning
psychosocial adjustment and behavioral self-management of chronic disease. For example,
millions of emerging adults in the United States make the decision to move away from their
parents home each year. These emerging adults are often lost in the transition from pediatric
to adult care. For some conditions, youth are now surviving into adulthood, creating
challenges for adult care physicians who are not experienced with these pediatric conditions
or with the normative developmental context of emerging adults. Greater understanding of
the patient, illness, and health care provider factors that contribute to illness management is
needed for this understudied emerging adult population. This symposium will examine the
common and unique challenges emerging adults face in working with health care providers
and self-managing three chronic conditions: fibromyalgia, diabetes, and cancer. The chair will
provide a brief overview of the symposium and impact of chronic disease on emerging adults.
The first panelist will present original longitudinal research outlining trajectories of pain and
mood symptoms from adolescence to emerging adulthood among those with fibromyalgia.
Early psychosocial factors affecting adjustment to chronic pain and key risk factors to be
targeted in treatment will be examined. The second panelist will present original research
concerning the association of depressive symptoms and diabetes distress with glycemic
control in low-SES emerging adults with diabetes, and will discuss treatment implications. The
third presenter will discuss the role of the doctor-patient relationships and cognitive
development in cancer patients perceptions of quality of care, and the importance of using
particular clinical skills when treating emerging adult cancer patients. Finally, an integrated
overview of the challenges of chronic illness management in emerging adulthood will be
provided by the discussant.

Symposium 12A
S410 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

TRAJECTORIES OF PAIN AND MOOD SYMPTOMS IN A LONGITUDINAL COHORT OF YOUTH

WITH FIBROMYALGIA AND PREDICTORS OF LATER ADJUSTMENT

Dr. Susmita Kashikar-Zuck, PhD

Juvenile fibromyalgia (JFM) is one of the most poorly understood chronic pain conditions.
Until recently, very little attention was paid to the prognosis of JFM into adulthood, the long-
term impact of JFM on functioning or early indicators of adjustment to this potentially
debilitating pain disorder. In a prospective longitudinal study of adolescents with JFM (N =
116) and healthy controls (N = 43), we conducted a series of detailed assessments of physical
and psychosocial functioning at from adolescence to adulthood (Mean age = 15 years at Time
1 and 23.4 years at Time 4). Of the JFM group, the majority (>85%) continued to experience
pain or other JFM symptoms and nearly half (48%) met full criteria for fibromyalgia in
adulthood. This presentation focuses on results of longitudinal growth models showing 3
distinct trajectories of pain and 2 trajectories of depressive symptoms from adolescence to
adulthood in the JFM group, and the impact of these trajectories on physical functioning over
time. The subgroup of JFM participants with unremitting pain and worsening depressive
symptoms had the poorest outcomes. Prior research has shown that regular participation in
physical activity and using active psychological coping strategies are essential to successful
management of chronic pain. However, no studies have examined risk and resilience factors
that impact the management of pain and disability in the crucial transition from adolescence
to young adulthood. As part of the longitudinal study, we examined family functioning,
trauma history, social support, mood and physical activity participation in youth with JFM.
Several factors such as a controlling family environment, early exposure to traumatic events
and lower perceived social support were significantly related to psychosocial adjustment in
adulthood. Depressive symptoms and smaller social networks were associated with lower
physical activity participation. The presenter will discuss how key findings from this
longitudinal examination of JFM illuminate our understanding of specific risk factors that need
to be targeted for optimal long-term management of pain and disability in fibromyalgia,
especially as adolescents with JFM transition into adulthood.

Symposium 12B

ADOLESCENT AND EMERGING ADULT CANCER PATIENTS: THE IMPORTANCE OF DOCTOR-

PATIENT RELATIONSHIPS AND COGNITIVE DEVELOPMENT IN QUALITY OF CARE

Elizabeth J. Siembida, Ph.D.

Emerging adults are unique from their older and younger counterparts because of their more
sophisticated cognitive skills, such as improved decision making and information processing,
but the necessity of supportive adults in their lives. Emerging adults diagnosed with cancer
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S411

are not immune to this developmental trajectory, and the current study was undertaken to
examine the experience of quality of care in a sample of adolescent and emerging adult
cancer patients (diagnosed between the ages of 10 and 20). This cross-sectional survey
examined participants perception of engagement in care; the doctor-patient relationship;
quality of care; their cognitive development, specifically their decision making and evaluative
thinking skills; and readiness for transition into adulthood. Bivariate and multivariate analyses
found little difference in perceptions of quality of care based on age or transitional readiness.
Instead, increased cognitive development (=.226, t(94)=2.468, p=.015) and experiencing the
opportunity for bi-directional information provision with providers was associated with
increased quality of care (=.345, t(94)=3.324, p=.001). The lack of a significant relationship
between chronological age and quality of care points to the importance of developmental
milestones, rather than age, as an important factor in initiating the transition to adult
healthcare. The role of increased cognitive development highlights the need for a certain level
of decision making and evaluative skills that are only gained as an individual enters late
adolescence and begins the developmental transition into emerging adulthood. The
significant relationship with bi-directional information provision points to the important role
doctors play in the cancer experience for adolescents and emerging adults. In looking for ways
to improve the transition from adolescence to emerging adulthood within the oncology
context, doctors appear to be ideally situated to help provide support and guidance to their
patients as they begin to take on an increased role in their care.

CORRESPONDING AUTHORS: Elizabeth Siembida, PhD, University of Connecticut;

[email protected]; Kerry Moss, MD, Connecticut Children's Medical Center;
[email protected]; Nina Kadan-Lottick, MD, MSPH, Yale University School of
Medicine and Smilow Cancer Hospital at Yale-New Haven; [email protected]; Keith
Bellizzi, PhD, MPH, University of Connecticut; [email protected]

Symposium 12C

DEPRESSIVE SYMPTOMS, DIABETES DISTRESS AND GLYCEMIC CONTROL IN LOW-SES

EMERGING ADULTS WITH DIABETES

Cheryl LP. Vigen, PhD

Emerging adults (EAs) with diabetes experience high rates of elevated blood glucose (as
indexed by hemoglobin A1c), depressive symptoms (DS), and diabetes distress (DD), all of
which contribute to increased risk of diabetes complications, disability and mortality. We
examined relationships between DS, DD, and A1c among a cohort of low-SES, largely minority
EAs with diabetes, using baseline data from an occupational therapy diabetes management
intervention randomized controlled trial.
S412 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Participants were 81 EAs (75% type 1 diabetes; mean age 22.6 years; mean A1c 10.8%). We
assessed DS via the Patient Health Questionnaire (PHQ-8) and DD via the Problem Areas in
Diabetes-Short Form (PAID-5). Spearman correlations and p-values are presented; differences
in medians are evaluated with Wilcoxon rank sum tests.

Women had higher DS than men (median PHQ 6.0 vs 4.0, p=0.02); no other demographic
variables were associated with DS scores. In univariate analysis, DD was not significantly
associated with A1c (r=0.19, p=0.09). In contrast, DS was associated with increased A1c
(r=0.27, p=0.02). This relationship remained after adjusting for potential confounders and for
blood glucose monitoring and medication adherence, both strong behavioral predictors of
A1c.

EAs with a depression diagnosis in their medical records (n=14) did not have higher DS than
those who did not (median PHQ 6.0 vs 5.0, p=0.10); the association between A1c and DS
remained significant when restricted to those without a diagnosis. Among EAs with PHQ
scores below the cut-off for major depression (PHQ <10), there was no association between
DS and A1c (r=0.06, p=0.67), but for those with scores 10, the association was substantial
(r=0.47, p=0.03).

Overall, in a cohort of low-SES, largely minority EAs, DS was more closely associated with A1c
than was DD; furthermore, the relationship between DS and A1c was not explained by blood
glucose monitoring or medication adherence. The PHQ cut-off score for major depression may
be a threshold above which DS affects A1c. Given the high proportion of EAs with both DS and
elevated A1c, screening and providing psychosocial intervention for DS may be effective in
improving both physical and mental well-being in this population.

CORRESPONDING AUTHORS: Cheryl Vigen, PhD, University of Southern California;

[email protected]; Kristine Carandang, University of Southern California;
[email protected]; Jesus Diaz, PhD, University of Southern California;
[email protected]; Jeanine Blanchard, PhD, University of Southern California;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S413

Symposium 13 8:00 AM-9:15 AM

PROMOTING PHYSICAL ACTIVITY INTERVENTIONS AMONG PEOPLE WITH MENTAL DISORDERS:

A CONTINUUM

Ahmed Jerome Romain, PhD1, Felipe Barreto Schuch, PhD2, Simon Rosenbaum, PhD3, Joseph
Firth, BSc4, Philip B. Ward, BMedSc, PhD5

1
University of Montreal Hospital Research Centre, Montreal, PQ, Canada; 2Centro
universitrio la salle (unilasalle) rio grande do sul, Canoas, Rio Grande do Sul, Brazil; 3UNSW,
Sydney, New South Wales, Australia; 4University of Manchester, Manchester, England, United
Kingdom; 5UNSW Australia, Liverpool BC, New South Wales, Australia

Within the last years, there was a growing interest towards the benefits of physical activity
among people with mental disorders. Even though recent studies highlighted the positive
impact of physical activity, more work is needed to facilitate its initiation as to develop more
efficacious interventions to sustain it over time.

As such, the present international symposium (Australia, Brazil, England, and Canada) aims to
propose several presentations on the need for lifestyle interventions in people with mental
disorders at various stages and further propose new interventions to improve their physical
and mental health over the long term.

In a first presentation, we will show that the co-occurrence of obesity and mood disorders
was associated with poor psychological well-being, more health comorbidities, higher use of
health services and less adequate health behaviours. In a second presentation, we will present
results from a trial assessing whether physical activity is feasible in first-episode psychosis
individuals and if the health benefits could be maintained overtime. The other speaker will
talk about a project with 3 phases, developed in coordination with the national insurer for
workplace injury, on the implementation of physical activity in first-responders people
developing posttraumatic stress disorder. Then, the next speaker will discuss the results from
a 6-month RCT intervention, conducted among people with major depressive disorder, which
aimed to promote physical activity using wearable technology and a theoretical model of
motivation. Finally, given the necessity to have a reliable measure of physical activity, the last
talk will be on the first results from the international and cross-cultural validation of the
simple physical activity questionnaire in people experiencing mental illness.

All bring together, these presentations will cover several aspects of the lifestyle intervention
prism going from the prevention, through methodological and interventions studies with
S414 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

follow-up. Moreover, the symposium, as the presenters will have the objective to highlight
the necessity to develop more appropriate lifestyle interventions in people with mental
disorders to improve their functioning and quality of life.

Symposium 13A

IMPACTS OF OBESITY AND MOOD DISORDERS ON PSYCHOLOGICAL WELL-BEING,

COMORBIDITIES, HEALTH BEHAVIOURS AND USE OF SERVICES.

Dr. Ahmed Jerome Romain, PhD

Introduction: Obesity is highly prevalent in North-America people with mood disorders (e.g.,
bipolar disorder, major depressive disorders) are more likely to be affected. Even though
obesity and mood disorders frequently co-occur, few studies examined if the presence of a
mood disorder can be an aggravating factor of obesity-related conditions.

Objective: to analyze whether people with obesity and mood disorders (POMD) differ from
those without mood disorder (Ob) in terms of psychological well-being, comorbidities, health
behaviours and use of health services.

Methods: Cross-sectional study based on the Canadian Community Health Survey (CCHS)
conducted in 2007-2008. Sample of people with obesity (n = 1298) was weighted to represent
the inhabitants from the province of Quebec (Canada). The CCHS used diagnostic criteria
outlined in the DSM-IV-TR to screen respondents.

Results: POMD (n = 99) were more likely to be women than Ob (n = 1199), (65%, 95%CI[56.29-
75.05%] vs 45%, 95%CI[42.92-48.56%]). In terms of psychological well-being, POMD were
more likely to report poor mental health (40%, 95%CI[31.81-51.32%] vs 3.17%, 95%CI[2.32-
4.34%]) with more psychological distress (37.5%, 95%CI[28.96-48.55%] vs 4.25%, 95CI [3.23-
5.59%]) and anxiety disorders (39.79%, 95%CI[31.19-50.76%] vs 4.01%, 95%CI[3.04-5.28%]).

Regarding comorbidities, higher percentages of POMD reported arthritis (32.65%,

95%CI[24.57-43.39%] vs 16.58%, 95%CI [14.60-18.83%], low back pain (41%, 95%CI[32.41-
51.87%] vs 19.48%, 95%CI[17.36-21.86%]). No significant differences were found in migraine,
hypertension and type 2 diabetes.

In terms of use of services, compared to Ob, POMD had more frequents consultations to
doctors (86.87%, 95%CI[80.46-93.78%] vs 74.21%, 95%CI[71.77-76.73%]), nurses (27.27%,
95%CI[19.77-37.62%] vs 15.44%, 95%CI [13.52-17.63]) and psychologists (21.21%,
95%CI[14.51-31.0%] vs 3.08%, 95%CI[2.25-4.24]).

Regarding their health behaviours, higher percentages of POMD reported to be inactive

(79.59%, 95%CI[71.99-87.98%] vs 66.67%, 95%CI[64.05-69.39%]), or to be limited in their
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S415

physical activity (37.37%, 95%CI[28.96-48.23%] vs 13.90%, 95%CI[12.07-16.01%]). No

differences were found on tobacco and fruits/vegetables consumption.

Discussion: The presence of a mood disorder seems to be an aggravating factor in obesity

because it was associated with poor health in several areas. There is a need to more consider
obesity as health behaviours among POMD.

CORRESPONDING AUTHORS: Jacques Marleau, Direction of public health, Integrated Centre

for Health and Social Services in Outaouais; [email protected]; Aurelie Baillot,
professor, University of Quebec at Outaouais; [email protected]

Symposium 13B

PHYSICAL ACTIVITY PROMOTION IN PEOPLE WITH MAJOR DEPRESSIVE DISORDER USING

WEARABLE DEVICES: A CLINICAL TRIAL

Felipe Barreto Schuch, PhD

People with major depressive disorder (MDD) have decreased levels of physical activity(PA)
when compared to the population levels. Also, there is evidence that this relationship is
bidirectional, and that increasing PA reflects in the alleviation of depressive symptoms in
people with depression. Motivational and other cognitive strategies have been used to help
people with MDD to achieve the recommended PA levels, however, the results suggest the
need for additional strategies. The present study aims to increase PA levels of outpatients
with MDD (n=120). It is a randomized clinical trial with 6 months of intervention, single
blinded for the main outcome (depressive symptoms assessed by the clinician). Participants
are being randomized into two groups: an intervention group, that wore a wearable
pedometer, with daily goals of steps to achieve and received a contact using self-
determination theory elements. The aims of daily steps to be achieved increase every month
until reach the public health recommendations of 10.000 steps per day; and a control group
that received general recommendations of PA promotion. Both groups were under usual
treatments for depression (pharmacotherapy). Patients were contacted once per week by
telephone or face-to-face contacts. Analysis on the effects of the intervention on PA levels,
depressive symptoms (both self and clinician assessed), and quality of life will be evaluated.
Also, the study plans to evaluate potential factors that predict success/failure.

CORRESPONDING AUTHORS: Andrea Dunnal, Klein Blundel, Denver, US; [email protected];

Anny Dias, Hospital de clnicas de Porto Alegre, Porto Alegre, BR; [email protected];
Bruno Mosqueiro, University of Rio Grande do Sul; [email protected];
Fernanda Monteiro, Hospital de clnicas de Porto Alegre, Porto Alegre, BR;
[email protected]; Davy Vancampfort, University of Leuven, Department of
S416 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Rehabilitation Sciences, Leuven, Belgium;; [email protected]; Gabriella Torelly,

Hospital de clnicas de Porto Alegre, Porto Alegre, BR; [email protected]; Graziele Alves
Dias, Hospital de clnicas de Porto Alegre, Porto Alegre, BR; [email protected]; Marco
Antonio Caldieraro, University of Rio Grande Do Sul; [email protected]; Matheus Frizzo,
Hospital de clnicas de Porto Alegre, Porto Alegre, BR; [email protected]; Marcelo Pio de
Almeida Fleck, University of Rio Grande Do Sul; [email protected]

Symposium 13C

IMPLEMENTING PHYSICAL ACTIVITY IN PTSD: APPLYING EVIDENCE-BASED PROGRAMS FOR

WORKPLACE-RELATED PSYCHOLOGICAL INJURIES

Dr. Simon Rosenbaum, PhD

Emergency service workers and first-responders (police, fire, ambulance officers) are at
significantly higher risk of poor mental health, including developing posttraumatic stress
disorder (PTSD) and/or increased risk of suicide, in response to occupational exposure to
potentially traumatising experiences. PTSD is also associated with high rates of preventable
cardiometabolic disease including obesity and metabolic syndrome, with symptoms of
hyperarousal associated with lower physical activity participation. Despite increased
community and organisational recognition of the risks associated with poor mental health
among veteran populations, translation of targeted programs to address the specific physical
and mental health needs of first responders is in its infancy. This study aims to build on
findings from an RCT of exercise augmentation for inpatients with PTSD in which exercise
reduced both PTSD and depressive symptoms at a clinically significant level and improved
body composition, compared to treatment as usual. The goal of the present study was to
translate this exercise program for community-based first responders identifying as suffering
from psychologically injury. To facilitate translation of the intervention, a multi-phase
evaluation process was conducted. Phase 1 involved conducting focus groups with current
emergency service crews (n>50 participants; fire and rescue) to document their views
regarding mental health help-seeking behaviour, physical activity and opportunities for
promoting access to exercise interventions. Themes identified included the acceptability of
exercise as a coping strategy for first responders, and highlighted the need for interventions
aimed at reducing stigma towards mental health issues among first-responders. Phase 2
involved longitudinal monitoring of the physical and mental health of first responders using a
mobile phone application (n=17). A significant proportion of respondents (41%: 7/17) failed to
achieve basic physical activity guidelines (150 minutes of moderate to vigorous physical
activity per week). Results are being used to inform Phase 3, in which we will design and
implement an exercise program for first responders with psychological injury, in partnership
with the national insurer for workplace injury.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S417

CORRESPONDING AUTHORS: Philip Ward, University of New South Wales;

[email protected]; Andrea Fogarty, Black Dog Institute; [email protected]; David
Berle, University of New South Wales; [email protected].; Katherine Boydell, University of
New South Wales; [email protected]; Julia Lappin, University of New South Wales;
[email protected]; Zachary Steel, University of New South Wales; [email protected]

Symposium 13D

INVESTIGATING THE FEASIBILITY AND BENEFITS OF EXERCISE IN EARLY PSYCHOSIS

Mr. Joseph Firth, BSc

Exercise has previously been shown to reduce symptoms of schizophrenia in long-term

patients, along with improving their physical health and cognitive functioning. However, the
effects of exercise in the early stages of psychosis have not been widely investigated. This
study aimed to examine the feasibility of using individualized exercise training to improve
physical and mental health outcomes in people with first-episode psychosis (FEP), and to
explore if the benefits of exercise can be sustained overtime. Twenty-eight people with FEP
participated in 10-weeks of twice-weekly supervised exercise, using activities tailored to their
own choice. Participants were assessed at baseline, 10-weeks, and then six-months after the
supervised intervention, and compared to a group of patients with FEP who did not receive
exercise training. During the 10 weeks of supervised exercise, participants achieved 107
minutes of moderate-to-vigorous exercise per week. Post-intervention assessments found
significant improvements in psychiatric symptoms, verbal memory and social functioning.
After 6 months, 55% of participants had continued their exercise. Only those who had
maintained regular exercise after 6 months continued to show reductions in psychiatric
symptoms, whereas the symptoms of those who had ceased exercising had returned to
baseline scores. Previously-observed benefits of exercise for social functioning were also
maintained at the follow-up, although cognitive improvements were lost. Future research
should establish the effectiveness of exercise for early psychosis in randomized controlled
trials, along with exploring optimal methods for implementing exercise interventions into
clinical care, in order to improve symptomatic and functional recovery from the earliest stages
of illness.

CORRESPONDING AUTHORS: Rebekah Carney, University of Manchester;

[email protected]; Paul French, University of Liverpool;
[email protected]; Rebecca Elliot, University of Manchester;
[email protected]; Alison Yung, University of Manchester;
[email protected]

Symposium 13E
S418 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

EVALUATING PHYSICAL ACTIVITY INTERVENTIONS : DEVELOPMENT & VALIDATION OF THE

SIMPLE PHYSICAL ACTIVITY QUESTIONNAIRE (SIMPAQ)

Prof. Philip B. Ward, BMedSc, PhD

Physical activity is consistently associated with improved physical and mental health among
people experiencing mental illness. Given increasing interest physical activity as a component
of treatment, ensuring that valid, clinically useful and feasible measures of physical activity
are available for routine use is of paramount importance. Existing self-report measures fail to
address unique considerations and limitations faced by people experiencing mental illness.
The SIMPAQ is a 5item clinical tool taking between five and 10 minutes to complete,
designed to assess physical activity among populations at high risk of sedentary behaviour,
such as those experiencing mental illness. This study aimed to determine the reliability and
validity of the SIMPAQ among mental health service users against an objective measure of
physical activity (accelerometer). We hypothesise that the SIMPAQ will demonstrate
acceptable reliability and criterion validity (i.e. correlation > 0.3 or above for total physical
activity) against the objective measure.The SIMPAQ is currently being validated across >25
sites in >18 countries. Eligible participants are adults aged 18-65 years, currently treated at
participating treatment facilities, who meet DSM or ICD criteria for any mental disorder
(excluding eating disorders). Demographic and anthropometric information, psychiatric
diagnoses, somatic co-morbidities, medication usage, symptom severity (DSM-5 Cross-Cutting
tool) and cognitive capacity (Montreal Cognitive Assessment) are assessed. Participants
complete the SIMPAQ through a structured interview. Participants are given a tri-axial
accelerometer to wear during waking hours for seven days. This study aims to recruit
approximately 1000 participants across a variety of psychiatric treatment facilities.As of
September 2016, validation data has been collected from n=95 subjects in six countries
(Australia, Brazil, Spain, USA, India, Switzerland).Preliminary analysis demonstrates that the
SIMPAQ is clinically feasible and acceptable across a range of languages and clinical settings.

CORRESPONDING AUTHOR: Simon Rosenbaum, PhD, UNSW Australia;

[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S419

Symposium 14 8:00 AM-9:15 AM

METHODOLOGICAL CHALLENGES IN STUDIES WITH STIGMATIZED POPULATIONS OF WOMEN

Karen F. Stein, PhD1, Lisette Jacobson, PhD, MPA, MA2, Leigh Ann Simmons, Ph.D., M.F.T.3,
Laura L. Hayman, PhD, RN, FAAN, FAHA4

1
University of Rochester, School of Nursing, Rochester, NY; 2University of Kansas School of
Medicine-Wichita, Wichita, KS; 3Duke University School of Nursing, Durham, NC; 4University of
Massachusetts Boston, Boston, MA

Stigma is broadly recognized as a fundamental cause of inequities in health and health care.
While both men and women experience stigma, data show that women are frequently the
gatekeepers of their families health. Thus, women in groups that are stigmatized due to
racial/ethnic identity, immigrant status, geographic residence, and overweight and obesity
experience additional educational, economic, and social barriers that negatively affect not
only their own health outcomes but also those of their children and families. Moreover,
women in these stigmatized groups experience disproportionate rates of physical and mental
health disorders, yet remain among those least likely to seek health care services. However,
studies designed to address the health care needs of these underserved populations of
women encounter unique challenges in all aspects of the research process. These challenges
are far reaching and include accessing the population, recruitment, retention, valid and
reliable measurement tools, and culturally sensitive intervention design and implementation.
The purpose of this symposium is to describe methodological challenges associated with
conducting research with stigmatized populations of women and to provide practical solutions
to overcome these common barriers to research. Karen Stein will discuss two projects focused
on body weight perceptions, eating and weight control behaviors in undocumented, low
acculturated women of Mexican origin living in rural farming communities in Western New
York. Lisette Jacobson will discuss the process of completing two pilot studies to understand
the needs of an obstetrical population located in an extremely remote rural region of a
Midwestern state prior to designing a health promotion intervention. Leigh Ann Simmons will
discuss two interventions with women who experience weight stigma associated with
entering pregnancy overweight or obese or gaining excess weight during pregnancy.
Discussant Laura Hayman will provide a critical analysis and synthesis of presentations with
emphasis on directions for future research with stigmatized populations.

Learning objective 1: Identify and describe methodological challenges encountered in health

research focused on stigmatized population of women.
S420 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Learning objective 2: Describe strategies to improve recruitment, retention, construct

measurement, and other design challenges encountered in health research with stigmatized
populations of women.

Learning objective 3: Describe initial steps toward designing culturally sensitive interventions
that consider the intersection of stigma with other factors that affect health outcomes.

Symposium 14A

LIVING IN THE SHADOW: UNDOCUMENTED MEXICAN AMERICAN WOMEN AND RESEARCH

CHALLENGES.

Karen F. Stein, PhD

Objective: Undocumented Mexican women living in the U.S. are a stigmatized group who
experience significant health disparities and limited access to health care. Particularly in
border states, the threat of deportation and separation from children and family result in high
levels of fear and isolation. Consequently, despite their numerous health disparities and
significant unmet health care needs, undocumented Mexican American (MA) women remain
an understudied population. The primary purpose this paper is to describe the challenges
associated with conducting research with this population and present practical solutions
developed in two studies focused on body weight perceptions, eating and weight control
behaviors in samples of undocumented young adult Mexican American women.

Design and Methods: Two studies using qualitative (focus groups) and quantitative (ecological
momentary assessment) were completed with two samples of non-pregnant Spanish speaking
MA women ages 18-45 years (total N=75) with undocumented legal status, low acculturation,
low health literacy and living in rural farming communities in Western New York.
Methodological challenges common across studies included: (1) accessing the target
population, (2) addressing exceptional needs for anonymity, (3) engaging family gatekeepers,
(4) identifying meaningful participant incentives, (5) avoiding socially desirable responses
shaped by stigma and embarrassment. Practical solutions were developed through
collaboration with community member advocates, institutional review board representatives,
study participants and trial and error.

Conclusion: Research with undocumented populations of MA women requires attention to

their unique stigmatized social position and cultural nuances to ensure validity and reliability
of study results. Adjustments to study protocols that meaningfully address these challenges
are feasible and essential to successful completion of much needed research with this
understudied population of women.

Symposium 14B
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S421

PROPOSING AN INTERVENTION TARGETING UNDERSERVED RURAL PREGNANT WOMEN:

LESSONS LEARNED FROM TWO PILOT STUDIES

Dr. Lisette Jacobson, PhD, MPA, MA

Background: In the United States, pregnant women who live in rural locations are more likely
to experience adverse birth outcomes compared to women in urban locations. This is partially
due to limited access, availability, and delivery of obstetrical healthcare services in rural
regions. Serving an underserved population in a Midwestern state, a rural hospital with 300
deliveries per year experienced a high rate of complicated pregnancies and births. The
purpose of this presentation is to describe the methodological challenges associated with
studies involving the pregnant population in this rural area.

Methods: Two pilot studies, quantitative (i.e., obstetrical health assessment) followed by
qualitative (i.e., focus groups), were completed. Results from the health assessment (N=177)
indicated that the majority of pregnant women were Hispanic and of low socio-economic
status, were overweight/obese, did not engage in exercise, and had a family history of
diabetes. Further, focus group findings (N=17) showed that women reported having limited
access to health promotion programs and breastfeeding support.

Results: Common themes across both studies included: (1) buy-in from local stakeholders; (2)
importance of relationship building among local healthcare providers and researchers; (3)
ease of work flow and continuous follow-up with on-site clinic personnel; (4) being mindful of
the no-show factor; and (5) identifying meaningful incentives. Study findings were used in
application for a K-award with the National Institute of Diabetes and Digestive and Kidney
Diseases (NIDDK) testing the feasibility of a health promotion intervention that contains
physical activity, dietary habits, and breastfeeding support.

Conclusion: Proposing an intervention approach that targets underserved rural pregnant

women requires prior in-depth study of this population. Moreover, inclusion of these women,
local key stakeholders, and healthcare providers in discussions surrounding behavioral change
could be beneficial with recruitment, participation, and retention in future interventions.

CORRESPONDING AUTHORS: Tracie Collins, MD, MPH, University of Kansas School of

Medicine; [email protected]; David Grainger, MD, MPH, University of Kansas School of
Medicine; [email protected]

Symposium 14C

BEHAVIORAL WEIGHT MANAGEMENT TRIALS DURING PREGNANCY AND POSTPARTUM

AMONG OVERWEIGHT AND OBESE WOMEN: THE ROLE OF WEIGHT BIAS
S422 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Dr. Leigh Ann Simmons, Ph.D., M.F.T.

Objective: Weight bias contributes to negative healthcare experiences among pregnant and
postpartum women with overweight and obese BMIs. Studies of patients in maternity care
settings show they experience discomfort, guilt, and shame for their weight status, as well as
indifference, judgment, and disrespectful treatment from providers. Underscoring these
findings, research also shows that many obstetric providers hold negative attitudes toward
pregnant women with BMI>25, including less positive attitudes toward caring for these
patients and negative judgments about their self-management behaviors. Given the
pervasiveness of obesity stigma in obstetric care, designing interventions aimed at limiting
pregnancy weight gain or promoting postpartum weight loss for women with prepregnancy
BMI>25 requires consideration of how weight bias may influence their research participation.

Data and Methods: Design considerations and lessons learned regarding intervention
approaches and content were characterized for two studies: (1) a digital intervention to
promote healthy prenatal diet for women who entered pregnancy with a BMI>25; and (2) a
telephonic coaching intervention for women who gained gestational weight in excess of
Institute of Medicine recommendations for prepregnancy BMI.

Results: Interventions that address pregnancy weight gain and postpartum weight loss among
women with BMI>25 who may experience weight bias should consider the following
strategies to promote recruitment, adherence, and retention: (a) focus change behaviors on
meeting healthy dietary goals rather than numbers-based targets, such as weekly pounds
gained prenatally or lost postpartum; (b) use computer-mediated communication, mobile
technologies, or telephonic coaching instead of in-person behavioral support strategies to
protect participants visual anonymity; and (c) develop interventions that address the whole
person and recognize the transition to parenthood as a unique, challenging time versus
interventions that focus solely on weight-related behaviors like diet and exercise.

Conclusions: Among women with overweight and obese BMIs, weight bias and obesity stigma
likely affect research participation in behavioral weight management trials. Study designs that
consider participants holistically, reflect the unique nature of pregnancy and the postpartum
period, and offer women options for remaining visually anonymous to interventionists may
facilitate recruitment, adherence, and retention with this vulnerable group of participants.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S423

Symposium 15 8:00 AM-9:15 AM

PAIN MANAGEMENT: TRANSLATING POLICY AND GUIDELINES INTO PRACTICE

Martin Cheatle, PhD1, Robert Twillman, PhD, PhD2, Lara Dhingra, PhD3, Rollin Mac. Gallagher,
MD, MPH4

1
Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA; 2Academy
of Integrative Pain Management, Lenexa, KS; 3MJHS Institute for Innovation in Palliative Care,
New York, NY; 4University of Pennsylvania, Gladwyne, PA

Pain Management: Translating Policy and Guidelines into Practice

Approximately one-third of the adult American population experience chronic or recurrent

pain and this number increases each year. The annual combined cost of pain care (healthcare
costs, lost productivity and disability) has been estimated as reaching $600 billion which is far
greater than what is expended on other major diseases including heart, diabetes and cancer.
Effective pain management requires a balanced and multimodal approach which includes, in
some cases, the use of opioid analgesics. There has been a growing controversy, however,
over both the efficacy and safety of opioid analgesics in managing patients with chronic pain.
This debate has stimulated a great deal of scholarly activity on both developing strategies and
guidelines to identify and mitigate the risk of prescription opioid misuse, abuse and opioid-
related fatalities but also a broader discussion on National pain policies and developing more
effective pain care models. This multidisciplinary panel will discuss the current pain
management policy activity and proposed National pain strategies; a review and critique of
the recently published CDC guidelines of opioid prescribing in chronic pain and an integrated
pain care model developed in the VHA.

Symposium 15A

CDC GUIDELINE FOR PRESCRIBING OPIOIDS FOR CHRONIC PAIN: TRANSLATING GUIDELINES
INTO CLINICAL PRACTICE

Dr. Martin Cheatle, PhD

Over the past decade there has been a significant increase in the prescribing of opioids for
chronic noncancer pain which has been paralleled by an increase in opioid related deaths and
admissions to treatment facilities for opioid abuse. There has been considerable scholarly
S424 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

activity dedicated to developing risk assessment and mitigation strategies to curb the
burgeoning opioid epidemic while preserving patients right to pain relief including the use of
guidelines. This presentation will discuss the evolution of the opioid prescribing guidelines,
review the latest CDC guideline and examine the challenges in translating the guideline into
real-world clinical practice.

Symposium 15B

NATIONAL PAIN POLICY INITIATIVES

Dr. Robert Twillman, PhD, PhD

This presentation will present an overview of current policymaking activity related to pain
management. While there has been a great deal of activity in this arena for several years, an
increased focus on prescription drug abuse, addiction, and overdose on the part of the media
and citizens in general, coupled with the issuance of the CDC's guideline for the use of opioids
to treat chronic pain in primary care and the National Pain Strategy, has caused a new surge in
policymaking. The challenges of crafting policies that permit the appropriate practice of pain
management while simultaneously addressing problems related to prescription opioids will be
highlighted, and solutions that could address both problems simultaneously and successfully
will be proposed.

Symposium 15C

INTEGRATED STEPPED PAIN CARE: A POPULATION-BASED MODEL ADDRESSING THE PUBLIC

HEALTH PROBLEMS OF CHRONIC PAIN & OPIOID OVER-USE

Dr. Rollin Mac. Gallagher, MD, MPH

Over 50 % of Veterans suffer chronic pain often complicated by PTSD and traumatic brain
injury related to combat exposure and multiple deployments. As policy the VHA adopted a
patient-centered, integrated, team-based, biopsychosocial Stepped Care Model based on the
foundation of an informed Veteran learning self-management skills with integrated primary
care teams supported by behavioral and integrative health and interdisciplinary pain specialty
programs. This presentation describes implementation challenges, such as training deficits in
pain management and safe opioid prescribing and the programs designed to address these
deficits, such as behavioral and integrative health in primary care, the Pain Mini-Residency,
and academic detailing in the Opioid Safety Initiative and SCAN-ECHO.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S425

Symposium 16 8:00 AM-9:15 AM

THINK OF ME: SUBSTANCE USE IDENTITIES IN SUBSTANCE USE

Andrew Hertel, PhD1, Kristen P. Lindgren, Ph.D., ABPP2, Alexander W. Sokolovsky, MA3,
Andrew Hertel, PhD1, Robin Mermelstein, PhD4

1
Knox College, Galesburg, IL; 2University of Washington, Seattle, WA; 3Brown University,
Chicago, IL; 4University of Illinois at Chicago, Chicago, IL

Basic identity research demonstrates that our identities influence our behaviors and vice
versa. Extensions of this research into the domain of substance use provide a more
comprehensive view of substance use, as identity encompasses a set of correlates that are
distinct from other known cognitive, affective, motivational, and neurobiological correlates. In
particular, an emerging body of research demonstrates that substance use identities, which
reflect that a behavior is no longer exclusively perceived as simply something one does but
also as something that defines who one is (e.g., I smoke vs. I am a smoker), can emerge
from and contribute to substance use, from uptake through cessation. In this symposium, we
present research on substance use identities, both explicit and implicit, from cross-sectional
and longitudinal observational investigations in three different behavioral domains (alcohol
consumption, electronic nicotine delivery system (ENDS) use, and cigarette smoking). The first
presenter will show a within-subject reciprocal relationship between alcohol consumption and
implicit drinking identity, using data from a longitudinal, repeated measures study among
college students. Using data from studies aimed at development of an ENDS use continuation
motives measure for young adult ENDS users, the second presenter will show a unique ENDS
user identity factor and its relationships with ENDS continuation. The third presenter will
show that implicit and explicit drinker and smoker identities are associated with craving for
alcohol and cigarettes as well as attentional bias towards alcohol and cigarette cues, using
data from observational studies among college student moderate drinkers and smokers as
well as adult heavy smokers. Combined, the presentations inform understanding of the
relationships between substance use identities and behaviors and indicate the value of
including identity in explanations of substance use. The discussant will integrate the findings,
discuss how they contribute to our understanding of substance use and inform intervention
efforts, and suggest future avenues of inquiry.

Symposium 16A

WITHIN-PERSON CHANGE IN IMPLICIT DRINKING IDENTITY: CHANGES IN DRINKING IDENTITY

PREDICT CHANGES IN DRINKING AND VICE-VERSA
S426 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Dr. Kristen P. Lindgren, Ph.D., ABPP

There has been a surge of interest in evaluating the role of implicit (fast/reflexive/impulsive)
cognitions in substance use. In the domain of alcohol use, attention has been increasingly
devoted to evaluating implicit cognitions about alcohols association with the self (i.e., implicit
drinking identity or associations with drinking and me vs. not me, Lindgren et al., 2013a).
There is evidence that implicit drinking identity is a robust predictor of drinking cross-
sectionally (Lindgren et al., 2013a, 2013b), across a wide age range (Lindgren et al., 2016a),
and over time (Lindgren et al., 2016b). A critical next step toward establishing whether
implicit drinking identity may represent a mechanism of behavior change and/or intervention
target is to evaluate whether changes in implicit drinking identity are associated with changes
in drinking behavior and vice-versa. Thus, the goal of the current study was to investigate
whether changes in implicit drinking identity were associated with subsequent changes in
drinking. Participants included 506 undergraduates (57% women), who completed
assessments of their drinking behaviors and implicit drinking identity at 3-month intervals for
a total of 8 assessments. Due to extreme non-normality in drinking behavior (50% of
participants reported no drinking at Time 1), participants drinking behaviors were converted
to categories based on NIAAAs criteria: non-drinkers, low risk drinkers (men: no more than
14 drinks/week; women: no more than 7 drinks/week), and high risk (men: > 14 drinks/week;
women: >7 drinks/week). Models, thus, evaluated whether within-person changes in implicit
drinking identity were associated with changes in (i.e., increasingly risky) drinker categories
and vice-versa. Cross lagged panel models indicated that changes in implicit drinking identity
were positively associated with changes in drinker categories (p < .001). Further, changes in
drinker category were also positively associated with changes in implicit drinking identity (p <
.001). Results will be discussed in terms of implications for theory and intervention.

CORRESPONDING AUTHORS: Cecilia Olin, University of Washington; [email protected]; Scott

Baldwin, PhD, Brigham Young University; [email protected]; Clayton Neighbors, PhD,
University of Houston; [email protected]

Symposium 16B

MOTIVES FOR CONTINUATION AMONG YOUNG ADULT USERS OF ELECTRONIC NICOTINE

DELIVERY SYSTEMS: THE ROLE OF IDENTITY

Mr. Alexander W. Sokolovsky, MA

Substance users, including users of electronic nicotine delivery systems (ENDS) can develop
identities associated with their use behavior that then contribute to maintenance and
escalation of use. Substance use identities thus represent one of many possible social,
emotional, environmental, and behavioral motives that are associated with continued
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S427

substance use. Multifactorial measures developed to index cigarette smoking continuation

motives often query behaviors unique to cigarette smoking that may not represent
experiences of ENDS users. We used a mixed-methods approach to develop a novel measure
of motives for ENDS use continuation among young adults (18-24). In a qualitative study,
young adults (N = 15; 8 male; 8 non-Hispanic white; 2 African American; 2 Hispanic; 3 Asian)
who reported past-week ENDS use and self-identified as regular ENDS users were invited to
complete intensive longitudinal interviews in a private clinical setting. Emergent themes
related to motives for continuation included ENDS user identity, as well as snacking, sensory
experience, affiliative attachment, tolerance, craving, positive reinforcement (mood and
social), negative reinforcement (mood and withdrawal), cognitive enhancement, task/skill
mastery (i.e. hobbyism, modding, or device construction), automaticity, and cue reactivity
(social and environmental). These themes informed the development of an item pool that
composes this studys focal measure. In a psychometric study, a separate sample of young
adults (N = 304; 61.5% male; 62.7% white; 21.7% Asian; 5.9% black; 81.6% non-Hispanic or
Latino) self-identifying as regular ENDS users completed a survey on the Amazon MTurk
platform. We probed for distinct factors and investigated validity and reliability of these
factors. A unique ENDS user identity factor emerged. When controlling for DSM tobacco use
disorder symptoms, ENDS user identity was associated with past-month days used ( = .71, SE
= .15, p < .05), lifetime-use (OR = 1.08, 95% CI = 1.02-1.14, p < .05), and escalation (i.e.
controlling for prior-year report of outcome) in nicotine concentration ( = .19, SE = .08, p
<.05). These results suggest that ENDS user identity is a salient and unique correlate of ENDS
use, independent of past use and dependence.

Symposium 16C

SUBSTANCE USE IDENTITIES, CRAVINGS, AND ATTENTIONAL BIAS IN CIGARETTE SMOKING

AND ALCOHOL CONSUMPTION

Dr. Andrew Hertel, PhD

Substance use identities (e.g., smoker identity) are associated with regular substance use, but
the reasons why are not well known. Understanding the correlates of substance use identities
could inform substance use intervention efforts. Substance use craving and attentional bias
(AB) toward substance use cues are proximal motivational and cognitive antecedents of
regular substance use. In two studies, we examined whether substance use identities were
associated with substance use cravings and AB toward substance use cues. In Study 1, college
students who consumed alcohol in the past 30 days, a subset of whom also smoked a
cigarette in the past 30 days (n = 65, age M = 20.1, 60.0% female, 70.6% White; n = 32
smokers), completed measures of explicit and implicit drinking- and smoking-related
identities, alcohol and cigarette craving, as well as alcohol cue and smoking cue AB (Stroop
tests). Alcohol craving was correlated with implicit drinker identity [r(63) = .29, p = .021 ] and
explicit future non-drinker identity [r(63) = -.31, p = .012], but not when controlling for
S428 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

drinking frequency. Both cigarette craving and smoking cue AB were correlated with explicit
smoker identity [r(30) = .67, p < .0001; r(29) = .42, p = .022] and explicit current non-smoker
identity [r(30) = -.40, p = .024 ; r(29) = -.42, p = .020]; when controlling for smoking frequency,
only cigarette craving was correlated with only explicit smoker identity [r(30) = .45, p = .011 ].
In Study 2, adult regular smokers with no intention of quitting in the next 30 days (n = 46, age
M = 39.2, 51.1% female, 80% White) completed measures of explicit and implicit smoking-
related identities, cigarette craving, and smoking cue AB (Stroop and visual dot probe tests).
Regardless of nicotine dependence level, cigarette craving was associated with explicit smoker
identity [r(42) = .43, p = .005 ] and implicit smoker identity [r(43) = .32, p = .038]; AB Stroop
was associated with implicit smoker identity [r(43) = .31, p = .044]; and AB visual dot probe
was correlated with explicit smoker identity [r(41) = .43, p = .006]. The findings indicate a
need for future investigations of the dynamic relationships between substance use identities,
craving, and attentional bias with respect to substance use as well as consideration of
substance use identities in substance use interventions.

CORRESPONDING AUTHOR: Kyle Baacke, BA, Knox College; [email protected]

ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S429

Symposium 17 8:00 AM-9:15 AM

CARING FOR TRAUMA SURVIVORS: CAREGIVER CHARACTERISTICS, COMPASSION FATIGUE,

AND OUTCOMES OF A SPIRITUAL INTERVENTION

Andrea D. Clements, PhD1, Jill E. Bormann, PhD, RN, FAAN2, Rose Constantino, PhD, JD, RN,
FAAN, FACFE, Fulbright Scholar3

1
East Tennessee State University, Johnson City, TN; 2VA San Diego Healthcare System Center
of Excellence for Stress and Mental Health (CESAMH), San Diego, CA; 3University of Pittsburgh
School of Nursing, Pittsburgh, PA

This symposium will present three papers addressing different aspects of religiousness/
spirituality applied to the realm of caregiving for trauma survivors. Religiousness/spirituality
will be explored as a characteristic of caregivers, as a buffer of compassion fatigue for
caregivers, and as an intervention in trauma survivors. The first paper addresses
characteristics of caregivers by occupation with particular emphasis placed on the caregiver's
own traumatic history through self-reported Adverse Childhood Experiences (ACE) scores,
self-reported level of compassion, and self-reported intrinsic religiosity. The second paper
offers a discussion of compassion fatigue in caregivers, pointing out the importance of self-
care and care for colleagues who experience compassion fatigue, endorsing the importance of
spiritual care among caregivers. The final paper will present findings from a prospective
randomized trial that tested the effects of a spiritual intervention with veterans diagnosed
with post-traumatic stress disorder (PTSD). It is hoped that a more general discussion will
follow regarding religious/spiritual contributions to the care of trauma survivors.

Symposium 17A

ACES, INTRINSIC RELIGIOSITY, AND COMPASSION IN HELPING PROFESSIONALS TARGETED

FOR TRAUMA-INFORMED CARE TRAINING

Dr. Andrea D. Clements, PhD

Background: Trauma-Informed Care (TIC), an empathy-centered paradigm focused on

ameliorating the impact of experienced trauma, is being disseminated to helping
professionals, but some occupational fields have been more receptive than others, inspiring
investigation of receptivity predictors. Total Adverse Childhood Experiences (ACE) score,
S430 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

intrinsic religiosity (IR), and compassion were compared across helping occupations,
conjecturing that individuals in resistant occupations may differ.

Objective: To investigate differences in ACE score, IR, and compassion across helping
professionals targeted to receive TIC training.

Methods: Surveys of TIC implementation and respondent characteristics were administered 6

months apart. Email invitations were sent to community crime reduction coalition members
(judicial, corrections, police, mental health, faith-based, medical, and other social service
organizations). Recipients were asked to complete the survey and forward the email to invite
others in these categories to complete the survey. Self-report items included occupation, the
ACE Test, IR, compassion, and demographics.

Results: At time 1, 30 surveys were completed and at time 2, 43 were completed by currently
employed human service workers and volunteers. ACE scores differed significantly (F(5,85) =
2.50, p =.036). Judicial/Corrections/Police had the fewest ACEs, but only significantly differed
from the Other group. Compassion differed across occupations with Volunteers, Educators,
and Counselors reporting higher levels than Judicial/Corrections/Police (2(10, n=73) = 18.88, p
= .04). IR differed by occupation, with Judicial/Corrections/Police, Volunteers, and Clergy
reporting high levels of IR and Counselors, Judicial/Corrections/Police, and University Faculty
reporting the lowest levels of IR (2(20, n=72) = 45.82, p = .001).

Discussion: Judicial/Corrections/Police and Educators have been more resistant to TIC training
than other groups such as Counselors and Social Workers. Judicial/Corrections/Police
employees tended to be lower in ACEs, IR,and compassion than those in other fields. This
could begin to explain the lack of TIC receptivity in probation/parole employees. Teachers
were high in compassion, so it seems that compassion alone may not predict receptivity to
TIC.

CORRESPONDING AUTHORS: Becky Haas, AS, Johnson City Police Department;

[email protected]; Valerie Hoots, MA, East Tennessee State University;
[email protected]

Symposium 17B

EFFECTS OF MANTRAM REPETITION ON SPIRITUAL WELLBEING AND MINDFULNESS IN

VETERANS WITH PTSD: A RANDOMIZED TRIAL

Dr. Jill E. Bormann, PhD, RN, FAAN

Objective: Few spiritually-based interventions have been tested for effects on spiritual
wellbeing and mindfulness in Veterans with posttraumatic stress disorder (PTSD). We
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S431

compared 8-weekly, one-hour individually delivered sessions of the Mantram Repetition

Program (MRP) with Present Centered Therapy (PCT) on subscales of spiritual wellbeing and
mindfulness.

Method: A prospective, randomized trial was conducted in 169 veterans with PTSD randomly
assigned to either MRP (n=87) or PCT (n=82). The Functional Assessment of Chronic Illness
Therapy-Spiritual Wellbeing (FACIT-Sp-12) and Five Facet Mindfulness Questionnaire (FFMQ)
subscales were assessed at baseline and post-treatment. Repeated measures analysis of
variance (ANOVAs) were used for analysis.

Results: Participants were 85% male, 65% white, 34% married/partnered with average age of
49 years (SD=14.46). There were group by time interaction effects showing that MRP
participants had significantly greater improvements in the subscales of FACIT-Peace,
F(1,132)=4.24, pF(1,132)=3.96, pF(1,133)=11.15, p=.001.

Conclusion: The MRP significantly improved levels of spiritual peace and mindful non-reacting
compared to PCT in Veterans with PTSD, suggesting possible for mechanisms of action. Both
MRP and PCT resulted in improvements over time on spiritual meaning.

CORRESPONDING AUTHORS: Princess Osei-Bonsu, PhD, Minneapolis VA Healtcare System &

University of Minnesota; princess.osei-bonsu; Danielle J. Beck, MPH, CCRC, VA San Diego
Healthcare System; [email protected]; Dorothy Plumb, MA, Edith Norse Rogers Memorial
Veteran Hospital; dorothy.plumb; Shahrokh Golshan, PhD, University of San Diego California;
[email protected]; A. Rani Elwy, PhD, VA Boston Healthcare System & Boston University of
Public Health; [email protected]

Symposium 17C

TITLE: EXPLORING THE CONSEQUENCES OF COMPASSION FATIGUE IN HEALTHCARE

PROVIDERS AND HEALING IN SPIRITUALITY

Dr. Rose Constantino, PhD, JD, RN, FAAN, FACFE, Fulbright Scholar

Purpose: The purpose of this presentation is to inform the attendees of the signs, symptoms,
and consequences of Compassion Fatigue (CF), the various ways it is displayed and the healing
effects of spirituality. The signs and symptoms of CF are not readily seen by the untrained eye
especially by the sufferer; it is an insidious and gradual lessening of compassion over time to
the observant, trauma-informed, and compassionate co-worker, supervisor or partner.

Literature Review: We present applicable evidence based literature, examine protective

factors and spirituality in both the healthcare provider and the CF sufferer. For example, Berg
and others (2016) studied CF in a trauma team and found that team members were not
managing work stresses as well as they perceived themselves to be. Wong (2004) presented a
S432 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

vision of positive holistic care through compassionate and spiritual care.He urgedhealthcare
providers to consider the benefits of a "bio-psychosocial-spiritual model" which is effective,
accessible and efficient.

Case Study: An extreme case of tension and preoccupation with the suffering of those being
helped to the degree that it traumatizes the helper describes CF. It has been suggested that
we enter into the healthcare giving field already with CF. CF differs from burnout as burnout is
exhaustion from the chronic tedium of the job (Tabor 2011). The clinician in the case study
will perform her duties and provide compassionate care to the patient devoid of CF. We will
leave it to the audience to imagine what if the healthcare provider in this case was suffering
from CF.

Implications: Whether we like it or not, believe or not, realize it or not, healthcare providers
are in the ministry of compassionate and spiritual care because we "help" and "care";
dedicated to serving the sick and relief of human suffering. Without compassionate and
spiritual care we become heartless and soulless. Spiritual care prods us to suffer with the
others, to enter the dark places where it hurts. To healthcare providers in CF, they are unable
to extricate themselves from those places and people who hurt; they stay there. Let's help
them through compassionate and spiritual care.

Conclusions: No one is immune to job-related trauma and stress. We need to care for each
other. We are indeed our brothers'/sisters' keeper. Being a compassionate and spiritual
healthcare professional is not expensive nor does it remove anything from us in fact it adds
something to the vast knowledge, information and evidence-based practice we already
possess. Watch out for CF in ourselves, in our family, co-workers and healthcare providers.

CORRESPONDING AUTHORS: Corrie Berk, University of Pittsburgh; [email protected]; Beth

Reefer, University of Pittsburgh; [email protected]; Linden Wu, University of Pittsburgh;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S433

Symposium 18 8:00 AM-9:15 AM

PROVIDING BRIEF EVIDENCE-BASED INTERVENTIONS IN HOSPITAL AND MEDICAL SETTINGS

Katherine Eisen, PhD1, Neda Kharrazi, PsyD1, Valerie Hoover, PhD2

1
Stanford Health Care, Stanford, CA; 2Stanford University, School of Medicine, Division of
Cardiovascular Medicine, Stanford, CA

With a move towards biopsychosocial conceptualization and integrated health care,

psychologists have taken on additional roles in medical settings. Psychologists across various
medical settings are paving the way for new roles and treatments to best meet the needs of
highly diverse individuals who seek and/or require psychological services. Brief psychological
interventions have increased in hospital settings, yet the empirical evidence for Cognitive
Behavioral Therapy interventions remains heavily dominated by studies with multiple sessions
(e.g. 12) and are often outpatient samples. There is a demand for very brief psychological
interventions to support patient care in medical settings, as well as some research to support
that such interventions can improve patient outcomes and be cost effective for health care
systems. This symposium will discuss some of the current literature, as well as highlight the
needs for future research.

Two presenters will focus on acute inpatient interventions for psychiatric patients. One
presenter will focus on psychosocial needs specific to cardiac patients, as well as a model for
integrated health care in cardiac care. Together, these presenters will explore three main
areas of discussion: (1) how to navigate best practices when psychological intervention time is
quite limited in medical settings; (2) how to collaborate effectively as a psychologist in an
interdisciplinary medical setting; and (3) areas for future research. A combination of empirical
and anecdotal evidence will be shared.

Symposium 18A

BUILDING AN EFFECTIVE COGNITIVE BEHAVIORAL THERAPY GROUP IN AN ACUTE INPATIENT

PSYCHIATRIC SETTING

Dr. Katherine Eisen, PhD

S434 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Implementing an effective CBT group program in an inpatient psychiatric hospital setting

presents myriad challenges. Rapport must be established rapidly, groups have frequent
patient turnover, and group members present with high levels of symptomology, variable
presenting problems, and varying degrees of motivation. However, the inpatient setting
presents some unique advantages as well. The entire therapeutic milieu can reinforce the
work done in therapy groups, and because group members are inpatients, the frequency of
sessions can be much more intensive. Research shows that adding CBT group programming to
inpatient psychiatric units is correlated with improvements in patient satisfaction and patient
outcomes, reduced length of stay, and lowered readmission rates in the 30 days following
discharge. Because CBT is a collaborative, action-oriented, and goal directed form of
psychotherapy, it lends itself especially well to brief interventions. Effective inpatient CBT
groups maximize impact by targeting themes that will be relevant to individuals who are in
crisis, regardless of specific diagnosis. In this presentation, the author will briefly review the
literature and will draw from relevant clinical experiences to provide recommendations for
adapting traditional CBT group formats to meet the unique needs of acute inpatient settings.
Topics will include: 1) Identifying individuals most likely to benefit from CBT groups, 2)
Maximizing the impact of brief interventions with selection of appropriate high-yield group
topics, 3) Integration of DBT, ACT, and other mindfulness-based approaches into a more
traditional CBT-based group curriculum, 4) Selection of appropriate homework topics that
allow patients to continue their work outside of group hours, and 5) Strategies for
communicating effectively with multidisciplinary staff about group content and homework
assignments so that themes covered in group are reinforced throughout the milieu.

CORRESPONDING AUTHOR: Neda Kharrazi, PsyD, Stanford Health Care;

[email protected]

Symposium 18B

PROVIDING MEANINGFUL & EFFECTIVE BRIEF INDIVIDUAL PSYCHOLOGICAL INTERVENTIONS

IN AN ACUTE INPATIENT SETTING

Dr. Neda Kharrazi, PsyD

Despite the demand for very brief psychological interventions in acute inpatient hospital
settings, much of the empirical evidence for Cognitive Behavioral Therapy (CBT) interventions
is based on both (1) outpatient treatment and (2) a relatively set number of weekly sessions
(e.g., 12 weeks). There is a growing evidence base to support brief CBT interventions in acute
inpatient psychiatric settings, but given the highly variable presenting problems, lengths of
stay, and multicultural factors, within the context of a multidisciplinary team approach, there
is a need for continued research and discussion about what brief individual treatments are
most effective for various individuals in acute inpatient settings.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S435

CBT interventions lend themselves well to brief interventions, given the nature of the theory,
which can be practiced flexibly to best meet the needs of the particular individual. During this
presentation, current brief individual CBT approaches will be explored as they apply within an
acute inpatient hospital setting. Given the importance of building rapport quickly under time
constraints in such settings, foundational therapeutic factors will be discussed. Case examples
will be used to illustrate both the utility of CBT in individual cases, as well as some of the
challenges.

CORRESPONDING AUTHOR: Katherine Eisen, PhD, Stanford Health Care;

[email protected]

Symposium 18C

INTEGRATED BEHAVIORAL HEALTH CARE IN AN OUTPATIENT CARDIOLOGY SETTING:

OPPORTUNITIES, CHALLENGES AND CASE ILLUSTRATION

Dr. Valerie Hoover, PhD

Recent policy changes with the Affordable Care Act are increasingly moving the U.S.
healthcare system towards integration of mental and physical health services. Broadly,
research on integrated care has been shown to improve disease management and early
identification of mental health needs, and decrease health care utilization and associated
costs to individuals and systems.

Research shows that certain psychiatric disorders occur at higher rates among cardiac patients
than in the general population. Furthermore, the presence of certain psychosocial factors and
psychiatric disorders most notably depression, anxiety, anger/hostility, low social support,
and chronic stress (including work, marital and caregiving strain) can place patients at
increased risk for adverse cardiac outcomes. Mechanisms are multifactorial and may include
chronic activation of the sympathetic nervous system and HPA axis, increased inflammation
and endothelial dysfunction, as well as indirect causes such as poor adherence to medical
regimens and heart-healthy lifestyle behaviors.

There is a relative paucity of data on integrated care models in the cardiology setting
compared to primary care. This presentation will provide a brief overview of common
psychosocial needs in cardiac patients, and describe one model of integrated behavioral
health care delivery in an outpatient cardiology setting. The presentation will include
discussion of clinic needs assessment data, development of a brief assessment and treatment
model, psychosocial intervention targets and brief treatment modalities, reimbursement and
funding considerations, and outcomes measurement.
S436 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Katharine Edwards, PhD, Stanford University School of Medicine;

[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S437

Paper Session 1: Is Daily Weighing Best: Research on Weight Tracking 10:45 AM-11:00 AM

THE TRACKING STUDY: LONG-TERM WEIGHT CHANGES BY WEIGHING FREQUENCY CONDITION

IN A RANDOMIZED BEHAVIORAL WEIGHT LOSS TRIAL

Jennifer A. Linde, PhD

University of Minnesota, Minneapolis, MN

Background: Weight control trials, community studies, and systematic reviews suggest daily
weight tracking promotes better weight loss outcomes than less frequent tracking. However,
this has yet to be examined in the context of a larger long-term randomized trial with
extended follow-up.

Purpose: The goal of the Tracking Study was to test the effects of study-assigned self-weighing
frequency on weight loss outcomes during a 12-month weight loss program with a 12-month
follow-up.

Methods: Three hundred thirty-nine adults were recruited and randomized to daily, weekly,
or no weighing conditions. All were enrolled in a 12-month weight loss program designed to
integrate each weight tracking instruction with standard behavioral weight loss techniques
(goal setting, self-monitoring, stimulus control, dietary and physical activity enhancements,
lifestyle modifications) and with wireless technology (scales and iPods) for the two tracking
conditions. All groups were scheduled for 32 sessions over one year. Participants completed
surveys and were weighed by trained study staff every six months from baseline to 24
months.

Results: Mean age of participants was 46.5 10.2 years; 65% were women. Sixty-eight
percent were married or partnered and 64% had a college degree. Mean baseline body mass
index (BMI) was 33.0 3.6 kg/m2. Study retention at 24 months was 83.6% (n=280), with no
differences between conditions. General linear models assessed 24-month weight change,
adjusting for baseline weight. On average, participants regained 3.3 kg from 12 to 24 months.
For women, a weak dose-response association for weight change by weighing frequency was
observed, such that daily weighers had better weight outcomes on average than weekly
weighers, followed by those who did not weigh (-3.87 kg vs. -3.63 kg vs. -2.76 kg,
respectively); for men, weekly weighing forestalled weight regain better than daily or no
weighing (-7.41 kg vs. -5.22 kg vs. -5.39 kg, respectively).

Conclusions: Results suggest that self-weighing promotes weight loss for adults seeking to lose
weight. A potential gender difference in response to weighing remains to be explored.
S438 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Jennifer A. Linde, PhD, University of Minnesota, Minneapolis, MN,

55454; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S439

Paper Session 1: Is Daily Weighing Best: Research on Weight Tracking 11:00 AM-11:15 AM

CITATION AWARD WINNER

ASSOCIATIONS OF WEIGHT TRACKING FREQUENCY WITH PSYCHOSOCIAL OUTCOMES OVER 24

MONTHS IN A RANDOMIZED WEIGHT LOSS TRIAL

Jennifer A. Linde, PhD

University of Minnesota, Minneapolis, MN

Background: Some weight control experts argue that frequent body weight tracking may carry
adverse psychological consequences, yet scant data have been reported on the effects of
differential weight tracking frequencies on psychosocial outcomes over time.

Methods: 339 adults were randomly assigned to track weight daily, weekly, or never during a
12-month behavioral weight loss program with a 12-month follow-up. Participants completed
surveys and height/weight measurements every six months from baseline to 24 months.
Depression, anxiety, body image, binge eating, and barriers to weight tracking were assessed
at the same time points; participants assigned to weighing conditions reported weight
tracking perceptions at 6, 12, 18, and 24 months. Repeated measures models assessed
changes over time and differences by weighing frequency condition.

Results: Mean age of participants was 46.5 years; the sample was 65% female. Mean baseline
body mass index was 33.0 kg/m2, and 4% of the sample reported binge eating symptoms at
baseline. Study retention at 24 months was 83.6%; participants achieved mean weight losses
of 8% at 12 months (post-intervention) and 4.4% at 24 months (final follow-up). There were
no differences in changes in depression, anxiety, body image, or self-esteem by weighing
frequency condition over time (p=.46-.87). On average, depression and anxiety remained in
the low normal range over the intervention and through follow-up. Body image suggested a
neutral to slightly unfavorable opinion of physical appearance at baseline but transitioned to a
slightly favorable and stable opinion over time (pppp=.77-.86).

Conclusion: The study was highly successful at assigning weight self-monitoring at varying
frequencies without adverse psychological effects.

CORRESPONDING AUTHOR: Jennifer A. Linde, PhD, University of Minnesota, Minneapolis, MN,

55454; [email protected]
S440 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 1: Is Daily Weighing Best: Research on Weight Tracking 11:15 AM-11:30 AM

PREDICTORS AND OUTCOMES OF LAPSES IN DAILY WEIGHING AND PHYSICAL ACTIVITY

TRACKING AMONG YOUNG ADULTS IN WEIGHT GAIN PREVENTION

Brooke T. Nezami, PhD, Carmina G. Valle, PhD, MPH, Deborah F. Tate, PhD

University of North Carolina at Chapel Hill, Chapel Hill, NC

Self-monitoring is a key strategy for weight loss and maintenance. However, adherence to
self-monitoring in weight control interventions generally declines over time and little is known
about predictors of lapses in adherence. The goal of this study was to examine predictors of
lapses in daily weighing and wearing PA trackers, and the relationships between lapses in self-
monitoring with weight and activity outcomes. Data were drawn from a sample of young
adults enrolled in ongoing interventions within the SNAP weight gain prevention study.
Participants included in this analysis (N = 161, BMI = 25.2 3.1 kg/m 2, 32.5 4.9 years) had at
least five observations of self-monitoring data from a wireless scale and PA tracker within a
90-day interval. A weighing lapse was identified as one or more days without weighing and a
PA tracker lapse was identified as one or more days without wearing the tracker. Lapse length
in days was calculated for weighing and tracker use. Additional data included transmitted
scale weights and steps/day. Generalized estimating equations accounting for repeated
observations within individuals were used to evaluate associations between weighing and PA
tracker lapses and changes in weight and steps/day. There were 1629 lapses in weighing and
514 lapses in tracker use, with mean lapse lengths of 3.2 ( 4.2) and 3.3 ( 6.1) days,
respectively. For weight observations with no lapses (i.e. daily weighing), there was a mean
weight change of -0.09 lbs. per day. Every 1-day increase in lapse length was associated with a
0.06-lb. increase in weight since the prior weight observation (p < .0001). Predictors of a
weighing lapse included any weight gain (vs. loss) since the prior weight observation (OR =
1.20, 95% CI: 1.09, 1.32) and weight change over the prior 7 days, with an increase in odds of
a weighing lapse of 1.07 for each 1-lb. increase in weight (95% CI: 1.04, 1.10). PA tracker lapse
length was not associated with changes in steps/day or weight between wear days. The only
predictor of a PA tracker lapse was weight change over the prior 7 days, with an increase in
odds of a PA tracker lapse of 1.08 for each 1-lb. increase in weight (95% CI: 1.02, 1.14). These
results suggest that there may be greater barriers to daily weighing as compared to wearing
PA trackers, and that weight gain may discourage adherence to self-monitoring. Future
research could examine the type and frequency of messaging required to reduce the
frequency and length of lapses in self-monitoring.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S441

CORRESPONDING AUTHOR: Brooke T. Nezami, PhD, University of North Carolina at Chapel

Hill, Chapel Hill, NC, 27514; [email protected]
S442 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 1: Is Daily Weighing Best: Research on Weight Tracking 11:30 AM-11:45 AM

CITATION AWARD WINNER

NONADHERENCE TO DAILY WEIGHING AND ACTIVITY TRACKING IS ASSOCIATED WITH WEIGHT

FLUCTUATIONS AMONG BREAST CANCER SURVIVORS

Carmina G. Valle, PhD, MPH, Chantel L. Martin, PhD, MSPH, Deborah F. Tate, PhD

University of North Carolina at Chapel Hill, Chapel Hill, NC

Background: Daily self-weighing (DSW) and daily activity tracking (DAT) have recently been
shown to be useful strategies for preventing weight gain among African American breast
cancer survivors. However, self-monitoring behaviors vary over time, increasing an
individuals risk of weight gain. This study examined the association of nonadherence to DSW
and DAT with corresponding weight fluctuations among a sample of African American breast
cancer survivors participating in a 6-month pilot intervention trial focused on DSW to promote
weight gain prevention.

Methods: Women were randomized into one of three groups: 1) DSW (INT); 2) DSW + DAT
(INT+); or 3) delayed intervention control. Intervention participants received a wireless scale
and activity tracker (INT+ only) that transferred objective data to a mobile app and website,
email lessons on weight control, and tailored feedback on objective weight (and activity data).
The intervention encouraged use of DSW as the primary self-monitoring strategy to prevent
weight gain. Participants in the INT+ group additionally were encouraged to track their
physical activity daily. Nonadherence to DSW was defined as one or more days without a
weight measurement, and nonadherence to DAT was defined as one or more days without a
bout of physical activity. Generalized estimating equations (GEE) accounting for repeated
measurements within individuals were used to examine the association between
nonadherence to DSW and weight fluctuations (kg) among women included in both the INT
and INT+ groups (n=24; n=2,438 weight measurements). We also used GEE to assess the
association of nonadherence to DAT in relation to weight fluctuations among women in the
INT+ only group (n=11; n=1,214 weight measurements).

Results: At baseline, women were 52.4 8.2 years of age with BMI of 33.3 5.7 kg/m 2. Over
the 6-month study period, women provided on average 119.2 46.0 weight measurements
and 120.3 53.5 days with bouts of physical activity. Interruptions in DSW and DAT were
associated with weight fluctuations. For every 1-day increase in nonadherence to DSW,
weight increased by 0.031 kg (95% CI: 0.012, 0.050; p < 0.01). Additionally, weight increased
by 0.035 kg (95% CI: 0.019, 0.050; p < 0.0001) for every 1-day increase in nonadherence to
DAT. During periods of DSW, weight change was, on average, -0.03 kg (-0.04, -0.01; p <
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S443

0.0001), while during periods of DAT, mean weight change was -0.05 kg (-0.08, -0.02; p <
0.01).

Conclusion: Our findings demonstrate, among a sample of African American breast cancer
survivors, the significance of DSW and DAT for weight gain prevention. Further research is
needed to better understand predictors of nonadherence to DSW and physical activity
tracking in larger study populations.

CORRESPONDING AUTHOR: Carmina G. Valle, PhD, MPH, University of North Carolina at

Chapel Hill, Chapel Hill, NC, 27599-7294; [email protected]
S444 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 2: Work Site Wellness Programs to Promote Physical Activity and Prevent
Obesity 10:45 AM-11:00 AM

CITATION AND MERITORIOUS AWARD WINNER

A COMPARISON OF REACH, RETENTION, UPTAKE AND SATISFACTION BETWEEN THREE

DELIVERY MODES OF FUEL YOUR LIFE

Heather M. Padilla, MS, RDN, LD1, Heather S. Zuercher, B.S., MPH2, Melissa Mitchell, M.S.3,
David DeJoy, PhD4, Mark G. Wilson, HSD5

1
Workplace Health Group, University of Georgia, Athens, GA; 2University of Georgia,
Jefferson, GA; 3University of Georgia, Athens, GA; 4College of Public Health, University of
Georgia, Athens, GA; 5University of Georgia/College of Public Health, Athens, GA

Background: A randomized control trial has shown that a worksite translation of the Diabetes
Prevention Program, FUEL Your Life, can be effectively delivered using three different modes
phone coaching, small group coaching, and self-study. Participants in all conditions lost weight
with the phone condition demonstrating significantly greater weight loss and higher costs
than the other two conditions. The purpose of this presentation is to describe the feasibility of
each delivery modality including strengths and weaknesses.

Methods: The process evaluation framework by Saunders, Evans, and Joshi (2005) was used to
compare reach, retention, dose received (uptake), and satisfaction among the three delivery
modes. Sources of data include participant surveys and health coach records. Differences
between delivery modes were tested using Pearson Chi-Square test and one-way ANOVA.

Results: A total of 660 employees enrolled in the study at three worksites. Twenty-one
percent of eligible employees enrolled at the worksite assigned to the phone condition
compared to 15% for the small group condition and 9% for the self-study condition. Retention
in the program at the end of the intervention was higher in the self-study condition than in
the phone and small group conditions and the small group condition had higher retention
than the phone condition (p < 0.01). The mean number of lessons read was higher for the
phone (13.1) and small group (12.4) conditions compared to the self-study condition (9.2; p <
0.01). A higher percentage of participants in the group (48%) and phone (43%) conditions
reported keeping track of their weight weekly compared to the self-study condition (36%),
although these differences were not statistically different (p=0.13). More than 90% of
participants in all conditions reported that the program provided the information that they
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S445

needed to lose or maintain weight and overall satisfaction.

Discussion: Reach in this worksite weight management study was higher for phone and small
group coaching conditions compared to the self-study condition. Additionally, participants in
the phone and group coaching conditions also reported higher intervention uptake likely
attributable to the coaches holding participants accountable for adherence to these elements.
Retention was higher for the self-study condition compared to the other two modes of
delivery and may differ due to the level of effort needed to participate in the self-study
compared to the other conditions and the lack of ongoing contact with a health coach.
Despite differences in uptake and retention, satisfaction was similar among the groups. It is
important to consider reach, retention, uptake and satisfaction when comparing intervention
effectiveness and determining the best way to offer interventions in the future. Differences
between the modalities and implications for future research will be discussed.

CORRESPONDING AUTHOR: Heather M. Padilla, MS, RDN, LD, Workplace Health Group,
University of Georgia, Athens, GA, 30602; [email protected]
S446 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 2: Work Site Wellness Programs to Promote Physical Activity and Prevent
Obesity 11:00 AM-11:15 AM

A HOLIDAY WEIGHT MANAGEMENT PROGRAM HELPS EMPLOYEES LOSE WEIGHT

Heather M. Padilla, MS, RDN, LD1, Mark G. Wilson, HSD2, Michael Bien, BS1, Carmen Daniel,
BS3

1
Workplace Health Group, University of Georgia, Athens, GA; 2University of Georgia/College
of Public Health, Athens, GA; 3Georgia Department of Public Health, Atlanta, GA

Introduction: Adult weight gain during the holidays (Nov.-Jan.) averages 0.5 kg though it
varies as much as -7 to 4.1 kg among individuals. Holiday weight gain appears to be due to
both excess energy intake and decreased expenditure. Previous studies suggest that the
weight gained is retained thus likely contributing to the obesity epidemic. This study examined
the effects of a team-based weight management program offered in a worksite setting on
prevention of holiday weight gain.

Methods: Holiday Survivor is a 10-week program offered to employees at the Georgia

Department of Public Health in November January. Participants form teams of four with
coworkers to support one another with mindful eating and physical activity through the
holiday season. Pre-test and post-test weight and waist circumference are measured by
wellness staff. Additionally participants complete a pre- and post-test survey on physical
activity, eating behaviors, social support, and program participation.

Results: In 2015, 102 employees enrolled in the program. Participants had a 2.45 kg reduction
in mean body weight (p < .0001) and a 1.1 inch reduction in waist circumference (p < .0001).
Overall, 82% of participants lost some weight and an additional 3% maintained their weight.
Self-reported physical activity increased, fast food consumption decreased and fruit and
vegetable intake increased. Participants reported that weigh-ins, fitness classes, cash prizes
for winners, walk/runs with their team, giveaways, and logging meals were important to their
success.

Discussion: A team-based behavioral weight management program was effective for weight
loss during the holiday season. Increased physical activity and improved eating behaviors
offset typical holiday weight gain. Holiday weight gain should not be considered inevitable.
Future research will examine the effects of the program on maintenance of weight loss and
positive eating and activity behaviors.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S447

CORRESPONDING AUTHOR: Heather M. Padilla, MS, RDN, LD, Workplace Health Group,
University of Georgia, Athens, GA, 30602; [email protected]
S448 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 2: Work Site Wellness Programs to Promote Physical Activity and Prevent
Obesity 11:15 AM-11:30 AM

TAKING WELLNESS TO THE BANK: AN INTEGRATED SERVICES WORKSITE WELLNESS PROGRAM

FOR BANK EMPLOYEES

Morgan S. Lee, PhD, MPH, CPH1, Paige Whitney, Health Promotion, MSc2, Mary Ann Kluge, BS,
MS, PhD3

1
University of Colorado - Colorado Springs, Colorado Springs, CO; 2Assistant Program Director,
Colorado Springs, CO; 3Beth-El College of Nursing and Health Sciences University of Colorado
Colorado Springs, Colorado Springs, CO

American adults spend half their waking time at work, often in highly stressful and sedentary
roles. Thus, the workplace is a critical environment for healthy lifestyle choices regarding
stress management, physical activity, and diet. Increasingly, employers are adopting worksite
wellness programs (WWPs) to promote healthy behaviors and disease management. Although
experts have called for multidimensional and integrated approaches to WWPs, few of these
programs meet such standards.

The Kirkpatrick Bank Workplace Wellness program (KBWWP) uses an integrated model to
provide comprehensive assessment of several domains of health and wellbeing, personalized
behavior change recommendations and coaching, and an array of related services. KBWWP
aims to help participants become more educated, mindful, and proactive about healthy
lifestyle choices and to make sustainable health behavior changes that produce long-term
benefits including improved health and reduced healthcare utilization. Additionally, KBWWP
will evaluate which program components have the greatest impact, what intervention dose is
needed to produce significant improvements, and how behavior change influences work
stress, work productivity, and healthcare utilization.

The KBWWP process begins with a health coach consultation as well as physical activity (PA)
and nutritional assessments, after which the providers recommendations are used to create
an integrated wellness plan. Participants complete an 8-week engagement phase with
opportunities including wellness services (e.g., personal training and nutrition education
classes) and additional appointments with service providers. A mid-program evaluation is then
conducted, and new recommendations are formulated for an 8-week maintenance phase,
which is followed by a final assessment.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S449

The KBWWP has reached the end of the 8-week engagement phase, and 16 of the 19
participants enrolled in the program have completed both baseline and mid-program
assessments. Participants weekly minutes of moderate and vigorous PA increased from
baseline (M = 173.00, SD = 163.43) to mid-program (M = 217.97, SD = 132.07), and sugar-
sweetened beverage intake decreased from baseline (M = 0.78, SD = 0.79) to mid-program (M
= 0.50, SD = 0.63). Participants were confident they could maintain the behavior changes they
made in the engagement phase (M = 8.15, SD = 1.68; 1-10 scale), and additional results are
forthcoming at the end of the maintenance phase.

CORRESPONDING AUTHOR: Morgan S. Lee, PhD, MPH, CPH, University of Colorado - Colorado
Springs, Colorado Springs, CO, 80905; [email protected]
S450 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 2: Work Site Wellness Programs to Promote Physical Activity and Prevent
Obesity 11:15 AM-11:30 AM

DOES A WORKPLACE PHYSICAL ACTIVITY PROGRAM WORK?

Rebecca Ellis, PhD, Duke Biber, MA

Georgia State University, Atlanta, GA

Objectives: Desire2Move (D2M) is an 8-week team-based program designed to encourage

university employees to meet the physical activity (PA) guidelines through participation in
moderate-to-vigorous physical activity (MVPA). The purpose of this study was to evaluate
D2M using the RE-AIM framework. Methods: During D2M, departments competed as teams
to accumulate the greatest average mins of MVPA. Each team selected a captain who sent
weekly emails with reminders, motivational tips, and team standings. Each team member
recorded minutes of MVPA using the MapMyRun mobile application or website. Graduate
assistants were assigned to specific teams to collect minutes. One week after D2M, an
electronic survey link was emailed to assess adverse events, program satisfaction and
implementation (1 = strongly disagree, 5 = strongly agree), and individual maintenance of
MVPA (Godin Leisure Time Exercise Questionnaire). Results: Twelve departments were invited
to participate in D2M, but only nine agreed (75.0% adoption rate). From these departments,
157 employees volunteered for the 8-week program (26.7% participation rate), and 72
consented to complete the follow-up survey (45.9% response rate). During the program,
participants recorded a weekly average of 233.50 mins of MVPA and only 1 participant
reported an adverse event. Participants also reported the program encouraged them to be
more active than usual (M = 3.74, SD = 1.27). Participants reported satisfaction with D2M by
indicating they would participate in future programs (M = 4.35, SD = 1.05) and would
recommend it to their co-workers (M = 4.40, SD = 1.08). Participants perceived the program
was well implemented by captains emailing information about how to record PA (M = 4.51, SD
= 0.94), weekly reminders (M = 4.54, SD = 0.95), and weekly team standings (M = 4.62, SD =
0.93). Finally, participants reported maintaining their PA since the end of the program (M =
3.06, SD = 0.67) and reported an average of 53.51 METS (SD = 39.37) of MVPA. Conclusions:
The RE-AIM evaluation revealed the program effectively, safely, and satisfactorily encouraged
university employees to meet the PA guidelines while demonstrating a very good adoption
rate, but a low participation rate. The results provide support for the translation of D2M to
other workplace settings, but strategies that improve reach must be identified.

CORRESPONDING AUTHOR: Rebecca Ellis, PhD, Georgia State University, Atlanta, GA, 30252;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S451

Paper Session 3: Sun Safety and Preventing Skin Cancer Among Youth 10:45 AM-11:00 AM

CITATION AWARD WINNER

UV PHOTOGRAPHY INCREASES SUN PROTECTION INTENTIONS IN YOUNG ADOLESCENTS

Mary K. Tripp, PhD, MPH1, Jian Wang, PhD2, Payal Pandit Talati, MPH1, Martha A. Askins,
Ph.D.1, Robert Dellavalle, MD, PhD, MSPH3, Carmen Galvan, MA4, Michael Davies, MD, PhD5,
Jeffrey E. Gershenwald, MD1, Susan Peterson, PhD, MPH5

1
The University of Texas MD Anderson Cancer Center, Houston, TX; 2UT MD Anderson Cancer
Center, Houston, TX; 3University of Colorado Denver, Denver, CO; 4The University of Texas
MD Anderson Cancer Center, Pearland, TX; 5MD Anderson Cancer Center, Houston, TX

Ultraviolet (UV) exposure from the sun or indoor tanning devices during adolescence
increases skin cancer risk. A UV light photograph reveals skin damage caused by UV exposure.
This skin damage is not visible in a standard (visible light) photograph. Immediately after
viewing a UV photograph of the face, adults report increased sun protection intentions,
stronger perceptions of photoaging susceptibility and more negative tanning attitudes. To our
knowledge, our study is the first randomized controlled trial (RCT) to examine immediate
effects of UV photography on socio-cognitive factors in young adolescents. We conducted a 2-
arm cluster-RCT in 12 Houston-area middle schools. Schools were pair-matched and
randomized to intervention (UV photography and brief education) or comparison (brief
education only) before participants (N=290 7th and 8th graders) completed baseline.
Intervention-school participants were shown their standard and UV photographs side-by-side
for comparison. Study personnel reviewed with participants the photographs, including any
skin damage visible in the UV photograph. Personnel then provided brief education by
reviewing with participants a skin cancer prevention handout. Comparison-school participants
received only brief education. Immediately after the intervention/comparison session,
participants completed a post-intervention survey. The trial design was conducted in 2015 (4
schools, N=98) and 2016 (8 schools, N=192) and data were combined. Generalized linear
mixed models were conducted, with treatment condition as a fixed effect and a school-
specific random intercept to account for the nested design. Analyses were adjusted for age,
race, ethnicity and study year. The majority of participants were female (71.7%), white
(73.6%) and not Hispanic or Latino (73.7%). Mean age was 13.2y (SD=.75, range 12-15y). There
was a significant, positive intervention effect on sun protection intentions (effect=.16, SE=.05,
S452 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

P < .01): the intervention group showed a greater positive change from baseline to post-
intervention in the sun protection intentions score, which was the average of four items that
assessed intention to use sunscreen, wear a sleeved shirt, wear a hat and stay in the shade.
There were no significant effects on intentions to sunbathe or indoor tan, tanning outcome
expectations, sociocultural influences to tan, perceived severity of photoaging or perceived
skin cancer risk. Results suggest UV photography may be a valuable intervention for increasing
sun protection intentions in young adolescents. UV photography may be less effective in
immediately affecting other socio-cognitive factors, such as those that may require more time
to demonstrate change. A 3-month follow-up is underway to examine longer term effects of
UV photography on socio-cognitive factors and UV exposure and protection behaviors in
adolescents.

CORRESPONDING AUTHOR: Mary K. Tripp, PhD, MPH, The University of Texas MD Anderson
Cancer Center, Houston, TX, 77230-1439; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S453

Paper Session 3: Sun Safety and Preventing Skin Cancer among Youth 11:00 AM-11:15 AM

INDOOR TANNING AMONG DIVERSE U.S. YOUTH: RESULTS FROM A NATIONAL SAMPLE

Aaron Blashill, Ph.D.

San Diego State University, San Diego, CA

Sexual minority males have one of the highest known prevalence rates of skin cancer, at 4.3%
to 6.6% (an increased odds of 1.5 to 2.0) compared to heterosexual males. One likely
explanation for this health disparity is sexual minority males use of indoor tanning, a Group 1
carcinogen. However, no known studies have examined indoor tanning by sex, sexual
orientation, and race/ethnicity, casting uncertainty regarding which groups are most
vulnerable for developing skin cancer.

Data were used from the 2015 Youth Risk Behavior Survey (YRBS), a nationally representative
survey that examines the prevalence of health risk behaviors among 9-12 grade public and
private school students. Sexual minority status was defined from responses to two items:
sexual identity and sex of sexual partners, with sexual minorities denoted as participants who
reported a non-heterosexual identity or reported sex with a member of their own sex (a
common approach in the field). Past 12 month indoor tanning was dichotomized as one or
more times vs. none. The total employed sample was N = 10,660, with 1,256 sexual minority
participants (21.0% females; 5.6% males). Logistic regressions were used, with independent
variables of sex, sexual orientation and race/ethnicity (i.e., dummy coded Black, Hispanic, with
White as the referent). Complex Samples in SPSS 24 was used to account for weighting,
cluster, and stratification.

Significant three-way interactions were revealed, thus, analyses were stratified by

race/ethnicity. Among Black participants, there was a main effect of sexual orientation (OR =
6.5, 95% CI: 3.3, 12.9, p < .01) and sex (OR = 2.9, 95% CI: 1.2, 7.3, p = .02), with sexual
minorities and males reporting elevated indoor tanning. Among Hispanic participants, there
was a main effect of sexual orientation (OR = 5.7, 95% CI: 2.3, 14.2, p < .01), with sexual
minorities reporting elevated indoor tanning. Among White participants, there was a sexual
orientation by sex interaction (F1,35 = 28.0 p < .01). Follow-ups revealed that sexual minority
status was a risk factor within males (OR = 3.1, 95% CI: 1.3, 7.7, p = .01), and a buffer variable
within females (OR = .41, 95% CI: .24, .73, p < .01).

Results highlight the need to incorporate sex, sexual orientation, and race/ethnicity when
developing skin cancer prevention programs for youth. For example, Black sexual minority
males reported the highest prevalence of indoor tanning, a rate equivalent if not higher than
White females. Clinicians working with sexual minority males, particularly males of color
should consider assessing use of indoor tanning during routine evaluations. Future research
S454 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

would benefit from exploring motivations to tan among diverse groups of adolescents, as
sexual minorities may use indoor tanning as strategy to regulate excess negative affect.

CORRESPONDING AUTHOR: Aaron Blashill, Ph.D. , San Diego State University, San Diego, CA,
92120; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S455

Paper Session 3: Sun Safety and Preventing Skin Cancer among Youth 11:15 AM-11:30 AM

SUNBURN, SUN PROTECTION AND TANNING BEHAVIOR IN YOUNG ADOLESCENTS

Mary K. Tripp, PhD, MPH1, Jian Wang, PhD2, Payal Pandit Talati, MPH1, Martha A. Askins,
Ph.D.1, Robert Dellavalle, MD, PhD, MSPH3, Carmen Galvan, MA4, Michael Davies, MD, PhD5,
Jeffrey E. Gershenwald, MD1, Susan Peterson, PhD, MPH5

1
The University of Texas MD Anderson Cancer Center, Houston, TX; 2UT MD Anderson Cancer
Center, Houston, TX; 3University of Colorado Denver, Denver, CO; 4The University of Texas
MD Anderson Cancer Center, Pearland, TX; 5MD Anderson Cancer Center, Houston, TX

Overexposure to ultraviolet (UV) radiation from the sun and UV exposure from indoor tanning
during adolescence increases skin cancer risk. Very little is known about UV exposure and sun
protection behaviors in young adolescents in the US. We analyzed baseline data from
participants (N=290) in 12 Houston-area middle schools enrolled in a group-randomized trial
of a UV photography intervention. The trial design was conducted in 2015 (4 schools, N=98)
and 2016 (8 schools, N=192) and data were combined. The majority of adolescents were
female (71.7%), white (73.6%) and not Hispanic or Latino (73.7%). Mean age was 13.2y
(SD=.75, range 12-15y). Sun protection practice most of the time or always varied by
behavior: using sunscreen (31.4%), wearing a shirt with sleeves covering the shoulders
(68.7%), wearing a hat (5.9%) and staying in the shade (13.1%). Averaging behavior scores (1-5
never to always scale) showed adolescents sometimes practiced sun protection
(M=2.88, SD=.52). More adolescents reported intentions to use sun protection in the next 3
months: sunscreen (87.6%), sleeved shirt (84.5%), hat (49%) and shade (64.3%). A 4-point
scale showed moderate intentions overall (M=2.87, SD=.45). Sunburn in the past year was
reported by 81.1%. One adolescent (.3%) reported using an indoor tanning device in the past
year. Five adolescents (1.7%) intended to indoor tan in the next 3 months. Sunbathing often
or always was reported by 10.4%. Intention to sunbathe in the next 3 months was reported
by 33.5%. Some wished they had a tan like people on TV (14.1%) or responded that friends
said they looked good when tan (35.0%). On 4-point scales, adolescents perceived negative
tanning outcome expectations (mean=2.10, SD=.45) and moderate perceptions about the
severity of photoaging (mean=2.75, SD=.68). Most (75%) perceived themselves unlikely to
develop skin cancer. In multivariable analyses, sun protection behavior correlates that
remained significant were stronger sun protection intentions, lighter skin color and male sex.
Significant sunbathing correlates were stronger sunbathing intentions, weaker sun protection
intentions, more positive tanning outcome expectations and female sex. Results show that
early adolescence is a critical time for intervention to reduce UV exposure. Adolescents
reported sunbathing, significant sunburn history and inadequate sun protection. Reported
S456 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

intentions to sunbathe and indoor tan suggest that more adolescents planned to practice
these risky behaviors over the next few months. Intentions to sunbathe and use sun
protection and tanning outcome expectations are important factors to influence with
interventions designed to reduce UV exposure in adolescents.

CORRESPONDING AUTHOR: Mary K. Tripp, PhD, MPH, The University of Texas MD Anderson
Cancer Center, Houston, TX, 77230-1439; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S457

Paper Session 3: Sun Safety and Preventing Skin Cancer among Youth 11:30 AM-11:45 AM

PHYSICAL ACTIVITY MOTIVES AND SUN SAFETY AMONG TEENS

Kasey Morris, PhD1, Frank Perna, EdD, PhD2

1
National Cancer Institute, Bethesda, MD; 2National Cancer Institute, Rockville, MD

Recent evidence suggests that physically active adults have higher melanoma rates, ostensibly
via relatively greater incidence of sun exposure and sunburn. Adolescents, especially active
teens, may also experience greater sun exposure than adults, and sunburn during youth is
especially predictive of melanoma later in life. Using data on adolescents (N = 1661) from the
National Cancer Institutes Family Life, Activity, Sun, Health, and Eating (FLASHE) national
survey, our purpose was to determine if physical activity level and motivational differences for
physical activity are associated with sun exposure and sun safety. After controlling for
demographic factors, results demonstrated a 3-way interaction between physical activity,
sports performance motivation, and history of sunburn on sunscreen use, b = .001, p = .029.
Among highly active teens with a history of sunburn (at least one in the past 12 months),
sunscreen use was reportedly greater among those high on sports performance motivation.
However, sports performance motivation for physical activity was also associated with
increased likelihood of sunburn (OR = 1.23, p = .008), regardless of physical activity level.
There was also an interaction between appearance motivation and sunbathing on sunscreen
use, b = .120, p = .023. Appearance motivation for physical activity was positively associated
with sunscreen use, but only among adolescents who engage in sunbathing (b = .078, p =
.029). Motivation for physical activity may play an important role in understanding why
adolescents engage in sun protection strategies. For at risk teens (i.e., those with a history of
sunburn), physical activity and high sports motivation was associated with greater reported
sunscreen use, but this relationship was not found among those without a history of sunburn.
This may reflect an inattentiveness to sun exposure when involved in sports and outdoor
physical activity, suggesting the need for greater vigilance and infrastructure to support sun
safety in physical activity contexts (e.g., sunscreen availability). In contrast, adolescents with
high appearance motivation for physical activity and a sunbathing history reported greater
sunscreen use, suggesting a possible strategic form of sun protection (i.e., to tan, but not
burn). These results highlight the importance of considering multiple motives for physical
activity and possible behavioral phenotypes for effective policy change and health behavior
promotion related to sun safety.

CORRESPONDING AUTHOR: Kasey Morris, PhD, National Cancer Institute, Bethesda, MD,
20892; [email protected]
S458 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 4: What Drives Activity? Motivation Science 10:45 AM-11:00 AM

USING SELF-DETERMINATION THEORY TO EXAMINE PROSPECTIVELY THE RELATION OF BODY

SATISFACTION WITH FITBIT-ASSESSED EXERCISE

Miriam H. Eisenberg, Ph.D.1, L. Alison Phillips, PhD2

1
Eunice Kennedy Shriver National Institute of Child Health and Human Development, NICHD,
ARLINGTON, VA; 2Iowa State University, Ames, IA

Regular physical activity (PA) is critical for promoting health and preventing disease but it
declines in young adulthood. Body dissatisfaction, which increases in young adulthood, could
exacerbate this decline in PA. Research is needed to test this association and examine the
mechanisms through which this occurs. Previous research shows that components of Self-
Determination Theory (SDT), mainly autonomous motivation and psychological needs
satisfaction, mediate this association cross-sectionally using self-reported PA. However,
intrinsic goals (e.g., exercising to improve health versus appearance) may mediate this
association as well, which would indicate an additional point for intervention. Further,
objective PA measurement and longitudinal data are needed to assess these associations
more precisely. The current study examined the extent to which intrinsic goals, psychological
needs satisfaction, and autonomous motivation mediated the association of body satisfaction
with PA measured objectively.

Participants (n=121, 73% White, mage=25y, BMI =24kg/m2, 75% female) reported their body
satisfaction and motivation, goals, and psychological needs satisfaction for PA. Fitbits
recorded their PA over the following two weeks; PA was operationalized as the proportion of
days with an exercise session/days wearing the Fitbit. Bootstrapping process macro 6 was
used to test whether the proposed SDT framework that included intrinsic goals mediated the
association of body satisfaction with PA adjusting for gender and BMI.

Body satisfaction was significantly associated with PA ( = 0.23, p=0.03). Bootstrapping

revealed support for the hypothesized model where intrinsic goals, psychological needs
satisfaction, and autonomous motivation all mediated the association of body satisfaction
with PA (Standardized Indirect Effect =0.0003, 95%CI: 0.0000, 0.0013).

Findings support and extend previous literature by using objective and prospective data to
identify intrinsic goals as an additional point for intervention. Based on these findings,
researchers and health professionals could improve PA in this age group by promoting body
satisfaction or targeting goals, needs satisfaction, or autonomous motivation directly. For
example, researchers could promote health-focused (intrinsic) versus appearance focused
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S459

(extrinsic) PA goals. The strengths and limitations to using Fitbits for data collection will also
be discussed.

CORRESPONDING AUTHOR: Miriam H. Eisenberg, Ph.D., Eunice Kennedy Shriver National

Institute of Child Health and Human Development, NICHD, ARLINGTON, VA, 22201-6500;
[email protected]
S460 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 4: What Drives Activity? Motivation Science 11:00 AM-11:15 AM

MOTIVATIONAL PREDICTORS OF PHYSICAL ACTIVITY AND SEDENTARY BEHAVIORS AFTER

REHABILITATION: A 6-MONTH PROSPECTIVE STUDY

Guillaume Chevance, PhD Student1, Nelly Hraud, PhD, Research director2, Nicolas Oliver,
Medical Doctor2, Julie Boich, PhD, Assistant Professor1

1
University of Montpellier, Montpellier, Languedoc-Roussillon, France; 2Groupe 5 Sant,
Lodeve, Languedoc-Roussillon, France

Two-third of the patients who followed a Pulmonary Rehabilitation (PR) program is

insufficiently active 6 months after. Previous studies have consistently highlighted that
exercise tolerance was a significant predictor of active behaviors after PR. Beyond this
physiological variable, motivation may also influence patients behaviors. Grounded in
contemporary dual models, this study sought to examine the contribution of controlled
(intentions) and automatic (implicit attitudes) motivational processes in the physical activity
(PA) and sedentary behaviors (SB) adopted after PR.

Patients with chronic diseases (N=62, Mage=61 years, MBMI=29kg/m) were

recruited at the end of a 5-week PR. Patients were evaluated on their exercise tolerance with
walking tests. Intentions to practice regular PA and to limit SB were measured by
questionnaires. Implicit attitudes (PA versus SB) were measured with an Implicit Association
Test. PA and SB were evaluated by questionnaires 6 months after PR. Regression were
conducted for PA and SB, respectively, with exercise tolerance, intentions and implicit
attitudes as independent variables, and behaviors as dependent variables.

Regarding PA, exercise tolerance (=.43; P < .001) and implicit attitudes (=.29;
P=.020) were significant predictors. Intentions were not associated to PA (=.10; P=.416). The
equation was statistically significant: F(3,53)=7.13; Radjusted=.30; P < .001. Regarding SB,
exercise tolerance (=-.45; P < .001) was a significant predictor. A statistical tendency was
observed for implicit attitudes (=-.21; P=.097), but intentions were not associated to SB
(=.15; P=.247). The equation was statistically significant: F(3,51)=5.98; Radjusted=.27;
P=.002.

Implicit attitudes are significantly associated with PA 6 months after PR,

independently from patients exercise tolerance. Hence, this variable may represent a
relevant target for future behavioral interventions. Studying independently implicit attitudes
for PA and SB is now needed to understand the patterns of relationships between those
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S461

motivational processes and behaviors. A replication study with an objective measure of

behaviors is also required.

CORRESPONDING AUTHOR: Guillaume Chevance, PhD Student, University of Montpellier,

Montpellier, Languedoc-Roussillon, 34000; [email protected]
S462 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 4: What Drives Activity? Motivation Science 11:15 AM-11:30 AM

SELF-REGULATION AND MOTIVATIONAL RESULTS OF A WEARABLE ACTIVITY MONITOR

INTERVENTION

Elizabeth J. Lyons, PhD1, Maria Swartz, PhD, MPH, RD, LD2, Zakkoyya Lewis, BS3, Eloisa
Martinez, BS, CCRP4

1
The University of Texas Medical Branch, Galveston, TX; 2UTMB, Missouri City, TX; 3University
of Texas Medical Branch, Bellflower, CA; 4University of Texas Medical Branch-Galveston, Texas
City, TX

Background: As adults age, their physical activity decreases, leading to increased risk of
numerous negative health outcomes. Increasing intrinsic motivation (i.e., motivation related
to intrinsic enjoyment) and self-regulation can increase physical activity. It has been
hypothesized that mobile applications (apps) combined with wearable activity monitors may
be an effective method of increasing both motivation and self-regulation. However, to our
knowledge, this hypothesis has not been tested. Thus, the purpose of this study was to
investigate the effect of an intervention that used an app and wearable monitor on intrinsic
motivation and self-regulation (self-efficacy, goal setting, and planning).

Methods: Adults (N = 40) aged 55 79 who self-reported as sedentary (< 60 minutes physical
activity/week) were randomized to a 12-week intervention or to a wait-list control. A Jawbone
Up24 wearable monitor, a tablet with the Up app installed, and weekly brief telephone
counseling were provided for the intervention group. Intrinsic motivation was tested using the
Behavioral Regulations in Exercise-2 questionnaire. Barrier and task self-efficacy were
assessed using the Barriers Self-Efficacy Scale and Rogers and colleagues Task Self-Efficacy
Scale. Goal setting and planning was assessed using Rovniak and colleagues measure.

Results: Participants were 61.5 5.6 years old, with a BMI of 30.3 3.5 kg/m2, 85% female
and 65% White. Analyses of covariance comparing the intervention to the wait-list group,
controlling for baseline values, found significant and large effects for intrinsic motivation (p =
.007, partial 2 = .18), task self-efficacy (p = .009, partial 2 = .17), barrier self-efficacy (p =
.003, partial 2 = .22), goal setting (p < .001, partial 2 = .41), and planning (p = .012, partial 2
= .16). The intervention group at 12 weeks reported 65% confidence to overcome barriers,
79% confidence to complete activity tasks, and intrinsic motivation of 3.0 on a 1 to 5 scale.

Discussion: As hypothesized, an intervention using an app and wearable monitor increased

intrinsic motivation, self-efficacy, goal setting, and planning. The large effect sizes found in
this pilot study suggest that these systems hold promise for targeting specific theoretical
constructs associated with behavior change.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S463

CORRESPONDING AUTHOR: Elizabeth J. Lyons, PhD, The University of Texas Medical Branch,
Galveston, TX, 77555; [email protected]
S464 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 4: What Drives Activity? Motivation Science 11:30 AM-11:45 AM

THE RELATIONSHIP BETWEEN SELF-EFFICACY, MOTIVATION, SOCIAL-SUPPORT AND EXERCISE

IDENTITY

Zakkoyya Lewis, BS1, Maria Swartz, PhD, MPH, RD, LD2, Eloisa Martinez, BS, CCRP3, Elizabeth J.
Lyons, PhD4

1
University of Texas Medical Branch, Bellflower, CA; 2UTMB, Missouri City, TX; 3University of
Texas Medical Branch-Galveston, Texas City, TX; 4The University of Texas Medical Branch,
Galveston, TX

Introduction: Physical inactivity is a prominent health concern. Physical activity interventions

that integrate social support or studies that are grounded in behavioral theory (e.g. Social
Cognitive Theory, Self-Determination Theory) have been implemented to change behavior.
One concept that is not often evaluated in these studies is exercise identity. Identity may work
with other behavioral techniques to add meaning and value to exercise which can elicit
behavior change. The purpose of this investigation was to investigate the relationship
between self-efficacy, social support, motivation and exercise identity.

Methods: Baseline data from a physical activity intervention were used. Participants (n=40)
enrolled in the intervention were 18-69 years of age, inactive, and overweight or obese (body
mass index 25-35 kg/m2). Self-reported measures were used to assess self-efficacy,
motivation, social support, and exercise identity. Self-efficacy was divided into task and
barrier self-efficacy using the scale developed by Rogers et. al. The Behavioral Regulation in
Exercise Questionnaire-2 was used for motivation and included subscales for amotivation,
external, introjected, identified, and intrinsic regulation. Perceived family and friend social
support was estimated from the Social Support and Exercise Survey. Exercise identity was
based off the Anderson measure. Bivarate Pearson correlations were used to identify
significant relationships between questionnaire subscales and exercise identity.

Results: On average, participants were 46.7 years old with a body mass index of 29.9 kg/m2.
Most participants were female (55%) and non-Hispanic White (65%). Barrier self-efficacy
(r=0.53; p

Conclusion: Exercise identity is strongly correlated to particular subsets of commonly used

behavioral constructs and strategies. Future research should evaluate the causal pathway
between these variables and investigate the influential differences between barrier and task
self-efficacy in addition to family and friend support. Furthermore, researchers should identify
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S465

methods to incorporate barrier self-efficacy and family support within the context of Self-
Determination Theory to enhance motivation.

CORRESPONDING AUTHOR: Zakkoyya Lewis, BS, University of Texas Medical Branch,

Bellflower, CA, 90706; [email protected]
S466 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 5: Depression in Diverse Populations 10:45 AM-11:00 AM

RELAPSE PREVENTION IN MAJOR DEPRESSIVE DISORDER: 24-MONTH FOLLOW UP OF

MINDFULNESS-BASED COGNITIVE THERAPY VS. ACTIVE CONTROL

Amanda J. Shallcross, ND, MPH1, Emily C. Willroth, MA2, Aaron Fisher, PhD3, Iris B. Mauss,
PhD4

1
New York University School of Medicine, New York, NY; 2University of California Berkeley,
Berkeley, CA; 3University of California, Berkeley, berkeley, CA; 4University of California,
Berkeley, Berkeley, CA

Background: Studies demonstrating the efficacy of mindfulness-based cognitive therapy

(MBCT) for depression relapse prevention and symptom reduction have primarily compared it
to treatment as usual and have scarcely examined outcomes beyond 12 months.

Objective: We evaluated the comparative effectiveness of MBCT versus an active control

condition (ACC) for depression relapse prevention and depressive symptom reduction over a
2-year follow up.

Method: Participants (n=92) in remission from major depressive disorder with residual
depressive symptoms were randomized to an 8-week MBCT or a validated ACC intervention
that is structurally equivalent to MBCT and controls for non-specific effects (e.g., interaction
with a facilitator, perceived social support, outcome expectations). MBCT is a standardized
intervention that combines training in mindfulness and cognitive therapy. The ACC was based
on the 'Health Enhancement Program' developed as a control condition for mindfulness
interventions by researchers at the U. of Wisconsin and NIH. It combines training in physical
activity, nutrition, and music therapy. Both interventions were delivered according to their
published manuals.

Results: Intention-to-treat analyses indicated no differences between MBCT and ACC in

incidence or time to relapse over a 2-year follow up (Fishers p=.65; Hazard ratio=.77, CI [.35-
1.68], p=.52). Piecewise analyses using mixed effects regression indicated significant
reductions in depressive symptoms for both groups over year 1, after which both groups
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S467

experienced a marginal rebound of depressive symptoms. Group x Time interactions indicated

differential symptom trajectories up to year 1. During the treatment period, the ACC group
experienced steeper declines in symptoms. From post-treatment to year 1, the MBCT, but not
the ACC group continued to experience symptom declines.

Conclusions: Over a 2-year follow up, MBCT did not differ from an ACC on depression relapse
rates or symptom reduction, suggesting that the ACC is as effective as MBCT for improving
depression outcomes. The trajectory of symptomatic improvement that favored MBCT after
the treatment period was not maintained after year 1. These findings underscore the
importance of comparing psychotherapeutic treatments to ACCs to isolate specific versus
nonspecific therapeutic components and to test the comparative effectiveness of approaches
that may have differential cost and dissemination advantages.

CORRESPONDING AUTHOR: Amanda J. Shallcross, ND, MPH, New York University School of
Medicine, New York, NY, 10016; [email protected]
S468 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 5: Depression in Diverse Populations 11:00 AM-11:15 AM

THE IMPORTANCE OF BASELINE PAIN, FATIGUE, SLEEP & PHYSICAL ACTIVITY: PREDICTING
CHANGE IN DEPRESSION IN ADULTS WITH MS

Karlyn A. Edwards, B.A.

University of New Mexico, Albuquerque, NM

Individuals with Multiple Sclerosis (MS) often suffer from a myriad of symptoms that place
them at higher risk for developing depression. These symptoms include pain, fatigue, and
poor sleep quality, and limited engagement in physical activity. However, little is known about
the impact of these symptoms on the course of depression over time. The current study
examined baseline levels of pain, fatigue, sleep quality, and physical activity level to
determine if they predicted either clinical improvement or development of depression 3.5
years later. The study involved 489 adults with a diagnosis of MS, who received two mail in
surveys that were 3.5 years apart. A PHQ-9 cut off score of 10 was used to indicate probable
major depression and scores < 10 indicating a sub-clinical level of depression. For the analysis,
two binary linear regressions were conducted. The first regression examined the role of
baseline pain, fatigue, sleep quality, and physical activity in predicting the likelihood that non-
depressed individuals at baseline later became depressed at follow up. The second binary
linear regression examined the role of baseline pain, fatigue, sleep quality, and physical
activity in predicting the likelihood that depressed individuals at baseline clinically improved
at follow up. Both regressions also controlled for age, sex, and diagnosis severity. The sample
was primarily female (82%), White/Caucasian (94%), highly educated (with 83% receiving at
least some college), and diagnosed with relapsing-remitting MS (53%). Our findings indicated
that poorer sleep quality and increased fatigue at baseline predicted the development of
depression (PHQ-9 10) 3.5 years later among those not depressed baseline. None of the
study factors predicted the improvement of depression among those classified as depressed
at baseline. These findings indicate that clinical interventions should emphasize treatment of
fatigue levels and sleep quality among those with MS to help prevent the future development
of depression, and that pain and physical activity level may not be as important. Concurrently,
the findings indicate that these factors may also not be important when examining the long
term improvement of depression among those with MS.

Learning Objective 1: Describe the impact of pain, fatigue, sleep quality, and physical activity
on depression among those with MS.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S469

Learning Objective 2: Discuss the implications of sleep quality and fatigue contributing to the
future development of depression over and above pain and physical activity.

Learning Objective 3: Discuss other possible factors (besides pain, fatigue, sleep quality, and
physical activity) that may contribute to the improvement of depression.

CORRESPONDING AUTHOR: Karlyn A. Edwards, B.A. , University of New Mexico, Albuquerque,

NM, 87108; [email protected]
S470 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 5: Depression in Diverse Populations 11:15 AM-11:30 AM

MATERNAL DEPRESSIVE SYMPTOMS AND LENGTH OF RESIDENCE IN THE US ARE POSITIVELY

RELATED TO OBESITY AMONG LOW-INCOME LATINA MOTHERS

Ana Cristina. Lindsay, DDS, MPH, Dr.PH1, Mary L. Greaney, PhD2, Julie A. Wright, PhD3, Sherrie
Wallington, PhD4, Anne Hunt, Sc.D.5

1
University of Massachusetts Boston, Boston, MA; 2University of Rhode Island, Kingstom, RI;
3
UMass Boston, Boston, MA; 4Georgetown University Medical Center, Lombardi
Comprehensive Cancer Center, Rockville, MD; 5Hunt Consulting Associates, Thetford Center,
VT

Background: Latinos are the largest and fastest growing minority population group in the
United States (US). Low-income Latina women are at elevated risk of being obese, making
obesity prevention a pressing public health concern in this population. Both depressive
symptoms and obesity are modifiable factors that can improve the health of women of
childbearing-age and their children. The purpose of this study was to examine the association
between maternal depressive symptoms and obesity among low-income, Latina women of
childbearing age.

Methods: Participants (n=147) were low-income mothers enrolled in the Latina Mothers
Child Feeding Practices and Styles (LMCFPS) Study, a cross-sectional study examining
psychosocial and cultural influences on child feeding practices and styles of low-income Latina
mothers of preschool-aged children. Women were eligible to participate if they were Latina,
mothers, eligible for or enrolled in the Special Supplemental Nutrition Program for Women,
Infants and Children and living in Rhode Island. Enrolled participants completed an
interviewer-administered survey in Spanish administered by bilingual interviewers in the state
of Rhode Island.

Results: Obesity (34%; 30kg/m(2)) and elevated depressive symptoms (28.3%; CES-D > 16)
were prevalent in this low-income, majority immigrant (69.4%) female Latina sample. Mothers
with elevated depressive symptoms (Center for Epidemiologic Studies Depression Scale score
> 16) had increased odds of being obese [odds ratio (OR)=2.80, 95% CI: 1.246.33].
Additionally, length of residence in the US was associated with increased odds of obesity
(OR=1.05, 95% CI: 1.021.09).
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S471

Conclusions: The present study adds to the understanding of the importance of maternal
depression as a modifiable risk factor for obesity in Latina women of childbearing-age.
Findings underscore the need for screening and interventions for maternal depressive
symptoms for Latina women of childbearing-age.

CORRESPONDING AUTHOR: Ana Cristina. Lindsay, DDS, MPH, Dr.PH, University of

Massachusetts Boston, Boston, MA, 02125; [email protected]
S472 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 5: Depression in Diverse Populations 11:30 AM-11:45 AM

COMPLEX PATIENTS' DESCRIPTIONS OF DEPRESSION SYMPTOMS, CAUSES, AND ILLNESS

INTERACTIONS: A QUALITATIVE PRIMARY CARE STUDY

Ernesto Moralez, PhD1, Main Debbi, PhD2

1
New Mexico State University, Las Cruces, NM; 2University of Colorado, Denver, CO

The recognition and treatment of depression among complex patients (patients with
concomitant illnesses) remains a significant challenge for primary care providers.
Understanding of the complex patients experience with depression is often missing during
the clinical encounter, improve patient-provider dialogue, and inform treatment options and
clinical interventions. The aim of this qualitative study was to elucidate how complex patients
describe causes and symptoms of depression and how it impacts their health. A purposive
sample of 19 complex patients receiving their primary care from Denver Health and Hospitals
were interviewed. Questions focused on how patients understand, experience, and negotiate
depression, along with their perceptions of the relationship between their mental and
physical health. Each participant had at least one chronic illness and were screened for
depression using the PHQ-9. All handling of the data was done using ATLAS.ti. The semi
structured interviews were analyzed thematically using the Iterative Process Model and
organized using constructs discussed in Kleinmans Illness Narratives. Three primary themes
emerged about the connection between physical illness and depression: 1) the burden of
physical illness is a contributing source of depression, 2) simply being a complex patient
causes depressive symptoms, and 3) the increasing severity of physical illnesses causes
feelings of distress and hopelessness. Specific examples of how patients described the
relationship included feelings of sadness because of a loss of social and daily functioning (e.g.,
loneliness) normalcy (e.g., unemployment), and hopelessness (e.g., feeling defeated because
of having illnesses with no hope for remission). Patients identified other causes including
financial troubles (e.g., health care costs associated with multiple illnesses), family stressors
(e.g., divorce, family estrangement), and bereavement/personal loss (e.g., death of a loved
one). The results indicate that patients perceive the causes and symptoms of their depression
as connected to contextual factors often not mutable in primary care treatment. Traditional
approaches for depression may not be impactful for complex patients if they perceive their
depression as a result of their health status and not as a separate coexisting illness. Given this,
providers should ask patients to describe their experience with depression to effectively
target the patients perceived sources of depression.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S473

CORRESPONDING AUTHOR: Ernesto Moralez, PhD, New Mexico State University, Las Cruces,
NM, 88005; [email protected]
S474 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 6: Advancing Understanding of PREP Use and ART Adherence in the Context of
HIV 10:45 AM-11:00 AM

CITATION AWARD WINNER

DAILY RELATIONSHIPS BETWEEN RESPONSES TO PROBES OF METHAMPHETAMINE USE AND

ANTIRETROVIRAL MEDICATION ADHERENCE

Elizabeth Pasipanodya, Ph.D.1, Jessica L. Montoya, M.S.2, Anya Umlauf, MS3, Alexandra
Rooney, B.A.4, Ben Gouaux, BA1, David J. Moore, Ph.D.5

1
University of California, San Diego, San Diego, CA; 2SDSU/UCSD Joint Doctoral Program in
Clinical Psychology, San Diego, CA; 3UCSD, San Diego, CA; 4University of California San Diego,
San Diego, CA; 5UCSD HIV Neurobehavioral Research Program, San Diego, CA

Background: Adherence to antiretroviral therapy (ART) is key to improving the health

outcomes of those living with HIV, as good ART adherence aids HIV suppression and prevents
adverse events like ART resistance. Methamphetamine (meth) use, however, poses a
significant barrier to ART adherence. Method: In an intervention study that aimed to promote
ART adherence, a text message assessing meth use was sent to 71 (22 control and 49
intervention) HIV+ meth users for 42 days. We examined the daily associations between self-
report of meth use via text message (no, yes, or non-response) and ART adherence
measured by the Medication Event Monitoring System (MEMS) by conducting concurrent and
lagged analyses. Overall, participants reported meth use on 19.2% of days and abstinence on
55.8% of days. On the remaining (25.1%) days, participants did not provide a response to the
meth use text (non-response). Daily adherence was modeled using logistic regression and
average daily adherence across the study period was modeled continuously; these within-
person and between-person effects were simultaneously estimated in the same multilevel
model. Results: When examined concurrently, no association was seen between daily ART
adherence and self-reports of meth use. Across individuals, having a higher proportion of non-
responses, relative to no responses, to the meth use texts was associated with decreases in
average daily adherence (B= -3.997, p= 0.002). Furthermore, those that responded with a
greater proportion of no responses to the meth texts had marginally greater average daily
adherence as compared to those that responded yes (B= 3.062, p= 0.050). When the daily
link between categories of meth use responses and adherence was examined using a lagged
model, the contrast between no responses to the meth use texts and a non-response
showed that a no response on one day was associated with increased odds of adherence
the next day (B= 0.303, OR= 1.354, p= 0.032), controlling for the previous days adherence. No
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S475

differences were found in this lagged daily relationship when meth use was endorsed yes as
compared to a non-response. Conclusion: In the context of a text messaging intervention
among HIV+ meth users, non-responsiveness to daily text messages about substance use may
be prognostic of next-day ART adherence. These results may have implications for
interventions targeting those at risk for ART non-adherence due to substance abuse, as a lack
of engagement may be more predictive of adherence difficulties than endorsement of
substance use.

CORRESPONDING AUTHOR: Elizabeth Pasipanodya, Ph.D., University of California, San Diego,

San Diego, CA, 92103; [email protected]
S476 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 6: Advancing Understanding of PREP Use and ART Adherence in the Context of
HIV 11:00 AM-11:15 AM

FAMILY COMMUNICATION AND HIV DISCLOSURE PREDICTING ART ADHERENCE AMONG A

VULNERABLE POPULATION LIVING WITH HIV/AIDS

Mary M. Mitchell, PhD1, Tuo-Yen Tseng, MA2, Amy Knowlton, MPH, ScD2

1
Johns Hopkins University Bloomberg School of Public Health, Nottingham, MD; 2Johns
Hopkins Bloomberg School of Public Health, Baltimore, MD

Background: Adherence to antiretrovial therapy (ART) is vital for reducing morbidity and
mortality among people living with HIV/AIDS (PLHIV). Maintaining adequate ART adherence
is especially important among disadvantaged, current or former drug using PLHIV. While
individual characteristics have been studied in relationship to ART adherence, few studies
have examined social network factors that influence adherence. We hypothesized that
negative family communication and drug use at baseline would predict poorer adherence at
12-month follow-up, while family problem solving and HIV disclosure to network members
would predict better adherence at follow-up.

Methods: We used data (n=313) from the BEACON study, which sampled primarily
disadvantaged, inner city African Americans.

Results: Using factor analysis and structural equation modeling, we found support for our
hypotheses such that negative family communication and current substance use predicted a
lower likelihood of adherence at 12-month follow-up, while controlling for baseline
adherence; and family problem solving and PLHIVs greater disclosure of their HIV serostatus
to their network members predicted an increased likelihood of adherence at follow-up.

Discussion: These results indicate the importance of family and social network factors in
targeting and developing interventions to increase ART adherence among this disadvantaged
population of PLHIV. ART adherence interventions that target social network members,
especially family members, and promote family/network communication and problem solving
may help to improve ART outcomes in this population.

CORRESPONDING AUTHOR: Mary M. Mitchell, PhD, Johns Hopkins University Bloomberg

School of Public Health, Nottingham, MD, 21236; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S477

Paper Session 6: Advancing Understanding of PREP Use and ART Adherence in the Context of
HIV 11:15 AM-11:30 AM

BEYOND TEXT MESSAGE REMINDERS: TRANSLATING BASIC NEUROSCIENCE INTO A NEW

ADHERENCE INTERVENTION FOR YOUTH LIVING WITH HIV

Amy L. Pennar, PhD1, Sylvie Naar, PhD2, Angulique Y. Outlaw, PhD2, Steven Woods, Psy.D.3,
Thomas Templin, PhD2, Karen MacDonell, PhD2

1
Wayne State University School of Medicine, Detroit, MI; 2Wayne State University, Detroit,
MI; 3University of Houston, Houston, TX

Despite inadequate medication adherence among youth living with HIV (YLH), and the
importance of adherence for virologic suppression and reductions in morbidity and mortality,
few interventions have been tested with youth. The purpose of this study was to test the
proof-of-concept of a text-delivered prospective memory (PM) intervention for taking ART
among YLH with suboptimal adherence. PM, the neurocognitive capacity to successfully form,
maintain, and execute an intention at a particular point in the future, can be used to target
basic neurocognitive processes involved in remembering to take medications.

Twenty-four, primarily male (87.5%), youth participated and were a mean age of 21.38 years
(SD = 2.32). Participants were prescribed a median of 3 ART medications. Baseline mean CD4-
1 was 552.83 cells/mm3 (SD = 286.06) and log10 transformed viral load was 2.49 (SD = .81).
Baseline measurements included a brief neuropsychological assessment, weekly phone-based
pill counts, and the Visual Analogue Scale (VAS). The intervention included a one-time in-
home session utilizing motivational interviewing and visualization to form implementation
intentions that paired taking medication with a salient cue in participants daily routines. The
intervention session was followed by a 12-week tapering text-based reminder of the cue-
intention pairing. Bi-weekly pill-counts and the VAS continued through the intervention and
follow-up periods. Viral load and hair sample assays were obtained at baseline, 18 and 24
weeks.

From baseline to 18-weeks, 46.7% of youth evidenced a log drop 1 in viral load (0.5 drop
considered clinically significant). By 24-weeks, 73.3% of youth evidenced a log drop 1. Data
from the hair assays, pill counts and VAS were inconclusive. Rates of metabolizing medications
and ceiling effects in reports of adherence may account for these findings, respectively.
Qualitative analysis of exit-interviews supported the feasibility and acceptability of the
S478 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

intervention via self-reports of acceptance of frequency of text messaging, feeling supported

through their involvement, and enhancement of their ability to develop a medication routine
and utilize visual cues to remember to take medications.

Targeting PM via brief interventions holds promise to improve health outcomes among YLH
and reduce transmission via viral load suppression. Findings suggest the intervention is worthy
of a controlled pilot trial as the next step while finding alternative adherence measures.

CORRESPONDING AUTHOR: Amy L. Pennar, PhD, Wayne State University School of Medicine,
Detroit, MI, 48202; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S479

Paper Session 6: Advancing Understanding of PREP Use and ART Adherence in the Context of
HIV 11:30 AM-11:45 AM

CORRELATES OF WILLINGNESS TO USE PRE-EXPOSURE PROPHYLAXIS (PREP) AMONG HIGH-

RISK DRUG USERS IN TREATMENT

Roman Shrestha, MPH1, Frederick Altice, PhD2, Pramila Karki, RN3, Michael Copenhaver, PhD3

1
University of Connecticut Health Center, Willimantic, CT; 2Yale University, New Haven, CT;
3
University of Connecticut, Storrs, CT

Introduction: Although people who use drugs (PWUDs) are one of the key risk populations
who could benefit from the use of pre-exposure prophylaxis (PrEP), to date, very little if any
attention has been given to incorporating PrEP into HIV prevention approaches targeting this
underserved group. This study aimed to fill a conspicuous gap in literature surrounding the
knowledge base and acceptability of PrEP use among high-risk PWUDs in a common
treatment setting.

Methods: A cross-sectional study was conducted among 400 HIV-negative individuals who
reported drug- and/or sex-related risk behaviors and were enrolled in methadone
maintenance program in New Haven, CT. Univariate and multivariate logistic regression
analyses were used to assess self-reported demographic and behavioral correlates of
willingness to use PrEP and the likelihood of drug- and sex-related risk reduction while on
PrEP among high-risk PWUDs in treatment.

Results: Of the total participants, only 18% reported to have heard of PrEP and 62.7%
reported that they would be willing to use PrEP. Willingness to use PrEP was independently
associated with neurocognitive impairment status (aOR = 2.143, p = .013), and higher
perceived risk for HIV infection (aOR = 3.811, p < .0001). Among participants who were willing
to use PrEP, only 12.5% and 28.2% indicated they would always use condom and not share
needles or works while on PrEP, respectively. The likelihood of consistent condom use was
associated with higher income (aOR = 5.743, p = .041) and always using condoms with casual
partners (aOR = 6.597, p = .001) and not sharing or needles or works was associated with age
(aOR = .952, p = .043) and high perceived risk for HIV infection (aOR = .191, p = .019).

Conclusions: The results show that the majority of participants were unware of PrEP;
however, after being informed about the basics of the medication, the majority indicated that
they would be willing to take PrEP as part of HIV prevention. Our findings suggest that,
because the use of PrEP has the potential to lead to risk compensation, the next generation of
interventionists must carefully incorporate intervention content that simultaneously focuses
on both HIV risk reduction and PrEP adherence over time.
S480 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Roman Shrestha, MPH, University of Connecticut Health Center,

Willimantic, CT, 06226; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S481

Paper Session 7: Health Behaviors and Psychological States in Cardiac Rehabilitation 10:45
AM-11:00 AM

INTERGENERATIONAL EFFECTS IN CARDIAC REHABILITATION PATIENTS

Stephanie Leung, Ph.D.1, Joseph Zander, Ph.D.2, Lawrence C. Perlmuter, Ph.D.3

1
VA Connecticut Healthcare System, New Haven, CT; 2Advocate Medical Group, Park Ridge, IL;
3
Rosalind Franklin University of Medicine and Science, North Chicago, IL

Intergenerational transmission of illness behaviors from parents to children and health

outcomes have been explored in individuals with chronic medical conditions such as
cardiovascular disease. Research has demonstrated a relationship between maladaptive
illness behaviorsi.e., preoccupation with illness or denial of diseaseand poor
cardiovascular outcomes, such as increased severity ofcardiac disease. Illness behaviors may
be transmitted from parents to offspring via behavior modeling. Parental modeling of illness
behavior has also been found to impact healthcare utilization (e.g., frequency of outpatient
physician visits). In this study, intergenerational transmission of illness behavior was evaluated
in a sample of outpatient cardiac rehabilitation patients. Adult participants illness behavior
was hypothesized to mediate the relationship between perceived parental illness behavior
and participants cardiac disease severity. Additionally, participants illness behavior was
hypothesized to mediate the relationship between perceived parental illness behavior and
participants healthcare utilization. The primary outcome variables were cardiac disease
severity, measured by left ventricular ejection fraction, and healthcare utilization, which was
assessed by number of outpatient physician visits and number of emergency room visits.
Cardiac rehabilitation patients (n=111; 81 males, 30 females) completed measures regarding
their illness behaviors and recollection of their parents illness behaviors. Mediation analyses
yielded significant relationships between perceived parental illness behavior and participants
illness behavior: recollection of perceived maternal illness behavior was significantly
associated with participants illness behavior (coefficient=.33, SE=.07, p < .001), and
recollection of perceived paternal illness behavior significantly predicted participants illness
behavior (coefficient=.40, SE=.08, p < .001). No evidence was found for indirect effects or
direct effects on outcome variables. When maternal illness behavior served as the predictor,
participant illness behavior was significantly associated with higher frequency of ER visits
(coefficient=.18, SE=.09, p=.05), suggesting that mothers behavior may play a role in
predicting ER presentations. These results implicate the role of parental modeling in illness
behavior and warrant further research on intergenerational relationships in cardiac-related
outcomes and healthcare utilization.
S482 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Stephanie Leung, Ph.D., VA Connecticut Healthcare System, New

Haven, CT, 06511; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S483

Paper Session 7: Health Behaviors and Psychological States in Cardiac Rehabilitation 11:00
AM-11:15 AM

A QUALITATIVE STUDY OF VALUES AND MOTIVATIONAL FACTORS IN CARDIAC AND

PULMONARY REHABILITATION PATIENTS

Jordan Ellis, MA, Matthew C. Whited, PhD, Licensed Psychologist, John T. Freeman, B.A.,
Ansley Taylor Corson, M.A.

East Carolina University, Greenville, NC

Participation in cardiovascular and pulmonary rehabilitation programs (CVPR) can lead to

improved functional abilities, improved quality of life, and reduction of lifestyle related risks
for those who have chronic cardiac or pulmonary conditions. Unfortunately, attendance and
adherence to these programs remain suboptimal. Third wave behavioral therapies, including
Behavioral Activation Treatment for Depression and Acceptance and Commitment Therapy,
have emphasized the importance of life value identification as a guide for goal setting and
behavior change for both psychological and physical health conditions. Individuals who
choose to engage in behaviors that align with their life values are thought to be intrinsically
reinforced. Thus, value identification and clarification could serve as a simple and
individualized intervention strategy for CVPR patients to establish reason and meaning for
engaging in lifestyle behavior change, which could improve attendance and adherence
outcomes. The purpose of the following qualitative study was to interview CVPR patients
about their own life values and motivating factors related to healthy behavior changes.

Thirty cardiac or pulmonary patients were recruited from a CVPR program in the Southeastern
US, participated in a 20 to 30-minute interview, and completed a 10-item values engagement
measure. Qualitative interview questions elicited conversation about values and motivating
factors that guide the patients engagement in the lifestyle change behaviors (e.g., exercise
and dietary change) monitored by the CVPR program. The interviews were audio recorded
and transcribed, and then coded thematically by the research team in NVivo.

Results showed that participants identified a wide range of values (17 in total) related to
program engagement. It was predicted that most participants would directly identify their
health improvement/maintenance as a value, given that this is the focus of goal setting at
CVPR; however, only half of the patients did so. The most frequently endorsed life values (N=
17 total values) included, being active (70.0%), family (63.3%), and independence (56.7%). The
interviews indicated that while patients make lifestyle changes in the program to improve
their physical functioning, there are often other values that primarily guide their choice to
engage in, and maintain, lifestyle behaviors. Results also indicated that approximately half of
S484 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

the patients had experienced a value shift, such that the most prominent values driving
lifestyle behavior change shifted to values more directly associated with the things they enjoy
in life (e.g. being active with family members).

Life values can serve as a powerful guide for individual behavior change, and the present
study suggests that the piloting of brief values interventions early in CVPR treatment is
warranted, and have the potential to improve patient outcomes.

CORRESPONDING AUTHOR: Jordan Ellis, MA, East Carolina University, Greenville, NC, 27858;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S485

Paper Session 7: Health Behaviors and Psychological States in Cardiac Rehabilitation 11:15
AM-11:30 AM

COPING, OPTIMISM, AND PHYSIOLOGICAL OUTCOMES IN CARDIAC REHABILITATION

Victoria Bangieva, PhD1, Lawrence C. Perlmuter, Ph.D.2, Joseph Zander, Ph.D.3

1
Stanford University School of Medicine, San Francisco, CA; 2Rosalind Franklin University of
Medicine and Science, North Chicago, IL; 3Advocate Medical Group, Park Ridge, IL

Coping, Optimism, and Health-Related Outcomes in Cardiac Rehabilitation

Among Cardiac Rehabilitation (Rehab) patients the relationships between psychosocial factors
and disease onset and progression are varied and complex. Coping strategies and optimism
are among the psychosocial variables of interest in this population. The present study sought
to assess relationships among four coping strategies (Distraction, Instrumental, Palliative, and
Emotion-focused) in relation to physiological outcomes (metabolic equivalent task (MET),
blood pressure, and BMI) in Cardiac Rehab patients. Additionally, the impact of optimism in
these relationships was explored through moderation analyses. Physiological outcome data
from 120 Cardiac Rehab patients was gathered at the start and completion of the 90-day
program. Participants also completed psychosocial questionnaires on coping strategies,
optimism, and depression at the start of the program. Results showed that Palliative coping
has a deleterious effect on METs in Cardiac Rehab patients ( = -.21, t(117) = -2.07, p = .04),
while Instrumental coping is associated with improvements in METs ( = .21, t(117) = 2.35, p =
.02). Thus, Palliative and Instrumental coping have a distinctly different impact on health-
related outcomes in Cardiac Rehab patients; passive behaviors are damaging to Cardiac Rehab
patients while problem-focused behaviors are associated with better outcomes. Also,
Emotional Preoccupation coping was predictive of the change in BMI ( = .27, t(117) = 2.63, p
= .01), which suggests that participants who reported more emotion-focused coping behaviors
gained weight during enrollment in Cardiac Rehab. Further, optimism was especially
important in the relationship between Palliative coping and weight-related outcomes, where
as optimism increased the relationship between Palliative coping and weight gain was
weakened (b = -.00, t = -2.14, p < .05). Therefore, low optimism appears to be a risk factor for
individuals engaging in passive coping behaviors. Consequently, this study has implications for
cardiac patient care in the assessment and treatment of cardiovascular disease patients.
S486 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Victoria Bangieva, PhD, Stanford University School of Medicine,

San Francisco, CA, 94108; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S487

Paper Session 7: Health Behaviors and Psychological States in Cardiac Rehabilitation 11:30
AM-11:45 AM

THE EFFECTS OF VARYING LEVELS OF OPTIMISM ON PHYSICAL ACTIVITY AMONG CARDIAC

PATIENTS

Rachel A. Millstein, PhD, MHS, Christopher Celano, MD, Eleanor Beale, BA, Jeff Huffman, MD

Massachusetts General Hospital, Boston, MA

Background: Over 1 million Americans suffer an acute coronary syndrome (ACS: myocardial
infarction or unstable angina) each year. Positive psychological constructs may improve
cardiac-related outcomes. Optimism has been associated with adherence to health behaviors,
like physical activity, that can lead to improved functioning and cardiac health. We sought to
examine the role of different levels of optimism on physical activity. Methods: Data from 156
patients was collected at 2 weeks and 6 months following hospitalization for an ACS as part of
an observational study. We measured optimism using the 6-item Life Orientation Test-
Revised (LOT-R) and divided the scores into quartiles. Objectively measured physical activity
(average steps per day) was collected using the Fitlinxx Pebble uniaxial accelerometer.
ANOVA with post-hoc tests and t-tests were used to explore univariate associations between
levels of baseline optimism and physical activity 6 months post-ACS. A preliminary Latent
Profile Analysis (LPA) was run to determine unique groups of patients with respect to their
optimism scores. Results: LOT-R scores were not normally distributed and ranged from 7-30
(mean=23.72, SD=5.61). Quartiles were created: Q1 (scores 7-19), Q2 (scores 20-24), Q3
(scores 25-27), Q4 (scores 28-30). Participants in the highest quartile of optimism scores had
the highest mean steps per day, and those in the second quartile had the lowest mean steps
per day (Q1: 5190.56 steps/day, Q2: 4240.79, Q3: 4806.27, Q4: 5987.13). Omnibus ANOVA
showed that LOT-R quartiles were significantly associated with steps (F (148)=2.78, p=.043).
Post-hoc tests showed that Q4 mean steps were significantly higher than Q2 (Bonferroni
comparison p=.039). Q4 mean steps were also significantly higher than Qs1-3 combined (t
(150)=-2.51, p=.013). Initial LPA results suggest that there are three distinct latent classes of
optimism scores. Conclusions: Cardiac patients with the highest optimism scores at 2-weeks
post-ACS were most physically active 6 months later, compared to all other quartiles. These
findings support prior literature indicating the importance of optimism for cardiac health and
preventive behaviors. These findings suggest that interventions to promote optimism may
have promise for helping cardiac patients improve physical activity after hospitalization. Next
steps will include using LPA results to further define optimism profiles and understand their
relationships with physical activity.
S488 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Rachel A. Millstein, PhD, MHS, Massachusetts General Hospital,

Boston, MA, 02114; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S489

Thursday

March 30, 2017

11:45 AM-12:45 PM
S490 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F001 11:45 AM-12:45 PM

HEALS: A FAITH-BASED HYPERTENSION CONTROL PROGRAM FOR AFRICAN AMERICAN

COMMUNITIES: RESULTS, CHALLENGES AND LESSONS LEARNED

Sunita Dodani, MD, MSc, PhD1, Claudia Sealey-Potts, PhD2, Sahel Arora, MD3

1
Mayo Clinic, Jacksonville, FL; 2University of North Florida, Jacksonville, FL; 3University of
Florida, Jacksonville, FL

Background. Health disparities related to cardiovascular diseases have remained higher in

African Americans (AAs) than many other populations. In this study,we examined the
feasibility and efficacy of a multi-level, faith-based hypertension control and prevention
program, Healthy Eating And Living Spiritually (HEALS).

Design. In a pre-post single arm study design with a community based participatory research
(CBPR) approach, we developed and implemented HEALS in a low-income urban AA church.
Study participants were high-risk adult AA church members with known or newly diagnosed
hypertension or pre-hypertension. The HEALS intervention included 3 months of weekly
sessions (1 session per week for 12 weeks) followed by a 6 month maintenance period that
included monthly booster sessions. The program was delivered by trained lay church
members and had group, individual, and church level components. Fruits & Vegetables were
also provided for the first 5 months of the program. Data were analyzed using repeated
measures ANOVA at baseline, 3 months, 9 months, and monthly follow up blood pressure
readings were also recorded. The primary outcome of interest was change in systolic blood
pressure (SBP).

Results. Of the 52 church participants screened, 51 met the eligibility criteria and 37 provided
written informed consent (n=37). A retention rate of 92% (34/37) and 73% (28/37) was
observed at 3 and 9 months post-baseline, respectively. A SBP reduction of 12.6 mm Hg
(p=0.0194) was observed at 3 months and 1.7 mm Hg (p=0.9998) at 9 months. The SBP
reduction was greatest at 5 months (13.2 mm Hg from baseline; p=0.04)
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S491

Conclusion and Implications. Our study presents a feasible and efficacious model of a
PREMIER /DASH modified CBPR program for hypertension control and prevention in AA
communities, a group disproportionally affected by health disparities. More research is
needed to test this intervention experimentally and to barriers to sustaining the lowered SBP.

CORRESPONDING AUTHOR: Sunita Dodani, MD, MSc, PhD, Mayo Clinic, Jacksonville, FL,
32207; [email protected]
S492 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F002 11:45 AM-12:45 PM

UNITED MOTHERS: DEVELOPING AN EFFECTIVE PRENATAL INTERVENTION

Paola Ilescas, Student1, Raquel C. Good, MA (in progress)2, Kimberly D'Anna-Hernandez, PhD1

1
CSU San Marcos, San Marcos, CA; 2California State University San Marcos, El Cajon, CA

Gestational diabetes mellitus (GDM) is a growing public health concern that has detrimental
effects. Affected women have a higher risk of developing type 2 diabetes. Maternal obesity
and a diet of high fat, correlated with GDM, increases the susceptibility of the offspring
developing health complications, including diabetes and obesity, and mental health and
behavioral disorders. There is a health disparity found among pregnant minority women of
which Mexican-American seem to be most affected. Between 1990 to 2001, a 96% increase of
GDM cases was observed among Mexican mothers, the most rapid growth of GDM in any
ethnic group. It is imperative to serve this population with a culturally sensitive prenatal
program that aims to reduce these health disparities in GDM.

We are engaging in community-engaged research with a local community clinic for a pilot
prenatal program tailored to the needs of the community with the goal of reducing
gestational weight gain. First, six qualitative was gathered with clinic providers and patients
data to understand those needs. The three clinic staff focus groups included 3 health
educators, 3 medical assistants and two providers. The patient focused groups included
pregnant women who were at risk for obesity-related complications during birth. One woman
had previously participated in a prenatal program, 3 women were English-speaking and 5
women were Spanish speakers. The average age was 30 years and average BMI was 34.4,
obese by NIH standards. Focus groups were led by research assistants with questions related
to experiences in the clinic and feedback for future prenatal programs. After analyzing the
discussions, the themes were patient-provider interactions, negotiating cultural barriers and
patient empowerment. For pregnant women, we discovered waiting times and continuity of
care were the main barriers to receive adequate prenatal care. Others were lack of emotional
support when diagnosed with GDM and lack of access to more information, control over own
medical care and cultural competence. Clinic providers expressed concerns over the lack of
motivation, participation, empowerment and networking among pregnant women. After
discussion with clinic leadership, we are proposing a culturally-competent group medical visit
prenatal program. This program will be adapted for GDM and focus on continuity of care,
patient empowerment amongst providers and patients, and increasing perinatal knowledge
related to GDM. It will also provide psychoeducation related to fetal brain development and
the maternal-fetal programming hypothesis as it relates to GDM - a need discovered by the
focus groups. This will be one of the first prenatal group visit models to incorporate GDM,
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S493

culture and psychoeducation of neuroscience early in life to decrease gestational weight gain
and promote healthy perinatal outcomes in Mexican-American pregnant women to reduce
health disparities.

CORRESPONDING AUTHOR: Paola Ilescas, Student, CSU San Marcos, San Marcos, CA, 92069;
[email protected]
S494 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F003 11:45 AM-12:45 PM

ACCESS TO PSYCHOSOCIAL SERVICES AMONG LATINO CANCER PATIENTS: MENTAL HEALTH

PROVIDERS' PERSPECTIVES

Emely Perez Ramirez, BA1, Emely Perez Ramirez, B.A.2, Rosario Costas-Muiz, PhD3, Olga
Garduo-Ortega, B.A.3, Xiomara Singh, MD4, Javier Gonzlez, MFA4, Normarie Torres-Blasco,
MS5, Francesca Gany, MD, MS6

1
University of Puerto Rico at Rio Piedras Campus, Penuelas, N/A, Puerto Rico; 2University of
Puerto Rico/Memorial Sloan Kettering Cancer Center, Penuelas, N/A, Puerto Rico; 3Memorial
Sloan Kettering Cancer Center, New York, NY; 4Memorial Sloan-Kettering Cancer Center, New
York, NY; 5Ponce Health Science University, Yauco, N/A, Puerto Rico; 6Immigrant Health and
Cancer Disparities Center, Memorial Sloan Kettering Cancer Center, New York, NY

Title: Access to psychosocial services among Latino cancer patients: A Mental Health
Providers perspective

Background/Purpose:
Research suggests that multiple structural, organizational, and psychological factors
negatively impact Latinos access to psychosocial services utilization (PSU) (1-4). This study
aimed to determine the most common barriers identified by Mental Health Providers (MHP)
from the United States (US), Latin America (LA) and Puerto Rico (PR) for PSU among Latino
cancer patients (LCP) and to provide recommendations to improve psychosocial services (PS)
access for Latinos in the US, LA, and PR.

Methods:
MHPs were invited to complete an online survey about barriers to PSU among LCP.
Descriptive and chi-squares were used to identify common barriers and compare barriers by
location of practice.

Results:
A total of 64 MHP participated. The most common barriers reported, included lack of:
availability of MHP who speak Spanish, knowledge about how PS can help them, knowledge
about where to receive PS, availability of MHP who understand patients culture and health
insurance, and also the belief that PS are too expensive. In analysis by location of practice (US
vs. LA) the lack of availability of MHP who speak Spanish and the inconvenience of counseling
centers location for cancer patients were significantly greater barriers in the US.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S495

Conclusions:
In sum, the most common barriers are access to (e.g. lack of availability of providers with
cultural sensitiveness and language skills) and education about PS (e.g. lack of knowledge
about where to receive PS or how PS can help). Only linguistic and location barriers were
more common in US than LA. The majority of the barriers were encountered across locations
with similar proportions.

Implications:
These results support the need for more culturally and linguistically competent MHP to treat LCP
in the US. Our findings demonstrate the importance of developing culturally and linguistically
adapted PS for LCP. To promote the utilization and access to PS in Latinos patient activation and
the integrated Behavioral Health Care model can be employed by MHP.

Psychosocial Services Utilization (PSU)

Mental Health Providers (MHP)
United States (US)
Latin America (LA)
Puerto Rico (PR)
Psychosocial Services (PS)
Latino Cancer Patients (LCP)

CORRESPONDING AUTHOR: Emely Perez Ramirez, BA, University of Puerto Rico at Rio Piedras
Campus, Penuelas, N/A, 00624; [email protected]
S496 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F004 11:45 AM-12:45 PM

ASSOCIATION BETWEEN SEVERITY OF PRETREATMENT SYMPTOMS AND COMPLIANCE WITH

RADIATION THERAPY IN HEAD AND NECK CANCER PATIENTS

Joseph C. Edwards, MD Candidate1, John Schmidt, PhD; MS2, Julie Bauman, MD, MPH3,
Dwight E. Heron, MD; FACRO; FACR4, Robert L. Ferris, MD; PhD5, Dana Bovbjerg, PhD6

1
University of Pittsburgh School of Medicine, Pittsburgh, PA; 2Navy Medicine Professional
Development Center Walter Reed National Military Medical Center, Bethesda, MD;
3
University of Arizona College of Medicine - Tucson, Department of Medicine, Tucson, AZ;
4
UPMC Shadyside; University of Pittsburgh School of Medicine, Pittsburgh, PA; 5University of
Pittsburgh Cancer Institute; University of Pittsburgh School of Medicine, Pittsburgh, PA;
6
University of Pittsburgh Cancer Institute, Pittsburgh, PA

Introduction: Radiation therapy (RT) is one of the primary modes of treatment for head and
neck cancer (HNC). Nonadherence to RT treatment (received dose < total planned dose or > 1
week break in treatment) has been shown to be associated with inferior locoregional control
and survival. Although previous studies have reported that higher pretreatment levels of
several emotional and physical symptoms, and poorer quality of life, predict inferior survival
in this patient population, little attention has been paid to possible effects on treatment
adherence. The goal of this prospective pilot study was to explore differences in pretreatment
symptoms between patients who were subsequently adherent or not adherent to RT, as a
first critical test of our overarching hypothesis that such associations may mediate
relationships between pretreatment patient-centered variables and survival rates.

Methods: HNC patients scheduled for RT were recruited, signed informed consent, and
completed self-report measures prior to RT. Validated questionnaires assessed pre-RT
variables previously reported to be related to inferior clinical outcomes including levels of
pain, fatigue, sleep, anxiety, and depression, as well as demographic data and quality of life.
Clinical data were collected via chart review, and RT adherence defined as above.
Relationships between pre-RT variables and RT adherence were explored with Mann-Whitney
U tests in a sample of 15 patients (3 women, mean age 57 yr).

Results: Four of the 15 participants (27%) were nonadherent to RT. There were no significant
differences in demographic or tumor characteristics between the two groups. Significant
differences (p < 0.05) were found in baseline levels of: fatigue, pain, appetite loss, insomnia,
sensory problems, physical function, role function, cognitive function, and pain self-efficacy.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S497

Discussion: Prior literature has shown associations between patients pretreatment symptom
levels and HNC outcomes including survival, but few studies have included RT adherence in
their analyses. Of the 11 pre-RT variables previously reported to be associated with poor
clinical outcomes, 9 were found in this study to be significantly associated with nonadherence
to RT. Although our results are correlative in a small sample, they suggest the importance of
future research to evaluate RT compliance as a mediating factor in studies of the relationships
between pre-RT patient experience and clinical outcomes including regional control of HNC
and survival.

CORRESPONDING AUTHOR: Joseph C. Edwards, MD Candidate, University of Pittsburgh School

of Medicine, Pittsburgh, PA, 15206; [email protected]
S498 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F005 11:45 AM-12:45 PM

FACTORS THAT INFLUENCE PROSTATE CANCER SCREENING IN HISPANIC MEN: A SYSTEMATIC

REVIEW

Jacob Martinez, BSN, RN

University of Miami School of Nursing and Health Studies- The University of Texas at El Paso
School of Nursing, El Paso, TX

In a time when the Hispanic population continues to grow and prostate cancer screening
recommendations have changed, it becomes important to systematically review the available
literature for factors that influence men to screen. This article employed the socioecological
model (SEM) to identify factors that influence prostate screening in Hispanic men at the
intrapersonal, interpersonal, and community/societal levels. Academic Search Premier,
CINAHL Plus, PsycINFO, PubMed, MEDLINE, and Google Scholar were queried without a
timeframe; 21 articles met inclusion criteria and were selected. At the intrapersonal-level low
health literacy, machismo, and other past cancer screening experiences were identified to
influence prostate cancer screening in Hispanic men. At the interpersonal level, both the
patient-physician relationship and the involvement of immediate family influenced screening.
Lastly, at the community/societal level the following were identified as factors that influence
prostate cancer screening: (a) access to care, (b) lack of cultural competence & tailoring, and
(c) providers not engaging clients in prostate cancer screening discussions. Cultural norms,
such as machismo, and socio-economic depravation were factors that manifested through all
levels of the SEM. The results from this review will help healthcare professionals tailor their
message when advocating and communicating the importance of screening in this
underserved population.

CORRESPONDING AUTHOR: Jacob Martinez, BSN, RN, University of Miami School of Nursing
and Health Studies- The University of Texas at El Paso School of Nursing, El Paso, TX, 79936;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S499

F006 11:45 AM-12:45 PM

NURSES READINESS TO ADVISE BREAST CANCER SURVIVORS WITH COGNITIVE PROBLEMS: A

META-SYNTHESIS STUDY

Sook Jung Kang, PhD, FNP-C

Ewha Womans University, Seoul, N/A, Republic of Korea

Purpose: After the completion of chemotherapy, 16 to 50% of breast cancer survivors

experience cognitive problems. Since this health problems hampers survivors back-to-normal
life, it needs to be intervened based on their needs. The purpose of this study was to
synthesize the result of qualitative studies related to cognitive problems of breast cancer
survivors by conducting meta-synthesis review.

Methods: Literature search was conducted from 2007 to 2016 published in nursing,
education, preventive medicine, and psychology journals using MEDLINE, CINAHL, EMBASE,
PsycINFO, and Pubmed databases. Inclusion criteria were: 1) research focusing on the aspects
of cognitive dysfunction in breast cancer survivors among chemotherapy; 2) papers published
by peer reviewed journals; 3) research involving qualitative or mixed-method methodologies.
Meta-synthesis was guided by Sandelowski & Barroso (2007), which is comprised of meta-
summary and meta-synthesis.

Result: From past 10 years of search, 125 publications were reviewed and 9 articles were
finally selected. After the quality assessment using Critical Appraisal Skills Program (CASP),
articles were reviewed and analyzed by two separate investigators. For the meta-summary,
four themes were revealed: 1) experience of cognitive decline; 2) impact on life after finishing
the treatments; 3) coping strategies used to overcome the difficulties; and 4) lack of care and
education from health care providers. One additional theme was found by meta-synthesis,
which is emotional struggle through cognitive decline.

Conclusion: According to the review, breast cancer survivors with cognitive impairments are
struggling with their daily life and they feel they need more care and education from nurses
regarding this issue. Although there are some guidelines to be utilized when counseling
survivors with cognitive decline, more detailed and patient-tailored educational resources
need to be organized and disseminated so that health care providers including nurses can
advise breast cancer survivors with cognitive problems.

CORRESPONDING AUTHOR: Sook Jung Kang, PhD, FNP-C, Ewha Womans University, Seoul,
N/A, 03760; [email protected]
S500 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F007 11:45 AM-12:45 PM

PROFESSIONAL IDENTITY AND TRAINING NEEDS OF LAY PATIENT NAVIGATORS

mariana Lara, Associate in Psychology

San Diego State University, san ysidro, CA

PROFESSIONAL IDENTITY AND TRAINING NEEDS OF LAY PATIENT NAVIGATORS

Purpose: Lay patient navigators work with people to navigate the health care system and
access appropriate social and financial services. The objective of this qualitative study was to
describe lay health workers perceptions of their professional identity and suggestions for
patient navigator training.

Methods: Using a standardized focus group guide, 8 focus groups were conducted with lay
health workers in four U.S. states (n=51; 90% female; 75% 40 years or older). Participants
were recruited through email and word of mouth. All focus groups were audio recorded,
transcribed verbatim, and analyzed using content analysis.

Results: Two specific themes emerged from the focus groups related to lay health workers,
the perceptions of their professional identity and needs for training. Focus group participants
expressed confusion regarding the qualifications one must possess to be a lay patient
navigator. Most participants identified as being a health promoter rather than being a patient
navigator. Lay health workers provided many suggestions regarding topics of training that
should be included in a lay patient navigation curricula, including: 1) physical and mental
health ; 2) community and organizational resources that can facilitate access to care; and 3)
communication and people skills, especially in assessment and intervention.

Conclusions: Data from focus groups indicated that lay health workers may not identify as
being a patient navigator, despite having qualification of a lay patient navigator. Focus groups
also indicated that there is substantial need for training of lay patient navigators and that
training is needed across a wide range of topics.

CORRESPONDING AUTHOR: mariana Lara, Associate in Psychology, San Diego State University,
san ysidro, CA, 92173; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S501

F008 11:45 AM-12:45 PM

STRENGTH OF CLINICIAN RECOMMENDATION OF HPV VACCINATION AND VACCINE DELIVERY

RATES

Lila J. Finney-Rutten, PhD, MPH1, Jennifer L. St. Sauver, MPH/PhD1, Timothy Beebe, PhD2,
Patrick M. Wilson, MPH1, Debra J. Jacobson, MS1, Chun Fan, MS3, Carmen Radecki Breitkopf,
PhD1, Susan Vadaparampil, PhD MPH4, Robert M. Jacobson, MD1

1
Mayo Clinic, Rochester, MN; 2University of MN, Minneapolis, MN; 3MAYO CLINIC,
ROCHESTER, MN; 4Moffitt Cancer Center, Tampa, FL

PURPOSE: We tested the hypothesis that strength of clinician recommendation of the human
papillomavirus (HPV) vaccination would be associated with rates of vaccine delivery at the
practice-site level.

METHODS: We conducted a survey of primary care clinicians (n=280) in a 27-county

geographic region from October 2015 through January 2016 to assess clinician behaviors
regarding HPV vaccination. The overall response rate for the survey was 41.0%, with 52
clinical sites represented. Additionally, we used the clinical data linkage infrastructure of the
Rochester Epidemiology Project to ascertain actual HPV vaccination rates among patients
residing in the same 27-county geographic region at the site level. We examined associations
of clinician survey data on strength of recommendation for HPV vaccination for male and
female patients aged 9 to 18 years with rates of initiation (1 dose) and completion (3 doses
the recommendation during our study period) among patients age 9 to 18 years (n=68,272) at
the clinical site level.

RESULTS: Among clinicians who completed the survey, 71.1% indicated that they strongly
recommend the HPV vaccine to their female patients while 58.2% strongly recommend to
their male patients. The percentage of clinicians who strongly recommend the HPV vaccine
was significantly higher among those practicing in pediatrics and board certified in pediatrics
compared to those practicing in and certified in other primary care settings. Results of Poisson
regression models were consistent with our hypothesis. Strength of recommendation of the
HPV vaccination was associated with higher rates of HPV vaccination initiation and
completion. Higher rates of initiation [Incident Rate Ratio (IRR)=1.15; 95% CI (1.08-1.23)] and
completion [IRR=1.12; CI (1.02, 1.23)] were associated with stronger recommendations for
vaccination among females; similar associations were observed for initiation [IRR=1.29; CI
(1.21,1.38)] and completion [IRR=1.19; 95% CI (1.09, 1.31)] among males. All models were
S502 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

adjusted for the following site-level characteristics: percent white, percent female, percent
ages 9-13, and percent with government insurance or self-pay.

CONCLUSIONS: Clinicians more frequently strongly recommended the HPV vaccine to their
female patients than to their male patients. Strength of clinician recommendation of the HPV
vaccine is associated with HPV vaccine delivery rates-initiation and completion. These data
can guide efforts to improve HPV vaccine delivery in clinical settings.

CORRESPONDING AUTHOR: Lila J. Finney-Rutten, PhD, MPH, Mayo Clinic, Rochester, MN,
55905; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S503

F009 11:45 AM-12:45 PM

ANTIDEPRESSANT USAGE AT TIME OF STROKE IS RELATED TO BETTER MENTAL FUNCTIONING

Teresa M. Damush, Ph.D.1, Lincoln Flossy, MPH2, Chandan K. Saha, PhD3, James Slaven, MS4,
Laura Myers, Ph.D.5, Jason Mackey, MD6, Linda Williams, MD7

1
Indiana University School of Medicine/Roudebush VAMC PRISM QUERI Center, Fishers, IN;
2
Roudebush VAMC, Indianapolis, IN; 3Indiana University School of Medicine, Indianapolis, IN;
4
Indiana University School of Medicine, Dept of Biostatistics, Indianapolis, IN; 5Roudebush
VAMC/PRIS-M QUERI Center, Indianapolis, IN; 6Indiana University School of Medicine, Dept of
Neurology, Indianapolis, IN; 7Roudebush VAMC/Indiana University School of Medicine,
Indianapolis, IN

Objective: An acute stroke may serve as a major life event that disrupts health-related quality
of life. Post stroke depressant is prevalent and related to poorer outcomes. We examined the
potential buffering relationship of antidepressants at time of stroke on mental health
functioning among patients discharged from two health care systems: Veterans Health
Administration (VHA) and a Joint Commission Stroke Center.

Methods: We enrolled 258 participants into a randomized controlled trial to evaluate a stroke
self-management program which targeted functional recovery and risk factor management.
All participants completed written consent and a baseline assessment which included
demographics, anxiety (GAD), depressive symptoms (PHQ9), antidepressant usage, stroke
specific, health-related quality of life (SSQoL) and pain (BPI). We conducted multivariate
analyses to evaluate the relationship of antidepressant usage with baseline mental
functioning.

Results: We enrolled our sample, on average, within 81.5 days after hospital discharge for
stroke/TIA. Our sample included 19% women with an average age of 61.7 (10.8) years. The
average NIH stroke scale score was 3.0 indicating minor stroke and did not differ among those
with and without antidepressants. Demographics were not significantly related to SSQoL in
the multivariate models. Among our cohort, 201 (78%) reported baseline usage. Among those
using antidepressants, a significantly less proportion (26.4%) reported moderate to severe
depressive symptoms compared to those who reported no usage (53.7%) (p

Conclusions: Stroke survivors reporting antidepressant usage reported moderate to severe

depression symptoms at half the rate of stroke survivors without antidepressant usage.
Antidepressant usage at the time of a stroke event may buffer the development of poorer
S504 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

mental health functioning. Post stroke depression is prevalent and associated with morbidity
and mortality. Therefore, post stroke programs which target depression at the early onset of
survivorship may enhance stroke specific, quality of life.

CORRESPONDING AUTHOR: Teresa M. Damush, Ph.D., Indiana University School of

Medicine/Roudebush VAMC PRISM QUERI Center, Fishers, IN, 46040; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S505

F010 11:45 AM-12:45 PM

EFFECTS OF SLEEP, DIET, AND PHYSICAL ACTIVITY ON CARDIOVASCULAR DISEASE IN THE

UNITED STATES: AN AGENT-BASED MODEL SIMULATION

Azizi Seixas, PhD, MS1, Ferdinand Zizi, MBA2, Girardin Jean-Louis, PhD2

1
New York University School of Medicine, New York, NY; 2NYU School of Medicine, NY, NY

INTRODUCTION: Cardiovascular disease (CVD) is the number one reason for health-related
morbidity and mortality in the United States. To reduce the burden of CVD, several behavioral
prevention strategies (eliminate smoking, healthy diet, and increased physical activity) have
been implemented. Unfortunately, these efforts have only yielded modest improvements,
and in some groups (racial/ethnic minorities) the burden of CVD is mounting. Recent evidence
that sleep is associated with increased CVD risk, acknowledges sleep as a novel and potentially
impactful target to reduce CVD burden. The current study explored the long-term impact
reducing the prevalence of insufficient sleep can have on the burden of CVD (diabetes, stroke,
and myocardial infarction) in the US, compared to healthy diet (eating fruits and vegetables)
and moderate physical activity.

METHOD: Using a representative sample of 100,000 hypothetical participants who do not

have sufficient sleep in an Agent-Based Model simulation (a dynamic simulation technique),
we investigated how reducing the prevalence of insufficient sleep (6 hrs.) and improving the
prevalence of average sleep duration (7-8 hrs.), health diet and physical activity by 10% and
20% impact the prevalence of CVD (including diabetes, myocardial infarction, and stroke) over
a 10 year period. We also compared the relative effect sleep, healthy diet, and physical
activity have on diabetes and CVD over a 10-year period.

RESULTS: Based on the simulation, reducing the prevalence of insufficient sleep by 10% and
20% could potentially reduce the prevalence of diabetes among insufficient sleepers by 1.7%
and 2.6%, respectively, in 10 years. In comparison, improving healthy diet or physical activity
by 10% and 20% activity had much less significant effect; the prevalence of diabetes was only
reduced by less than 1% in all simulated scenarios related to improved diet and physical
activity. In addition, reducing insufficient sleep could reduce the prevalence of stroke and
myocardial infarction by 0.5% in 10 years.

CONCLUSION: Improving sleep could be more impactful in preventing or reducing prevalence

of diabetes, myocardial infarction and stroke, compared to healthy diet and physical activity.
S506 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Future research should examine the impact of improving sleep in other population groups,
especially those with high risks of diabetes, myocardial infarction, and stroke.

CORRESPONDING AUTHOR: Azizi Seixas, PhD, MS, New York University School of Medicine,
New York, NY, 10016; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S507

F011 11:45 AM-12:45 PM

MOMMIO: A VIDEOGAME FOSTERING THE USE OF EFFECTIVE VEGETABLE PARENTING

PRACTICES WITH PRESCHOOLERS

Sheryl Hughes, PhD1, R Buday, FAIA2, J Baranowski, MPH1, Teresia O'Connor, MD, MPH1, T
Nicklas, Dr.P.H. 1, A Beltran, MS1, H Dadabhoy, MS, RD1, Tom Baranowski, PhD1

1
Baylor College of Medicine, Houston, TX; 2Archimage, Inc., Houston, TX

Vegetable consumption is related to better child diet quality; however, many parents report
that their children dislike vegetables. Simulation is one way to produce a fun and engaging
experience for parents to learn new skills to facilitate child vegetable consumption. The
Mommio game was designed to produce such an experience, especially for mothers who
oversee food parenting. Mommio was designed to be played on smartphones in short time
frames (e.g. 5-10 min) when busy mothers have time to play. The three dimensional
environment includes a home, back yard, automobile, and a grocery store. The player can
choose specific characteristics for the child avatar (Kiddio) such as gender, hairstyle, and skin
color. Artificial intelligence governs the childs behavior. The mother character is controlled by
the player and can explore all aspects of the three dimensional food environment including
kitchen cabinets, refrigerator, and recipe box. During gameplay, Kiddio becomes hungry and
the player is required to search for vegetables in the home (or at the store) and prepare them
for consumption. When Kiddio refuses to eat the vegetables, the player must select from a list
of effective and ineffective parenting practices to deal with the child. Choosing more effective
vegetable practices results in winning whereas choosing more ineffective practices results in
losing. The player can also set a goal to use an effective practice later with their real world
child. Episodes include a lunch, a dinner, riding in the car, shopping at the grocery store, and
preparing a vegetable dish with the child. Extensive formative research was conducted to
guide the design of the videogame. Tests of alternative forms of providing feedback on
ineffective practice choices revealed that feedback should avoid vague warnings and
provocation of guilt, use succinct but correct grammar, and clearly identify why the choice
was problematic. Early alpha testing revealed mothers desires for more extensive recipes in
the recipe box. Tested recipes met their need for quick easy to prepare dishes that tasted
good to their child and self. Two novellas, Totally Frobisher and CHA, provide backstory.

CORRESPONDING AUTHOR: Sheryl Hughes, PhD, Baylor College of Medicine, Houston, TX,
77030; [email protected]
S508 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F012 11:45 AM-12:45 PM

PARENTS' OPINIONS ABOUT USING THEIR CHILDREN'S EMR FOR RESEARCH: A CONTINUUM
OF PERSPECTIVES ACROSS THREE GENETIC GROUPS

Melissa Raspa, PhD1, Sara Andrews, MPH, RD1, Anne Edwards, BA1, Lauren Turner-Brown,
PhD2, Alexandra Alvarez, BA2, Rebecca Moultrie, BA1, Laura Wagner, MPH1, Anne Wheeler,
PhD1

1
RTI International, RTP, NC; 2UNC Chapel Hill, Chapel Hill, NC

Introduction: The advent of electronic medical records (EMR) has led to a tremendous shift in
health care delivery and the ability to mine large amounts of clinical data. Simultaneously,
advances in genomic science have led to increased understanding of rare diseases. These
developments elicit the need to examine parents preferences for using their childrens EMR
for research purposes. Specifically, there is a need to understand the perspectives of those
with suspected or known genetic conditions; these groups may have the most to gain from
research but also face increased risks.

Methods: Using a semi-structured guide, we conducted focus groups in the triangle area of
North Carolina with parents of typically developing (TD) children with no known genetic
conditions (n=8); parents of children with autism spectrum disorder (ASD), a suspected
genetic condition (n=5); and parents of children with fragile X syndrome (FXS), a known
genetic condition (n=6). The majority were mothers (74%). Participants were 63% White, 26%
Black or African American, and 11% Asian; 5% identified as Hispanic. Approximately 74% were
married and 53% reported family incomes of $75,000 or less. Focus groups were recorded,
transcribed, and analyzed for thematic content.

Results: Almost all participants were in favor of sharing their childrens EMR for research
purposes, although there was a continuum of perspectives. Parents of TD children were the
most conservative in their preferences. Parents of children with ASD were the most open to
sharing, stating that research was the only way to better understand ASD. Parents of children
with FXS were also supportive of EMR research for similar reasons, but had more caveats.
Both known (FXS) and suspected (ASD) genetic groups were willing to share sensitive
information about their children, such as genetic test results and mental health issues, which
they considered part of their childrens conditions. It was also important to these groups to
receive research results. A main concern of all groups was the risk of personal information
being accessed and misused for non-research purposes (e.g., identity theft, stigmatization,
discrimination). All groups also had reservations about pharmaceutical companies using their
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S509

childrens EMR for research, with many worrying that data could be used for ulterior motives.
The known and suspected genetic groups were more willing if the work was conducted in
partnership with trusted researchers. Perceived benefits of sharing EMR data were altruistic in
nature, with the two genetic groups also seeking personal benefit for their children.

Conclusions: Overall the genetic groups believed EMR research would offer more benefits
than risks. Ensuring that parents have sufficient information to make an informed decision
about their childrens participation was critical to all groups preferences.

CORRESPONDING AUTHOR: Melissa Raspa, PhD, RTI International, RTP, NC, 27709;
[email protected]
S510 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F013 11:45 AM-12:45 PM

EXAMINING THE RELATIONSHIP BETWEEN OMEGA-3 FATTY ACIDS AND ANXIETY: POSSIBLE
MECHANISMS AT PLAY

Ashley R. Polokowski, M.A.1, Haque Shakil, Student2, Laura C. Reigada, Ph.D.2

1
The Graduate Center, City University of New York, Brooklyn, NY; 2Brooklyn College, City
University of New York, Brooklyn, NY

Background: One of the most under-recognized factors in the development of mental health
disorders is the role of nutrition. Omega-3 fatty acids (-3 FAs) possess numerous physical
and mental health benefits and are a necessary component of a healthy lifestyle. Given the
anti-inflammatory properties of -3 FAs, they are being tested as a potential treatment
option for various physiological and psychological diseases characterized by high levels of
inflammation. Much of the literature examining -3 FAs and mental health has targeted
depression with growing support that -3 FA supplementation has therapeutic benefits in
reducing depressive symptoms but has been studied much less within anxiety. The aim of the
present systematic review is to assess the current understanding of the possible mechanisms
underlying the relationship between -3 FAs and anxiety. Methods: 1,277 citations were
identified as a result of the initial literature search, after the removal of duplicates (n=388)
and irrelevant literature (n=712), the full-text of 177 studies were screened for eligibility
criteria. Six studies met inclusion criteria for the systematic review. Results: The six studies
were grouped into three sections determined by clustering of primary outcomes reported by
each study (Physiological Correlates, Inflammatory Response, and Neuroendocrine Processes).
Several mechanisms (e.g., IL-6, salivary cortisol) have been noted to potentially explain the
association between -3 FAs and anxiety. No support was found for other variables (e.g.,
heart rate variability, IL-) as possible mechanisms. Future Implications: This systematic
review is the first to examine the potential mechanisms underlying the relationship between
-3 FAs and anxiety. Findings showed a large amount of heterogeneity between studies and
future work should implement more rigorous methodology (e.g., consistent dosage and
duration of -3 FA supplements, assessment of dietary patterns). While it is not possible to
make robust conclusions given the literature is in its early stages, initial empirical findings of
this review provide a preliminary understanding of possible mechanisms by which -3 FAs
influence anxiety. These findings indicate that future research should aim to directly test
mechanisms of physiological correlates (e.g., cortisol) and the inflammatory response (e.g., IL-
6) underlying this relationship. Elucidating these mechanisms can help improve the
prevention and treatment of anxiety.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S511

CORRESPONDING AUTHOR: Ashley R. Polokowski, M.A., The Graduate Center, City University
of New York, Brooklyn, NY, 11229; [email protected]
S512 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F014 11:45 AM-12:45 PM

IMPACT OF A DECISION AID ON SEVERE COPD PATIENTS MOTIVATION TO ENGAGE IN

ADVANCE CARE PLANNING

Melissa Basile, PhD1, Johanna Andrews, MPH1, Sonia Jacome, MSCH1, Meng Zhang, PhD1,
Michael A. Diefenbach, Ph.D.1, Negin Hajizadeh, MD, MPH2

1
Northwell Health, Manhasset, NY; 2Northwell. Health, Manhasset, NY

Purpose: To assess impact of the InformedTogether COPD decision aid on patient motivation
to engage in advance care planning (ACP) i.e. wanting life support vs. comfort care for a
severe breathing exacerbation.

Methods: Severe COPD patients were identified through our health systems EHR and
recruited at 7 outpatient clinics. Patients viewed the decision aid with their clinician, and
were interviewed prior to, immediately after, and 1 month after viewing the decision aid. To
assess motivation to make an ACP, patients were asked at each of these time-points about the
likelihood that they would make an ACP within the next year, and were scored using a 5 point
Likert scale. Clinical encounters using the decision aid, and 1 month follow up interviews were
recorded and transcribed for qualitative analysis.

Results: N=38 patients were enrolled. Mean age: 66 10; majority education: 9 th-12th grade;
race: n=19 White/n=7 Black/n=11 Hispanic/n=1 other; mean health: good/fair; numeracy:
n=19 high/n=19 low. At baseline, 17 people (45%) did not have an ACP or were not highly
motivated to make one. Of those 17, by 1 month: 12 had increased motivation, 3 stayed the
same, 2 had decreased motivation (4 had made, and 5 said they definitely would make, an
ACP). Qualitative analysis showed that 9 people (53%) discussed ACP with their families as a
result of using the decision aid; and the decision aid had helped them to focus on the process
of ACP. At 1 month, one patient stated: You know, throughout the years Ive said, I want
this, I want this, but after seeing the decision aid, I became serious and discussed exactly
what I want. Another stated: You know, [since seeing the decision aid] it has been really
weighing on my mind. Almost like, its just made me feel like, you gotta get things done.

Conclusion: Our decision aid was intended to increase patient motivation to discuss end of
life care preferences, but not necessarily require that a decision be made. For patients with
advanced COPD, ACP facilitated by a decision aid may serve as one variable motivating an
individual to prepare for end of life decision making. For many people, specific events trigger
opportunities for deliberation and planning at different periods in time. Our decision aid
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S513

created one such opportunity for patients by allowing them to introduce a difficult topic with
others involved in their care.

CORRESPONDING AUTHOR: Melissa Basile, PhD, Northwell Health, Manhasset, NY, 11030;
[email protected]
S514 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F016 11:45 AM-12:45 PM

RISKY BEHAVIORS AND DIABETES OUTCOMES IN THE TRANSITION FROM ADOLESCENCE TO

EMERGING ADULTHOOD WITH TYPE 1 DIABETES

Eunjin Lee, Ph.D.1, Cynthia Berg, Ph.D.1, Ashley C. Baker, B.S.2, Daniel Mello, N/A3, Bryce A.
Huntbach, Psychology, BS (Expected Fall 2016) 1, Michelle Litchman, PhD, FNP-BC, FAANP1,
Deborah Wiebe, MPH, PhD3

1
University of Utah, Salt Lake City, UT; 2University of California Merced, Merced, CA;
3
University of California, Merced, Merced, CA

Adolescence and emerging adulthood are important developmental periods for establishing
health-promoting behaviors, though youth are also vulnerable to risky behaviors (e.g.,
smoking, drinking alcohol, and unhealthy weight loss behaviors). Such risky behaviors are
especially problematic for youth with type 1 diabetes, as they increase the risk for developing
health complications in adulthood. This study explored the concurrent and longitudinal
associations between risky behaviors (i.e., smoking, drinking, and skipping insulin to lose
weight) and diabetes management (i.e. adherence and glycemic control). We examined these
associations during the transition from late adolescence to emerging adulthood for individuals
with type 1 diabetes who were in their last year of high school (N = 247, M age = 17.76) at
Time 1 and one year later (Time 2). Adherence was measured by the Diabetes BehaviorRating
Scale (DBRS) and glycemic control was assessed through HbA1c assay kits. As engaging in risky
behaviors is related to higher externalizing behavior problems, all analyses controlled for
oppositional behaviors (as assessed by Conners Rating Scales) to isolate the effect of risky
behaviors in addition to gender, race (white/non-white), pump status, and years since
diagnosis. Furthermore, as smoking and drinking are highly associated with one another,
analyses of these variables were examined simultaneously. Hierarchical regression analyses
revealed that participants (time 1) who had smoked a cigarette in the past six months had
poorer HbA1c levels. At time 2, participants who had smoked a cigarette in the past six
months had lower adherence. From late adolescence to emerging adulthood, participants
who had smoked a cigarette in the past six months at time 1 showed larger decreases in
adherence (time 2, after controlling for adherence at time 1). Drinking did not account for
significant variance in either HbA1c or adherence. Finally, at time 2, participants who skipped
insulin to lose weight had poorer HbA1c. From late adolescence to emerging adulthood, those
who skipped insulin to lose weight more at time 1 changed over time to have poorer HbA1c
(time 2, while controlling` for time 1 HbA1c). The risk behaviors of smoking and skipping
insulin appear to be especially detrimental for adherence and HbA1c in the transition from
late adolescence to emerging adulthood both concurrently and longitudinally.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S515

CORRESPONDING AUTHOR: Eunjin Lee, Ph.D., University of Utah, Salt Lake City, UT, 84103;
[email protected]
S516 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F017 11:45 AM-12:45 PM

THE ONE DROP MOBILE APP AND EXPERTS PROGRAM IS EVIDENCE-BASED AND IMPROVES
BLOOD GLUCOSE

Chandra Osborn, PhD, MPH1, Mark Heyman, PhD, CDE2, Jeff Dachis, MA3

1
One Drop, Informed Data Systems Inc., Nashville, TN; 2University of California, San Diego,
Solana Beach, CA; 3Informed Data Systems Inc., New York, NY

Overview: Mobile health apps must incorporate best practices for clear health
communication, behavior change, psychosocial care and education. The One Drop iOS and
Android app is used by over 120,000 people worldwide. The app facilitates blood glucose (BG)
and self-care tracking, and provides recipes, insights, tips and community support (likes,
stickers, data sharing). In December 2016, One Drop released the Experts coaching service.
Certified Diabetes Educators deliver a structured in-app program and offer on call support.
Here, we describe One Drops grounding in scientific evidence, incorporation of best care
practices, and present its initial efficacy.

Mobile App: The apps self-care tracking feature is designed to enhance understanding
(information) of the link between behaviors and BG. The combination of in-app recipes,
inspirational content and community data sharing and praise is designed to enhance favorable
attitudes and social normative support that, combined, fuel motivation. The medication and
BG check reminders, data-driven insights, and health tips are designed to enhance self-
efficacy or behavioral skills. In these and other ways, the One Drop app is grounded in the
empirically-supported InformationMotivationBehavioral Skills model of behavior change.
Finally, all content is presented clearly, including numerical information.

Experts: Experts uses best practices for clear health communication, omitting jargon,
simplifying instructions and lessons, and using the teach-back technique to confirm
understanding. Experts teaches evidence-based behavioral techniques (e.g., setting realistic
goals, using environmental cues to remember behaviors). It also provides emotional and
instrumental support, and behavior change accountability. Finally, Experts provides
strategies to overcome psycho-social barriers to performing self-care.

Initial Efficacy: After one month, people tracking food and BG with the One Drop App +
Experts:

6. Reduced their average BG by 15% (27 mg/dL)

7. Lowered their average BG from 185 mg/dL (A1c 8.1%) to 158 mg/dL (A1c 7.1%)
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S517

8. Reduced their percentage of high BG readings from 18.5% to 3.7%

9. Achieved 40% more in-range BG readings

Conclusions: One Drop incorporates critical programmatic elements associated with self-care
and A1c in the scientific literature. At one month, the One Drop App + Experts significantly
improved BG. Evidence-based and effective mobile health programs may serve an important
public health role in providing self-care support in a scalable manner.

CORRESPONDING AUTHOR: Chandra Osborn, PhD, MPH, One Drop, Informed Data Systems
Inc., Nashville, TN, 37206; [email protected]
S518 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F018 11:45 AM-12:45 PM

BODY MASS INDEX OF COLLEGE STUDENTS EXHIBITING AVOIDANT/RESTRICTIVE FOOD INTAKE

DISORDER AND PICKY EATING BEHAVIOR

Jordan Ellis, MA1, Hana F. Zickgraf, MA2, Matthew C. Whited, PhD, Licensed Psychologist1,
Amy T. Galloway, Ph.D.3

1
East Carolina University, Greenville, NC; 2University of Pennsylvania, Philadelphia, PA;
3
Appalachian State University, Boone, NC

Avoidant/Restrictive Food Intake Disorder (ARFID) is a new DSM-5 diagnosis designed to

capture patients with clinically significant restrictive eating problems that result in persistent
failure to meet nutritional/energy need or cause psychosocial difficulties. Picky eating (PE)
behaviors are one manifestation of ARFID. PE is characterized by the consumption of a limited
variety of food, through the avoidance or rejection of familiar and unfamiliar foods. PE
research has largely focused on children but PE can persist in to adulthood, and develop into
adult ARFID. It is not known if picky eating among adults results in failure to meet energy need
as it can in children.

The current study examined relationships between Body Mass Index (BMI), ARFID, and adult
PE, to explore differences between PE and traditional ED symptoms in regards to BMI.
Previous research has shown clear relationships between BMI and traditional EDs, but
research on BMI and PE has been mixed. Some studies have indicated that childhood picky
eaters are more likely to be underweight, but these relationships remain unclear in adults and
are important given BMIs status as a health indicator.

Participants in a college sample (n = 1219; 4.1% underweight, 63.2% normal weight, 21.0%
overweight; 11.8% obese) completed the Nine Item ARFID Screener (NIAS), Adult Picky Eating
Questionnaire (APEQ), Eating Disorders Diagnostic Scale (EDDS), and ARFID Symptom
Checklist (ARFID-cl). The ARFID-cl is used to confirm an ARFID diagnosis based on DSM-5
criteria and was used to classify a probable ARFID diagnostic group (n = 31), subclinical ARFID
group (n = 86), group who screened for some picky eating behaviors (n = 799), and typical
eating group (n = 272).

Pearson correlations indicated no significant associations between BMI and the NIAS (r = -.01,
p = .84), or APEQ (r = .05, p = .10). A bootstrapped ANCOVA, with the EDDS composite score as
a covariate, showed BMI did not differ across ARFID groups, F(3, 1202) = .70, p = .55. ARFID
groups were also compared across BMI groups (i.e. underweight, normal weight, and
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S519

overweight/obese), and revealed no significant differences by group (2 (6) = 1.76, p = .94) As

expected, the EDDS was positively correlated with BMI (r = .29, p < .001), and ED risk groups
were associated with BMI in the expected directions.

In the current sample, BMI was not associated with ARFID group, indicating that adult PE
behaviors do not appear to contribute to failure to meet energy need.

CORRESPONDING AUTHOR: Jordan Ellis, MA, East Carolina University, Greenville, NC, 27858;
[email protected]
S520 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F019 11:45 AM-12:45 PM

PREGNANT WOMENS KNOWLEDGE OF GESTATIONAL WEIGHT GAIN: ONLINE NEEDS

ASSESSMENT

Shannon Mellon, Student Researcher1, Alicia A. Dahl, MS2, Brie Turner-McGrievy, PhD, MS,
RD3

1
University of South Carolina: Arnold School of Public Health, Bristol, CT; 2University of South
Carolina, West Columbia, SC; 3University of South Carolina, Columbia, SC

BACKGROUND: The Institute of Medicine (IOM) recommends a range for healthy gestational
weight gain (GWG) based on pre-pregnancy Body Mass Index (BMI). Health practitioners are
the primary source of health information during prenatal appointments. Mobile applications
(apps) and websites hold potential as convenient and available resource for pregnant women.
Electronic health (e-health) interventions could be used as a parallel or supplemental tool to
provide pregnancy weight-related information, specifically IOM recommendations for GWG.
Understanding pregnant womens knowledge and perceptions regarding GWG is essential for
the development of meaningful resources.
METHODS: Using Facebook and Twitter social media sites, pregnant women living in the
United States who owned a smartphone were recruited to complete a brief one-time online
survey regarding the needs and interests of pregnant women to guide the development of an
e-health intervention. The survey included questions regarding pregnancy history, weight-
related knowledge and behaviors, topics discussed with health care practitioners, information
sought by pregnant women, and interests in electronically delivered health information.
Descriptive statistics, chi-square, and regression analyses were used to summarize the data.
RESULTS: A total of 324 pregnant women completed the survey and self-reported pre-
pregnancy height and weight (mean BMI =26.1+6.5). Majority of the sample was normal
weight (53.1%). Half of the participants (50.6%) reported that they did not receive
recommendations for healthy GWG from their physicians during pregnancy. Provider
recommendations were not associated with pre-pregnancy BMI classification (X2=0.189,
p=0.979). When asked about how much weight would be healthy to gain during pregnancy, 46
participants (14.5%) cited GWG below recommendations for their BMI class, 200 (62.9%) cited
within recommendations and 72 (22.6%) cited above recommendations. Normal weight
women most often cited GWG within IOM recommendations (N=114, 35.8%), while
overweight/obese women most often cited GWG above recommendations (N=50, 41%).
CONCLUSIONS: Normal weight participants frequently responded with a GWG
recommendation within IOM guidelines, while overweight/obese participants responded with
a GWG that exceeded IOM guidelines. This indicates that more targeted resources and efforts
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S521

are needed to increase overweight/obese pregnant womens knowledge of healthy GWG.

Physician-provided information to address IOM guidelines may help close that gap in
knowledge, but mobile apps or websites geared towards pregnancy can address GWG
recommendations as well.

CORRESPONDING AUTHOR: Shannon Mellon, Student Researcher, University of South

Carolina: Arnold School of Public Health, Bristol, CT, 06010; [email protected]
S522 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F020 11:45 AM-12:45 PM

THE EFFECTIVENESS OF INTERVENTIONS TO SUPPORT HEALTHIER FOODS AND BEVERAGES IN

SCHOOLS: NEW FINDINGS FROM THE COMMUNITY GUIDE

Holly Wethington, PhD1, Ramona Finnie, DrPH, MPH, CHES1, Devon Okasako-Schmucker,
MPH1, Leigh Buchanan, PhD2, Shawna Mercer, PhD, MSc1

1
Centers for Disease Control and Prevention, Atlanta, GA; 2Cherokee Nation Assurance,
Atlanta, GA

Purpose: Schools can play an important role in promoting healthier dietary habits and
preventing obesity because most US children attend school six hours a day and consume as
much as half of their daily calories at school. This systematic review evaluated the evidence on
effectiveness of interventions to support healthier foods and beverages in schools.

Methods: This review included the subset of studies from the Agency for Healthcare Research
and Qualitys Childhood Obesity Prevention Programs: Comparative Effectiveness Review and
Meta-Analysis that focused on dietary-only approaches in schools, combined with an
electronic database search through February 2016. Using the CDC Comprehensive Framework
for Addressing the School Nutrition Environment and Services, studies were stratified into
four intervention types: 1) Meal and fruit and vegetable snack interventions to increase the
availability of healthier foods and beverages provided by schools (MFVS), 2) Interventions
supporting healthier snack foods and beverages sold or offered as a reward in schools (SFB),
3) Multicomponent interventions to increase availability of healthier foods and beverages in
schools (multicomponent), and 4) Interventions to increase water access in schools (WA).
Community Guide systematic review methods were used to identify studies and judge quality
of execution. Evidence was evaluated using measures of dietary habits and weight outcomes
including, but not limited to, fruit/vegetable consumption and obesity prevalence. When
possible, relative percent change, medians and interquartile intervals (IQI) were calculated for
each outcome.

Results: Thirty-five studies were included. Highlights of dietary and weight-related outcomes
include that MFBS interventions were effective in increasing fruit/vegetable consumption (5
studies/6 arms; relative percent increase of 25.7%, IQI: 0.25%-55.2%) and multicomponent
interventions were effective for stopping the increase in overweight/obesity prevalence (6
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S523

studies, favorable findings for 4). Evidence was insufficient for SFB and WA, due to too few
studies.

Conclusions: Interventions to support healthier foods and beverages in schools show

improvements in several dietary and weight-related outcomes for MFBS and multicomponent
interventions. More research is needed for SFB and WA, and behavioral researchers are
encouraged to develop consistent measures of diet and weight outcomes to allow for
synthesizing and strengthening the evidence base.

CORRESPONDING AUTHOR: Holly Wethington, PhD, Centers for Disease Control and
Prevention, Atlanta, GA, 30329; [email protected]
S524 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F021 11:45 AM-12:45 PM

DO SOCIOECONOMICALLY DISADVANTAGED POPULATIONS USE TECHNOLOGY?

Michele G. Lanpher, MA1, Jacob Christy, BS2, Sandy Askew, MPH3, Gary G. Bennett, PhD3

1
Psychology and Neuroscience, Duke University, Durham, NC; 2Duke Global Digital Health
Science Center, Duke University, Durham, NC; 3Duke University, Durham, NC

Background: Socioeconomically disadvantaged and racial/ethnic minority populations

experience disproportionately high rates of obesity and have the least access to medical care.
Obesity treatment delivered via mobile technologies may be a viable option for this
population. We sought to examine trends in technology use among a medically vulnerable
population enrolled in a behavioral weight loss program.

Methods: Participants (N=351) were enrolled in Track, a 12-month randomized controlled

weight loss trial in primary care. The obesity treatment intervention included weekly self-
monitoring via interactive voice response and text messages, 18 coaching calls with dietitians,
e-scales for daily self-weighing, and skills training. Using baseline data, we analyzed the
proportion of Track participants that used technology and whether use varied by
sociodemographic variables or was associated with engagement in the intervention (i.e.,
percentage of coaching calls and self-monitoring interactions (IVR or text) completed).

Results: Over half (54%) of participants were Black, only 10% received a college degree and
30% were living below the U.S. poverty threshold. All participants were required to own a
text-enabled cell phone, but 9% of individuals reported no SMS use. Over half (53%) owned a
smartphone, 57% had social media (Twitter or Facebook), 75% owned computers (laptop or
desktop), and 85% used Internet or email, while 86% reported using at least one of these
technologies. Individuals with greater education were more likely to use smartphones, social
media, computers and Internet/email (all ps < .01). Those at or above the poverty threshold
were more likely to use the Internet/email (90% vs. 81%) and computers (83% vs. 67%; all ps <
.01). Individuals who were younger were more likely to use all forms of technology (all ps <
.001), as were those with higher health literacy (all ps < .01). Blacks were less likely to have
social media, compared to non-Blacks (51% vs. 64%; p < .05). There was no difference in
technology use by gender or BMI. Among intervention participants, there was no association
between technology use and either of the engagement variables.

Conclusion: As obesity is highly prevalent in medically vulnerable populations, innovative

strategies are needed that are both accessible and engaging. Digital health solutions may be
an appropriate treatment option. Tracks population displayed high uptake of technologies;
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S525

these rates are comparable to those found among U.S. adults. Surprisingly, engagement in the
intervention did not vary by any of the technology use variables, suggesting that limited
technology use did not impede an individuals engagement with the intervention. Obesity
interventions targeting socioeconomically disadvantaged populations should consider using
mobile technologies for treatment delivery.

CORRESPONDING AUTHOR: Michele G. Lanpher, MA, Psychology and Neuroscience, Duke

University, Durham, NC, 27708; [email protected]
S526 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F022 11:45 AM-12:45 PM

EXPLORING ETHNIC/RACIAL DIVERSITY IN RECRUITMENT AND RETENTION IN A MOBILE

HEALTH INTERVENTION TARGETING OBESITY

Kinnari (Nina) Jhaveri, BA, MBA1, Samantha Schleicher, BA2, Michael A. Cohn, PhD3, Ashley E.
Mason, PhD2

1
University of Southern California, San Francisco, CA; 2University of California San Francisco,
San Francisco, CA; 3Uber Technologies, San Francisco, CA

Background: Although the mobile delivery of health behavior interventions may address
treatment barriers, such as inabilities to physically present to a clinic due to cost, childcare, or
work, the extent to which mobile interventions are sought after and utilized across individuals
of differing ethnic background remains unknown. Here, we sought to explore ethnic/racial
differences in recruitment, enrollment, and completion of a 28-day (28-lesson) smartphone-
delivered mindful eating intervention targeting reductions in craving-related eating.

Methods: Participants (n=104) learned of the study via social media (e.g., Facebook, Craigslist)
and UCSF communications (e.g., university-wide email, paper patient letters). Participants
completed in-person visits and text-based assessments before, and one month after
competing, the 28-day smartphone-delivered mindful eating intervention. We used Monte
Carlo estimates of exact tests to examine whether the following differed significantly by
race/ethnicity: 1) initial eligibility, defined as BMI25, age18 yrs, female, smartphone-
owning, locale status, no diabetes, no history of eating disorders, and intractable food
cravings most days of the week; 2) enrollment status, defined as enrolling or not; and 3)
intervention completion per protocol [90 days] or at all [150 days]. We used one-way
ANOVAs to examine whether race/ethnicity was associated with 4) number of lessons
completed, and 5) number of days to complete all 28 lessons.

Results: Of 1651 participants who completed an online screener, 719 met initial eligibility
criteria, and 104 completed a baseline visit and began the intervention. Of these 104
participants (M BMI=31.2; M Age=46.1), 71 (68.3%) identified as White, 11 (10.6%) as
Hispanic/Latino, 10 (9.6%) Asian/Pacific Islander, 5 (4.8%) as Black, 6 (5.8%) as Mixed
Race/Other, and 0 (0 %) as Native American/Native Alaskan. Participants completed an
average of 22.9 days (lessons). The following variables did not significantly differ by
race/ethnicity: Initial eligibility (p=.124); enrollment status (p=.675); intervention completion
per protocol (90 days; n=64; p=.412), or at all (150 days; n=77; p=.368); number of lessons
completed (F=1.46, p=.222); and days to complete intervention (F=0.26, p=.905).
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S527

Conclusions: Diverse populations seek out, are eligible for, and participate in, smartphone-
delivered interventions targeting health behavior change. The steadily increasing rate of
smartphone ownership in the US will render these interventions highly disseminable.
Researchers and clinicians should consider smartphone-delivery as a feasible and acceptable
means of intervention delivery in diverse populations, especially for individuals who are
unable to regularly attend traditional in-person behavioral interventions.

CORRESPONDING AUTHOR: Kinnari (Nina) Jhaveri, BA, MBA, University of Southern California,
San Francisco, CA, 94115; [email protected]
S528 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F023 11:45 AM-12:45 PM

THEORETICAL FACTORS ASSOCIATED WITH REAL-TIME USE OF AN MHEALTH APP DESIGNED

FOR HIV SELF-MANAGEMENT

Maya Baumann, MSN, MPH1, Marcia Holstad, PhD, FNP-BC2

1
Emory University, Peachtree Corners, GA; 2Emory University, Atlanta, GA

Background: Despite the ubiquity of mobile health (mHealth) apps, few mobile phone users
consistently integrate them into their daily lives. Not many empirical studies explore reasons
behind this lack of sustained interest by focusing on theory-based factors that potentially
influence mHealth app acceptance/usage from an end-user perspective.

Purpose: Guided by the modified Unified Theory of Acceptance and Use of Technology
(UTAUT2), this secondary analysis of data from the Music for Health Project (MFHP) an
mHealth app-based HIV self-management intervention (iApp) conducted in rural Georgia
evaluated theoretical factors associated with the acceptance and real-time adoption of the
iApp.

Methods: We compared the results of UTAUT2, smartphone experience (SPexp), and

electronic health (eHealth) literacy surveys with iApp usage among 34 MFHP persons living
with HIV/AIDS (PLWHA) in the first 100 days of study participation. Except for SPexp (baseline
only), surveys were administered at baseline/three months and measured UTAUT2 constructs
(behavioral intention [BI], effort expectancy [EE], hedonic motivation [HM], and performance
expectancy [PE]) and eHealth literacy (confidence finding/using Internet-based health
information). Usage metrics included frequency of iApp openings and duration of time spent
in the app.

Results: At baseline and three months, most scored at or near the highest attainable in all
surveys. UTAUT2 subscales, eHealth literacy, and SPexp were positively intercorrelated with
each other (all p.05) but not with frequency or duration of iApp usage. Younger participants
(Mdn 31-33 years) scored the highest in the UTAUT2 survey, indicating the most intent (BI) to
adopt mHealth apps and reporting stronger beliefs that mHealth apps could be easy to use
(EE), enjoyable (HM), and helpful to maintain health ([PE], all p < .025). Despite the near
universal willingness to adopt mHealth, almost half (14/34, 41%) did not open the iApp - these
individuals were typically newly diagnosed with HIV less than six months prior to entering the
MFHP (p < .025). Among those who did open the app, both frequency and duration of use
peaked in the first four weeks of the study, then declined to almost zero over the next nine
weeks.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S529

Conclusions: Findings indicate that a persons theoretical willingness and his/her behavioral
readiness to use mHealth are not the same. Thus, the intention to adopt mHealth does not
necessarily translate into initial or sustained action. Factors moderating this intention-
behavior mismatch should be more fully evaluated. Moreover, a "one-app-fits-all" approach
might not be the most effective way to improve disease self-management equally among all
HIV-positive patients. Newly diagnosed PLWHA may require an mHealth app specifically
geared to their needs in order to improve disease outcomes and engagement-in-care in this
vulnerable group.

CORRESPONDING AUTHOR: Maya Baumann, MSN, MPH, Emory University, Peachtree

Corners, GA, 30092; [email protected]
S530 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F024 11:45 AM-12:45 PM

USING FOCUS GROUPS TO INFORM DEVELOPMENT OF A FACEBOOK INTERVENTION FOR

TOBACCO USE AND HEAVY EPISODIC DRINKING

Danielle Ramo, PhD1, Manpreet Kaur, M.A2, Derek Satre, PhD2, Judith J. Prochaska, PhD,
MPH3, Ella Corpuz, B.S2, Sandra A. Brown, Ph.D.4, Nicolas Sheon, Ph.D.2

1
University of California, San Francisco, San Francisco, CA; 2University of California San
Francisco (UCSF), San Francisco, CA; 3Stanford University, Stanford, CA; 4University of
California San Diego (UCSD), La Jolla, CA

Introduction. Intervention addressing smoking and heavy episodic drinking (HED) may result in
better tobacco and alcohol outcomes for young adults who exhibit both behaviors. We
analyzed qualitative data from focus groups to inform development of an intervention for
smoking and HED on Facebook. Method: Young adults (N=25, M age = 20.3, 12% female, 72%
non-Hispanic White) aged 18-25, who reported past month smoking and HED (5+ for men, 4+
for women) were recruited through Facebook and assigned to one of the three 90-minute
focus groups on Facebook. Focus groups followed a semi-structured guide facilitating
discussion around smoking, drinking, and social media intervention. Content was
independently coded by two research staff using a grounded theory approach until saturation
was reached. Results. Participants were 40% daily smokers who smoked an average of 5
cigarettes per day on 4 days per week. Participants reported drinking 9 (SD=6.6) days and 8
(SD=6.2) HED episodes in the past month on average. Drinking and smoking was reported
most commonly in social contexts including with friends and at parties. Few reported ever
trying to change smoking or drinking and none had ever tried an online group to quit.
Participants were more receptive to quitting smoking than drinking, and preferred an
intervention targeting one substance over a combined intervention. Participants had mixed
reactions to joining a group on Facebook targeting smoking and/or drinking, with main
concerns being related to privacy and not being ready to change use of either substance.
However, there were many suggestions as to what would be appealing in an intervention,
including about one social media post daily that would address smoking and drinking
separately. Most people were inclined towards seeing pictures and videos in posts.
Participants were interested in live sessions with a counselor and suggested once a week for
20-30 minutes. Most participants reported they were not aware of or not concerned about
Facebook privacy policies, but had concerns about sharing information about substance use
on FB. Conclusions/Discussion: Given low motivation to change tobacco use and HED, social
media intervention for combined tobacco use and HED should focus on motivational
enhancement and engagement. Integrated intervention is warranted and should take into
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S531

account the social context of combined tobacco use and HED, but early focus of an
intervention may need to separate information about tobacco and alcohol use for highest
receptivity.

CORRESPONDING AUTHOR: Danielle Ramo, PhD, University of California, San Francisco, San
Francisco, CA, 94114; [email protected]
S532 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F025 11:45 AM-12:45 PM

MEDIA COVERAGE OF A CRISIS RESPONSE: EXPLORING EBOLA AND TRAVELER HEALTH

Federico P. Feldstein, JD, MA1, Sarah Ray, MA2, Susana Peinado, PhD3, Brittany Zulkiewicz,
BS4, Brian G. Southwell, PhD3, Heather Bair-Brake, MS DVM DACVPM 5

1
CDC/NCEZID/DGMQ/QB, Atlanta, GA; 2RTI International, Atlanta, GA; 3RTI International,
Research Park Triangle, NC; 4RTI international, Waltham, MA; 5CDC/NCEZID/DGMQ/QBHSB,
Atlanta, GA

Media Coverage of a Crisis Response: Exploring Ebola and Traveler Health

Federico Feldstein1, Sarah Ray2, Susana Peinado2, Brittany Zulkiewicz2, Brian Southwell2,
Heather Bair-Brake1

CDC
RTI International

The Check and Report Ebola (CARE) program, implemented by CDC in October 2014, was
designed to support monitoring of travelers entering the US from a country with an Ebola
outbreak. One component of the evaluation of the CARE program is an effort to understand
the information environment in which CARE was created and implemented, including how
media coverage may have been related to key events in the epidemic. We analyzed the
volume of coverage related to CDCs response and the CARE program, focusing on enhanced
entry risk assessment and post-arrival monitoring. We searched US newspapers and TV news
transcripts available through Nexis between July 1, 2014-June 30, 2015, which corresponded
with the major events of the outbreak and response in the US. We used Google Trends to
examine the association between news coverage and online searches to provide insight into
how the public responded to media coverage by looking for additional information. The
volume of media coverage and online searches corresponded with major events in the US
throughout the timeline of the Ebola crisis, peaking in early October 2014 with the diagnosis
of the first imported case of Ebola in the US. By the time entry risk assessment and monitoring
began for all travelers from countries with outbreaks, media coverage and online searches
had begun to decrease, largely declining by the first week of November. Among TV news
sources, CNN was the main driver of coverage, beginning with early coverage of two American
healthcare workers brought to the US for treatment through CDCs announcement of entry
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S533

risk assessment and monitoring procedures and the implementation of the CARE program.
Among newspapers, The Washington Post included the most coverage of key topics
throughout the selected timeframe. Understanding how the volume of coverage changed
over time provides important context to describe the environment in which this response
took place. This analysis provides program planners with information that can be used for
media advocacy and strategic and emergency risk communication in future crises. Next steps
may include a focused analysis of the content of media coverage, as well as an analysis of
relevant conversations on social media.

CORRESPONDING AUTHOR: Federico P. Feldstein, JD, MA, CDC/NCEZID/DGMQ/QB, Atlanta,

GA, 30333; [email protected]
S534 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F026 11:45 AM-12:45 PM

SIMULATION FOR CHRONIC PAIN OPIOID MANAGEMENT: SYMPATHY AFFECTS PHYSICIAN

PRESCRIBING DECISIONS

Marissa Heirich, BS1, Sean Mackey, MD, PhD1, Jordan Newmark, MD2

1
Stanford University School of Medicine, Palo Alto, CA; 2Stanford University School of
Medicine, Redwood City, CA

Opioid related deaths are a major public health issue. Physician overprescribing has been
blamed as a major cause. While many physicians prescribe safely, some do not. Physician
characteristics and patient-physician interaction factors must be assessed to determine what
leads to improper prescribing. Simulation is an emerging tool in health education, which
allows for observation and feedback in a consistent clinical setting. Here, we used simulation
to characterize factors associated with safe and unsafe opioid prescribing.

Twenty pain medicine trainees participated in an opioid simulation at the Stanford Center for
Immersive and Simulation-based Learning. Participants were oriented to a simulated patient
(SP) story, which indicated risk for abuse, and to on-site Risk Evaluation and Mitigation
Strategy tools. Participants were given 45 minutes to interact with the SP and type an
encounter note. SPs provided feedback on a standardized instrument, which included a
patient-provider interaction score (PPI) and overall score (OS). Participants also completed
the Thomas-Kilmann Conflict Mode Instrument (TKI), which measures five conflict-handling
modes. All interactions were video recorded and sent to RIASWorks for coding using the Roter
Interaction Analysis System (RIAS).

Both PPI (M=5.87, SD=1.02) and OS (M=6.05, SD=1.13) were scored on a 9-point Likert scale.
Participants with higher scores for RIAS global affect ratings such as interest, warmth, respect,
and sympathy showed strong and significant relationships with higher SP scores for both PPI
and OS. Higher physician affect ratings of anxiety and anger reflected lower ratings for PPI and
OS as well. Participants who scored higher on the avoiding TKI mode had lower emotional
rapport-building scores (r= -.692, p=.001), and made less empathetic and reassuring
statements. Higher collaborating TKI scores were reflective of higher OS scores (r =.453, p
=.045) and engaged affect (r=.491, p=.028). Physicians who were rated with higher sympathy
scores (M=4.56, SD=.727) were more likely to prescribe opioids than those with lower scores
(M=2.75, SD=.957); t(18) = 3.539, p=.025.

Our pilot findings suggest physicians global affect and conflict-handling have effects on both
patient-reported outcomes and decision making in a simulated opioid interaction. While
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S535

sympathy played a role in building social rapport and higher patient satisfaction, it also
influenced physicians prescribing decisions for a high risk patient. Further simulation studies
are needed in order to better characterize these factors and translate this knowledge into
improved education on safe opioid prescribing.

CORRESPONDING AUTHOR: Marissa Heirich, BS, Stanford University School of Medicine, Palo
Alto, CA, 94303; [email protected]
S536 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F027 11:45 AM-12:45 PM

WHY HELP OTHERS WHEN YOU CAN HELP YOURSELF? A STUDY INVESTIGATING THE
EFFECTIVENESS OF FLU VACCINATION MESSAGES

Melissa A. Shepherd, PhD1, Mary A. Gerend, PhD2, E. Ashby Plant, PhD3

1
University of Montevallo, Montevallo, AL; 2Northwestern University, Chicago, IL; 3Florida
State University, Tallahassee, FL

Influenza infection is associated with social and financial concerns (e.g., missing events or
work due to illness) as well as negative health outcomes (e.g., hospitalization, premature
death). By receiving the flu vaccine people not only protect themselves from the virus; they
also protect others. Certain vulnerable populations (e.g., infants, the elderly, immuno-
compromised patients) may be unable to receive the vaccine and thus rely on herd immunity
(i.e., high rates of vaccination in a population to reduce risk of contracting the virus). We
conducted two experiments to examine whether educating individuals about the benefits of
flu vaccination for themselves only (self-focused) or for themselves and vulnerable others
(herd-focused) would lead to more favorable attitudes and stronger intentions to get
vaccinated. We hypothesized that the herd-focused message would be more effective than
the self-focused message. Furthermore, we hypothesized that the herd-focused message
would be even more effective for some people (e.g., females, democrats, people who had
previously been vaccinated, collectivists). Study 1 included 186 Americans from Amazons
MTURK (ages 19 to 75, 58% male, 83% white). Study 2 included 78 American college students
(ages 18-23, 73% female, 79% white). The method was identical for studies 1 and 2.
Participants reported baseline vaccination attitudes, vaccination history, and personal
characteristics. Participants were randomly assigned to read either a self-focused or herd-
focused flu vaccination message. Lastly, participants reported their vaccination attitudes and
intentions. Contrary to our hypotheses, participants who read the self-focused or herd-
focused message reported equivalent vaccination attitudes and intentions. Additionally,
personal characteristics like gender, political orientation, vaccination history, and
collectivism/individualism did not interact with message type. Study 2 replicated these null
findings. Findings suggest that adding information about herd immunity may not increase the
effectiveness of flu vaccination messages. Results may have differed if herd-focused messages
mentioned benefits to specific family members and/or friends rather than people in general.

CORRESPONDING AUTHOR: Melissa A. Shepherd, PhD, University of Montevallo, Montevallo,

AL, 35115; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S537

F028 11:45 AM-12:45 PM

CITATION AND MERITORIOUS AWARD WINNER

MAPPING A SYNDEMIC OF PSYCHOSOCIAL RISKS DURING PREGNANCY AMONG HIV-INFECTED

WOMEN IN SOUTH AFRICA

Karmel W. Choi, MA, PhD Candidate1, Jennifer A. Smit, PhD2, Nzwakie Mosery, BA2, David R.
Bangsberg, M.D.3, Steven A. Safren, PhD4, Christina Psaros, PhD1

1
Massachusetts General Hospital/Harvard Medical School, Boston, MA; 2Maternal,
Adolescent, and Child Health Research Unit, Department of Obstetrics and Gynecology,
Faculty of Health Sciences, University of the Witwatersrand, Durban, KwaZulu-Natal, South
Africa; 3Harvard Medical School/Massachusetts General Hospital, Boston, MA; 4Department
of Psychology, University of Miami, Coral Gables, FL

Background: A variety of risk factors are known to negatively impact maternal and child
health across the perinatal period in resource-limited settings, and HIV-infected women often
bear a heavy burden of these perinatal risk factors. According to syndemic theory, when risk
factors cluster together, the combination of these risk factors tends to produce worse
outcomes than any risk factor alone. This study sought to use network modeling to
characterize the co-occurrence of psychosocial risks during pregnancy among HIV-infected
women living in KwaZulu-Natal (KZN), South Africa, which has the worlds highest rates of HIV
among pregnant women.

Methods: A cohort of 200 HIV-positive pregnant women was recruited from antenatal clinics
in KZN. Various risk factors were assessed during the third trimester, including antenatal
depression; HIV-related stigma; exposure to physical violence; low-income status; low social
support; unintended pregnancy; distress upon pregnancy recognition; and younger age at
childbearing. Network analysis was conducted in R using mixed graphical modeling to visualize
associations between risk variables, adjusting for variable type(s) and the presence of all other
variables in the network. Centrality statistics were examined to assess relative importance of
each node in the network.

Results: The resulting network revealed complex relationships between risk factors. Of note,
antenatal depression was associated with distress upon pregnancy recognition, which was
strongly tied to unintended pregnancy. Antenatal depression was also connected to low social
support and marginally to physical violence exposure. In turn, low social support was
associated with low-income status and HIV-related stigma. Furthermore, HIV-related stigma
S538 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

was independently associated with distress upon pregnancy recognition, as was younger age.
The most central risk factor in the network was distress upon pregnancy recognition, which
served to connect several other factors.

Conclusions: Our results characterize a network of psychosocial risks during pregnancy in a

sample of HIV-positive women in South Africa. Emotional aspects of unintended pregnancy
emerged as a central feature of this network, suggesting that unintended pregnancy in
contexts of poverty, younger age, HIV-related stigma, and low social support is particularly
distressing and may even contribute to further risks to maternal and child health, such as
depression. Prevention of unintended pregnancies and interventions for coping with
unplanned pregnancies may be especially critical in populations where multiple risks intersect.
In addition, efforts targeting single risk factors (e.g., depression) should recognize the
influence of other contextual variables and consider an integrated, multilevel approach to
supporting the wellbeing of HIV-infected women during the perinatal period.

CORRESPONDING AUTHOR: Karmel W. Choi, MA, PhD Candidate, Massachusetts General

Hospital/Harvard Medical School, Boston, MA, 02114; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S539

F029 11:45 AM-12:45 PM

AN INNOVATIVE POPULATION HEALTH MODEL REDUCES EMERGENCY DEPARTMENT

UTILIZATION IN A LARGE URBAN HOSPITAL SYSTEM

Kenneth Halloran, MPH1, Heather Kitzman, PhD 2, Donald E. Wesson, MD MBA3, Lindsey
Philpot, PhD MPH4

1
Baylor Scott and White Health and Wellness Center, dallas, TX; 2Baylor Scott and White
Health and Wellness Center, Dallas, TX; 3Baylor Scott & White Health, Dallas, TX; 4Mayo Clinic
Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery, Rochester, MN

Background: Individuals who are uninsured and from lower income communities have higher
rates of chronic disease and are typically greater utilizers of emergency department (ED) and
inpatient (IP) health services. Population health approaches that integrate individual,
community, and environmental factors with access to a patient centered medical home have
the potential to improve health and reduce health inequities and healthcare costs. This study
evaluated whether individuals who accessed the Baylor Scott and White Health and Wellness
Center (BSW HWC) had reduced ED or IP health services at Baylor Scott & White Health, a
large urban hospital system in Dallas, TX.

Methods: The BSW HWC, a population health approach of Baylor Scott and White Health,
located in an underserved urban community, contains: 1) a medical clinic, 2) a city maintained
recreational center, 3) a teaching kitchen, 4) prevention programs for diabetes, weight
management, and wellness, 5) access to community health workers for navigation, 6) social
services, 7) weekly farm stands, 8) community outreach staff, and 9) academic and research
staff. Participants with two or more visits to the BSW HWC between January 2010 and
January 2015 were included in this analysis. Baylor Scott & White Health ED and IP utilization
prior to and following initiation of care at the BSW HWC was evaluated with t-test and chi-
square statistics.

Results: Data included 906 and 322 participants for the ED and IP utilization analysis,
respectively (N=1056, 56% African American, 31% Hispanic, 8% Caucasian, 65% female, 65%
uninsured, 14% Medicaid/Medicare, mean age = 50.013.3). The average number of Baylor
Scott & White Health ED encounters significantly decreased by 17.7% (p < .001) following
initiation of care at BSW HWC with an average cost decrease of 34.5%. The average number of
Baylor Scott & White Health IP encounters decreased 36.7% (p < .0001) following initiation of
care at BSW HWC with an average cost decrease of 54.4%.
S540 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusion: As rates of chronic disease rise, particularly in lower income communities,

population health approaches that improve health outcomes and act cost effectively will need
to be adopted. Approaches that address social determinants of health, access to health care,
prevention and management services, and ancillary health services can reduce ED and IP
utilization in large urban healthcare systems leading to better population health and reduced
costs.

CORRESPONDING AUTHOR: Kenneth Halloran, MPH, Baylor Scott and White Health and
Wellness Center, dallas, TX, 75210; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S541

F030 11:45 AM-12:45 PM

COMORBIDITY OF ANXIETY AND DEPRESSION IN A CLINICAL SAMPLE OF INTERNALIZING

YOUTH

Angela W. Wang, M.Ed1, Argero A. Zerr, Ph.D.2, Karen T.G. Schwartz, M.S. 3, V. Robin W.
Weersing, Ph.D.3

1
San Diego State University, San Diego, CA; 2San Diego State University, Suite 200, CA;
3
SDSU/UC San Diego Joint Doctoral Program in Clinical Psychology, San Diego, CA

Comorbidity and Severity of Anxiety and Depression in a Primary Care Sample of Anxious
and/or Depressed Youth

Objectives: To examine the sociodemographic and clinical characteristics of youths with

anxiety compared to youths with comorbid anxiety and depression in a primary care sample
of internalizing youths.

Brief rationale: One-third of adolescents meet DSM-IV criteria for an anxiety disorder, and
11.4% for depression (Merikangas et al., 2010). Furthermore, 10-15% of youths meet criteria
for both (Nolen-Hoeksema & Hilt, 2010), putting them at risk for increased impairment over
time (Garber & Weersing, 2010). Most internalizing youths are less likely to be identified by
parents and teachers, compared to externalizing youths; thus, primary care service providers
are usually the first line of defense in identifying this population (Dwyer, 2006). Therefore, the
present study, based in a primary care setting, evaluated the additional burden that comorbid
depression places on anxious youths, in terms of current symptom severity and impairment.

Methods: The current study evaluated baseline characteristics of 185 youths participating in a
clinical trial in primary care (61.1% anxiety only, 32.4% comorbid depression). Clinical and
sociodemographic characteristics between these two groups were compared.

Results: Youths meeting DSM-IV criteria for both anxiety and depression scored significantly
and consistently higher than those in the anxiety only group on self- (Mc=37.85; Ma=27.96),
S542 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

parent- (Mc=34.28; Ma=29.88), and clinician- (Mc=45.35; Ma=16.08) rated measures of

anxiety and depression symptoms (all p < .05), respectively. Comorbid youths were also more
severe on many secondary symptom domains, including hyperactivity (Mc=4.83; Ma=3.81),
emotional symptoms (Mc=6.05; Ma=3.56), problems with peers (Mc=3.10; Ma=2.41), and
suicidal risk (Mc=.03; Ma= -.65), (all p < .05). Furthermore, anxious youths without depression
scored significantly higher on current levels of general functioning (Mc=52.17;
Ma=58.22).Analyses supported that a comorbid presentation was associated with increased
symptom severity and clinical complexity (Bernstein, 1991; Garber & Weersing, 2010). The
final poster will explore treatment response in an effort to inform intervention selection in
comorbid cases. Additional sociodemographic comparisons (e.g., gender, race, education
level) may also help address baseline health disparities in primary care settings, which may
contribute to the translation of mental health behavioral medicine research into policy and
practice.

CORRESPONDING AUTHOR: Angela W. Wang, M.Ed, San Diego State University, San Diego, CA,
92108; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S543

F031 11:45 AM-12:45 PM

INNOVATIVE PLATFORM (MEVOKED) USES DIGITAL FOOTPRINT TO SUPPORT DEPRESSION

MANAGEMENT IN PRIMARY CARE

Kristin Vickers Douglas, Ph.D., ABPP1, Julie Hathaway, MA2, Arun Ravi, MBS3, Juan Pablo
Ferrer, PsyD, EdM 4, Angela Mattson, D.N.P., M.S., R.N.5, Mark Williams, M.D.6

1
Mayo Clinic Department of Psychiatry and Psychology, Rochester, MN; 2Mayo Clinic,
Rochester, MN; 3Mevoked Inc., San Francisco, CA; 4Mevoked, INC., San Francisco, CA;
5
Department of Nursing, Mayo Clinic, Rochester, MN; 6Integrated Behavioral Health, Mayo
Clinic, Rochester, MN

Effective depression management in primary care often includes use of a clinical registry to
track patient symptom severity and response to treatment over time, nurse care
coordination, and patient self-management support. While accurate and repeated depression
assessment data are necessary for optimizing clinical treatment, patients with depression
have barriers to adherence (Wing et al., 2002), including completion of self-report measures
and clinical data collection with calls and clinic visits. MEVOKED program uses the individuals
digital footprint to collect data relevant to a depressed persons self-management (e.g.,
minutes of online activity during the night, emotional valence of online content). Additionally,
pop-ups collect quick and simple ratings of mood, quality of life, and medication adherence.
MEVOKED provides a weekly summary of online behavior and self-assessment ratings emailed
to the patient. In this study, a team of mental health providers, technology developers, and
patient education researchers, assessed the feasibility and acceptability of MEVOKED within a
primary care clinic already successfully implementing depression care management. Patient
and provider use and satisfaction with MEVOKED were assessed with individual qualitative
interviews. Data sources also included user behavior metrics, such as pop-up response rate
and rating scales. A convenience sample of patients provided informed consent and
qualitative and quantitative data to assess the addition of MEVOKED to their current
depression management program. Of 35 participants using MEVOKED, 21 completed a
qualitative interview (week 6 and 12). Subgroup analysis of clinical data (PHQ-9 scores,
diagnoses, psychiatric medication) indicated participants were representative of patients
enrolled in depression management. Qualitative interview data were analyzed by behavioral
researchers experienced in NVivo qualitative data analysis software and methods of content
analysis. Predominant patient themes included increased awareness of mood states and
usefulness of behavioral prompts for medication adherence. Participant mean ratings (0-10; 0
= very negative/none; 10 = very positive/high) for MEVOKED program at end of study: 7.36
(SD = 2.92) for overall impression of MEVOKED); 7.54 ( 2.81) for perceived added value
S544 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

treatment and 8.50 (2.96) for likelihood of recommending MEVOKED to others. User
difficulties with technology were communicated to MEVOKED team allowing for responsive
program refinement. Provider feedback emphasized integrating MEVOKED data into existing
clinical systems to minimize burden. For those actively using MEVOKED, individualized reports
of online behavior and mood ratings contributed to depression self-management and were
perceived by many as a valuable addition to evidence-based depression management.

CORRESPONDING AUTHOR: Kristin Vickers Douglas, Ph.D., ABPP, Mayo Clinic Department of
Psychiatry and Psychology, Rochester, MN, 55901; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S545

F032 11:45 AM-12:45 PM

THE IPE DEBRIEFING HOUR: CREATING SPACE WITHIN INTERPROFESSIONAL EDUCATION TO

FACILITATE AND PROCESS LEARNING

Leila Islam, Ph.D.1, Kathleen Pincus, PharmD, BCPS2

1
University of Maryland-Baltimore School of Medicine, Baltimore, MD; 2Department of
Pharmacy Practice and Science University of Maryland School of Pharmacy, Baltimore, MD

We developed an interprofessional education (IPE) experience where students from schools

of pharmacy, medicine, and social work longitudinally to serve as a care coordination team for
patients identified complex and at high risk for unplanned healthcare utilization. Over the
course of 10 four-week blocks, students and medical and pharmacy residents cared for 36
patients within a high volume clinic. Participating students (N = 29) in the pilot phase of its
development were in their late 20s (M = 29 years old), mostly women (72%), and largely
White (38%) or Asian (38%). Students completed a pre-and-post measure evaluating their
attitudes toward health care teams (Heinemann, 1999), including assessment of their
attitudes toward the quality of care achieved through collaborative care and physician
centrality. No significant change in attitude was demonstrated among the participants, likely
because participants involved in the learning experience were to some degree interested in
interprofessional care and thus provided scores favorable of IPE.

Lessons learned from this pilot project stemmed largely from the debriefing sessions
facilitated by one or both of the authors, held for one hour at the end of Weeks 1, 3, and 4.
These sessions provided the opportunity for learners to reflect in formal debriefing sessions
on the experience of working as an interprofessional team. Debriefings included structured
activities, including getting to know you exercises and discussion of assigned articles, and
the creation of a safe space in which students learners processed their experience of IPE. We
aim to share the barriers and facilitators experienced throughout the development of this
program, which we have continued into its second year, with the goal to aid others interested
in creating similar IPE programs at their own institutions.

CORRESPONDING AUTHOR: Leila Islam, Ph.D., University of Maryland-Baltimore School of

Medicine, Baltimore, MD, 21230; [email protected]
S546 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F033 11:45 AM-12:45 PM

EXAMINING JOHN HENRYISM AND MEDICATION ADHERENCE AMONG AFRICAN-AMERICAN

WOMEN WITH HYPERTENSION

Yendelela Cuffee, Ph.D., MPH1, Jeroan Allison, MD, MSci2

1
Penn State College of Medicine, Hershey, PA; 2University of Massachusetts Medical School,
Worcester, MA

John Henryism is a measure of maladaptive active coping that has been linked with health
conditions including hypertension, diabetes, and prostate cancer. Few studies have examined
the relationship between John Henryism and health behaviors among individuals with
hypertension. Forty-seven percent of African-American women have hypertension, and
approximately half have controlled hypertension. Adherence to antihypertensive medications
is essential for hypertension control and preventing hypertension-related complications. The
objective of this study was to examine the relationship between John Henryism and
medication adherence among African-Americans women with hypertension. Data were
obtained from the TRUST study, and included 553 African-American women with
hypertension receiving care at a safety-net hospital in the Deep South. Medication adherence
was measured using the 4-item Morisky Medication Adherence scale and John Henryism was
measured using the 12-item John Henryism Active Coping Scale. Associations were quantified
using ordinal logistic regression and adjusted for age, education, income and physical and
mental health status. Within our sample of African-American women, individuals with high
John Henryism scores reported lower adherence to medication (OR: 0.95, CI: 0.92-0.99). Our
findings also indicated that higher self-reported mental health status (OR: 1.04, CI: 1.02-1.05)
was linked with greater adherence. This study of John Henryism provides insights to the
complex psychological and behavioral responses to stress seen among African-American
women with hypertension. The findings of this study support the need for interventions that
promote coping strategies and tailored approaches for improving medication adherence
among African-American women.

CORRESPONDING AUTHOR: Yendelela Cuffee, Ph.D., MPH, Penn State College of Medicine,
Hershey, PA, 17033; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S547

F034 11:45 AM-12:45 PM

PATIENT ATTITUDES TOWARDS AN INGESTIBLE BIOSENSOR MEDICATION ADHERENCE SYSTEM

Peter R. Chai, MD MMS1, Brendan Innes, BS2, Stephanie Carreiro, MD2, Brittany Chapman,
BS2, Sae Jin Oh, BS2, Theodore Mooncai, BS2, Edward W. Boyer, MD PhD3

1
Brigham and Women's Hospital/Harvard Medical School, Boston, MA; 2University of
Massachusetts Medical School, Worcester, MA; 3Department of Emergency Medicine,
Brigham and Women's Hospital, Boston, MA

Introduction: Measuring medication adherence in the real-world is fraught with difficulty.

Direct measures of adherence monitoring (eg. directly observed therapy) are one of the few
methods that provide reliable confirmation of medication ingestion and adherence. A novel
direct measure of adherence, such as a digital pill, can provide reliable ingestion data in the
real-world. Digital pills are composed of a radiofrequency emitter in a standard gelatin
capsule encapsulating a desired medication. Upon ingestion, the digital pill is activated,
transmitting real-time ingestion data to a cloud-based program that displays adherence data
to clinicians. We have previously reported successful deployment of digital pills for opioid
ingestion pattern monitoring from the emergency department (ED). Here, we report initial
qualitative data regarding patient attitudes, beliefs and preferences towards use of this
system for medication adherence.

Methods: We conducted our study at an academic, urban emergency department (ED) in

individuals who were >18 years old, English speaking, without a history of substance abuse,
and sustained acute fractures that would be treated on an outpatient basis. Study participants
were outfitted with a digital pill (eTectRx, Newbury FL) and received brief instruction on
operating the system prior to discharge. Participants used the digital pill for one week, and
returned to our clinical research center for a guide-based open-ended interview regarding
attitudes/beliefs/preferences of the ingestible biosensor system. Interviews were transcribed,
reviewed and analyzed using the technique of applied thematic analysis.

Results: During our study period, 26 individuals were eligible for the study. We approached
and consented 12 participants. Eleven participants (92%) reported acceptance of the digital
pill and the monitoring device, finding the technology easy to operate after the training
session in the ED. Nine participants (75%) perceived real-time adherence monitoring through
a digital pill as acceptable and valuable. Nine participants (75%) found reassurance in
confirmatory text messages, and would have preferred consistent reminder messages to
reinforce adherence. Seven participants (58%) reported they would have preferred increased
S548 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

battery life and a smaller hip monitor. No participants reported concerns regarding privacy of
data transmission.

Conclusion: Digital pills are acceptable as a medication adherence monitoring device among
ED patients. Mobile phone-based confirmatory text messages are an acceptable method to
deliver real-time adherence data to users of digital pills.

CORRESPONDING AUTHOR: Peter R. Chai, MD MMS, Brigham and Women's Hospital/Harvard

Medical School, Boston, MA, 02115; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S549

F035 11:45 AM-12:45 PM

CHANGES IN BOTH DIET AND EXERCISE IMPROVE DEPRESSION SYMPTOMS AMONG

PERMANENT SUPPORTIVE HOUSING RESIDENTS

Whitney Mollenhauer, MA1, Scott Walters, PhD2, Subhash Aryal, PhD1, Esther Holmes, MPH,
CHES1

1
University of North Texas Health Science Center, Fort Worth, TX; 2University of North Texas
Health Science Center, Ft. Worth, TX

Background: There is increasing evidence supporting lifestyle interventions for depression. For
example, there is strong evidence for an inverse relationship between physical activity and
depressive symptoms. In addition, there is growing evidence supporting dietary interventions
as an effective treatment approach for depression. While depression disproportionately
affects people with low income or a history of homelessness, it is unclear if exercise, diet, or a
combined treatment approach should be emphasized in interventions for depression among
this population.

Methods: We used baseline and 6-month follow-up assessments from 315 people who were
participating in m.chat, a technology-assisted health coaching program for people with mental
health symptoms residing in permanent supportive housing. Subjects were categorized into 4
groups based on their improvement in diet and/or exercise over this 6-month period: Group 1
consisted of subjects who improved both diet and physical activity; Group 2 consisted of
subjects who improved diet only; Group 3 consisted of subjects who increased physical
activity only and Group 4 consisted of subjects who did not improve on either measure.

Results: We calculated the change in Patient Health Questionnaire (PHQ-9) scores between
baseline and 6 months and compared the difference between the 4 groups using a one-way
ANOVA. A statistically significant difference was observed between the groups (F3,311 = 3.12,
p = 0.0263). Next, we conducted pairwise comparisons among the 4 groups and observed a
significant difference between groups 1 and 4 (Mean difference = 3.076, 95% CI 0.154 -5.997)
after making Bonferroni corrections for multiple comparison. On average, PHQ-9 scores were
3 points lower for subjects who made improvements on both diet and exercise compared to
those who did not make any improvement in either category.

Conclusion: Lifestyle interventions for depression that target both physical activity and diet
may be more effective than interventions that target either physical activity or diet alone in
members of disadvantaged populations.
S550 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Whitney Mollenhauer, MA, University of North Texas Health

Science Center, Fort Worth, TX, 76107; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S551

F036 11:45 AM-12:45 PM

EVALUATION OF AN MHEALTH APP: A PILOT TRIAL OF DESTRESSIFY ON UNIVERSITY

STUDENT MENTAL HEALTH

Rebecca A. Lee, BScH1, Mary E. Jung, PhD2

1
University of British Columbia, Vancouver, BC, Canada; 2University of British Columbia,
Kelowna, BC, Canada

1 in 5 Canadians experience mental health issues (Smetanin et al., 2011), with those aged 15
to 24 being most at risk (Statistics Canada, 2013). Mental health issues are of a particular
concern for university students as they have shown significantly higher rates of mental health
problems than the general public (Stallman, 2010). Current university support services are
limited by factors such as available staff and finances, and social stigma has frequently been
identified as an additional barrier that prevents students from accessing these resources
(Eisenberg, Golberstein, & Gollust, 2007; Sasaki & Sasaki, 2014). Mobile health (mHealth) apps
are one form of alternative health support that is discrete and accessible to students, and
while they are recognized as a promising alternative, the efficacy of such apps has yet to be
demonstrated through research (Donker et al., 2013).

The efficacy of the commercially available app Destressify Pro was tested in a sample of 163
undergraduate students from the University of British Columbia Okanagan using validated
measures of quality of life, stress, anxiety, depressive symptomatology, sleep quality, and
work productivity. Individuals in the experimental group (n=77) received Destressify Pro and
were instructed to use the app for 4 weeks before completing a follow-up survey. Participants
in the wait list control group (n=86) received no intervention material.

The experimental and control groups were statistically similar at baseline in terms of age, sex,
ethnicity, program enrolment, mental health diagnosis percentage, and mental health service
use (p>.05). Responses were compared between treatment groups using ANCOVA, repeated
measures (RM) ANOVA, and RM MANOVA. App use was associated with lower trait anxiety
(p=.005), higher energy and less fatigue (p=.005), as well as greater emotional wellbeing
(p=.002). These differences were all of small effect size (hp2.05).
S552 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

This research provides insight into the efficacy of mHealth apps in improving stress, anxiety,
and related symptomatology in university students. The results of this study may be used by
student-focused mental health support services to develop future initiatives using an
evidence-based mHealth app to best support students. While changes in stress were not
statistically different between groups (p=.06), mean changes over time were noteworthy in
that the experimental group appeared to be decreasing in stress values (M pre = 18.58; Mpost =
17.78) as compared to control group (Mpre =19.60; Mpost =19.84). Similarly, depressive
symptomatology was not statistically different between groups but also showed similar trends
(Mexppre = 8.36; Mexppost = 6.36; Mconpre = 8.08; Mconpost = 7.37). Future research may benefit
from examining Destressify Pro and similar mHealth apps over a longer period of time and in
university staff and faculty.

CORRESPONDING AUTHOR: Rebecca A. Lee, BScH, University of British Columbia, Vancouver,

BC, V6T 1W9; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S553

F037 11:45 AM-12:45 PM

MENTAL HEALTH STIGMA AND ATTITUDES TOWARD SEEKING MENTAL HEALTH SERVICES IN
TAIWANESE IN TAIWAN AND US

Yung Fang Irene Chung Chavez, M.A.

California School of Professional Psychology at Alliant International University - Los Angeles,

San Gabriel, CA

Limited research has examined mental health in Taiwanese in Taiwan and Taiwanese descent
in the US. The aim of the current study is to examine: 1) the relationship between mental
health stigma (MHS) and mental health help seeking attitudes (MHHSA) in Taiwanese in
Taiwan and in the US; 2) the association of acculturation with MHS and MHHSA; 3) whether
loss of face moderates the relationship between MHS and MHHSA, and 4) factors that should
be addressed by culturally appropriate interventions to reduce MHS and improve MHHSA in
Taiwanese in Taiwan and the US. The sample consisted of 243 participants (114 Taiwanese in
Taiwan and 129 Taiwanese descent in the US). Participants completed an online survey
assessing mental health self-stigma, public stigma, loss of face, acculturation (for US sample
only), and MHHSA. Independent-Samples t-Test and Hierarchical Multiple Regression analyses
were conducted. The results showed that Taiwanese in Taiwan reported greater levels of self-
stigma (M = 2.34, SD = .54) than Taiwanese in the US (M = 2.18, SD = .61), p < .05. However,
differences in public stigma was not found between the two groups. Taiwanese in the US (M =
3.12, SD = .47) reported significantly more positive MHHSA than their counterparts in Taiwan
(M = 2.91, SD = .46), p =.001. Increased levels of self-stigma and public stigma significantly
predicted less positive MHHSA for both groups. Acculturation as measured by levels of
immersion into Taiwanese and US mainstream cultures predicted MHHSA. Specifically, greater
levels of immersion into Taiwanese culture significantly predicted less MHHSA (B = -.01, SE =
.002, p = .001) whereas greater levels of immersion into US culture significantly predicted
more positive MHHSA, B = .01, SE = .002, p = .005. Loss of face did not moderate the
relationship between MHS and MHHSA in both Taiwan and US samples. The results of this
study inform research on the relationship between MHS and MHHSA. This finding calls for
future examination of other factors that may moderate the relationship between MHS and
MHHSA. Findings also highlight the significance of developing interventions to reduce mental
health stigma and also interventions that are sensitive to cross-cultural and acculturative
differences across Asian and other immigrant populations.

CORRESPONDING AUTHOR: Yung Fang Irene Chung Chavez, M.A., California School of
Professional Psychology at Alliant International University - Los Angeles, San Gabriel, CA,
91775; [email protected]
S554 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F038 11:45 AM-12:45 PM

REACTIONS TO ACUTE MEDICAL EVENTS: TRAUMATIC VS. NON-TRAUMATIC STRESSORS

Anna-Leigh Powell, B.S., Kelsey C. Hewitt, M.S., D. Brian. Haver, B.S., B.A., Meghan W. Cody,
Ph.D.

Mercer University, Atlanta, GA

Posttraumatic stress disorder (PTSD) is one of the few psychopathologies requiring a specific
antecedent for the diagnosis. Criterion A for PTSD in the DSM-5 defines a traumatic event as
one that involves exposure to actual or threatened death, serious injury, or sexual violence.
Operationalizing what constitutes a trauma is challenging as individual perceptions and
appraisals differ and can alter the impact of the event (Stein, Wilmot, & Solomon, 2016). One
area where this distinction is especially difficult to discern is medical illness. Research on PTSD
due to acute medical events has largely focused on heart attacks and strokes, two life-
threatening occurrences that do meet the Criterion A threshold. However, some non-
traumatic medical stressors may be associated with similar psychological effects (Edmondson,
2014). The purpose of the current study is to explore the impact different types of acute
medical events/stressors (AMES) have on subjective responses and psychological symptoms.

Participants (N = 51) completed an online study of reactions to AMES, defined as serious acute
illness or stressful diagnostic/treatment procedures. Participants first reported which AMES
they had experienced (out of a list of 10) and identified the one they considered most
distressing. Next, they answered questions about their subjective responses to the event,
including pain, helplessness, fear, belief that they would die, and current distress. AMES were
classified into categories to reflect whether or not they met Criterion A. Traumas (e.g., heart
attack, miscarriage; n = 17) were compared to non-traumatic stressors (e.g., surgery,
diagnostic procedures; n = 34) to identify differences in subjective responses and scores on
measures of PTSD (PTSD Checklist; PCL-5) and depression (Beck Depression Inventory; BDI-II).

Results showed no significant differences between trauma groups for subjective responses,
PTSD symptoms, or depression symptoms. However, regressions predicting symptoms from
subjective response scales were significant; PCL: R2 = .38, p < .001; BDI: R2 = .24, p = .018.
Belief that one would die was the only significant predictor (at the time of the event) for both
PTSD symptoms ( = .38, p = .004) and depression ( = .35, p = .015). This suggest that
whether a medical stressor meets PTSD Criterion A is less important than the patients
interpretation of the event in determining its long-term effects on mental health.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S555

CORRESPONDING AUTHOR: Anna-Leigh Powell, B.S., Mercer University, Atlanta, GA, 30329;
[email protected]
S556 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F039 11:45 AM-12:45 PM

SCREENING OF MOOD SYMPTOMS IN NEW PATIENTS PRESENTING TO REPRODUCTIVE

ENDOCRINOLOGY CLINIC: TIME FOR EARLY INTERVENTION?

Abbey Kruper, Psy.D.

Medical College of Wisconsin, Milwaukee, WI

AIM
Women are at greatest risk for mood disorders during childbearing years and infertility is a
significant and stressful life event further impacting emotional well-being. The aim of this
study was to determine presence of clinically significant anxiety or depression in new patients
establishing care within reproductive endocrinology clinic with goal of early intervention with
clinical health psychologist.

METHODS
Data was collected for all new patients presenting to the Reproductive Endocrinology clinic for
consultation regarding infertility. A total of 170 patients were assessed during a 3-month time
period from April 2016 through June 2016. Fertility preservation consults were excluded.
New patients were given the Patient Health Questionnaire-9 (PHQ-9) and the Generalized
Anxiety Disorder-7 (GAD-7) scale by nursing staff to complete prior to their physician
appointment. Scores were calculated for the presence of mild, moderate, or severe
symptomatology. A cut-off of 10 indicating moderate anxious or depressed symptoms was
set to prompt physician discussion regarding emotional well-being with patients and referral
for psychological intervention early in treatment process.

RESULTS
Of the 170 patients evaluated, 126 of the patients (74%) scored below 5 on both measures
indicating minimal to no emotional distress at this early time point in the fertility process. 33
patients (19%) endorsed mild symptoms of depression or anxiety and the remaining 11
patients (7%) indicated moderate symptoms. No patient endorsed severe symptoms of either
depression or anxiety at this time point. Of the 11 patients indicating moderate symptoms, 7
patients were referred for initial psychological evaluation and 3 elected to schedule and
attend the appointment.

CONCLUSIONS
Despite the incidence of mood disorders in general population and the stress associated with
reproductive time period for women, a lower than expected amount of patients endorsed
clinically relevant mood symptoms at their initial infertility consult. Of those that endorsed
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S557

clinically significant mood symptoms, relatively few followed through with establishing with
health psychology. Ongoing assessment of whether there is clinical utility in initial
psychological distress measures is warranted. Based on these results, for next steps clinical
providers modified the threshold for referral to mild symptoms and initiated re-evaluation
time points, including after failed treatment interventions when emotional distress is likely to
increase. Barriers to early intervention with health psychologist should also be identified.

CORRESPONDING AUTHOR: Abbey Kruper, Psy.D., Medical College of Wisconsin, Milwaukee,

WI, 53211; [email protected]
S558 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F040 11:45 AM-12:45 PM

DEVELOPMENT AND ASSESSMENT OF HEALTH SURVEYS: FITTING A GRADED RESPONSE

MODEL

Sarah Depaoli, PhD., Jitske Tiemensma, PhD, John Felt, MA

University of California, Merced, Merced, CA

Objective: The graded response model, an item response theory (IRT) model, can be used to
improve the development of surveys, even when sample sizes are restricted. Typically, health-
based survey development utilizes classical statistical techniques (e.g., reliability and factor
analysis). In a review of four prominent journals within the field of Health Psychology, we
found that IRT-based models were used in less than 10% of the studies examining scale
development or assessment. However, implementing IRT-based methods can provide more
details about individual survey items, which is useful when determining the final item content
of surveys. An example using a disease specific quality of life (QoL) questionnaire for patients
treated for Cushings syndrome is discussed. Cushings syndrome is a rare hormonal disease
characterized by chronic exposure to elevated cortisol levels. Patients suffering from
Cushings syndrome serve as a disease model for chronic stress.

Design: Patients treated for Cushings syndrome were recruited from a patient organizations
(The Cushings Research and Support Foundation) email listserv and Facebook. Patients
completed an online version of the Cushings Quality of Life (QoL) questionnaire (CushingQoL;
English version) and demographic questions.

Results: Patients with Cushings syndrome (n = 397) completed the CushingQoL. Results from
the graded response model supported a 2-subscale scoring process for the survey [M2(17) =
19.25, p = .31). All items were deemed as worthy contributors to the survey (all slopes > 1.0).

Conclusion: The graded response model can accommodate unidimensional or

multidimensional scales, be used with relatively lower sample sizes, and is implemented in
free software. Use of this model can help to improve the quality of health-based scales being
developed within Health Psychology.

CORRESPONDING AUTHOR: Sarah Depaoli, PhD., University of California, Merced, Merced, CA,
95343; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S559

F041 11:45 AM-12:45 PM

FACTORIAL INVARIANCE AND RELIABILITY OF THE DECISIONAL BALANCE AND SELF EFFICACY
SCALES FOR EXERCISE

Iman Almohammadi, Behavioral Science Doctoral Student1, Wayne Velicer, Professor, Co-
Director of Cancer Prevention Research Center2, Joseph Rossi, Professor; Director of the
Behavior Change Research Center2, Colleen Redding, Research Professor, Cancer Prevention
Research Center2

1
University of Rhode Island, North Kingstown, RI; 2University of Rhode Island, Kingston,, RI

Two crucial psychometric requirements for any measure are scale reliability and factorial
invariance. This paper evaluates the Decisional Balance for Exercise and the Self-Efficacy Scale
for Exercise, using a sample of parents or legal guardians of 9th-grade students (N = 347)
across Rhode Island. Three levels of invariance were tested, from least to most restrictive:
Configural Invariance, Pattern Identity Invariance, and Strong Factorial Invariance. In Study I,
the Decisional Balance Scale for Exercise, consisting of two uncorrelated factors with three
items each for (Pros) and (Cons) was evaluated. Strong Factorial Invariance was a good fit for
the model across educational level (CFI: 0.92), Pattern Identity Invariance across gender (CFI:
0.94), annual income (CFI: 0.93), and stage of change (CFI: 0.91), and Configural Invariance
across age (CFI: 0.95). Internal consistency (Coefficient Alpha) was .85 for Pros, and .69 for
Cons. In Study II, the Self Efficacy Scale for Exercise had a hierarchical model that consists of
two factors with two items for (Internal Confidence) and four items for (External Confidence).
Strong Factorial Invariance was a good fit for the model across gender (CFI: 0.99), age (CFI:
0.92), educational level (CFI: 0.98), annual income (CFI: 1.00), and stage of change (CFI: 0.91).
Coefficient Alpha was .36 for Internal Confidence, .68 for External Confidence, and .66 for the
Total Scale. Together, invariance and reliability provide empirical support for the validity of
these two measures.

CORRESPONDING AUTHOR: Iman Almohammadi, Behavioral Science Doctoral Student,

University of Rhode Island, North Kingstown, RI, 02852; [email protected]
S560 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F042 11:45 AM-12:45 PM

IDENTIFYING ATYPICAL RESPONSES IN THE CUSHING'S QUALITY OF LIFE QUESTIONNAIRE:

USING PERSON FIT STATISTICS TO DETECT OUTLIERS

John Felt, MA, Ruben Castaneda, NA, Sarah Depaoli, PhD., Jitske Tiemensma, PhD

University of California, Merced, Merced, CA

Context: Cushings syndrome is a rare hormonal disorder characterized by the chronic

overproduction of cortisol. Symptoms include physical, psychological, and health
impairments. Upon remission, symptoms typically improve, however quality of life (QoL)
remains impaired despite long-term remission. Because impairments in QoL are persistent in
patients treated for Cushings syndrome, the Cushings QoL (CushingQoL) questionnaire was
developed to evaluate disease specific QoL concerns. When evaluating QoL using the
CushingQoL questionnaire, outlier detection is important. However, traditional methods of
outlier detection (e.g., boxplots) can miss patients with atypical responses to the questions
that otherwise have similar total (subscale) scores. In addition to detecting outliers, it can be
of clinical importance to determine the reason for the outlier status, or atypical response.

Objective: The aim of the current study was to illustrate how to derive validity scores for
outlier detection through a statistical method examining person fit.

Design and Patients: Patients treated for Cushings syndrome (n=394) were recruited from the
Cushings Support and Research Foundations listserv and Facebook page. All patients were
asked to complete an online version of the CushingQoL.

Results: Conventional outlier detections methods revealed no outliers reflecting extreme

scores on the subscales of the CushingQoL. However, person fit statistics identified 18
patients with atypical response patterns, which would have been otherwise missed (Zh >
|2.00|). Twelve of the patients with atypical responses tended to endorse the more
extreme always or never responses and were deemed as invalid responses. The other 6
patients tended to endorse the sometimes or somewhat responses and were deemed as
overfitting.

Conclusion: While the conventional methods of outlier detection indicated no outliers, person
fit statistics identified several patients with atypical response patterns who otherwise
appeared average. Person fit statistics allow healthcare providers to delve further into the
underlying problems experienced by these atypical patients treated for Cushings syndrome.
Person fit statistics are an important tool for physicians and researchers when working with a
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S561

variety of patient groups. This method can also be extended to other types of health-related
questionnaires.

CORRESPONDING AUTHOR: John Felt, MA, University of California, Merced, Merced, CA,
95340; [email protected]
S562 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F043 11:45 AM-12:45 PM

LATENT GROWTH CURVE MODELS FOR BIOMARKERS OF THE STRESS RESPONSE

John Felt, MA, Sarah Depaoli, PhD., Jitske Tiemensma, PhD

University of California, Merced, Merced, CA

Objective: The stress response is a dynamic process that can be characterized by predictable
biochemical and psychological changes. Biomarkers of the stress response are typically
measured over time and require statistical methods that can model change over time.
Through a search in the journal Health Psychology since 2000, the most commonly used
methods to evaluate change over time in biomarkers were ANOVA and multilevel modeling
approaches. Despite its flexibility, we identified no studies using latent growth curve models
(LGCMs). Stress researchers may be unaware of how these methods can be useful. Therefore,
our purpose is to provide an overview of LGCMs in the context of stress research.

Methods: The use of LGCMs are discussed in the context of the Trier Social Stress Test (TSST),
a common laboratory acute stress paradigm. Results: The following four specifications of the
LGCM are described: basic LGCM, LGCM for multiple groups (observed or unobserved),
piecewise LGCM (PLGCM), and LGCM for two parallel processes. The basic LGCM is best suited
for evaluating the rate of change over time. The LGCN for multiple groups can evaluate
differences in growth trajectories for measured groups (e.g., gender, experimental groups) or
unobserved groups (e.g., nonresponders vs. high responders). The piecewise LGCM is best
suited for evaluating nonlinear change where the location of the bend (i.e., time-point where
growth becomes nonlinear) in the growth trajectory line is known a priori. The LGCM for two
parallel processes is best suited to model the relationships between two different
simultaneous measures of the stress response.

Conclusions: The LGCM and the alternative specifications of the LGCM are a useful tool for
stress researchers evaluating change over time of biomarkers such as cortisol or alpha-
amylase. Ultimately, the choice of which statistical method to use will be determined by the
research question and characteristics of the data. A link to example code of each specification
of the LGCM will be provided on the poster.

CORRESPONDING AUTHOR: John Felt, MA, University of California, Merced, Merced, CA,
95340; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S563

F044 11:45 AM-12:45 PM

MODEL (MIS)FIT IN A CONFIRMATORY FACTOR ANALYSIS OF THE RELIGIOUS ORIENTATION

SCALE REVISED

Cecilia Bess, M.A.1, Andrew Littlefield, PhD1, Jennifer L. Brown, Ph.D.2, Amelia Talley, PhD1

1
Texas Tech University, Lubbock, TX; 2University of Cincinnati, Cincinnati, OH

The Religious Orientation Scale-Revised (ROS-R) was developed to assess religious motivation
with two sub-scales (i.e., intrinsic religion as an end, an individual who lives his/her life
according to their religious beliefs; extrinsic religion as a means to an end, an individual who
uses religion to achieve personal non-religious goals). The ROS-R is a 14-item questionnaire
derived from the 20-item original measure; the ROS-Rs two-factor structure was originally
tested, via confirmatory factor analysis, in a sample of students who attended secular and
religious colleges. The ROS and ROS-R are the most commonly used measures when
examining religious orientation in relation to health. Within the broader health literature,
individuals who are intrinsically oriented are less likely to partake in unhealthy behaviors and
experience fewer anxiety symptoms than those who are extrinsically oriented. The aim of the
present study was to confirm the factor structure of the ROS-R in a large sample of college
students (N = 810; 71.3% female, M age = 20.1, SD = 2.9). Confirmatory factor analysis
evaluated the goodness of fit of the previously validated, two-factor structure. Comparative
fit index (CFI), root-mean-square error of approximation (RMSEA), and weighted root mean
square residual (WRMR) indicators of model fit were considered. Results indicated poor
model fit for all models tested (e.g., significant 2 values; CFIs < .90, RMSEAs > .10; Hu &
Bentler, 1999). Results suggest greater caution when health researchers consider using the
suggested two-factor structure of the ROS-R. Future research should be conducted to further
examine the factor structure and psychometrics of the ROS-R in diverse samples.

CORRESPONDING AUTHOR: Cecilia Bess, M.A., Texas Tech University, Lubbock, TX, 79416;
[email protected]
S564 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F045 11:45 AM-12:45 PM

IMPLEMENTATION OF A TBI SCREENING CLINIC FOR OEF/OIF VETERANS

Jenna Gress Smith, PhD1, MaryLu Bushnell, PsyD2

1
Phoenix VA Medical Center, Gilbert, AZ; 2Phoenix VA Medical Center, Phoenix, AZ

Approximately 6.7% (Taylor, 2011) of Operation Enduring Freedom/Operation Iraqi Freedom

(OEF/OIF) Veterans are returning from deployment with traumatic brain injury (TBI), also
known as the "signature injury" of this conflict. TBI can be associated with a host of
neuropsychological deficits such as problems with attention, memory, and executive
functioning. Although most individuals who sustain a mild TBI have full recovery within days
to weeks post injury, a small proportion have cognitive deficits that extend beyond one year.
Reports of cognitive complaints are often complicated by co-occurring mental health
symptoms. The current study sought to understand the prevalence and relationships among
TBI, post-traumatic stress disorder PTSD, and depression among OEF/OIF Veterans within the
context of a neuropsychological screening evaluation which was conducted during the same
appointment as the nationally mandated second level TBI evaluation. The purpose of the clinic
was to improve access and quality of care by conducting brief (one hour) neuropsychological
screening exams, the results of which were used to connect Veterans to the most appropriate
level of care. The majority of the 138 participants was male (95%) and Caucasian (71%); 44%
were married, 48% employed, and the number of deployments ranged from 1 to 14 (M=1.79
SD=1.40). On average, Veterans reported clinically significant levels of depression (Beck
Depression Inventory-II; M=23.76, SD=13.34) and PTSD (PTSD Checklist, military version;
M=51.74, SD=16.02) symptoms. Averages for neuropsychological subscales, as measured by
the Neuropsychological Assessment Battery (NAB), were within normal limits (reported as
standard scores between 85-115) for attention, memory, and executive functioning.
Correlations were also examined; while PTSD symptoms were negatively correlated with all
three subscales of the NAB, depression symptoms were only negatively correlated with
memory (all ps < .05). After this brief assessment, 20% of veterans were referred for full
neuropsychological batteries and 70% to mental health services. The results of this study
highlight the potential benefits of a brief screening assessment to improve services, access,
and provide psychoeducation to newly returned Veterans.

CORRESPONDING AUTHOR: Jenna Gress Smith, PhD, Phoenix VA Medical Center, Gilbert, AZ,
85295; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S565

F046 11:45 AM-12:45 PM

MIXED-EFFECTS MODEL ANALYSIS OF REDUNDANCY IN LONGITUDINAL ELECTRONIC HEALTH

RECORDS

Jihoon Kim, MS1, Hai Yang, MS1, Kellie Avery, BS2, Hyeon-Eui Kim, RN,PhD1, Tania Zamora,
BA3, Zia Agha, MD4, Lucila Ohno-Machado, MD,PhD1

1
University of California, San Diego, La Jolla, CA; 2VA San Diego Healthcare System, La Jolla,
CA; 3VA San Diego Healthcare System, San Diego, CA; 4West Health, La Jolla, CA

Widely adopted copy-and-paste function in electronic health records (EHRs) system impacts
negatively on patient care by introduction of errors, decreasing the efficiency in patient care
due to redundant information in EHR, and imposing cognitive burden on physicians requiring
unnecessary sifting through the clinical notes in a limited time during patient encounter.
Previous studies on redundancy in EHRs have limitations. They only focused on the whole
note even though nested sections have different redundancy or used only a small number of
notes, no systematic regression approach was applied that can correctly capture covariance
structure and multi-levels inherent in EHR redundancy data. We collected 950 clinical notes
from 100 patients seen by 18 physicians over 31 time-points from Veterans Affairs (VA) San
Diego hospital. Each note was tagged and split into sections of assessment plan, labs,
medications, past medical history, physical exam, vital sign, and a whole note. Preprocessing
was applied with natural language processing (NLP) that includes lower-casing,
punctuation/comma removal, stop-word removal, and lexical normalization. A series is
defined as a set of notes for a single patient in a chronological order. Within a series, the note
created at the earliest time-point was deemed as a reference note, from which a distance to
all other notes, within a series, was measured. Three distance functions are normalized
Levenshtein edit distance, Needleman-Wunsch (NW) global alignment based distance, and
Smith-Waterman (SW) local alignment based distance. A mixed-effects model was fit using
DISTANCE as an outcome, PATIENT identifier as a group variable, TIME as a nested/correlated
within a patient, and three covariates; document type (DOCTYPE), ALGORITHM for distance,
preprocessing of either NLP-processed or not. Python 2.7 and Natural Lanuage Toolkit (NLTK)
3.0 were used for data management and R 3.3.1 was used for statistical analysis. Starting with
the null model, each covariate was sequentially added if ANOVA test of a full model and a
nested model was significant. The resulting random effects model retained 4 covariates,
DOCTYPE, ALGORITHM, NLP, and TIME, all significant at p-value threshold 0.05. Profiling per-
section trend of redundancy in an individual patient showed the marked difference of pattern
for each section in reference to the whole note. For example, the past medical history section
level pattern showed plateau pattern while the whole note level pattern fluctuated along the
S566 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

time. Hence different sections contribute differently to the redundancy between two clinical
notes. When the rank-based correlation was measured between distance algorithms,
Levenshtein and NW showed high correlation ranging [0.87, 0.97]. Our study shows that
section instead of a whole note is preferred unit of redundancy analysis, and demonstrates
the usefulness of mixed-effects model.

CORRESPONDING AUTHOR: Jihoon Kim, MS, University of California, San Diego, La Jolla, CA,
92093; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S567

F047 11:45 AM-12:45 PM

ROBOTIC THERAPY IN A VA LONG TERM CARE FACILITY - EFFECTIVENESS DATA ON AN ITEM

LEVEL

Geoffrey W. Lane, PhD / ABPP 1, Delilah Noronha, PsyD2, Carolyn Resner, BA3

1
VA Palo Alto Healthcare System - Livermore Division, Livermore, CA; 2VA Palo Alto Healthcare
System - San Jose Clinic, San Jose, CA; 3Palo Alto University, Livermore, CA

Robotic technology has become increasingly widespread in its adoption and acceptance in
mainstream society. One of the most recent developments in the clinical use of robotics is
that of the socially assistive robot, with the most prominent and early entrant to this social
robotics niche taking the form of a baby harp seal, dubbed 'Paro.'

Paro has been studied and used with a variety of populations, and most extensively, with
older adults in residential and custodial care settings. Paro has been previously found
effective in reducing depression, subjective loneliness, and improving cognitive function in
older adults, and there has been evidence suggesting Paro is effective in reducing behavioral
and psychological symptoms of dementia. Recently, the Department of Veterans Affairs has
become interested in using social robotics with older adult long-term care veterans as a
psychosocial intervention at their Community Living Centers (AKA, VA nursing homes).

The current proposal involves an item-level reanalysis of previously published data

demonstrating the broad effectiveness of Paro in improving mood and behavior
in residents at a specialty geropsychiatric CLC facility. Previous data has found that use of Paro
tends to be more effective in improving observations of positive mood and prosocial behavior,
rather than decreasing negative mood states and antisocial behaviors (e.g., both internalizing
and externalizing in nature). It is thought that this is because introduction of a
nonpharmacological intervention tool like Paro is effective for subjects who are not acutely
distressed, and therefore are more receptive to introduction of a companion robot.

What is as yet unclear is for what specific kinds of behavioral and mood symptoms and states
the introduction of Paro might be most or least effective for in veteran long-term-care
residents. This kind of information could help to guide clinicians in helping to structure and
target their introduction of such an intervention to patients and residents for whom they
stand the best chance of success.

Results suggest that while the greatest magnitude of change in behavioral and mood
descriptors observed by nursing staff was in positive mood and behavioral descriptors such as
S568 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

conversing, and bright affect, there was evidence that Paro was effective in reducing
nursing observations of negative states such as anxiety and even externalizing behaviors
such as wandering. It should be noted that, in line with expectations, observations such as
pacing, yelling, or complaints of pain were essentially flat.

Reasons for these findings will be briefly discussed, as well as clinical implications for veteran
and non-veteran populations, and suggestions for further research.

CORRESPONDING AUTHOR: Geoffrey W. Lane, PhD / ABPP , VA Palo Alto Healthcare System -
Livermore Division, Livermore, CA, 94550; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S569

F048 11:45 AM-12:45 PM

SMOKING AND DRINKING BEHAVIORS OF MILITARY SPOUSES: FINDINGS FROM THE

MILLENNIUM COHORT FAMILY STUDY

Daniel Trone, PhD1, Teresa Powell, MS2, Amber D. Seelig, MPH3, Arthur V. Peterson, Jr.,
Ph.D.4, Alyson J. Littman, PhD5, Emily Williams, PhD, MPH6, Charles Maynard, PhD7, Jonathan
Bricker, PHD 4, Edward J. Boyko, MD, MPH8

1
Military Population Health Directorate / Naval Health Research Center, San Diego, CA; 2Henry
Jackson Foundation, San Diego, CA; 3Seattle Epidemiologic Research and Information Center,
Va Puget Sound Health Care System, Seattle, WA; 4Fred Hutchinson Cancer Research Center,
Seattle, WA; 5VA Puget Sound Health Care System and the University of Washington, Seattle,
WA; 6VA Puget Sound Health Services Research & Development and University of Washington
Department of Health Services, Seattle, WA; 7Department of Veteans Affairs, Seattle, WA; 8VA
Puget Sound Health Care System, Seattle, WA

The association between stressful military experiences and tobacco use and alcohol misuse
among service members is well-documented. However, little is known about whether
stressful military experiences are associated with tobacco use and alcohol misuse among
military spouses. Using 9872 service member-spouse dyads enrolled in the Millennium Cohort
Family Study, we employed logistic regression to estimate the odds of self-reported cigarette
smoking, risky drinking, and problem drinking among spouses by service member deployment
status; communication regarding deployment; and stress associated with military-related
experiences while adjusting for demography, mental health, military experiences, and service
member military characteristics. Current cigarette smoking was reported by 17.2%, risky
drinking by 36.3%, and problem drinking by 7.3% of military spouses. Current deployment was
not found to be associated with spouse smoking or drinking behaviors. Communication about
deployment experiences with spouses was associated with lower odds of smoking, but not
with risky or problem drinking. Spouses bothered by communicated deployment experiences,
and those who reported feeling very stressed by a combat-related deployment or duty
assignment, had consistently higher odds of both risky and problem drinking. These results
demonstrate how a deployment experienced by military spouses is associated with their
smoking and alcohol misuse. A focus on deployment occurrence without consideration of a
spouses experience may miss important correlates of this exposure.
S570 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Daniel Trone, PhD, Military Population Health Directorate / Naval
Health Research Center, San Diego, CA, 92106-3521; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S571

F049 11:45 AM-12:45 PM

ASSOCIATIONS BETWEEN EXPERIENTIAL AVOIDANCE AND PSYCHOLOGICAL DISTRESS IN

OVERWEIGHT AND OBESE INDIVIDUALS

Danielle Wischenka, MA

Yeshiva University, lawrence, NY

Background: Experiential avoidance (EA) is a pattern whereby individuals attempt to avoid

experiencing physical sensation, thoughts and feelings. This avoidance pattern is theoretically
related to exercise dependence, disordered eating, depression, anxiety, and stress; however,
these relationships remains unclear in subclinical populations. Understanding the associations
between these variables in groups such as pre-obese and obese individuals can help clinicians
and researchers better individualize recommendations to reduce weight and effectively
change behavior. Thus, the purpose of this study was to examine experiential avoidance
strategies and other related variables in a sample of approximately 30% overweight and 32%
obese individuals using BMI and body dissatisfaction as weight measures.

Method: Participants (N=1075; Mean Age = 42.05, SD = 15.37, 60% female) completed an
online survey that included the Multidimensional Experiential Avoidance Questionnaire
(MEAQ), Eating Aptitudes Test (EAT-26), and the Depression Anxiety Stress Scale (DASS-21).
Correlation and regression analyses were used to identify associations between experiential
avoidance, eating pathology, psychological distress variables and weight measures.

Results: BMI and body dissatisfaction were related to depression (r= .107, p< .001, N= 1068;
r= .135, p< .001, N= 1067), anxiety (r= .121, p< .001, N= 1069; r= .122, p< .001, N= 1068),
stress (r= .067, p= .027, N= 1068; r= .087, p= .005, N= 1067), and overall psychological distress
(r= .103, p= .001, N= 1061; r= .124, p< .001, N= 1060) levels in the sample. Mean comparisons
further indicated that overall psychological distress levels increased across BMI categories as
participants BMI increase. Further analyses indicate that individuals psychological distress
level was related to their levels of experiential avoidance (r= .454, p< .001, N= 1057). An
inverse relationship between BMI and obligatory exercise was found (r= -.191, p< .001, N=
1057), indicating that individuals with higher BMI have less compulsive desire to engage in
exercise. A multiple regression model with disordered eating pathology, overall experiential
avoidance, and specific experiential avoidance strategies of procrastination, distraction and
suppression, and behavioral avoidance as predictors of psychological distress: R2= .305, F(5,
1056) = 92.59, p< .001 .

Discussion: Results suggest the associations among disordered eating pathology,

psychological distress, and experiential avoidance require further research and understanding,
S572 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

particularly in the treatment and care of overweight and obese individuals. This study may
inform researchers and clinicians on ways to better assess and treat individuals with weight
and eating difficulties, and further analyses are expected to explain the relationship between
these theoretically related constructs.

CORRESPONDING AUTHOR: Danielle Wischenka, MA, Yeshiva University, lawrence, NY, 11559;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S573

F050 11:45 AM-12:45 PM

EFFECTS OF BEHAVIORAL WEIGHT-LOSS TREATMENT ON SYMPTOMS OF FOOD ADDICTION IN

ADULTS WITH OBESITY

Eliza L. Gordon, M.S., Vivian Bauman, B.S., Aviva H. Ariel-Donges, M.S., Kathryn Ross, PhD,
MPh, Michael Perri, PhD, ABPP

University of Florida, Gainesville, FL

Introduction: Food addiction (FA) involves the compulsive consumption, despite adverse
consequences, of highly palatable foods such as those high in sugar or fat. In this study, we
examined the effect of behavioral weight-loss treatment on FA symptoms, and we explored
potential correlates of decreases in FA symptoms, including weight loss and changes in the
consumption of sweets and fruits.

Methods: The study sample included 127 adults (82.7% women, 83.5% Caucasian; MSD; age
= 55.610.4 years, BMI = 36.5 .5 kg/m2) who completed a 4-month behavioral weight-loss
program. The Yale FA Scale (YFAS), a 25-item self-report questionnaire using modified DSM-IV
substance dependence criteria, was administered pre- and posttreatment. Consumption of
sweets and fruits were measured by the Rapid Eating Assessment for Participants-Shortened
Version (REAPS). Analyses were conducted using bootstrapped one-way ANOVAs and
bootstrapped hierarchical regression.

Results: At baseline, participants endorsed a mean of 2.4 out of 7 FA symptoms, and 14.2%
met the YFAS criteria for FA (3 symptoms coupled with distress/impairment). The
intervention produced a mean weight loss of 8.7% initial body weight (p < .001). From pre- to
posttreatment, the mean number of FA symptoms fell to 1.5 (p < .001) and the proportion of
participants who met the YFAS diagnostic criteria decreased to 2.4% (p = .001). Reduction in
FA symptoms was associated with reduced frequency of sweets consumption (B = -.446, p =
.011) and increased fruit intake (B = -.408, p = .005), but weight change was not significantly
related to FA symptom change (B = -.332, p = .900). The following blocks were entered in the
hierarchical regression: 1) racial status, baseline BMI; 2) weight change; 3) change in sweets
consumption, change in fruit intake. Racial status and baseline BMI explained a significant
amount of the variability in FA symptom changes (R2 = .073, p = .009); weight change did not
contribute significantly (R2 change = .012, p = .201), but changes in consumption of sweets
and fruit explained a significant amount of variability beyond weight change (R2 change =
.112, p < .001).
S574 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Discussion: The findings in this study showed that participation in behavioral weight-loss
treatment significantly decreased FA symptoms. However, the reduction in FA symptoms was
not related to weight loss but was associated with increased intake of fruits and decreased
consumption of sweets.

CORRESPONDING AUTHOR: Eliza L. Gordon, M.S., University of Florida, Gainesville, FL, 32611;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S575

F051 11:45 AM-12:45 PM

EVALUATING A TARGETED, UNIVERSAL MIDDLE SCHOOL PROGRAM FOR CHILDHOOD

OVERWEIGHT AND OBESITY: STAYINGFIT

Myra Altman, MA1, Jacqueline Hayes, MA1, C. Barr Taylor, Ph.D.2, Denise Wilfley, Ph.D.1

1
Washington University in St. Louis, St. Louis, MO; 2Palo Alto University/Stanford Medical
School, Palo Alto, CA

Childhood obesity is a pressing public health concern associated with significant medical and
psychosocial comorbidities. Intervention is crucial, and schools are often suggested as an
optimal venue through which to intervene, although the results of such interventions are
inconsistent and mixed. The present study sought to expand on the literature by evaluating
the effectiveness of an evidence-based, online, targeted and universal school-based obesity
intervention, StayingFit, in three middle schools in a rural community of low-socioeconomic
status. The intervention was tested in both a cluster-randomized (Study 1, N=524) and pre-
post (Study 2, N=307) design. StayingFit was feasible to implement and was generally liked by
students and teachers, however, outcome results were limited and inconsistent between
Study 1 and 2.In Study 1, the interventionresulted in decreases in fruit juice (B=-0.01,
pStayingFit was ineffective in creating changes in relative weight, diet, physical activity, sleep,
or psychosocial outcomes (all other ps>.05). Possible reasons for the limited effectiveness of
the intervention include implementation challenges, the nature of the population and
community, insufficient use of behavioral strategies and technology, limited program
duration, low completion rates, and lack of parental engagement. This study contributes to
the mixed findings in the school-based intervention literature, and highlights many of the
challenges encountered when implementing interventions in high-risk communities. There is a
pressing need for further research to design and test multi-level and multi-sector
interventions that can create meaningful and necessary change in these communities.

CORRESPONDING AUTHOR: Myra Altman, MA, Washington University in St. Louis, St. Louis,
MO, 63110; [email protected]
S576 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F052 11:45 AM-12:45 PM

FOOD SELECTION DURING BINGE EPISODES IN HISPANICS WITH BINGE EATING DISORDER

Abigail Alido, B.A.1, Sylvia Herbozo, Ph.D.2, Dean Lim, M.A.3

1
Department of Psychology, Loma Linda University, Eastvale, CA; 2Loma Linda University,
Loma Linda, CA; 3Department of Psychology, Loma Linda University, Rancho Cucamonga, CA

Binge eating disorder (BED) has been identified as the most prevalent eating disorder among
Hispanics. This ethnic group has also shown high rates of obesity, a medical condition strongly
associated with BED. Yet, few studies have examined the type of foods consumed during
binge episodes and such studies have consisted of primarily Caucasian samples limiting the
generalizability of the findings to other ethnic groups. Given that recent research has shown
no ethnic differences in BED, studies investigating nutritional aspects of binge eating episodes
among ethnic minorities, especially Hispanics who are obese, are needed. We examined the
food selection during binge episodes for Hispanics with BED and obesity. Participants were 14
Hispanic adults aged 20 to 62 (M = 47.21, SD = 13.17) with a mean body mass index of 36.46
(SD = 6.34). Approximately 93% of the sample was born in the U.S. Participants were
administered the Eating Disorder Examination in which they were asked to describe foods
consumed during a binge episode. Two raters coded the binge episodes to identify which food
pyramid categories (e.g. grains, fruits, vegetables, meats and beans) and food types (e.g.
Bread/pasta/tortillas, potatoes, rice, sweet snacks, salty snacks, high-fat meat items) were
consumed. The two raters reached 100% agreement after discussing the few coding
discrepancies. The predominant food consumed during the binge episode was identified by
food pyramid category as well as food type. Participants had an average of 3.22 binge
episodes per week (SD = 2.69) for the past three months. Of the binge episodes, 85%
contained high-fat meat items, 68% contained sweet snacks, and 64% contained
breads/pastas/tortillas. For food pyramid categories, the most common predominant food
consumed during a binge episode was meats and beans (64%) followed by grains (21%).
Specifically, when examining type of food consumed during a binge episode, high-fat meat
items were the most common predominant food (53%) followed by bread/pasta/tortillas
(14%). Study findings indicate that high-fat meats and sweet snacks are common binge foods
in Hispanics with BED and obesity which partially supports prior research with primarily
Caucasian samples. Additional research is warranted to further determine food selection
patterns during binge episodes in this ethnic minority group. Such research could provide
valuable information that can be utilized to modify BED treatments for Hispanics.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S577

CORRESPONDING AUTHOR: Abigail Alido, B.A., Department of Psychology, Loma Linda

University, Eastvale, CA, 92880; [email protected]
S578 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F053 11:45 AM-12:45 PM

OBESITY STIGMA: EFFORT, ANXIETY, AND COGNITIVE FUNCTIONING

Jacob M. Burmeister, PhD1, Afton M. Koball, PhD, ABPP2, Sarah Domoff, PhD3, Robert Carels,
Ph.D., MBA4

1
University of Dayton, Dayton, OH; 2Gundersen Health System, La Crosse, WI; 3Central
Michigan University, Mt. Pleasant, MI; 4East Carolina University, Greenville, NC

Background: Health-related stigma is a fundamental yet understudied contributor to health

problems. Among health conditions that are stigmatized, obesity ranks near the top. Negative
effects of obesity stigma have been found in healthcare, education, employment, and
interpersonal relationships. Health behaviors have been shown to be directly affected by
experiences of stigma as demonstrated by several experiments in which acute stigmatization
leads individuals to eat unhealthy foods and engage in less physical activity. Less is known
about the mental processes that mediate these effects and how experiencing stigmatization
might affect cognitive functioning. Two experiments explored these effects.

Methods & Results: Experiment 1 tested the hypothesis that being stigmatized for having
obesity acutely affects overweight individuals' effort compared to normal weight participants
and controls. In a laboratory setting, participants (N = 97 ) were randomly assigned 1 of 3
conditions. They received either mildly stigmatizing information about their weight status,
strongly stigmatizing information about obesity stereotypes, or no stigmatizing information.
The dependent variable of effort was measured by the amount of time participants spent
attempting to solve very difficult anagram tasks. No between group differences were
observed (p > .05). Experiment 2 was of similar design (N = 155), but with different dependent
variables: implicit bias toward anxiety-related words and working memory. Deleterious
effects were hypothesized for overweight individuals vs normal weight participants and
controls. Implicit bias toward anxiety-related words was measured using a computerized dot
probe task. Working memory was measured using a computerized operation span memory
task. No between group differences were observed (p > .05).

Conclusions: In both experiments, participants who were overweight fared no worse in these
realms than other participants after being stigmatized for their weight. All participants who
were stigmatized performed as well as controls regardless of level of stigmatization. These
findings suggest a limit exists to how badly individuals are affected by experiencing acute
stigma. This evidence contradicts other studies that do show individuals suffer from acute
stigmatization and a broad literature that suggests cumulative effects of stigma are
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S579

detrimental. It is possible that chronic stigmatization leads to problems not apparent in acute
stigmatizing episodes. It may also be that health stigma only affects individuals in specific
relevant behaviors as in stereotype threat phenomena.

CORRESPONDING AUTHOR: Jacob M. Burmeister, PhD, University of Dayton, Dayton, OH,

45469; [email protected]
S580 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F054 11:45 AM-12:45 PM

UNDERSTANDING OF DYNAMIC INTERRELATIONS BETWEEN POSITIVE EMOTIONS AND

CRAVINGS FOR UNHEALTHY SNACKS AND DRINKS

Bin Xie, PhD1, Kim Reynolds, PhD1, Anna Lee, MPH, MA1, Zhongzheng Niu, MS, MD1, Stacey
Doan, PhD2, Alan Stacy, PhD1

1
Claremont Graduate University, Claremont, CA; 2Claremont McKenna College, Claremont, CA

Background: Few efforts have been made to investigate dynamic interrelations between
positive emotions and cravings for unhealthy snacks and drinks among adolescents.
Ecological momentary assessment (EMA) has been recommended to acquire ecologically valid
real-time information on snacking behaviors and related psychosocial factors. The time-
varying effect model (TVEM) allows regression coefficients, which quantify behavioral
associations, to vary over time and to be expressed dynamically and graphically, as smooth
functions of time. This model therefore provides a novel approach for modeling dynamic
interrelations. In this study we analyzed data from a longitudinal assessment of momentary
ratings of four types of positive emotions and craving for unhealthy snacks and drinks among
adolescents with the purpose of advancing understanding of their dynamic interrelations over
time.

Methods: EMA data were collected during a 7-day period on handheld PDA devices among
158 adolescents aged 14 to 17 years. Time-varying measurements included momentary self-
reported scores of positive emotions (i.e. feeling relaxed, happy, energetic and
cheerful, scores ranging 0-100) and self-reported scores of three craving outcomes (i.e.
craving for a sweet snack, a salty snack and a sweetened drink, scores ranging 0-100).
Multilevel random-effect models and TVEM models were employed in the longitudinal
analysis with adjustment for gender, ethnicity, age and obesity status.

Results: Results from the multilevel random-effect models revealed significant positive
associations of positive emotion scores with scores of craving for a sweet snack (feeling
relaxed: =0.04, p=0.03; feeling happy: =0.07, p=0.003; feeling energetic: =0.10, p0.0001;
feeling cheerful: =0.08, p=0.001) and with scores of craving for a sweetened drink (feeling
relaxed: =0.08, p=0.0001; feeling happy: =0.14, p

Conclusions: Our study demonstrated an initial effort to understand the process of time-
varying interrelations between momentary assessments of positive emotions and craving for
unhealthy snacks and drinks among adolescents. A better understanding of the underlying
mechanisms for this dynamic process may help foster development of effective interventions.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S581

CORRESPONDING AUTHOR: Bin Xie, PhD, Claremont Graduate University, Claremont, CA,
91709; [email protected]
S582 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F055 11:45 AM-12:45 PM

USE OF A NOVEL TECHNOLOGY-BASED APPROACH TO HARNESS SOCIAL NETWORKS FOR

WEIGHT LOSS IN ADULTS: PILOT RANDOMIZED CONTROLLED TRIAL

Courtney M. Monroe, PhD, EP-C, Chelsea Larsen, MPH, CHES, Delia S. West, PhD

University of South Carolina, Columbia, SC

Engaging social support and giving individuals tangible resources are methods that have been
found to improve weight loss outcomes. This study examined the effect of leveraging existing
social networks through the provision of advanced technologies for increasing weight losses
over current best practices. Adults were randomized to a 16-week, evidence-based, standard
behavioral weight loss treatment (SBT; n=18) or SBT plus social support resources for weight
loss (Enhanced; n=18). Both groups received weekly, in-person group counseling sessions
targeting dietary and physical activity (PA) behavior change, a Fitbit Zip PA tracker, digital
body weight scale, and access to a study website. The Enhanced group also received
additional Fitbit Zips and digital scales which they were asked to share with up to two people
of their choice within their social circle. Repeated measures ANOVA was used to compare
baseline and 16-week measured weight, social support for diet and PA (measured on a 5-point
Likert scale with higher scores indicating greater support), and reported number of
appropriate weight control practices. X2analysis was used to measure the group difference in
the proportion of participants who lost >5% of their initial weight. Participants averaged
45+8.9 years, had a BMI of 36.1+7.3 kg/m2 and were 94% female and 39% African-American.
Intent-to-treat analyses showed significant weight decreases (p < .001) in both SBT (-3.8+3.6
kg) and Enhanced (-4.6+5.1 kg), with no difference between groups (p=.56). The proportion
losing >5% of initial weight was similar between groups (33% for SBT vs 50% for Enhanced,
p=.31). In SBT and Enhanced, respectively, the increases (ps < .05) in social support from pre
(3.0+0.6 and 3.3+0.4 for diet; 2.9+0.9 and 2.9+0.6 for PA) to post (3.4+0.7 and 3.6+0.6 for diet;
3.2+0.9 and 3.4+0.7 for PA) were not different (ps>.05). At 4 months, SBT reported using
14.9+3.0 appropriate weight control practices (out of 23) versus 13.9+3.7 for Enhanced, an
increase from baseline (p < .001), with no between group difference (p=.60). Simply providing
advanced technologies to participant-selected support partners did not enhance weight losses
beyond current best practices. A similar but less minimalistic approach (e.g., providing
resources and support partner guidance) may be more effective. Future research should
continue to explore how to best capitalize on the joint influence of technologies and social
support to increase weight losses in larger samples.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S583

CORRESPONDING AUTHOR: Courtney M. Monroe, PhD, EP-C, University of South Carolina,

Columbia, SC, 29208; [email protected]
S584 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F056 11:45 AM-12:45 PM

DETERMINANTS OF PYSCHOSOCIAL VARIABLES AND PHYSICAL ACTIVITY AMONG ADULTS

WITH KNEE OA PAIN: A SYSTEMATIC REVIEW OF THE EVIDENCE

Erica Anderson, BS, Bridgett Rahim-Williams, MA, MPH, PhD

Bethune-Cookman University, Daytona Beach, FL

Background: Knee OA is a chronic degenerative disease affecting functional mobility and

psychosocial variables. While it is recognized that psychosocial variables such as depression
and anxiety are associated with individuals who have knee OA pain, no systematic literature
review has focused on examining the effect physical activity has on these psychosocial
variables. The aim of this systematic literature review was to determine whether physical
activity has an effect on psychosocial factors such as depression and anxiety among adults
with knee OA pain.

Methods: A literature search using PubMed, CINAHL, Research gate, Sage database, and
references of retrieved articles was conducted. Studies published from 2006-2016(through
September) using a predefined search criteria: women with knee OA, mention of an
anxiety/depression scale, pain assessment, addressed the role of physical activity were
reviewed.

Results: The search terms: Physical Activity, Knee OA yielded 5,621 studies. After adding key
terms: Women, Anxiety and Depression, and Pain, 537 studies resulted. Of the 537 studies,
530 studies were excluded. The exclusion criteria included such studies: that did not mention
an anxiety or depression scale, physical activity, knee OA was not specifically mentioned,
pre/post-operative total knee replacement and somatic pain. Seven (7) studies were deemed
appropriate for the present review and included a diverse population by race/ethnicity, age,
sex/gender, geographic location and type of physical activity (Tai Chi, Yoga, Exergames, and
Walking). Two studies only included women. One study included both men and women, and
four studies did not mention sex/gender. Racial/ethnic diversity included African Americans,
Hispanics, Asians and Caucasians.

Conclusion: Current published studies revealed a positive relationship between psychosocial

variables, physical activity and knee OA pain. However, there is a dearth of literature
investigating the effect of physical activity and psychosocial variables in this population.

CORRESPONDING AUTHOR: Erica Anderson, BS, Bethune-Cookman University, Daytona Beach,

FL, 32114; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S585

F057 11:45 AM-12:45 PM

EVALUATION OF INTEGRATIVE PAIN MANAGEMENT ON MOOD, TREATMENT ENGAGEMENT,

AND TRADITIONAL MEDICAL VISITS IN VETERANS

Melody Mickens, Ph.D., Leah Farrell-Carnahan, Ph.D., Meghan Rooney, Psy.D., Jay Wiles, M.S.,
Jina Wang, Pharm.D.

Hunter Holmes McGuire VA Medical Center, Richmond, VA

Background: Innovative chronic pain management programs, such as the Veterans Integrative
Pain (VIP) Center, are developing within Veterans Affairs Medical Centers. This short-term
outpatient program blends traditional medical, complementary, and alternative approaches.
It emphasizes self-management training (e.g., cognitive behavioral therapy, mindfulness
meditation, biofeedback, dietary change) supplemented with passive approaches (e.g.,
acupuncture) aiming to empower Veterans with non-pharmacologic options. At intake,
Veterans select from 13 treatments and engage in multiple services. The program aims to
improve Veterans functional capacity, mood, anxiety, and quality of life. The current study is
an exploratory analysis of these outcomes.

Methods: Participants were Veterans with chronic pain (N=38) who participated in the
program for at least 3 months. Demographic characteristics, program engagement, and
number of medical visits three months before and three months after intake were obtained
via chart review. Measures of pain catastrophizing (PCS), pain interference (POQ), and
mood/anxiety symptoms (BAI, PHQ-9) were obtained at intake. Descriptive statistics were
performed to characterize all variables. T-tests were conducted to explore changes in number
of medical visits before and during the program participation period.

Findings: Veterans were predominately African-American (58%) males (71%) with a mean age
of 49 years (SD=15). On average, Veterans engaged in 3 services (SD=2; range = 0-7),
predominately auricular acupuncture (68%), biofeedback (42%), full body acupuncture (39%),
anti-inflammatory diet education (24%), and mindfulness meditation (21%). Veterans had 0-6
medical visits prior to program and this reduced to 0-3 visits post program. However, these
differences were not statistically significant. At intake, they reported subclinical pain
catastrophizing (M=26, SD=11), mild functional interference (M=91, SD=29), moderate
depression (M=13, SD=6), and severe anxiety (M=19, SD=12). Veterans reporting higher
anxiety were more likely to engage in auricular acupuncture, t(27) = 2.85, p < .01 and those
reporting higher pain catastrophizing and anxiety were significantly more likely to engage in
auricular acupuncture, t(23) = 3.42, p < .05 and qigong, t(23)=2.98, p < .05.
S586 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions: As anxiety and pain catastrophizing increased, engagement in integrative

approaches did as well. The absence of a significant decrease in traditional medical office
visits in conjunction with participation is not surprising given the medical-complexity of this
sample. Future investigations should investigate this further and consider other variables
including disease comorbidity and change in opioids use.

CORRESPONDING AUTHOR: Melody Mickens, Ph.D., Hunter Holmes McGuire VA Medical

Center, Richmond, VA, 23224; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S587

F058 11:45 AM-12:45 PM

PATTERNS AND PREDICTORS OF PAIN FOLLOWING LUNG TRANSPLANTATION

Julia Farquhar, B.S., MD candidate1, Patrick Smith, PhD2, James Blumenthal, PhD2

1
Duke University School of Medicine, Durham, NC; 2Department of Psychiatry and Behavioral
Sciences/Duke University Medical Center, Durham, NC

Introduction: Lung transplantation, not unlike other major surgeries, is associated with
significant acute pain for most patients. However, variability in pain levels during
hospitalization and predictors of pain levels following hospital discharge have not been
systematically evaluated.

Methods: We conducted a retrospective study of a series of 167 patients (39% female; 88%
Caucasian) transplanted at Duke University Hospital between January 2015 and September
2016. Patients subjective ratings of pain intensity using a 0-10 Numeric Rating Scale (NRS)
were obtained daily during patients hospitalization and 2 months after hospital discharge.
Demographic variables (age, gender, and ethnicity), native lung disease (e.g., cystic fibrosis,
chronic obstructive pulmonary disease, interstitial lung disease), history of alcohol, narcotic,
or cocaine abuse, psychiatric conditions, and chronic pain, and co-morbid medical conditions
were examined as predictors of pain after transplant surgery.

Results: Of 167 patients who were transplanted, 150 were discharged (median length of stay
= 13 days [IQR = 15 days]), 10 died, and 7 remained hospitalized. In a mixed-effects, multiple
regression model, pain ratings decreased over time (P < .001) and were higher among patients
undergoing bilateral lung transplant (P =.012). Among pre-transplant predictors, history of
depression was predictive of higher pain levels and slower resolution of postoperative pain
(P=0.021). Two months after discharge, 18% of patients continued to report pain and a third
remained on narcotic pain medications. Cystic fibrosis patients reported higher levels of pain
at follow-up compared to patients with other native lung diseases (P=.006).

Conclusion: Although acute pain resolved following surgery for most lung recipients, nearly
one in five patients continued to experience persistent pain 2 months following hospital
discharge. Patients with cystic fibrosis or a history of depression appear especially vulnerable
to chronic pain and may require more intensive efforts to help them manage pain once
discharged.

CORRESPONDING AUTHOR: Julia Farquhar, B.S., MD candidate, Duke University School of

Medicine, Durham, NC, 27705; [email protected]
S588 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F059 11:45 AM-12:45 PM

PREDICTING ABERRANT MEDICATION-RELATED BEHAVIOR IN ADOLESCENT CANDIDATES FOR

OPIOID THERAPY: UPDATE ON SCREENER DEVELOPMENT

Kelly A. Manser, B.A.1, Stacey McCaffrey, PhD1, Kimberlee J. Trudeau, Ph.D.1, Deirdre Logan,
PhD 2, Jean C. Solodiuk, RN PhD3, Ryan Black, PH.D.1, Jennifer Harman, PhD4, Kelly Smith, BA5,
Stephen Butler, PhD6

1
Inflexxion, Inc., Waltham, MA; 2Boston Children's Hospital, Boston, MA; 3Boston Childrens
Hospital, Boston, MA; 4St. Jude Childen's Research Hospital, Memphis, TN; 5Boston Children's
Hospital, Waltham, MA; 6Inflexxion, Inc, Waltham, MA

Pediatric providers have voiced a need for a screening tool to assess risk of aberrant
medication-related behavior (AMRB) in adolescents who may receive opioid therapy for
chronic pain. Having previously identified indicators of ongoing AMRB in this population
(Trudeau et. al., 2016), we interviewed clinicians as well as current/former patients and
caregivers (CFP) to identify observable predictors (red flags) that suggest elevated risk of
future AMRB in adolescents with chronic pain.

Thirteen pediatric clinicians (experts) and 29 CFP completed semi-structured interviews.

Experts discussed red flags they associate with future AMRB risk; CFP described their
experiences and/or difficulties related to opioid therapy. AMRB-predictive statements were
extracted from interviews, coded by content, and consolidated. Fifteen clinicians used group
concept mapping to sort the final 101 statements into thematic piles and rated on a 5-point
Likert scale how confident they were (1=Not at all, 5=Extremely) that each statement was
predictive of AMRB.

Concept maps with 5-9 clusters were generated; results supported an 8-cluster model.
Clusters were: 1) Daily functioning, M=2.53; 2) Resistance in treatment context, M=2.78; 3)
Pre-existing illicit substance use, M=3.67; 4) Psychological distress, M=2.66; 5) Maladaptive
family dynamics, M=2.78; 6) Concerning opioid-specific patient behaviors, M=4.05; 7)
Behaviors causing concern re: opioid use at home, M=3.92; 8) Caregiver not engaging
appropriately, M=3.33. Clusters 6 and 7 were each rated significantly higher (i.e., more likely
to predict AMRB) when compared individually to each of the other six clusters (all ts < 13.2,
all ps < .005). Statements derived from CFP interviews were rated significantly higher than
those originating from experts, t(99)=2.09, p=.04.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S589

This study extends our previous findings: providers see risky behavior specifically related to
substance (in this case, opioid) use as highly correlated with AMRB in adolescent pain
patients. The psychosocial issues represented in Clusters 1, 2, and 4 seem to be more sensitive
than specific as AMRB predictors; some (e.g., amotivation, social isolation) may be a natural,
expected reaction to pediatric chronic illness. Lastly, CFP willing to discuss opioid use may
provide information that is especially predictive of AMRB. These findings continue to inform
our development of a risk assessment tool; a conceptually-derived checklist is currently under
expert review.

CORRESPONDING AUTHOR: Kelly A. Manser, B.A., Inflexxion, Inc., Waltham, MA, 02451;
[email protected]
S590 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F060 11:45 AM-12:45 PM

PRE-SURGICAL CORRELATES OF POST-SURGICAL PAIN ACCEPTANCE AMONG UTAH WORKERS

COMPENSATION PATIENTS UNDERGOING LUMBAR FUSION

M. Scott DeBerard, Ph.D.1, Jessica Gundy Cuneo, Ph.D.2, Cassie Dance, M.A.1, Julie L. Murray,
BA1, Seth Seifert, M.A.1, JoAnn Tschanz, Ph.D.1

1
Utah State University, Logan, UT; 2UC San Diego Health System, San Diego, CA

Lumbar fusion surgery is an increasingly prevalent and costly procedure in the U.S. Given
outcomes following this surgery are quite variable, there is an effort to identify pre-surgical
biopsychosocial variables that may predict lumbar fusion outcomes (e.g., litigation,
depression, pre-surgical pain, and etc.). Pain acceptance is a more recently examined
construct shown to correlate with multiple patient outcomes. Identifying pre-surgical
predictors of pain acceptance is likely relevant for purposes of pre-surgical screening. The
present study examines how pre-surgical demographic, health status, and compensation
variables correlate with post-surgical measures of pain acceptance. Participants were 101
Utah workers compensation patients (average age = 42.4 yrs., 75.2% male) who underwent
lumbar fusion surgery from 1987-2007 and were at least two-years post-surgery at time of
follow-up. A retrospective cohort design was utilized consisting of a review of pre-surgical
records and a postsurgical telephone survey. Presurgical variables included: gender, age,
educational level, smoking, BMI, prior history of depression, and litigation status. The
outcome measure was the Chronic Pain Acceptance Questionnaire (CPAQ) that assessed
patient willingness to experiences pain despite the associated negative cognitive and
psychological consequences and to continue to engage valued life activities. This measure
included a total summary scale as well as subscales of Pain Willingness (PW) and Activity
Engagement (AE). Correlational analyses revealed that male gender, lower levels of
education, active smoking, positive history of depression and presence of litigation were all
statistically significant correlates of lower CPAQ total and subscale scores. Pre-surgical
variables were used to predict CPAQ scores using regression analyses. Statistically
significant amounts of variance were predicted in CPAQ Total (R2 = 0.14), PW (R2 = 0.17) and
AE (R2 = 0.12%) scales. The benefits of integrating pre-surgical predictors of pain acceptance
into patient selection protocols are discussed.

CORRESPONDING AUTHOR: M. Scott DeBerard, Ph.D., Utah State University, Logan, UT,
84322-2810; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S591

F061 11:45 AM-12:45 PM

PSYCHOMETRIC PROPERTIES OF THE MINDFUL ATTENTION AWARENESS SCALE (MAAS) IN A

GREEK-CYPRIOT SAMPLE OF CHRONIC PAIN PATIENTS

Marianna Zacharia, M.Sc., Vasilis Vasileiou, M.A., Maria Karekla, Ph.D.

University of Cyprus, Nicosia, Nicosia, Cyprus

Mindfulness as a concept has received great interest from the therapeutic community,
especially in the area of chronic pain alleviation. However, well validated measures to assess
changes in mindfulness especially in the Greek language are lacking. The purpose of this paper
is to examine the factor structure and psychometric properties of the Mindful Attention
Awareness Scale (MAAS) in Greek-Cypriot chronic pain patients. The sample (N = 133) was
derived mainly from the Antirheumatoid Association of Cyprus and from various other health
care centers. The forward-backward translation method was applied. Initially, and following
the reported structure of the original English version, all 15 items were forced to a single-
factor solution using exploratory factor analysis (EFA), which explained 32% of the total
variance. EFA was then conducted without forcing factor extraction to investigate the
existence of other factors explaining a larger percentage of the variance. Results supported a
four-factor solution based on Kaisers criterion, explaining 59.50% of the total variance. This
finding was inconsistent with the sudden change on the scree plot chart emerging after the
first factor. The clustering of items into the four factors does not appear to make sense
conceptually. Parallel analysis was also performed, which corroborated the ambiguity of the
scree plot, suggesting either two-factor or three-factor solutions. Therefore, items were
forced into two-factor and three-factor solutions using EFA. Results indicated that the
difference between the eigenvalues of the first factor and subsequent factors was large, with
subsequent factors either showing negative or weak loadings and being comprised of a small
number of items. Thus, consistent with previous empirical studies, a single-factor solution
appears to be most suitable. G-MAAS exhibited good internal consistency (Cronbachs =
.83). G-MAAS showed relevant convergent validity with CAMS-R, a mindfulness measure and
theoretically coherent significant negative correlations (discriminant validity) with measures
of anxiety (HADS-anxiety), distress and non-acceptance (AAQ-II). G-MAAS appears to be a
valid and reliable measure for assessing mindfulness in Greek-speaking chronic pain patients.
The dimensionality of the G-MAAS, directions for future research and clinical implications are
discussed.

Keywords: Mindfulness, MAAS-G, Reliability, Validity, Chronic Pain Patients

S592 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Marianna Zacharia, M.Sc., University of Cyprus, Nicosia, Nicosia,

1678; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S593

F062 11:45 AM-12:45 PM

AN IN-HOME VIRTUAL COACHING PROGRAM TO PROMOTE PHYSICAL ACTIVITY IN DIVERSE

POPULATIONS: USER TESTING AND ACCEPTABILITY

Aldo Chazaro, BS1, Sandra Jane. Winter, PhD, MHA2, Vincent ChangChien, M.S.3, Bo Yu Hou,
M.S.4, Abby C. King, PhD1

1
Stanford School of Medicine, Palo Alto, CA; 2Stanford University School of Medicine, Palo
Alto, CA; 3Stanford University School of Medicine, Chutung, Hsinchu, Taiwan (Republic of
China); 4Industrial Technology Research Institude, Chutung, Hsinchu, Taiwan (Republic of
China)

Background: Physical inactivity due to lack of access to gyms, personal trainers, fitness
programs and other physical activity (PA) resources can lead to increased risk of chronic
disease. To address this, researchers at Stanford University worked with a team from the
Industrial Technology Research Institute of Taiwan (ITRI) to test a virtual PA coaching program
(vCoach), with the aim of encouraging people to engage in PA in the comfort of their own
homes. The purpose of this pilot was to learn if this type of technology is acceptable to US
residents.

Setting: vCoach testing was conducted in an exercise laboratory at Stanford. The vCoach
system, located in a device roughly the size of a cable box, was connected to a tabletop 40
HD TV, that was used to display exercise modules provided by ITRI to participants. A camera
sensor and wearable smart watch were utilized to display information to participants on the
TV screen, such as calories burned, heart-rate, and similarity rating of participants exercise
movements to the v-coach model.

Method: Individuals, recruited via online ads, completed the PAR-Q to ensure that they could
safely perform PA. Participants were categorized according to self-reported PA of either low,
medium or high intensity. One-time testing was conducted at the exercise lab by trained staff
and lasted 1-2 hours. All participants did warm up and cool down exercises and activity
modules matched to their self-reported intensity levels. On completion, all participants
provided feedback on the modules and vCoach software.

Results: Participants (n=15, mean age=44) included 7 women and 7 Latinos. 93% rated their
health as good or better. They reported their usual PA as low (n=6), medium (4), or high (5).
93% reported that the vCoach was appropriate for them, and 93% said they would use it if
they had access. 66% reported they preferred seeing a human coach rather than an avatar
S594 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

coach, and 80% had no preference for gender, age, or ethnicity of the coach. Reported
barriers to using the software included technical issues, lack of efficacy in using technology
more generally, and a preference for increased visual and audio instruction.

Conclusion: While almost all participants expressed a strong interest and desire to use the
vCoach system, some participants reported technical barriers and lack of ease of use as
potential deterrents to using the software at home. Next steps include incorporating the
feedback to enhance the system and further in-home user testing.

CORRESPONDING AUTHOR: Aldo Chazaro, BS, Stanford School of Medicine, Palo Alto, CA,
94304; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S595

F063 11:45 AM-12:45 PM

ASSOCIATION OF SEDENTARY BEHAVIOR WITH SALIVARY ESTRADIOL LEVEL AMONG AFRICAN-

AMERICAN WOMEN WHO ARE OVERWEIGHT

Md Abdullah Mamun, MS1, Heather Kitzman-Ulrich, PhD2, Leilani Dodgen, MPH1, Surendra
Reddy Mandapati, BDS1, Kisa Gant, BS1, Ike Eke, BA3

1
University of North Texas Health Science Center, Fort Worth, TX; 2Baylor Scott & White
Health, Dallas, TX; 3University of North Texas Health Science Center, fort Worth, TX

Objective

Recent studies have reported that sedentary behavior may have multiple adverse health
outcomes in adults, and is related to elevated levels of estradiol that are associated with
breast cancer, ovarian cancer, and endometrial cancer. A growing body of research has
assessed the association of sedentary behavior with estradiol levels, however few studies
have been in African-American (AA) overweight women who are disproportionally diagnosed
with cancer. The objective of this study was to assess the association of self-reported
sedentary behavior of AA overweight women with salivary estradiol level.

Method

We recruited 263 AA overweight (BMI>25, mean BMI 36.8 (SD 8.5), mean age 49.4 (SD 11.6)
years) women from a faith-based weight reduction program from 2014-2016 in Dallas, Texas.
Approximately 4mL of saliva was collected over four consecutive weeks. Saliva was stored in a
sub 800C freezer and sent to a lab to assay. Weekly sedentary behavior was collected with a
valid and reliable survey for weekdays and weekends that included hours of sitting to watch
television, working on the computer, riding in a car, bus, or train, and other sitting practices.
We used Pearsons-correlation and linear models to estimate the unadjusted and adjusted
association of sedentary behavior with estradiol level.

Results
S596 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Average weekly sedentary time was 40.6 hours (SD = 17.5). The range of estradiol level was
0.50 pg/mL to 3.50 pg/mL with mean 1.13 pg/mL (SD = 0.55). Sedentary time was positively
associated with estradiol level (correlation coefficient = 0.18, p < 0.01). After adjusting for age,
BMI, and menstrual status, we found a significant positive association between sedentary
time and estradiol level (R2 = 8.4%, p < 0.01).

Conclusion

Self-reported sedentary hours for AA overweight women were less than the national average
(5.8 hours vs. 7.7 hours per day). We found that sedentary time was a predictive factor for
estradiol level, and higher sedentary time was associated with elevated estradiol levels.
Future research should further explore how sedentary behavior and elevated estradiol levels
influence cancer risk in AA women.

CORRESPONDING AUTHOR: Md Abdullah Mamun, MS, University of North Texas Health

Science Center, Fort Worth, TX, 76107; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S597

F064 11:45 AM-12:45 PM

CITATION AND MERITORIOUS AWARD WINNER

ASSOCIATIONS BETWEEN SEDENTARY AND STANDING BEHAVIORS WITH COGNITIVE

FUNCTION AMONG BREAST CANCER SURVIVORS.

Sandahl H. Nelson, MS1, Catherine R. Marinac, PhD2, Lisa Cadmus-Bertram, PhD3, Jacqueline
Kerr, PhD, MSc4, Loki Natarajan, PhD4, Suneeta Godbole, MPH4, Sheri J. Hartman, PhD4

1
University of San Diego, California, La Jolla, CA; 2Harvard TH Chan School of Public Health,
Dana-Farber Cancer Institute, Boston, MA; 3Department of Kinesiology University of
Wisconsin, Madison, WI; 4Department of Family Medicine and Public Health, UC San Diego, La
Jolla, CA, USA, La Jolla, CA

Introduction: Many breast cancer survivors experience problems with cognitive functioning
following cancer treatment. Based on research in older adults, a potential risk factor for
cognitive impairments is time spent in sedentary behaviors. Since breast cancer survivors
spend the majority of their day sedentary it is an important risk factor to assess. This pilot
study examined associations between objectively measured sedentary behaviors and
neuropsychological testing in breast cancer survivors (n=30).

Methods: For 7 days participants wore a hip accelerometer to measure total sedentary time
and time in long sedentary bouts (>20min) and a thigh inclinometer to measure sit-to-stand
transitions and total standing time. Cognitive functioning was measured with the NIH
Toolbox, which provided a composite score of fluid cognitive abilities and 5 domain scores
(executive functioning, visual episodic memory, auditory episodic memory, working memory,
and processing speed), all adjusted for age. Partially adjusted linear regression models
controlled for measurement device wear time only, and fully adjusted models controlled for
measurement device wear time plus education, employment status, and moderate to
vigorous physical activity. There were no meaningful differences between the 2 models;
therefore, the fully adjusted models are presented.

Results: Participants (n=30) were, on average 62 years of age (SD=8) and 2.6 years post-
diagnosis (SD=1.1); 17 (56%) received chemotherapy. Visual episodic memory was positively
associated with time (min/day) in sedentary bouts (b=0.11, p=0.004), indicating that each 30
min/day increase in sedentary bouts was associated with a 3.3-point increase on the visual
episodic memory test. Processing speed was also positively associated with total standing
time (b=0.09, p=0.031), indicating that each 30 min per day increase in standing time is
S598 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

associated with a 2.7 point increase on the processing speed test. Executive function, working
memory, and auditory episodic memory were not significantly associated with sedentary
behaviors.

Conclusion: Longer standing time and greater time in sedentary bouts were positively
associated with episodic memory and processing speed in breast cancer survivors, with or
without controlling for education, employment status, and physical activity. Given the
counterintuitive positive association between sedentary time and cognition future studies
should examine what a person is doing while sedentary to aid in understanding this positive
association.

CORRESPONDING AUTHOR: Sandahl H. Nelson, MS, University of San Diego, California, La

Jolla, CA, 92093; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S599

F065 11:45 AM-12:45 PM

CAN AN EXERCISE PRESCRIPTION PROGRAM BE IMPLEMENTED IN FAITH BASED SETTINGS?

Corliss Solomon, MPH1, Jessica Montana, MPH2, Marcella Beam, BS2, Elizabeth Racine, PhD,
RD1, Greg Stewart, BS2, Jennifer West, MPH2

1
University of North Carolina at Charlotte, Charlotte, NC; 2Cabarrus Health Alliance,
Kannapolis, NC

Background: Improved opportunities for chronic disease prevention, risk reduction or

management through community linkages have long been a public health priority, especially
in relation to decreasing health disparities. One such strategy is increasing physical activity
through exercise prescriptions and community referrals.

Objective: The main objective of exercise prescription programs are to nudge healthcare
workers to include physical activity when developing treatment plans and referring patients to
beneficial and nearby exercise programming. Traditionally, exercise prescription programs
have been implemented in clinical settings, but this study explores the feasibility and
effectiveness of an exercise prescription program in faith based settings with Faith
Community Nurses (FCNs).

Methods: The exercise prescription program was implemented and evaluated in churches
from January to December 2016 in a North Carolina county. Churches were eligible if they had
a FCN and a predominately African American and/or Hispanic population. As a part of this
exercise prescription program, the nurses provided exercise prescriptions as a part of their
clinical encounters with church congregants. Additionally, referrals were given to local
community resources to help patients fill their exercise prescription. Prescription data was
collected in May with a 3 month follow up in September. Additionally, the public health
effects of this exercise prescription program was assessed using the RE-AIM (Reach,
Effectiveness, Adoption, Implementation and Maintenance) framework of program
evaluation.

Results: Four Faith Community Nurses at four churches with predominately Black and
Hispanic populations were trained and implemented exercise prescription programs within
their congregations. Of those patients who received exercise prescriptions (n=72), 93% were
overweight or obese, 75% had prehypertension or hypertension and more than half exercised
moderately or vigorously 2 days/week or less. The average age of participants was 60 with the
youngest participant being 21 and the oldest being 90. While there were no significant clinical
changes in weight and blood pressure, the participants physical activity levels improved.
S600 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions and Implications: While exercise prescription programs are found to be feasible,
it was not significantly effective on the health of participants and showed modest
improvement in the physical activity of participants. This may be partly due to an older sample
that have an assortment of health issues and significant barriers to physical activity. Thus,
there are challenges in capturing clinical success. Additional research is needed and should
include a larger and younger sample to explore exercise prescription programs feasibility and
effectiveness in faith settings.

CORRESPONDING AUTHOR: Corliss Solomon, MPH, University of North Carolina at Charlotte,

Charlotte, NC, 28216; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S601

F066 11:45 AM-12:45 PM

SELF-COMPASSION AND MOTIVATION FOR EXERCISE

Mandrila Das, B.A.

Texas Tech University, Lubbock, TX

Background: In a previous study, which was accepted for a paper presentation at SBM (Self-
Compassion and Mindfulness: Examining the Relationship with Disordered Eating,
Appearance-Motivated Exercise, and Depression Among College Students), researchers found
that self-compassion negatively predicted appearance-motivated exercise. Given this novel
and significant finding, researchers ran a post hoc analysis to understand how self-compassion
may predict other motivations for exercise in addition to appearance-motivated exercise.

Recent research has focused on compulsive exercise to influence body weight or shape
(Greenleaf, Petrie, Carter, & Reel, 2009). While exercise has typically been associated with
positive mental health outcomes (Stephens, 1988), this relationship may differ if one engages
in disordered eating behavior. For example, in a study with college students, exercise was
related to positive mental health outcomes for individuals who did not engage in disordered
eating, but negative outcomes for those who did engage in disordered eating (De Young &
Anderson, 2010). One explanation for this discrepancy may be embedded in the motivation
for exercise. Past research suggests that individuals who engage in exercise for appearance
management reasons may be more likely to develop eating pathology than individuals who
exercise for more health and fitness oriented reasons (La Page & Crowther, 2010).

Participants were 732 college students, both male (46%) and female (54%), ages 18-25
(M=19). Regression analysis was conducted to test main effects of self-compassion on
motivation for exercise.

Results: Self-compassion was a significant positive predictor of exercising for more positive
health-related reasons including to improve strength, to improve endurance, to increase
resistance to disease and illness, to maintain physical well-being, and to have fun. However,
self-compassion was a significant negative predictor of exercising to change ones body,
appear attractive to the opposite sex, to be sexually desirable, to cope with sadness, and to
cope with anxiety,

Conclusion: Self-compassion may increase the desire to engage in self-nourishing behaviors

and decrease the desire to exercise for externally validating reasons (i.e. to be sexually
desirable). One possible explanation, which builds on previous research, is that self-
compassion promotes a healthier relationship with ones body, which in turn may lead
S602 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

individuals to engage in exercise for more internal, physical health related reasons. While
research on self-compassion in the exercise domain is fairly limited, this study provides initial
support for further research in this area.

CORRESPONDING AUTHOR: Mandrila Das, B.A., Texas Tech University, Lubbock, TX, 79415;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S603

F067 11:45 AM-12:45 PM

THE ROLE OF EXERCISE AND SELF-ESTEEM ON BODY SELF PERCEPTION

Elisabeth O'Rourke, B.S.1, Renata Vaysman, PhD2, Leslie Halpern, PhD1

1
University at Albany, SUNY, Albany, NY; 2New York Police Department, Brooklyn, NY

Introduction. Research has shown that low self-esteem is related to poor body image;
however, whether exercise plays a role in this relationship is unclear. This study extends the
literature by examining the moderating effect of exercise on the relation between self-esteem
and body image. We hypothesized that self-esteem and BMI would be related to body image in
both males and females. Moreover, we expected that after controlling for BMI, exercise
frequency would moderate the relation between self-esteem and body image such that increased
exercise frequency would be related to better body image as self-esteem increased.

Method. The participants were 201 undergraduates, [M(age)= 19.11, N(males)= 123].
Participants completed the body areas satisfaction subscale (BAS) of the Multidimensional
Body Self Relations Questionnaire (MBSRQ; Brown, Cash, & Mikulka, 1990) and reported on
self-esteem using the Rosenberg Self-Esteem Scale (Rosenberg, 1989). BMI was calculated
based on self-reported height and weight, and participants reported the frequency of vigorous to
moderate exercise they engaged in during an average week.

Results. T-tests revealed significant sex differences on several measures. Males (M=3.55,
SD=.75) reported greater BAS than females (M=3.32, SD=.66, respectively) and higher self-
esteem than females (M=32.39, SD=5.43; M=30.88, SD=5.15, respectively). Males (M=11.52,
SD=4.41) also reported engaging in physical activity more frequently than females (M=9.67,
SD=3.98). BMI was negatively correlated with BAS in males only (r(123)=-.27, p < .01). For
males and females, BAS was positively correlated with self-esteem (r(123)=.55, p < .01;
r(75)=.42, p < .01, respectively). Frequency of physical activity was not significantly correlated
with any of the variables.

Hierarchical multiple regression analyses were used to examine the moderating effect of
exercise on the relation between self-esteem and body image in men and women. BMI was
included as a covariate on the first step of each regression. Both regressions were significant (p
< .05). Self-esteem and exercise frequency accounted for a significant amount of variance in
body image for both males and females (R2 = .322, F(1, 121) = 31.76, p < .01; R2 = .223,
F(1, 73) = 21.55, p < .01, respectively). In the final step, the interaction term was significant
only for males: R2 = .025, F(1, 119) = 5.14, p < .05. Examination of the interaction plot
showed that as self-esteem increases, BAS increases for individuals who exercise more.
S604 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Discussion. For both males and females, greater overall self-esteem relates to positive body
image, however, a moderating effect was found for exercise in males. The results indicate that
for males, exercise can enhance how they positively perceive and evaluate their bodies.
Research shows that having a positive body image has a significant positive impact on
psychological well-being and willingness to participate in certain health behviors. Therefore
these findings have widespread implications for intervention and research in behavioral
medicine and mental health. Why exercise did not show a moderating effect for females
requires further investigation.

CORRESPONDING AUTHOR: Elisabeth O'Rourke, B.S., University at Albany, SUNY, Albany, NY,
12203; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S605

F068 11:45 AM-12:45 PM

THE SELF-EFFICACY FOR PHYSICAL ACTIVITY SCALE: VALIDITY AND PSYCHOMETRICS AMONG A
SAMPLE OF LATINA WOMEN

Andrea Mendoza-Vasconez, MPH1, Bess H. Marcus, Ph.D.2

1
UC San Diego/San Diego State University, San Diego, CA; 2University of California, San Diego,
La Jolla, CA

Even though Latinos have become a priority population for the promotion of physical activity (PA),
several of the most widely used scales in PA promotion research have not been validated among this
population. The objective of this study is to assess the validity and other psychometrics of the Self
Efficacy for Physical Activity (SEPA) scale among a sample of Latino women who participated in the
Pasos Hacia La Salud study. Data from 205 women were analyzed. Item Response Theory (IRT)
methods were used to model response options for the SEPA scale. Internal consistency (Cronbachs
alpha) and unidimensionality (factor analysis) were assessed for the original and a modified version of
the scale, both with baseline and 6-month data. Concurrent and predictive validity of the original and
modified SEPA scales were assessed against accelerometer-measured moderate to vigorous PA
(MVPA), using correlations and linear regressions. Based on plots of Option Characteristic Curves
(OCC), a modified version of the SEPA scale was created by collapsing the five response options into
three. Internal consistency for this modified scale remained acceptable (.71 - .76), and factor loadings
improved compared to the original scale (.54 - .81 compared to .40 - .86). Self-efficacy at 6-months,
measured with both the original and modified scales, was significantly correlated with MVPA at 6
months (r=.30 and r=.32, respectively, p < 0.001). However, self-efficacy at baseline was not correlated
with MVPA at baseline, possibly due to the narrow range and restricted values of baseline MVPA.
There were no significant effects of baseline self-efficacy (measured with the original and modified
scales) on MVPA at 6 months. In conclusion, this study confirmed the SEPA scales reliability and
unidimensionality, and revealed that the scales accuracy improves when response items are
collapsed, which is an important finding for future research among populations with low literacy
levels. Additionally, the study provided some evidence of concurrent validity of the SEPA scale against
MVPA, but no evidence of predictive validity, possibly due to the long period of time elapsed between
baseline and follow-up measurements. Thus, further research to validate the SEPA scale among this
population is warranted.

CORRESPONDING AUTHOR: Andrea Mendoza-Vasconez, MPH, UC San Diego/San Diego State

University, San Diego, CA, 92116; [email protected]
S606 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F069 11:45 AM-12:45 PM

MONEY MATTERS: FINANCIAL STRESS AMONG YOUNGER ADULTS WITH CANCER

Sugandha Gupta, BS1, Amanda Marin-Chollom, MA2, Tracey A. Revenson, PhD3

1
The Graduate Center, City University of New York, New York, NY; 2Columbia University
Medical Center Mailman School of Public Health & The Graduate Center, City University of
New York, New York, NY; 3Hunter College, New York, NY

Background: Younger adult survivors, like all cancer survivors, face life stressors that are a
result the illness and its treatment, but they also face specific concerns that emerge because
their illness is off-time in the normative life cycle. Little research has examined how these
specific stressors affect quality of life for younger survivors.

Aim: This mixed-methods study identifies the unique types of financial stress experienced by
younger survivors and examines the relationship between financial stress and psychological
distress.

Methods: Sixty hematological cancer survivors aged 20-49, 65% female, diagnosed within the
past 5 years, were interviewed about the age-related challenges of their cancer diagnosis.
Transcripts were coded for four types of financial stressors: financial burden on family; issues
with health insurance; spillover to other life domains, such as plans for employment and
education; and emotional sequelae. Participants also completed a self-report measure of
depressive symptoms (CESD; Radloff, 1977).

Results: Although not directly asked about their financial situation, over 2/3 of the sample
described at least one type of financial stressor during the interview. In 70% of the cases,
these financial stains were directly linked to the survivors life stage, including not finishing
school, becoming unemployed, or envisioning an uncertain financial future. Interesting, those
who mentioned financial consequences had lower education (p < 0.04) and tended to be in
the 30s and 40s vs. 20s (p < .10). Half of the sample raised issues with health insurance
(52.5%), including the strains of no insurance or inadequate coverage. Nearly half the sample
(45%) described their financial strains as creating a burden on family and friends, often a
result of having to move back home or take a loan, which created feelings of inadequacy and
threats to independence. Over half the sample (55%) met the criterion for clinical depression
(CESD score > 16) although depressive symptoms were unrelated to reporting any type of
financial stress. Surprisingly, some younger survivors construed benefits in the way their
financial status affected their lives.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S607

Conclusions: Financial strains are clearly a concern for younger adult cancer patients, often
related to interruptions in school and work responsibilities as well creating constraints on
dating and friendships. These are concerns that should be discussed in counseling or peer
forums for young adult survivors.

CORRESPONDING AUTHOR: Sugandha Gupta, BS, The Graduate Center, City University of New
York, New York, NY, 10035; [email protected]
S608 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F070 11:45 AM-12:45 PM

THE ASSOCIAION BETWEEN FAMILY SES AND CHILDREN'S HEALTH-RELATED QUALITY OF LIFE:
THE ROLE OF OBJECTIVE AND SUBJECTIVE SES

Kay Kim, B.A.1, Jan Wallander, PhD2

1
University of California, Merced, Merced, CA; 2Univ. California, Merced, Los Gatos, CA

Objective: To our knowledge, no study has compared the relationships that objective social
status (OSS) and subjective social status (SSS) have with childrens general health and health-
related quality of life (HRQOL) across racial/ethnic groups. The present study tested the
hypotheses that (a) OSS and SSS each have unique associations but (b) SSS mediates a portion
of the association that OSS has with childrens health outcomes; and (c) these associations are
present across Black, Latino, and White children.

Method: Data were from the Healthy PassagesTM study, which enrolled 4,824 Black, Latino,
and White 5th graders (ages 10-11) in three U.S. metropolitan areas. OSS was measured as a
latent variable of parent reported educational attainment and total household income. SSS
was measured also a latent variable of community and U.S. standing on the MacArthur Scale
of Subjective Social Status. HRQOL was measured with the self-reported form of the PedsQL
Physical and Psychosocial scales. Childs health status was measured by both childs self-
report and parents report of his or her health status on 5-point scale from Poor to Excellent.

Results: Structural equation modeling of path analysis was conducted using Mplus version 7.
OSS had a stronger relationship to childrens health outcomes compared to SSS. Because SSS
did not have significant associations with the outcomes, there was little support for SSS
mediating between OSS and health outcomes. The results of path analysis significantly
differed across race/ethnicity, and only Black childrens health status, when self-reported, was
mediated by SSS.

Conclusion: Unlike emerging research on adults that SSS influences their health, the current
study showed that parents SSS did not play the same role for childrens health. Rather, OSS
was confirmed to have a stronger association with childrens health status. Research is
needed into the role of SSS for childrens health, for example how and at what age childrens
own perception of socioeconomic status influences their health.

CORRESPONDING AUTHOR: Kay Kim, B.A., University of California, Merced, Merced, CA,
95340; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S609

F071 11:45 AM-12:45 PM

ETHNIC DIFFERENCES IN EARLY FAMILY ADVERSITY AND HPA AXIS FUNCTIONING AMONG
LATINO AND EUROPEAN AMERICAN ADOLESCENTS

Ahra C. Ko, BA1, Jessica Chiang, PhD2, Julienne Bower, PhD3, Shelley Taylor, PhD4, Andrew
Fuligni, PhD3

1
University of Southern California, Los Angeles, CA; 2Northwestern University, Evanston, IL;
3
UCLA Departments of Psychology and Psychiatry/Biobehavioral Sciences; Cousins Center for
Psychoneuroimmunology, Los Angeles, CA; 4Department of Psychology, UCLA, Los Angeles, CA

Previous research indicates that early family adversity can lead to blunted responding of the
hypothalamic-pituitary-adrenal (HPA) axis (i.e., primary biological stress response system),
thereby increasing risk for poor health outcomes such as cardiovascular disease, anxiety, and
depression. Latinos are known to strongly value familism and collectivism whereas European
Americans promote independence. Thus, the enduring consequences of early family adversity
may be different between the two ethnic groups. However, the role of ethnicity and culture
has often been neglected in past studies. Furthermore, examination of adolescents has been
relatively rare, despite the fact that numerous biological and psychosocial changes occur
during the adolescent period. The current study investigated whether the association
between early family adversity and HPA response (i.e., cortisol production) differ between
Latino- and European-American late adolescents.

Participants (N = 91; Mage = 18.37 years, SD = .51; 57.14% female; 37.36% European and
62.64% Latino) completed the Risky Families Questionnaire (Taylor et al., 2006) and
participated in the Trier Social Stress Task (TSST), a widely used laboratory-based stressor that
activates the HPA axis. Six saliva samples were collected (baseline, and 0, 30, 45, 60, 75, and
90 minutes post-TSST) and assayed for cortisol. Total area of the curve with respect to ground
was computed to index total cortisol output (Pruessner et al., 2003).

Hierarchical regression analyses controlling for age, gender, and parental education revealed
that early family adversity was unrelated to cortisol output. However, a significant interaction
between early family adversity ethnicity emerged (b(SE) = -.62.29 (17.28), p = .001). Tests of
simple slopes indicated that early family adversity was associated with blunted cortisol output
among Latinos (b(SE) = -.33.82 (7.53), p < .001) but not among European-Americans (b(SE) =
28.47 (15.56), p = .07). Results suggest that the impact of early family adversity on HPA axis
functioning may be more pronounced in Latinos compared to European-Americans, which
may be due to their culturally rooted values of familism. This study highlights the importance
S610 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

of considering the role of ethnicity and culture when examining the effects of family-related
adversity.

CORRESPONDING AUTHOR: Ahra C. Ko, BA, University of Southern California, Los Angeles, CA,
90005; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S611

F072 11:45 AM-12:45 PM

THE IMPACT OF HEALTHCARE DISCRIMINATION ON BELIEFS ABOUT HEALTH PROFESSIONALS

AND CANCER SCREENING EMOTIONS

Daniel B. Betancourt, B.S. candidate1, Patricia M. Flynn, Ph.D., M.P.H.2

1
La Sierra University, Riverside, CA; 2Loma Linda University, Loma Linda, CA

Ethnic minorities and lower socioeconomic status (SES) populations report more negative
healthcare encounters and perceived discrimination. The aim of this research was to study the
role of perceived healthcare discrimination as an antecedent of patients negative beliefs
about health professionals. These beliefs about health professionals have been associated
with emotional reactions to clinical encounters that negatively impact continuity of cancer
screening care (Betancourt, et al., 2011). It is proposed that negative beliefs about health
professionals are the result of prior experiences of healthcare discrimination, which are
expected to be associated with cancer screening emotions such as fear and anxiety about
future clinical breast exams (CBE).

Participants were 171 Anglo American (Anglo) and 164 Latino American (Latino) women of
varying SES levels from Southern California. They responded to a series of scales on perceived
experiences of healthcare discrimination, negative beliefs about health professionals, and
cancer screening emotions. Of the 335 participants, 73% of Anglo and 79% of Latino women
reported prior experiences of healthcare discrimination, mainly based on insurance, SES, and
ethnicity. Scores on negative beliefs about health professionals were higher for participants
who experienced discrimination than those who did not (Latino: t = 3.86, p = .00; Anglo: t =
3.78, p = .00). In addition, compared to participants that did not report a prior experience of
healthcare discrimination, those that did had higher levels of fear (Latino: t = 3.20, p = .00;
Anglo: t = 3.01, p = .00) and anxiety (Latino: t = 2.60, p = .01; Anglo: t = 2.07, p = .04) in
anticipation of a CBE.

Results are discussed in terms of implications for addressing health disparities and the
translation of research into practice. This research demonstrates that prior experiences of
healthcare discrimination elicit negative beliefs about health professionals and negative
emotions associated with cancer screening, which according to previous research impact
discontinuity of cancer screening care (Betancourt et al., 2011; Flynn et al., 2011). Training
health professionals to use evidence-based bias reduction strategies to prevent the negative
implications of their own biases about ethnic minorities and low SES patients could contribute
to more positive clinical encounters and reduce perceptions of discrimination on the part of
S612 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

patients. This, in turn, may improve cancer screening emotions and screening rates overall,
but particularly among ethnic minorities and low SES groups.

CORRESPONDING AUTHOR: Daniel B. Betancourt, B.S. candidate, La Sierra University,

Riverside, CA, 92505; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S613

F073 11:45 AM-12:45 PM

THE ROLE OF ACCULTURATION AND STRESS IN RISK FOR OBESITY IN PREGNANT MEXICAN
AMERICAN WOMEN

Amiel Maldonado, Student, Kimberly D'Anna-Hernandez, PhD, Paola Ilescas, Student

CSU San Marcos, San Marcos, CA

Pregnancy is a vulnerable time period where any influence on maternal wellbeing affects the
development and outcome of offspring. A common issue of concern is obesity, which greatly
increases the risk for complications, with intergenerational effects extending well into the
offspring's adult life. Nationwide, obesity disproportionately affects Mexican Americans, one
of the fastest growing minorities. Acculturation, or the process of changes in beliefs and
customs as the result of cultural contact, has been associated with obesity. As Mexican
traditions clash with American mainstream culture, unhealthy behaviors, stress, and mental
health symptoms increase. This population also experiences cultural-specific stressors, such
as discrimination, or the negative treatment and attitudes towards an individual or group,
which further augment perinatal health risks. Though we know acculturation is associated
with disparate health outcomes, such as obesity, we have yet to discover the psychological or
social mechanism by which it is occurring during te sensitive period of pregnancy. The purpose
of this study is to investigate the roles of depression, stress, and perceived discrimination in
the relationship between acculturation and obesity during pregnancy. It was hypothesized
that these factors would moderate the acculturation and obesity relationship, where higher
levels of each would strengthen the association during pregnancy. Data was gathered as part
of an ongoing longitudinal study among 133 Mexican American women with singleton
pregnancies who were administered self-report surveys for symptoms of depression,
perceived stress and discrimination, and acculturation via adherence to Mexican vs. Anglo
cultural values. Pre-pregnancy BMI was determined via medical records and categorized as
per WHO guidelines (< BMI 25 as normal weight, >BMI >25 as overweight/obese). T-test
analyses revealed that overweight/obese women were less acculturated, i.e. less likely to
adhere to Anglo values (t=2.53, p=0.01), but BMI was not associated with stress, depressive
symptoms or discrimination. Stress positively moderated the relationship between
acculturative values and BMI (B=-0.105, SE=0.045, t, p=0.02), such that women who adhered
to less acculturated values and had more perceived stress, had higher pre-pregnancy BMIs.
These results suggest that the role of acculturation is complex in shaping the health outcomes
of minorities, including obesity risk. Women may be struggling to hold onto Mexican cultural
values in U.S. society, which may result in more perceived stress with consequences for
gestational physiology and may place them at risk for obesity-related outcomes in
pregnancy. Since weight control during this sensitive time period is paramount, our results
S614 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

have implications for culturally sensitive interventions to minimize obesity prevalence in the
fast-growing Mexican-American population.

CORRESPONDING AUTHOR: Amiel Maldonado, Student, CSU San Marcos, San Marcos, CA,
92069; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S615

F074 11:45 AM-12:45 PM

RELIGIOSITY, SHAME DUE TO HETEROSEXISM & FORGIVENESS: CORRELATES OF SEXUAL

ATTITUDES IN OLDER GAY MEN AND LESBIANS

Rikki S. Willingham, Research Assitant

University of North Texas, Denton, TX

Religiosity, Shame due to Heterosexism & Forgiveness:

Correlates of Sexual Attitudes in Older Gay Men and Lesbians

Rikki Willingham, Tosha Griggs & Mark Vosvick Ph.D.

University of North Texas

Religion may influence older gay mens and lesbians (LG) attitudes about sex (Bonds-Raacke,
2011). While attending religious services older LGs may experience negative messages about
their sexual minority status (Lease et al., 2005). Older LGs may internalize these negative
messages and feel ashamed of their sexuality (DAugelli, 2001). High scores on aging sexual
knowledge and attitudes scale indicate negative attitudes toward sexual activity in older
adults (White, 1998). Forgiveness may help older LGs let go of shame and strengthen self-
acceptance (Enright, 1996). Through a minority stress (Meyer, 2003) lens, we hypothesize
religiosity and shame due to heterosexism (SDH) are positively associated with negative
sexual attitudes. Forgiveness of self is negatively associated with negative sexual attitudes.
Religiosity, SDH and forgiveness of self account for a significant proportion of variance in
negative sexual attitudes. After IRB approval, we recruited 50 gay men and 50 lesbians from
DFW, (Mage = 58.99, SD= 6.48). Participants were 68% European American, 17% African
American and 15% other ethnicity. A hierarchical regression analysis (F[10, 89]= 4.17, pR2=.24,
p

Older LGs reported favorable attitudes toward sexual activity this, which suggests that even
though sexual activity may decline with age, older LGs maintain a positive attitude towards
sexual activity (Lindau, 2007). African Americans reported less favorable attitudes toward
sexual activity; this may be due to minority stress associated with ethnicity and sexual
minority status (Ramos, 2016). Also, older LGs that scored high on SDH reported less favorable
sexual attitudes towards sexual activity in older adults this findin, which g supports that SDH is
associated with negative mental health outcomes in some older gay men and lesbians
S616 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

(Balsam, 2006). Therapeutic Iinterventions should encourage forgiveness of self and positive
sexual attitudes towards sexual activity. aAs well as, address the unique challenges faced by
African American older gay men and lesbians.

CORRESPONDING AUTHOR: Rikki S. Willingham, Research Assitant , University of North Texas,

Denton, TX, 76210; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S617

F075 11:45 AM-12:45 PM

ASSOCIATION OF THE PITTSBURGH SLEEP QUALITY INDEX AND SLEEP MEASURED USING
ACTIGRAPHY IN INDIVIDUALS WITH MULTIPLE SCLEROSIS

Christina S. Khou, MA1, Mayis Aldughmi, PT, PhD2, Melike Kahya, PT, MS3, Catherine
Siengsukon, PT, PhD2

1
University of Kansas, Lawrence, KS; 2University of Kansas Medical Center, Kansas City, KS;
3
University of Kansas Medical Center, Mission, KS

Background: The Pittsburgh Sleep Quality Index (PSQI) is a commonly used measure of self-
report sleep quality. Studies have found that the PSQI is not significantly associated with sleep
measured objectively using actigraphy; however, this relationship has not been explored
amongst individuals diagnosed with multiple sclerosis (MS). The purpose of this study was to
examine the relationship between scores on the PSQI and sleep measured using actigraphy in
this population.

Methods: Data were analyzed from a study investigating sleep in individuals diagnosed with
MS. Participants wore actigraphs (GTX3+) for 7 consecutive days and completed the PSQI.
Total score and subscale scores on the PSQI and average actigraphy time in bed (TIB), sleep
onset latency (SOL), total sleep time (TST), wake after sleep onset (WASO), and number of
awakenings (NWAK) were used for analyses. Correlation and multiple linear regressions were
conducted to examine the relationship between scores on the PSQI and the sleep variables
measured using actigraphy.

Results: A majority of participants (N=76; Mage= 48.4 years-old, SD= 11.3) were female (84%)
and diagnosed with relapsing remitting MS (86%; 14% secondary progressive MS). PSQI total
scores did not significantly predict any actigraphy measures. However, subscales of the PSQI
that were selected based upon bivariate correlations did significantly predict sleep variables
measured using actigraphy. PSQI sleep medication use significantly predicted SOL (=-.26, p <
0.05). PSQI sleep duration significantly predicted both TIB (=-.33, p < 0.05) and TST (=-.36, p
< 0.05). PSQI sleep disturbance significantly predicted NWAK (=-.27, p < 0.05).

Conclusions: These data indicate that there is an inverse relationship between subscales of
the PSQI and sleep measured using actigraphy. Results suggest these two forms of assessment
may be assessing different constructs of sleep. Therefore, self-report and objective sleep data
should be interpreted and used differently in clinical and research settings. Research should
continue to investigate the relationship between self-reported and objectively measured
S618 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

sleep in order to further characterize this relationship among individuals diagnosed with MS
and other chronic conditions.

CORRESPONDING AUTHOR: Christina S. Khou, MA, University of Kansas, Lawrence, KS, 66049;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S619

F076 11:45 AM-12:45 PM

THE IMPACT OF A SUGARED BEVERAGE SALES BAN WITH VERSUS WITHOUT A BRIEF
INTERVENTION ON SUGAR SWEETENED BEVERAGE CONSUMPTION

Alison R. Hartman, BA1, Angela Wackerly, BFA2, Cindy W Leung, ScD MPH2, Ashley E. Mason,
PhD2, Laura Ishkanian, MPH1, Leeane Jensen, MA1, Janet Wojcicki, PhD, MPH1, Samantha J.
Schilf, BA1, Samantha Schleicher, BA2, Robert Lustig, MD1, Laura Schmidt, PhD1, Elissa S. Epel,
PhD1

1
University of California, San Francisco, San Francisco, CA; 2University of California San
Francisco, San Francisco, CA

The Impact of a Sugared Beverage Sales Ban With versus Without a Brief Intervention on
Sugar Sweetened Beverage Consumption

Background: Heavy consumption of sugar-sweetened beverages (SSBs) is linked to chronic

disease in large epidemiological studies. In response to this strong evidence base, academic
campuses and other workplaces have begun to implement policies to reduce SSB
consumption. One such example is the UCSF Healthy Beverage Initiative (HBI), which
eliminated the sale of sugar-sweetened beverages at all UCSF campuses and hospitals. The
outcomes of this policy, however, are unknown.

Research question: Does the addition of motivational interviewing counseling to an

environmental intervention (i.e., banning sales of sugar-sweetened beverages) improve
metabolic benefits to employees following a ban on sales of sugared beverages in the
workplace.

Methods: We recruited 214 employees who self-reported consuming greater than 12 oz. per
day of SSBs to participate in an evaluation of the UCSF HBI. In two in-person visits (pre-sales
ban, and ~10 months later), we collected anthropomorphic measurements, a fasted blood
draw for a metabolic panel, and self-reported reward-driven eating. We randomized half of
the participants to a motivational interviewing-like counseling group (intervention) or a
control group (no counseling). The intervention group had one 10-minute session with a
health counselor covering psychoeducation about effects of sugar intake and realistic goal
setting, and had two 5-10-minute follow-up phone calls discussing goals and overcoming
obstacles.

Results: The intervention group (n=108) reduced their SSB consumption by an average of 23.7
oz/day, while the control group (n= 106) reduced their SSB consumption by an average of 4
S620 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

oz/day. Specifically, the intervention group consumed significantly fewer ounces of SSBs than
the control group both at work (m decrease=12 oz; p=.0001) and at home (m decrease=11.8
oz; p=.0007). Both groups trended toward non-significant improvements in waist
circumference, sagittal diameter, and reward-driven eating from baseline to follow-up, and
the intervention group also showed small changes in BMI (m BMI change=-1.68; p = .27).

Conclusions: After implementation of the UCSF HBI, both intervention and control groups
showed significant decreases in SSB consumption both at work and at home, which suggests
that a workplace sales ban on SSBs alone may reduce SSB consumption. Brief counseling may
boost effectiveness of a sales ban on SSB consumption. Combining both an environmental
change (sales ban) and a brief intervention may be a highly economical and effective public
health intervention for institutions to reduce widespread SSB consumption.

CORRESPONDING AUTHOR: Alison R. Hartman, BA, University of California, San Francisco, San
Francisco, CA, 94118; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S621

F077 11:45 AM-12:45 PM

FRESHMAN 15: THE PRESSURE TO BE THIN

CeCe Wolfner, N/A, Rucha Kandlur, N/A, Lauren Boddy, BA, Nancy A. Hamilton, PhD

University of Kansas, Lawrence, KS

Body imageaffects both mental and physical aspects of women in their daily lives. The socially
constructedthin idealmay lead normal weight individuals (BMI=18.5-24.9) to fall prey to
the pressures of maintaining a certain weight. These concerns may become more distressing
during college and freshman women may be especially vulnerable to these social pressures
during emerging adulthood. Thus, it was predicted that among normal weight freshmen
women, distress (measured by the Beck Depression Inventory, (BDI-II)and Perceived Stress
Scale (PSS)) would be related with believing that one needed to lose weight. To test this
hypothesis, data from alarger, longitudinal study were collected from a sample
ofprimarilyCaucasian(83 %)college freshmenwomen (N=48; mean age=18.2yrs). Consistent
with this hypothesis, among normal weight women (n=33), those that reported they were
trying to lose weight had significantly higher depression and perceivedstressscores (BDI:
M=9.12, SD =9.02; PSS: M=19.29, SD=5.90) compared to those who indicated they were not
trying to lose weight (BDI: M=2.06, SD=2.02; PSS: M=11.38, SD=6.29), BDI: t(31)=3.06, p < .01,
PSS: t(31) =3.73, p < .01. Interestingly, depression and perceived stress scores were also
related to stage of change for weight loss. The highest indicators of distress were found
amongwomen who have a plan to lose weight this semester(BDI: M=12, PSS: M=20.5)and
who have recently lost weight and were trying to keep it off(BDI: M=27, PSS: M=28). The
lowest depression and perceived stress scores were found among women who were happy
with their weight and have not tried to change it (BDI: M=1.6, PSS: M=9.8). Statistically
significant group differences were found amongst women at different stages of weight loss for
indicators of distress (BDI: F(5,27)=4.83, p < .01, PSS: F(5,27)=3.71, p < .05).
Thesepreliminaryresults demonstrate that the desire to lose weight is significantly related to
indicators of distress for female emerging adults. During a time of identity formation and
transition to college, these body weight concerns may increase vulnerability to disordered
eating behaviors and other maladaptive behavioral health patterns. Limitations of these pilot
data include lack of racial diversity, small sample size, and limited age variation.

CORRESPONDING AUTHOR: CeCe Wolfner, N/A, University of Kansas, Lawrence, KS, 66044;
[email protected]
S622 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

F078 11:45 AM-12:45 PM

OPTIMIZATION OF MODERN TECHNOLOGY IN COPING WITH STRESS & DEPRESSION

Emily Wong, Student1, John Hunter, BA2, Sarah D. Pressman, PhD3

1
University of California, Irvine, Temecula, CA; 2UC Irvine, Irvine, CA; 3University of California-
Irvine, Irvine, CA

The majority of research concerning the modern use of technology, specifically cellphones,
has focused on the degenerative effects it has on users. Previous research has also focused on
its specific ties to depressive symptoms. However, few studies have examined the potential
benefits that cellphone use can have in stressful social situations. Participants (n=143)
underwent the Yale Interpersonal Stressor (YIPS), in which two confederates socially exclude
one participant during a brief interaction. The experiment consisted of three cellphone-use
conditions: 1) the ability to use a cellphone during the YIPS, 2) the presence of a cellphone,
but use was restricted during the YIPS, and 3) the absence of a cellphone during the YIPS.
Moreover, the level of subclinical depressive symptoms was considered a moderator, as
depression and depressive symptoms have shown to alter stress-responses. Though self-
reported data on feelings of stress did not support our stress-buffering hypothesis,
participants whose phones were present during the YIPS, felt significantly less excluded. And
lastly, the level of subclinical depressive symptoms had no serious moderating effect on the
main effect. So, contrary to previous research, those with higher levels of depressive
symptoms did not experience more exclusion than those with low levels of depressive
symptoms. Thus, those with higher levels of depressive symptoms still benefited from
cellphone presence. This implies that cellphones do have potential benefits in stressful social
situations. However, future research needs to further explore physiological stress responses
to gain a more comprehensive picture of how cellphone use can positively influence our
physiological functioning, especially for those with higher levels of subclinical depressive
symptoms; this allows for possible interventions that optimize the function of our modern
technology in coping with stress.

CORRESPONDING AUTHOR: Emily Wong, Student, University of California, Irvine, Temecula,

CA, 92591; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S623

F079 11:45 AM-12:45 PM

THE EFFECT OF POSITIVE EMOTIONS ON CARDIOVASCULAR REACTIVITY AND RECOVERY FROM

STRESS

Casey E. Cavanagh, Ph.D.1, Kevin T. Larkin, Ph.D.2

1
VA Connecticut Healthcare System, Milford, CT; 2West Virginia University, Morgantown, WV

Despite the recent interest in the role of positive emotions on physical health, the
mechanisms explaining this relation are not well understood. The undoing hypothesis, which
proposes that positive emotions speed physiological recovery from stress (Fredrickson &
Levenson, 1998), has emerged as a potential mechanism explaining this relation. However,
the literature is mixed in demonstrating this effect. Therefore, the purpose of this study was
to examine two methods of inducing positive emotions to determine how positive emotions
affected cardiovascular reactivity and recovery from stress.

This study employed a 2 x 2 between-subjects design in which the two methods for inducing
positive emotions included engagement in a Duchenne-smile manipulation and exposure to a
photograph of a pleasant memory. Participants included 61 undergraduates who were
randomized to one of four groups: Smile-Happy (Duchenne-smile + personally-relevant
pleasant photograph), Smile-Neutral (Duchenne-smile + neutral photographs), No-Smile-
Happy (Neutral expression + pleasant photographs), and No-Smile-Neutral (Neutral
expression + neutral photographs). Participants engaged in these emotion induction tasks
simultaneously while completing a mental arithmetic stressor task.

Cardiovascular reactivity to the stressor task was examined using Smile by Event ANCOVAs on
systolic and diastolic blood pressure (SBP and DBP respectively). Results demonstrated a
significant main effect of Event for SBP with the personally-relevant photograph groups
demonstrating greater reactivity as compared to the neutral photograph groups, F(1, 55) =
4.30, p = .04. Additionally, results also demonstrated a significant main effect for Smile for
DBP with greater reactivity in the Smile groups as compared to the No-Smile groups, F(1, 55) =
7.56, p = .008. It is possible that the elicitation of specific positive emotions results in
increased cardiovascular reactivity (Yogo, Hama, Yogo, & Matsuyama, 1995). Finally, a series
of between-subjects ANCOVAs were conducted on area under the curve for SBP and DBP.
Results revealed no group differences in SBP and DBP recovery, suggesting that positive
emotions did not improve cardiovascular recovery from stress as measured by SBP and DBP
and thereby fails to support the undoing effect. Although there was no evidence supporting
the undoing effect, the evidence of increased reactivity to the positive-emotion induction
tasks suggests the importance of examining emotion specific effects on cardiovascular
S624 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

reactivity and recovery. Future research should also consider conducting reproducibility
studies of the original undoing effect and further explore the clinical utility of the undoing
effect.

CORRESPONDING AUTHOR: Casey E. Cavanagh, Ph.D., VA Connecticut Healthcare System,

Milford, CT, 06460; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S625

F080 11:45 AM-12:45 PM

YOGA FOR TEACHERS

Christine Glissmann, BS in Nutrition and Exercise Physiology 1, Jennifer Huberty, PhD 2, Ali
Owens, BS in Communication Studies and World Arts and Cultures 3

1
Arizona State University, Tempe, AZ; 2Arizona State University School of Nutrition and Health
Promotion, Phoenix, AZ; 3YogaWorks, Los Angeles, CA

Introduction:
Nearly 40% of elementary teachers leave their profession due to heavy workloads and stress.
Although there is research that has focused on understanding educator stress and the
associated consequences on their well-being, there is little research focused on the best way
to mitigate stress and improve well-being. Contemplative practices (i.e., mindfulness) have
been tested as a means to prevent and alleviate teacher stress with several studies
demonstrating their effectiveness. The purpose of this study was to determine interest in
using online yoga (Udaya.com) as a means to reduce stress and improve health. Findings from
the survey will be used to design an online yoga intervention for teachers before, during, or
afterschool.

Methods:
This study was approved by an Institutional Review Board. The 17-question online survey was
administered via Qualtrics to preschool and elementary school teachers in a Southwestern
state. Teachers were recruited using social media (i.e., Facebook), fliers at schools, and email
list servs. Questions were related to teachers interest in participating in online yoga and
perceived principal support for yoga participation.

Results:
Of the 117 who completed the survey, most (78%; n=91/117) were not currently participating
in yoga but 75% (n=68/91) had participated in yoga in the past, with, 57% (n=45/79) only ever
participating in a few sessions. When asked about interest in participating in online yoga, 84%
(n=90/108) were very interested/interested. Almost 90% of teachers (n=96/109) were
interested in participating in online yoga in a common area amongst other teachers/staff. The
top three reasons teachers were interested in yoga included: stress relief (91%; n=86/94),
relaxation (82%; n=77/94), and general health (63%; n=59/94). Teachers perceived their
principal would support online yoga participation if it was after school (24%; n=26/109) or
before or after school (39%; n=43/109).
S626 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusion:
Teachers were interested in having online yoga at their school to reduce stress, and improve
their health and felt principals would support online yoga. Online yoga may be a reasonable
solution to reducing time restraints, mitigating cost, and providing accessibility to yoga for
teachers. Interventions to determine feasibility and initial effectiveness of online yoga for
teachers in the school setting are warranted.

CORRESPONDING AUTHOR: Christine Glissmann, BS in Nutrition and Exercise Physiology ,

Arizona State University, Tempe, AZ, 85281; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S627

F081 11:45 AM-12:45 PM

A GENOME-WIDE ASSOCIATION STUDY TO INVESTIGATE SHARED GENETIC VARIATION FOR

BINGE EATING AND BINGE DRINKING IN COLLEGE STUDENTS

Carley G. Lovell, MS, MA, BS, BA, RN, WHNP-BC1, Fazil Aliev, Dr.2, Kenneth Kendler, M.D.3,
Danielle Dick, Ph. D.4, Amy Adkins, PhD5

1
Virginia Commonwealth University, Henrico, VA; 2Virginia Commonwealth University
Department of Psychology, Richmond, VA; 3Virginia Institute for Psychiatric and Behavioral
Genetics, Richmond, VA; 4Devlopmental and Clinical Psychology; Virginia Commonwealth
University, Richmond, VA; 5Dept. of Psychology, Virginia Commonwealth University,
Richmond, VA

Purpose: To investigate shared genetic variation between binge drinking and binge eating in a
large, college-aged sample using a Genome-Wide Association Study (GWAS) approach.

Design: Researchers will perform a secondary analysis of data in a large, longitudinal

database consisting of approximately 10,000 individuals, which aims to understand how genes
and the environment come together to influence substance use and mental health. Fall and
spring surveys from the 2011-2013 cohorts first year of college were analysed (n~4100).
Surveys queried mental and behavioural health outcomes as well as environmental variables.
Saliva samples were collected and DNA isolated.

Results: Initial analyses showed that students who binge eat are 1.7 times more likely to
report parental problem drinking than students who do not binge eat (95% CI= 1.5, 1.9); and
students who binge drink are 1.5 times more likely to report parental problem drinking than
students who do not binge drink (95% CI= 1.3, 1.7). The DNA samples were genotyped on the
Affymetrix Biobank Version 2 Array, which contains both rare variation, as well as an
imputation GWAS grid.

Conclusions: Results offer preliminary steps to determine if previous findings may be due to a
genetic susceptibility to binge behaviours. Current GWAS analysis, with considerations of
Linkage Disequilibrium and a gene-based analysis, are being conducted to study this
hypothesis.

Clinical Relevance: The overall objective of this study is to advance understanding of shared
genetic variation between binge eating and drinking, and to contribute to the long-term
S628 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

objective of informing and designing effective and efficacious prevention and intervention
strategies for binge behaviours for college age youth.

CORRESPONDING AUTHOR: Carley G. Lovell, MS, MA, BS, BA, RN, WHNP-BC, Virginia
Commonwealth University, Henrico, VA, 23231; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S629

F082 11:45 AM-12:45 PM

ASSESSING HETEROGENEITY IN THE ADOLESCENT PHYSICAL ACTIVITY-SMOKING RELATION: A

SEQUENTIAL PROCESSES GROWTH MIXTURE MODEL

Daniel Rodriguez, Ph.D.

La Salle University, Philadelphia, PA

Significance. Adolescence is a key period for the initiation of cigarette smoking (smoking) and
progression to a regular habit. 9 of 10 adult smokers initiated smoking by age 18. One factor
that may protect adolescents from smoking initiation and progressing to a regular habit is
physical activity (PA). Yet, not all PA may be equally predictive. We sought to better
understand the PA-smoking relation by testing for developmental heterogeneity.

Methods. Participants were 1361 high school students (50% female; 73% White) taking part in
a prospective study (4 years, 8 waves) of the relation between adolescent PA and smoking.
We employed Sequential Processes Growth Mixture Modeling (SPGMM) to analyze our data.
SPGMM allows for identification of homogenous subpopulations (conjoint classes) on
repeated measures of 2 variables (e.g., smoking and PA), and characterizing the classes on
select covariates with multinomial logistic regression. We employed Bayesian Information
Criterion (BIC), Adjusted BIC (aBIC), Bootstrap Likelihood Ratio Test (BLRT), average probability
of correct classification, and substantive criteria to identify the correct number of conjoint
classes.

Results. Four classes were identified: High PA/Low smoking (HPA/LS); Low PA/Consistently
high smoking (LPA/CHS); Low PA/Increasing smoking (LPA/IS); moderate PA/Low smoking
(MPA/LS). Having greater expectations of the benefits of smoking was associated with an
increase in the odds of belonging to all other classes compared to HPA/LS. Participation in
alternative PA (e.g., skateboarding) was associated with an increase in the odds of being in
LPA/CHS and LPA/IS compared to HPA/LS. Using alternative tobacco products and having
peers who smoke were associated with an increase in the odds of belonging to LPA/CHS and
LPA/IS versus HPA/LS. By contrast, having PA support and positive expectations of the benefits
of PA were associated with a decrease in the odds of belonging to LPA/CHS versus HPA/LS.

Conclusions. The findings of this study suggest the existence of multiple relations among
smoking and PA, and that the best predictors of higher smoking and lower PA are smoking
expectations and White race. Finally, not all PA is equally protective against smoking. Various
physical activities (e.g., skateboarding) may actually increase the odds of smoking.
S630 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Daniel Rodriguez, Ph.D., La Salle University, Philadelphia, PA,

19141; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S631

F083 11:45 AM-12:45 PM

TEXT & MOBILE MEDIA SMOKING CESSATION SERVICE FOR YOUNG ADULTS IN SOUTH TEXAS

Patricia Chalela, DrPH1, Alfred McAlister, PhD2, Kip Gallion, MA1, Edgar Muoz, MS1, Cliff
Despres, BJ1, David Akopian, PhD3, Andrea Fernandez, MPH1, Rafael Diaz, BS1, Amelie
Ramirez, DrPH1

1
UT Health San Antonio, San Antonio, TX; 2The University of Texas, School of Public Health,
Austin, TX; 3University of Texas at San Antonio, San Antonio, TX

Smoking among Latino young adults (18-29) in South Texas is high (23.2% to 25.7%),
representing a serious public health problem. Yet few are reached by services to help them
quit smoking. Young adults are heavy users of mobile devices for texting and access to mobile
media. These have an extraordinary potential for assisting smoking cessation by providing
peer modeling and eliciting social reinforcement for behavior change. We present preliminary
results of Quitxt a bilingual text messaging and mobile media service to help young adults quit
smoking.

Methods: We constructed a bilingual texting and mobile media system that was promoted in
South Texas via social media advertising and other recruitment channels. The ads, which
featured couples with different themes (disgust with cigarettes or confidence in quitting
success) and styles (cowboy, metro/urban, geek, punk and graphic novel), asked potential
participants who showed interest in quitting smoking to text a code to our system
corresponding to the channel of recruitment. Text messages include links to web pages with
additional content and YouTube videos with peer modeling of reasons and skills to quit
smoking.

Preliminary Results: Preliminary results showed that enrollments were achieved for 798
participants with a mean age of 29.3 and 55% were below the age of 30. More men (57%)
than women (43%) enrolled and 36% identified themselves as Hispanic or Latino. The mean
number of cigarettes consumed per day was 11.5. Seven-month texted follow up found that
21% (171) of the enrollees reported abstinence at that point. This is consistent with high rates
of success found in studies of telephone counseling for young adults and confirms that text
and mobile media service specifically designed for young adults provide a feasible and cost-
effective approach to promoting cessation.

Conclusion:Preliminary results provide evidence that young adult smokers in South Texas can
be reached via mobile media service. The anticipated outcome is a scalable, culturally
S632 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

relevant, evidence-based and cost-effective service with broad national reach to help young
adult Latinos stop smoking, with the potential to reduce health care costs, reduce chronic
disease burden and improve quality of life among this young, fast-growing, at-risk population.

CORRESPONDING AUTHOR: Patricia Chalela, DrPH, UT Health San Antonio, San Antonio, TX,
78229; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S633

F084 11:45 AM-12:45 PM

THE ADDITIVE ASSOCIATION OF INDOOR CIGARETTE AND MARIJUANA SMOKING ON

POTENTIAL EXPOSURE TO FINE PARTICLES

Alexander Posis, BS1, Neil Klepeis, PhD2, John Bellettiere, MA, MPH, PhDc2, Sandy Liles, MPH2,
Vincent Berardi, MS, PhD2, Ben Nguyen, MPH2, Suzanne Hughes, MPH, PhD2, Melbourne
Hovell, PhD3

1
Center for Behavioral Epidemiology and Community Health (C-BEACH), Graduate School of
Public Health, San Diego State University, san diego, CA; 2Center for Behavioral Epidemiology
and Community Health (C-BEACH), Graduate School of Public Health, San Diego State
University, San Diego, CA; 3San Diego State University, San Diego, CA

Introduction: Exposure to fine particles (<=2.5 microns) from any source can induce
respiratory damage and exacerbate bronchitis and asthma. The substantial impact of cigarette
smoking on indoor fine particle levels is well known, but the impact of marijuana smoking has
not been described. That description is an essential first step toward understanding the public
health implications of increasing marijuana use. Using data from 193 households, we
examined associations of indoor marijuana and cigarette smoking with fine particle
concentrations and tested whether the two particle generating events (PGEs) have a
synergistic effect on fine particle levels.

Methods: As part of a larger trial, air particle monitors were placed in 298 homes of families
with at least 1 cigarette smoker and 1 child under the age of 14. After monitors continuously
measured fine particle counts (0.5 to 2.5 microns) for at least 7 days, participants were
interviewed about past 7-day frequency of cigarette and marijuana smoking, other PGEs (e.g.,
burning candles, burning food) and ventilation activities such as use of exhaust fans.
Dichotomized survey responses (any vs. none) from 193 homes that provided data for all
relevant PGEs and ventilation activities were used along with mean air particle counts from
the 7-day period. Unadjusted geometric means (GMs) were computed to summarize air
particle concentrations. After log transformation of mean air particle counts, multivariable
linear regression models were used to adjust for other factors associated with fine particle
levels. Synergistic effects were then tested by including a multiplicative (cigarette
smoking)*(marijuana smoking) interaction term in the regression model.

Results: Homes without indoor smoking had GM counts of 1985 particles/0.01 ft^3 while
homes with marijuana smoking only, cigarette smoking only, and both cigarette and
marijuana smoking had GM counts of 2869, 3019, and 5250. Regression models adjusted for
S634 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

home size, other PGEs, and ventilation activities indicated that marijuana smoking was
associated with a 68% increase in GM counts while indoor cigarette smoking was associated
with a 57% increase (ps < 0.001). No synergistic effect was detected between marijuana and
cigarette smoking (p=0.586).

Conclusion: Indoor marijuana and cigarette smoking were significantly associated with
elevated fine particle concentrations; their relationship was additive (not synergistic).

CORRESPONDING AUTHOR: Alexander Posis, BS, Center for Behavioral Epidemiology and
Community Health (C-BEACH), Graduate School of Public Health, San Diego State University,
san diego, CA, 92123; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S635

F085 11:45 AM-12:45 PM

TOBACCO USE AND PREFERENCES FOR WELLNESS PROGRAMS AMONG HEALTH AIDES AND
OTHER EMPLOYEES OF AN ALASKA NATIVE HEALTH CORPORATION

Tabetha A. Brockman, M.A.1, Christi A. Patten, Ph.D.1, Carrie A. Bronars, Ph.D.2, Matthew
Scott, MBA, BSN, RN3, Rahnia Boyer, BS3, Harry Lando, Ph.D.4, Matthew Clark, Ph.D.1, Ken
Resnicow, Ph.D.5, Paul A. Decker, B.S.1, Christina M. Smith, BA1, Agnes Roland, M.A.3, Marcelo
M. Hanza, M.A.1

1
Mayo Clinic, Rochester, MN; 2Veterans Affairs Medical Center, Minneapolis, MN; 3Yukon
Kuskokwim Health Corporation, Bethel, AK; 4University of Minnesota, Minneapolis, MN;
5
University of Michigan, Ann Arbor, MI

Objective: To assess health behaviors and preferences for health and wellness programs,
among employees of an Alaska Native-serving health corporation in western Alaska. Village-
based health aides were compared with other employees on health indicators and program
preferences. Methods: An electronic cross-sectional survey was distributed to all 1290 staff
employed at the Yukon Kuskokwim Delta Corporation (YKHC). The 30-item survey was
developed using a community-based participatory approach. Items assessed demographics,
tobacco use, physical activity level, body mass index (BMI), and included the Perceived Stress
Scale and Brief Resilience Scale. Preferences for health and wellness topics and for program
delivery formats were also assessed. Results: 467 employees responded (36%) with a final
sample 429 of usable data; of which 77% were women and 57% were Alaska Native. The 7-day
point prevalence of current tobacco use was 39%. Health aides (n=46) were more likely than
other employees (n=383) to currently use tobacco (59% vs. 36%, p = 0.003) and more likely to
use a homemade form of smokeless tobacco known as Iqmik (37% vs. 15%, p < 0.001). After
adjusting for covariates, greater stress levels was associated (p=0.013) with increased
likelihood of tobacco use. Overall, employees reported lower than recommended levels of
physical activity and motivation for exercise, 74% had a BMI indicating overweight or obese,
and no differences between health aides and other respondents. Top preferences for wellness
topics were for eating healthy, physical activity, weight loss, reducing stress, and better sleep.
Preferred program delivery formats were in-person, DVD, e-mails and posters. Health aides
reported greater interest in tobacco cessation than did other employees (37% vs. 21%,
p=0.016) and were less likely to prefer programs delivered by e-mail (28% vs. 50%, p=0.006).
Health aides reported higher perceived stress (p=0.006) and lower resiliency (p < 0.001) than
other employees. Conclusions: This study provides novel information about tobacco use and
interest in worksite wellness programs among employees of an Alaska Native-serving health
S636 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

corporation. These findings are helpful in developing tailored wellness promotion programs
for YKHC employees. Promoting healthy lifestyles among health aides and other employees
could ultimately have downstream effects on the health of Alaska Native patients and
communities that YKHC serves.

CORRESPONDING AUTHOR: Tabetha A. Brockman, M.A., Mayo Clinic, Rochester, MN, 55905;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S637

Panel Discussion 1 12:15 PM-1:15 PM

USING DISSEMINATION AND IMPLEMENTATION SCIENCE TO ADDRESS DISPARITIES IN OBESITY

PREVENTION AND REDUCTION IN PRIMARY CARE

Chanita Hughes-Halbert, PHD1, Cathy Melvin, PhD1, Melanie Jefferson, PhD1, Gary Bennett,
PhD2

1
Medical University of South Carolina, Charleston, SC; 2Duke University, Durham, NC

Background: Provider counseling about lifestyle modification and health behavior change
counseling is important for addressing obesity and excess weight, but rates of provider
counseling is low. As such, implementation of evidence-based interventions for health
behavior change counseling may be sub-optimal.

Objective: The purpose of this panel is to present the results of a demonstration project
that was funded by the AHRQ to use the principles of community-based participatory research
to disseminate and implement evidence-based strategies for health behavior change
counseling into primary care. The following key issues will be addressed: (1) what lifestyle
counseling and behavioral change interventions are effective in changing patient outcomes
(diet, physical activity and weight loss and/or BMI) for all patients, especially those from racial
and ethnic minority groups, in primary care practice settings; (2) what are patient and
provider priorities for health behavior change in primary care; (3) what are patient
motivations and barriers for health behavior change; (4) to what extent are primary care
patients engaged in shared decision making about weight loss/management?

Results: Through an evidence synthesis review, we found that intensive group and
individualized coaching combined with pharmacological therapy yielded the best weight loss
results in primary care. But, few interventions have specifically targeted racial/ethnic groups
and the differences observed between usual care and intervention groups were minimal in
minority populations. We also found that interventions that address diet and physical activity
are a priority among patients and providers for managing chronic conditions. Both patients
and providers indicated that tailored interventions are most likely to be effective, but there
was discordance about how the effects of these interventions should be monitored. Patients
wanted to be held accountable to providers, but providers wanted patients to be accountable
to themselves. These beliefs may explain why 57% of patients were very/completely
confident about their ability to lose/manage weight and 62% were making weight loss efforts.
Patient perceptions about shared decision-making about weight loss/management were also
limited. Only about 25% of patients reported that they selected a weight management option
together with their physician, their provider helped them understand weight management, or
that their physician explained the advantages and disadvantages for weight managemt.
S638 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions: It is important to actively engage providers and patients in the development of

implementation plans that are based on the evidence about the effects of interventions and
their priorities and preferences for behavioral interventions to increase the adoption of
evidence-based interventions for health behavior change counseling in primary care.

CORRESPONDING AUTHOR: Chanita Hughes-Halbert, PHD, Medical University of South

Carolina, Charleston, SC, 29425; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S639

Panel Discussion 2 12:15 PM-1:15 PM

ETHICAL USE OF MOBILE IMAGING, PERVASIVE SENSING, SOCIAL MEDIA, AND LOCATION
TRACKING (MISST) TECHNOLOGIES IN RESEARCH

Camille Nebeker, Ed.D., M.S.1, Katie Crist, MPH1, Job Godino, Ph.D.1, Gina Merchant, PhD2

1
UC San Diego, La Jolla, CA; 2University of California San Diego, La Jolla, CA

Mobile imaging, pervasive sensing, social media, and location tracking (MISST) technologies
show promise in providing more efficient and accurate methods for measuring and
intervening on health behaviors. Data from these devices can aid researchers to better
understand participant behavior in real world settings and to provide feedback in real time.
However, these ubiquitous technologies also present new challenges for human research
protections that must be considered to ensure that public health research utilizing them is
carried out in a manner that is socially and scientifically responsible. This panel will
demonstrate current MISST devices, explain how they are used in public health research and
the ethical challenges they entail, and suggested best practices for data collection, data
management, and participant protection. The panel will present information on: the
combined use of person worn cameras (SenseCam camera), accelerometry, and GPS and GIS
(Presenter 1); combined sensors and consumer-level wearables (Presenter 2); social media
and networks (Presenter 3); and ethical and regulatory challenges and solutions (Presenter 4).
Examples will be drawn from several large interventional and observational studies (e.g.,
http://cwphs.ucsd.edu/miparc ; http://cwphs.ucsd.edu/smart ; http://abcdstudy.org/).
Special emphasis will be placed on older adults, young adults, and adolescents, as sample
populations range in age from 6 to 102 years.

CORRESPONDING AUTHOR: Camille Nebeker, Ed.D., M.S., UC San Diego, La Jolla, CA, 92093;
[email protected]
S640 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Panel Discussion 3 12:15 PM-1:15 PM

DEVELOPING AND IMPLEMENTING EVIDENCE-BASED CLINICAL PRACTICE GUIDELINES:

DEPRESSION AND OBESITY

Jason Satterfield, PhD1, John R. McQuaid, Ph.D2, Hollie Raynor, PhD3, Lynn F. Bufka, PhD4

1
UCSF, San Francisco, CA; 2San Francisco VA Health Care System/University of California, San
Francisco, San Francisco, CA; 3University of Tennessee, Knoxville, TN; 4American Psychological
Association, Washington, DC

When utilized in health care systems, high quality clinical practice guidelines have the
potential to improve clinical practice and increase access to appropriate services. The
development of such guidelines synthesizes research evidence in order to facilitate clinical
decision-making and assist patients in making choices about treatment options. Although
widely used in medicine, fewer practice guidelines address interventions specific to mental
and behavioral health. The American Psychological Association (APA) and its partners are
generating clinical practice guidelines for psychological and health disorders. The goal is to
provide trustworthy, evidence-based guidelines the public can use to select among available
treatment options. The APA charged an advisory steering committee, including SBM
leadership and members, with the task of recommending transparent, best practices for
guideline generation and identifying strategies to promote dissemination and
implementation. The steering committee studied models adopted elsewhere such as the NICE
guidelines and recommended adhering closely to the recent National Academy of Medicine
(IOM) reports describing current best practices with regard to guideline generation. The APA
has new, draft guidelines for depression and childhood obesity that will be presented and
discussed by the panel. In each instance, every effort was made to deal with bias by ensuring
that competing perspectives were represented on the guideline panels, a process known as
adversarial collaboration. Similarly, both guidelines seek to integrate the best existing
empirical evidence with the expertise of the provider and the preferences and values of the
recipient of the intervention. This multi-organizational panel first describes the process of
guideline development and steps taken to minimize bias and improve quality. Speakers will
then briefly present near final findings from the guidelines in development. Strategies to
effectively disseminate and implement clinical practice guidelines within health care systems
and by individual providers and by patients will also be discussed. A final discussant will guide
a conversation about the implications of this project and how to best improve quality of care
for all.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S641

CORRESPONDING AUTHOR: Jason Satterfield, PhD, UCSF, San Francisco, CA, 94115;
[email protected]
S642 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Panel Discussion 4 12:15 PM-1:15 PM

NEW TECHNOLOGIES AND ANALYTIC TECHNIQUES: TOWARDS INNOVATIONS IN

MONITORING, MODELING, AND MODIFYING FOOD INTAKE

Donna Spruijt-Metz, MFA, PhD1, Melanie D. Hingle, BS; MPH; PhD2, Kayla de la Haye, Ph.D.1,
John Lach, PhD3, Jack Stankovic, PhD3, Stephen Kobourov, MSC, PhD4, Mihai Surdeanu, PhD4

1
University of Southern California, Los Angeles, CA; 2University of Arizona, Tucson, AZ;
3
University of Virginia, Charlottesville, VA; 4University of Arizona, Tuson, AZ

Dietary behavior, including accurate assessment of intake and valid observations the
influences on food-related behavior, continues to pose one of the wickedest problems in
obesity research. This panel brings together experts in nutrition, dietary intake, mobile health
and social networks to discuss the difficulties and opportunities in assessing and
understanding food intake. The considerations in the fields of nutrition obesity that inform
which characteristics of measures are considered necessary and sufficient will be revisited and
challenged. Some of the solutions that new technologies might offer will then be reviewed.
The use of digital breadcrumbs scraped from the web, crowdsourcing and other app and
web-based solutions will be compared and contrasted with photography + machine learning
approaches. We will examine new work in cyber-physical systems to understand eating
behaviors in real time, with a discussion on the kind of data that this technology can offer, and
the challenges of modeling that data. The challenges of using it to intervene in real time will
be examined. Finally, the very nature of the questions that we need to ask and the data that
we need to understand eating behavior will be queried, with a particular emphasis on the
real-time, socially networked nature of food consumption.

CORRESPONDING AUTHOR: Donna Spruijt-Metz, MFA, PhD, University of Southern California,

Los Angeles, CA, 90004; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S643

Panel Discussion 5 12:15 PM-1:15 PM

USING HUMAN COACHES IN DIGITAL HEALTH INTERVENTIONS: A DISCUSSION OF METHODS

AND MODELS

Emily Lattie, Ph.D.1, Kathryn Noth Tomasino, PhD2, Dori M. Steinberg, PhD, RD3, Ruth Q.
Wolever, PhD4

1
Northwestern University, Chicago, IL; 2Northwestern University, Feinberg School of
Medicine, Chicago, IL; 3Duke University, Durham, NC; 4Vanderbilt University Medical Center,
Durham, NC

The use of digital health interventions, such as those delivered online and via mobile phones,
is an increasingly common strategy to extend the reach and availability of behavioral medicine
services. A growing body of research suggests benefits of incorporating human coach support
into the provision of these services in order to increase patient engagement and enhance
outcomes of digital interventions. However, because including human support is costly, we
need to develop a common understanding of effective coaching methods and models in order
to best design and implement engaging and scalable digital health interventions. Further, it is
important to understand and utilize practical assessment strategies for measuring patient
engagement. During this panel, we will review heterogeneous definitions of health
coaching, discuss various health coaching programs, digital delivery methods, and models of
providing coach support. Panelists include behavioral scientists with expertise in developing
health coaching programs for general clinical practice and for use in digital health intervention
studies. Each panelist will provide a brief overview of the work they have conducted in this
field. This panel will provide an opportunity for the panelists and attendees to discuss
methods for enhancing patient engagement in digital health interventions in order to broaden
the reach of behavioral medicine services in an efficient and effective manner. Attendees will
be encouraged to think critically about digital health intervention design and a substantial
portion of time will be dedicated to discussion.

CORRESPONDING AUTHOR: Emily Lattie, Ph.D., Northwestern University, Chicago, IL, 60611;
[email protected]
S644 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Panel Discussion 6 12:15 PM-1:15 PM

GOTTA CATCH EM ALL: HARNESSING VIRALITY OF TECHNOLOGY AND TRENDS TO CATAPULT

OUR RESEARCH FORWARD

Carly M. Goldstein, PhD1, Claudio R. Nigg, PhD, FSBM2, Eric Hekler, PhD3, Madeleine P. Ball,
PhD4, Alexander M. Biel, BA 5, Ernesto Ramirez, PhD, MS6

1
Weight Control and Diabetes Research Center; Brown Medical School, Providence, RI;
2
University of Hawaii, Honolulu, HI; 3Arizona State University, Phoenix, AZ; 4Open Humans
Foundation, Astoria, NY; 5Arizona State University, Tempe, AZ; 6Fitabase, San Diego, CA

Increased social media use has propagated videos, photographs, fundraisers, and challenges
to a viral level. Going viral refers to rapid and wide electronic dissemination of material to
many users so that it becomes popularized. The typical research timeline has historically been
incompatible with studying viral trends, but some researchers have successfully harnessed
virality to recruit participants, conduct studies on timely topics, and generate data for national
policies. The purpose of this panel is to focus on targeting virality of technology and trends to
benefit research and just-in-time studies that could add substantially to the literature. This
panel will feature presentations and a moderated discussion with a team composed of: 1) Two
behavioral researchers (one faculty, one graduate student) who created and oversaw two
Pokemon, Go! studies shortly following the games creation and widespread popularity; 2) the
creator of the Open Humans research platform designed to connect researchers with
participants willing to gather and share their own data including through projects like the
Personal Genome Project; 3) an industry-academic partner with Fitabase specializing in
efficiently gathering health and behavioral data by integrating and working with consumer
tools, apps, and products; and 4) a behavioral scientist framing this discussion in the context
of taking advantage of emerging trends and early-and-often sharing as facets of agile science.
This panels aim is to provide instruction and suggestions for researchers aiming to harness
virality to recruit participants, gather data, or inform policy in a timely manner.

CORRESPONDING AUTHOR: Carly M. Goldstein, PhD, Weight Control and Diabetes Research
Center; Brown Medical School, Providence, RI, 02903; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S645

Panel Discussion 7 12:15 PM-1:15 PM

EXPANDING HORIZONS OF RESEARCH IN INTEGRATED PRIMARY CARE: CHALLENGES, LESSONS

LEARNED AND FUTURE DIRECTIONS

Kathryn E. Kanzler, PsyD, ABPP1, Deepu George, PhD2, Mario Nicolas, Ph.D.3, Jeffrey Goodie,
Ph.D., ABPP4, William Isler, PhD5

1
University of Texas Health San Antonio, San Antonio, TX; 2UT Rio Grande Valley School of
Medicine, McAllen, TX; 3Knowesis, Inc. (Contractor for US AIr Force), San Antonio, TX;
4
Uniformed Services University, Gaithersburg, MD; 5US Air Force, JBSA Lackland, TX

Integrating behavioral health clinicians in primary care settings has emerged as an essential
component of patient centered medical homes. Research assessing the most effective
methods for how to best address the needs of primary care populations is mounting.
However, conducting clinical research in these settings can be quite challenging due to issues
such as IRB hurdles, clinic workflow, time constraints, and limited resources. Our three
panelists are currently conducting research in the Primary Care Behavioral Health (PCBH)
model of service delivery and will share their unique experiences, to include all the ups and
downs of this rewarding clinical research. Our first panelist will discuss challenges with
conducting research in a large military healthcare system. They will describe difficulties
encountered when attempting to implement a study on shared medical appointments, and
how lessons learned from this experience were used to plan and execute a weight-loss
intervention study. Our second panelist will discuss their grant-funded pilot study, involving
examination of a clinical pathway for patients with persistent pain using Acceptance and
Commitment Therapy techniques in a patient centered medical home. Our third panelist will
describe their work doing a grant-funded multi-site quasi-experimental study of the PCBH
model of service delivery in different primary care settings. Our final speaker has decades of
experience in integrated primary care practice and research, and they will facilitate discussion
and Q&A, as well as highlight the key challenges, lessons learned and future directions for
research in integrated primary care.

CORRESPONDING AUTHOR: Kathryn E. Kanzler, PsyD, ABPP, University of Texas Health San
Antonio, San Antonio, TX, 78209; [email protected]
S646 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Panel Discussion 8 12:15 PM-1:15 PM

THE MULTIPLE FACES OF COMMUNITY OUTREACH: THE TULANE PREVENTION RESEARCH

CENTER

Carolyn C. Johnson, PhD MS1, Keelia O'Malley, MPH2, Catherine Haywood, BSW1, Naomi
Englar, BA3

1
Tulane University School of Public Health and Tropical Medicine, New Orleans, LA; 2Tulane
Prevention Research Center, New Orleans, LA; 3Tulane Prevention Research Center at Tulane
School of Public Health and Tropical Medicine, New Orleans, LA

Community outreach, engagement and partnership are the foundations upon which
community behavior change programs are based. Without engaging community residents in
health promotion efforts, little to no expectation of success can be achieved. Community
engagement can provide strong and relevant input on such issues as methods, media
messages, timing, priorities, and cultural sensitivity. A total of 26 Prevention Research
Centers (PRCs), funded by the Centers for Disease Control and Prevention (CDC), provide
tremendous outreach efforts across the nation to enhance health and decrease disease
burden. The Tulane PRC, which has been working in the New Orleans community since 1998,
has the mission of preventing or reducing overweight and obesity by addressing the physical
and social environmental factors that influence diet and physical activity. The Tulane PRC
strives to achieve its mission through community outreach, engagement and partnering.
Experience has shown that only multiple methods at various levels can achieve specified goals
and objectives; therefore, the purpose of this session is to demonstrate and discuss
community outreach methods, focused on: 1) technology via social media; 2) social
marketing; 3) grassroots engagement through the Tulane PRC Neighborhood Ambassador
Program; 4) policy research; and 5) education of stakeholders and local government. The
Tulane PRC community-based efforts have included the development and use of community
coalitions for specific programs, evaluating and promoting new bike lanes, providing input on
revised zoning around schools, facilitating joint-use agreements between schools and the city,
and promoting farmers' and mobile markets. One example is the building of a walking path
which provided initial impetus for a total neighborhood revitalization program. Various other
examples of successes and challenges will be presented, along with group discussion. The
lessons learned are very clear: engage community residents as your partners for any
community-based programming through any combination of methods that work.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S647

CORRESPONDING AUTHOR: Carolyn C. Johnson, PhD MS, Tulane University School of Public
Health and Tropical Medicine, New Orleans, LA, 70112; [email protected]
S648 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Panel Discussion 9 12:15 PM-1:15 PM

ENHANCING THE EFFECTIVENESS OF PHYSICAL ACTIVITY INTERVENTIONS IN MILITARY

VETERANS THROUGH CULTURAL TAILORING

David E. Goodrich, EdD, MA, MS1, Katherine S. Hall, PhD2, Ken Resnicow, Ph.D.3

1
Center for Clinical Management Research, VA Ann Arbor HCS, Ann Arbor, MI; 2Durham VA
and Duke University Medical Centers, Durham, NC; 3University of Michigan, Ann Arbor, MI

Military veterans represent almost one in ten adult Americans and yet represent a health
population not widely understood by those who have not served this group. Veterans are also
a medically complex and vulnerable population that has been exposed to a number of
psychological, physical, and environmental stressors as part of their military service. These
stressors, such as participation in combat, are often unique to the service member population
and contribute to a disparate distribution of chronic disease and debilitating conditions
among veterans. While active duty service members are healthier than nonveterans due to
mandatory fitness standards, many veterans become insufficiently active and develop health
problems soon after separation from the military. This critical life transition may be best
characterized as a culture shock for some veterans who struggle to adapt without the
sociocultural benefits and supports of military life, and constitutes a key time for behavioral
health interventions.

This panel will help clinicians and researchers understand the importance of cultural tailoring
to promote physical activity and identify innovative strategies for delivering targeted activity
programs in veteran populations. Panelists will illustrate how tailoring can go beyond surface
elements of culture such as veterans preferred physical activities or known barriers to
exercise, to deeper cultural structures of the warrior identity such as mission, values, and
military as family that can enhance motivation. The first panelist will present salient
characteristics of military culture that can be used to enhance program design and messaging
and increase veteran engagement in physical activity programs. The second panelist will
review current determinants of physical activity participation for veterans in the Veterans
Health Administration and community settings, highlighting areas for improvement and
examples where cultural tailoring in program development and implementation has been
successful. These presentations will be followed by an interactive discussion with the
audience moderated by an innovator in culturally-tailored health messaging and community-
based interventions that will emphasize application to real world interventions.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S649

CORRESPONDING AUTHOR: David E. Goodrich, EdD, MA, MS, Center for Clinical Management
Research, VA Ann Arbor HCS, Ann Arbor, MI, 48105; [email protected]
S650 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Panel Discussion 10 12:15 PM-1:15 PM

MINDFULNESS TRAINING AND CARDIOVASCULAR DISEASE PREVENTION: FACTS AND MYTHS.

Elena Salmoirago-Blotcher, M.D., Ph.D.1, Eric B. Loucks, PhD2, Judson A. Brewer, MD/PhD3

1
The Miriam Hospital & Alpert Medical School of Brown University, Providence, RI; 2Brown
University School of Public Health, Providence, RI; 3UMass Medical School, Worcester, MA

Nearly half of the observed reduction in coronary heart disease deaths over the past several
decades is attributable to improvements in coronary risk factors such as smoking,
hypertension, and high cholesterol, with the other half due to progress in medical treatment
and procedures. Despite these encouraging results, much remains to be done in the realm of
cardiovascular disease prevention, as it remains the number one cause of death worldwide.
First, adherence to lifesaving medications like statins and blood pressure lowering drugs is not
optimal. Second, maintaining healthy lifestyle changes over the long term remains an elusive
goal. Intense research efforts have been directed into finding novel and creative approaches
to promote lasting changes in these important behavioral risk factors.

Recent, rigorous meta-analyses have shown that mindfulness training (MT) is effective in
preventing depression relapse with moderate to large effect sizes. But can it impact the
physical health realm, including cardiovascular health? This panel discussion will focus on a
novel and controversial aspect, namely the possible role of MT in the promotion of
cardiovascular health. Uniquely, we will examine this topic from an epidemiological,
neurophysiological, and medical perspective.

We will first present the current evidence supporting the role of mindfulness in the promotion
of cardiovascular health i.e., through improvements in important cardiovascular risk factors
such as hypertension, smoking, diet, obesity, and physical activity.

Next, we will offer an overview of the posited neurophysiological and psychological

mechanisms by which MT can promote cardiovascular health namely its effects on known
psychological contributors as well as its effects in changing unhealthy habits.

Finally, we will present the cardiologists perspective with a focus on challenges to the
clinical application of this approach and barriers to implementation in a clinical setting. We
will conclude with a panelist discussion and Q&A session to obtain panelists and audiences
feedback and input on this novel approach to cardiovascular prevention.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S651

CORRESPONDING AUTHOR: Elena Salmoirago-Blotcher, M.D., Ph.D., The Miriam Hospital &
Alpert Medical School of Brown University, Providence, RI, 02903; elena_salmoirago-
[email protected]
S652 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Panel Discussion 11 12:15 PM-1:15 PM

BEHAVIORAL SLEEP MEDICINE: CAREER PATHS AND FUTURE CAREER OPPORUTNITIES

Michelle Drerup, PsyD1, Kelly Glazer. Baron, PhD, MPH, CBSM2, Christina McCrae, PhD3,
Stephanie A. Hooker, M.S., M.P.H.2, Samantha Domingo, Psy.D.1

1
Cleveland Clinic, Cleveland, OH; 2Rush University Medical Center, Chicago, IL; 3University of
Missouri-Columbia, Columbia, MO

Behavioral Sleep Medicine is an area of tremendous opportunities in behavioral medicine

because common sleep disorders such as insomnia and obstructive sleep apnea lower quality
of life and contribute to physical and mental health problems. Sleep disorders and lack of
sleep are an often underdiagnosed and under treated threat to public health. Behavioral
Sleep Medicine is a field of clinical practice and scientific inquiry that encompasses the study
of behavioral, psychological, and physiological factors underlying normal and disordered sleep
across the life span; and, the development and application of evidence-based behavioral and
psychological approaches to the prevention and treatment of sleep disorders and co-existing
conditions. The panel members will discuss the different settings that Behavioral Sleep
Medicine (BSM) providers are embedded including academia, private practice and
multidisciplinary sleep disorder centers housed in hospitals or medical centers. The panel
members will offer their advice on a wide range of professional and career development
issues in the area of BSM, including graduate and post-graduate training opportunities,
interdisciplinary collaboration, and future growth opportunities in the field.

CORRESPONDING AUTHOR: Michelle Drerup, PsyD, Cleveland Clinic, Cleveland, OH, 44195;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S653

Panel Discussion 12 12:15 PM-1:15 PM

MID-CAREER IN BEHAVIORAL MEDICINE: STAYING PRODUCTIVE AND CREATIVE AMIDST

CAREER CHALLENGES, TRANSITIONS, AND ADVANCEMENTS

Karen Oliver, Ph.D.1, Catherine Alfano, PhD2, Tracey A. Revenson, PhD3, Christine Rini, PhD4

1
Brown Alpert Medical School/ Providence VAMC, Providence, RI; 2American Cancer Society,
Washington, DC, DC; 3Hunter College, New York, NY; 4University of North Carolina at Chapel
Hill, Chapel Hill, NC

This panel discussion is geared toward professionals in behavioral medicine whose work
includes combinations of research, teaching, clinical care, and/or administration, and who are
navigatingor preparing to navigatethe challenges of mid-career. Professional paths in
mid-career are often less well-defined than paths through early career, and the challenges of
this career stage are often quite different. Challenges and opportunities in mid-career may
involve issues surrounding promotion, giving and getting mentoring, leadership, funding,
institutional or leadership transitions, balancing increasing demands on time or performance,
and potential changes in work setting, position, or environment. Moreover, mentoring
regarding decisions, goals, and challenges is often less available to mid-career professionals
than to those just beginning their career in behavioral medicine. Questions regarding how to
stay creative, generative, and motivated often arise, as do questions about career trajectories
and goals, and how to learn about and prepare for them. Questions regarding practical issues
to consider when preparing for career advancement are also common and answers are not
always easy to find. This panel will discuss these issues across a variety of behavioral medicine
careers. Panelists will share personal experiences and strategies for successfully coping with
mid-career challenges while capitalizing on opportunities. Discussion will also include lessons
learned and will provide strategies for audience members to consider when faced with their
own unique mid-career challenges. This panel is designed to be interactive and there will be
time for questions and discussion.

CORRESPONDING AUTHOR: Karen Oliver, Ph.D., Brown Alpert Medical School/ Providence
VAMC, Providence, RI, 02908; [email protected]
S654 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Panel Discussion 13 12:15 PM-1:15 PM

NATUROPATHIC MEDICINE 101 - INTRODUCTION TO THEORY AND PRACTICE

Brad S. Lichtenstein, Doctorate of Naturopathic Medicine 1, Cyndi Hope, Doctorate of

Naturopathic Medicine 2

1
Bastyr University, Lake Forest Park, WA; 2Bastyr University, San Diego, CA

When Engel proposed his bio-psycho-social model in the late 1970s to address the limits of
biomedicine, the field of naturopathic medicine had only 2 medical schools, and small number
of licensed providers spread out across the US and Canada. Currently, with seven
naturopathic medical schools in North America and more in development, the past two
decades has seen a growing interest in this comprehensive medical discipline that is grounded
in a whole person, behavioral medicine approach to care. 17 states, DC and the US territories
of Puerto Rico and the US Virgin Islands have licensing or regulation laws for naturopathic
doctors (NDs), which include graduation from an accredited four-year residential naturopathic
medical school and passing postdoctoral board examination (NPLEX).

As a distinct, integrative form of primary health care profession, naturopathic medicine

emphasizes prevention, treatment, and optimal health through the use of therapeutic
methods and substances that encourage individuals inherent self-healing process, and
acknowledges the importance of behavioral, socioculturalpolitical, cognitive, and emotional
factors to treat the whole person. Naturopathic doctors diagnose, treat, and help prevent
diseases using a system of practice that is based on the natural healing capacity of individuals.
Trained in a wide-array of modalities and therapies, including physiological, psychological or
mechanical methods,naturopathic physicians use natural medicines, prescription or legend
drugs, foods, herbs, and other natural remedies to treat their patients.

Basic knowledge about naturopathic medicine can benefit patients and providers alike. This
presentation will examine the theory, practice and scope of naturopathic medicine, and will
introduce the guiding principles of discipline which include The Healing Power of Nature (Vis
Medicatrix Naturae), Above all Else, Do No Harm (Primum Non Nocere), Treat the Cause (Tolle
Causum), Treat the Whole (Tolle Totum), Doctor as Teacher (Docere), and Prevention
(Preventare). The naturopathic therapeutic order will be outlined and case examples will be
used to illustrate the concepts.

CORRESPONDING AUTHOR: Brad S. Lichtenstein, Doctorate of Naturopathic Medicine , Bastyr

University, Lake Forest Park, WA, 98155; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S655

Symposium 19 1:30 PM-2:45 PM

UNPACKING RISK PERCEPTIONS: HOW, WHEN, AND FOR WHOM FEELINGS OF RISK MOTIVATE
BEHAVIOR CHANGE

Erin M. Ellis, PhD, MPH1, Rebecca A. Ferrer, PhD1, Linda D. Cameron, Ph. D. 2, Noel T. Brewer,
PhD3

1
National Cancer Institute, Rockville, MD; 2University of California, Merced, Merced, CA;
3
University of North Carolina, Chapel Hill, Chapel Hill, NC

Many health decision-making theories center on perceived risk of health threats as a

motivator of health behavior change. Recent work has explored the unique influences of
cognitive and affective subcomponents of risk perceptions, often finding that affective risk
perceptions such as worry can exceed, moderate, or mediate the effect of cognitively-based
risk perceptions on behavior. It is therefore necessary to examine methodological, contextual,
and person-level factors that may shape the strength of the associations between affective
risk perceptions and behavior. The aim of this symposium is to highlight the role of affective
risk perceptions as a key driver of health behavior, and to extend this work to examine how
the cognitive-affective interplay varies across people or behaviors, key measurement and
methodology decisions in understanding these different components of risk perceptions, and
implications for health communication interventions. The first presentation reports on a
multidimensional model of risk perceptions used to predict motivation to engage in 32
different health protective behaviors. Although affective risk perceptions were a consistently
strong predictor of motivation to engage in health behavior, they were particularly predictive
for threats perceived to be less severe and for people who habitually used specific emotion
regulation techniques, suggesting a role for both contextual and person-level characteristics.
In the second presentation, willingness to take a disease-preventive medication that caused
side effects was associated with cognitive and affective risk perceptions, but only when these
risk perceptions were conditioned on whether or not one takes the medication, highlighting
the importance of conditionality. In the third study, theoretical work disentangling cognitive
and affective risk perceptions is applied in a health communication intervention that utilized
metaphoric messaging about skin cancer. The results suggest that affective risk perceptions
are important as both a mediator and moderator of the interventions effect on sun
protection motivation. A discussant whose expertise in risk perceptions spans descriptive,
intervention, and meta-analytic research across numerous behavioral contexts will synthesize
these studies, highlighting key themes and remaining questions. Together, these innovative
S656 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

state-of-the-science presentations connect research in social psychology, health psychology,

and public health to extend risk perception theory and intervention efforts by highlighting
when and for whom affective risk perceptions motivate behavior change.

Symposium 19A

RISK PERCEPTIONS AND MEDICAL DECISIONS INVOLVING TRADE-OFFS

Dr. Erin M. Ellis, PhD, MPH

Background: Perceived risk of a health threat can motivate health-protective behaviors such
as using medications that reduce disease risk. However, these medications are often
accompanied by side effects, including elevated risks for other adverse health effects, and
perceived risk of these side effects may also influence treatment decisions. Moreover, risk
perceptions are complex, with deliberative and affective subcomponents whose relation to
behavior can differ depending on whether they are queried as conditional or non-conditional
on engaging in behaviors that may influence ones risk (e.g., smoking). This study explored the
role of deliberative and affective risk perceptions in a medical trade-off decision involving a
medication that lowered risk for one disease while increasing risk for another, and whether
these associations differed when risk perceptions were queried as conditional on taking or not
taking the medication.

Methods: Participants (N = 310) read about a hypothetical medication that reduced their risk
for a target condition (e.g., cancer), but increased their risk for a side effect (e.g., heart
disease). They provided their deliberative and affective risk perceptions about the target
condition and side effect, both conditional and not conditional on whether they took the
medication. Then they indicated their willingness to take the medication.

Results: When risk perceptions were not queried as conditional on whether the medication
was taken, neither perceived risk of the target condition nor side effect was associated with
willingness to take the medication, regardless of whether they were deliberative, target
condition: b = 0.030, p = .82 95% CI (-0.23, 0.29); side effect: b = 0.032, p = .82, 95% CI (-0.23,
0.30), or affective in nature, target condition: b = 0.18, p = .13 95% CI (-0.056, 0.43); side
effect: b = -0.10, p = .40, 95% CI (-0.34, 0.14). On the other hand, greater perceived risk of the
target condition conditioned on not taking the medication was associated greater willingness
to take it, deliberative: b = 0.30, p = .041; 95% CI (0.012, 0.58); affective: b = 0.23, p = .030,
95% CI (0.023, 0.44). For the side effect, greater perceptions of risk if one did take the
medication were associated with less willingness to take it, deliberative: b = -0.35, p = .006,
95% CI (-0.60, -0.10); affective: b = -0.25, p = .026, 95% CI (-0.46, -0.030).

Conclusions: Thoughts and feelings of risk about a medications targeted health condition if
one did not take a medication were associated with greater willingness to take it, whereas
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S657

thoughts and feelings of risk about its side effects if one did take the medication reduced
willingness to take it. These findings suggest conditioning both deliberative and affective risk
perceptions on specific behavioral contexts may better capture risk perceptions role in
medical decisions that involve trade-offs.

CORRESPONDING AUTHORS: Rebecca Ferrer, PhD, National Cancer Institute;

[email protected]; William Klein, PhD, National Cancer Institute; [email protected]

Symposium 19B

WHEN DO DELIBERATIVE, AFFECTIVE, AND EXPERIENTIAL RISK PERCEPTIONS PREDICT

PROTECTION MOTIVATION FOR HEALTH THREATS?

Rebecca A. Ferrer, PhD

Although risk perception is a key concept in many health behavior theories, little research has
explicitly tested when different types of risk perception (i.e., deliberative, affective, and
experiential) predict motivation to take protective action against a health threat (protection
motivation). The present study tackled this question by (a) adopting a multidimensional model
of risk perception that comprises deliberative, affective, and experiential components (the
TRIRISK model), and (b) taking a person-by-situation approach. We leveraged a highly
intensive paradigm to test health threat features (i.e., perceived severity) and individual
differences (e.g., emotion reappraisal) as moderators of the relationship between risk
perception and protection motivation. Multi-level modeling of 2968 observations (32 health
threats across 94 individuals) demonstrated that all three risk perceptions were associated
with protection motivation, with affective risk perception being the most powerful predictor
(B = 0.95, p < .001, d = 1.14), compared to deliberative (B = -0.40, p < .001, d = -0.41) and
experiential (B = 0.11, p =.007, d = 0.10) risk perceptions. Analyses also showed interactions
among the TRIRISK components and moderation both by person-level and situational factors.
For instance, affective risk perception better predicted protection motivation when
deliberative risk perception was high (B = 0.05, p < .001, d = 0.13), when the threat was less
severe (B = -0.03, p < .001, d = -0.17), and among participants who engage less in emotional
reappraisal (B = -0.06, p = .002, d = -0.11). Experiential risk perceptions were more strongly
associated with protection motivation when deliberative risk perception was low (B = -
0.04, p < .001, d = 0.10), and when cognitive reflection was low (B = -0.05, p = .009, d = -
0.10). These findings support the TRIRISK model and offer new insights into when deliberative,
affective, and experiential risk perceptions predict health protection motivation.

CORRESPONDING AUTHORS: William Klein, PhD, National Cancer Institute;

[email protected]; Aya Avishai-Yitshak, University of North Carolina, Chapel Hill;
[email protected]; Katelyn Jones, University of North Carolina, Chapel Hill;
S658 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

[email protected]; Paschal Sheeran, PhD, University of North Carolina, Chapel Hill;

[email protected]

Symposium 19C

METAPHOR USE IN SKIN CANCER PREVENTION MESSAGES: MEDIATING AND MODERATING

ROLES OF AFFECTIVE RISK APPRAISALS

Dr. Linda D. Cameron, Ph. D.

Framing health messages with potent metaphors represents a promising communication

strategy for motivating protection behavior, yet little is known about how, when, and for
whom they do so. Guided by conceptual metaphor theory, we tested whether messages
framing a health risk metaphorically as a familiar hazard (versus literally) increases worry and
protection motivations, and whether these effects are specific to individuals with high (versus
low) fear of that hazard. We also tested if messages combining matched
(metaphoric/metaphoric) versus mismatched (metaphoric/literal; literal/metaphoric)
framings of risk and protection actions increase worry and response engagement. In an
intervention trial, we tested messages about sun exposure risk framed metaphorically as an
enemy sun attacking the skin (versus framed literally) and sun-safe products as superheroes
that protect users (versus framed literally). We predicted that an inherent fear of enemies
would moderate (enhance) their effects on skin cancer worry, sunscreen intentions, and
willingness to pay (WTP) for sunscreen; and that skin cancer worry (but not perceived risk)
would mediate their effects on sunscreen intentions and WTP. Adults (N = 174) recruited at
outdoor recreational areas during summer completed an enemy fear measure; read a flyer
with one of the four sets of risk and response messages; and completed measures of worry,
perceived risk, sunscreen intentions and WTP. Regression and Process model analyses
supported predictions. When the message framed UV risk metaphorically as an enemy sun
and sun protection metaphorically as superheroes, recipients high in enemy fear were more
worried about skin cancer, more motivated to use sunscreen, and willing to pay higher prices
for sunscreen (enemy fear X risk framing X response framing interaction ts > 2.16, ps < .03).
For those low in enemy fear, the metaphoric messages did not affect worry, intentions, or
WTP. Worry mediated the 3-way interaction effects on sunscreen intentions, 95% CI = [0.03,
0.33], and WTP, 95% CI = [0.04, 0.50]. Neither metaphoric message influenced perceived risk.
Affective risk appraisals (metaphor-relevant fear and worry) can moderate and mediate
metaphors effects on protection motivations.

CORRESPONDING AUTHORS: Mark Landau, PhD, University of Kansas; [email protected];

Jamie Arndt, PhD, University of Missouri; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S659

Symposium 20 1:30 PM-2:45 PM

LA BUENA VISTA: TRANSLATING 40 YEARS OF CANCER INTERVENTION RESEARCH INTO POLICY

AND PRACTICE TO IMPROVE HEALTH OUTCOMES

Shawna Ehlers, PhD, ABPP1, Barbara L. Andersen, PhD2, Qian Lu, MD, PhD3

1
Mayo Clinic, Rochester, MN; 2The Ohio State University, Columbus, OH; 3University of
Houston, Houston, TX

Cancer distress screening and referral for intervention are now required to maintain
accreditation by the American College of Surgeons, thus providing "teeth" to longstanding
cancer distress screening guidelines. National efforts to implement these guidelines have
resulted in significant opportunities to advance implementation science via the creation of
natural laboratories in cancer health care settings. This presentation will use the NIH-
translational pathway to define the trajectory of 40 years of cancer intervention research. Dr.
Ehlers will serve as Chair with introduction of speakers and the structure of the presentation.
Dr. Andersen will review 40 years of research, inclusive of promising advancements, stalling
points, and current opportunities to speed translation to "real world" health care settings. Dr.
Lu will then focus in on the importance of broadening research efforts to reduce health
disparities, presenting a contextual, ecologically derived, systems model to understand
culture and thus to develop culturally sensitive interventions. Dr. Ehlers will serve as
Discussant, including discussion of possible approaches to align stakeholders focused on late
stage translational research, evidence-based practice, and the "Triple Aim" of health care
delivery.

Symposium 20A

THE TRAJECTORY OF CANCER DISTRESS INTERVENTION RESEARCH: CELEBRATING

ADVANCEMENTS AND NAVIGATING BARRIERS TO TRANSLATION

Dr. Barbara L. Andersen, PhD

We have 40 years of cancer intervention research. The milestones of this research body will
be reviewed to establish our trajectory as a framework for the presentation, utilizing major
meta-analytic and review papers. Highlights include progress in research designs, patient
selection, intervention development, and broadening outcomes to a biobehavioral model.
S660 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Then a perspective will be offered on recent research efforts, including advancements in

knowledge and the redundancy of small impact results. The presentation will conclude with a
perspective designed to utilize these advancements and propel translation forward to
dissemination in modern day oncology health care settings.

Symposium 20B

TRANSLATION OF CANCER INTERVENTION RESEARCH TO PROVIDE CULTURALLY SENSITIVE

PSYCHOSOCIAL INTERVENTIONS THAT REDUCE CANCER HEALTH DISPARITIES

Qian Lu, MD, PhD

Considerable cancer health disparities still exist and persist in the U.S. Elimination of health
disparities requires a theoretically guided, multidimensional, culturally sensitive approach to
illness prevention and behavioral intervention. Culture is the core and fundamental system of
life, designed to ensure the survival and well-being of its members. As a result, culture is part
of a complex system that is challenging to define and study. Using a contextual, ecologically
derived, systems model to understand culture and to develop culturally sensitive intervention
may be the key to reduce and eliminate health disparities. By focusing on understudied and
underserved populations, this presentation aims to illustrate that contextual models hold the
promise to build cultural competence and move the science forward to understand and
eliminate health disparity in cancer. The presenter will first review psychosocial interventions
among minority cancer survivors for the past 40 years, and then provide examples of using a
community based participatory approach to develop and implement culturally sensitive
interventions among Asian American breast cancer survivors. An example of intervention
development will be provided in examination of Chinese American breast cancer survivors
post-treatment (N=19) who were randomly assigned to a usual care control group, or an
intervention group designed to provide emotional and informational support through peer
mentoring and education. Compared with control group, the intervention group reported
improvements in health outcomes in medium-to-large effect sizes (including overall,
emotional, functional and additional concerns quality of life) post-intervention. Future studies
with other minority groups are warranted to better understand the efficacy and the
mechanisms of social support interventions across culture.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S661

Symposium 21 1:30 PM-2:45 PM

SOCIAL DETERMINANTS OF PHYSICAL ACTIVITY AMONG LATINOS: FROM OBSERVATIONAL

RESEARCH TO TRANSLATIONAL INTERVENTIONS

Sandra Echeverria, PhD1, Rosenda Murillo, PhD MPH2, Elizabeth Vasquez, DrPH3, Sandra
Echeverria, PhD1, Ana F. Abraido-Lanza, PhD 4

1
CUNY Graduate School of Public Health and Health Policy, New York, NY; 2University of
Houston, Houston, TX; 3SUNY School of Public Health, Albany, NY; 4Columbia University, New
York, NY

Latinos are the fastest growing population in the United States and are disproportionately
affected by overweight/ obesity and participate in low levels of physical activity (PA). A well-
documented body of evidence has indicated the social patterning of cardiovascular risk
factors, including physical inactivity. However, limited research exists investigating social
determinants of active living among Latinos. In this symposium, we present findings using
nationally representative data (National Health Interview Survey) to examine associations
between key social factors- nativity, occupational status, and neighborhood perceptions- and
distinct measures of physical activity. Specifically, we examine how neighborhood perceptions
are associated with mobility in older Latino adults, and then examine differences in walking
for transport and walking for leisure by nativity status and U.S.-based census occupational
categories in middle-aged adults. Building on the population-based observational studies
presented, the symposium provides an example of a pilot study that takes into account
complex cultural, social, and organizational factors in a communtiy-based PA intervention
designed for Latinos. The discussant closes the session highlighting the importance of
integrating observational and translational research to more effectively promote adoption of
PA across the lifespan in this vulnerable population.

Symposium 21A

NATIVITY AND OCCUPATIONAL STATUS INFLUENCING WALKING BEHAVIORS AMONG LATINOS

Rosenda Murillo, PhD MPH

Walking has long been associated with numerous health benefits, including lower all-cause
and cardiovascular mortality. Only select studies have investigated walking patterns among
Latinos, and no study has investigated if walking purpose differs by key social determinants
S662 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

such as race/ ethnicity, nativity and occupational status. We analyzed the National Health
Interview Survey (NHIS), a yearly nationally representative household survey of the non-
institutionalized United States (US) population, which in 2010 collected data on walking for
leisure and walking for transport. We classified Latinos (n=5,158) by nativity status and
number of years in the US and compared them to non-Latino Whites (n=15,570). The
prevalence of each walking behavior was calculated, and duration and frequency of walks in
the past week were used to estimate if individuals met national physical activity (PA)
recommendations of engaging in at least 150+ minutes of moderate physical activity per
week. A clear gradient in walking for transport was observed, with 42% of immigrants living in
the US less than 10 years walking for transport and dropping to 30% in US-born Latinos. In
logistic regression models adjusted for age, sex, education, poverty level, and obesity, Latinos
showed a significantly higher probability of being sufficiently active through transportation
activity compared to non-Latino whites [Prevalence Ratio (PR)= 1.62, 95% Confidence Interval
(CI)= 1.17, 2.22)]. For leisure walking, none of the models were statistically significant. Tests
for interaction showed that occupational status modified observed associations between
nativity status and walking for transport and leisure, showing more pronounced differences
for lower status and physically demanding occupations. Walking initiatives need to be tailored
to working poor and culturally diverse populations.

Symposium 21B

NEIGHBORHOOD SOCIAL COHESION AND MOBILITY LIMITATIONS AMONG OLDER RACIALLY/

ETHNICALLY ADULTS IN THE UNITED STATES

Elizabeth Vasquez, DrPH

Background: Unrestricted walking and climbing ability are markers of functional performance
and health in older adults. Several studies have shown that neighborhood contexts can
influence mobility in older populations, but little evidence exists on elderly Latinos. This study
examines associations between neighborhood social cohesion and mobility limitations in a
national sample of older adults and specifically tested if associations differed by race/ethnicity
and present disparities faced by Latino older adults.

Methods: We combined the 2013 and 2014 years of the National Health Interview Survey
(NHIS), a nationally representative survey of non-institutionalized adults of the United States,
and selected adults 60 years and older. Mobility limitation with walking, climbing or both
(total) were determined from NHIS questions asking if the participant experienced difficulty
walking or climbing steps. Self-reported neighborhood social cohesion was evaluated with
four questions about whether neighbors help one another, can be counted on, can be trusted,
or are close-knit. All four questions answers were summed and divided into three categories
representing high, medium and low neighborhood social cohesion. Prevalence of mobility
limitations were calculated for each neighborhood category and adjusted for age, sex,
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S663

education, and physical activity. Models were further stratified by race/ ethnicity and nativity
status for Latinos.

Results: Non-Latino whites (NLW) had the highest proportion of individuals living in
neighborhoods with high neighborhood social cohesion (42.5%), followed by Latinos (28.5%),
and non-Latino blacks (28.4%).When compared with low neighborhood social cohesion living
in a neighborhood with high social cohesion conferred a protective effect for total mobility
limitation (OR: 0.70; 95% CI: 0.62, 0.78), walking (OR: 0.59; 95% CI: 0.53, 0.67) and climbing
limitations (OR: 0.66; 95% CI: 0.59, 0.74) adjusting for race/ethnicity. The association
remained statistically significant after further adjustment for sex, age, physical activity, and
education. In stratified analyses, foreign-born Latinos living in socially cohesive neighborhoods
had lower mobility limitation than the US born.

Conclusion: Neighborhood conditions may be modifiable targets for interventions to reduce

socioeconomic and racial/ethnic disparities in mobility limitations and subsequent disability in
older adults.

Symposium 21C

PROMOTING PHYSICAL ACTIVITY AMONG LATINOS WITH DIABETES: THE PHYSICAL ACTIVITY
SYSTEM OF SUPPORT (PASOS)

Sandra Echeverria, PhD

Latinos are less likely to be physically active than their non-Latino White peers and have a
disproportionate burden of Type II diabetes. They are also at increased risk of diabetes-
related health complications and mortality. Randomized clinical trials have shown that
increasing physical activity (PA) for patients with diabetes can improve glucose and lipid
levels, insulin resistance, and weight reduction goals. Nonetheless, there is limited evidence
on the effectiveness of physical activity interventions targeting underserved groups,
particularly Latinos with low income and limited English proficiency who may face greater
challenges in adopting physical activity behavior change. We developed the Physical Activity
System of Support (PASOS) program, a community-based intervention designed to increase
physical activity among Latinos living with diabetes. The intervention consisted of a culturally-
tailored physical activity intervention consisting of scheduled exercises offered twice per week
over an 8 week period. We undertook a collaborative process to develop the intervention and
describe the multi-level nature of the intervention which incorporates community,
organizational, and group-based behavior change approaches. The feasibility of the design
and implementation arepresented. A total of 30 participants were enrolled in the study and
nearly all participants were poor, had limited English proficiency and had no or limited health
insurance coverage. There was over a 70% participation rate and pre-post intervention
S664 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

measures showed significant increases in self-reported PA. A 'systems' model is proposed to

support long-term adoption of PA among Latinos.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S665

Symposium 22 1:30 PM-2:45 PM

ADVANCING BEHAVIORAL MEDICINE RESEARCH USING DIGITAL TOOLS TO IMPROVE HEALTH

E. Amy Janke, PhD1, Danielle E. Jake-Schoffman, PhD2, Claudio R. Nigg, PhD, FSBM3, Kassandra
I. Alcaraz, PhD, MPH4, Jeffrey Haibach, PhD, MPH5, Beth Lewis, PhD6, Jennifer Duffecy, Ph.D.7

1
University of the Sciences, Philadelphia, PA; 2University of Massachusetts Medical School,
Worcest, MA; 3University of Hawaii, Honolulu, HI; 4American Cancer Society, Atlanta, GA;
5
U.S. Department of Veterans Affairs, Washington, DC; 6University of Minnesota, Minneapolis,
MN; 7University of Illinois at Chicago, Chicago, IL

The Society of Behavioral Medicine is uniquely positioned to lead the way in underscoring the
dynamic nature of behavioral medicine to advance the nations health and provide direction
for the next generation of leaders in the field. At the forefront of these efforts are SBMs
Special Interest Groups (SIGs) that allow members to connect with others who share common
interests. During the past year, SBMs SIG membership has been working on outlining a call to
action from the perspective of each of their unique topic areas to promote research
advancing behavioral science to inform policy and practice. Many of these forward-thinking
papers will be highlighted in an upcoming SIG-focused special issue on Advancing Behavioral
Medicine Research: A Call to Action to be published in the Journal of Behavioral Medicine.
However, despite the great variety in our SIGs, some common themes emerged across all
special issue papers, and one salient theme was the important role of digital tools to improve
health. This symposium will highlight the importance of technology to the future of behavioral
medicine research and practice as outlined by 4 different SIGs: Ethnic Minority and
Multicultural Health (EMMH), Military and Veterans Health, Physical Activity, and Evidence
Based Behavioral Medicine (EBBM). The EMMH SIG will present their ConNECT Framework as
a model for linking the sciences of health equity and behavioral medicine and discuss the
ConNECT core principle of Harnessing Communication Technology as a means to assess,
understand, document/track, and intervene to advance health equity. The Military and
Veterans Health SIG will discuss the current landscape of mobile health and other
technologies being implemented with military servicemembers, veterans, and their families,
and review the unique opportunities and potential barriers to research and practice using
such technology in these populations. The PA SIG will review emerging areas of PA
intervention research, including the important role technology plays in innovation. Finally, the
EBBM SIG will discuss how innovative intervention designs and data collection approaches
integrate technology to increase the public health impact of behavioral medicine.

Symposium 22A
S666 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

HARNESSING COMMUNICATION TECHNOLOGY TO ADVANCE HEALTH EQUITY: PROMISE,

POTENTIAL AND POSSIBILITIES FOR THE 21ST CENTURY

Dr. Kassandra I. Alcaraz, PhD, MPH

Colleagues from SBMs Ethnic Minority and Multicultural Health SIG recently introduced the
ConNECT Framework as an actionable model for linking the sciences of health equity and
behavioral medicine with the goal of achieving equitable health and outcomes in the 21 st
century. One of ConNECTs core principles, Harnessing Communication Technology,
recognizes the potential of technology in assessing, understanding, documenting/tracking,
and intervening to advance health equity. Although utilization of emerging technology in
behavioral medicine is becoming commonplace, critical gaps remain in employing these
innovations to address health disparities, especially among ethnic minority and understudied
populations. Using the ConNECT Framework as a guide, this presentation covers four topics:
(1) the need for culturally, contextually, and linguistically/literacy relevant content and
approaches in communication technology interventions; (2) the promise of Connected Health
for broad dissemination of health information and interventions to reduce health disparities;
(3) ethical, cultural, and infrastructure considerations for ensuring equitable access,
functionality, and usability of technology for multicultural and multilingual groups; and (4)
specific recommendations for leveraging emerging communication technologies to enhance
reach and dissemination. Communication technology has alteredand continues to
revolutionizethe landscape of behavioral medicine. Focused efforts to engage and benefit
medically vulnerable populations in this technological era are vital for advancing not only
health equity but behavioral medicine as a whole.

CORRESPONDING AUTHORS: Linda Fleisher, PhD, MPH, Children's Hospital of Philadelphia;

[email protected]; Qian Lu, MD, PhD, University of Houston; [email protected]; Sabrina
Ford, PhD, Michigan State University; [email protected]; Cathy Meade, PhD, RN,
FAAN, Moffitt Cancer Center; [email protected]; Jamilia Sly, PhD, Icahn School of
Medicine at Mount Sinai; [email protected]; Kimlin Ashing, PhD, City of Hope; KAshing-
[email protected]; Virginia Gil-Rivas, PhD, University of North Carolina at Charlotte;
[email protected]; Jean Yi, PhD, Fred Hutchinson Cancer Research Center; [email protected];
Usha Menon, PhD, RN, FAAN, The University of Arizona; [email protected];
Clement Gwede, PhD, MPH, RN, FAAN, Moffitt Cancer Center; [email protected]

Symposium 22B

MOBILE HEALTH AND OTHER TECHNOLOGIES FOR MILITARY AND VETERAN HEALTH BEHAVIOR
CHANGE

Jeffrey Haibach, PhD, MPH

ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S667

Mobile health and other electronic health technologies provide opportunities to link DoDs 2.1
million servicemembers, VAs 8.9 million enrolled veterans, and millions more veterans and
family members to online resources for health behavior change. Primary areas in current
research and practice include nutrition counseling, physical activity promotion, smoking
cessation, reducing alcohol and other substance misuse, and broader health promotion and
coaching. E-health interventions provide opportunities to promote health through health risk
appraisals, monitoring changes in health status, social support, self-management, and
behavior change with motivational messages, problem-solving, and goal setting with
feedback. These technologies provide benefits to help servicemembers and veterans
overcome access barriers and stigmas related to treatment for obesity, substance abuse, and
mental health disorders such as depression or PTSD. There are further opportunities to
support continuity among transitioning servicemembers, augment clinical care, and reach
more of the population VA and DoD serve such as deployed servicemembers and veterans in
rural locations. While there are many potential benefits of using mobile technologies in
research and practice, there are also unique considerations for protection of DoD/VA patient
information and emerging performance enhancing technologies with unknown implications.
Additionally, some servicemembers and veterans do not have access to these technologies,
respectfully do not want not use them, and, more generally, excess screen time and
technology misuse are associated with negative health consequences. Overall and with
caveats in mind, there is an immense opportunity through research and practice with mobile
health and other e-health technologies to promote positive health behavior change for the
health of military servicemembers, veterans, and their families.

Symposium 22C

THE FUTURE OF PHYSICAL ACTIVITY INTERVENTION RESEARCH: EXPANDING FOCUS TO

SEDENTARY BEHAVIOR, TECHNOLOGY, AND DISSEMINATION

Beth Lewis, PhD

Despite the increased health risks of a sedentary lifestyle, only 49% of American adults
participate in physical activity (PA) at the recommended levels based on self-report data.
Objective data indicates that only 10% of Americans meet the PA recommended levels. In an
effort to move the PA field forward, we briefly review three emerging areas of PA intervention
research. First, new intervention research has focused on not only increasing PA but also on
decreasing sedentary behavior. Americans sit on average 7.5 hours per day as measured by
an accelerometer, which has significant implications for disease risk. Researchers should
utilize randomized controlled trials, common terminology, investigate which behaviors should
replace sedentary behaviors, evaluate long-term outcomes, and examine interventions across
the lifespan. Second, technology has contributed to an increase in sedentary behavior but has
also led to innovative PA interventions. With the prevailing reach of technology, future
research is needed to better understand which technological components can contribute to
S668 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

PA behavior change. PA technology research should focus on large randomized trials with
evidence-based components, explore social networking and innovative apps, improve PA
monitoring, consider the lifespan, and be grounded in theory. Finally, recently there has been
an increased emphasis on disseminating evidence-based physical activity and other health
interventions. In an effort to maximize public health impact, dissemination efforts should
include dissemination and implementation models, health disparities, and assessments of
intervention costs. As PA researchers, we need to challenge the status quo and explore new
ideas as strategies to improve PA interventions.

CORRESPONDING AUTHORS: Melissa Napolitano, PhD, The George Washington University;

[email protected]; Matthew Buman, PhD, Arizona State University;
[email protected]; David Williams, PhD, Brown University; [email protected];
Claudio Nigg, PhD, University of Hawaii; [email protected]

Symposium 22D

USING TECHNOLOGY TO INCREASE THE PUBLIC HEALTH IMPACT OF EVIDENCE-BASED

INTERVENTIONS IN BEHAVIORAL MEDICINE

Dr. Jennifer Duffecy, Ph.D.

The dissemination and implementation of evidence-based behavioral medicine interventions

into real world practice has been limited. Specific limitations of current behavioral medicine
research will be discussed within the context of the RE-AIM framework. Potential
opportunities to increase public health by focusing on novel intervention designs and data
collection approaches are key. Novel intervention designs include the MOST framework,
which aims to optimize multicomponent behavioral and bio-behavioral interventions, and
SMART designs, imbedded within the MOST framework, to optimize adaptive interventions.
Novel data collection approaches also have the potential to improve the public health
dissemination. This presentation will focus on how these innovative methods can use
technology to increase the public health impact of behavioral medicine. An example of an
effective technology delivered MOST trial will be covered as well as a discussion of technology
to improve data collection and dissemination, including technology based strategies to
increase retention in trials, increase adherence to healthcare recommendations, decrease
patient burden and improve fidelity of interventions delivered. Innovative methods, notably
using technology, may have the ability to increase the public health impact of evidence-based
behavioral interventions to promote health.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S669

CORRESPONDING AUTHORS: Joanna Buscemi, Ph.D., DePaul University;

[email protected]; E. Amy Janke, Ph.D., University of the Sciences;
[email protected]; Kari C. Kugler, Ph.D., Pennsylvania State University; [email protected];
Thelma J. Mielenz, Ph.D., Columbia University; [email protected]; Sara St. George,
Ph.D., University of Miami; [email protected]; Sherri N Sheinfeld Gorin, Ph.D., New
York Physicians against Cancer; [email protected]
S670 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 23 1:30 PM-2:45 PM

EXAMINING ROOT CONTRIBUTORS TO WEIGHT CONTROL: APPLYING SOCIAL PSYCHOLOGICAL

THEORIES IN WEIGHT MANAGEMENT PROGRAMS

Corrine Voils, PhD1, Danielle Arigo, Ph.D., Licensed Psychologist2, Tricia M. Leahey, PhD3, Amy
A. Gorin, Ph.D.4, Shannon S. Taylor, Ph.D.5, Kathryn I. Pollak, PhD6

1
Durham Veterans Affairs Medical Center, Duke University Medical Center, Durham, NC; 2The
University of Scranton, Scranton, PA; 3University of Connecticut, Storrs, CT; 4University of
Connecticut; Institution for Collaboration on Health, Intervention, and Policy, Storrs, CT;
5
Durham VA Medical Center, Durham, NC; 6Duke University, DURHAM, NC

Accumulating evidence demonstrates that the social environment and ones social
perceptions influence weight management behaviors (e.g., food choice, physical activity), and
thus, weight loss and maintenance. However, most research efforts focus on the individual in
isolation and do not take into account social processes or perceptions. As a result, none has
identified the best method(s) for optimizing social processes for weight control. Social
psychological theories help to identify relevant social processes to guide assessment and
intervention efforts in behavioral weight loss treatment to improve obesity outcomes. The
goal of this symposium is to describe three types of social influences on weight control:
intrapersonal processes (i.e., perceptions of the self, relative to others), interpersonal
processes (i.e., dyadic interactions), and group dynamics. Speakers in this session will present
empirical evidence from their weight management trials from each of these perspectives. At
the intrapersonal level, Dr. Arigo will report on the frequency of distinct social comparisons
(with respect to occurrence, direction, and dimension), and prospective relations between
comparisons and session attendance/weight loss/physical activity, from a daily diary study of
participants in a weight loss trial. With respect to dyadic interaction, Dr. Leahey will present
evidence for Social Accountability Theory, demonstrating that social accountability (i.e.,
from professional and peer coaches) is indeed an important construct within the context of an
e-health weight loss maintenance intervention. Dr. Gorin will then discuss the relationship
between autonomy support from a spouse/partner and weight loss in two different
behavioral programs. Finally, at the level of group dynamics, Dr. Taylor will describe the role
of group cohesion in predicting weight loss in a group-based intervention. The discussant, Dr.
Kathryn Pollak, will integrate these findings and discuss their implications for understanding
and improving behavioral weight control interventions.

Symposium 23A
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S671

DAILY SOCIAL COMPARISONS, WEIGHT LOSS, AND PHYSICAL ACTIVITY CONSISTENCY IN A

BEHAVIORAL WEIGHT LOSS PROGRAM

Dr. Danielle Arigo, Ph.D., Licensed Psychologist

Social comparisons (i.e., self-appraisals relative to others) made in daily life may influence
weight control outcomes such as treatment attendance, total weight loss, and physical
activity. However, the types of comparisons that may be beneficial for weight control are
unclear. The goals of this daily diary study were to (1) characterize the types of comparisons
made by weight loss treatment participants in daily life, and (2) to examine both between-
and within-person relations between daily comparisons and weight control outcomes.
Participants were adults seeking behavioral weight loss treatment (n=186, MAge=52,
MBMI=36.9 kg/m2). For 7 days at the start, middle, and end of a 1-year weight loss program,
participants responded to nightly prompts via mobile application; participants categorized
comparisons made each day in terms of domain (appearance, financial/work status, other)
and direction (comparisons to others deemed superior/upward, inferior/downward, or
similar/lateral to the self). Participants wore tri-axial waistband accelerometers to capture
minutes in bouts of moderate-to-vigorous physical activity (MVPA). Overall, comparisons were
reported on 17% of days and the majority of participants (63%) reported at least one
comparison (M=3.11); comparisons were more frequent at baseline (vs. mid- and end-of-
treatment; 2=15.49, pps>0.10). The overall number of comparisons reported was positively
associated with treatment session attendance (r=.20), though greater frequency of upward
comparisons was negatively associated with percent weight loss at 1 year (r=-0.20, ps<.01). At
the day level, upward comparisons were associated with greater between-day variability in
MVPA, relative to downward and lateral comparisons (2=9.50, p<.01). A significant
interaction between reports of skipping planned MVPA and upward appearance comparisons
showed that participants who skipped and made (vs. did not make) upward appearance
comparisons engaged in 15 additional minutes of MVPA the following day (F=13.34, p<.01).
Findings suggest that daily comparisons predict critical weight control outcomes. Additional
work is necessary to determine the mechanisms of these relationships, as well as best
practices for harnessing the benefits of comparisons in weight loss treatment programs.

CORRESPONDING AUTHORS: Leah Schumacher, M.S., Drexel University; [email protected];

Meghan Butryn, Ph.D., Drexel University; [email protected]; Evan Forman, Ph.D., Drexel
University; [email protected]

Symposium 23B

SUPPORTIVE ACCOUNTABILITY THEORY: EVIDENCE FROM AN E-HEALTH WEIGHT LOSS

MAINTENANCE TRIAL
S672 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Dr. Tricia M. Leahey, PhD

Supportive Accountability Theory asserts that e-health interventions with human involvement
enhance patient support and accountability, which consequently yields better treatment
outcomes. However, this theory is not well tested, and, to our knowledge, no one has
examined whether the construct of supportive accountability is indeed associated with e-
health treatment outcomes. The purpose of this study was to test Supportive Accountability
Theory. We examined whether the inclusion of human support in an e-health intervention for
weight loss maintenance does in fact enhance participant reported supportive accountability.
Further, we tested whether supportive accountability is associated with weight maintenance
outcomes. Participants (N= 72, 85% female, 48.5 years, 31.5 BMI) with a recent 5% weight loss
were randomized to 1 of 3, 10-month, weight loss maintenance interventions: (1)
electronically delivered maintenance strategies plus weekly professional e-coaching (Pro e-
coach), electronically delivered maintenance strategies plus weekly Peer e-coaching (Peer e-
coach), or maintenance strategies alone (No e-coach). At the start and end of the 10 month
maintenance program weight was objectively assessed and supportive accountability was
measured via self-report. Participants in Pro e-coach and Peer e-coach reported significantly
higher levels of supportive accountability (5.72.2, 6.82.2, ps.05) compared to No e-coach
(4.71.8) and did not differ from one another (p=.16). Weight regain showed a similar pattern.
Pro and Peer arms lost an additional 1.87.0kg and 0.56.4kg during the maintenance
intervention (p=.47) whereas No e-coach regained 3.55.7kg (ps<.05). Moreover, higher
levels of supportive accountability were associated with less weight regain during treatment
(r=-.28, p=.018). This study is the first to provide evidence for Supportive Accountability
Theory. In addition, results suggest that peer e-coaching may be an effective and perhaps
cost-effective method for providing support and accountability and promoting weight loss
maintenance.

CORRESPONDING AUTHOR: Rena Wing, Ph.D., Brown University; [email protected]

Symposium 23C

HELPING AND HINDERING: EXPLORING DIFFERENT TYPES OF SPOUSE/PARTNER SUPPORT IN

WEIGHT LOSS

Dr. Amy A. Gorin, Ph.D.

Background: Social support is often cited as a critical ingredient for successful weight
management; however, weight loss interventions designed to enhance support among
spouses and romantic partners have had mixed effects. Theory-based examinations that
pinpoint the specific types of support related to weight loss outcomes can guide the
development of more effective couples weight loss programs. This study explored
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S673

associations between spouse/partner support and weight loss outcomes through the lens of
Self Determination Theory (SDT). SDT posits that autonomy support (e.g., providing choices,
valuing anothers perspective) creates an environment conducive to health behavior change
whereas more controlling forms of support (e.g., reminding, criticizing) can hinder goal
progress.

Methods: Participants (N=130, MAGE=53.40, MBMI=33.93, 68.5% female) in a larger weight

loss trial were randomized to 6 months of Weight Watchers (WW) or a self-guided control
group (SG) and assessed at baseline, 3, and 6 months. Measures included the modified
Important Others Questionnaire (Autonomy Support) and the two subscales of the Sallis
Social Support for Eating Questionnaire (Encouragement of Healthy Eating: reminding about
healthy choices, commenting on relapses; Discouragement of Healthy Eating: getting angry
about slips, sabotaging). A series of regressions were tested for each predictor and percent
weight change. Interactions with intervention group were also examined.

Results:Consistent with SDT, if autonomy support increased from baseline to 3 months,

participants experienced a larger percent weight loss at 6 months, B=-.18, p=.01. A similar
pattern was found for encouragement of healthy eating, B=-.29, p=.01. Increases in
discouragement of healthy eating predicted less weight loss, B=.27, p=.03. This finding was
qualified by an interaction with condition, p=.02, which indicated that this effect was
significant only in WW.

Discussion: These results suggest that autonomy support from ones spouse/partner can
facilitate weight loss success. Interventions that foster autonomy support in couples while
decreasing negative remarks and sabotaging behaviors have the potential to bolster goal
progress. Rigorous designs are now needed to develop and test SDT-informed weight
management programs.

CORRESPONDING AUTHORS: Talea Cornelius, MA, University of Connecticut;

[email protected]; Erin Lenz, M.A., University of Connecticut; [email protected];
Alexis Wojtanowski, Weight Watchers, Inc.; [email protected]; Gary
Foster, Ph.D., Weight Watchers, Inc.; [email protected]

Symposium 23D

COHESION IN GROUP-BASED BEHAVIORAL WEIGHT LOSS INTERVENTIONS

Shannon S. Taylor, Ph.D.

Behavioral weight loss interventions are often delivered in a group format, and it is believed
that group cohesion enhances treatment adherence and, thereby, weight loss. However,
group social dynamics are rarely measured. The goals of these secondary analyses were to:
S674 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

1a) assess whether group cohesion measured in the 3rd group session of a behavioral weight
loss intervention predicts program attendance and weight loss outcomes, and if so, 1b)
explore whether attendance mediates the link between group cohesion and weight loss; 2)
assess whether cohesion late in the intervention is associated with weight loss; and 3)
characterize the association between change in group cohesion and weight loss over the
course of the intervention. 504 Veterans (48.6% Caucasian, 47% African-American; 79.2%
Male; M age = 58.9, SD = 10.1; M BMI = 35.9, SD = 5.6) initiated a 16-week, group-based
behavioral weight loss program involving in-person group visits every two weeks. Group
cohesion was measured during the 3rd and 8th group visits. In linear regression models, early
group cohesion was not related to group attendance (R2 = 0.008, F (1,323) = 2.54, p = .11) or
weight loss (R2 = 0.001, F (1,323) = 0.27, p = .60). Late group cohesion was also unrelated to
weight loss (R2 = 0.001, F (1,236) = 0.28, p = .60). Although there was a significant increase in
group cohesion during the intervention (M3rd group= 3.16, SD = 0.48; M8th group = 3.33; SD =
0.44; t (190) = 7.59, p < .001), change in group cohesion was not associated with weight loss
(R2 = 0.007, F (1,190) = 1.43, p = .23). These findings are consistent with the limited literature;
however, they are inconsistent with theoretical assertions and clinical observations of the
influence of group factors on outcomes. Thoughtful consideration to measurement of group
dynamics including timing, level of analysis, and psychometric quality of scales is advised for
future studies.

CORRESPONDING AUTHORS: Maren Olsen, Ph.D., Duke University Medical Center & Durham
VA; [email protected]; Janet Grubber, M.S.PH, Durham VA; [email protected];
Megan McVay, Ph.D., Duke University Medical Center; [email protected]; Jennifer
Gierisch, Ph.D. M.PH., Duke University Medical Center & Durham VA; [email protected];
William Yancy, M.D. MPH, Duke University Medical Center & Durham VA;
[email protected]; Corrine Voils, Ph.D., Duke University Medical Center & Durham VA;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S675

Symposium 24 1:30 PM-2:45 PM

INNOVATIVE APPROACHES TO IMPROVING ENGAGEMENT AND EFFECTIVENESS OF

INTERVENTIONS IN INTEGRATED PRIMARY CARE

Jennifer S. Funderburk, PhD1, Kyle Possemato, Ph.D.2, Jeffrey Goodie, Ph.D., ABPP3, Robyn L.
Shepardson, PhD1

1
VA Center for Integrated Healthcare, Syracuse, NY; 2Syracuse Veterans Affairs Medical
Center, Syracuse, NY; 3Uniformed Services University, Gaithersburg, MD

Due to the high rates of patients presenting with mental and behavioral health concerns in
primary care, primary care is increasingly viewed as an ideal setting for delivering behavioral
interventions. As such, the integration of behavioral health providers into primary care to
provide these interventions is becoming more common as primary care shifts towards team-
based service delivery models. However, the literature on the types of evidence-based
interventions that behavioral health providers can employ in the unique primary care setting
is still in its infancy. This symposium will present three research studies that have explored
ways to improve the engagement and effectiveness of brief interventions in integrated care in
an effort to promote future research on innovative approaches to intervention in primary
care. The first presentation will share results from a study that examined the impact of various
treatment factors, such as treatment duration, modality, and focus, on treatment preferences
among primary care patients. The second presentation will summarize the results from a pilot
RCT examining the impact of peer support specialists as an augment to a Web-based CBT
intervention for primary care Veterans with PTSD. The third presentation will share
acceptability data from a group medical visit for patients with hypertension and depressive
symptoms to illustrate an innovative approach to engaging patients in care. The discussant
will summarize common themes to directly inform and/or stimulate new clinical intervention
work to help continue the development of a strong evidence base for interventions that can
be implemented in this setting.

Symposium 24A

WHICH TREATMENT FACTORS PREDICT TREATMENT PREFERENCE FOR BEHAVIORAL

INTERVENTIONS: A CONJOINT ANALYSIS

Dr. Jennifer S. Funderburk, PhD

S676 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Patient treatment preferences may dramatically impact engagement in, and adherence to,
behavioral treatment. Thus, examining treatment preferences is a vital initial step toward
developing effective behavioral interventions suitable for primary care. Although previous
research has examined several treatment attributes separately and their impact on treatment
preference, no research has looked at these attributes simultaneously to understand how
they are evaluated as a whole by patients. This type of decision making is more characteristic
of real-life, where specific trade-offs are evaluated among multiple attributes to determine
the patients preference. There are several attributes characteristic of behavioral treatments
offered to primary care patients including: the setting (e.g., primary care, behavioral health
clinic), treatment duration (e.g., single, less than 6 sessions), treatment modality (e.g.,
individual, group), and treatment method (e.g., telephone, in-person). In addition, no research
to date has examined an important treatment attribute: focus of treatment (e.g., the
presenting problem the patient is interested in changing). This attribute is important as most
primary care patients present with multiple concerns, so the focus of treatment may be an
important attribute impacting patient preference. The objective of this study was to better
understand how these attributes were evaluated as a whole by patients and examine the use
of conjoint analysis to help provide a more refined picture of treatment preferences within
primary care. Using a random sample of primary care patients who were not engaged in
specialty mental healthcare, a total of 216 (76% male, 91% White) primary care patients
completed a mailed survey (28% response rate). A majority of participants when ranking
attributes independently, rated a preference for individual (86%) and in-person (80%)
appointments. For treatment focus, participants identified weight, pain, and sleep as their
most preferred treatment foci as compared to other foci (e.g., sadness, anxiety, alcohol,
tobacco use, etc). Of those who returned the survey, results of a conjoint analysis will be
presented using data from the 173 participants who completed rating their preferences for all
27 scenarios depicting treatment options that varied based on the five attributes to help
understand how patients made decisions when all attributes were evaluated together. We
will discuss the implications to future research and clinical practices within primary care.

CORRESPONDING AUTHORS: Possemato Kyle, PhD, VA Center for Integrated Healthcare;

[email protected]; Gregory Beehler, PhD, VA Center for Integrated Healthcare;
[email protected]; Wilfred Pigeon, PhD, Center for Excellence for Suicide Prevention;
[email protected]

Symposium 24B

COMBINING PEER SUPPORT AND WEB-BASED TREATMENT TO BOOST ENGAGEMENT IN

PRIMARY CARE- MENTAL HEALTH SERVICES

Dr. Kyle Possemato, Ph.D.

ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S677

Veterans who served in Iraq and Afghanistan have high rates of co-morbid PTSD symptoms
and hazardous alcohol use. Evidence-based treatments for these problems are available, but
multiple barriers lead to low utilization of these treatments. Innovative treatment
approaches are needed. Previous research supports that 1) primary care-mental health
integration services are associated with increased treatment engagement, 2) technology-
based interventions are well accepted by Veterans and are more effective when delivered
with professional support, and 3) veteran peer services are associated with improved
treatment engagement and overall functioning. We added Peer-Support to an existing Web-
based Cognitive Behavioral Treatment, called Thinking Forward, for PTSD and alcohol misuse
in response to our previous data that indicated that Veterans wanted interpersonal
interaction while receiving Thinking Forward and needed help engaging in the treatment more
fully. A pilot randomized controlled trial compared Self-Managed Thinking Forward to Peer-
Supported Thinking Forward. Thinking Forward consisted of 24 brief self-guided modules that
taught cognitive and behavioral skills to manage trauma-related thoughts and reduce alcohol
use. Peer Support services involved meeting with a Veteran peer support specialist as needed
over 12 weeks. Peer support specialists discussed the content of Moving Forward modules to
help participants process the material and apply it to their own life. Preliminary results
indicated that Peer-supported participants averaged a 12-point decrease on the PTSD
Checklist at a mid-treatment assessment compared to a 5-point decrease in Self-Managed
participants. Peer-support participants also completed 42% more Thinking Forward modules
than self-managed participants. Preliminary data indicates that among primary care patients
with PTSD and alcohol misuse, combining peer support services with web-based treatment
may lead to better treatment engagement and clinical outcomes than web-based treatment
alone. Final results will be presented in March.

CORRESPONDING AUTHORS: Emily Johnson, Ph.D., Syracuse Veterans Affairs Medical Center;
[email protected]; Kimblerly Barrie, LCSW, MPH, Syracuse Veterans Affairs Medical
Center; [email protected]; J. Bronte Emery, B.A., Syracuse Veterans Affairs Medical
Center; [email protected]; Stephen Maisto, Ph.D., Syracuse University; [email protected]

Symposium 24C

PATIENT AND PROVIDER ACCEPTABILITY OF GROUP MEDICAL VISITS FOR DEPRESSION AND
UNCONTROLLED HYPERTENSION

Dr. Robyn L. Shepardson, PhD

Within the team-based environment of integrated primary care, group medical visits offer an
innovative way for the primary care team to collaborate in providing both medical and
behavioral health interventions to take advantage of the interdisciplinary skillset. Including a
physical health condition (e.g., hypertension) as a focus of groups targeting a behavioral
health concern (e.g., depression) can reduce stigma to increase engagement and acceptability
S678 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

of treatment. This presentation summarizes feasibility and acceptability pilot data from an 8-
session group medical visit for Veterans (N=12, M age=64.1, SD=13.1) with uncontrolled
hypertension and depression. The integrated behavioral health provider led the visits (50-90
minutes) and implemented behavioral activation, but was joined at each visit by a member of
the primary care team (primary care provider, nurse, physical therapist, or clinical
pharmacist), who discussed medications and/or the role of physical activity in managing
hypertension. Participants reported high levels of satisfaction (M=27.1, SD=3.6) with the
group on the Client Satisfaction Questionnaire, and 66% attended all 8 visits. Qualitative
interviews revealed participants felt the group decreased stigma associated with seeking help,
provided camaraderie and accountability, and served as a source of new ideas to try to
improve their symptoms. Participants shared their likes and dislikes as well as suggestions for
how to engage reluctant patients in group medical visits. All providers reported overall
positive feelings about the group medical visit and their willingness to do it again. Providers
suggested minor modifications to improve flow within and across visits. We will discuss
implications for future research as well as how group medical visits may be used to help
increase engagement with behavioral health interventions in primary care.

CORRESPONDING AUTHORS: Jennifer Funderburk, PhD, VA Center for Integrated Healthcare;

[email protected]; Kyle Possemato, PhD, VA Center for Integrated Healthcare;
[email protected]; Kim Barrie, MSW, VA Center for Integrated Healthcare;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S679

Symposium 25 1:30 PM-2:45 PM

WOMEN AND CARDIOVASCULAR DISEASE: CURRENT APPROACHES AND FUTURE

OPPORTUNITIES FOR INTERDISCIPLINARY COLLABORATION

Valerie Hoover, PhD1, Elizabeth Jackson, MD MPH2, Elena Salmoirago-Blotcher, M.D., Ph.D.3,
Katharine S. Edwards, Ph.D.4, Linda Hudson, ScD ScM MSPH5

1
Stanford University, School of Medicine, Division of Cardiovascular Medicine, Stanford, CA;
2
University of Michigan, Ann Arbor, MI; 3The Miriam Hospital & Alpert Medical School of
Brown University, Providence, RI; 4Stanford University School of Medicine, Stanford, CA;
5
Tufts University School of Medicine, Boston, MA

Cardiovascular disease (CVD) is the leading cause of death globally and women bear a
significant proportion of the disease burden. While CVD mortality in the U.S. has decreased in
both men and women in the past couple decades, the rate of decline has been slower in
women than men. Research suggests the presence of important sex differences in
cardiovascular risk and biopsychosocial factors, symptom presentation, diagnosis and
treatment. Appropriate management of women with CVD requires an interdisciplinary
approach that incorporates an understanding of these factors. This symposium will provide a
multidisciplinary perspective on key sex differences in CVD risk factors and conditions, areas
that would particularly benefit from an interdisciplinary approach, and presentation of new
qualitative data from focus groups and interventions designed to identify and meet the
unique needs of this population. The first talk will provide an overview of the current
research on sex differences in CVD and areas for interdisciplinary collaboration. The second
talk will present data from a pilot study examining the acceptability and feasibility of a CBT-
based intervention to address the psychosocial needs of women with Spontaneous Coronary
Artery Dissection, a condition that occurs most often in women. The third and fourth talks
will present and discuss qualitative data from focus groups on, respectively, Tai Chi as an
alternative exercise option for female survivors of acute coronary events, and the
development of a culturally tailored heart disease prevention program for African American
women. Together, these talks will present qualitative data on tailored interventions for the
prevention and treatment of CVD in women; education on sex differences in biopsychosocial
factors related to the prevention, diagnosis and management of CVD; and discussion of
opportunities for interdisciplinary collaboration.

Symposium 25A
S680 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

WOMENS CARDIOVASCULAR HEALTH: THE UNIQUE NEEDS OF WOMEN

Dr. Elizabeth Jackson, MD MPH

Womens Cardiovascular Health: the Unique Needs of Women

Women comprise a significant percentage of patients living with cardiovascular disease.

Women experience cardiovascular risk factors and related events such as coronary artery
disease in different ways than man. More women than men experience angina and
myocardial infarction without significant coronary artery disease. More women than men
experience heart failure with preserved ejection fraction (HFpEF) which carries significant
disability and mortality risk. Women with atrial fibrillation can differ from men in terms of
symptom burden and risk for stroke. For cardiovascular disease conditions, the presentation,
diagnosis and management frequently differs for women compared to men. These differences
frequently increase the psychological burden of CVD among women. Adequate management
of CVD must include management of health behaviors and emotional health for both men and
women; however an improved understanding of the unique aspects of women living with CVD
is required in order to improve quality of care.

This lecture will outline the key cardiovascular risk factors and conditions in which important
gender-related differences exist. A discussion of the epidemiology of these risk factors and
conditions will be followed by a discussion of the symptom presentations and management
strategies unique to women. Finally, a summary of the related critical areas in which
collaboratory research is needed will be discussed. The goal of the lecture will be improve the
audiences understanding of clinical cardiovascular medicine specifically related to women
and to promote a discussion of multidisciplinary research in the area of cardiovascular health
for women.

Symposium 25B

TAI CHI AS AN EXERCISE OPTION FOR FEMALE SURVIVORS OF ACUTE CORONARY EVENTS:
QUALITATIVE FINDINGS FROM THE GENTLE REHAB STUDY

Elena Salmoirago-Blotcher, M.D., Ph.D.

Background. Cardiac rehabilitation (CR) reduces cardiovascular mortality in survivors of acute

coronary events (ACE), but referrals are low especially among women. Tai Chi a practice
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S681

based on light/moderate aerobic exercise accompanied by meditative components could be

an alternative exercise option after an ACE.

Methods. To garner patients perspectives regarding the usefulness of Tai Chi we conducted 3
focus groups among participants enrolled in the Gentle Rehab Study a pilot RCT study
assessing the feasibility and adequate dose of Tai Chi for ACE survivors who did not attend CR.
All participants in the parent study were invited to participate and provided informed consent.
A semi-structured agenda allowed for the discovery of emergent themes as well as barriers
and benefits to use of the tai chi program. LC led the focus groups with input from co-
facilitators ES and JK. Interviews were transcribed and analysis was directed primarily by a
grounded theory approach.

Results. Of the original cohort, 21/29 (72.4 %) agreed to participate(16M, 5F; age range 44-82
years). Overall, participants reported that Tai chi was enjoyable and easy to practice, while
such aspects are missing from traditional CR. Most subjects reported they would like to
continue with Tai chi and they appreciated the social support provided by the Tai Chi group.
Women uniquely appreciated the meditative components of Tai Chi, and reported reduced
stress and anxiety. Women reported that Tai chi allowed them to build strength in a safe and
comfortable manner. Women also mentioned that their balance improved, reducing their fear
of completing physically challenging activities of daily living. As well, women appreciated how
Tai chi exercises are easy to fit into everyday life. Direct quotes and measures of agreement
on specific themes will be presented.

Conclusions. Tai Chi offers benefits that might not be directly attainable from CR, with women
reporting benefits from the meditative aspect of the practice. The gentleness of the exercises
appeals to women in particular, who may not have experience or interest in more vigorous
exercise such as that offered in CR.

CORRESPONDING AUTHORS: Shira Dunsiger, PhD, The Miriam Hospital & Brown University;
[email protected]; Julie Krol, MA, The Miriam Hospital; [email protected]; Lisa Conboy,
PhD, Harvard Medical School; [email protected]

Symposium 25C

PSYCHOSOCIAL TREATMENT FOR PATIENTS WITH SPONTANEOUS CORONARY ARTERY

DISSECTION

Dr. Katharine S. Edwards, Ph.D.

Background: Symptoms of psychological distress are common following myocardial infarction

(MI) and depression is highest among young women (Mallik et al., 2006). Spontaneous
Coronary Artery Dissection (SCAD) is a rare vascular condition that occurs in patients without
S682 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

traditional cardiac risk factors, disproportionately affects young women (mean age 42, 80%
female), and can lead to MI or sudden death (Hayes, 2014; Giacoppo et al., 2014). There is no
known treatment for SCAD and recurrence rates are high. Preliminary studies suggest that
SCAD patients experience higher levels of psychological distress after MI than patients with
coronary artery disease (Liang et al., 2013; Silber et al., 2015; Starovoytov et al., 2016). The
aim of the current pilot study was to develop an acceptable and feasible CBT-based
intervention to target the unique psychosocial needs of SCAD patients. Method: Pre-post
questionnaire data were collected from a small sample of SCAD patients (n=8) who completed
an eight-week CBT-based psychosocial group intervention. Measures included the Seattle
Angina Questionnaire, SF-12, Multidimensional Health Locus of Control scale, STAI, PSS, ISI,
PHQ-9, and IES-R as well as between-session feedback forms and post-treatment acceptability
interviews. Patients met once per week for two hours for a structured intervention that
included mindfulness practice, education about the existing evidence on SCAD, group
discussion, training in CBT skills, acceptance and values work, and reviewing homework
assignments. Results: Qualitative, post-treatment data from this pilot study indicate that
women found the social support aspect of the group particularly helpful. Four women
reported clinically significant decreases in symptoms of depression, anxiety and trauma, and
sleep at 3-month follow up. Perceived stress did not change significantly from pre- to
posttreatment or follow-up. Conclusion: Results from this pilot study suggest that participants
found the group to be both acceptable and feasible and experienced reductions in
psychological distress. Additional findings, feedback from participants, and implications for
expanding psychosocial group offerings to female cardiac patients through the use of online
and mobile technology will be discussed.

CORRESPONDING AUTHORS: Kelsey Clemson, M.S., PGSP-Stanford Psy.D. Consortium;

[email protected]; Jennifer Tremmel, M.D., Stanford University School of Medicine;
[email protected]

Symposium 25D

CULTURAL ADAPTATION OF THE STRONGWOMEN HEALTHY HEARTS PROGRAM FOR

AFRICAN AMERICAN WOMEN

Linda Hudson, ScD ScM MSPH

Purpose: The StrongWomen Healthy Hearts Program is a 12-week community-based

program designed to improve heart health among women. Evaluation of national
dissemination revealed differential outcomes for African American women leading the
authors to explore a cultural adaptation. The purpose of this study was to conduct qualitative
research to understand deep structure factors and to incorporate these into a culturally
adapted curriculum.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S683

Methods: The African American Obesity Research Network (AACORN) paradigm served as a
basis for qualitative inquiry of deep structure factors. Five focus groups were conducted
with a total of 22 participants (3-7 participants per group). Women were eligible if they self-
identified as Black or African American, were over the age of 18, not currently engaged in
regular physical activity, and had a BMI 25. Recruitment was conducted via flyers,
announcements to the Boston NAACP membership and local churches, and word of mouth. A
$25 gift card was given to participants. Groups were audio-recorded and transcribed. An initial
codebook was developed based on the discussion guide and then refined through an iterative
and collaborative process with the study team. Data were coded using NVivo (QSR
International) and interpreted with the aim of uncovering the values, attitudes, and beliefs of
participants and to better understand their worldview, experiences, and cultural perspectives.

Results: Major themes included the perceived severity of CVD, the impact of stress, the role of
inequitable access to resources and structural barriers of racism influencing ability to make
positive decisions about health, the role of family and friends as both supports for and
barriers to engaging in health behaviors, mid-to-high confidence in being able to change
behavior, and a desire for more knowledge and resources.

Discussion: The adapted curriculum, Healthy Hearts for an Abundant Life, incorporates
themes related to the AACORN expanded knowledge domains. It includes stress-reducing
activities, tools for self-care, metaphorical language that is based in spirituality,
acknowledgement of the role of racism, and strategies and tools consistent with available
resources.

CORRESPONDING AUTHOR: Sata Folta, PhD, Friedman School of Nutrition Science and Policy
Tufts University; [email protected]
S684 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 26 1:30 PM-2:45 PM

WHAT DOES LIVING WELL WITH ILLNESS LOOK LIKE?: EXAMINATIONS IN CHRONIC AND LIFE-
THREATENING CONDITIONS

Kevin E. Vowles, PhD1, Donald McGeary, PhD2, Laurie E. Steffen, PhD3

1
University of New Mexico, Albuquerque, NM; 2University of Texas Health Science Center San
Antonio, San Antonio, TX; 3Wake Forest University School of Medicine, Winston-Salem, NC

Over the past few decades, increasing interest has focused upon a model of positive or
healthy adaptation to illness, as opposed to previous models which tended to prioritize the
elimination of an illness state or minimization of a certain set of symptoms. This interest has
grown out of multiple domains, including positive psychology, the patient activation
movement, therapeutic model developments focusing on constructs such as acceptance and
resilience, as well as a pragmatic clinical need to help individuals in successfully adapting to
chronic illness. Yet, in many areas, the characteristics of healthy adaptation to, or living well
with, chronic illness remain unclear. For example, when one experiences chronic symptoms
and when treatments have failed to reduce them, is living well even a possibility and, if so,
what are its characteristics? The data presented in this symposium will detail the qualities of
living well with illness in two significant healthcare conditions, chronic pain and late stage
cancer. Dr McGeary will describe cross-sectional findings in 160 active duty service members
with chronic pain, which found evidence of two participant clusters. The first reported high
levels of pain acceptance and low levels pain catastrophizing and the second reported the
opposite. Cluster membership was a strong and significant predictor of depression, PTSD,
suicide risk, and healthcare utilization. Dr Steffen will present a daily assessment study of
patients with late stage lung cancer, which examined how daily goal setting, including
identification and monitoring of progress, related to daily symptoms, treatment engagement,
and quality of life. Results, drawn from over 1000 days of data, indicated that those who set
goals and worked towards them had higher daily quality of life, less time in bed, better social
interactions, more engagement with important activities, and fewer negative side effects from
treatment. Dr Vowles will present cross-sectional and longitudinal data collected from
approximately 500 community-dwelling individuals with chronic pain, who were clustered
according to time spent engaging in struggles for pain control and engagement in valued
activities. Findings indicated that individuals who spent the least amount of time struggling for
pain control and who were engaged in an adequate level of valued activities were functioning
significantly better than those who were not. In combination, the presented data provide
guidance on characteristics of living well with chronic illness they include greater
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S685

acceptance, more frequent goal setting and monitoring of progress, less time struggling for
symptom control, and more time engaging in meaningful activity.

Symposium 26A

PROVIDING A REFERENT FOR ANALYSES OF CLINICALLY SIGNIFICANT CHANGE: WHAT DOES

LIVING WELL WITH CHRONIC PAIN LOOK LIKE?

Dr. Kevin E. Vowles, PhD

In 1984, Jacobson, Follette, & Revenstrof provided a two-part definition of clinically significant
change as it pertained to behavioral interventions. The first, reliable change, referred to an
assessment of whether amount of change on a specific measure, observed in a single
individual, exceeded change that could be accounted for by measurement error alone. Thus,
reliable change is readily calculated using indices of temporal stability and score dispersion.
The second, clinical significance, required that, by the end of treatment, an individuals score
had moved outside the range of the dysfunctional (Jacobson et al.s terminology)
population and within a "functional" population range. For example, fulfillment of this
criterion would occur when the post-treatment depression score in a previously depressed
individual was indistinguishable from a never-depressed individual's score. In chronic pain,
Jacobsen et al.'s first criterion has been successfully evaluated, but the second has not, as
there are no data appropriate for comparison. Pain-free individuals are not an adequate
comparison group for those with ongoing pain and there are no data available representing a
high functioning group of individuals with chronic pain. The present analyses provide initial
normative information for such a high functioning group, defined here as the presence of
chronic pain in the absence of significant functional disruption. Cross sectional and
longitudinal data were collected from approximately 500 community dwelling individuals with
chronic pain (data collection is ongoing; current N = 437), each of whom completed a battery
of assessments, including a measure of self-rated membership in a dysfunctional or
functional population with four response options, assessing the degree to which pain was
adversely impacting functioning (e.g., not at all, a little). This screening measure has
previously been used to categorize individuals in this way with group membership having
significant associations with current depression and disability. Ratings on this assessment
were used to categorize individuals into a functional, dysfunctional, or intermediate cluster
and between cluster differences were examined across a number of measures of adaptive
(e.g., pain acceptance, values, self-compassion) and maladaptive (e.g., disability, depression,
substance misuse) functioning. Significant between-group differences were indicated across
these same measures and were particularly pronounced for adaptive functioning. They were
also present in both cross-sectional and longitudinal analyses. These results provide a
reference point for the second criterion of clinically significant change and may provide a
benchmark for assessing it in individuals with chronic pain.
S686 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 26B

THE CONTRIBUTION OF ACCEPTANCE TO CHRONIC PAIN RESILIENCY: A SECONDARY ANALYSIS

OF THE ESCAPE STUDY

Dr. Donald McGeary, PhD

There is plentiful evidence showing the long-term negative outcomes associated with poor
pain coping and treatment failure (McGeary et al., 2006). Ameliorating variables widely
believed to be implicated in poor pain coping may not be enough to improve outcomes
(Nicholas et al., 2012), but studies of Acceptance and Commitment Therapy for pain
management are starting demonstrate that the promotion of healthy pain coping strategies
may prove effective (Vowles et al., 2014). The present study explored the potential of healthy
coping interventions in the uniquely complicated population of military service members and
veterans with chronic pain (cf. McGeary et al., 2011). Using a similar cluster analysis to that
published by Vowles et al. (2008), this study examined how service members differ in how
they cope with pain (based on measures of pain acceptance/catastrophizing) and how these
differences reflect on pain and psychosocial outcomes using a secondary analysis of the
Evaluation of Suicidality Cognitions and Pain Experience (ESCAPE) study (Kanzler et al., 2012).
The ESCAPE study included 160 active duty military members and veterans treated for chronic
pain through a large military treatment facility in the Southern U.S. who were asked to
complete a one-time comprehensive assessment of pain and psychosocial coping. Details can
be found in Kanzler and colleagues (2012). A two-step cluster analysis was used to identify
coping clusters using data from 99 ESCAPE participants who completed measures of both pain
acceptance (CPAQ) and pain catastrophizing (PCS). Cluster analysis revealed two distinct
clusters (labelled Living Well and Living Ill) with good cluster quality based on fit indices
and silhouettes. The Living Well [LW; n=75] cluster was characterized by high levels of chronic
pain acceptance (high CPAQ Willingness and Activity scores) and low levels of pain
catastrophizing (low PCS Helplessness, Magnification, and Rumination scores). The Living Ill
[LI; n=36] cluster was comprised of a diametric opposite score pattern. The two clusters were
compared on a number of pain and psychosocial variables. Participants in the LW cluster
reported significantly lower levels of depression, PTSD, suicide cognitions, and
burdensomeness to others (all at p<.001) compared to LI cluster. The LW cluster also reported
significantly lower levels of disability (p<.001), pain intensity (p<.001), and 40% fewer
healthcare visits in the past year than the LI cluster. Results of this secondary analysis showed
a similar pattern of healthy coping as that found by Vowles et al. (2008). These findings
suggest that military pain management programs targeting increased pain acceptance have
the potential to improve pain-related outcomes and to ameliorate common psychosocial
comorbidities of military pain like depression, PTSD, and suicidality.

CORRESPONDING AUTHORS: Kathryn Kanzler, PsyD, UT Health Science Center San Antonio;
[email protected]; Craig Bryan, PsyD, University of Utah; [email protected]; Cindy
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S687

McGeary, PhD, UT Health Science Center San Antonio; [email protected]; Robert

Villarreal, M.S., UT Health Science Center San Antonio; [email protected]; Willie Hale,
PhD, UT Health Science Center San Antonio; [email protected]

Symposium 26C

THE IMPACT OF HOPE ON DAILY QUALITY OF LIFE IN LUNG CANCER PATIENTS

Laurie E. Steffen, PhD

Goal-setting, problem-solving, and values-based action are common components of quality of

life interventions for cancer (Graves, 2003). These components are sometimes referred to as
hope, which has been associated with quality of life across multiple types of cancer. The
utility of hope has not, however, been tested longitudinally in lung cancer where quality of life
is a key treatment outcome given low survival rates and high symptom and treatment burden.
The goal of this study was to study relations between hope and quality of life (e.g., social
functioning, palliative wellbeing) in lung cancer patients undergoing treatment. Patients with
higher hope were expected to report better quality of life and daily levels of hope were
hypothesized to predict daily quality of life. Fifty lung cancer patients (58% female, 78% non-
small cell, 66% metastatic disease, average age = 68.66, SD = 8.78) completed a baseline
questionnaire and 21 daily assessments (M = 20.3 days, SD = 1.3; 1,042 days of data).
Multilevel modeling was used to analyze same-day and next-day relationships. Pre-specified
models controlled for disease (e.g., time since diagnosis, treatment type, daily disease
symptoms) and psychological (e.g., depression, daily affect) factors. Intra-class correlations
ranged from 0.64 to 0.75. Patients with higher levels of hope reported higher social and role
functioning (estimate = 3.37, SE = 0.90, 95% CI = 1.60, 5.14) and higher palliative wellbeing
(estimate = 0.88, SE = 0.24, 95% CI = 0.40, 1.35). On days where patients reported higher
hope, they reported higher social and role functioning (estimate = 2.36, SE = 0.70, 95% CI =
1.00, 3.73), better physical functioning (estimate = 2.27, SE = 0.81, 95% CI = 0.68, 3.87), and
higher palliative wellbeing (estimate = 0.91, SE = 0.13, 95% CI = 0.66, 1.16). Daily hope
appeared to attenuate the negative impact of treatment days on functioning (estimate = 3.33,
SE = 1.25, 95% CI = 0.88, 5.77). Daily hope did not, however, predict quality of life in any next-
day models. Daily hope was not predicted by cancer symptoms, which suggests that hope may
support quality of life regardless of symptom burden.

CORRESPONDING AUTHORS: Bruce Smith, PhD, University of New Mexico;

[email protected]; Kevin Vowles, PhD, University of New Mexico; [email protected]
S688 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 27 1:30 PM-2:45 PM

PEER SUPPORT TO PROMOTE HEALTH BEHAVIOR CHANGE AMONG VETERANS

Katherine D. Hoerster, PhD, MPH1, Karin Nelson, MD, MSHS2, Michele Heisler, MD MPA3, Amy
N. Cohen, PhD4, Edwin Fisher, PhD5

1
VA Puget Sound Healthcare, Seattle and University of Washington, Seattle, WA; 2VA Puget
Sound, Univesity of Washington, Seattle, WA; 3University of Michigan, Ann arbor, MI;
4
Greater Los Angeles VA Healthcare Center, Los Angeles, CA; 5The University of North Carolina
at Chapel Hill, Gillings School of Global Public Heealth, Chapel Hill, NC

Veterans often have multiple comorbid medical problems, including high rates of obesity and
diabetes, which can be effectively managed with activity and diet. It is therefore important for
Veteran-serving organizations to identify effective ways to improve Veterans lifestyle
behaviors. Peer support counselors--individuals with lived experience shared with a particular
target population--have been utilized effectively in healthcare delivery to address lifestyle
behaviors and help patients manage a host of medical conditions. As such, the Veterans
Health Administration (VHA) has begun to invest considerable resources in the delivery of
peer-delivered and peer-supported services. For example, VHAs Office of Mental Health
recently hired over 1000 peer support counselors who are in recovery from mental illness to
address well-being among Veterans with psychiatric conditions. VHA is also beginning to
expand peer support services to target conditions such as diabetes and obesity in primary
care. While research from the general population suggests peer support services may be
beneficial in the management of Veteran health and health behaviors, research is limited and
thus requires further investigation. Moreover, given that Veterans often have multiple
comorbidities, the application of peer support services among Veterans has relevance to
addressing lifestyle health behaviors among medically complex non-Veteran patient
populations.

This symposium presents a range of perspectives related to peer support services targeting
Veterans health behaviors with implications for Veteran and non-Veteran populations. KH
will present preliminary pilot data from a peer support-delivered behavioral weight
management program for Veterans with PTSD, and will discuss approaches to intervention
development such as iterative refinement and partnered research. KH also will highlight
insights into real world health promotion treatment complexities, given the medical-
psychiatric condition comorbidity being addressed. KN will discuss a novel community-based
peer counselor recruitment strategy for a large trial, highlighting issues related to peer
support counselor employment, training, and supervision. MH and AC will present data from
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S689

two large trials of technology- and peer-based behavioral diabetes and weight management
interventions among Veterans. Lastly, EF, leader of Peers for Progress, which promotes peer
support worldwide, will discuss how the papers exemplify important themes in disseminating
peer support services, and healthcare/policy implications for varied settings. This symposium
will highlight benefits and challenges of peer support care, and will yield insights regarding
ways to optimize patient-centered care to promote healthy lifestyle behaviors using peer-
based services.

Symposium 27A

DEVELOPING A PEER SUPPORT WEIGHT LOSS PROGRAM: A MODEL FOR ITERATIVE

REFINEMENT AND PARTNERED RESEARCH

Dr. Katherine D. Hoerster, PhD, MPH

PTSD is common in Veterans and associated with increased obesity risk, in part due to unique
activity and healthy diet barriers. VAs 16-week MOVE! weight management program is less
effective for Veterans with PTSD. We developed and piloted a 16-week peer-led weight
management program tailored for Veterans with PTSD called MOVE!+UP (Improving MOVE!
for Veterans with PTSD Using Peer Support). MOVE!+UP was delivered in 4 group sessions by
a paid peer support counselor who addressed PTSD-specific weight loss barriers and led
Veterans in community walks, followed by six biweekly phone sessions. Three cohorts
initiated participation (12/2015-4/2016; N=23). Baseline, 4-week, 16-week, and 6-month
quantitative and qualitative data were collected. We obtained input from numerous
stakeholders (e.g., peer support counselor, investigative team, clinical stakeholders, national
VA leadership). All MOVE!+UP components (e.g., facilitator and participant manuals; fidelity,
safety, training, and supervision procedures) were modified after each cohort to better
address weight, physical activity, and mediator targets. Veterans reported strong benefits
from the support received from other group members and the peer support counselor. Group
walks were cited as a particularly helpful aspect. Stakeholders made numerous suggestions
for changes, many of which were integrated during iterative refinement. Weight did not
significantly decrease from baseline to 4 weeks. However, diet quality, mental health quality
of life, and PTSD symptom severity improved (ps <.05). Several mediators improved
significantly (e.g., diet and activity change strategies). Most 4-week findings were sustained at
subsequent timepoints. Although not significant, average weekly physical activity minutes
increased from 122 to 186, p=.19. A modified MOVE!+UP will be tested in an uncontrolled
pilot and subsequent large-scale trial. This study provides data on a peer-led weight loss
intervention addressing mental and physical health of Veterans with PTSD, and a model for
iterative refinement and partnered research. This approach may facilitate readiness for future
implementation and dissemination.
S690 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHORS: Karin Nelson, MD, MSHS, VA Puget Sound Healthcare System,
Seattle Division; [email protected]; Lamont Tanksley, VA Puget Sound Healthcare System,
Seattle Division; [email protected]; Tracy Simpson, PhD, VA Puget Sound Healthcare
System, Seattle Division; [email protected]; Brian Saelens, PhD, University of
Washington; VA Puget Sound Healthcare System, Seattle Division; Gayle Reiber, PhD,
University of Washington; [email protected]; Jurgen Unutzer, University of Washington;
[email protected]

Symposium 27B

DEVELOPING A PEER SUPPORT PROGRAM TO TARGET HIGH RISK PATIENTS IN PRIMARY CARE

Karin Nelson, MD, MSHS

Cardiovascular disease (CVD) is the leading cause of mortality among veterans, and CVD risk
factors remain sub-optimally controlled. The goal of the Vet-Coach (Veteran peer Coaches
Optimizing and Advancing Cardiac Health) study is to test the effectiveness of a home-visit,
peer health coach intervention to promote health behavior change among veterans with
multiple CVD risk factors. Veteran peer coaches (Vet-Coaches) will provide social support,
health education, and link veterans to clinic and community resources.

The presentation will describe a novel strategy to recruit Vet-Coaches who live in targeted
high risk areas. Using geo-coded patient addresses, we generated spatial distributions of
hypertension rates by census tract within a 5-mile radius of our clinic and selected areas with
the highest rates of poorly controlled hypertension, which comprised 60 census tracts. Each
primary care clinic team was given a list of patients living in the target areas with a diagnosis
of hypertension. During a staff meeting, we asked providers to review these lists and
nominate veterans based on the following criteria: successful use of VA services, successful
management of hypertension or other chronic conditions, and experience utilizing community
resources. We sent the nominated patients (n=72) an introductory letter prior to conducting
follow-up calls to assess interest in becoming a Vet-Coach, which provides reimbursement
based on an hourly rate. Vet-Coach interviews are ongoing. To date, we have interviewed n=9
veterans. Results of the Vet-Coach recruitment will be discussed along with ideas and lessons
learned, with a specific focus on working in primary care clinics and developing a peer support
program that is integrated into primary care clinic work flow.

CORRESPONDING AUTHORS: Jennifer Williams, VA Puget Sound Healthcare System;

[email protected]; Marie Lutton, VA Puget Sound Healthcare System;
[email protected]; Christopher Vanderwarker, VA Puget Sound Healthcare System;
[email protected]; Julie Silverman, VA Puget Sound Healthcare System;
[email protected]; Kamala Jain, VA Puget Sound Healthcare System;
[email protected]; Tiffanie Fennell, VA Puget Sound Healthcare System;
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S691

[email protected]; Matthew Augustine, VA Puget Sound Healthcare System;

[email protected]; Kristen Gray, VA Puget Sound Healthcare System;
[email protected]

Symposium 27C

TECHNOLOGICALLY ENHANCED COACHING (TEC): A PEER SUPPORT MODEL TO IMPROVE

DIABETES OUTCOMES

Dr. Michele Heisler, MD MPA

This presentation will describe several approaches we have tested to enhance peer coach
interventions with interactive, e-Health tools. In several completed VA and non-VA RCTs we
have found that both mutual peer support and peer coaching models are more effective than
usual care, financial incentives, and nurse care management in improving glycemic control
and other clinical and patient-centered outcomes in high-risk adults with diabetes. While
peers can be trained in effective approaches to support other patients self-management
behaviors, such coaches necessarily lack the diabetes and diabetes medications content
expertise to help patients better share in treatment decisions and goal-setting with their
health care providers. Accordingly, in a prior study we developed and tested in an inner-city,
predominantly ethnic minority population a tailored, interactive diabetes and diabetes
medication information tool that outreach workers and peers could use to facilitate
discussions with patients. Such tools can enhance the sustainability and effectiveness of
coaching programs to better prepare adults to set self-management goals and action plans,
and to discuss treatment options with their providers. We then adapted this tool for the VA
and VA patients and are using mixed methods to evaluate over 6 and 12 months peer
coaching supported with this tool, peer coaching alone, and usual care among 312
predominantly African American Veterans with diabetes and poor glycemic control at the
Detroit VA. Recruitment has been completed and we will have final results in January, 2017 to
report. In this presentation, we will describe the e-Health tool, present results of in-depth
qualitative interviews with peer coaches and patients on experiences with the intervention,
and present the final results of this comparative effectiveness pragmatic clinical trial.

Symposium 27D

ONLINE WEIGHT MANAGEMENT WITH PEER SUPPORTS FOR PEOPLE WITH SERIOUS MENTAL
ILLNESS

Dr. Amy N. Cohen, PhD

S692 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

People with serious mental illness are at high risk for obesity, related cardiovascular illness,
and premature death. In-person weight management programs result in lower weight, and
are found in treatment guidelines. However, these programs are often not provided, and
patient enrollment and retention is low. Programs require substantial clinician time and
frequent clinic visits. Barriers could be addressed using computerized provision of diet and
exercise services, combined with motivation and support from peer coaches. 276 overweight
patients were recruited from a Veterans Affairs mental health clinic, and assigned to 1) online
weight management with peer coaching, 2) in-person clinician-led services, or 3) treatment as
usual. At 6 months, patient outcomes were assessed and interviews conducted. There was a
significant group by time interaction (F=4.02, p=.02). In the online plus peer coaching group,
the model estimated an average change in body mass index (BMI) from baseline to six months
of 34.9.4 to 34.1.4 . This change corresponds to a 2.8 kg (6.2 lbs) weight loss (t=3.2,
p=.001). No significant change in BMI was seen in either the in-person (t=.10, p=.92), or
treatment as usual (t=-.25, p=.80) group. There was a significant difference in how many
participants completed the intervention with 31% of the online group but none of the in-
person group completing the intervention (2=24.2; p<.0001). Those in the online group
received, on average, 7.9 (SD=6.8) peer coaching calls.

Addressing physical health in mental health clinics has been a challenge. Online weight
management with peer supports is feasible, and well received. An online program with peer
support produces weight loss, and may have greater effectiveness than clinician-led services.
With this program, there is little burden on clinicians, costs are low, and service provision is
patient-centered.

CORRESPONDING AUTHORS: Alexander Young, Greater Los Angeles VA Healthcare Center;

[email protected]; Richard Goldberg, VA Capitol Healthcare Network Mental Illness Research,
Education, and Clinical Center (MIRECC); [email protected]; Julie Kreyenbuhl, VA
Capitol Healthcare Network Mental Illness Research, Education, and Clinical Center (MIRECC);
[email protected]; Noosha Niv, VA Desert Pacific Mental Illness Research, Education,
and Clinical Center (MIRECC); [email protected]; Nancy Nowlin-Finch, Edelman Clinic--
Los Angeles County Department of Mental Health; [email protected]; Gerhard
Hellemann, UCLA Semel Institute Biostatistics Core; [email protected]; Fiona
Whelan, UCLA Semel Institute Biostatistics Core; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S693

Symposium 28 1:30 PM-2:45 PM

LESSONS LEARNED FROM THE DEVELOPMENT AND DISSEMINATION OF THREE HEALTH POLICY
BRIEFS

Joanna Buscemi, Ph.D.1, Caryn E. Peterson, PhD, MS2, Karriem S. Watson, DHSc, MS, MPH3, E.
Amy Janke, PhD4

1
DePaul University, Chicago, IL; 2University of Illinois at Chicago, Chicago, IL; 3University of
Illinois at Chicago (UIC), Chicago, IL; 4University of the Sciences, Philadelphia, PA

The Society of Behavioral Medicine's (SBM's) Health Policy Committee (HPC) provides
structured and guided opportunities for SBM members and SBM special interest groups (SIGs)
to increase the impact of their work through the development and dissemination of health
policy briefs. The purpose of this symposium is to 1) present three of the health policy briefs
that have been developed in the past year and 2) provide members with lessons learned
and tips for writing and disseminating the briefs. First, Dr. Karriem Watson will discuss a
health policy brief highlighting the importance of supporting lung cancer screening among
high-risk populations. Dr. Watson will discuss the process of engaging academic and like-
minded organizations and partners in the development and dissemination of the brief. Next,
Dr. Caryn Peterson will describe a brief in support of efforts to reduce barriers to Human
Papillomavirus (HPV) vaccination services in order to increase vaccine coverage. Dr. Peterson
will discuss how to develop recommendations to address the most pressing areas in need of
change and how to describe the epidemiologic, clinical, and public health features of the
problem in a manner appropriate to a policy brief. Third, Dr. Amy Janke will present a brief on
improving psychological care for pain to reduce risk for opioid misuse. Dr. Janke will discuss
how to collaborate across disciples to reach consensus on the framing of the brief and the
recommendations, and the process of using an initial brief as a springboard for future related
briefs, papers, presentations and outreach. The discussant and Chair of the Health Policy
Committee, Dr. Joanna Buscemi, will close with remarks on key strategies for developing
authorship teams, engaging organizational partners, and increasing the reach of briefs
through broad dissemination.

Symposium 28A

THE DEVELOPMENT OF SBMS POLICY BRIEF: INCREASING HPV VACCINATION UPTAKE: AN

URGENT OPPORTUNITY FOR CANCER PREVENTION
S694 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Dr. Caryn E. Peterson, PhD, MS

HPV infection is common in the U.S. and globally, and affects both males and females through
its association with oral cancer, anal cancer, genital warts, and cervical cancer. HPV
vaccination is a safe and effective primary prevention strategy to reduce the infections that
can lead to these cancers. However, vaccination coverage remains low and subject to barriers
at multiple levels.

SBM supports the goals outlined by Healthy People 2020, the Presidents Cancer Panel, and
the National Vaccine Advisory Committee to increase vaccination coverage among both males
and females. The purpose of this brief was to present a set of recommendations for
legislators, policymakers, healthcare providers, and health systems in order to address critical
barriers to HPV vaccination.

Recommendations for legislators and policymakers focus on ensuring adequate resources to

achieve the nations HPV vaccination goals. Recommendations for healthcare providers and
systems center on strengthening the quality and consistency of HPV vaccination
recommendations for all eligible patients.

This presentation will first describe the critical need for increased HPV vaccination among
males and females, outline barriers to vaccine uptake, and present SBM recommendations.
Then it will describe the process of developing the brief. Key aspects of this process include:
gathering a multidisciplinary team of co-authors in order to describe the issue in its most
current form; developing recommendations to address the most pressing areas in need of
change; and describing the epidemiologic, clinical, and public health features of the problem
in a manner appropriate to a policy brief.

Symposium 28B

ENCOURAGING LUNG CANCER SCREENING IN HIGH RISK POPULATIONS: LESSONS LEARNED

FROM THE DEVELOPMENT OF A HEALTH POLICY BRIEF

Dr. Karriem S. Watson, DHSc, MS, MPH

Informing key stakeholders of policy implications that impact health outcomes is vital to
advocacy, research and interventions aimed to mitigate health inequities. Collaborations with
multi-stakeholders including community, academic and advocacy partners to develop policy
briefs is an essential tool to support dissemination of policy recommendations. Several
academic stakeholders along with policy and advocacy partners aligned with the Society of
Behavioral Medicine (SBM) to address United States Preventive Services Task Force (USPSTF)
recommendation of low-dose computed tomography (LDCT) screening of the chest for eligible
populations to reduce lung cancer mortality. This group assembled to support efforts that
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S695

further advances the translation of research findings into practical applications for health care
providers and healthcare systems. Lung cancer mortality exceeds that of breast, colorectal
and prostate cancer mortality combined. The assembly of academic, advocacy and policy
stakeholders resulted in policy recommendations for both healthcare providers and
healthcare systems as well as policy makers. The primary lessons learned from this
collaborative development of a policy brief included: 1) the importance of identifying
academic partners who value a multi-disciplinary approach to examining healthcare policy
implications; 2) the benefit of including the expertise of recognized thought leaders as part of
the collaboration; 3) the advantage of leveraging existing partnerships to identify common
alignments such as multi-institutional grant projects; and 4) the incorporation of principles of
shared governance to ensure that each stakeholder has an active voice in the process.

Symposium 28C

IMPROVING ACCESS TO PSYCHOLOGICAL CARE FOR INDIVIDUALS WITH PERSISTENT PAIN

E. Amy Janke, PhD

Objective: Present SBMs first policy brief addressing psychological care for chronic pain and
provide an overview of lessons learned from writing this brief.

Background: Growing concerns regarding prescription opioid misuse, abuse and opioid-
related fatalities have initiated a re-evaluation of the long-term efficacy and potential risks of
this class of medications in the management of pain, and are part of a larger dialogue on the
way pain is managed in U.S. Policy makers have articulated a need for clear, evidence-based
guidance to help inform pain policy. Persistent pain is common, expensive, and debilitating,
and requires comprehensive assessment and treatment planning. Recently released opioid
prescribing guidelines by the CDC (2015) emphasize the importance of using non-opioid
therapies before considering opioid treatment for those without a malignant illness. The
favorable risk-benefit ratio of psychological approaches suggests these be considered early
on, prior to other treatments that carry more risks. Unfortunately, despite persuasive
evidence supporting the efficacy of many non-opioid therapies, access to these interventions
is severely limited. Access to psychological care including pain assessment, intervention, and
prevention approaches needs to be available in all clinical contexts, for all populations, at all
points during the entire trajectory of the illness. To achieve this, we must prioritize
reimbursement of evidence-based psychological approaches for pain management and
improve provider training and competencies to implement these approaches

Approach: This presentation will provide an overview of the pain policy brief and discuss how
the core team of authors was cultivated from the SBM membership. The authorship team
identified many needs to address with regards to pain policy, and the presentation will review
the process of conceptualizing ideas and narrowing the briefs focus to develop a specific,
S696 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

targeted message. The process of writing the brief and engaging stakeholdersfrom within
SBM and other organizationswill be discussed as will ensuring the briefs synergy within the
larger landscape of emerging pain policy. Opportunities identified by the authorship team to
use the brief as a foundation for ongoing activities in the area of pain policy (outreach, briefs,
presentations, papers) will be discussed.

CORRESPONDING AUTHORS: Martin Cheatle, PhD, Center for Studies of Addiction Department
of Psychiatry Perelman School of Medicine University of Pennsylvania;
[email protected]; Lara Dhingra, PhD, Department of Family and Social Medicine
Albert Einstein College of Medicine; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S697

Symposium 29 1:30 PM-2:45 PM

STIGMA AS A PATHWAY TO OBESITY AND NEGATIVE HEALTH OUTCOMES IN UNDERSERVED

POPULATIONS

Emily Panza, MS1, Lisa Shank, M.S.2, Janet Tomiyama, Ph.D. 3, Mary S. Himmelstein, PhD4,
Kristina M. Decker, M.A.5

1
Rutgers, the State University of New Jersey, Highland Park, NJ; 2Uniformed Services
University of the Health Sciences, Bethesda, MD; 3University of California, Los Angeles, Los
Angeles, CA; 4University of Connecticut, Hartford, CT; 5The University of Memphis, Memphis,
TN

Obesity is associated with serious negative health consequences, ranging from cardiovascular
disease to early mortality, and is disproportionately prevalent among gender, racial, ethnic,
and sexual minority groups. Given the wide prevalence and multi-determined nature of
obesity, it is necessary to examine how multiple aspects of identity (e.g. race, sexual
orientation) intersect with social contextual factors, emotional experiences, and behaviors to
heighten risk for obesity.

The research presented here examines how one social contextual factor, societal
stigmatization, cuts across multiple aspects of identity and serves as a pathway to the
development and exacerbation of obesity in a variety of underserved groups. This is
highlighted in the first symposium presentation, which identifies weight stigma as a risk factor
for the development of physical health problems and medical conditions in a group of
overweight active duty military, a specialized population with a cultural emphasis on fitness.
The second presentation builds on these findings by examining how multiple sources of
stigma contribute to the weight disparity among a triply oppressed group, overweight and
obese sexual minority women. This study uses EMA methods to examine whether lifetime and
daily exposure to stigma based on sexual orientation, gender and weight increases risk for
overeating and binge eating. These findings are extended in the next presentation, which
examines strategies for coping with weight stigma across racial and gender minority groups.
Using population-based data, this presentation reveals how race and gender interact to
promote either effective or harmful coping, displaying how strategies for coping with stigma
impact obesity. The final presentation will explore the relation between stigma and health
outcomes in emerging adults, demonstrating that greater stigma and higher BMI are
associated with more exercise avoidance, depression, and poorer physical health in young
adults with low self-reported resilience.
S698 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

The symposium discussant, an expert on stress, weight stigma and obesity, will tie together
these diverse presentations, which employ a variety of methods to explore obesity-related
outcomes in multiple underserved groups, collectively highlighting the relevance of social
stigmatization as a pathway to the development and worsening of obesity in marginalized
populations. This has important clinical implications, suggesting that efforts to reduce stigma
and to teach effective strategies for coping with stigma may enhance the effectiveness of
obesity interventions and improve health in all populations.

Symposium 29A

MINORITY STRESS AND EATING BEHAVIOR AMONG OVERWEIGHT AND OBESE SEXUAL
MINORITY WOMEN

Emily Panza, MS

Background. Sexual minority women are nearly three times more likely to be overweight or
obese than their heterosexual counterparts, but little research has investigated potential
mechanisms that underlie this disparity. One such mechanism may be minority stress. Sexual
minority women are members of multiple socially stigmatized groups: being non-
heterosexual, being female, and for 60% of sexual minority women, being overweight. As a
result, they are chronically exposed to minority stressors like social stigmatization (i.e. being
treated unfairly or differently). In an effort to reduce distress, sexual minority women may use
overeating and binge eating to cope, behaviors that, over time, promote weight gain and
confer risk for obesity.

Methods. The current study used cross-sectional and Ecological Momentary Assessment
(EMA) methods to examine whether lifetime and acute experiences of stigmatization heighten
risk for overeating and binge eating in overweight and obese sexual minority women. 30
overweight and obese (BMI>25) sexual minority (e.g. lesbian, bisexual, queer) women were
recruited from the local community. Participants completed baseline assessments of eating
behavior and minority stress related to sexual orientation, weight and gender, and for five
days, used a smartphone application to report instances of perceived stigmatization,
overeating, binge eating, stress, and negative emotion five times daily.

Results. Using hierarchical linear modeling,we expect to find significant links between
stigmatization, overeating and binge eating behaviors, such that greater lifetime
stigmatization, as well as daily acute stigmatization, will be associated with greater rates of
overeating and binge eating behaviors among overweight and obese sexual minority women.
We also expect that daily episodes of stigmatization will increase risk for concurrent and
future elevations in stress, negative emotion, overeating and binge eating.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S699

Conclusion. This is the first study to examine the intersecting effects of sexual orientation,
gender, and weight-based minority stress on eating behavior among overweight and obese
sexual minority women. Given the lack of research investigating the origins of the weight
disparity among sexual minority women, findings will facilitate obesity prevention efforts and
improve our ability to develop effective and culturally tailored obesity interventions for this
multiply marginalized group.

CORRESPONDING AUTHORS: Kara Fehling, M.S., Rutgers, the State University of New Jersey;
[email protected]; Yasmine Omar, M.S., Rutgers, the State University of New Jersey;
[email protected]; Maribel Plasencia, B.A., Rutgers, the State University of New
Jersey; [email protected]; Allison Borges, B.A., Rutgers, the State University of
New Jersey; [email protected]; Samira Dodson, B.A., Rutgers, the State University of
New Jersey; [email protected]; Edward Selby, Ph.D., Rutgers, the State University of
New Jersey; [email protected]

Symposium 29B

THE RELATIONSHIP BETWEEN WEIGHT STIGMA AND PHYSICAL HEALTH IN MILITARY

PERSONNEL WITH OR AT HIGH-RISK OF OVERWEIGHT/OBESITY

Ms. Lisa Shank, M.S.

Perceived weight stigma and discrimination are associated with adverse health indices, such
as elevated cortisol and inflammation, lipid/glucose dysregulation, and poorer self-rated
health. The relationship between weight stigma and physical health may be particularly
relevant for military personnel, given the increasing prevalence of overweight/obesity
(OW/OB), a cultural emphasis on fitness and weight/shape, and a high rate of reported weight
stigma. Therefore, we investigated the relationship between weight stigma and physical
health in 117 active duty service personnel (66.7% male; 35.0% Non-White; age: 30.87.4y;
BMI: 29.52.5 kg/m2) prior to enrollment in an obesity prevention program. All had OW/OB
and/or a family history of OW/OB (family history only: n = 7). Participants completed an
adapted Experiences and Sources of Weight Stigma Questionnaire which produced a total
score (WS) and a military-specific subscale (WS-M), and the Weight Bias Internalization Scale-
Modified (WBIS). Participants reported the presence of medical conditions (e.g., type 2
diabetes, hypertension, asthma) and responses were dichotomized to indicate presence ( 1;
n = 57) or absence (n = 60). After adjusting for age (y), sex, height (cm), and body fat (%), WS
and WS-M were significantly associated with the presence of a medical condition (ORs = 2.63,
2.76, respectively, ps = .04). WBIS did not moderate the relationship between WS or WS-M
and physical health (ps > .05). Similar patterns were observed when BMI was used as a
covariate instead of height and body fat. Findings suggest that weight stigma may contribute
to the development or maintenance of medical conditions, perhaps due to associated
biochemical stress. Alternatively, active duty personnel with a medical condition might
S700 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

experience more frequent weight stigma; for example, when visiting healthcare providers.
Longitudinal studies should prospectively examine the relationship between weight stigma
and physical health among service members, who may be a particularly high-risk subgroup.

CORRESPONDING AUTHORS: Natasha Schvey, Ph.D., Uniformed Services University of the

Health Sciences; [email protected]; Kendra Oldham, M.S., Uniformed Services
University of the Health Sciences; [email protected]; Marissa Barmine, B.S.,
Uniformed Services University of the Health Sciences; [email protected]; Dawn
Bates, Madigan Army Medical Center; [email protected]; Douglas Maurer, DO,
MPH, FAAFP, Madigan Army Medical Center; [email protected]; Mark
Stephens, M.D., Uniformed Services University of the Health Sciences;
[email protected]; Marian Tanofsky-Kraff, Ph.D., Uniformed Services University of
the Health Sciences; [email protected]; Tracy Sbrocco, Ph.D., Uniformed
Services University of the Health Sciences; [email protected]

Symposium 29C

MEET ME AT THE CROSSROADS: WEIGHT-STIGMA AT THE INTERSECTIONS OF RACE AND

GENDER

Dr. Mary S. Himmelstein, PhD

A large literature demonstrates the harmful psychological, behavioral, and physical health
outcomes associated with weight-stigma, but few studies have systematically examined
intersectionality in weight-stigma. Studies have disproportionately focused on White women,
with little attention given to weight-stigma in Asian, Black, or Hispanic women or stigma in
men. Examining weight-stigma in non-White women and men is essential because the
prevalence of obesity varies along racial lines with greater prevalence of obesity in Black and
Hispanic populations. This study examined intersectionality (race, gender) using a diverse
national sample of 2,378 individuals matching demographics from the US census. Participants
identifying as Asian, Black, Hispanic, and White responded to questions about their
experiences with weight-stigma and strategies employed to cope with weight-stigma. Women
reported experiencing more weight-stigma than men, but no differences in stigma emerged as
a function of race. In coping with weight-stigma, Black women were less likely to cope with
stigma using negative affect and disordered eating strategies, while Hispanic women were
more likely to cope with weight-stigma via disordered eating strategies. Black men were more
likely to cope with stigma via increased eating. No coping differences emerged for Asian
women or Hispanic men. These results suggest potential protective factors against poor
health outcomes associated with weight-stigma for Black women, but potentially exacerbating
factors for Hispanic women and Black men. This study represents the first systematic
exploration of weight-stigma and coping in non-White men and women. Given we found no
differences in experienced weight-stigma as a function of race, it is imperative that more
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S701

research be focused on weight-stigma in underserved populations in order to understand the

unique stigma processes these populations face.

CORRESPONDING AUTHORS: Rebecca Puhl, PhD, Uconn Rudd Center for Food Policy and
Obesity; [email protected]; Diane Quinn, PhD, University of Connecticut;
[email protected]

Symposium 29D

PROFILES OF WEIGHT STIGMA AND MENTAL HEALTH: ASSOCIATION WITH HEALTH

OUTCOMES AMONG EMERGING ADULTS WITH OVERWEIGHT/OBESITY

Kristina M. Decker, M.A.

Objective. Rates of obesity continue to soar among emerging adults. Weight-based stigma
and poor mental health are associated with negative physical health outcomes among young
people with obesity. We aimed to examine how varying patterns of risk (stigma, mental
health, stress) and protective (resilience) factors relate to health outcomes.

Method. 196 emerging adults (61% White, 16% Black, 12% Hispanic; 60% female) enrolled in
our study via Amazon Turk and a university subject pool in the U.S. MidSouth. These 18-25-
year-olds (Mage=22.38, SD=2.01) completed measures of Weight Self-Stigma, Stigmatizing
Situations, Depression, Anxiety, Perceived Stress, Resilience, and Exercise Avoidance. Body
Mass Index (BMI) was calculated from self-reported height and weight. Participants reported
perceived health status (How would you rate your overall physical health?) from poor to
excellent. In Mplus 7.4, Latent Profile Analysis (LPA) was used to generate profiles of stigma,
mental health, stress, and resilience and examine how the resulting profiles were associated
with exercise avoidance, BMI, and perceived health.

Results. Based on standard fit statistics, the following 3 classes emerged: 1) low stigma, low
risk, high resilience (LLH; N=58); 2) high stigma, high risk, low resilience (HHL; N=50); 3)
moderate stigma, moderate risk, moderate resilience (MMM; N=88). Exercise avoidance
significantly differed across classes with LLH class members reporting the least exercise
avoidance (M=11.31, SE=1.07) and HHL members reporting the most avoidance (M=32.54,
SE=1.89). Individuals in HHL (M=31.43, SE=.78) and MMM (M=31.61, SE= .86) reported
significantly higher BMIs than those in LLH (M=28.10, SE=.66). Individuals in LLH reported
significantly better physical health (M=3.60, SE=.14) than those in HHL (M=2.65, SE=.14) and
MMM (M=2.91, SE=.11).

Conclusions. Person-centered findings reveal significant health disparities among young

people with overweight/obesity with those most at risk (i.e., have higher BMI, poorer physical
health, greater exercise avoidance) reporting higher depression, anxiety, stress and lower
S702 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

resilience (i.e., ability to bounce back from distress). Such health disparities should be
prioritized in obesity intervention development.

CORRESPONDING AUTHORS: Idia Thurston, PhD, The University of Memphis;

[email protected]; Robin Hardin, M.A., The University of Memphis;
[email protected]; Mollie Anderson, PhD, The University of Memphis;
[email protected]; Courtney Maclin-Akinyemi, M.S., The University of Memphis;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S703

Symposium 30 1:30 PM-2:45 PM

FERTILITY, PREGNANCY, AND BIRTH IN THE AGE OF REPRODUCTIVE TECHNOLOGIES:

CHALLENGES AND OPPORTUNITIES FOR BEHAVIORAL MEDICINE

Marci Lobel, Ph.D.1, Lauri Pasch, PhD2, Lisa Rubin, Ph.D. 3, Jennifer Nicoloro SantaBarbara,
MSW, MA4, Marci Lobel, Ph.D.1

1
Stony Brook University, Stony Brook, NY; 2University of California San Francisco, San
Francisco, CA; 3New School for Social Research, New York, NY; 4Stony Brook University, Long
Island CIty, NY

The extraordinary evolution of reproductive technologies in recent years has included

methods to help infertile couples conceive and bear children, and to reduce the likelihood
that offspring will suffer from heritable diseases for which their parents are at-risk. Although
such technologies offer great promise to improve human health and well-being, they also
pose potent moral, emotional, behavioral, and tangible challenges for people considering
their use. The increased availability of reproductive technologies and their impact on
individuals, families, and society make it imperative to understand the psychosocial milieu of
these technologies. For example, what are the emotional consequences of using assisted
reproductive technologies such as in-vitro fertilization and pre-implantation genetic diagnosis
(PGD)? Conversely, do emotional factors including decisional conflict or distress affect the
success of these technologies? What mental health services might be needed by individuals
using assisted reproductive technologies? This symposium will feature cutting-edge
behavioral medicine research on assisted reproductive technologies to further scientific
investigation and to inform the development of appropriate clinical services. The first
presenter will describe results from a large longitudinal study of individuals undergoing
fertility treatment, focusing on identification of clinically-significant distress and use of mental
health services. The second presenter will discuss findings from in-depth interviews focused
on decisional distress surrounding PGD in individuals of reproductive age who carry a
heritable genetic marker for breast cancer. The third presenter will describe a meta-analytic
review of studies to evaluate the association of emotional distress with outcomes of assisted
reproductive technologies. The Discussant will highlight points of agreement and
disagreement from the findings presented, offer a conceptual overview to integrate the three
presentations and ground them in research and theory on stress, and highlight how this work
can be used to advance research and treatment for people using assisted reproductive
technologies.

Symposium 30A
S704 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

PSYCHOLOGICAL DISTRESS AND ASSISTED REPRODUCTIVE TECHNOLOGY USE

Lauri Pasch, PhD

Although there is no doubt that the use of fertility treatments offer the opportunity to
conceive for many couples who would not otherwise have been able to, they are clearly a
mixed blessing. Despite significant advances in success rates, treatment cycles still fail more
often than they succeed and many couples endure multiple failures. We present data from a
large prospective cohort study of couples seeking fertility treatment funded by the NIH to
address three main research goals. First, we describe the psychological consequences of
treatment failure. Second, we investigate risk factors for greater distress. Third, we examine
the use of mental health services for fertility patients and the extent to which they target
those at greatest psychological risk. From five fertility practices in Northern California, we
recruited 352 women and 274 men at the start of treatment for infertility and followed them
over the subsequent 18 months. Depression, anxiety, fertility treatment outcome, and
mental health service use were assessed over the course of the research. Major results
include: 1) rates of psychological distress following fertility treatment failure are extremely
high, 2) psychological distress does not predict the outcome of fertility treatment, but the
outcome of fertility treatment predicts subsequent increases in psychological distress, 3) the
most important predictor of whether someone will experience a major depressive episode
during fertility treatment is not how distressed they are at the beginning, but instead their
lifetime history of major depression, 4) most patients report that they are not offered and do
not use mental health services during fertility treatment, and high risk patients are not
targeted for mental health service referrals. We conclude that clinically significant
psychological distress is very common during fertility treatment and after failure, and that
those are particular risk can be readily identified. However, most patients and partners do
not receive and are not referred for help, even those at high risk for serious psychological
distress. Strategies for addressing these findings will be discussed.

CORRESPONDING AUTHORS: Sarah Holley, Ph.D., San Francisco State University;

[email protected]; Patricia Katz, Ph.D., University of California, San Francisco;
[email protected]; Nancy Adler, Ph.D., University of California, San Francisco;
[email protected]

Symposium 30B

IS IT WORTH IT?: DECISIONAL CONFLICT IN CONSIDERING PRE-IMPLANTATION GENETIC

DIAGNOSIS FOR HEREDITARY CANCER RISK

Dr. Lisa Rubin, Ph.D.

ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S705

Background: BRCA1 and BRCA2 gene variants are associated with increased susceptibility of
breast and ovarian cancer ranging from 20-76%, contingent on other genetic and
environmental factors. Further, individuals with BRCA1/2 gene variants face a 50% chance
that their gene variants will be transmitted to their genetic offspring. To address this risk of
transmission, reproductive options, such as preimplantation genetic diagnosis (PGD), a
technology in which in vitro fertilized (IVF) embryos are tested before implantation (with only
unaffected embryos transferred to the uterus), are increasingly being offered to BRCA1/2
gene variant carriers. However, for many individuals and couples, deciding whether to use
PGD and, if unsuccessful, when to stop, are fraught decisions, where costs and benefits are
uncertain and difficult to weigh.

Methods: Qualitative interviews were conducted with 39 reproductive-age BRCA1/2 gene

variant carriers about their views of PGD. Participants were predominantly female-identified
(89%) and partnered (76%); 42% had children, with 21% reporting that they had completed
their plans for childbearing. Data were analyzed using inductive thematic analysis.

Results: Among the 15 (38%) participants who had considered PGD, decisional conflict was a
key theme. This report emphasizes the types of conflicts reported by participants, and the
contexts in which these conflicts occur. Conflicts emerged in the contexts of participants (1)
reproductive goals, experiences, and pressures; (2) BRCA-related risk perceptions; (3)
perceptions of assisted reproductive technologies (ARTs); (4) financial resources; and (5)
personal cancer risk management strategies. Particularly for those choosing PGD, conflicts
were iterative and dynamic, as participants faced unexpected difficulties and ethical
dilemmas.

Conclusions: The rapid expansion of ARTs, including the use of PGD for inherited disease
susceptibility markers, threatens to outpace clinical and scientific understanding of the
associated decisional tasks and dilemmas. Qualitative studies highlighting the content and
contexts of decisional conflicts in PGD can inform research at the intersection of decision
science and ARTs, as well as clinical interventions to reduce patient conflict and increase
decision satisfaction.

CORRESPONDING AUTHORS: Margaux Genoff, B.S., New School for Social Research;
[email protected]; Emily Breitkopf, M.A., New School for Social Research;
[email protected]

Symposium 30C

DOES MATERNAL DISTRESS PREDICT ASSISTED REPRODUCTIVE TECHNOLOGY OUTCOMES? A

META-ANALYTIC INVESTIGATION
S706 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Jennifer Nicoloro SantaBarbara, MSW, MA

Nearly 1.5 million (6%) married women in the U.S. between the ages of 15-44 are infertile.
Infertility can leave a woman feeling defective, angry, out of step with her peers, and lonely
(Benyamini, Gozlan & Kokia, 2009; Frederiksen, Farver-Vestergaard, Skovgard, Ingersiev, &
Zacharia, 2015). In the U.S., half of infertile women seek treatment via assisted reproductive
technology (ART), which for many is physically and emotionally taxing (Demyttenaere, Nijs,
Evers-Kiebooms, & Koninckx, 1991; Klonoff-Cohen 2008). Stress and related negative
emotions have the potential to reduce ART success through both biological and behavioral
mechanisms that affect the likelihood of pregnancy. However, systematic analysis of this topic
is lacking. This review evaluated the association of womens emotional distress with likelihood
of treatment success via ART. Meta-analysis using a random-effects model was conducted on
prospective studies (k = 21) that examined levels of anxiety, depressive symptoms, or
perceived stress before and during treatment in women undergoing ART who achieved
successful outcomes versus women who did not (total N = 4,801). This is the first known
meta-analysis to examine depressive symptoms and the first to distinguish pre-treatment
from procedural distress in evaluating their association with ART outcomes. Results indicate
that women who experienced greater anxiety, depressive symptoms, or perceived stress
before treatment were less likely to have favorable outcomes. Effect sizes were small but
statistically significant. Emotional distress during treatment did not predict outcomes. Fail-
safe N values provided little evidence of publication bias. ART is considered a last resort for
infertile couples who seek to have a biological child and its use is becoming increasingly
common. Thus, there is a need for rigorous research to elucidate factors that affect and are
affected by ART outcomes and to identify the mechanisms responsible for these effects.
Based on our review of existing research, this presentation will offer recommendations for the
design of future studies. Improving our understanding of the experience and outcomes of ART
is a critical first step to reduce the considerable personal, familial, and societal costs
associated with infertility and its treatment.

CORRESPONDING AUTHORS: Marci Lobel, Ph.D., Stony Brook University;

[email protected]; Anne Moyer, Ph.D., Stony Brook University;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S707

Symposium 31 1:30 PM-2:45 PM

THE SOCIAL SIDE OF HEALTH: COMMUNAL COPING WITH ILLNESS CAN IMPROVE HEALTH
BEHAVIOR AND OUTCOMES

Kelly Rentscher, MA1, Vicki Helgeson, PhD2, Caitlin Kelly, M.A.3, Tracey A. Revenson, PhD4

1
University of Arizona, Madison, WI; 2Carnegie Mellon University, Pittsburgh, PA; 3University
of Utah, Salt Lake City, UT; 4Hunter College, New York, NY

This symposium brings together innovative research on communal copinga relationship

process in which family members view a health problem as shared and engage in collaborative
action to address it. These presentations showcase family research across laboratory and real
world settings with a diverse set of methods (observation, self-report, text analysis) and
health samples (diabetes, alcohol use, smoking-related heart and lung disease) to suggest that
a communal approach to addressing health problems can impact chronic illness management
and intervention outcomes. Rentscher presents data on how increases in patient and spouse
communal coping behaviors and language use during couple-focused interventions for health-
compromised smoking and alcohol use predict successful treatment outcomes. Helgeson
presents data on how patient and spouse communal illness appraisals are associated with
patient self-care behavior, perceived support, and illness-related distress among couples
coping with type 1 diabetes. Kelly presents data on how collaboration with parents matters
for the daily management of diabetes among emerging adults that live within and outside of
the parental home. Finally, Revenson will provide commentary and discussion. Together,
these presentations and commentary highlight communal coping as an important potential
mechanism through which relationships influence health behavior and outcomes.

Symposium 31A

PARTNER COMMUNAL COPING IN THE CONTEXT OF COUPLE-FOCUSED INTERVENTIONS FOR

HEALTH PROBLEMS

Kelly Rentscher, MA

Communal coping a relational process in which partners view a problem interdependently

as ours rather than yours or mine, and take collaborative action to address it (Lyons et al.,
1998) has emerged as an important correlate of relationship quality and predictor of partner
S708 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

health. Several studies also link couples first-person plural pronoun use (we-talk), a linguistic
marker of communal coping, to change processes in couple-focused interventions for health
problems and addictions (Hallgren & McCrady, 2015; Rentscher et al., 2015; Rohrbaugh et al.,
2012). Here we present findings from two intervention studies with a combined sample of 56
couples, in which one partner either abused alcohol or continued to smoke cigarettes despite
having heart or lung disease. Specifically, this study involved a process analysis of specific
interview techniques designed to activate communal coping, where the objective was to
identify session-level changes in we-talk and observed couple behavior from before to after
the therapists intervention. Pronoun measures for each partner were obtained via
computerized text analysis from transcripts of partner speech during the target therapy
sessions. Teams of trained raters also observed the target sessions and made independent
ratings of couple communal coping behaviors. Patient and spouse we-talk and communal
coping scores were significantly correlated (rs = .44 to .82, ps < .05). Both patients and
spouses showed increases in we-talk (F = 43.00, p < .001) and communal coping behavior (F =
61.94, p < .001) from a baseline problem-focused therapy block to the active solution-
focused therapy block. Partner increases in we-talk were of similar magnitude, whereas
patients showed greater increases in communal coping behavior than spouses (F = 6.26, p =
.02). Preliminary findings also suggest that within-session increases in spouse communal
coping behaviors (accounting for baseline levels) predict successful patient treatment
outcomes (B = 1.09, p = .02). Results highlight communal coping as a potential mechanism of
action in couple-focused interventions for health problems.

CORRESPONDING AUTHORS: Michael Rohrbaugh, PhD, George Washington University;

[email protected]; Matthias Mehl, PhD, University of Arizona;
[email protected]

Symposium 31B

LINKS OF PATIENT AND SPOUSE COMMUNAL APPRAISALS TO RELATIONAL AND HEALTH

OUTCOMES

Vicki Helgeson, PhD

We investigated diabetes appraisals by asking patients with type 1 diabetes and their spouses
(n = 143 couples) separately whether they perceived diabetes to be the patients problem or a
shared problem (i.e., communal). Patients were 61% female, average age 46 (range 25-75 yrs),
and 95% Non-Hispanic White. Patients completed online questionnaires that assessed
diabetes distress, self-care behavior, and reports of supportive and unsupportive behavior
from spouses. Whereas 71% of spouses said that diabetes was a shared problem, only 26% of
patients said diabetes was shared. Those who scored high on independence centrality (i.e.,
preferences for independence in handling illness) were less likely to report communal
appraisals (p < .01). Although patient communal appraisal was unrelated to self-care behavior,
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S709

spouse communal appraisal was marginally related to better patient self-care behavior (p =
.10). Patient communal appraisals were related to lower diabetes distress (p < .05) and overall
higher marital quality (p < .05). Patient communal appraisals also were related to greater
receipt of emotional support, instrumental support, and collaboration with spouses (all ps <
.05). Spouse communal appraisals were related to lower patient regimen distress, and patient
receipt of greater emotional support, instrumental support, and collaboration (all ps < .05).
We also investigated whether patient appraisals altered the relation of support to outcomes,
focusing on self-care behavior and diabetes distress. There was some evidence of stronger
associations among patients who held communal versus noncommunal appraisals. For
example, emotional support was related to better self-care when patients perceived diabetes
as shared (r = .48, p < .05) but unrelated to self-care when they did not (r = .11, n.s.), z = 2.09,
p < .05. Instrumental support was related to reduced diabetes distress when patients
appraised the illness as shared (r = .35, p < .05) but unrelated to distress when they did not (r
= .06, n.s.), z = 2.15, p < .05. These results suggest that communal illness appraisals might
enhance patient outcomes and optimize the support process.

CORRESPONDING AUTHORS: Meredith Van Vleet, PhD, Carnegie Mellon University;

[email protected]; Caitlin Kelly, MA, University of Utah;
[email protected]; Mary Korytkowski, MD, University of Pittsburgh;
[email protected]; Cynthia Berg, PhD, University of Utah; [email protected]

Symposium 31C

PARENTAL INVOLVEMENT IN EMERGING ADULTS' DAILY DIABETES MANAGEMENT

Caitlin Kelly, M.A.

Little is known about parental involvement with type 1 diabetes (T1D) management during
emerging adulthood. In late adolescence, parental involvement with T1D occurs via diabetes
information-sharing (disclosure to parents, parental solicitation). Parents may be less involved
in daily diabetes management as emerging adults move out of the parental home, but
diabetes care may still benefit from parents involvement. The goal of this study was to
examine the daily frequency of emerging adults disclosure and parental solicitation of
diabetes-related information, the link between parental involvement and T1D management,
and whether these processes differed among participants who lived in the parental home
versus those who did not. Emerging adults with T1D (N = 212; Mage = 18.81; 64.5% female)
reported daily for 14 days on parents involvement with diabetes care (disclosure to parents,
parents solicitation of information, parents knowledge of T1D management, frequency and
type of parental contact) one year after their senior year of high school. Across the 14 days,
63.8% of participants reported contact with their mothers and 46.3% reported contact with
their fathers. Participants reported their parents knew only a little bit about their daily
S710 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

management on average (Mmother = 2.03; Mfather = 1.70 on a 5-point scale). Participants who
lived outside of the parental home reported lower levels of parental contact (Mdiff mother = -
2.64; Mdiff father = -1.44), solicitation (Mdiff mother = -.11; Mdiff father = -.09), knowledge
(Mdiff mother = -.63; Mdiff father = -.49), and mothers, but not fathers, helpfulness (Mdiff = -
.63) (all ps < .05). No differences emerged for disclosure of diabetes-related information.
Notably, when parental involvement did occur, it was associated with improved management,
regardless of whether or not participants lived in the parental home. When mothers solicited
information and emerging adults disclosed diabetes-related information, adherence increased
(p < .05). This suggests that, although the frequency of parental involvement may decrease in
emerging adulthood, especially after leaving the parental home, parents are still an important
source of support for diabetes management.

CORRESPONDING AUTHORS: Amy Hughes-Lansing, PhD, Dartmouth University, Geisel School

of Medicine; [email protected]; Eunjin Lee, PhD, University of Utah,
Department of Psychology; [email protected]; Sara Turner, MS, University of Utah,
Department of Psychology; [email protected]; Deborah Wiebe, PhD, University of
California-Merced,Psychological Sciences and Health Sciences Research Institute; d wiebe @
ucmerced.edu; Cynthia Berg, PhD, University of Utah, Department of Psychology;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S711

Symposium 32 1:30 PM-2:45 PM

IDENTIFYING AND PREVENTING EATING DISORDERS AS A WAY TO DECREASE OBESITY IN

PRIMARY CARE

Jessica Y. Breland, PhD1, Rebecca K. Blais, PhD2, Shira Maguen, Associate Professor/Ph.D.3,
Niloofar Afari, PhD4

1
VA Palo Alto & Stanford University School of Medicine, Menlo Park, CA; 2Utah State
University, Logan, UT; 3San Francisco VA Medical Center and University of California, San
Francisco, San Francisco, CA; 4University of CA, San Diego & VA San Diego Healthcare System,
La Jolla, CA

Disordered eating is fairly common in military and veteran populations, but eating disorder
identification and prevention in this population is hampered by the lack of screening tools
regularly utilized in primary care, as well as a dearth of information on how different military
stressors impact eating disorder risk. To address this gap, the first presentation used cross-
sectional data from 407 women veterans to determine the associations between combat
exposure, military sexual trauma (MST), and eating disorders. The authors found that MST,
but not combat exposure, was significantly associated with eating disorders (OR: 2.03; CI:
1.03-3.98). Asian race was also associated with increased risk (OR: 3.36; CI: 1.26-8.97),
suggesting these factors are important prevention/intervention moderators. The second
presentation examined the longitudinal association between MST and eating disorders among
595,525 Iraq/Afghanistan veterans enrolled in VA care, and whether this association varied by
sex. At one and five years, administrative data demonstrated that risk for eating disorders was
nearly two times higher among those with a positive screen for MST (1-year adjusted odds
ratio [AOR]: 1.99; CI: 1.57-2.40 and 5-year AOR: 1.86; CI: 1.49-2.32). Male veterans with a
history of MST were at differentially greater risk for eating disorders than women with a
history of MST in the 1-year cohort only (AOR: 2.13; CI: 1.01-4.50). The third presentation
reviewed the utility of two existing and one newly proposed eating disorder screens in
primary care. Data were examined in a cohort of 407 women veterans, and positive and
negative predictive values for binge eating disorder, bulimia, and anorexia were determined
for each screen. The authors determined that one new and one existing screen outperformed
a primary-care specific eating disorder screen. Efforts to identify and treat eating disorders
among male and female veterans should focus on veterans who report MST, including men
with MST. Having a strong screening tool in primary care can help identify veterans who
report eating disorder symptoms, thereby assisting with early intervention and prevention of
obesity as well as other weight-related health problems.
S712 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 32A

MILITARY SEXUAL TRAUMA: A UNIQUE EATING DISORDER RISK FACTOR

Jessica Y. Breland, PhD

Background. Eating disorders are more common among veterans than other populations,
suggesting that military service affects eating habits. However, efforts to prevent or reduce
eating disorders and obesity are hampered by a lack of information on how different military
stressors affect eating disorder risk.

Objective. To determine whether military sexual trauma and combat exposure are
independent predictors of eating disorders among a high-risk group: women veterans.

Methods. Women veterans (N=407) who used VA care completed self-report questionnaires
assessing demographics, presence of current eating disorders (using the Eating Disorder
Examination Questionnaire and SCOFF), history of military sexual trauma, history of combat
exposure, depression (using the PHQ-9), and post-traumatic stress disorder (using the PCL-5).
We used logistic regression to predict the presence of any eating disorder (anorexia or bulimia
nervosa, or binge eating disorder) as a function of military sexual trauma and combat
exposure, adjusting for demographic variables. As in past, published work we did not use
mental health conditions as predictors, given overlap in symptoms and unknown times of
onset.

Results. Sixty-three women (15%) met eating disorder criteria, mostly for binge eating
disorder (N=47), 270 women (66%) reported military sexual trauma, and 130 women (32%)
reported combat exposure. Mean age was 49 years (SD=13); 40% of women were veterans of
color. Women who reported military sexual trauma had twice the odds of having an eating
disorder compared to women who did not (OR: 2.03; CI: 1.03-3.98). Combat exposure was not
associated with eating disorders. Of all demographics, only Asian race was associated with
eating disorders. Asian women had much higher odds of having an eating disorder than white
women (OR: 3.36; CI: 1.26-8.97).

Conclusion. Women in this study reported high rates of eating disorders and military sexual
trauma compared to past research, highlighting the importance of screening. For efficient
implementation, screening programs could focus on women reporting military sexual trauma
and, possibly, race. Given associations among eating disorders, obesity, and mortality, such
efforts could greatly improve veteran health.

CORRESPONDING AUTHORS: Rosemary Donalson, MPH, San Francisco VA;

[email protected]; Yongme Li, PhD, San Francisco VA; [email protected]; Claire
Hebenstreit, PhD, San Francisco VA & University of California, San Francisco;
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S713

[email protected]; Lizabeth Goldstein, PhD, San Francisco VA & University of

California, San Francisco; [email protected]; Shira Maguen, PhD, San Francisco VA
& University of California, San Francisco; [email protected]

Symposium 32B

MILITARY SEXUAL TRAUMA IS ASSOCIATED WITH INCREASED RISK FOR EATING DISORDERS
AMONG AFGHANISTAN AND IRAQ VETERANS

Dr. Rebecca K. Blais, PhD

Military sexual trauma (MST) and eating disorders (ED) are underestimated public health
concerns in Veterans, particularly in males. The association between MST and ED, and
whether this association varies by sex, is not well understood. The current study evaluated
whether MST screen status was associated with increased risk for ED in 595,525
Iraq/Afghanistan Veterans within 1- and 5-years after initiating Veterans Health
Administration (VHA) care, and whether the association between MST and ED varied by sex.
Three percent (n=18,488) screened positive for MST. At 1- and 5-year follow up, 0.1% (n= 513,
74% female), and 0.2% (n= 504, 71% female) were diagnosed with ED. In regression models
adjusted for demographic, military service and psychiatric comorbidities, risk for ED was
nearly 2 times higher among those with a positive screen for MST in the 1-year (adjusted odds
ratio [AOR]=1.94, 95% confidence interval [CI]=1.57-2.40) and 5-year (AOR=1.86, 95% CI=1.49-
2.32) cohorts. Male Veterans with a history of MST were at differentially greater risk for ED
than women with a history of MST in the 1-year cohort only (AOR=2.13, 95%CI=1.01-4.50). A
positive screen for MST, particularly in male Veterans, is independently associated with nearly
a 2-fold increased risk for ED. Both MST and ED are historically considered problems that
mainly impact women; however, current results suggest that it is critical to screen for ED in
both men and women Veterans, especially in those who report MST. Early detection and
intervention may reduce the negative sequelae of both MST and ED thereby resulting in more
positive post-deployment adjustment.

CORRESPONDING AUTHORS: Emily Brignone, B.S., Utah State University;

[email protected]; Shira Maguen, PhD, VA Medical Center;
[email protected]; Marjorie Carter, MSPH, VA Medical Center; [email protected];
Jamison Fargo, PhD MSEpi, Utah State University; [email protected]; Adi Gundlapalli,
MD, PhD, VA Medical Center/University of Utah; [email protected]

Symposium 32C

DISCRIMINATIVE ACCURACY OF A NEW SCREENING INSTRUMENT FOR EATING DISORDERS IN

VETERANS
S714 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Dr. Shira Maguen, Associate Professor/Ph.D.

Although there is evidence that eating disorders are more prevalent in veterans than in the
general population, veterans are not routinely screened for eating disorders in primary care.
The goal of this study was to develop an eating disorders screening instrument for veterans in
primary care and compare it to existing screening instruments. Existing instruments were
developed prior to Binge Eating Disorder (BED) becoming a formal diagnosis and have not
been validated with veteran populations. Prevalence rates of BED, Bulimia Nervosa (BN),
Anorexia Nervosa (AN) and Any Eating Disorder (AED) were obtained in a sample of 407
females who completed the Eating Disorder Examination Questionnaire (EDE-Q), the self-
report version of the Eating Disorder Examination, considered the gold-standard for
diagnosing eating disorders. To assess the discriminative accuracy of the new screening
instrument, participants completed the new screener, as well as two existing screeners, the
Eating Disorders Screen for Primary Care (EDS-PC), and the SCOFF screen. EDE-Q prevalence
rates of probable eating disorders in our sample were 12% for BED, 2.2% for BN, 2% for AN,
and 16% for AED. The sensitivity, specificity, receiver operating characteristic curve (ROC), and
positive and negative predictive values of the new screener for BED, BN, AN, and AED were
compared to those of the EDS-PC and the SCOFF. The new screener correctly classified 87.2%
(CI: 74.3%-95.2%) of BED cases, all cases of BN and AN, and 90.5% (CI: 80.4%-96.4%) of cases
with AED. This discriminative power was higher than that of the SCOFF, which correctly
identified 69.6% (CI: 54.2%-82.3%) of BED, 77.8% (CI: 40.0%-97.2%) of BN, 37.5% (CI: 8.52%-
75.5%) of AN, and 66.1% (CI: 53%-77.7%) of AED. While the EDS-PC had slightly higher
sensitivity than the new instrument, the new screener had much better specificity. ROC
results showed that the new instrument was better able to distinguish cases from non-cases
than the EDS-PC (e.g., 0.740; CI: 0.695-0.785 vs. 0.684; CI: 0.649-0.72 for AED). The new
screener outperformed the SCOFF in classifying true cases, outperformed the EDS-PC in
classifying true non-cases, and outperformed the EDS-PC in distinguishing cases from non-
cases of BED, BN, AN and AED. The new screener can be used in primary care to help detect
BED and other eating disorders among veterans, supporting early intervention efforts as well
as the prevention of obesity and other weight-related health issues.

CORRESPONDING AUTHORS: Claire Hebenstreit, PhD, SFVAMC; [email protected];

Yongmei Li, PhD, SFVAMC; [email protected]; Rosemary Donalson, MPH, Highland Hospital;
[email protected]; Julie Dinh, BA, Rice University; [email protected]; Sarah
Dalton, BA, SFVAMC; [email protected]; Emma Rubin, MPH, SFVAMC;
[email protected]; Robin Masheb, PhD, Yale University; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S715

Symposium 33 1:30 PM-2:45 PM

STIGMA AND SUBSTANCE USE: IMPLICATIONS FOR RESEARCH, PRACTICE, AND POLICY

Valerie A. Earnshaw, PhD1, Kristi E. Gamarel, PhD, EdM2, H. Jonathon Rendina, PhD, MPH3,
John F. Kelly, PhD4, Laramie R. Smith, PhD5

1
University of Delaware, Newark, DE; 2Brown University School of Public Health, Providence,
RI; 3Hunter College of the City University of New York (CUNY), New York, NY; 4Mass. General
Hospital/Harvard Medical School, Boston, MA; 5University of California San Diego School of
Medicine, La Jolla, CA

Researchers, clinicians, and policy makers are increasingly recognizing the role of stigma
(social devaluation and discrediting) in substance use epidemics. Stigma can lead to substance
use and misuse among people living with a range of socially devalued characteristics (e.g.,
sexual minority orientation, chronic illnesses including HIV), and undermine the treatment
and wellbeing of people living with substance use disorders. To date, efforts to address the
role of stigma in substance use have met limited success. This symposium brings together
cutting-edge research to foster greater understanding of associations between stigma and
substance use across the lifespan, and discuss ways that this research could be translated to
impact practice and policy aimed at improving the wellbeing of people at risk of or living with
substance use disorders. The first two talks will present studies exploring ways in which
stigma leads to greater substance use, including greater cigarette susceptibility among sexual
minority youth, and greater use of alcohol, marijuana, and other drugs among people living
with HIV. The third talk will present a study exploring ways in which stigma associated with
substance use undermines the wellbeing of people living with HIV. Together, these talks
advance research on stigma and substance use by drawing on advanced methodologies (e.g.,
ecological momentary assessment, moderation and mediation analyses), addressing stigma
associated with a range of characteristics (i.e., sexual minority status, HIV, substance use),
investigating different experiences of stigma (i.e., enacted and internalized stigma), examining
several types of substance use (i.e., cigarette, alcohol, marijuana, and other drug use), and
exploring different stages of the lifespan (i.e., youth, adulthood). The final talk will present a
systematic review of the empirical literature on stigma and substance use and introduce a
theoretical framework linking stigma with substance use. Discussion will be led by a
researcher and clinician in the field of addiction medicine, and will focus on future directions
for research, practice, and policy related to stigma and substance use.

Symposium 33A
S716 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

SUBSTANCE USE STIGMA AMONG PEOPLE LIVING WITH HIV

Dr. Valerie A. Earnshaw, PhD

People with substance use disorders (SUDs) experience pronounced stigma from others.
Although research has largely focused on experiences of SUD stigma from healthcare workers,
recent qualitative work suggests that enacted SUD stigma (i.e., experiences of discrimination
associated with SUDs in the past or present) from family members is particularly distressing.
The current study builds on this work by quantitatively examining enacted SUD stigma from
family members and healthcare providers using the Substance Use Stigma Mechanisms Scale.
Eighty-five people living with HIV who reported current substance use or a history of
problematic substance use were recruited from a community clinic. Participants reported
histories of alcohol (87.1%), marijuana (72.9%), cocaine (62.4%), crack (42.4%), and opiate
(43.5%) use. Approximately one-third (35.1%) reported a history of opioid replacement
therapy. Participants reported more frequent experiences of enacted SUD stigma from family
members than from healthcare providers [t(83)=12.81, p<0.001]. Regression-based analyses
conducted with the PROCESS macro demonstrate that enacted SUD stigma from family
members [B(SE)=0.43(0.21), p=0.04], but not healthcare providers [B(SE)=-0.12(0.38), p=0.77],
is associated with greater depressive symptoms even after controlling for internalized SUD
stigma and socio-demographic characteristics related to depressive symptoms. Moreover,
enacted SUD stigma from family members is associated with decreased motivation to engage
in HIV care via the mediator of greater depressive symptoms [indirect effect (SE)=-0.79(0.42),
CI=-0.12, -1.99]. Findings lend support to previous qualitative work suggesting that enacted
SUD stigma from family members may be more detrimental than enacted SUD stigma from
healthcare providers, and further suggest that enacted SUD stigma from family members may
impact motivation to engage in health behaviors, including those associated with HIV.

CORRESPONDING AUTHOR: Laramie Smith, PhD, University of California, San Diego;

[email protected]

Symposium 33B

THE ROLE OF DISCRIMINATION IN CIGARETTE SUSCEPTIBILITY AMONG SEXUAL MINORITY

YOUTH

Dr. Kristi E. Gamarel, PhD, EdM

Background: Sexual minority youth are more likely to smoke cigarettes compared to their
heterosexual counterparts. However, less is known about sexual orientation differences in
smoking susceptibility (i.e., willingness to use cigarettes if offered by friends), which in an
important precursors to initiation that is necessary to guide prevention programs.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S717

Discrimination has been consistently associated with substance use and may serve as
contributing factor in understanding disparities in cigarette susceptibility among sexual
minority youth. The purpose of this study was to examine the potential role of perceived
discrimination on the association between sexual orientation and cigarette susceptibility.

Methods: The sample was 443 youth (26.6% sexual minority, 59.5% female, 72.0% White,
Mage = 16.7 years) enrolled in a prospective study of substance use initiation and progression.
Participants self-reported their sexual identity and attraction, everyday discrimination
experiences, cigarette susceptibility (i.e., willingness to use if offered by a group of friends),
and whether they had ever smoked.

Results: Accounting for sociodemographic variables, multivariate linear regression models

revealed that sexual minority youth were significantly more likely to report cigarette
susceptibility compared to their heterosexual counterparts (B=0.11, SE=0.06, p<0.05), but
discrimination was not associated with sexual minority status or cigarette susceptibility.
Discrimination moderated associations between sexual orientation and susceptibility (B=0.01,
SE=0.00, p<0.05), such that discrimination was associated with greater cigarette susceptibility
for sexual minority youth, but not among their heterosexual counterparts.

Conclusions: Peers may play a critical role in decisions about cigarette use for sexual minority
youth, and discrimination may be a contributing factor in explaining cigarette susceptibility
among sexual minority adolescents. Future research and prevention programs which target
stigma and peer contexts will be discussed to guide effective tobacco control and prevention
for sexual minority adolescents.

CORRESPONDING AUTHORS: Ethan Mereish, PhD, American University;

[email protected]; Suzanne Colby, PhD, Brown University;
[email protected]; Nancy Barnett, PhD, Brown University;
[email protected]; Kerri Hayes, Brown University; [email protected]; Kristina
Jackson, PhD, Brown University; [email protected]

Symposium 33C

THE CRITICAL INFLUENCE OF DAILY EXPERIENCES OF INTERNALIZED HIV STIGMA ON

SUBSTANCE USE AMONG HIV-POSITIVE GAY AND BISEXUAL MEN

Dr. H. Jonathon Rendina, PhD, MPH

Background: Research has shown associations between levels of HIV stigma and substance
use outcomes such as overall frequency of use. However, the co-occurrence of event-level HIV
stigma and substance use has rarely, if ever, been examined.
S718 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Method: We analyzed 666 daily reports of 47 HIV-positive gay and bisexual men (GBM)
collected as part of a three-week prospective online diary study. We utilized multilevel logistic
models to predict daily use of stimulants, marijuana, and heavy drinking (separate models).
Models were fit with an AR(1) covariance structure and random intercept, and adjusted for
day of diary cycle, whether the report was on a weekend, time since HIV diagnosis, race,
relationship status, and age. Measures of HIV-related forms of internalized stigma (e.g., guilt,
shame) and anticipated/social stigma (e.g., wanting to hide, worry about judgement) were
given daily. These daily measures were disaggregated into individual-level averages and
situational fluctuations around those averages and used to predict daily ART non-adherence.

Results: The sample was diverse with regard to race/ethnicity (75% men of color),
employment status (58% unemployed), and educational background (68% less than college).
Mean age was 38 years (Mdn=33.5) and mean years since diagnosis was 10 (Mdn=9). Overall,
stimulant use was reported on 13.2% of days, marijuana on 23.7%. and heavy drinking on
9.0%. Daily increases in internalized HIV stigma were associated with substantially increased
odds of both stimulant use (AOR = 7.67, p < 0.001) and heavy drinking (AOR = 3.89, p = 0.01),
but was not significantly associated with marijuana use (AOR = 1.34, p = 0.49); daily
fluctuations in anticipated/social stigma were not significantly associated with stimulant use
(AOR = 0.76, p = 0.47), heavy drinking (AOR = 0.64, p = 0.27), or marijuana use (AOR = 0.89, p =
0.74).

Conclusions: Fluctuating levels of internalized HIV stigma (e.g., guilt, shame) from day-to-day
were associated with significantly higher odds of stimulant use and heavy (i.e., binge)
drinking, even after adjusting for daily fluctuations in social stigma as well as individual-level
measures of both forms of stigma (i.e., average tendencies to experience each form of
stigma). Mobile interventions that can be designed to recognize and intervene upon increases
in internalized HIV stigma may provide meaningful improvements in substance use outcomes
for HIV-positive GBM.

CORRESPONDING AUTHORS: Sitaji Gurung, MD, CHEST; [email protected]; Chloe Mirzayi,

MPH, CHEST; [email protected]; Jonathan Lopez Matos, MA, CHEST;
[email protected]; Douglas Keeler, BA, CHEST; [email protected]; Jeffrey
Parsons, PhD, Hunter college, CUNY; [email protected]

Symposium 33D

STIGMA AND SUBSTANCE USE: A SYSTEMATIC REVIEW AND THEORY-BUILDING PROCESS

MODEL

Dr. Laramie R. Smith, PhD

ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S719

The deleterious health effects of stigma and substance use are well documented; however,
the process by which stigma and substance use work to undermine health and well-being is
understudied. Stigma may both lead to increased substance use and risk for developing
substance use disorders (SUDs), as well as undermine the health and treatment of those living
with SUDs. We conducted a systematic review of the extant literature on stigma and
substance use to assess the evidence for a process by which social stigma influences both
substance use and substance use-related health outcomes. Through this review, we evaluate
the ways in which various social stigmas (e.g., sexual orientation, race/ethnicity, mental
illness, incarceration history) place members of these stigmatized groups at increased risk of
developing a SUD. We further evaluate how stigma associated with SUDs, once acquired,
undermine the health of people living with SUDs including their engagement in the SUD
diagnosis, treatment, and recovery process. In this theory-building process, we examine ways
stigma associated with SUDs and other stigmatized characteristics manifest at the structural
(e.g., criminalization laws), interpersonal (i.e., persons who stigmatize), and intrapersonal (i.e.,
persons who are stigmatized) levels. We also examine how intrapersonal manifestations of
stigma (e.g., enacted, anticipated, internalized) affect substance use and treatment behaviors,
and ways such relationships may be mediated by internalizing (e.g., depression, anxiety) or
externalizing (e.g., aggression, sensation seeking) psychosocial mechanisms. We further assess
the available research examining resilience resources (e.g., housing policies, social support,
positive coping strategies) that may moderate the relationship between stigma and negative
substance use outcomes. Discussion will focus on evaluating the depth and quality of available
research on policies and translational practices that serve to reinforce or buffer against the
effects of stigma on substance use-related outcomes.

CORRESPONDING AUTHOR: Valerie Earnshaw, PhD, Univ. of Delaware;

[email protected]
S720 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 34 1:30 PM-2:45 PM

BREAST DENSITY AND MANDATED DENSITY NOTIFICATIONS: RACIAL DIFFERENCES AND

OPPORTUNITIES FOR INTERVENTION

Mark A. Manning, PhD1, Suzanne O'Neill, PhD2, Kristen S. Purrington, PhD, MPH3, Sara J.
Knight, PhD4

1
Karmanos Cancer Institute/Wayne State University School of Medicine, Detroit, MI;
2
Georgetown University, Washington, DC; 3Wayne State University School of Medicine,
Detroit, MI; 4University of Alabama at Birmingham, Birmingham, AL

Breast density (i.e., ratio of fibroglandular to fatty breast tissue) is one of the strongest risk
factors for breast cancer. Over half of US states have implemented mandatory breast density
notification laws to inform women of their density status. Millions of women receive density
notifications each year, even as specific follow-up guidelines remain elusive and as the
medical/epidemiological evidence base regarding breast density as a cancer risk factor
continues to develop. Much remains to be clarified when it comes to breast density-attributed
breast cancer risk, particularly regarding racial differences in the magnitude of risk. Further,
we know little regarding womens reactions to receiving breast density notifications,
particularly regarding racial differences in psychological and behavioral responses to receiving
the notifications, or regarding potential interventions that utilize breast density notifications
to reach women in need of guideline-informed care. We present data from three separate
breast density studies to address these questions. The first speaker will present data showing
between-race differences in the distribution of breast density and corresponding breast
cancer risk for African American and European American women. The second speaker will
present data elucidating between-race differences in psychological responses and decision
making processes following receipt of breast density notifications. The third speaker will
present data on the acceptability and efficacy of a web-based decision support intervention
for women with clinically elevated breast cancer risk due to breast density and other risk
factors, such as family history. The transdisciplinary session will conclude with a panel
discussion led by our discussant, identifying unifying themes related to the state of breast
density research, dissemination of breast density information and decision-making following
receipt of notifications.

Symposium 34A
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S721

BETWEEN-RACE DIFFERENCES IN DECISION MAKING RELATED TO PHYSICIAN

COMMUNICATION FOLLOWING BREAST DENSITY NOTIFICATION

Dr. Mark A. Manning, PhD

Background: A number of states have enacted laws mandating disclosure of dense breasts to
women via their mammogram reports; however, no studies have examined womens
psychological and behavioral responses to receiving notifications. Additionally, no one has
examined how between-race (African American vs. European American) differences in
relevant cognition and emotion are related to psychological and behavioral responses to
receiving BD notifications. Guided by the theory of planned behavior, we examined how
womens knowledge, emotions and perceptions of discrimination influence their intentions
and behaviors related to discussing recently received density notifications with their
physicians.

Methods: We sent survey invitations to screen negative African American and European
American women within two weeks of their screening mammograms. Consenting women
completed measures assessing demographic variables (e.g., age, income, education, and
marital status), breast density and breast cancer risk knowledge, breast density anxiety,
breast cancer worry, perceptions of group-based medical mistrust, perceptions of
discrimination, and attitudes, norm, control perceptions and intentions towards talking to
physicians about their breast density notification. Three months later, we sent women a
follow-up survey assessing whether they talked to their physicians about the breast density
notifications. We used path models to examine the influence of cognition, emotion and
perceptions of discrimination on womens behavior, mediated by the cognitive processes
implied by the theory of planned behavior.

Results: African American women had stronger intentions to discuss their breast density
notifications with their physicians; however, there were no between-race differences in actual
behaviors. Nonetheless, there were significant between-race differences in the processes that
lead to behavior. Intentions predicted behaviors for European American women, but not for
African American women. Rather, cognition, emotion and perceptions of discrimination had
significant direct effects on behaviors for African American women.

Conclusion: These data suggest that there are significant between-race differences in the
cognitive processes related to physician communication among women who receive breast
density notification.

CORRESPONDING AUTHORS: Terrance L. Albrecht, PhD, Karmanos Cancer Institute/Wayne

State University School of Medicine; [email protected]; Julie Shultz, MD, Karmanos
Cancer Institute/Wayne State University School of Medicine; [email protected]; Zeynep
Yilmaz-Saab, MD, Karmanos Cancer Institute/Wayne State University School of Medicine;
S722 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

[email protected]; Kristen Purrington, PhD, Karmanos Cancer Institute/Wayne State

University School of Medicine; [email protected]

Symposium 34B

PRELIMINARY FINDINGS OF A DECISION SUPPORT TOOL FOR WOMEN WITH DENSE BREASTS

Suzanne O'Neill, PhD

Background: Mammographic breast density is one of the strongest risk factors for breast
cancer. Mandatory density disclosure policies are increasing nationally without established
best practices or evidence of clinical benefit. Density disclosure could serve as an opportunity
to communicate with women with clinically elevated risk for breast cancer. Guided by
Protection Motivation Theory and the RE-AIM framework, we are testing a personalized, web-
based breast cancer risk communication and decision support tool for women with dense
breasts whose elevated cancer risk makes them eligible to consider guideline-informed risk
management (chemoprevention, breast MRI).

Methods: We have engaged multiple stakeholders within an integrated healthcare system

throughout the tool development and trial planning process (patient focus groups and
usability pre-testing; health care provider and delivery system leader interviews). We are
currently pilot testing the tool with 20 women in our target population prior to trial launch in
January 2017. Pilot outcomes will include knowledge, perceived breast cancer risk and
severity, cancer worry, intentions for screening and chemoprevention and the response cost
and response efficacy of these risk management options.

Results: Focus group results suggested that most women were aware of breast density, but
unfamiliar with its role as a breast cancer risk factor. Participants in usability pre-testing rated
the tool highly on acceptability and ease of use. Health care providers and clinical leaders
highlighted key areas for provider engagement prior to the trials launch, such as clarifying
pathways for referral and monitoring of clinical impact within the health system. Results of
pretesting and preliminary experiences with the larger trial will be shared in the symposium.

Conclusions: Preliminary results suggest acceptability of our web-based tool by patients and
providers within an integrated healthcare system.

CORRESPONDING AUTHORS: Karen Wernli, PhD, Group Health Research Institute;

[email protected]; Sarah Knerr, PhD, University of Washington; [email protected]; Kelly
Ehrlich, MS, Group Health Research Institute; [email protected]; Amanda Graham, PhD, Truth
Initiative; [email protected]; David Farrell, MPH, People Designs, Inc.;
[email protected]; Chalanda Evans, BS, Georgetown University;
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S723

[email protected]; Marc Schwartz, PhD, Georgetown University;

[email protected]; Kathleen Leppig, MD, Group Health Cooperative;
[email protected]

Symposium 34C

RACIAL DIFFERENCES IN BREAST DENSITY DISTRIBUTION AND CANCER RISK IN AN URBAN

HOSPITAL-BASED COHORT

Kristen S. Purrington, PhD, MPH

Breast density (BD) is a strong risk factor for BC (BC). Women with the highest density (75%)
are at a 4.6-fold increase in risk of BC compared to women with the lowest density (5%). BD
and associated BC risk has not been well studied for African American (AA) women. Among
studies that reported data for AA women, results are conflicting on whether AA women tend
to have higher or lower BD than white women or whether the BC risks associated with BD are
the same these two groups. We conducted a retrospective study to explore racial differences
BD patterns and resulting cancer risk at the Karmanos Cancer Center Imaging Department. We
have currently abstracted BI-RADS BD (a: almost entirely fatty, b: scattered fibroglandular
density, c: heterogeneous fibroglandular density, d: extremely dense) from radiology reports
for 16,167 out of 20,699 women (82% AA; 18% white) who underwent routine
mammographic screening from 2012-2014. Subsequent BCs as well as previous history of BC
were determined by linkage with the Metropolitan Detroit Cancer Surveillance System
database as well as radiology records. BI-RADS BD and volumetric BD estimates from Volpara
Research software were also obtained for an additional 5,484 women for whom raw
mammogram images were available. The overall distribution of BD was significantly different
for white and AA women, where AA women were less likely than white women to have
scattered [Odds ratio (OR)=0.79, 95% confidence interval (CI) 0.71-0.89], heterogeneous
[OR=0.73, 95% CI 0.62-0.86], or entirely dense breasts [OR=0.62, 95% CI 0.47-0.82] compared
to fatty breasts, adjusted for age and body mass index (BMI). Similarly, although increasing BD
is associated with an increase in BC risk for both AA and white women, the effects of BD
adjusted for age were stronger for AA [Scattered vs fatty: OR=2.23; Heterogeneous/dense vs
fatty: OR=3.65] compared to white women [Scattered vs fatty: OR=1.14;
Heterogeneous/dense vs fatty: OR=1.58]. Data were sparse for white BC cases and for women
with entirely dense breasts. We also explored changes in BI-RADS BD over time and the
agreement between BI-RADS and volumetric BD. Among 1,884 women who had 2
mammograms within 2 years, 521 (27.6%) were assigned a different BI-RADS classification at
each time point and 185 (9.8%) of these women were assigned a category that would shift
them from being considered to have dense vs. non-dense breasts. Similarly, 46.4% of women
were assigned to a different BI-RADS classification using volumetric breast density vs
subjective assessment, and 12.1% were assigned a category that would shift them from being
considered to have dense vs. non-dense breasts. These data confirm that BD definitions are
S724 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

subjective and dependent on the measurement tool, which has implications for recently
implemented BD notification laws across the United States.

CORRESPONDING AUTHORS: Mark Manning, PhD, Wayne State University;

[email protected]; Terrance Albrecht, PhD, Wayne State University;
[email protected]; Zeynep Yilmaz-Saab, MD, Karmanos Cancer Institute;
[email protected]; Julie Shultz, MD, Karmanos Cancer Institute; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S725

Symposium 35 1:30 PM-2:45 PM

RELIGIOUSNESS/SPIRITUALITY AND HEALTH BEHAVIORS ACROSS THE CANCER CONTINUUM

Crystal L. Park, PhD1, John M. Salsman, PhD2, Crystal L. Park, PhD1, Melinda Stolley, MA, PhD3,
Cheryl L. Holt, PhD4, Sarah D. Mills, MS, MPH5

1
University of Connecticut, Storrs Mansfield, CT; 2Wake Forest School of Medicine, Winston
Salem, NC; 3Medical College of Wisconsin, Milwuakee, WI; 4University of Maryland, College
Park, MD; 5San Diego State University/University of California, San Diego Joint Doctoral
Program in Clinical Psychology, Los Angeles, CA

Spirituality has been shown in many studies to be fairly consistently related to mental and
physical health (see Koenig, 2012 for a review), but the mechanisms explaining these linkages
are not well-understood. Many different pathways of influence have been proposed, but few
have been tested. Adding complexity, spirituality comprises many different dimensions,
including beliefs, behaviors, motivation, emotions, and relationships, and each may relate
differently to health. One important but understudied pathway is that of health behaviors.
This symposium focuses on how different dimensions of spirituality may link with health
behaviors across the cancer continuum, from healthy populations to those at higher risk of
certain cancers to cancer survivors. Sarah Mills will describe relationships between beliefs
about religious influences on health and health behaviors in a sample of churchgoing Latina
women. Cheryl Holt will present results of a national longitudinal study of African American
adults examining whether religious social support mediates relationships between multiple
aspects of religious involvement and cancer control behaviors in African Americans. Melinda
Stolley will present results of a study of relationships between religiosity/spirituality and
health behaviors in minority adult childhood cancer survivors compared to White childhood
cancer survivors and to non-cancer controls. Crystal Park will discuss relationships between
multiple dimensions of spirituality and health behaviors in survivors of breast cancer, focusing
on potential negative as well as positive links. Our discussant, John Salsman, an expert in
spirituality and health, will draw out common themes across the presentations, focusing on
the mechanisms that may account for these linkages.

Symposium 35A
S726 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

DIMENSIONS OF SPIRITUALITY ARE DIFFERENTIALLY ASSOCIATED WITH DIET AND PHYSICAL

ACTIVITY IN BREAST CANCER SURVIVORS

Dr. Crystal L. Park, PhD

Health behaviors, while important for everyone in terms of morbidity and mortality risks, are
especially important for cancer survivors, who are at heightened risk not only for recurrence
and additional primary cancers, but also for other conditions such as metabolic syndrome and
cardiovascular disease. Some psychosocial factors, such as social support and self-efficacy,
have been identified as promoting healthful lifestyles, but much remains to be learned. One
potentially important factor, spirituality, has been suggested as promoting better health, but
few studies have examined these associations. The present study investigated links between
three dimensions of spirituality (behaviors, beliefs, and identity) and health behaviors in a
sample of 172 recent survivors of breast cancer enrolled in a lifestyle change intervention.
Present analyses are from baseline surveys completed prior to randomization. Participants
were primarily white (93%), married (73%), and well-educated (74% had a college degree or
higher); mean age was 54 (SD = 10). Results of hierarchical multiple regression analyses
controlling for age and education indicated that frequency of service attendance was related
to less moderate/vigorous physical activity while frequency of private prayer was related to
consumption of more fruits and vegetables (but not total calories consumed). Degree of
identity as religious was negatively related to moderate/vigorous activity and higher BMI but
also with lower total calorie consumption. Beliefs in God were unrelated to health behaviors,
but belief in an afterlife was related to less moderate/vigorous physical activity and higher
BMI. These results suggest that spirituality has some complex but potentially important
associations with health behaviors in breast cancer survivors. Further, different dimensions
have different effects and many of these effects are in an unexpectedly adverse direction.
More research is needed to understand these relationships and to determine how spiritual
dimensions may play a useful role in lifestyle change interventions.

CORRESPONDING AUTHOR: Dalnim Cho, PhD, City of Hope; [email protected]

Symposium 35B

SPIRITUALITY AND HEALTH BEHAVIORS IN ADULT MINORITY CHILDHOOD CANCER SURVIVORS

Melinda Stolley, MA, PhD

Background. Over 75% of childhood cancer patients survive long-term yet over 60% of
survivors will experience adverse late effects including cardiovascular disease (CVD),
osteoporosis and secondary cancers. Practicing healthy lifestyles is important and may be
particularly salient for African-American and Hispanic survivors given increased risk for chronic
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S727

disease. Few studies have examined health behaviors in minority childhood cancer survivors
(CCS) and fewer still have examined correlates of health promoting behaviors outside of
demographic factors. This presentation examines spirituality and health behaviors in a diverse
cohort of adult CCS and non-cancer controls.

Methods. The Chicago Healthy Living Study was a cross sectional study with 450 adult CCS
(n=150 African-American, 152 Hispanic, 150 white) recruited via Chicago area hospital cancer
registries and 375 racial/ethnic matched controls recruited via targeted digit dial. Participants
completed a one-time 2-hour interview with validated measures of body mass index (BMI),
health behaviors (diet, physical activity, smoking, alcohol use) and the brief Systems of Belief
Inventory (SBI-15). Analyses adjusted for age, sex and SES examined differences in health
behaviors and spirituality within racial/ethnic groups between CCS and controls and between
racial/ethnic groups within CCS. We also explored associations between spirituality and
obesity, current smoking and adherence to dietary (percent calories from fat and saturated
fat, sodium, calcium) and physical activity recommendations (150 weekly mins of moderate
activity).

Results. Overall BMI and health behaviors were similar between CCS and controls across
racial/ethnic groups. Within CCS, minority survivors were more likely to be obese and not
adhere to dietary recommendations. Spirituality was similar between CCS and controls among
African-Americans and whites, but Hispanic CCS SBI scores were higher than controls
(p<.005). Within survivors, spirituality was highest among African-Americans followed by
Hispanics and then whites (p<.001). Initial analyses show no association between spirituality
and health behaviors among CCS, but sub-group analyses may reveal more.

Discussion. Spirituality was highest among minority survivors, yet there were no associations
with adherence to healthy lifestyle behaviors. Other factors such as knowledge, perceived risk
or engagement in long-term follow-up care may be more relevan to health behaviors,
whereas spirituality may be more tied to quality of life in this populaiton. Understanding
correlates of health behaviors among this high risk group will help to inform the development
of relevant interventions.

CORRESPONDING AUTHORS: Patricia Sheean, PhD, RD, Loyola University; [email protected];

Linda Schiffer, MS, University of Illinois at Chicago; [email protected]; Anjishnu Banarjee, PhD,
Medical College of Wisconsin; [email protected]; Lisa Sharp, PhD, University of Illinois at
Chicago; [email protected]

Symposium 35C

DOES RELIGIOUS SOCIAL SUPPORT MEDIATE THE RELATIONSHIP BETWEEN RELIGIOSITY AND
CANCER CONTROL BEHAVIORS IN AFRICAN AMERICANS?

Dr. Cheryl L. Holt, PhD

S728 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Abundant research has examined the association between religious involvement and health-
related outcomes. Additional research has examined reasons for this religion-health
connection, including studies testing theoretical models that cover a variety of potential
explanatory mechanisms. However, most of this research has been cross-sectional in nature.
Given the relatively high levels of religious involvement among African Americans and
research suggesting that faith communities provide a unique form of religious social support
that may impact health, the present study aimed to test a model of religious social support as
a mediator of religious involvement and a variety of cancer control behaviors in a national
sample of African Americans. The RHIAA (Religion and Health In African Americans) study
conducted three waves of telephone interviews with N=3173 African Americans over a 5-year
period. N=766 participants provided data at all three waves. Measurement models were fit to
the data followed by longitudinal structural models. Religious beliefs were not associated with
a change in religious social support over time. However, religious behaviors (e.g., church
attendance) were associated with a significant increase in religious social support over time.
Religious social support was associated with decreased binge drinking and vegetable
consumption, and increased reports among men of receiving prostate cancer screening. There
was no evidence of a mediation effect of religious social support in the association between
religious beliefs and health-related outcomes. However, the relationship between religious
behaviors and decreased binge drinking over time was mediated by increases in religious
social support. Implications for theory and practice are discussed.

CORRESPONDING AUTHORS: Jin Huang, PhD, Johns Hopkins University; [email protected];

Eddie Clark, PhD, Saint Louis University; [email protected]; David Roth, PhD, Johns Hopkins
University; [email protected]; Crystal Park, PhD, University of Connecticut;
[email protected]

Symposium 35D

THE PERCEIVED RELIGIOUS INFLUENCE ON HEALTH BEHAVIOR SCALE AND THE ILLNESS AS
PUNISHMENT FOR SIN SCALE AMONG CHURCHGOING LATINAS

Ms. Sarah D. Mills, MS, MPH

The present study 1) evaluated the psychometric properties of the Spanish versions of the
Perceived Religious Influence on Health Behavior scale, a measure of the extent to which an
individuals health behaviors are affected by his or her religion, and the Illness as Punishment
for Sin scale, a measure of the extent to which illness is believed to be an act of retribution,
and 2) examined how these measures relate to health behaviors (i.e., physical activity,
food/beverage consumption, mammography screening, sleep), in a sample of churchgoing
Latina women (N = 404). For the Perceived Religious Influence on Health Behavior scale,
confirmatory factor analysis provided support for the expected one-factor model, internal
consistency reliability was good, and there was evidence of convergent validity. For the Illness
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S729

as Punishment for Sin scale, confirmatory factor analysis provided support for the expected
one-factor model, but on a revised, seven-item version of the measure. Internal consistency
reliability for this revised version was good, but evidence of convergent validity was mixed.
For the second aim, no significant relationships (p > .05) were found between the measures
and the health behaviors examined. It is recommended that future studies use the Perceived
Religious Influence on Health Behavior scale and the revised Illness as Punishment for Sin
scale when examining these constructs among Latina women. Additional research is needed
to clarify why no significant relationships were found between these measures and the health
behaviors examined.

CORRESPONDING AUTHORS: Elva M. Arredondo, PhD, Graduate School of Public Health, San
Diego State University, Institute for Behavioral and Community Health (IBACH);
[email protected]; Lilian G. Perez, MPH, Graduate School of Public Health, San Diego
State University, Institute for Behavioral and Community Health (IBACH);
[email protected]; Jessica Haughton, MPH, MA, Graduate School of Public
Health, San Diego State University, Institute for Behavioral and Community Health (IBACH);
[email protected]; Scott C. Roesch, PhD, San Diego State University/University of
California, San Diego Joint Doctoral Program in Clinical Psychology; [email protected];
Vanessa L. Malcarne, PhD, San Diego State University/University of California, San Diego Joint
Doctoral Program in Clinical Psychology; [email protected]
S730 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 36 1:30 PM-2:45 PM

BASIC HEALTH DECISION RESEARCH IN DIVERSE POPULATIONS: ONGOING WORK AND FUTURE
DIRECTIONS

Jennifer Hay, PhD1, Austin S. Baldwin, PhD2, Kimberly Kaphingst, ScD3, Amber Emanuel,
Ph.D.4, Betina Yanez, Ph.D.5

1
Memorial Sloan Kettering Cancer Center, New York, NY; 2Southern Methodist University,
Dallas, TX; 3University of Utah and Huntsman Cancer Institute, Salt Lake City, UT; 4University
of Florida, Gainesville, FL; 5Northwestern University Feinberg School of Medicine, Chicago, IL

The science of decision making has important implications for behavioral medicine. Problems
such as how, when, and among whom health protective behaviors are adopted and
maintained, and how risk behaviors can be best eliminated have been informed by a thriving
body of work. One example is the finding that people make different health decisions based
on whether the consequences of a behavior are framed as gaining something positive or
avoiding something negative. However, much of the work regarding the basic science of
decision making has been conducted in academic samples of convenience, where decisional
experiments can be easily mounted, or using on-line survey panels, where large samples can
be collected with modest resources. Such samples tend to be highly educated and are not
racially or ethnically diverse enough to be representative of the US population. Measures,
findings, and interventions developed in these ways are often adapted and applied across
more diverse settings, yet a growing body of evidence indicates that basic decisional
principles may differ in profound ways across different population subgroups. Consequently,
there is untapped potential to develop novel theory and decision principles based on data
from more diverse samples and to thereby increase the likelihood that subsequent
interventions are effective in underserved populations. In this Symposium, jointly sponsored
by Health Decision Making and Ethnic Minority and Multicultural Health Special Interest
Groups, we will provide three examples of basic health decision research in diverse
populations. Austin Baldwin will present research regarding self-persuasion principles in a
safety-net population with lower education and literacy levels. Amber Emanuel will present
findings regarding the use and limits of self-affirmation interventions in a community sample.
Kim Kaphingst will present work examining the complex influences of decisional conflict, trust,
and race on intentions to donate to a cancer biobank in a diverse population. Finally, Betina
Yanez, an expert in racial/ethnic minority populations, will provide discussion regarding the
use of basic decision research to enhance health across diverse populations. This Symposium
will encourage applied and basic researchers to conduct basic decision research in diverse
community and clinic settings, and with US subpopulations that vary across socioeconomic
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S731

status as well as racial and ethnic subgroups. Such work will enhance basic science and lead
to new opportunities for health promotion.

Symposium 36A

DEVELOPING AND TESTING A SELF-PERSUASION INTERVENTION IN A SAFETY-NET CLINIC

POPULATION

Prof. Austin S. Baldwin, PhD

Objective: Self-persuasion is an effective behavior change strategy (Stice et al., 2008).

However, self-persuasion interventions have typically been used among well-educated
populations (Baldwin et al., 2013) and have not been developed for lower income and diverse
populations or to promote human papillomavirus (HPV) vaccination. We developed a tablet-
based application with voiceover narration (in English and Spanish) to elicit parental self-
persuasion for adolescent HPV vaccination. We then evaluated the feasibility of the self-
persuasion tasks (generate and verbally articulate ones own arguments for HPV vaccination)
and their effect on parents decision-making in a racially and ethnically diverse safety-net
population with lower education and literacy levels.

Methods: Parents (N=45) of age-eligible adolescents used the self-persuasion application.

Then, during cognitive interviews, staff gathered quantitative and qualitative feedback on the
self-persuasion tasks and on parental decision stage.

Results: The self-persuasion tasks were rated as easy to complete (M range 1.00 to 2.32 on a
5-point difficulty scale) and helpful (M range 4.33 to 4.98 on a 5-point helpfulness scale).
Among the 33 parents who had not previously thought about or were undecided about
getting the vaccine for their adolescent, 27 (81.8%; 2(1) = 13.36, p < .001, = .64) reported
deciding to get their adolescent vaccinated after completing the self-persuasion tasks, a
significant and large effect. No results differed between Spanish- and English-speakers.

Conclusions: The self-persuasion application was feasible and facilitated parents verbalizing
their own reasons for HPV vaccination in a low literacy, diverse safety-net population. Its use
also resulted in a change in parents decision stage. In a current randomized controlled trial
(RCT) conducted in this lower income and diverse population, we are testing (1) the efficacy of
the tablet-based application on HPV vaccine series initiation and completion, and (2) whether
basic cognitive and motivational mechanisms (i.e., deep cognitive processing, autonomous
motivation) hypothesized to underlie self-persuasion mediate its effect.

CORRESPONDING AUTHOR: Jasmin Tiro, PhD, University of Texas Southwestern Medical

Center; [email protected]
S732 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 36B

IMPACT OF RACE, DECISIONAL CONFLICT, AND TRUST ON BIOBANK INTENTIONS IN AN

UNDERSERVED POPULATION

Kimberly Kaphingst, ScD

Introduction. Biobanks are an essential resource for research in the era of precision medicine.
Participation of all population subgroups is critical so that findings reflect the needs of diverse
communities and address disparities. Research is needed to examine factors that impact
decisions to participate in a biobank among underserved populations.

Methods. We conducted a randomized experiment to examine how model of consent for

secondary use of biospecimens affected intentions to donate to a cancer-related biobank. We
recruited 356 women aged 40 and older stratified by race; 56% were Black and 44% were
White. Participants reviewed three informed consent documents in a randomized order based
on different models of consent: notice (informs that samples will be used in future research);
broad (asks permission once for multiple future studies); and study-specific (asks permission
before each future study). After each document, we assessed intention to donate with a four-
point scale from definitely not to definitely. We also assessed trust in medical research
and decisional conflict as factors that might affect intention. We used multivariable logistic
regression to examine predictors of intention for each consent model.

Results. 47% of participants had an annual household income less than $20,000; 31% had no
education beyond high school. The proportion who definitely intended to donate was 58%
for study-specific; 44% for broad; and 41% for notice. In multivariable analyses, Black women
were significantly less likely to intend to donate for each consent model than White women
(p<.05). Those with higher decisional conflict were significantly less likely to intend to donate
for each consent model (p<.05). After adding trust, race was no longer a significant predictor
of intention for the notice and broad consent models.

Discussion. Intentions to donate to a cancer biobank differed by race; this association was
partly mediated by trust. Decisional conflict also lowered intentions to donate. These findings
suggest that strategies to assist informed decision making are essential for underserved
populations and that trust may be critical to decisions about donating to biobanks.

CORRESPONDING AUTHORS: Katherine Brown, MPH, Washington University in St. Louis;

[email protected]; Sarah Lyons, MPH, Washington University in St. Louis;
[email protected]; Bettina Drake, PhD, Washington University in St. Louis;
[email protected]

Symposium 36C
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S733

USING SELF-AFFIRMATION IN A COMMUNITY SAMPLE TO INCREASE ORAL CANCER

SCREENINGS

Amber Emanuel, Ph.D.

Objective: Self-affirmations, when an individual affirms a valued aspect of his or her self,
allow individuals to accept threatening health information. Accepting this threatening health
information may then allow individuals to decide to take action related to their health. In a
study of low-income men a self-affirmation manipulation was conducted to examine if after
being self-affirmed, high-risk men would be more accepting of health information related to
their risk of oral cancer.

Methods: Participants (n = 152) were Black or White men over the age of 40 who had smoked
at least 100 cigarettes in their lifetime. Participants were recruited by word of mouth and
interviews were conducted in community settings, such as churches or classrooms at local
colleges. Participants answered questions about their health behaviors, completed a standard
self-affirmation manipulation, watched a video that explained the risks of oral cancer and
detailed the process of an oral cancer examination, and then answered questions related to
their attitudes and intentions. Participants could then schedule a free oral cancer screening.

Results: The majority of participants were low-income (49% of participants reported income
of less than $20,000) and of lower education (61% of participants had a High School degree or
less). Approximately half of participants completed a self-affirmation condition or a control
condition. However, the self-affirmation manipulation failed to increase self-affirmative
thoughts and feelings. Participants in the self-affirmation condition, following the self-
affirmation manipulation, did not rate themselves as being more affirmed (M = 4.56, SD = .97)
than those participants in the control condition (M = 4.46, SD = 1.19), t (176) = .62, p = .54.
Across both conditions, participants reported low defensive avoidance of information related
to oral cancers (M = 1.12 (out of 5), SD = .49). However, men who did report high information
avoidance were more likely to make risky health-related decisions. Specifically, defensive
avoidance was related to lower intentions to schedule a free oral cancer screening (r = -.18, p
= .02).

Conclusions: Since self-affirmation usually dampens the defensive response against

threatening health information, the finding that defensive avoidance of information led to
lower intentions highlights the need for successful implementations of self-affirmations in this
population. In the current study, self-affirmations did not predict lower defensive information
avoidance, but the manipulation used for self-affirmations in the current study may not be as
ideal for community populations as they are for the college undergraduate populations.
Suggestions for how to successfully implement self-affirmations in low-income or low-literacy
populations so individuals can make better health related decisions will be discussed.

CORRESPONDING AUTHOR: James Shepperd, Ph.D, University of Florida; [email protected]

S734 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 37 1:30 PM-2:45 PM

LUNG CANCER SURVIVORS: A GROWING POPULATION WITH UNIQUE PSYCHOLOGICAL AND

BEHAVIORAL NEEDS

Jamie Ostroff, PhD1, Heidi Hamann, PhD2, Elyse R. Park, PhD, MPH3, Jamie L. Studts, PhD4,
Maureen Rigney, MSW5

1
Memorial Sloan Kettering Cancer Center, New York, NY; 2University of Arizona, Tucson, AZ;
3
Massachusetts General Hospital, Boston, MA; 4University of Kentucky College of Medicine,
Lexington, KY; 5Lung Cancer Alliance, Washington, DC

Introduction: Lung cancer is the leading cancer killer for both women and men. However,
recent advances in low dose CT screening, multimodality management, targeted
chemotherapy, and immunotherapy are responsible for an estimated 384,000 survivors of
lung cancers. Although only a small proportion of cases are diagnosed at an early stage each
year, of the 225,000 people diagnosed annually with lung cancers, more than 32,000 will
survive for at least five years. Lung cancer survivors face unique challenges, including stigma
(the perception and internalization of negative appraisal and devaluation from others), risk of
persistent smoking, and a dearth of specialized survivorship resources.

Methods: Here, we report on data from lung cancer survivors that address these concerns
and provide guidance for future research and service. Each presentation addresses unique,
yet interconnected themes, related to lung cancer survivorship.

Results: The first presentation describes a patient-focused process of developing a measure of

lung cancer stigma. The next presentation focuses on both short- and long-term smoking and
cessation data from a national cohort of lung cancer survivors and public responses to these
data. Presentations conclude with an introduction to a novel survivorship program aimed at
the unique needs of individuals with lung cancer. Our discussant offers future directions for
research, clinical practice, and advocacy to address the unmet needs of lung cancer survivors.

Discussion: Lung cancer survivors are a growing proportion of the overall cancer survivorship
population with unique psychosocial and behavioral challenges. Increasing our focus on
stigma, smoking cessation, and survivorship interventions represents crucial steps in
addressing the needs of lung cancer survivors.

Symposium 37A
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S735

DEVELOPMENT AND PRELIMINARY PSYCHOMETRIC EVALUATION OF A PATIENT-REPORTED

OUTCOME MEASURE FOR LUNG CANCER STIGMA

Dr. Heidi Hamann, PhD

Introduction: The robust causal connection between smoking and lung cancer underscores
the relevance of smoking prevention and cessation interventions. However, it may also
inadvertently result in stigma (negative appraisal and devaluation) toward lung cancer
patients and survivors. Internalization of this stigma can have negative psychosocial
consequences, including depressive symptoms, lower quality of life, and reduced satisfaction
with provider communication. Increased recognition of lung cancer stigma calls for attention
to identifying valid and reliable measurement tools that incorporate significant patient and
expert input.

Method: We report on the multiphase process of measure development to assess patient-

reported stigma. Our earlier work described the process of patient and survivor interviews
and development of a conceptual framework for lung cancer stigma. Here, we describe the
process of item generation/refinement, and initial psychometric scale evaluation for the Lung
Cancer Stigma Inventory (LCSI).

Results: We started with a pool of 240 items culled from existing stigma scales and new items
based on our conceptual framework. Item refinement (to 49 items) was achieved through a
multistep process including: a) expert input on item relevance and clarity, and b) cognitive
interviews with 20 lung cancer patients and survivors. A multi-site sample of 231 lung cancer
patients and survivors completed the 49-item survey. Exploratory factor analysis (EFA) was
used to identify a 25-item scale with three primary factors: Internalized Stigma (alpha=.90),
Perceived Stigma (alpha=.74), and Constrained Disclosure (alpha=.82). Two-week test-retest
correlation was high (r = .90), suggesting strong stability of measurement over time.

Conclusions: Based on current measurement guidelines that emphasize significant patient

input and our conceptual model, we have developed a reliable new measure of lung cancer
stigma, the Lung Cancer Stigma Inventory (LCSI). Two factors reflected our concept elicitation
work identifying relevant themes of negative appraisal and devaluation from others
(Perceived Stigma), as well as self-blame, guilt, and regret (Internalized Stigma). A third factor,
Constrained Disclosure, reflects information withholding and avoidance that may suggest
interpersonal intervention pathways. Subsequent psychometric work is needed to establish
further evidence of validity and clinically meaningful change.

CORRESPONDING AUTHORS: Megan Shen, PhD, Weill Cornell Medical College;

[email protected]; Anna Thomas, PhD, Baylor Medical Center;
[email protected]; Simon Lee, PhD, UT Southwestern Medical Center;
[email protected]; Jamie Ostroff, PhD, MSKCC; [email protected]
S736 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 37B

SMOKING RATES AMONG LUNG CANCER SURVIVORS AND CORRESPONDING PUBLIC

RESPONSE

Dr. Elyse R. Park, PhD, MPH

Background: Smoking is the leading risk factor for lung cancer; continued smoking after
diagnosis may adversely affect treatment outcomes, subsequent cancers, and QOL. We
assessed short and long term smoking rates among a population-based cohort of lung cancer
survivors (LCS) and explored public opinion to better understand stigmatization toward them.

Methods: We examined smoking rates, at 5-months and 6 years post-diagnosis (n=2455 and
198, respectively; median age=65; 69% White; 45% Stage I/II), among LCS from the Cancer
Care Outcomes Research and Surveillance cohort. Short-term results were published online,
and 432 comments were posted on a public discussion board. We used qualitative content
analysis to establish a theoretical framework to explain public attitudes about LCS smoking.

Results: 38% of LCS were current smokers during the year diagnosis; 14% were smokers at 5
months post-diagnosis. 6 post-diagnosis, 16.2% of LCS reported tobacco use, but only 6.4%
participated in a smoking cessation program, in the past year. Public responses to LCS
smoking were divided into sympathetic vs. stigmatizing attitudes. Attitudes were shaped by
personal experiences with smoking and familiarity with lung cancer. Personal experiences
influenced beliefs about cancer (e.g., causes and fatalistic outcomes) and smoking (e.g., ability
to quit, benefits of quitting after cancer). Stigmatizing beliefs toward LCS who smoked were
expressed by individuals less familiar with smoking and lung cancer; these individuals tended
to endorse the inevitability of smoking causing cancer, were pessimistic about the benefits of
quitting post diagnosis, and emphasized personal choice over nicotine addiction

Conclusions/Implications: A sizable minority of LCS did not quit smoking and, at long-term
follow-up, continued to use tobacco; few sought support. LCS may be vulnerable to public
stigma, particularly from individuals unfamiliar with the struggles of quitting and its benefits.
This work provides insight into challenges LCS might face in their attempts to quit.

CORRESPONDING AUTHORS: Christina Luberto, PhD, Massachusetts General Hospital;

[email protected]; Giselle Perez, PhD, Massachusetts General Hospital;
[email protected]; Emily Friedman, Massachusetts General Hospital;
[email protected]

Symposium 37C
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S737

DEVELOPMENT OF THE KENTUCKY LEADS COLLABORATIVE LUNG CANCER SURVIVORSHIP

CARE PROGRAM

Dr. Jamie L. Studts, PhD

Introduction: Individuals diagnosed with lung cancer commonly suffer numerous threats to
preserving quality of life, including substantial symptom burden, clinically significant distress,
and considerable stigma/bias secondary to their tobacco-related malignancy. Despite these
survivorship challenges, relatively little effort has been allocated to developing interventions
that address the unique experience of lung cancer survivors and their caregivers.

Method: To expand lung cancer survivorship care, the investigative team developed a
targeted and tailored psychosocial intervention to address the diverse needs of lung cancer
survivors and the challenges faced by caregivers. Principles of motivational interviewing and
shared decision making guided development of the Kentucky LEADS Collaborative Lung Cancer
Survivorship Care program.

Results: Selection of intervention content was based on a review of the literature, input from
lung cancer survivorship experts, and feedback from an engaged community advisory board.
During the development phase, the investigative team confronted the challenge of designing
a psychosocial intervention that is feasible to deliver in diverse cancer care settings and would
be acceptable to this underserved and difficult to reach population. To promote
implementation feasibility, the investigative team designed for dissemination, creating an
intervention that was discipline, delivery mode, and setting neutral. To enhance program
acceptability, design focused on common unmet needs, but also incorporated a content menu
that allows lung cancer survivors to select modules aligned with their personal concerns. After
conducting online and in-person interventionist training, program feasibility, acceptability,
and preliminary efficacy are being evaluated in a single-arm trial in 10 Kentucky cancer care
facilities. Approximately 300 lung cancer survivors and caregivers will complete validated
patient-reported outcome measures, including program acceptability assessments.

Conclusion/Implications: Based on results, the investigative team plans to modify the

intervention and training program and then conduct a group randomized trial in collaboration
with the Kentucky Clinical Trials Network.
S738 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 8: Neighborhood and Environmental Influences on Physical Activity 3:00 PM-
3:15 PM

MODERATORS OF OBJECTIVELY MEASURED NEIGHBORHOOD SAFETY AND PHYSICAL ACTIVITY

AMONG ADOLESCENTS

Jessa K. Engelberg, BA1, James Sallis, PhD1, Terry L. Conway, PhD1, Kelli Cain, MA2, Carie
Geremia, N/A3, Larry Frank, PhD4, Karen Glanz, PhD, MPH5, Brian E. Saelens, PhD6

1
University of California, San Diego, San Diego, CA; 2University California San Diego, San
Diego, CA; 3UCSD, San Diego, CA; 4University of British Columbia, Vancouver, BC, Canada;
5
University of Pennsylvania, Philadelphia, PA; 6University of Washington, Seattle, WA

Objective: The relation between neighborhood safety and physical activity (PA) and the
factors that moderate this relation are unclear. The current study examined this among
adolescents. Methods: Participants were aged 12-17 years (n=725) from the Baltimore, MD
and Seattle, WA regions, with 50% male and 33% non-white. Moderate-to-vigorous PA
(MVPA) in the neighborhood was determined through accelerometers and GPS. Objective
neighborhood safety data were collected using a validated observational measure of
streetscapes (MAPS). Survey-reported (from parent and adolescents) moderators at multiple
ecological levels were examined: individual (e.g. gender, BMI), psychosocial (e.g. self-efficacy,
social support), home environment (e.g. electronics ownership, home PA equipment) and
perceived neighborhood safety (e.g., traffic safety, safety from crime). A final multilevel
generalized linear regression model identified main and cross-level moderator effects.
Results: There were 7 main effects (pConclusions: Main effects or moderators were identified
from all tested levels of the ecological model, supporting the ecological approach and
implying that multi-level interventions are promising. Findings suggest that living in an
objectively safer neighborhood may benefit White non-Hispanics and help those with low
social support or moderate perceived environmental barriers overcome these limitations and
engage in PA in the neighborhood.

CORRESPONDING AUTHOR: Jessa K. Engelberg, BA, University of California, San Diego, San
Diego, CA, 92103; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S739

Paper Session 8: Neighborhood and Environmental Influences on Physical Activity 3:16 PM-
3:30 PM

FIVE-YEAR POST-INTERVENTION FOLLOW-UP OF PHYSICAL ACTIVITY AND ASSOCIATIONS WITH

THE BUILT ENVIRONMENT

Heather J. Leach, PhD1, Daniel P. O'Connor, PhD2, Richard J. Simpson, PhD3, Hanadi Rifai,
PhD3, Scherezade K. Mama, DrPH4, Rebecca E. Lee, PhD5

1
Colorado State University, Fort Collins, CO; 2University of Houtson, Houston, TX; 3University
of Houston, Houston, TX; 4The Pennsylvania State University, State College, PA; 5Arizona State
University, Phoenix, AZ

Background: Few studies have examined long-term maintenance of physical activity (PA) after
completion of a PA intervention. Ecologic models of PA posit that extra-individual factors,
such as the built environment, have a direct influence on PA, potentially extending the success
of behavioral interventions. This study examined PA five-years after completion of a
community-based, randomized controlled trial (RCT) and the association of built environment
characteristics with PA changes. Method: African American women who completed the RCT
were eligible to participate. The International PA Questionnaire (IPAQ) measured PA at
baseline, post-intervention and five-year follow up. Built environment characteristics included
the presence and quality of PA resources, neighborhood walkability, and traffic and crime
safety. Paired t-tests examined changes in PA. Stepwise linear regression models examined
associations between built environment variables and PA change scores (five-year follow-up
minus post-intervention PA). Results: Of the 140 women who completed the RCT, 21.4%
(N=30) participated in the follow-up study. The majority (39%, N=55) were lost due to
outdated contact information. From post-intervention to five-year follow-up, walking (p=.008,
=-461.9 MET-min/week) and leisure-time (p=.015, =-154.4 MET-min/week) PA decreased.
There were no significant changes in moderate, vigorous or total PA, but trends showed that
moderate PA increased, vigorous PA decreased and total PA remained the same. When built
environment variables were added to regression models, they explained a greater proportion
of the variance in PA change, (R2 increase range = 002-.478, all p>.05). Conclusions: Long-term
maintenance of PA following an intervention may depend on the domain or type of activity.
Future studies should include objective measures of PA and plan a priori for long-term follow-
up by ensuring up-to-date participant contact information is maintained. Further research on
PA maintenance is needed, and should utilize more advanced statistical techniques to explore
individual and psychosocial factors as mediators or moderators of built environment influence
on long-term PA.
S740 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Heather J. Leach, PhD, Colorado State University, Fort Collins, CO,
80523; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S741

Paper Session 8: Neighborhood and Environmental Influences on Physical Activity 3:46 PM-
4:00 PM

TAILORED INTERVENTION CHANGES WALKING BEHAVIOR AND PERCEIVED LOCAL

NEIGHBORHOOD SIZE

Katie M. Heinrich, PhD1, Joseph Lightner, MPH2, Kara Saiki, MPH3, Cheryl Albright, PhD, MPH4

1
Kansas State University, Manhattan, KS; 2University of Kansas Medical Center, Kansas City,
MO; 3University of Hawai`i at Manoa, School of Nursing and Dental Hygiene, Honolulu, HI;
4
Univeristy of Hawaii at Manoa, Honolulu, HI

Various neighborhood sizes have been studied for health outcomes. Perceived local (close to
home) neighborhoods are affected by physical, social, and personal characteristics including
daily activities like walking. It is unknown how a perceived neighborhood area might change
over time. This study examined changes in perceived neighborhood size before and after a
moderate-to-vigorous physical activity (MVPA) intervention for new mothers.

In the Na Mikimiki study, conducted 2008-2012 in Honolulu, HI, healthy but underactive new
mothers (N = 278; mean age = 32 6 years, 84% ethnic minority) with infants 2-12 months at
entry were randomized to either a standard, information only (comparison) condition that
included MVPA print/website materials or a tailored (intervention) telephone counseling and
eHealth technology condition, to promote MVPA, largely walking. Baseline and 12-month
measures included sociodemographics and the Active Australia survey (with specific walking
questions). Participants drew an enclosed boundary around their home address to indicate
their neighborhood at each time point. If they reported walking in their neighborhood, they
also drew in their walking route. ArcGIS 10.4 and SAS 9.3 were used for analysis.

Complete data were available from 124 women (comparison = 68, intervention = 56). Minutes
walked/week significantly increased from baseline to 12-months in the comparison group (57
71-min to 88 134-min; t = 6.28, p < 0.001), and the intervention group (54 66-min to 141
150-min; t = 3.43, p = 0.001). Average neighborhood size decreased significantly from 1,057-
km2 to 1,025-km2 in the comparison group (t = 6.49, p < 0.001) and increased significantly
from 557-km2 to 573-km2 in the intervention group (t = 6.05, p < 0.001). Neighborhood
walking route distance did not change in the comparison group (p = 0.80) but significantly
increased from 2.5-km2 to 4.3-km2 in the intervention group (t = 5.09, p < 0.001). Intervention
main effects were significant for change in neighborhood size (t = 1.93, p = 0.05).
S742 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Walking increased for all participants, but 107% more for the tailored intervention group.
Accordingly, average neighborhood size and the neighborhood walking route distance
significantly increased in the intervention group. Researchers should consider that perceived
neighborhood size increases for women when they walk more in their neighborhoods.

CORRESPONDING AUTHOR: Katie M. Heinrich, PhD, Kansas State University, Manhattan, KS,
66506; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S743

Paper Session 8: Neighborhood and Environmental Influences on Physical Activity 4:01 PM-
4:15 PM

MEDIATORS OF A TWO-YEAR PHYSICAL ACTIVITY INTERVENTION FOR CHURCHGOING LATINAS

PARTICIPATING IN FE EN ACCIN/FAITH IN ACTION

Elva M. Arredondo, PhD1, Jessica Haughton, MPH, MA2, Guadalupe X. Ayala, Ph.D. MPH1,
Donald J. Slymen, PhD1, James Sallis, PhD3, Lilian Perez, MPH4, Natalicio Serrano, MPH5, John
Elder, PhD, MPH1

1
San Diego State University, San Diego, CA; 2San Diego State University/UC San Diego Moores
Cancer Center Partnership Institute for Behavioral and Community Health, San Diego, CA;
3
University of California, San Diego, San Diego, CA; 4UC San Diego/San Diego State University,
San Diego, CA; 5Washington University at St. Louis, San Diego, CA

Objectives: Hispanic/Latina women are less likely to engage in leisure time physical activity
than women from other ethnic/racial backgrounds. Faith-based organizations are a promising
setting for promoting preventive health behaviors among Latinas. This cluster randomized
controlled trial evaluates an intervention involving community health workers (CHWs)
targeting different levels of the ecological model to promote physical activity.

Design: Sixteen churches were randomly assigned to a physical activity intervention or cancer
screening comparison condition and about 27 women per church were recruited to serve on
an evaluation cohort.

Measurements: We enrolled 436 Latinas (aged 18-65 years) who engaged in less than 250
min/wk of moderate to vigorous physical activity (MVPA) at baseline as assessed by
accelerometer. At baseline, 12- and 24-months follow-up, we collected accelerometer
assessed MVPA, socio-demographic data, self-reported leisure time MVPA, and potential
mediator variables like strategies for engaging in physical activity and social support for
physical activity.

Results: After adjusting for socio-demographic factors, findings of the mixed effects analyses
suggested significant increases in self-reported leisure time MVPA (p
S744 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Summary: Encouraging participants to buy exercise equipment, seek social support, and set
goals for physical activity may be key areas to target in future physical activity programs.
Current study findings suggest that a faith-based intervention led by CHWs can be a culturally
sensitive approach to promoting physical activity and other health behaviors among Latinas.

CORRESPONDING AUTHOR: Elva M. Arredondo, PhD, San Diego State University, San Diego,
CA, 92123; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S745

Paper Session 8: Neighborhood and Environmental Influences on Physical Activity 4:16 PM-
4:30 PM

CITATION AWARD WINNER

ENVIRONMENTAL AND INDIVIDUAL INFLUENCES ON CHANGE IN ACTIVITY AMONG NON-

ATHLETIC MIDDLE-SCHOOL STUDENTS; GENDER DIFFERENCES

Margaret Schneider, Ph.D.

University of California, Irvine, Irvine, CA

Background: Adolescence is characterized by a general trend toward declining physical

activity. By the time adolescents enter high school, approximately 57% of females and 38% of
males do not meet current recommended levels of moderate-to-vigorous physical activity
(MVPA). Factors at multiple levels (i.e., physical environment, motivations to exercise, and
physiological characteristics) are related to MVPA, but little is known about the relative
influence of these different factors on change in MVPA over time among middle-school
students.

Methods: Middle-school students (N = 127, 49% male, 48% Latino) who reported not
engaging in organized sport at the time of recruitment completed assessments of motivations
to exercise (MPAM-R), body mass index and cardiorespiratory fitness (VO2peak; graded cycle
ergometer) in the fall of their 6th-grade year. Parent/guardians reported their perceptions of
neighborhood safety in the fall. MVPA (ActiGraph) was assessed in the fall and again in the
spring. Hierarchical regression analysis examined the independent associations of each
influencing factor on change in MVPA over time for males and females separately.

Results: At baseline, mean daily minutes of MPVA were 51.83 (SD = 20.76) for boys and 38.66
(SD = 14.80) for girls. The mean change in MVPA from fall to spring was 2.69 (SD=18.67) for
boys and -.38 (SD = 10.94) for girls. Controlling for baseline MVPA, neighborhood safety
predicted spring MVPA among boys only (B = -4.19, p< .05). Greater perceptions of
neighborhood threats (e.g., gangs, traffic, danger) were associated with decreased MVPA
from fall to spring. Among females, none of the influencing factors significantly predicted
change in MVPA.

Conclusions: At baseline, boys were engaging in greater MVPA than girls, and this difference
persisted into the spring. However, the change in activity from fall to spring was more
variable among boys than among girls. Moreover, neighborhood safety predicted change in
activity among boys and none of the individual-level factors predicted change in activity
S746 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

among either boys or girls. Among middle-school boys who do not participate in organized
sports, neighborhood safety has a stronger influence on activity levels over the school year
than intrinsic motivation to exercise, body composition, or cardiorespiratory fitness.

CORRESPONDING AUTHOR: Margaret Schneider, Ph.D., University of California, Irvine, Irvine,

CA, 92617; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S747

Paper Session 9: Understanding Obesity: Cells to Society 3:00 PM-3:18 PM

CITATION AND MERITORIOUS AWARD WINNER

ROLE OF METABOLIC HORMONES IN THE OBESITY-DEPRESSION-INFLAMMATION LINK

Yesenia Cabrera, B.S.1, Tiefu Cheng, M.S.2, Nicholas Guay-Ross, B.S.2, Suzi Hong, Ph.D.2

1
University of California Los Angeles, Los Angeles, CA; 2University of California San Diego, La
Jolla, CA

Obesity is marked by a state of chronic inflammation as well as metabolic dysregulation.

Knowledge in the underlying mechanisms of obesity-related chronic inflammation is
incomplete, perpetuating limited options for efficacious therapeutics, including behavioral
interventions. With up to 50% of overweight or obese individuals reporting clinically relevant
depressive symptoms, challenges in implementing health behavior interventions are only
compounded. Depression and obesity both present with elevated inflammation. Adipose
tissue acts as an endocrine organ by secreting inflammatory cytokines and adipokines such as
leptin. We previously reported that dysregulation of inflammatory processes mediated by
glucocorticoid receptors (GRs) underlies both obesity and elevated depressive symptoms. In
the current study we investigated the role of leptin and insulin resistance in the obesity-
depression-inflammation link among 129 asymptomatic men (49.6%) and women (50.4%) of
18 65 years of age. Depressive symptoms were measured using Beck Depression Inventory
(BDI). Fasting leptin and insulin levels in plasma were measured using a multiplexed
immunoassay. Homeostatic model assessment was calculated to indicate insulin resistance
(HOMA-IR). Thirty-one, 40% and 30% of the participants were lean, overweight and obese,
respectively. The average total BDI score was 5.7 ( 6.2) with a range of 0 - 31. Levels of leptin
and insulin were 19.06 ( 24.04) ng/mL and 342.7 ( 346.4) pg/mL. Mean HOMA-IR was 1.78
( 1.96), indicating normal insulin resistance on average, while 8% and 6% of the sample
showing HOMA-IR > 3 and > 5 exhibited moderate and severe IR, respectively. BDI total and
somatic scores were correlated with BMI (r= .30 for total; .35 for somatic), leptin levels (r= .28,
.34), HOMA-IR (r= .27, .32), and GR-mediated cellular inflammation regulation (r= .45, .56) (ps
.05 to .001). In multivariate models HOMA-IR (= .24, p< .05) remained a significant predictor
for BDI somatic scores even after adjusting for covariates (age, gender, ethnicity, mean blood
pressure) and BMI. However, leptin no longer predicted depression once BMI was entered
into the model. We found that even subclinical insulin resistance is associated with depressive
mood and underlying inflammation, with or without obesity in consideration. The insulin
pathway may be a promising target in scrutinizing the depression-inflammation link in
metabolic and endocrine disorders.
S748 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Yesenia Cabrera, B.S., University of California Los Angeles, Los
Angeles, CA, 90095; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S749

Paper Session 9: Understanding Obesity: Cells to Society 3:19 PM-3:36 PM

FEASIBILITY, ACCEPTABILITY, AND INITIAL OUTCOMES OF A NETWORKED BEHAVIORAL

WEIGHT LOSS PROGRAM

Christine C. Call, A.B., Leah M. Schumacher, M.S., Jocelyn E. Remmert, B.A., Evan M. Forman,
PhD, Meghan L. Butryn, PhD

Drexel University, Philadelphia, PA

There is a critical need to enhance the efficacy of behavioral weight loss (BWL) for obesity.
Enrolling individuals in BWL with members of their social networks (e.g., friends, family
members) may improve adherence to weight control efforts by increasing positive social
influence and decreasing negative social influence related to eating and physical activity. BWL
delivered to social networks (networked BWL) has not been evaluated using a randomized
controlled design, and it is unclear if individuals will be willing to enroll in a networked BWL
research trial knowing that there is a 50% chance that they will be treated without their social
network. The current study examined the feasibility and acceptability of a randomized
networked BWL trial and explored initial outcomes. Overweight or obese adults interested in
a 4-session BWL program were required to recruit two or more overweight or obese network
members who also wanted to lose weight. Index participants (i.e., the individual first
contacting the study) were randomized to receive treatment with or without their recruited
network members. Seventeen index participants and 41 network members enrolled. Nine
index participants were assigned to BWL with their 24 network members. Eight index
participants were assigned to BWL without their network; their 17 network members received
a delayed intervention. Recruitment rates were consistent with other weight loss studies,
demonstrating that index members were willing to recruit network members despite the
possibility of being treated separately. Attendance and retention across groups was high
(>90%), supporting the feasibility of a randomized networked design. Qualitative interviews
and focus groups revealed high treatment acceptability. Ninety-three percent of participants
in networked BWL reported being satisfied or very satisfied with treatment and 92.6%
reported that networked treatment very much or somewhat helped them lose weight.
Weight losses from session 1 to 4 were as follows: index participants in networked BWL=3.5%,
index participants in standard BWL=2.9%, network members in networked BWL=2.4%. Future
research is warranted to determine if weight losses differ across conditions with a larger
sample and greater dose of treatment.

CORRESPONDING AUTHOR: Christine C. Call, A.B., Drexel University, Philadelphia, PA, 19104;
[email protected]
S750 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 9: Understanding Obesity: Cells to Society 3:37 PM-3:54 PM

ASSOCIATIONS BETWEEN HEALTHFUL FOOD PURCHASING AND BMI CHANGE VARY BY AGE

Brad Appelhans, PhD1, Christy Tangney, PhD1, Simone French, PhD2, Lisa Powell, PhD3, Yamin
Wang, MS1, Elizabeth Avery-Mamer, MS4

1
Rush University Medical Center, Chicago, IL; 2Div. Epidemiology & Community Health,
Minneapolis, MN; 3University of Illinois at Chicago, Chicago, IL; 4Rush University Medical
Center, Morton Grove, IL

Household food purchases represent a potentially important source of information about

individuals food preferences and diet quality, and less healthful food purchasing patterns
may be predictive of weight gain. Associations between the overall diet quality of household
food purchases and 12-month change in body mass were examined in a diverse sample of 112
primary household food shoppers [81% female; mean age=43.9 (SD=12.9); 36% African-
American, 35% non-Hispanic white, 16% Hispanic, 13% other/multiracial). Participants
collected and annotated receipts for all household food and beverage purchases (16,087
total) over 14 days. Research staff photographed purchased foods during four visits to
participants homes. The Healthy Eating Index-2010 (HEI-2010) scoring system was applied to
summarize the overall diet quality of food purchases. Body mass index (BMI) was calculated
from height and weight measured during the 14-day food purchase assessment period, and
after 12 months of follow-up (98% retention). Over 12 months, average change in BMI was
0.36 (SD=1.81) kg/m2, with 27% of participants exhibiting clinically meaningful 12-month BMI
change of 1.0 kg/m2 or more. Age was inversely related to observed weight change (r=-.18, p
< .05) such that younger participants demonstrated greater BMI increases. HEI-2010 scores
for food purchases varied widely (M= 60.3, SD=16.2 on a 0-100 scale), but did not differ by age
(r=.04, p=.65). The odds of a BMI gain 1.0 kg/m2 was predicted by an interaction between
HEI-2010 scores and age (p=.03) in logistic regression models adjusting for baseline BMI,
demographic variables, and the presence of children in the household. Every 10-point
increase in HEI-2010 score was associated with reduced odds of demonstrating BMI gain 1.0
kg/m2 among subjects ages 21-34 y (n=37; OR=0.18, 95%CI: 0.04, 0.85), but not in subjects
ages 35-49 y (n=36; p=.80) or age 50 y (n=39; p=.99). The interaction between age and food
purchasing was also observed when BMI change was modeled as a continuous outcome
(p=.003). Food purchasing patterns may be an important intervention target for obesity
prevention in younger adults. Funded by R01HL117804. ClinicalTrials.gov: NCT02073643.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S751

CORRESPONDING AUTHOR: Brad Appelhans, PhD, Rush University Medical Center, Chicago, IL,
60612; [email protected]
S752 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 9: Understanding Obesity: Cells to Society 3:55 PM-4:12 PM

BEHAVIORAL WEIGHT LOSS VS STEPPED MULTI-MODAL TREATMENT FOR BINGE EATING

DISORDER: ACUTE AND LONGER-TERM 18-MONTH OUTCOMES

Carlos M. Grilo, PhD

Yale University School of Medicine, New Haven, CT

Objective: Randomized controlled trial tested the effectiveness of a stepped-care multi-modal

approach versus behavioral weight loss (BWL) for patients with obesity and binge eating
disorder (BED).

Methods: 191 patients (mean age 48, 71% female, 79% white) with BED and co-morbid
obesity (mean BMI 39) were randomly assigned to 6 months of BWL (N=39) or stepped-care
(N=152). Within stepped-care, patients started with BWL for one month; treatment-
responders continued BWL while non-responders switched to specialist treatment (CBT) and
all stepped-care patients were additionally randomized (double-blind) to anti-obesity
medication or placebo for the remaining five months. Independent assessments were
performed at baseline, during treatment, post-treatment (6 months), and 6- and 12-month
follow-ups after completing treatments (through 18 months) with reliably-administered
structured interviews and measures.

Results: ITT analyses of abstinence rates (zero binges/month) revealed BWL and stepped-care
did not differ significantly overall at post-treatment (74% vs 64%) or 12-month follow-up (45%
vs 41%). Mixed-models regression analyses of binge-eating frequency through post-treatment
revealed significant time effects but BWL and stepped-care did not differ overall; within
stepped-care, however, medication was significantly superior to placebo overall and among
initial non-responders switched to CBT. Mixed-models of binge-eating frequency during the
12-months after treatment revealed good maintenance that did not differ across treatments.
Mixed models revealed significant % weight loss through post-treatment but BWL and
stepped-care did not differ overall; within stepped-care, however, medication was
significantly superior to placebo overall and among initial responders continued on BWL and
non-responders switched to CBT. Mixed-models during follow-up revealed significant time
effects with % weight change larger at 6 than 12 months; % weight loss at 12-month follow-up
was 4% (BWL), 6% (BWL+placebo), and 7.5% (BWL+medication).

Conclusions: BWL and stepped-care treatments produced significant improvements in binge-

eating and weight loss that were maintained through 18 months in obese patients with BED.
Anti-obesity medication enhanced outcomes with behavioral treatments within stepped-care.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S753

CORRESPONDING AUTHOR: Carlos M. Grilo, PhD, Yale University School of Medicine, New
Haven, CT, 06519; [email protected]
S754 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 9: Understanding Obesity: Cells to Society 4:13 PM-4:30 PM

SODIUM, POTASSIUM, AND MARKERS OF OBESITY AMONG DIVERSE US HISPANIC/LATINOS:

RESULTS FROM HCHS/SOL.

Tali Elfassy, MS1, Yasmin Mossavar-Rahmani, PhD2, Linda Van Horn, PhD, RD3, Marc Gellman,
PhD1, Daniela Sotres-Alvarez, DrPH4, Neil Schneiderman, PhD1, Martha Daviglus, MD, PhD5,
Jeannette Beasley, PhD6, Maria LLabre, PhD7, Pamela Shaw, PhD8, Guillermo Prado, PhD, MS1,
Hermes Florez, MD PhD7, Adina Zeki Al Hazzouri, PhD7

1
University of Miami, Miami, FL; 2Albert Einstein College of Medicine, Bronx, NY;
3
Northwestern University, Chicago, IL; 4University of North Carolina at Chapel Hill, Chapel Hill,
NC; 5University of Illinois at Chicago, Chicago, IL; 6New York University, New York, NY;
7
University of Miami, MIami, FL; 8University of Pennsylvania, Philadelphia, PA

Introduction: Emerging evidence suggests sodium and potassium are associated with obesity
independent of energy intake. This relationship has been underexplored, especially among
immigrant minority populations with evolving dietary practices exhibited with longer duration
of US residence.

Objectives: We evaluated the independent associations of sodium and potassium with

measures of obesity among US Hispanics/Latinos and tested whether nativity/duration of US
residence moderated these associations.

Methods: The Hispanic Community Health Study/Study of Latinos is a population-based study

of 16,415 diverse Hispanics/Latinos aged 18-74 years from four US communities. Dietary
sodium, potassium, and energy intake were assessed from two interviewer- administered 24-
hour diet recalls in the full sample. Twenty-four hour urine served as the biomarker for
sodium and potassium intake and doubly labeled water for energy intake in the ancillary sub-
study: Study of Latinos Nutrition & Physical Activity Assessment Study (SOLNAS), n=447.
Obesity markers included: measured body mass index (BMI), waist circumference, and body
fat measured using doubly labled water. Nativity/duration of US residence was classified as:
US born, foreign born with 10 years in the US, or foreign born with < 10 years in the US.
Linear regression models adjusted for covariates which included energy intake were used to
determine the associations among sodium, potassium, with markers of obesity.

Results: In the full sample, mean age was 41 years (SE: 0.02); with 53% female (SE: 0.05).
From fully-adjusted linear regression models, a 500 mg/day higher dietary sodium intake was
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S755

associated with a higher BMI (0.07 kg/m2; 95% CI: 0.00, 0.15). Nativity/duration of US
residence moderated the association between dietary potassium and BMI (P < 0.05), so that
higher dietary potassium intake by 500 mg/day was associated with lower BMI among US
born participants (-0.62kg/m2; 95% CI: -0.92, -0.31), lower BMI among those foreign-born
living in the US for 10 years (-0.13 kg/m2; 95% CI: -0.25, -0.00), and with no significant
association among those foreign-born living in the US for < 10 years. Similar associations were
observed for waist circumference. Using biomarkers among SOLNAS participants, a 500
mg/day higher urinary sodium was associated with higher BMI (0.27 kg/m 2; 95%CI: 0.09, 0.43)
and more body fat (0.53 kg; 95% CI: 0.16, 0.91). Urinary potassium was not associated with
BMI or body fat.

Conclusions: Among diverse Hispanics/Latinos, sodium intake was positively associated with
BMI, waist circumference, and body fat, independent of energy intake, suggesting a direct
assocation between sodium intake and markers of obesity. The inverse relationships between
dietary potassium and obesity markers were stronger with longer duration of US residence.

CORRESPONDING AUTHOR: Tali Elfassy, MS, University of Miami, Miami, FL, 33136;
[email protected]
S756 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 10: mHealth Strategies in Women 3:00 PM-3:18 PM

PROMOTING PHYSICAL ACTIVITY IN DYADS OF AFRICAN AMERICAN AND HISPANIC WOMEN

THROUGH SOCIAL NETWORKING AND MOBILE TECHNOLOGY

Larkin L. Strong, PhD, MPH1, Karen Basen-Engquist, PhD, MPH2, Carrie R. Daniel, PhD, MPH3,
Lorna McNeill, PhD, MPH1, Diana S. Hoover, PhD1, Elizabeth Dominguez, MPH3, Israel Christie,
PhD2

1
University of Texas MD Anderson Cancer Center, Houston, TX; 2The University of Texas MD
Anderson Cancer Center, Houston, TX; 3UT MD Anderson Cancer Center, Houston, TX

Racial/ethnic minority adults have lower rates of physical activity (PA) than non-Hispanic
Whites, and women are less active than men. Given the strong influence of family and
friends, intervening upon womens existing social networks may represent an important
strategy for PA promotion. Here we describe preliminary findings from a pilot randomized
trial to test the effect of an intervention designed to build social support and behavioral skills
to increase PA in dyads of women. Dyads were randomly assigned to the 16-week
intervention, which consisted of dyad-based telephone counseling, an activity monitor with
social networking, and health education e-newsletters. Dyads assigned to the control
condition received e-newsletters during the study and the activity monitor upon study
completion. Participants were assessed in-person at baseline and 16 weeks, and completed an
online assessment at 8 weeks. Both objective and self-report measures of PA were collected.
Participants included 62 African American (58%) and Hispanic (37%) women (n=31 dyads)
aged 25-60 years (mean=44 years). At baseline, 79% of women were obese, 76% were not
meeting PA recommendations, 53% reported less than a college education, and 38% reported
an annual household income less than $50,000. From baseline to the 16-week follow up,
intervention and control participants reported an increase in average weekly minutes of
moderate-to-vigorous PA (MVPA) of 116 and 64 minutes, respectively. Preliminary analyses
involved mixed effects models to account for clustering within dyads and within participants
over time and controlled for age, race/ethnicity, education, and baseline body mass index.
Analyses showed that all participants increased their minutes of self-reported MVPA over the
16 weeks (=21.3, se=6.0, p < 0.001), and that MVPA was higher on average among
intervention participants than controls (=104.2, se=48.4, p=0.04). The group by time
interaction, however, did not reach statistical significance (=11.6, se=12.2, p=0.34). Analyses
assessing intervention effects on objective PA are ongoing. Findings provide preliminary
support for an intervention that builds upon womens existing social networks and uses
mobile technology to enhance social support and behavior change skills for increased PA.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S757

CORRESPONDING AUTHOR: Larkin L. Strong, PhD, MPH, University of Texas MD Anderson

Cancer Center, Houston, TX, 77030-3906; [email protected]
S758 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 10: mHealth Strategies in Women 3:19 PM-3:36 PM

COST EFFECTIVENESS OF A WEB-BASED PHYSICAL ACTIVITY INTERVENTION FOR SPANISH

SPEAKING LATINAS: PASOS HACIA LA SALUD

Britta Larsen, PhD1, Bess H. Marcus, Ph.D.2, Dori Pekmezi, PhD3, Sheri J. Hartman, PhD4, Todd
Gilmer, PhD5

1
University of California, San Diego, San Diego, CA; 2University of California, San Diego, La
Jolla, CA; 3University of Alabama at Birmingham, Birmingham, AL; 4Department of Family
Medicine and Public Health, UC San Diego, La Jolla, CA, USA, La Jolla, CA; 5UCSD, La Jolla, CA

Background: Latinas report greater rates of lifestyle-related conditions such as obesity and
diabetes, and concordantly report particularly low levels of physical activity (PA). There is a
need for interventions that can increase PA in this growing population in a broad-scale, cost-
effective manner. Given the scalability of web-based programs and the marked increase in
Internet use in Latinos, web-delivered interventions have great potential as a cost-effective
way to improve health in Latinas. The objective of this study was to examine the costs and
cost-effectiveness of a web-based Spanish language PA intervention for Latinas compared to a
contact control. Methods: Healthy adult Latinas (N=205) were recruited from the community
and randomly assigned to receive a Spanish language web-based individually-tailored PA
intervention (Intervention), or were given access to a website with content on wellness topics
other than PA (Control). PA was measured by 7-Day Physical Activity Recall Interview and
ActiGraph accelerometer at baseline, 6 months (post-intervention) and 12 months
(maintenance phase). Costs were estimated from a payer perspective and included all
features necessary to implement the intervention in a community setting, including staff time
(wage, benefits and overhead), materials, hardware, website hosting, and routine website
maintenance. Results: At 6 months, the cost of the Intervention and Control groups were
$17/month and $8/month, respectively. These fell to $12 and $6 at 12 months. Linear
interpolation showed Intervention participants increased their PA by 1362 total minutes at six
months (523 by accelerometer) compared to 715 minutes in Controls (186 by accelerometer).
At six months, each minute increase in PA for the Intervention cost $0.08 ($0.20 by
accelerometer), compared to $0.07 in Controls ($0.26 by accelerometer). Incremental costs
per minute increase associated with the intervention were $0.08 at six months and $0.04 at
12 months ($0.16 and $0.08 by accelerometer, respectively). Sensitivity analyses showed
variations in staffing costs or intervention effectiveness yielded only modest changes in
incremental costs. Conclusions: While the web-based PA intervention was more expensive
than the wellness control, both were quite low cost compared to face-to-face or mail-
delivered interventions. Cost-effectiveness ranged markedly based on physical activity
measure, and was similar between the two conditions. Overall, the web-based intervention
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S759

was effective and low cost, suggesting a promising channel for increasing PA on a large scale
in this at-risk population.

CORRESPONDING AUTHOR: Britta Larsen, PhD, University of California, San Diego, San Diego,
CA, 92093-0725; [email protected]
S760 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 10: mHealth Strategies in Women 3:37 PM-3:54 PM

ASSOCIATION BETWEEN PHYSICAL ACTIVITY (PA) INTERVENTION WEBSITE USE AND PA LEVELS
AMONG SPANISH-SPEAKING LATINAS

Sarah Linke, PhD, MPH1, Shira Dunsiger, PhD2, Kim M. Gans, PhD, MPH3, Sheri J. Hartman,
PhD4, Dori Pekmezi, PhD5, Tanya Benitez, PhD6, Bess H. Marcus, Ph.D.6

1
UC San Diego, La Jolla, CA; 2The Miriam Hospital and Brown University, Portsmouth, RI;
3
University of Connecticut, storrs, CT; 4Department of Family Medicine and Public Health, UC
San Diego, La Jolla, CA, USA, La Jolla, CA; 5University of Alabama at Birmingham, Birmingham,
AL; 6University of California, San Diego, La Jolla, CA

The Internets low cost and potential for high reach makes web-based channels prime for
delivering evidence-based physical activity (PA) interventions. Despite the well-studied
success of Internet-based PA interventions in primarily non-Hispanic white populations,
evidence about Spanish-speaking Latinas use of such interventions is lacking. The recent rise
in technology use among Latinas in the US, a population at heightened risk for low PA levels
and related conditions, suggests that they may benefit from web-based PA interventions
tailored to their cultural and language preferences. Pasos Hacia la Salud tested a Spanish-
language, culturally adapted, individually tailored, Internet-based PA intervention vs. a
Spanish-lanugage Internet-based Wellness Contact Control condition for under-active Latinas
(N=205, Mage=39.2 10.5, 84% Mexican-American). These analyses examined engagement
with the website and explored how use was associated with adoption and maintenance of PA
behavior change. Overall, participants logged on to the website an average of 22 times
(SD=28) over 12 months, with Intervention participants logging on significantly more than
Controls (29 vs. 14.7, p < .001). On average, participants spent more time on the website at
months 1, 4, and 6 compared to all other months, with max use at month 4. Both logins and
time spent on the website were significantly related to intervention success (achieving higher
mean minutes of MVPA/week at follow-up: b=.48, SE=.20, p=.02 for objectively measured
MVPA and b=.74, SE=.34, p=.03 for self-reported MVPA at 12 months, controlling for
baseline). Furthermore, those meeting ACSM guidelines for PA at 12 months (>=150
min/week of MVPA) logged on significantly more than those not meeting guidelines (35 vs 20
over 12 months, p=.002). Among participants in the intervention arm, goal setting features,
personal PA reports, and PA tips were the most utilized portions of the website. Higher use of
these features was associated with greater success in the program (more minutes of self-
reported MVPA at 12 months controlling for baseline, ps < .05). Specifically, one additional
use of these features/month over 12 months translated into an additional 34 min/week of
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S761

MVPA (goals feature), 12 min/week (PA tips), and 42 min/week (PA reports). These results
demonstrate that greater use of a tailored, web-based PA intervention, particularly certain
features on the site, was significantly related to increased PA levels in Latinas.

CORRESPONDING AUTHOR: Sarah Linke, PhD, MPH, UC San Diego, La Jolla, CA, 92093-0628;
[email protected]
S762 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 10: mHealth Strategies in Women 3:55 PM-4:12 PM

CAMINEMOS JUNTAS: A LOCATION-BASED SMARTPHONE APP FOR LATINAS TO CONNECT

WITH NEARBY WALKING PARTNERS

Valerie H. Myers, PhD1, Lyndsay Fluharty, MA2, Courtney Fultineer, MPH3, Sophia M.
Strickfaden, BA1

1
Klein Buendel, Inc., Golden, CO; 2Telligen, Greenwood Village, CO; 3Colorado Foundation for
Public Health and the Environment, Denver, CO

Health disparities for Latinas are high. They are more likely than their non-Hispanic White
counterparts to be overweight, have diabetes, and be physically inactive. Community-focused
walking interventions produce improvements in physical activity (PA) and are well-received by
Latinas when they are socioculturally sensitive. New technology has the ability to provide
Latinas with innovative ways to connect socially and increase PA. Location-based services
(LBS) are a popular technology that uses geographical positioning to allow individuals to use
their smartphones to connect to their surrounding environment. Caminemos Juntas is an NIH-
study developing a prototype smartphone app for use with 18-45 year old Latinas that uses
LBS to connect women within geographically proximal neighborhoods as a way to provide
social support for increased walking behaviors. The initial phase of the study used multi-
method formative research to guide app design and content prior to conducting field usability
testing. To guide prototype development, a national sample of Latinas (n=98; mean age 32.7
+/- 7.8 years; 45% primary Spanish speaking; 28.6% with annual income < $15,000) were
surveyed to better understand their preferences, usage, needs, and obstacles of current apps
in relation to health and PA. Latinas current PA behaviors and smartphone use, opinions on
health-related apps using LBS, and how often they access social networking sites on their
phone and their likelihood of using a social networking app to connect to others with
intentions to be physically active were also examined. Results revealed that 22.5%
never/rarely exercised, 73.5% accessed social networking sites daily with an average of 8
times a day, and 43.9% used LBS every day. Ease of use (82%), informationally accurate
(79.2%), and reliability (84.7%) were app features rated as highly important. Over 63%
reported high likelihood of using a social networking app to connect to others with the
intentions of being physically active, and 67.4% reported that this type of app would be very
helpful. Focus groups showed that the app was appealing. A month-long field usability study
(n=60) to accurately identify user location and return information to the user to facilitate real-
time decision making for partnered walking is currently underway. Survey and formative
research results as well as the quantitative outcomes from the completed pilot study will be
presented.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S763

CORRESPONDING AUTHOR: Valerie H. Myers, PhD, Klein Buendel, Inc., Golden, CO, 80401;
[email protected]
S764 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 10: mHealth Strategies in Women 4:13 PM-4:30 PM

ENGAGING BREAST CANCER SURVIVORS IN DEVELOPING A TECHNOLOGY-SUPPORTED

PHYSICAL ACTIVITY INTERVENTION: THE FIT2THRIVE STUDY

Siobhan Phillips, PhD, MPH1, Gillian Lloyd, BA1, Margaret R. Moran, MPH1, Danielle Blanch-
Hartigan, PhD, MPH2, David Cella, PhD1, Kerry S. Courneya, PhD3, Linda M. Collins, PhD4,
Frank Penedo, PhD5, Bonnie Spring, PhD5, Susan LeBailly, PhD, PhD6, Ronald T. Ackermann,
MD, MPH5

1
Northwestern University Feinberg School of Medicine, Chicago, IL; 2Bentley University,
Waltham, MA; 3University of Alberta, Edmonton, AB, Canada; 4Penn State, University Park,
PA; 5Northwestern University, Chicago, IL; 6Northwestern U, Chicago, IL

Most physical activity (PA) interventions for breast cancer survivors (BCS) are intense, on-site
and simultaneously test multiple intervention components (i.e. exercise class, coach) designed
by research teams with very limited BCS input. Thus, these interventions have limited patient-
centeredness and scalability. To address this, the present study used a mixed-methods,
patient-centered approach to develop five remotely-delivered, technology-supported PA
intervention components. Following this formative phase, we will examine these components
in conjunction with a core intervention for feasibility, acceptability and preliminary effects
on PA using the Multiphase Optimization Strategy (MOST) in the Fit2Thrive trial. MOST uses a
highly efficient experimental design to evaluate individual and combined effects of
intervention components to determine which ones are, or are not, effective. A nationwide
sample of BCS [n=96; Mage=56 (SD=10.4] completed an online needs assessment. A subsample
was randomly selected to participate in audio-recorded, semi-structured interviews (n=28)
and a subsample of these BCS (n=15) participated in a consensus-building conference call.
Needs assessment data were analyzed using descriptive statistics. Transcribed interviews and
discussions were evaluated by the coding team using a thematic content analysis approach
and consensus review. Several key characteristics of intervention components emerged
including: being user-friendly, accessible, and encouraging; integrating the Fitbit with other
intervention components; providing a progressive PA prescription with measurable progress;
allowing interaction with other users and a clear linkage to evidence-based BCS information.
Based on consensus with BCS, the core Fit2Thrive intervention will consist of a Fitbit,
educational materials and a basic PA tracking mobile application (app). The five key
components identified for testing in the MOST trial are: a deluxe app including BCS specific
content and enhanced feedback, an online gym membership, assignment to a supportive
team of BCS; behavioral coaching calls and motivational text messages. Engaging BCS in the
development of Fit2Thrive intervention components has the potential to result in more
effective components with greater patient acceptance. Having elicited these key intervention
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S765

themes from BCS, we will now test the comparative, interactive, and optimal dose of these
five components in the Fit2Thrive trial using MOST.

CORRESPONDING AUTHOR: Siobhan Phillips, PhD, MPH, Northwestern University Feinberg

School of Medicine, Chicago, IL, 60611; [email protected]
S766 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 11: Approaches to Increase Cancer Screening 3:00 PM-3:18 PM

ACADEMIC-COMMUNITY PARTNERSHIP TO TEST STRATEGIES TO PROMOTE COLORECTAL

CANCER SCREENING AMONG PRIMARY CARE LATINO PATIENTS

Sheila F. Castaneda, PhD1, Balambal Bharti, Staff Res Assoc2, Rebeca Espinoza Giacinto,
MA/MPH, PhD3, Shawne O'Connell, Fatima Munoz, Sylvia Mercado, Marie Elena Rojas, Wendy
Rojas, Gregory Talavera, MD4, Samir Gupta, MD2

1
San Diego State University, San Diego, CA; 2University of California, San Diego, La Jolla, CA;
3
San Diego State University/UC San Diego, Auburn, CA; 4SDSU, Chula Vista, CA

Background: Colorectal cancer (CRC) is the second leading cause of cancer-related death in
the US. Screening for CRC can reduce incidence and mortality, but participation rates remain
low among low-income, unacculturated Latino adults.

Methods: This academic-community clinic partnership conducted two distinct pilot studies
testing the implementation of evidence-based interventions to promote fecal
immunochemical test (FIT) screening among Latino patients aged 50-75 years who were not
up-to-date with CRC screening (n=200) at a large Federally-Qualified Health Center (FQHC) in
the US-Mexico border region of San Diego, CA. One pilot focused on an inreach intervention
including a 30 educational minute session with a patient navigator, review of a flip-chart,
and a take-home FIT kit with instructions. The second pilot was an outreach intervention
consisting of mailed print materials (i.e., FIT kit, culturally and linguistically tailored
instructions, and a pre-paid return envelope). Both groups received patient navigator-based
follow-up to promote FIT completion and referrals for additional diagnostic follow-up and
treatment if needed. The primary outcome was measured by return of FIT kit within three
months follow-up and was assessed using electronic medical records (EMR).

Results: The in-reach pilot consisted of mostly insured (85%), women (82%) and Spanish-
speaking (88%) patients. The outreach pilot consisted of mostly insured (50%), women (64%)
and Spanish-speaking (73%) patients. At three months follow-up, screening completion was
76% for in-reach, and 19% for outreach. The median time for in-reach to have FIT results in
the EMR was 16 days as compared to 46 days for outreach.

Conclusion: These data demonstrate that evidence-based strategies to promote CRC

screening can be implemented successfully within FQHCs, but implementation may require
setting and population-specific optimization. Patient-, provider- and healthcare system
related implementation issues and lessons learned from this study will be discussed.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S767

CORRESPONDING AUTHOR: Sheila F. Castaneda, PhD, San Diego State University, San Diego,
CA, 92107; [email protected]
S768 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 11: Approaches to Increase Cancer Screening 3:19 PM-3:36 PM

AFFECTIVE COMPONENTS OF RISK MEDIATE THE RELATION BETWEEN COGNITIVELY-BASED

PERCEIVED RISK AND COLONOSCOPY INTENTIONS

Lynne B. Klasko-Foster, N/A1, Deborah O. Erwin, PhD2, Lina Jandorf, MA3, Marc T. Kiviniemi,
BA, PhD4

1
State University of New York at Buffalo, Buffalo, NY; 2Roswell Park Cancer Institute, Buffalo,
NY; 3Icahn School of Medicine at Mount Sinai, New York, NY; 4University at Buffalo, Buffalo,
NY

Early detection through routine colonoscopy has been linked to reduced mortality from
colorectal cancer (CRC), but uptake remains low, especially for African Americans as
compared to non-Hispanic whites. Cancer risk processing involves both affective components
(fear, worry) and cognitive components (beliefs about the chances of getting cancer). Previous
work has shown that colonoscopy uptake is related to perceived risk, but less is known about
how these risk components work together to impact decision-making. Guided by the
Behavioral Affective Associations Model (BAAM), this study examines how both affective and
cognitive components of risk influence colonoscopy intentions. 605 African American
participants, ages 50 and older and eligible for colonoscopy screening in the next year
completed surveys assessing affective (cancer worry and fear) and cognitively-based (absolute
and comparative) perceived risk of CRC, future screening intentions, and demographic
characteristics. Regression analysis was used to model both components of risk as predictors
of screening intentions, controlling for demographic variables. Bootstrap estimates for
modeling the indirect effects of affective perceived risk as a mediator of the relation between
cognitive perceived risk and screening intentions were conducted. Both affective and
cognitively-based perceived risk predicted screening intentions when controlling for
sociodemographic variables (all bs>0.14, all ps < 0.05). Cancer fear fully mediated the absolute
risk to intentions relation and both fear and cancer worry mediated the comparative risk to
intentions relation (all indirect effects>0.05, all ps < 0.05). These findings suggest that
affective components of risk play an important role in determining CRC screening intentions
for age-eligible African Americans. Given the potential impact of colonoscopy, educational
strategies highlighting the effectiveness of screening and the risks associated with non-
compliance are necessary to increase health equity. While cognitive risk appraisals also
influence decision-making, this study reaffirms findings of the BAAM that affectively-based
perceived risk influences decision-making by mediating cognitions and should be incorporated
in cancer screening interventions.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S769

CORRESPONDING AUTHOR: Lynne B. Klasko-Foster, N/A, State University of New York at

Buffalo, Buffalo, NY, 14214; [email protected]
S770 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 11: Approaches to Increase Cancer Screening 3:37 PM-3:54 PM

ACCEPTABILITY OF A VIRTUAL PATIENT EDUCATOR TABLET APPLICATION TO INCREASE

CERVICAL CANCER SCREENING KNOWLEDGE AMONG LATINAS

Janna Gordon, M.S. 1, Laura Barnes, PhD2, Gloria Arroyo, N/A3, Patricia Medina-Ramirez,
MPH, CPH4, Sara Proctor, PA, AA5, Kristen Wells, Ph.D., MPH6

1
SDSU/UC San Diego Joint Doctoral Program in Clinical Psychology, San Diego, CA; 2University
of Virginia, Charlottesville, VA; 3San Diego State University Research Foundation, San Diego,
CA; 4University of South Florida, Tampa, FL; 5Catholic Mobile Medical Services, Dover, FL; 6San
Diego State University, San Diego, CA

Latinas experience higher cervical cancer (CC) incidence and mortality compared to the U.S.
general population. Many Latinas lack access to health care and experience literacy,
communication, and knowledge barriers to CC screening. The current study developed a
Virtual Patient Educator (VPE), a Spanish-language, interactive multimedia application in
which an embodied conversational agent provides CC and human papillomavirus education.
Developed using user-centered design, the VPE includes 15 content segments presented by
an animated virtual human using oral explanations accompanied by diagrams. The VPE also
includes 4 segments assessing understanding of the VPE content. Participants not up to date
with recommended CC screening used the VPE and completed a Spanish-language adaptation
of the Unified Theory of Acceptance and Use of Technology questionnaire (UTAUT) and a
demographic survey. The UTAUT measures technology acceptability in the following areas
(scores range from 1 to 5): anxiety, attitude towards technology, perceived enjoyment,
perceived ease of use, perceived sociability, perceived usefulness, and trust. Exploratory
analyses evaluated associations of VPE acceptability with age, education, and employment
status. Patients (N = 32; mean age: 41.3; mean education: 5.4 years; 50% employed)
demonstrated high overall VPE acceptance (M = 3.96), as well as perceived enjoyment (M =
4.34), perceived sociability (M = 4.30), perceived usefulness (M = 4.25), and trust (M = 4.25).
Older patients had less positive attitudes about the VPE and were more anxious about using it
(ps < .05). Education was negatively associated with anxiety and positively associated with all
other subscales (ps < .05). Women who were currently employed rated the VPE as more
enjoyable and more trustworthy compared to unemployed women (ps < .05). In conclusion,
the VPE is accepted by patients and may be used to augment health education, especially
among younger, more educated, and employed Latinas.

CORRESPONDING AUTHOR: Janna Gordon, M.S. , SDSU/UC San Diego Joint Doctoral Program
in Clinical Psychology, San Diego, CA, 92120; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S771

Paper Session 11: Approaches to Increase Cancer Screening 3:55 PM-4:12 PM

RESULTS OF THE VIRGINIA COOPERATIVE EXTENSION 80% BY 2018 INITIATIVE: PROMOTING

COLORECTAL CANCER SCREENING AND RISK REDUCTION

Carlin Rafie, PhD, RD1, Bryan C. Anderson, N/A2, Noelle Voges, MPH3, Lindsay Hauser, MS4,
Lindsay Vander Molen, N/A 5

1
Virginia Polytechnic Institute and State University, Blacksburg, VA; 2Virginia Polytechnic
Institute and State University, Christiansburg, VA; 3University of Virginia Cancer Center,
Charlottesville, VA; 4University of Virginia, Charlottesville, VA; 5Virginia Tech, Blacksburg, VA

Background: Colorectal cancer (CRC) is the third leading cause of cancer-related death in the
U.S, yet is highly preventable and detectable through healthy lifestyles and screening. Virginia
Cooperative Extension (VCE) provides community education in all 107 counties in Virginia.
VCE joined the national 80% by 2018 CRC screening initiative in 2016, and implemented an
internal web-based CRC screening and risk reduction program.

Methods: An online survey informed by the Health Belief Model (HBM) was distributed to all
VCE personnel and the results were used to develop the organization-wide program. A
communications plan was developed to address screening barriers identified in the survey.
Four web-based programs were adapted from an existing Colorectal Cancer Free Zone
(CCFZ) intervention and addressed: 1) screening and insurance coverage, 2) diet, 3) physical
activity and weight management, and 4) personnel testimonials. Sessions were facilitated by
an Extension Specialist and topic expert. VCE CRC champions were recruited from each of 4
state VCE districts to encourage personnel engagement. A program website and Twitter
account with scheduled tweets were created to facilitate communication. Personal action
was encouraged, including signing a CCFZ pledge.

Results: The initial survey had a 33% response rate (288/ 872) and indicated positive
responses to HBM questions. Respondents expressed negative health beliefs, however, about
barriers to screening (22%) and perceived severity of CRC (26%). Communication efforts
targeted these two areas. The 4 web-based programs were viewed live by 139 participants,
representing 42 out of 107 Virginia counties. An additional 47 individuals viewed program
recordings available on the website. 122 (87.8%) viewers completed post-session
assessments. Of these, 50% signed the CCFZ pledge, 42% took action to get screened, and
65% spoke with someone about colorectal cancer. Actions taken after viewing the sessions
included increased fruit and vegetable intake (52%), decreased red and processed meats
(32%), increased fiber intake (32%), and increased exercise (35%).
S772 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions: A web-based colorectal cancer screening and prevention initiative implemented

within Virginia Cooperative Extension and informed by responses to a survey based on the
Health Beliefs Model resulted in a significant percentage of participants taking actions to get
screened, reduce risk, and tell others about colorectal cancer. Personal action toward
screening and healthy lifestyle change by Extension Agents is essential for effective
community education about colorectal cancer. A program toolkit of the initiative is being
developed for use by Extension Agents to promote colorectal cancer screening and risk
reduction in the communities they serve.

CORRESPONDING AUTHOR: Carlin Rafie, PhD, RD, Virginia Polytechnic Institute and State
University, Blacksburg, VA, 24061; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S773

Paper Session 11: Approaches to Increase Cancer Screening 4:13 PM-4:30 PM

USABILITY AND ACCEPTABILITY OF LUCAS CHOICES: A DECISION AID TO SUPPORT INFORMED

DECISION MAKING ABOUT LUNG CANCER SCREENING

Jamie L. Studts, PhD1, Christina R. Studts, PhD, LCSW2, Richard J. Thurer, MD3, Margaret M.
Byrne, PhD4

1
University of Kentucky College of Medicine, Lexington, KY; 2University of Kentucky College of
Public Health, Lexington, KY; 3Miller School of Medicine University of Miami, Miami, FL;
4
University of Miami, Miami, FL

Purpose: Decision support tools (DST) have been advocated as a vital component of
implementing high quality lung cancer screening (LCS), and patient decision aids (PtDA) have
been integrated as an essential component of federal lung cancer screening policy. Further
use of PtDAs during shared decision making consultations prior to screening constitute a
reimbursable health service. However, there are currently no targeted LCS available in the
literature that have been rigorously evaluated and determined to be evidence-based with
regard to their ability to improve knowledge, clarify values, and facilitate shared decision
making. This study assessed the usability and acceptability of LuCaS CHOICES, a web-based
LCS PtDA/DST to facilitate high quality lung cancer screening decisions, in comparison against
the NCI LCS website.

Methods: This study involved a randomized controlled feasibility and acceptability trial
comparing the LuCaS Lung Cancer Screening (LuCaS) CHOICES decision aid and the NCI
website for LCS. Outcomes addressed several aspects of usability and user acceptability and
were assessed at 2 weeks and 4 months after baseline. Participants (n=50) were recruited
from the Miami, Florida area and from rural areas in eastern Kentucky (Appalachia). Survey
measures administered at 2 week follow up included user ratings of PtDA usability and
acceptability as well as satisfaction ratings for individual components of LuCaS CHOICES.

Results: Fifty participants at increased risk for lung cancer were recruited from Miami (n=25)
and rural Kentucky (n=25). All completed baseline surveys, and 44 individuals completed the 2
week follow up (88% retention rate). Average age of participants was 52.6 years (SD 5.1), and
77% of participants were female. The sample was moderately diverse: 63% White, 34% Black,
and 16% Hispanic or Latino. Ratings of intervention usability, satisfaction and acceptability
were favorable for both websites, but there were no statistically significant differences
between the two PtDA conditions. The LuCaS CHOICES participants trended toward higher
decision facilitation (p=.09). Acceptability ratings for the LuCaS CHOICES risk calculator,
S774 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

question prompt list, and conjoint exercise were also favorable (means ~ 5.0/7.0).

Conclusions: Overall, there was a high level of usability and user satisfaction for both PtDAs.
Participant ratings suggested that both websites provided valuable information to facilitate
informed LCS decision making. Ongoing research will evaluate a range of decision efficacy
outcomes as well as the behavioral impact of these tools on patient-provider communication
and LCS uptake. There is a need for rapid development and rigorous testing of PtDAs to fulfill
the promise of lung cancer screening as a strategy to reduce lung cancer mortality.

CORRESPONDING AUTHOR: Jamie L. Studts, PhD, University of Kentucky College of Medicine,

Lexington, KY, 40536-0086; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S775

Paper Session 12: Innovative Health Communication Methods 3:00 PM-3:18 PM

PROVISION OF TECHNOLOGY TO FACILITATE POST-ARRIVAL MONITORING FOR EBOLA AMONG

TRAVELERS TO THE U.S.

Abbey Wojno, PhD1, Heather Joseph, MPH2, Christine E,. Prue, PhD2, Mihaela Johnson, PhD3,
Peyton Williams, N/A4, Brian G. Southwell, PhD5

1
Centers for Disease Control and Prevention, Decatur, GA; 2Centers for Disease Control and
Prevention, Atlanta, GA; 3RTI International, Research Triangle Park, NC; 4RTI International,
Research Triangle Parl, NC; 5RTI International, Research Park Triangle, NC

Although technology can facilitate symptom monitoring and self-reporting during infectious
disease outbreaks, relevant participants may not have access to such tools. We investigated
that possibility during the 2014-2016 Ebola epidemic after the Centers for Disease Control and
Prevention (CDC) recommended post-arrival monitoring for travelers arriving in the United
States from countries with Ebola outbreaks, including health checks and daily reporting to
public health authorities. To facilitate monitoring, CDC began the CARE+ (Check and Report
Ebola Plus) program to provide travelers with information, a cellular phone, and digital
thermometer. We analyzed three waves of interviews with travelers arriving to John F.
Kennedy and Washington Dulles International Airports in 2015: an airport intercept (n=1,195),
a first telephone follow-up (TFU1) 3-5 days post arrival (n=654), and a second follow-up
(TFU2) 1-2 days before the end of monitoring (n=319).

Approximately 51% of travelers indicated not having access to a thermometer upon arrival.
Travelers with passports from Sierra Leone, Liberia, and Guinea, compared to U.S. passport
holders, were significantly less likely to have a thermometer (p < .001). At TFU1, 98% of
travelers reported having used the CARE+ thermometer.

Roughly 36% of travelers reported not having a working cell phone other than the CARE+
phone provided. Travelers with passports from countries with Ebola outbreaks were less likely
to have a cell phone than U.S. passport holders (p < .001). Travelers reported high use of the
CARE+ phone. At TFU1, 81% of respondents had been contacted by a public health authority
via CARE+ phone; 67% of travelers at TFU1 and 77% of travelers at TFU2 used the CARE+
phone to report daily to the health department. U.S. passport holders reported less CARE+
phone usage compared to those with passports from countries with Ebola outbreaks (p <
.001).

Providing cell phones and thermometers met travelers reported needs and facilitated
monitoring behavior.
S776 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Abbey Wojno, PhD, Centers for Disease Control and Prevention,
Decatur, GA, 30033; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S777

Paper Session 12: Innovative Health Communication Methods 3:19 PM-3:36 PM

USER CENTERED DESIGN OF A REAL-TIME PATIENT SAFETY DASHBOARD

Bryan Gibson, DPT, PhD1, Jorie Butler, PhD2, Leah Yingllng, BS3, David Classen, MD MS4

1
University of Utah, Salt Lake City, UT; 2University of Utah, SALT LAKE CITY, UT; 3University or
Utah Department of Biomedical Informatics, Salt Lake City, UT; 4university of utah, salt lake
city, UT

Background: Nearly one third of patients are harmed in U.S. hospitals annually. Currently,
healthcare organizations are seeking to use Health Information Technology to reduce patient
harm through the early detection of safety problems and alerting of healthcare providers. One
existing initiative is a provider-facing patient safety dashboard that presents a list of safety
events and a related global safety risk score for each patient in a panel. The goal of this
project was to design a patient-facing real-time safety dashboard (My Safety Advisor) that
aims to educate and motivate patients to engage with their care team.

Methods: An interdisciplinary team of a health psychologist, an informaticist and a clinician,

conducted four user-centered design sessions with 32 recently hospitalized patients and their
caregivers between February and June 2016. Each semi-structured, moderated session began
with individuals sharing personal stories of recent hospitalizations and any safety events that
occurred. We then solicited input on the contemporaneous iteration of the interface design.
Design sessions were audio-recorded and transcribed. Transcriptions were loaded into
ATLAS.ti and analyzed for themes relevant to design and functionality of My Safety Advisor.

Results: Themes derived from the user-centered design sessions drove the iterative design of
the patient-facing My Safety Advisor dashboard. For example, it was observed that patients
fail to associate specific safety events with appropriate follow-up questions or corrective
actions (as providers do). To meet this need, the patient-facing dashboard design was
modified to include a table that connects each safety issue with links for Questions you
should ask and Things you can do, instantiating the theme show me the links. Similarly,
the detailed display of an individuals risk over time, originally designed for high numeracy
providers, was simplified to an icon-centric display intended for lower numeracy lay people,
S778 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

thus instantiating the theme keep it simple. Other themes included provide an analogy
and use simple language.

Conclusion: We present the user-centered design of a real-time patient safety dashboard. We

will discuss how this qualitative work led to an experiment of the effect of design features of
the dashboard on anxiety and risk perception and whether the themes derived from this work
may generalize to other presentations of risk-related information to patients.

CORRESPONDING AUTHOR: Bryan Gibson, DPT, PhD, University of Utah, Salt Lake City, UT,
84148; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S779

Paper Session 12: Innovative Health Communication Methods 3:55 PM-4:12 PM

USING SOCIAL MEDIA TO TARGET NICHE POPULATIONS: THE SEARCH FOR CONSUMERS OF
ONLINE DNA INTERPRETATION SERVICES

Tiernan J. Cahill, MA1, Blake Wertz, MA2, Qiankun Zhong, Master3, John Donegan, MBA, MA4,
Rebecca J. Forman, Master's Degree5, Supriya Manot, M.A.6, Andrew M. Parlato, M.A.
Emerging Media Studies7, Tianyi Wu, Master8, Yazhu Xu, master student9, James J. Cummings,
Ph.D.1, Tricia Norkunas, MA, MPH10, Catharine Wang, PhD, MSc11

1
Boston University, Boston, MA; 2Boston University: College of Communication, Boston, MA;
3
Emerging media studies, Boston, MA; 4Boston University, East Orleans, MA; 5Boston
University, Brighton, MA; 6Boston University College of Communication, Newton, MA;
7
Boston University College of Communication, Boston, MA; 8Boston University, Beijing, N/A,
China (People's Republic); 9School of Communication, Boston University, Tilburg, Noord-
Brabant, Netherlands; 10Boston University School of Public Health, Bostom, MA; 11Boston
University School of Public Health, Boston, MA

In order to study users of online third party raw DNA interpretation services a niche
population estimated to include only a fraction of a percent of the US population this study
explored recruitment via social media platforms. Multiple social media platforms were
compared to investigate the viability of distributing surveys via social media. Using a 2x2
factorial design, a survey was distributed on two of the most prominent social media
platforms, Facebook and Twitter, using two different types of advertising structures.
Advertisements were either cost-per-click (CPC) which allowed researchers to target those
who liked or commented about relevant topics such as genetics, or cost-per-conversion
(CPV) which began from the same targeting criteria used in cost-per-click methods but
iteratively re-targeted the survey based on the characteristics of users who submitted
completed questionnaires. Payment occurred either each time an advertisement link was
clicked (CPC) or for each completed survey (CPV). Researchers also distributed the same
survey for free on Reddit, by identifying and posting a link to the survey on relevant sub-
communities (subreddits). A total of 545 surveys were completed over a two-week
distribution period, 100 ($2.50/each) from Facebook CPC, 120 ($2.08/each) from Facebook
CPV, 35 ($7.04/each) from Twitter CPC, 155 ($1.61/each) from Twitter CPV and 135 ($0/each)
from Reddit, making Twitter CPV the most cost-effective paid method. Although Reddit
distribution was free, this organic distribution could not easily be achieved on other platforms
due to systemic features, and there is reason to conclude that this number was relatively
close to saturation while the same cannot be said for advertisement. Platforms differed on
S780 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

participant demographics, yet no significant differences were found within methods using the
same platform. Most notably, 83% of respondents on Facebook and Twitter were female (17%
male), in contrast to 39% respondents from Reddit (61% male). The average age of Reddit
respondents (35 years) was lower than that of respondents from Twitter (53 years) or
Facebook (49 years). Researchers interested in conducting surveys on otherwise hard to
identify populations may want to consider various social media platforms as an effective
means of surveying attitudes and behaviors, mindful of the potential limits and non-
representativeness of the respondents.

CORRESPONDING AUTHOR: Tiernan J. Cahill, MA, Boston University, Boston, MA, 02215;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S781

Paper Session 12: Innovative Health Communication Methods 4:13 PM-4:30 PM

EFFECTIVENESS OF DVD VS. GROUP-INITIATED DIABETES PREVENTION ON INFORMATION

UPTAKE FOR HIGH & LOW HEALTH LITERACY PARTICIPANTS

Cody Goessl, MS, LAT1, Fabio Almeida, PhD1, Wen You, Ph.D.2, Paul A. Estabrooks, PhD1

1
University of Nebraska Medical Center, Omaha, NE; 2Virginia TEch, Blacksburg, VA

Purpose: To evaluate DVD versus group-initiated diabetes prevention to enhance patient

comprehension of educational objectives between high (HHL) & low health literacy (LHL)
patients.

Methods: Participants received evidence-based diabetes prevention content through either a

DVD or in-person group class. The intervention included physical activity, nutrition, and
behavioral components to facilitate a 5 to 10% body weight loss. The content was consistent
between conditions, only delivery channel differed. After baseline assessment & watching a
DVD or attending a class both focused on diabetes prevention, all participants (n=442;
BMI=37.06.7; 52.312.1 years; 68% women, 17% African American) completed teach-back
strategies used to reinforce intervention content. Approximately 51.0% (n=225) and 49.0%
(n=217) were assigned to a DVD & class, respectively. Comprehension was operationalized as
the proportion of participants correctly answering all the questions in each round. More
rounds would reflect lower initial comprehension post-DVD or class. Health literacy was
assessed by the Newest Vital Sign test at baseline. Descriptive & chi-square analyses were
completed for all comparisons.

Results: The proportion of HHL (n=361) and LHL (n=81) participants was similar between DVD
(n=187 HHL; n=38 LHL) and class (n=174-HHL; n=43 LHL) conditions. Overall, the proportion of
HHL participants that answered round 1 of the teach back questions correctly was 23%
compared to 9% of LHL participants (p < 0.01). This pattern of results was consistent for both
DVD participants (24% HHL and 11% LHL; p < .05) and class participants (23% HHL and 7% LHL;
p < .05). In the second round of teach-back questions, for participants who answered round 1
incorrectly, 82% of both HHL and LHL participants answered correctly, these numbers
increased to 100% and 86%, respectively for round 3. Only 2 LHL participants did not answer
all of the questions correctly by round 3.

Conclusion: DVD and class delivery of diabetes prevention materials do not appear to lead to
differential uptake of information. However, LHL participants demonstrated a higher need for
teach-back, yet, even participants with HHL benefited from the strategy. The use of teach-
back as a strategy among diabetes prevention patients may improve knowledge
S782 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

reinforcement. Thus, teach-back may be a valuable component to any health education

intervention.

CORRESPONDING AUTHOR: Cody Goessl, MS, LAT, University of Nebraska Medical Center,
Omaha, NE, 68105; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S783

Paper Session 13: Diabetes Across the Lifespan 3:00 PM-3:18 PM

BENEFITS OF COMMUNAL COPING IN COUPLES WITH TYPE 2 DIABETES LIMITED BY SPOUSE

UNMITIGATED COMMUNION

Vicki Helgeson, PhD1, Howard Seltman, Ph.D.1, Mary Korytkowski, M.D.2, Leslie Hausmann,
PhD3

1
Carnegie Mellon University, Pittsburgh, PA; 2University of Pittsburgh, Pittsbugh, PA; 3VA
Pittsburgh Healthcare System, Pittsburgh, PA

When one person in a couple faces a chronic illness, both members of the couple are likely to
be affected. Increasingly, research has taken into consideration the social context of coping
with chronic disease. Communal coping is one such approach and consists of a shared illness
appraisal (i.e., it is our problem instead of my or his/her problem) and collaboration to
manage the illness (i.e., joint problem solving). Research has linked communal coping to
better patient outcomes. The present study aimed to examine whether a personality trait in
spouses, unmitigated communion, moderated those relations. Unmitigated communion refers
to over-involvement in others problems leading to self-neglect. It is characterized by being
overly nurturant, intrusive, and easily exploitable. The extent to which spouse unmitigated
communion moderated the relation of spouse communal coping to patient outcomes was
examined in a study of 125 couples in which one person was recently diagnosed with type 2
diabetes (mean age = 54 years, 59% white/41% black, 55% male). The study consisted of
separate patient and spouse interviews and a behavioral observation task in which couples
discussed difficulties managing diabetes. Two independent raters coded spouse communal
coping from the behavioral observation task. Results showed that observer ratings of spouse
communal coping were associated with higher patient ratings of marital quality, higher
patient self-efficacy, higher patient medication adherence, and reduced patient distress. The
majority of these relations were moderated by spouse unmitigated communion, such that
relations were reduced or eliminated when spouses were high in unmitigated communion.
Spouses who were high in unmitigated communion also had higher levels of distress, and their
own communal coping was related to fewer benefits in terms of relationship and
psychological health if they scored high on unmitigated communion. These results suggest
that spouse involvement in illness is not beneficial for patients or spouses when spouses score
high on a dispositional variable that reflects intrusion and immersion in relationships.

CORRESPONDING AUTHOR: Vicki Helgeson, PhD, Carnegie Mellon University, Pittsburgh, PA,
15213; [email protected]
S784 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 13: Diabetes Across the Lifespan 3:19 PM-3:36 PM

DAILY STRESS AND SLEEP SATISFACTION IN COUPLES WITH TYPE 1 DIABETES: EVIDENCE OF
BIDIRECTIONAL EFFECTS

Eunjin Lee, Ph.D.1, Cynthia Berg, Ph.D.1, Katherine J.W. Baucom, Ph.D.1, Vicki Helgeson, PhD2

1
University of Utah, Salt Lake City, UT; 2Carnegie Mellon University, Pittsburgh, PA

Sleep and stress play pivotal roles in living a healthy life and are especially important as
individuals cope with a chronic illness. When individuals experience a chronic illness such as
type 1 diabetes in the context of an intimate relationship, the relation between sleep and
stress may be interdependent and experienced as a dyad. We investigated the bidirectional
relations of both daily variation in and average levels of diabetes stress, general stress, and
sleep satisfaction across 14 days among couples in which one member had type 1 diabetes (n
= 122 couples M age=45.48). Patients and spouses completed a diabetes stressor checklist
indicating whether patients experienced diabetes stressors, whether they experienced
general stressors, and rated their own sleep satisfaction. Multilevel modeling using the actor
partner interdependence model indicated that when patients reported having higher than
their average sleep satisfaction, patients diabetes stress and spouses report of patients
diabetes stress were lower the next day. In regards to the effect of sleep satisfaction on
general stress, both actor and partner effects were found at the average and daily level. For
instance, when patients and spouses reported higher sleep satisfaction than their average
level, both their own and their spouses general stress was lower the next day. Further, when
spouses reported patients diabetes stress was greater than their average, patients levels of
sleep satisfaction were higher, yet their own levels of sleep satisfaction were lower. There
may be benefits to patients when their spouses are aware of their diabetes stress. Patients
and spouses higher average levels of daily general stress predicted lower levels of their own
sleep satisfaction. These results suggest the relation between daily stress and sleep
satisfaction is bidirectional, and supports the link between stress and sleep as a dyadic
process among couples coping with type 1 diabetes.

CORRESPONDING AUTHOR: Eunjin Lee, Ph.D., University of Utah, Salt Lake City, UT, 84103;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S785

Paper Session 13: Diabetes Across the Lifespan 3:37 PM-3:54 PM

MERITORIOUS AWARD WINNER

DIABETES RISK PERCEPTION AMONG RACIAL/ETHNIC MINORITY AND IMMIGRANT GROUPS:

DATA FROM THE NATIONAL HEALTH AND NUTRITION SURVEY

Loretta Hsueh, M.A.1, Juan M. Pea, B.A.2, Adam T. Hirsh, PhD1, Mary de Groot, PhD3, Jesse C.
Stewart, PhD1

1
Indiana University-Purdue University Indianapolis, Indianapolis, IN; 2University of New
Mexico, Albuquerque, NM; 3Indiana University School of Medicine, Indianapolis, IN

Individuals from racial/ethnic minority groups are at an increased risk of developing diabetes.
Emerging research shows that immigrant (foreign-born) individuals are also at an increased
diabetes risk, but less is understood about risk perception in this vulnerable group. Our aim
was to examine associations of immigrant status and racial/ethnic minority status with risk
perception for developing diabetes among a nationally representative sample of people
without diabetes.

Respondents were the 7,983 adults (mean age=47 years, 52% female) from NHANES 2011-
2014 who reported no history of diabetes. We coded immigrant status as foreign-born
(n=2,246, 28%) or U.S.-born (n=5,737, 72%). We used the following six NHANES racial/ethnic
categories: non-Hispanic White (n=3,413, 43%), non-Hispanic Black (n=1,722, 22%), Mexican
American (n=864, 11%), other Hispanic (n=731, 9%), non-Hispanic Asian (n=1,015, 13%), and
other/multi-racial (3%). Diabetes risk perception (yes/no) was assessed by the question: Do
you feel you could be at risk for diabetes or prediabetes? Immigrant status and five dummy-
coded variables comparing each racial/ethnic group to the non-Hispanic White group were
simultaneously entered as predictor variables into logistic regression models with diabetes
risk perception as the outcome. Models were adjusted for demographic factors (age, gender,
education, income) and health risk factors, including diabetes risk factors (HbA1c, self-
reported prediabetes diagnosis, family history of diabetes, BMI, CVD, hypertension,
hypercholesterolemia, smoking).

A total of 2,360 (30%) reported perceiving that they were at risk for diabetes. Both immigrant
status and racial/ethnic minority status were associated with diabetes risk perception. Being
foreign-born, versus U.S.-born, was associated with decreased odds (OR=0.72, 95% CI: 0.58-
0.91, p=0.005) of perceiving a risk for diabetes, while being Mexican American (OR=1.80, 95%
CI: 1.36-2.37), non-Hispanic Asian (OR=1.47, 95% CI: 1.12-1.94), and other/multi-racial
(OR=2.17, 95% CI: 1.62-2.90), versus non-Hispanic White, were associated with increased odds
S786 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

of perceiving a risk for diabetes (all ps < 0.05). Our findings suggest that, after controlling for
actual diabetes risk status, foreign-born adults are less likely to perceive that they are at risk
for diabetes than U.S.-born adults. Importantly, lower diabetes risk perception among foreign-
born individuals could result in later detection and poorer management of diabetes, thus
contributing to immigrant diabetes disparities. Finally, further research is needed to examine
the mechanisms underlying the unexpected finding of higher diabetes risk perception among
Mexican American and Asian persons without diabetes.

CORRESPONDING AUTHOR: Loretta Hsueh, M.A., Indiana University-Purdue University

Indianapolis, Indianapolis, IN, 46204; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S787

Paper Session 13: Diabetes Across the Lifespan 3:55 PM-4:12 PM

FAMILY BARRIERS TO DIABETES MEDICATION ADHERENCE

Lindsay S. Mayberry, MS, PhD, Kryseana Harper, M.Ed

Vanderbilt University Medical Center, Nashville, TN

Family members and close friends (FF) perform harmful diabetes-specific actions which
interfere with adults self-management of type 2 diabetes (T2D). Experiencing more diabetes-
specific harmful actions from FF has been consistently linked to less diabetes medication
adherence. However, most identified harmful FF behaviors pertain to diet, exercise, and self-
monitoring of blood glucose. Therefore, we sought to identify medication adherence-specific
harmful FF actions experienced by adults with T2D.

We recruited adults with T2D (n=102) who were receiving care from Federally Qualified
Health Centers (FQHCs) in Nashville, TN to complete a survey and a lab drawn A1c test. We
identified six ways FF may interfere with patients diabetes medication adherence and
participants indicated which they had experienced within the past 30 days. We examined
bivariate associations between the number of harmful FF actions endorsed by participants
and age, gender, race/ethnicity, marital status, depressive symptoms (Patient Health
Questionnaire-2), diabetes duration, insulin status, medication adherence (Adherence to
Refills and Medications Scale for Diabetes) and A1c using non-parametric tests (Spearmans
/Mann-Whitney U). Last, we regressed medication adherence and, separately, A1c on to the
number of harmful FF actions endorsed, adjusted for a priori covariates (age, gender,
race/ethnicity, marital status, depressive symptoms, diabetes duration, insulin status).

Participants reported family priorities made it difficult to take medication regularly (27%), FF
gave annoying medication reminders (21%), FF attempts to help with patient adherence led to
conflict (20%), and FF judged patients for taking diabetes medications (17%). Less frequently,
participants reported FF gave inaccurate information about (6%) or disapproved of (6%)
medications. Fifty-seven percent endorsed at least one diabetes medication-specific harmful
FF action and endorsing more of these actions was associated with younger age (=-.23, p <
.05), higher depressive symptoms (=.22, p < .05), less medication adherence (=-.35, p <
.001), and worse A1c (=.28, p < .01). In adjusted regression models, more FF harmful actions
remained significantly associated with worse medication adherence (=-.58, p < .01) but not
with A1c.

Over half of our patient sample reported recently experiencing one or more harmful FF
actions specific to diabetes medication adherence. Experiencing more harmful FF actions was
associated with more depressive symptoms and worse A1c and independently associated with
worse medication adherence. Family interventions for adults with T2D should attend to these
medication-specific harmful FF actions to support improved patient adherence.
S788 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Lindsay S. Mayberry, MS, PhD, Vanderbilt University Medical

Center, Nashville, TN, 37203; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S789

Paper Session 13: Diabetes Across the Lifespan 4:13 PM-4:30 PM

DEPRESSIVE SYMPTOMS MODERATE LINKS BETWEEN SELF-REGULATION AND TYPE 1

DIABETES MANAGEMENT IN LATE ADOLESCENTS

Deborah Wiebe, MPH, PhD1, Katherine J.W. Baucom, Ph.D.2, Kristin Wolfe, MRC3, Seth Silva,
Bachelor of Arts, Psychology4, Bryce A. Huntbach, Psychology, BS (Expected Fall 2016) 2,
Cynthia Berg, Ph.D.2, Yana Suchy, Ph.D.2

1
University of California, Merced, Merced, CA; 2University of Utah, Salt Lake City, UT;
3
University of Texas, Southwestern Medical Center, Dallas, TX; 4University of California,
Merced, Turlock, CA

Late adolescents and emerging adults have poorer management of type 1 diabetes compared
to younger adolescents and adults. This may be because they have high levels of depressive
symptoms (DS) that undermine the self-regulatory skills and processes necessary to manage
diabetes. Prior research demonstrated that those with lower self-regulatory capacity (i.e.,
problems with executive functions; EF) have poorer management, but that even those with
good EF display daily self-regulation failures (e.g., forgetting to test blood glucose) that disrupt
diabetes adherence on a daily basis. Such regulatory disruptions may be more common and
more damaging for those with DS. The present study used survey and daily diary methods to
examine whether DS: a) are associated with poorer self-regulation (i.e., EF skills and daily self-
regulation) and b) moderate associations between self-regulation and diabetes outcomes
(adherence, blood glucose control). Late adolescents with type 1 diabetes > 1 year were
recruited during their senior year of high school (N=247; 60% female; M = 17.8 yrs old).
Participants completed cognitive testing to obtain objective measures of EF, and surveys to
measure DS and adherence; a blood sample was provided to assay HbA1c as a measure of
glycemic control. Subsequently, participants completed a 14-day diary protocol, reporting on
daily self-regulation failures and adherence through brief surveys at the end of each day. DS
were correlated with poorer EF skills, r= -.22, adherence, r= -.33, and HbA1c, r=.24, ps < .001.
Interactions revealed higher DS were particularly problematic for adherence and glycemic
control among those with lower EF skills, ps < .05. Multilevel modeling of diary data indicated
those with higher DS had, on average, more daily self-regulation failures and poorer daily
adherence over the subsequent 2 weeks, ps < .001. On days participants had higher self-
regulation failures, they also reported poorer adherence, and this association was present
even after the effect of DS on average levels of self-regulation failures was covaried. Further,
DS moderated the daily association of self-regulation failures with poorer adherence,
indicating a stronger association among those with higher DS, t = -2.39, p < .05. Findings
S790 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

identify those who are more at risk for poor diabetes management (high DS, poor EF), and the
potential processes (disrupted daily self-regulation) that may be targeted for intervention.

CORRESPONDING AUTHOR: Deborah Wiebe, MPH, PhD, University of California, Merced,

Merced, CA, 95343; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S791

Paper Session 14: Life After Diagnosis: Planning for Recurrence and End of Life 3:00 PM-3:18
PM

ACCEPTANCE AND COMMITMENT THERAPY FOR CANCER SURVIVORS WITH FEAR OF

RECURRENCE: RESULTS OF A 3-ARM RANDOMIZED CONTROLLED PILOT

Shelley A. Johns, PsyD, ABPP1, Andrea A. Cohee, PhD2, Kathleen Beck-Coon, MD2, Linda
Brown, PhD3, Michelle L. LaPradd, MS MBA4, Patrick Monahan, PhD4

1
Indiana University School of Medicine, Indianapolis, IN; 2Indiana University School of Nursing,
Indianapolis, IN; 3Indiana University School of Medicine, Bloomington, IN; 4Indiana University,
Indianapolis, IN

Fear of cancer recurrence (FCR) is one of the most prevalent, persistent, and disruptive
sources of distress for adult cancer survivors. After completing treatment, 44-56% of survivors
continue to report clinically significant FCR well into disease-free survivorship. FCR is
associated with maladaptive coping and reduced quality of life. Despite being the most
frequently identified unmet supportive care need reported by survivors, few empirically
supported treatments exist for FCR.

The primary objective of this 3-arm randomized controlled pilot was to assess feasibility and
preliminary efficacy of Acceptance and Commitment Therapy (ACT) in reducing FCR and
cancer-related avoidant coping compared to a survivorship education (SE) intervention and to
enhanced usual care (EUC). The ACT and SE interventions were delivered in a group format 2
hours weekly for 6 weeks. EUC was largely self-administered with a packet of readings on
coping with FCR and other common sources of distress during post-treatment survivorship.
Breast cancer survivors (BCS; n=91) diagnosed at an early stage (0-III) who had completed
curative treatment and who reported clinically-significant FCR (Fear of Cancer Recurrence
Inventory-Short Form score 13) were randomized to 1 of the 3 arms in a 1:1:1 ratio.
Feasibility was assessed descriptively with percentage of eligible BCS who enrolled in the trial,
attendance across 6 weekly group sessions (ACT and SE arms), and trial retention (all arms).
Intent-to-treat ANCOVA was used to calculate pair-wise Cohen d effect sizes between the 3
arms on FCR and cancer-related avoidant coping at post-intervention (T2) and 1-month
follow-up (T3), controlling for baseline (T1) scores and covariates.

The majority of participants were White (83.5%) and college educated (64.8%) with an
average age of 58.7 years (SD = 10.7). Of eligible BCS, 61.7% enrolled in the trial. Retention
was 94.5% through T3, and attendance across 6 sessions was 4.95 sessions for ACT and 5.16
sessions for SE with no significant difference between groups (p = 0.47). At T2, ACT was
superior to SE in reducing FCR (d=0.69, p=0.035) and cancer-related avoidant coping (d = 0.66,
p = 0.037). Similarly at T3, ACT was superior to SE in reducing FCR (d = 0.66, p = 0.023) and
S792 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

cancer-related avoidant coping (d = 0.83, p = 0.002). The only significant difference favoring
ACT over EUC was on decreased cancer-related avoidant coping at T2 (d = 0.68, p = 0.046),
with marginal superiority of ACT over EUC on decreased cancer-related avoidant coping at T3
(d = 0.59, p = 0.078). No significant differences between SE and EUC emerged at any time
point.

Findings suggest that ACT is a promising intervention for BCS with clinically-significant FCR and
warrants a fully powered efficacy trial.

CORRESPONDING AUTHOR: Shelley A. Johns, PsyD, ABPP, Indiana University School of

Medicine, Indianapolis, IN, 46278; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S793

Paper Session 14: Life After Diagnosis: Planning for Recurrence and End of Life 3:19 PM-3:36
PM

PERFECTIONISM AND PSYCHOLOGICAL COMORBIDITY AMONG CANCER PATIENTS: A

LONGITUDINAL ANALYSIS

Jose Savard, Ph.D.1, Marie-Hlne Savard, Ph.D.2

1
Universit Laval, Qubec, PQ, Canada; 2CHU de Qubec-Universit Laval Research Center,
Qubec, PQ, Canada

Background: Symptoms experienced by cancer patients (e.g., anxiety, depression, insomnia)

frequently occur in comorbidity, suggesting a shared etiology. Data from the general
population suggests that perfectionism constitutes a possible psychological transdiagnostic
process underlying many cancer-related symptoms. The goal of this study was to investigate
whether perfectionist patients experience psychological symptoms to a greater extent during
the cancer trajectory. Methods: A total of 962 cancer patients (mixed sites) completed self-
report measures at the peri-operative phase (T1), as well as 2 (T2), 6 (T3) 10 (T4), 14 (T5) and
18 (T6) months after. Measures included the Multidimensional Perfectionism Scale (MPS;
baseline only), the Hospital Anxiety and Depression Scale (HADS), the Fear of Cancer
Recurrence Inventory (FCRI), the Multidimensional Fatigue Symptom Inventory (MFI), the
Cognitive Failures Questionnaire (CFQ) and the Insomnia Severity Index (ISI). Results: 24.6% of
the participants showed a clinical level of perfectionism (i.e. score 25 on the MPS Concern
over mistakes subscale) at baseline. This subset of patients consistently reported more severe
symptoms (depression, anxiety, fear of cancer recurrence, fatigue, cognitive disturbances and
insomnia) throughout the 18-month period. Moreover, perfectionist patients were at least
twice as likely as non-perfectionists to report clinical levels of 3 or 4 symptoms concurrently.
Conclusions. These results suggest that perfectionism is a risk factor of multiple cancer-
related psychological symptoms. Perfectionism may represent a critical common underlying
process that could be addressed in psychological interventions to reduce symptom
comorbidity and the overall burden of cancer on patients and society.

CORRESPONDING AUTHOR: Jose Savard, Ph.D., Universit Laval, Qubec, PQ, G1R 2J6;
[email protected]
S794 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 14: Life After Diagnosis: Planning for Recurrence and End of Life 3:37 PM-3:54
PM

EARLY INTEGRATED PALLIATIVE CARE FOR PATIENTS WITH ADVANCED LUNG AND
GASTROINTESTINAL CANCER: WHAT ARE THE KEY ELEMENTS?

Michael Hoerger, PhD, MSCR1, Jennifer S. Temel, MD2, Vicki A. Jackson, MD, MPH3, Elyse R.
Park, PhD, MPH3, William F. Pirl, MD, MPH4, Areej El-Jawahri, MD3, Emily R. Gallagher, RN3,
Teresa Hagan, PhD, RN3, Juliet Jacobsen, MD3, Joseph A. Greer, PhD2

1
Tulane University, New Orleans, LA; 2Massachusetts General Hospital Cancer Center/Harvard
Medical School, Boston, MA; 3Massachusetts General Hospital, Boston, MA; 4University of
Miami Hospital, Miami, FL

Background: Early integrated palliative care (PC) has been shown to improve quality of life
and mood in patients with advanced cancer. Yet, a barrier to dissemination is insufficient data
about the essential components of the care model. The goal of this study was to identify core
elements of integrated PC across the illness trajectory and examine whether variation in the
focus of PC visits was associated with changes in quality of life and mood at follow-up.

Methods: Participants (n = 175) were patients with incurable lung or non-colorectal

gastrointestinal cancer who were randomized to receive early integrated PC in the
intervention arm of an NIH-funded RCT. Patients completed psychosocial surveys at baseline,
12 weeks, and 24 weeks assessing quality of life (Functional Assessment of Cancer Therapy
General, FACT-G), depression (Hospital Anxiety and Depression Scale, HADS; Patient Health
Questionnaire 9, PHQ-9), and anxiety (HADS). Clinicians completed post-visit surveys
identifying the elements they focused on in each palliative care visit. Analyses examined the
proportion of visits with particular areas of focus (overall, initial 3 visits, middle visits, final 3
visits) and whether the proportion of visits-to-date at 12 and 24 weeks with a given focus was
associated with changes in quality of life and mood.

Results: Throughout the illness trajectory, PC visits focused predominantly on two core
elements: coping (64.7%) and symptom management (74.7%). Key coping processes were
counseling, behavioral strategies, and redirecting hope. Symptom management focused most
on pain and fatigue. Palliative care clinicians focused on building rapport in the initial 3 visits,
decision making and advanced care planning in the final 3 visits, and illness understanding
throughout. At 24 weeks, patients who had a higher percentage of their visits focused on
coping experienced more favorable changes in quality of life on the FACT-G (total score: =
.23, p = .006; social well-being: = .21, p = .02; functional well-being: = .17, p = .03) and
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S795

symptoms of depression on the HADS ( = -.27, p = .002) and PHQ-9 (PHQ-9: = -.24, p =
.002).

Conclusions: Focusing on coping and symptom management are two key hallmarks of early
integrated palliative care for patients with advanced cancer. The role of coping in the care
model warrants further attention, as focusing on coping was associated with favorable
changes in quality of life and mood.

CORRESPONDING AUTHOR: Michael Hoerger, PhD, MSCR, Tulane University, New Orleans, LA,
70118; [email protected]
S796 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 14: Life After Diagnosis: Planning for Recurrence and End of Life 3:55 PM-4:12
PM

ADVANCED CANCER PATIENT-CAREGIVER ACCURACY OF AND AGREEMENT ON LIFE

EXPECTANCY AND ASSOCIATIONS WITH ADVANCE CARE PLANNING

Kelly Trevino, PhD, Megan Shen, PhD, Holly Prigerson, PhD

Weill Cornell Medicine, New York, NY

Background. Many advanced cancer patients have an inaccurate understanding of their

prognosis, which is associated with less engagement in advance care planning (ACP). This
study examined patient-caregiver dyads accuracy of and agreement on estimates of patients
life expectancy (LE) and the relationship between agreement and accuracy and rates of ACP.

Method. Data were collected as part of a multi-site longitudinal study of terminally-ill cancer
patients (LE6 months) and their informal caregivers. Patients and caregivers (n=178 dyads)
completed an interviewer-administered survey (median 3.05 months before patients death)
assessing their estimates of patients LE. ACP was assessed through patient-report. Accuracy
was defined as a LE estimate for the patient of 6 months. Concordance was defined as
patient-caregiver dyads agreeing on LE estimates (6 months or >6 months). Logistic
regression analyses were conducted.

Results. Across all dyads, 14.6% of patients and 28.1% of caregivers reported accurate LE
estimates. A majority of the dyads (76.4%) reported concordant LE estimates. Only 9.6% of
dyads reported concordant and accurate LE estimates. Over two-thirds of dyads (66.9%)
reported concordant but inaccurate estimates. Approximately one-fifth (18.5%) of dyads were
discordant with an accurate caregiver estimate; 5.1% were discordant with an accurate
patient estimate. Patients were more likely to have a completed DNR order in concordant-
accurate (OR=10.80, p < .01) and discordant dyads (patient accurate: OR=16.20, p=.01;
caregiver accurate: OR=2.47, p=.03) than in concordant-inaccurate dyads. Patients were more
likely to have a HCP in concordant-accurate (OR=10.74, p=.02) and discordant dyads in which
the caregiver was accurate (OR=2.49, p=.050) than concordant-inaccurate dyads. Patients
were more likely to have a living will in concordant-accurate (OR=6.00, p=.02) and discordant
dyads in which the caregiver was accurate (OR=2.50, p=.04) than concordant-inaccurate
dyads.

Discussion. The majority of dyads share an inaccurate view of LE. Patients are more likely to
engage in ACP when they and their caregiver are concordant and accurate in their LE
estimates than concordant-inaccurate dyads. In discordant dyads, patient accuracy may
promote DNR order completion; caregiver accuracy may support HCP identification and living
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S797

will completion. Results suggest interventions targeting patient and caregiver accuracy of and
agreement on LE may improve ACP rates.

CORRESPONDING AUTHOR: Kelly Trevino, PhD, Weill Cornell Medicine, New York, NY, 10065;
[email protected]
S798 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 14: Life After Diagnosis: Planning for Recurrence and End of Life 4:13 PM-4:30
PM

FEAR OF CANCER RECURRENCE IN PROSTATE CANCER: EXAMINING PATIENT-SPOUSE DYADS

FROM PRE-TREATMENT TO 12 MONTHS POST

Heather L. McGinty, PhD1, Lisa M. Wu, PhD2, Ali Amidi, PhD3, Michael A. Diefenbach, Ph.D.4

1
The Ohio State University & James Cancer Hospital and Solove Research Institute, Columbus,
OH; 2Northwestern University, Feinberg School of Medicine, Chicago, IL; 3Aarhus University
Hospital, Aarhus, Midtjylland, Denmark; 4Northwell Health, Manhasset, NY

Background: Fear of cancer recurrence (FCR) in both cancer survivors and their family
caregivers is associated with reduced quality of life and increased psychological morbidity.
Furthermore, patients and caregivers may influence each other, potentially exacerbating
these problems. Currently, little is known about FCR and its trajectory among prostate cancer
patients and their spouses. This study aimed to investigate: i) the trajectory of FCR over time
from pre-treatment to post-treatment; ii) the influence of partner FCR over time in patient
spouse dyads; and iii) the possible moderating effect of treatment type upon later FCR.
Methods: A total of 71 men diagnosed with early-stage prostate cancer (PC) and their spouses
were followed at baseline prior to treatment (T1), and at two time points after treatment 6
months (T2) and 12 months after baseline (T3). Participants completed a 2-item measure
evaluating FCR on a 5-point Likert scale at each time point. Repeated measures ANOVA
examined differences within the dyads over time comparing longitudinal patterns for patients
and spouses. Separate actor-partner interdependence models (APIMs) were conducted to
examine actor and partner effects on subsequent FCR from T1-T2 and T2-T3, including the
effect of treatment type on later FCR.
Results: On average, both members of the dyads reported mild to moderate levels of FCR
across all 3 time points. Spouses demonstrated a trend for greater FCR at T1 (p = .06) and
reported significantly greater FCR than patients at T2 and T3 (ps > .01). Patient and spouse
FCR were moderately correlated over time (r = .36 at T1; r = .35 at T2; r = .19 at T3). Both
members of the dyads had significant decline in FCR over time (p < .001), with a trend for
patient FCR to decrease at a faster rate compared to spouses (p = .12). APIMs demonstrated
significant actor, but not partner effects for T1-T2 and T2-T3 models. Receiving surgery was
significantly related to less FCR in spouses at T2 (p = .01); surgery was unrelated to FCR at T3
for either member of the dyad.
Conclusions: FCR is common in patient-spouse dyads and gradually declines over the first year
following treatment for early-stage PC. Spouses report greater FCR than patients and tend to
exhibit slower decline in fear over time. Patient and spouse FCR, while moderately correlated,
does not appear to impact one another over time. The impact of surgical intervention appears
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S799

to reduce FCR for spouses more than for the patients. Future research should examine how
FCR relates to other types of distress and quality of life in dyads over time.

CORRESPONDING AUTHOR: Heather L. McGinty, PhD, The Ohio State University & James
Cancer Hospital and Solove Research Institute, Columbus, OH, 43215;
[email protected]
S800 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 15: Food Access and Availability 3:00 PM-3:18 PM

WORKING UPSTREAM TO IMPACT DIETARY BEHAVIOR CHANGE: CHARACTERIZING RURAL

FOOD ACCESS IN REMOTE AREAS

Courtney Pinard, PhD1, Chris Bardenhagen, JD2, Rich Pirog, M.S., B.A.2, Amy Yaroch, PhD1

1
Gretchen Swanson Center for Nutrition, Omaha, NE; 2Michigan State University Center for
Regional Food Systems, East Lansing, MI

Introduction: Diet-related environmental and policy interventions are advantageous at a

population level since individual change is more likely to be facilitated and sustained if the
environment supports healthful food options. The current study characterized rural food retail
and healthy food access in the rural northeast region of Michigan to provide more
information on specific needs within this predominately rural region.

Methods: A 15 county region of rural northeast Michigan was assessed through triangulating
a combination of three secondary data sources, store observations using a customized tool,
and key informant interviews.

Results: Secondary data yielded over 800 retail food establishments, which were categorized
into 15 food retail types (e.g., large chain supermarkets, limited assortment food marts, dollar
stores). In-person observations were conducted at 20 stores, which included small
grocers/convenience stores without gas (n=7), small or mid-sized grocers with gas (n=6), mid-
sized independent retailers (n=5), and limited assortment/food mart/gas stations (n=2).
Ready-to-eat items (e.g., sandwiches available in 90% of the stores) were readily available
across all store types. Across food categories, we found that the larger stores had a greater
variety and higher quality of produce, lean meats and dairy, and whole grains. In addition,
small grocers with gas had slightly more availability of processed snack foods (e.g., chips,
sugary drinks). Key informant interviews (N=10) emphasized the challenges that these
communities faced in terms of healthy food access, such as the drain of the large regional
supercenter and related economic challenges. Interviewees described that many of these
communities have an aging and dwindling population, however, they also have an interest in
supporting locally grown and healthier foods, but conveyed that this may be cost prohibitive
for local families. The need for community-driven solutions and for further engaging and
tailoring nutrition and health education for rural and remote communities was also
highlighted.

Conclusions: This study provides insights into how the rural food environment may shape
dietary behaviors and inform healthy food access strategies, but also describes the
development of methodology and tools relevant for the unique challenges that rural
communities face.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S801

CORRESPONDING AUTHOR: Courtney Pinard, PhD, Gretchen Swanson Center for Nutrition,
Omaha, NE, 68114; [email protected]
S802 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 15: Food Access and Availability 3:19 PM-3:36 PM

MERITORIOUS AWARD WINNER

SMARTPHONE-BASED ASSESSMENT OF FOOD ENVIRONMENT ASSOCIATIONS WITH SUGAR-

SWEETENED BEVERAGE (SSB) INTAKE IN SHANGHAI, CHINA

Jenna Hua, RD, MPH, PhD1, Lindsey Smith, PhD2, Victor Shia, PhD3, Shufa Du, PhD2, Jiajie
Zang, MD, PhD4, Edmund Seto, PhD5

1
Stanford Prevention Research Center, Stanford, CA; 2University of North Carolina, Chapel Hill,
Chapel Hill, NC; 3UC Berkeley, Berkeley, CA; 4Shanghai Municipal Center for Disease Control
and Prevention, Shanghai, Shanghai, China (People's Republic); 5University of Washington,
Seattle, WA

The rising obesity epidemic in China occurs at a time of fast economic development and built
environment changes is characterized by increasing availabilities of energy-dense foods with
added sugars, such as sugar-sweetened beverages (SSBs). As few studies have systematically
quantified and addressed the nutrition transition and changes in the built environment in
China, and how these shifts relate to changes in food environment, diet, physical activity and
obesity, the objective of this paper is to use smartphone-assisted diet and physical activity
assessments and mobility tracking to examine the relationships between food environment
exposures and SSB consumption in rural and urban Shanghai.

120 healthy adults aged 25-40 years with roughly 1:1 rural to urban and male to female ratios.
Participants SSB consumptions were assessed using smartphone-assisted 24-hour recalls on
three consecutive days, and physical activities were assessed via a custom app called CalFit,
which recorded date, time, latitude, longitude, and accelerometry counts at 10-second
intervals. Participants food environment exposures were assessed by extracting food
environment data along participants GPS tracks using a Staypoint algorithm and Google
Places, and three measures were generated including overall number of food establishments,
richness and Shannon Diversity Index. Other covariates were obtained via demographic
surveys and anthropometric measurements. Logistic regression models were used in
examining the relationships between food environment and SSB consumption. Multiplicative
interactions between food environment exposure variables and the several covariates (age,
sex, urbanicity, BMI, smoking, employment and parent status) were also explored in separate
models.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S803

23% of participants were overweight or obese; 33% were smokers; 44% had children, 54% had
college or university degrees, and 52% reported SSB consumption. In general, participants had
approximately 1814 Kcal of physical activities. They were exposed to 186 food establishments
per day, and had an average richness score of 14.5 and Shannon Diversity Index of 2.86.
Individuals food environment exposures were found to influence their consumption of SSBs.
Urban, older (aged 27-40), more educated, male gender and overweight participants had
higher odds of consuming SSBs. Moderation effects were observed depending on participants
weight status, as well as in participants who have children versus without children.

As this is the first and only study utilizing smartphones to assess the associations between
food environment and SSB consumption, more research is needed to replicate current study
findings. Due to the sample size and cross-sectional nature of the current study, future
longitudinal study with larger sample size would improve current study results.

CORRESPONDING AUTHOR: Jenna Hua, RD, MPH, PhD, Stanford Prevention Research Center,
Stanford, CA, 94305; [email protected]
S804 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 15: Food Access and Availability 3:37 PM-3:54 PM

ECONOMIC FACTORS ASSOCIATED WITH DIETARY QUALITY AND MEETING DIETARY

GUIDELINES IN UNDERSERVED NEIGHBORHOODS IN THE SOUTHEAST

Sara Wilcox, PhD, Patricia A. Sharpe, PhD, MPH, Angela D. Liese, PhD, MPH, Brent Hutto,
MSPH

University of South Carolina, Columbia, SC

Background: Geography and economics can influence availability of healthy food, but it is
unclear how these factors influence dietary intake in underserved, financially disadvantaged
neighborhoods. This study examined whether dietary quality and the proportion meeting
dietary guidelines differed by food security, education, and SNAP benefits in a predominantly
African American sample from underserved neighborhoods..

Methods: Participants were recruited from seven census tracts in two South Carolina cities.
Six met the USDA definition of an urban food desert. In-person interviews collected
sociodemographic and food security information (using the 18-item USDA Household Food
Security questionnaire). One unannounced 24-hour dietary recall was conducted by telephone
using Minnesotas NDSR protocols and software. Healthy Eating Index 2010 (HEI-2010) total
scores characterized dietary quality and conformance with federal dietary guidelines.
Participants were also classified as meeting vs. not meeting 2010 Dietary Guidelines for major
foods and macronutrients. Differences in mean dietary quality and percent meeting dietary
guidelines according to food security (low/very low or high/marginal), education (less than
high school, high school, or some college or greater), and receipt of SNAP benefits (yes or no)
were tested with Fishers exact test and ANOVA.

Results: Participants with dietary data (n=465) averaged 52.314.2 years; 92% were African
American, 79% overweight or obese, 37% high school graduates, 32% at least some college,
and 64% received SNAP benefits. HEI-2010 scores averaged 48.813.08 (max possible
score=100). Fewer than 50% of the sample met dietary guidelines for total fat, saturated fat,
carbohydrates, and grains. Fewer than 20% met guidelines for fiber, fruits, vegetables,
transfat, and 50% whole grains. Dietary quality was lower among those in SNAP households
(p=.06), those reporting high school or less education (p=.05), and among those in food
insecure households (p=.04). Fewer food insecure participants met guidelines for 50% of
grains being whole grains (8% vs. 15%, p=.03), but no other significant (p < .05) differences
were found for meeting dietary guidelines by food security, those in SNAP households, or
education.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S805

Conclusion: Within southeastern underserved neighborhoods, economic variables were

related to dietary quality but not meeting dietary guidelines. The lack of association with
meeting dietary guidelines might be attributed to low percentages meeting guidelines overall.

CORRESPONDING AUTHOR: Sara Wilcox, PhD, University of South Carolina, Columbia, SC,
29208; [email protected]
S806 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 15: Food Access and Availability 3:55 PM-4:12 PM

IS THERE A ROLE FOR SMALL FOOD RETAIL OUTLETS IN HEALTHY FOOD ACCESS?

Courtney Pinard, PhD, Leah Carpenter, MPH, Hollyanne Fricke, MPH, Teresa Smith, PhD, Amy
Yaroch, PhD

Gretchen Swanson Center for Nutrition, Omaha, NE

Introduction: One contributing factor to poor diet is an environment in which foods high in
sugar and fats are ubiquitous, coupled with a lower availability of more nutritious foods such
as fruits and vegetables. This study explored the role that smaller food retail outlets (e.g.,
corner stores) can play in healthy food access.

Methods: Qualitative interviews were guided by the Social Ecological Model and were
conducted with experts in the area of nutrition and public health (N=18) as well as with store-
owners at small food retail outlets (N=20) on current practices, challenges, and strategies
implemented. Interviews with content experts from across the United States (U.S.) were
conducted by phone, while store-owner interviews were conducted in-person across the state
of Nebraska. Interviews were recorded and transcribed verbatim, then coded for themes.

Results: Rural storeowners reported key challenges, including a dwindling population and
competition with larger chains in neighboring towns set in a sparsely populated landscape.
Urban storeowners reported challenges, including gang presence and safety concerns in their
communities, which deterred even local customers. Both urban and rural storeowners
described the constraints that manufacturer/distributor agreements place on their
businesses. Content experts described the focus of their work and unique settings in which
they implement interventions. Further, experts conveyed the opportunity small stores provide
to address food access concerns in low resource communities, specifically to meet
consumers where they are. Measurement gaps included difficulty in gaining access to and
working with sales data, limited sensitivity of observation tools (e.g., NEMS), inadequate
validity of secondary data and need for resource intensive primary data collection, and lack of
understanding of shopping behaviors.

Conclusions: This study adds to the growing body of literature about the role that small food
retail outlets can play in healthy food access, in particular the challenges and barriers that
must be addressed. The triangulation of qualitative data from store-owners in Nebraska and
content experts from across the U.S. provides a rich source of information to inform
discussion of best practices and future directions.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S807

CORRESPONDING AUTHOR: Courtney Pinard, PhD, Gretchen Swanson Center for Nutrition,
Omaha, NE, 68114; [email protected]
S808 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 15: Food Access and Availability 4:13 PM-4:30 PM

DOES THE ADOPTION OF THE HEALTHY EATING STANDARDS IMPACT SNACK QUALITY IN
LOCAL AFTER-SCHOOL PROGRAMS?

Meagan Helmick, CHES, MPH1, Abigail Esmond, MS2, Valisa Hedrick, PhD, RD3, Jamie Zoellner,
PhD, RD3, Wen You, Ph.D.4, Jennie L. Hill, PhD5

1
UNMC, Galax, VA; 2Virginia Tech, Middleburg, VA; 3Virginia Tech, Blacksburg, VA; 4Virginia
TEch, Blacksburg, VA; 5University of Nebraska Medical Center, Omaha, NE

Background. More than 8 million children participate in after-school programs (ASPs). In 2011,
the National Afterschool Association adopted the evidence-based Healthy Eating and Physical
Activity Quality Standards (HEPAQS) to address snack quality & physical activity in ASPs.
Although research has indicated promise in the adoption of these policies, little is known
about local adoption, implementation, and effectiveness of HEPAQS. This study occurs under
an ongoing CBPR initiative, as such, community partners responsible for local ASPs
approached researchers to evaluate the planned adoption of HEPAQS in local ASPs. This
allowed the partnership to take advantage of a natural experiment to 1) compare the quality
of snacks served at ASPs sites before & after the HEPAQS policies were implemented, 2)
determine the nutrient quality of snacks being offered, and 3) the impact of non-program
snacks brought in by children or purchased at onsite vending machines. Methods. An
interrupted time series design was used to measure childrens snack observations across all
sites during a pre-policy adoption period (n=531) & a post-policy adoption period (n=412) and
included 3 policy adopting sites and 2 non-policy adopting sites as comparison control.
Trained research staff collected snack type, brand, and amount consumed using the HAAND
tool and a modified quarter-waste method. Analysis on nutrient content was completed in
NDS-R software. Results. Adoption of the HEPAQS among the policy-adoption sites did not
result in better snack quality based on adherence to the standards. Policy-adoption sites were
only meeting 4 of the 9 adopted Healthy Eating standards post-policy, almost all of which
were already being met pre-policy. Although no significant changes were expected among
comparison sites, there was a significant increase in 100% fruit juice consumption and a
decrease in grams of trans-fat from pre- to post-policy data collection. Across all 5 sites,
program snacks were generally healthier than non-program snacks, as program snacks
contained less SSBs, sweets, and snack chips. Conclusions. The adoption of the HEPAQS did
not result in better snack quality thus pursuing additional HEPAQS of implementation, staff
training, and social and program support may be necessary to impact snack quality. ASPs may
improve non-program snack quality by addressing HEPAQS provided for ASPs or onsite
vending machines and guidelines provided to parents regarding non-program snacks.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S809

CORRESPONDING AUTHOR: Meagan Helmick, CHES, MPH, UNMC, Galax, VA, 24333;
[email protected]
S810 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 16: Leveraging Communities to Advance Health Promotion 3:00 PM-3:18 PM

HARNESSING COMMUNITY-ENGAGED CITIZEN SCIENCE TO PROMOTE SAFE ROUTES TO

SCHOOL IN RURAL PARTS OF SANTA CLARA COUNTY, CA

Nicole M. Rodriguez, MS 1, Sandra Jane. Winter, PhD, MHA2, Bonnie Broderick, RD, MPH3,
Alisa Arce, MPH4, Abby C. King, PhD1, Ann Banchoff, MSW, MPH1

1
Stanford School of Medicine, Palo Alto, CA; 2Stanford University School of Medicine, Palo
Alto, CA; 3Santa Clara County, San Jose, CA; 4Santa Clara County Public Health Department,
San Jose, CA

Background: Community-engaged citizen science can promote sustainable local change to

address health disparities and improve population health. By leveraging multi-sectoral
partnerships, this cost-efficient model builds capacity for tailored behavioral health
interventions. Through a partnership between Stanford School of Medicine and the Santa
Clara County Public Health Department (SCCPHD), citizen science has been incorporated into
local Safe Routes to School (SRTS) programs. This community-engaged process has been
employed to unite residents, public and private organizations, non-government agencies, and
researchers in driving the sustainable development of healthier communities. Methods:
Stanford researchers partnered with the SCCPHD and relevant local organizations to develop a
citizen science-engaged SRTS program in two rural schools in Gilroy, CA. The project involved
15 students at a middle school and 7 parents at an elementary school. Participants gathered
qualitative information about neighborhood barriers and facilitators to health and wellbeing
using an electronic tablet. Collectively, participants and partners prioritized neighborhood
issues to address, possible solutions and potential allies. Results: Projects resulted in
increased involvement of community residents in community health advocacy. Measurable
effects include 6 school parent task force meetings, 5 monthly walking Wednesdays, 9
walking/biking encouragement events, 2 meetings with local stakeholders, and 4 community
discussion meetings. In contrast, a neighboring school that did not involve the citizen science
process in their SRTS program resulted in less overall community involvement without task
forces and community discussion meetings, as well as fewer engagement events, with 2
walking/biking events and 2 walking school bus days. Conclusion: Incorporating citizen science
into school-based public health investigations has helped launch community-based programs,
task forces, and resident initiatives to improve walkability and bikeability in school settings.
Citizen science has been shown to be an effective strategy for translating research into
practice through advocacy and policy.

Learning objectives: Design projects that incorporate citizen science into school-based public
health interventions. Develop effective community engagement strategies to involve
community residents in health advocacy.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S811

CORRESPONDING AUTHOR: Nicole M. Rodriguez, MS , Stanford School of Medicine, Palo Alto,

CA, 94304; [email protected]
S812 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 16: Leveraging Communities to Advance Health Promotion 3:19 PM-3:36 PM

RESULTS OF A COMMUNITY-BASED APPROACH TO DECREASE GENDER-BASED PHYSICAL

ACTIVITY DISPARITIES

April Ennis Keippel, MA1, Nicholas Coombs, MS2, Amanda L. Golbeck, PhD3

1
St. Vincent Healthcare, Billings, MT; 2Billings Clinic, Billings, MT; 3University of Arkansas for
Medical Sciences, Little Rock, AR

The Healthy By Design Coalition sought to accommodate gender norms to increase leisure-
time physical activity for women and girls in Yellowstone County, Montana. Baseline data
from the 2010/2011 Community Health Needs Assessment (CHNA) indicated a statistically
significant disparity with 27.2% of women versus 17.2% of men reporting no leisure-time
physical activity in the past month. To address this disparity, the Coalition developed a
community-based approach including community-based participatory research, social
marketing, and policy development.

A gender-tailored intervention with quasi-experimental cluster-randomized intervention trial

was conducted. Groups were randomized to receive the standard, evidence-based Active
Living Every Day program, a 12-week lifestyle approach class designed to increase physical
activity in previously sedentary healthy adults, or the Gender-Tailored Active Living Every Day
class, the standard class with additional focus on the role gender plays on physical activity,
including specific strategies to overcome gender-based barriers to physical activity.

A social marketing campaign was developed to focus on gender-based activities in which

women engaged, but did not consider physical activity. The campaign theme, It all adds up
was designed to focus on accumulated activity. By accommodating existing gender-based
caregiving roles and activities, physical activity could be accumulated without challenging
existing societal expectations for women as caregivers.

Understanding gender was also influential in the development of worksite physical activity
policies. Barriers to activity such as appearance expectations and discomfort sweating in front
of others were taken into account as policies were drafted. The workplace physical activity
policies emphasized incorporating short 5-10 minute physical activity breaks into meetings
such as stretch breaks, icebreakers or other activities which would not require heavy sweating
or special clothing.

In the fall of 2016, a follow-up survey was conducted using the same methodology as the
2010/2011 CHNA. Results from the 2016/2017 CHNA reveal a significant decrease in gender-
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S813

based physical activity disparities with only 19.8% of women and 15.04% of men reporting no
leisure-time physical activity as compared to 27.2% of women and 17.2% of men in 2011.
These results indicate a collaborative community-based approach may lead to a decrease in
gender-based physical activity disparities.

CORRESPONDING AUTHOR: April Ennis Keippel, MA, St. Vincent Healthcare, Billings, MT,
59101; [email protected]
S814 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 16: Leveraging Communities to Advance Health Promotion 3:37 PM-3:54 PM

CULTURALLY-RELEVANT ONLINE EDUCATION EMPOWERS COMMUNITY HEALTH WORKERS TO

REDUCE CANCER RISK AND DISSEMINATE INFORMATION

Katie Cueva, MAT MPH1, Melany Cueva, RN, EdD2, Laura Revels, BA3, Anne Lanier, MD, MPH2,
Mark Dignan, PhD, MPH4

1
University of Alaska Anchorage, Anchorage, AK; 2Alaska Native Tribal Health Consortium,
Anchorage, AK; 3ANTHC, Anchorage, AK; 4University of Kentucky, Lexington, KY

INTRODUCTION

Alaskas community-based primary care providers requested information about cancer for
themselves and their communities that would overcome geographic and economic barriers by
being accessible via the Internet. In response, a research team at the Alaska Native Tribal
Health Consortium developed a successful application to the National Cancer Institute to
create a culturally-respectful distance-delivered cancer education course with, and for,
Alaskas Community Health Aides and Community Health Practitioners (CHA/Ps).

THEORETICAL FRAMEWORK

This intervention is framed by Community-Based Participatory Action Research that honors

Indigenous Ways of Knowing, and is grounded in Empowerment Theory. Empowerment
Theory includes three components of empowerment outcomes: intrapersonal (a change in
the way an individual thinks of themself), interactional (shifts in understanding about a given
concern), and behavioral (actions taken to directly influence outcomes).

APPROACH

Nine interactive online cancer education modules were created on topics including Womens
Health, Colorectal Cancer, Men's Health, Cancer and Our Genes, Tobacco, and Cancer Basics.
Each of these modules was reviewed by content experts and a community advisory board of
CHA/Ps and partners. The revised versions were published online through the Alaska
Community Health Aide Program Training Center as CHA/P continuing education. While
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S815

evaluation is iterative and ongoing, 338 module evaluations were collected between March
2015 and July 2016.

RESULTS

Similar to employed CHA/P demographics, individuals who provided evaluations were mostly
female (85%) and Alaska Native/American Indian (76%). Evaluation measured the three
components of Empowerment Theory; changes in comfort and confidence in talking about
cancer (intrapersonal), changes in knowledge (interactional), and intent to disseminate cancer
information and reduce cancer risk (behavioral). Preliminary findings include respondents
sharing they would use the information they learned in the cancer education modules with
their patients (85%) and their communities (61%); and that most respondents (80%) reported
the cancer education modules had made a difference in the ways they took care of
themselves, including motivating CHA/Ps to reduce or quit tobacco use, eat healthier, exercise
more, and schedule cancer screening exams.

DISCUSSION

CHA/Ps, and other community health workers, are optimal people to empower individual and
community-level behavior change due to their embeddedness within existing social networks,
and their centrality in each communitys health network. The impact of interventions with
community health workers can change actual, or perceived, social norms, and disseminates
those norms through social networks to promote community-wide health behavior change.

CORRESPONDING AUTHOR: Katie Cueva, MAT MPH, University of Alaska Anchorage,

Anchorage, AK, 99516; [email protected]
S816 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 16: Leveraging Communities to Advance Health Promotion 3:55 PM-4:12 PM

IT'S TIME TO ACT: ASSESSING STUDY DESIGN OF A HEALTH PROMOTION PILOT INTERVENTION
TO IMPACT CRC SCREENING AMONG AFRICAN AMERICANS

Crystal Y. Lumpkins, PhD, MA1, Adam G. Blackstock, Doctorate of Divinity2, W. Todd Moore,
MPS1, Christina Hester, PhD, MPH1, Rayven Smart, Masters of Science in Marriage and Family
Therapy 1, Jannette Berkley-Patton, PhD3, Christine M. Daley, PhD, MA, SM1

1
University of Kansas Medical Center, Kansas City, KS; 2Glory Bible Fellowship International
Church, Lees Summit, MO; 3University of Missouri-Kansas City, Kansas City, MS

Background: Colorectal cancer (CRC) incidence and morbidity disparities among African
Americans may be addressed through innovative health communication and health
promotion programs co-designed and co-delivered by the community. This study details the
feasibility of design, intervention and baseline characteristics of a Community-Based
Participatory Research (CBPR) faith-based pilot intervention to increase CRC screening among
African Americans. Method: A cluster randomized design was employed to assign 4 churches
to an intervention group where they received tailored materials through a socially marketed
communications campaign within the church (Its Time to ACT) and 4 churches to a control
group (non-tailored and non-socially marketed). Church members and community members
who were 45 years or older; not up-to-date with screening and who did not have a family
history of CRC were recruited for the study. All 8 churches were within a bi-state area in the
Midwest and were predominately African American congregations. Participant baseline
characteristics were measured to assess attitude toward screening, attitude toward the
tailored CRC screening communication materials within the campaign, attitude toward the
church co-sponsoring the tailored materials, CRC beliefs and also behavior intention. Primary
outcome was screening completion of an iFOBT (immunochemical fecal occult blood test).
Results: Church members and also community members affiliated with the study churches
were enrolled (N=175). At baseline 77% of the intervention group participants indicated that
they had intentions to complete CRC screening with an iFOBT. Based on 7-point semantic
scales, a majority of this group had a positive attitude toward their church as the co-sponsor
of a tailored CRC post card advertisement and rated it as credible (M=1.3). More than half of
the participants in the intervention (78%) and comparison groups (83%) also had a positive
attitude toward screening for CRC with a fecal occult blood test. Conclusions: Preliminary
results of baseline data show that church members recruited to participate in a CPBR health
communication pilot-intervention study focused on increasing CRC screening behavior among
African Americans is feasible. Employing faith-based organizations to co-develop and co-
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S817

deliver health communications programs within faith-based settings aimed at improving CRC
screening shows promise.

CORRESPONDING AUTHOR: Crystal Y. Lumpkins, PhD, MA, University of Kansas Medical

Center, Kansas City, KS, 66160; [email protected]
S818 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 16: Leveraging Communities to Advance Health Promotion 4:13 PM-4:30 PM

CITATION AWARD WINNER

INTEGRATING CIVIC ENGAGEMENT INTO A BEHAVIOR CHANGE INTERVENTION TO REDUCE

CARDIOVASCULAR DISEASE IN RURAL WOMEN

Rebecca Seguin, PhD, CSCS, Urshila Sriram, MSPH, Meredith Graham, MS

Cornell University, Ithaca, NY

Introduction: Rural women experience poorer health outcomes as compared to their urban
counterparts. These disparities are partly attributable to environmental conditions, including
limited access to healthy foods, opportunities for physical activity, and medical care. Civic
engagement is a potential strategy for improving community environments; however,
research on the effectiveness of this approach in rural contexts is limited. The present study
aimed to evaluate the implementation of a novel civic engagement curriculum (HEART Club)
designed to catalyze positive environmental change in rural communities.

Methods: The HEART Club curriculum was integrated into a six-month community-based
lifestyle modification intervention trial to reduce heart disease risk factors among
overweight/obese rural women in medically underserved communities. HEART Clubs were
implemented with the support of the local health educators running the intervention program
in eight rural towns in Montana and New York. At each site, participants worked to
collectively address an issue related to their local food or physical activity environment and
establish benchmarks for self-monitoring of progress.To assess HEART Club progress, focus
groups (n=8) and interviews (n=8) were conducted with participants and leaders post-
intervention. Leaders also completed an online evaluation survey three months post-
intervention.

Results: The majority of HEART Club groups in the eight towns documented satisfactory
progress and had accomplished two or more benchmarks post-intervention. Among the most
successful groups, facilitators of progress included support from local stakeholders, effective
leadership, and high group morale. Resource constraints, uncertain expectations, and
unfeasible timelines were cited as barriers to HEART Club implementation. Results also
indicate that participants felt overwhelmed by the prospect of community change while
simultaneously trying to make personal health improvements and their initial resistance likely
stalled progress.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S819

Conclusions: These findings highlight both the potential and challenges associated with rural
civic engagement initiatives within the context of a lifestyle intervention program among at
risk, underserved women. Findings can help guide improvements to ensure future success of
this approach.

CORRESPONDING AUTHOR: Rebecca Seguin, PhD, CSCS, Cornell University, Ithaca, NY, 14853;
[email protected]
S820 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 17: Mental and Physical Health of Veterans 3:00 PM-3:18 PM

ASSESSING DEPRESSION IN MILITARY MENTAL HEALTH: A COMPARISON OF THE PHQ-9 AND

PHQ-8

Jagruti Bhakta, PhD1, Vasudha Ram, MPH2, Scott Roesch, Ph.D3, Jennifer Webb-Murphy, PhD2,
Jeffrey Millegan, MD, MPH2, Scott Johnston, CAPT, PhD2

1
Naval Center for Combat & Operational Stress Control (NCCOSC), San Diego, CA; 2Naval
Center for Combat & Operational Stress Control, San Diego, CA; 3San Diego State University,
San Diego, CA

The World Health Organization (WHO) recognizes depression as a global public health concern
(WHO, 2012) and one of the most burdensome diseases in the world (WHO, 2002). It is one of
the leading causes of disability and reduced work productivity. Due to the high prevalence and
impact of this disease, valid scales for the screening and diagnosis of depression are essential
in healthcare. The Patient Health Questionnaire nine item (PHQ-9) scale is commonly used to
assess depression and is valid as both a diagnostic and severity measure. However, the ninth
item of the scale, which asks about suicide, has caused some concerns. Specifically, there is
sometimes apprehension to ask this question in a self-report fashion without immediate
opportunity to assess any positive endorsement. In the military population where increased
attention is being placed on suicide, institutional review boards are especially critical of the
ninth item (Wells, Horton, LeardMann, Jacobson & Boyko, 2013). Some studies have adopted
the PHQ-8, which excludes the ninth item and found it to be valid for assessing depression in
general (Kroenke, Strine, Spitzer, Williams, Berry & Mokdad, 2009) and military (Wells et al.,
2013) populations. The depression screening capabilities of the PHQ-8 versus PHQ-9 have yet
to be examined in a military mental health population.

In order to compare the PHQ-8 with the PHQ-9 and understand their ability to screen for
depression we conducted a retrospective analysis of self-report data from 2854 active-duty
service members (SMs) who completed the PHQ-9 as part of a standardized assessment
program, Psychological Health Pathways (PHP), in military mental health clinics from 2009 to
2012. Univariate analyses determined how often SMs screened positive or negative on both
measures, and how often they screened positive on PHQ-9 but negative on PHQ-8. Kappa
statistics at T1 (entry into PHP) and T2 (10 weeks post-PHP entry) established the degree of
non-random agreement and provided a cross-sectional comparison of performance at both
time points. Kappa statistics also determined level of agreement between PHQ-8 and PHQ-9
to detect clinically significant improvement (a 5-point or greater decrease in score from T1 to
T2). Only 49 (1.7%) of the 2854 SMs screened discordantly (PHQ-9 positive and PHQ-
8 negative) when using the PHQ-8 compared with the PHQ-9. At baseline, 1524 (53.4%) SMs
screened positive for depression on the PHQ-9, and of those, 1475 (96.8%) screened positive
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S821

using the PHQ-8. The PHQ-8 had greater than substantial agreement with the PHQ-9 (k=.97),
and it also compared well when assessing a 5-point or greater increase (k=.91) or decrease
(k=.95) in score from T1 to T2. Our results indicate that the PHQ-8 performs similarly to the
PHQ-9 when assessing depression in military mental health.

CORRESPONDING AUTHOR: Jagruti Bhakta, PhD, Naval Center for Combat & Operational
Stress Control (NCCOSC), San Diego, CA, 92131; [email protected]
S822 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 17: Mental and Physical Health of Veterans 3:19 PM-3:36 PM

VETERANS GROUP EXERCISE: INTEGRATING EXERCISE AND MINDFULNESS TO REDUCE POST-

TRAUMATIC STRESS

Lizabeth Goldstein, PhD1, Wolf Mehling, MD2, Margaret A. Chesney, PhD3, Thomas J. Metzler,
N/A4, Shira Maguen, PhD5, Chris Geronimo, CTRS6, Gary Agcaoili, BA / CTRS / RTC7, Jennifer
Hlavin, MS8, Thomas C. Neylan, MD9

1
San Francisco VA Medical Center, San Francisco, CA; 2University of California San Francisco,
San Francisco, CA; 3UCSF School of Medicine, Menlo Park, CA; 4San Francisco VA Health Care
System, San Francisco, CA; 5San Francisco VA Medical Center and UCSF, San Francisco, CA;
6
SFVAHCS, Belmont, CA; 7VHA VA Northern California Health Care System, Martinez, CA; 8San
Francico VA Medical Center, San Francisco, CA; 9University of California, San Francisco, San
Francisco, CA

Posttraumatic stress disorder (PTSD) is prevalent among military veterans (Ramchand et al.,
2010; Schlenger et al., 1992) and is associated with a number of negative physical health
outcomes, including cardiovascular disease (Ahmadi et al., 2001) and obesity (Bartoli et al.,
2015). Although empirically supported psychotherapies for PTSD have been disseminated in
the Veterans Health Administration (Karlin et al., 2010), only about 10% of veterans attend a
minimally adequate number of mental health appointments in the first year following PTSD
diagnosis (Seal et al., 2010). Veterans with mental health disorders endorse significant
barriers to traditional mental health treatment, including stigma (Hoge et al., 2004),
necessitating the development of alternative treatment options for PTSD. We developed an
integrative exercise (IE) program combining aerobic and resistance exercise, which is familiar
to veterans, with mindfulness-based practices, which are increasingly utilized in military
programs (Crawford et al., 2013). Veterans (N = 47) were randomized to either IE or waitlist
control (WL). The sample was 81% male and 47% racial/ethnic minority. Average duration of
PTSD was 18 years, and 40% of participants were concurrently taking psychiatric medication.
IE was offered in group format at a local YMCA for 3 1-hour sessions weekly for 12 weeks.
Exercise sessions emphasized aerobic exercise, strength training (weights and resistance
bands), and yoga within the context of mindfulness-based principles, such as nonjudgmental
attitude and acceptance. Seventy-six percent of veterans randomized to IE completed the
program and demonstrated an average reduction of 31 points on the Clinician Administered
PTSD Scale, significantly greater than WL (M = 15), d = .90, p = .038. In particular, symptoms
of hyperarousal improved substantially more in the IE group compared with WL, d = .80, p =
.044. There was also greater improvement in quality of life in the psychological domain, (d =
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S823

.53, p = .005) but no relative improvement in quality of life in the physical domain (d = .33, p =
.183), both as measured by the World Health Organization Quality of Life (WHOQOL-BREF)
subscales. This initial study suggests that IE is a novel approach to treating veterans with PTSD
that reduces symptoms of posttraumatic stress and improves psychological quality of life. This
approach to recovery may expand the reach of PTSD treatment into less-stigmatized activities
and settings.

CORRESPONDING AUTHOR: Lizabeth Goldstein, PhD, San Francisco VA Medical Center, San
Francisco, CA, 94121; [email protected]
S824 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 17: Mental and Physical Health of Veterans 3:55 PM-4:12 PM

FIVE-YEAR DIABETES COMPLICATION PROGRESSION LESS SEVERE IN VETERANS WITH MENTAL

HEALTH CONDITIONS

Eric Schmidt, PhD1, Cheng Chen, MS2, James Barnes, MD, MS3, Jessie Wong, PhD2, Alex Harris,
PhD4

1
HSR&D, VA Palo Alto Health Care System, Menlo Park, CA; 2VA Palo Alto Health Care System,
Menlo Park, CA; 3Stanford University, Stanford, CA; 4VA Palo Alto HCS, Palo Alto, CA

Mental health conditions (MHCs) are associated with higher risk for diabetes complications,
but their association with progression of diabetes complications over time is understudied.
Administrative data from the Veterans Health Administration (VHA) was used to identify
patients with new-onset diabetes in 2008 who were alive through 2013. Patients were
included (N = 91,687) if aged 18-75 years at the time of their first recorded diabetes diagnosis,
and if they used any VHA medical care in 2013. The Diabetes Complication Severity Index
(DCSI, score range 0-13) was calculated for two periods of time: (1) two years before diabetes
onset (baseline), and (2) the one-year period starting five years after the date of the
patients first diabetes diagnosis. Change in DCSI score from baseline to five years after
diabetes onset was regressed on baseline socio-demographics, comorbidities, and medical
and mental health care utilization. Having or not having a MHC at baseline was the main
comparison of interest. On average, DCSI scores increased by 0.63 (SD = 1.49) after five years.
Patients with MHCs at baseline (27.3%) were 2.8 years younger, more likely to be female,
unmarried, homeless, and had more medical comorbidities (p < .001 for all). Patients with
MHCs had higher DCSI scores at baseline (mean = 0.76, SD = 1.18) but lower scores five years
later (mean = 1.00, SD = 1.34) than did patients without MHCs (means = 0.37 and 1.15, SDs =
0.88 and 1.48, respectively). Contrary to our hypotheses, having a MHC was independently
associated with less progression of diabetes complications (b = -.39, SE = .02, p < .001) after
controlling for baseline socio-demographics, comorbidities, and health care utilization. More
outpatient visits to a mental health provider before a patients first diabetes diagnosis was
further associated with less of an increase in DCSI score after five years (b = -.005 per visit, SE
= .001, p < .001). Unique aspects of VHA health care delivered to people with MHCs may be
protective against diabetes complications, compared to care delivered to people without
MHCs. Future studies should use longitudinal designs with greater power to model
moderators and mediators of diabetes progression while investigating unique aspects of care
received by Veterans with MHCs and diabetes.

CORRESPONDING AUTHOR: Eric Schmidt, PhD, HSR&D, VA Palo Alto Health Care System,
Menlo Park, CA, 94025; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S825

Paper Session 17: Mental and Physical Health of Veterans 4:13 PM-4:30 PM

PERSISTENT POSTTRAUMATIC STRESS DISORDER AMONG U.S. MILITARY SERVICE MEMBERS

AND VETERANS

Richard F. Armenta, Epidemiologist1, Toni Rush, MPH2, Cynthia A. LeardMann, MPH3, Jeffrey
Millegan, MD, MPH4, Adam Cooper, CPT/PhD5, Charles W. Hoge, MD6

1
Naval Health Research Center/Henry Jackson Foundation, San Diego, CA; 2UCSD/SDSU Joint
Doctotal Program in Epidemiology, Oak Harbor, WA; 3Naval Health Research Center, San
Diego, CA; 4Naval Center for Combat & Operational Stress Control, San Diego, CA; 5Naval
health Research Center, San Diego, CA; 6Walter Reed Army Insitute of Research, Silver Spring,
MD

Objective: Chronic posttraumatic stress disorder (PTSD) can have long-term and far-reaching
impacts on health, social, and occupational functioning. This study examined factors
associated with persistent PTSD among U.S. service members and veterans.

Method: Using baseline and follow-up (20012013) questionnaire data from the Millennium
Cohort Study, multivariable logistic regression was used to determine factors associated with
persistent PTSD. Participants included those who screened positive for PTSD using the PTSD
ChecklistCivilian Version at baseline (n=2409). Participants were classified as having remitted
or persistent PTSD based on screening negative or positive, respectively, at follow-up.

Results: Almost half of the participants (47%) met criteria for persistent PTSD at the first
follow-up; among those, 71% also screened positive for PTSD at the second follow-up.
Multiple factors were independently associated with persistent PTSD in an adjusted model at
the first follow-up, including older age, deployment with high-combat exposure, enlisted rank,
initial PTSD severity, depression, history of physical assault, disabling injury/illness, and
somatic symptoms. Among those with persistent PTSD at the first follow-up, additional factors
of less sleep, separation from the military, and lack of social support were associated with
persistent PTSD at the second follow-up.

Conclusions: Combat experiences and PTSD severity were the most salient risk factors for
persistent PTSD. Comorbid conditions, including injury/illness, somatic symptoms, and sleep
problems, also played a significant role and should be addressed during treatment. The high
percentage of participants with persistent PTSD supports the need for more comprehensive
and accessible treatment, especially after separation from the military.
S826 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Richard F. Armenta, Epidemiologist, Naval Health Research

Center/Henry Jackson Foundation, San Diego, CA, 92126; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S827

Paper Session 18: Innovative Studies on Discrimination and Stress 3:00 PM-3:18 PM

CITATION AWARD WINNER

PERCEIVED WEIGHT, RACIAL, AND SEX DISCRIMINATION AND CARDIOVASCULAR DISEASE IN

UNITED STATES ADULTS

Carlos M. Grilo, PhD1, Tomoko Udo, PhD2

1
Yale University School of Medicine, New Haven, CT; 2SUNY Albany, Albany, NY

Objective: Psychosocial stress has received increased attention as a potential modifiable risk
factor for cardiovascular disease (CVD). Perceived discrimination is increasingly recognized as
a psychosocial stressor with a profound impact on physical and mental health. To date, much
of the research on discrimination in relation to CVD has focused on racial discrimination. The
current study investigated the relationship between three common forms of discrimination
(weight, race, and gender) and 3-year incidence of CVD in a large national sample of U.S.
adults.

Methods: 26,992 adults (55.5% women) who participated in the 2001-2002 and 2004-2005
National Epidemiologic Survey of Alcohol and Related Conditions (NESARC) were included in
this study. Multiple logistic regression analyses were used to calculate odds ratios (OR) and
95% confidence intervals (CI) for three forms of perceived discrimination (simultaneously
included in equations after adjusting for relevant confounds) for predicting CVD incidence at
Wave 2.

Results: Adjusting for sociodemographic variables and other confounders, perceived weight
and racial discriminations were associated with significantly greater likelihood of reporting
myocardial infarction (OR=2.56 [95% CI=1.31-4.98], OR=1.84 [95% CI=1.19-2.84], respectively)
and minor heart conditions (OR=1.48 [95% CI=1.11-1.98], OR=1.41 [95% CI=1.18-1.70],
respectively). Perceived racial discrimination was also significantly associated with greater
likelihood of reporting arteriosclerosis (OR=1.61 [95% CI=1.11-2.34]). Odds of diagnoses for
arteriosclerosis, myocardial infarction, and other minor heart disease were the largest for
individuals reporting multiple forms of discrimination.

Conclusion: Our findings suggest that adults who experience weight and racial discrimination,
and especially multiple forms of discrimination, may be at heightened risk for CVD. Our
findings add to the emerging literature on the detrimental effects of obesity stigma. Perceived
discrimination may be important to consider during assessment of life stressors by health
providers. Future research should address mechanisms linking discrimination and CVD and
S828 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

especially on public health and policy efforts to reduce discrimination which may contribute
to improved health outcomes.

CORRESPONDING AUTHOR: Carlos M. Grilo, PhD, Yale University School of Medicine, New
Haven, CT, 06519; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S829

Paper Session 18: Innovative Studies on Discrimination and Stress 3:19 PM-3:36 PM

MINORITY STRESS, PSYCHOSOCIAL RESOURCES, AND PSYCHOLOGICAL DISTRESS AMONG

SEXUAL MINORITY BREAST CANCER SURVIVORS

Charles Kamen, PhD, MPH1, Jennifer M. Jabson, PhD, MPH2, Karen Mustian, PHD, MPH, MS1,
liz margolies, LCSW3, Luke J. Peppone, PhD, MPH4, Matthew Asare, PhD, MBA, CHES5,
Michelle Janelsins, PhD, MPH1, Ulrike Boehmer, PhD6

1
University of Rochester Medical Center, Rochester, NY; 2University of Tennessee, Knoxville,
TN; 3National LGBT Cancer Network, New York, NY; 4University of Rochester, Wilmot Cancer
Institute, Rochester, NY; 5University of Rochester, Rochester, NY; 6Boston University School of
Public Health, Boston, MA

Background: Few studies have examined unique factors associated with psychological distress
among sexual minority (i.e., lesbian and bisexual) women following breast cancer diagnosis.
The minority stress model posits that heterosexist discrimination increases distress in sexual
minority persons, and that the effect of discrimination could be mediated by positive and
negative psychological factors. The present study assessed the associations between minority
stress, adaptive psychosocial resources, and distress (depression and anxiety) among sexual
minority breast cancer survivors.

Methods: 201 sexual minority women who had ductal carcinoma in situ (DCIS) or stage I-IV
breast cancer participated in this study through the Love/Avon Army of Women (AOW). Self-
report questionnaires were used to assess demographic and clinical factors, depression,
anxiety, minority stress factors (discrimination, minority identity development, outness), and
psychosocial resources (resilience, social support). These factors were included in a structural
equation model testing psychosocial resources as mediators between minority stress and
psychological distress (a latent variable derived from depression/anxiety).

Results: Over half (50.2%) of women in the sample had experienced heterosexist
discrimination; 15% reported elevated depression, and 29% reported elevated anxiety.
Discrimination, negative minority identity, resilience, and social support were correlated with
S830 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

depression and anxiety. While outness and negative minority identity remained positively
associated and social support remained negatively associated with distress in the final
structural equation model, resilience fully mediated the association between discrimination
and psychological distress. The final model predicted 56% of the variance in the latent variable
of psychological distress and demonstrated acceptable fit across all sexual minority women,
2 = 24.50, p > 0.05; confirmatory fit index = 0.99, root-mean-square error of approximation =
0.02.

Conclusions: Minority stress factors are associated with higher rates of psychological distress
among sexual minority women who are breast cancer survivors, while psychosocial resources
are associated with reduced distress. Exposure to discrimination is common in this population;
however, development of resilience mediates the relationship between discrimination and
distress. Enhancement of resilience could be a promising aim of interventions to improve
psychological well-being among sexual minority cancer survivors.

CORRESPONDING AUTHOR: Charles Kamen, PhD, MPH, University of Rochester Medical

Center, Rochester, NY, 14614; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S831

Paper Session 18: Innovative Studies on Discrimination and Stress 3:37 PM-3:54 PM

RACIAL DISCRIMINATION & ALLOSTATIC LOAD AMONG AFRICAN AMERICAN WOMEN:

ROUTINE VS. NON-ROUTINE EXPERIENCES OF DISCRIMINATION

Marilyn Thomas, MPH1, Eli Michaels, BA2, Alexis Reeves, MPH2, Okoye Uche, BS3, David Chae,
ScD4, Amani Nuru-Jeter, PhD2

1
University of California Berkeley School of Public Health, Redwood City, CA; 2University of
California Berkeley School of Public Health, Berkeley, CA; 3University of California Berkeley
School of Public Health, Carson, CA; 4Auburn University, Auburn, AL

Objective: To examine the association between racial discrimination and allostatic load (AL),
an index of physiologic functioning across biological systems, among midlife African American
women, and to assess potential variation in routine vs. non-routine experiences of
discrimination.

Methods: Data are from the African American Womens Heart & Health Study, which
recruited a community sample of 208 African American women ages 30-50 residing in the San
Francisco Bay Area. The Experiences of Discrimination scale was used to measure non-routine
racial discrimination (e.g., in housing, medical care, criminal justice contexts). Routine racial
discrimination was assessed using the 10-item Everyday Discrimination Scale (e.g., being
treated with less courtesy or respect due to race). AL was measured as a composite score
consisting of 15 biomarkers representing cardiovascular, inflammatory, neuroendocrine, and
metabolic function. Multivariate analyses were adjusted for socioeconomic indicators and
other known confounders.

Results: Compared to those reporting very high levels of discrimination (reference), there was
a significant positive association between non-routine racial discrimination (EOD) and AL
among those reporting no experiences of discrimination (=1.45, SE=.64, p

Conclusion: Findings raise methodological questions about how racial discrimination is

conceptualized and measured. Results suggest that reporting non-routine, major life
experiences of racial discrimination, is associated with AL but not in the expected direction.
Conversely, routine everyday experiences show no effect. Being able to recognize and report
racial discrimination in more major, domain-specific life situations may have self-protective
properties. This may be the case particularly in light of the more ambiguous and subtle ways
in which contemporary forms of racial discrimination manifest in institutional contexts. These
results call attention to the need to understand how different types of racial discrimination
S832 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

impact health, and to revisit presumptive associations. Using stress theory, we discuss
potential explanations for these findings and make recommendations for future direction in
research.

CORRESPONDING AUTHOR: Marilyn Thomas, MPH, University of California Berkeley School of

Public Health, Redwood City, CA, 94062; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S833

Paper Session 18: Innovative Studies on Discrimination and Stress 3:55 PM-4:12 PM

THE PROTECTIVE ROLE OF MINDFULNESS IN THE RELATIONSHIP BETWEEN PERCEIVED

DISCRIMINATION AND DEPRESSION

Amanda J. Shallcross, ND, MPH1, Tanya Spruill, PhD2

1
New York University School of Medicine, New York, NY; 2NYU School of Medicine, New York,
NY

Background: Perceived discrimination has substantial negative effects on well-being and is

associated with severe psychological outcomes, including depression. Similar to other types of
life stress, the consequences of discrimination may depend on an individual's ability to cope
with negative emotions triggered by a distressing event. Widespread evidence indicates that
mindfulness, deliberate attentiveness to internal and external experiences, is an effective
coping strategy that is associated with salutary outcomes in the face of stressful experiences,
including discrimination. Few studies, however, have examined the role of mindfulness in
minority samples. Thus, it is unknown whether the protective effects of mindfulness extend to
racial/ethnic minorities, who are most likely to experience discrimination.

Objective: The goal of the current analysis is to test whether trait mindfulness moderates the
relationship between perceived discrimination and depressive symptoms in a minority
sample.

Method: A heterogeneous sample of healthy adults (n=163) living in New York City were
recruited as a part of a larger study on stress, rumination, and blood pressure. A subset
(n=100) of exclusively minority participants (54% female; 48% Black/non-Hispanic, 35%
Hispanic, 14% Asian, 3% mixed race/ethnicity; 34 years=Mean age) was used for the current
analysis. Standardized inventories of trait mindfulness, perceived discrimination, and
depressive symptoms were assessed cross-sectionally.

Results: Regression analyses indicated a significant interaction between perceived

discrimination and trait mindfulness predicting depressive symptoms ( = -1.3, p =.04). Simple
slopes analyses revealed that participants lower in mindfulness exhibited increased
depressive symptoms with higher levels of discrimination ( = 2.3, p < .001). Discrimination
S834 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

was not associated with increased depressive symptoms among individuals higher in
mindfulness ( = -.05, p =.96).

Conclusions: This study is one of the first empirical studies of trait mindfulness in an
exclusively minority sample. Results suggest that mindfulness may mitigate the deleterious
effects of discrimination on depressive symptoms in racial/ethnic minority individuals. The
high prevalence and severity of discrimination among minorities, in particular, underscores
the importance of these findings. The use of mindfulness-based interventions may be a
promising approach to reduce the harmful psychological effects of discrimination.

CORRESPONDING AUTHOR: Amanda J. Shallcross, ND, MPH, New York University School of
Medicine, New York, NY, 10016; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S835

Paper Session 18: Innovative Studies on Discrimination and Stress 4:13 PM-4:30 PM

SOCIO-ENVIRONMENTAL STRESS EXPOSURES AND OBESITY: RESULTS FROM THE HCHS/SOL

SOCIOCULTURAL ANCILLARY STUDY

Adolfo Cuevas, PhD1, Cheryl Clark, MD, ScD2, Carmela Alcantara, PhD3, Sheila F. Castaneda,
PhD4, Martha Daviglus, MD, PhD5, Donglin Zeng, Samantha Reina, Mercedes Carnethon,
Ph.D.6, Linda C. Gallo, PhD7, Carmen R. Isasi, MD, PhD8, Robert Kaplan, William Arguelles

1
Harvard T.H. Chan School of Public Health, Boston, MA; 2Brigham and Women's Hospital,
Boston, MA; 3Columbia University, New York, NY; 4San Diego State University, San Diego, CA;
5
University of Illinois at Chicago, Chicago, IL; 6Northwestern University, Chicago, IL; 7San Diego
State University, Chula Vista, CA; 8Albert Einstein College of Medicine, Bronx, NY

Introduction: Hispanics/Latinos have a higher rate of obesity compared to non-Hispanic

Whites.1,2 Few studies have examined the obesogenic effects of three socio-environmental
stressors (ethnic discrimination, acculturative stress, neighborhood stress), and the combined
effect of the individual stressors.

Methods: Data derived from the Sociocultural Ancillary Study, which assessed sociocultural
and psychosocial influences in cardiovascular risks in a subset (n= 5,313) of the larger Hispanic
Community Health Study/Study of Latinos. BMI was used to define weight categories as
underweight/normal weight, overweight, and obese. The three socio-environmental stress
variables were each standardized to facilitate comparison across stress domains. A cumulative
stress variable was created by summing together each of the standardized socio-
environmental stressors. Multinomial logistic regression was used to model the effects of
each stressor, adjusting for sociodemographics, physical activity, and dietary behavior.

Results: Greater ethnic discrimination and greater acculturative stress were each associated
with a higher odds of being obese (OR=1.14, 95% CI=1.01-1.29 and OR=1.17, 95% CI=1.04-
1.31, respectively). Compared with no stress, those experiencing stress in a single domain had
a 1.84 times higher odds of being obese (95%CI=0.73-4.61), while those experiencing stress in
two or three domains had 1.94 (95%CI=0.78-4.83) and 2.29 (95%CI=0.94-5.61) times greater
odds of being obese, respectively.

Conclusion: Ethnic discrimination and acculturative stress may be important contributors to

obesity among Latinos/Hispanics. Stress management techniques may serve as
complementary methods to lifestyle modification interventions to prevent or treat obesity in
Hispanics/Latinos.
S836 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Adolfo Cuevas, PhD, Harvard T.H. Chan School of Public Health,
Boston, MA, 02215; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S837

Paper Session 19: Substance Use: Patterns and Treatments 3:00 PM-3:18 PM

SUBSTANCE USE PATTERNS IN HOMELESS YOUTH: RESULTS OF AN ECOLOGICAL MOMENTARY

ASSESSMENT

Diane Santa Maria, DrPH, MSN, RN, APHN-BC1, Julia Joohyun. Jung, BA2

1
University of Texas Health Science Center Houston, Houston, TX; 2School of Social Sciences
and Humanities, Houston, TX

Introduction: Homeless youth have high rates of substance use. While housing instability and
homelessness has been associated with higher rates of risky behaviors such as substance use,
there is limited literature assessing how substance use rates vary depending on day level
housing type. Yet, most studies exploring this association have been limited to cross-sectional
data.

To address this gap, we explored if the relation between shelter type and substance use using
Ecological Momentary Assessment (EMA) data from homeless youth in Houston.

Methods: We recruited 71 homeless youth aged 1824 from a drop-in center in Houston, TX,
between September 2015 and March 2016. We conducted daily EMAs for up to 3 weeks,
which measured substance use and shelter status. To determine the relation between
substance use and shelter status, we conducted chi-squared testing on daily EMAs (n = 71,
EMAs =2232).

Results: We collected a total of 861 daily assessments using EMA. Daily EMAs indicate that
221 (24.4%) were substance use days and 44 (5.1%) were alcohol use days. Days of substance
use per participants ranged from 0-18 days out of 21 possible days. In response to the
question regarding shelter status, youth most frequently reported unstable housing (i.e.
staying with friends, acquaintances, or relatives) (44% of days), staying on the streets (15%),
and in shelters (12%). Using chi-squared tests, we found a significant difference (p > 0.00) in
daily substance use by current shelter status. Days in which youth stayed with a friend or
acquaintance had the highest rates of substance use (35.6%) followed by street dwelling days
(34.4%) and days staying in the home of a romantic or sexual partner (33.7%). Shelter days
had the lowest rates of substance use (7.8%). We also found a significant difference (p > 0.00)
in daily alcohol use by shelter status. Days when youth stayed in a car had the highest rates of
alcohol use (11.11%) followed by days staying in a hotel or motel (10.7%) and on the streets
(9.4%). Shelter days had the lowest rates of alcohol use (1.9%) followed by days staying with
relatives (2.8%).

Conclusion: We found that shelter type is associated with higher rates of substance and
alcohol use. This data suggests that sheltering type may reduce the risk of substance use and
S838 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

that days when youth stay with others may be high risk substance use days. The social
network surrounding unstable housing circumstances and substance use warrants further
exploration.

CORRESPONDING AUTHOR: Diane Santa Maria, DrPH, MSN, RN, APHN-BC, University of Texas
Health Science Center Houston, Houston, TX, 77030; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S839

Paper Session 19: Substance Use: Patterns and Treatments 3:19 PM-3:36 PM

FACTORS CONTRIBUTING TO SUCCESSFUL INDUCTION IN MEDICATION ASSISTED TREATMENT

Emily Zhang, BA , Naomi Newman, BS, Emily Loscalzo, Psy.D., Robert Sterling, Ph.D.

Thomas Jefferson University, Philadelphia, PA

Factors Contributing to Successful Induction in Medication Assisted Treatment

Emily Zhang, Naomi Newman, Emily Loscalzo, & Robert Sterling

Thomas Jefferson University

While ASAMs expert panel suggests that it should take 4-6 weeks to achieve an effective
blocking dose for those in medication assisted treatment-MAT (Baxter et al., 2013), little
research is available on latency to abstinence in this treatment population. To address the gap
in the literature, this study examined latency to abstinence and possible predictors.
Abstinence was defined as four consecutive negative urine tests. Subjects were a subsample
of 132 consecutive admissions to a university sponsored, community based, outpatient, MAT
facility. Participants were individuals admitted between 7/1/2014-6/30/2015. Average age at
time of admission was 41.75 10.54; 72% were male; 68.9% were white. Mean age of first
drug use was 23.528.15; mean years of opiate use were 18.311.62; mean number of prior
drug treatment episodes was 3.372.88. For latency to secondary substance abstinence, only
20.7% of the 121 identified at admission as using a second drug were abstinent within one
year, indicating that individuals struggled to suspend secondary drug use. A series of
discriminant analyses were conducted to identify whether key demographic, life skill,
addiction, social support, and mental health variables were predictive of latency to opioid
abstinence. Analyses conducted on the entire sample failed to identify a set of predictor
variables. Thus, two discreet groups were identified those who achieved abstinence in
accord with the ASAM criteria (n=24) and those unable to achieve abstinence in the first year
of enrollment (n=24). A significant discriminant function (Wilks Lambda = .50, p=.02), which
correctly re-classified 79.2% of the initial cases, was identified. Examination of the
discriminant function coefficients suggest that a younger, more educated, later heroin use
onset, more resource laden individual with a sense of efficacy to achieve abstinence was less
likely to obtain abstinence. These results differentiate between who engages in and profits
more rapidly from treatments vs. who engages in bad outcomes. Data suggest that efforts to
increase the likelihood of profiting from treatment should focus on issues related to age,
education, financial resources, over-confidence, and opioid abuse history. Such individuals are
at risk for continued illicit substance use despite MAT enrollment.
S840 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Emily Zhang, BA , Thomas Jefferson University, Philadelphia, PA,

19107; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S841

Paper Session 19: Substance Use: Patterns and Treatments 3:37 PM-3:54 PM

MERITORIOUS AWARD WINNER

PERCEPTIONS AND BELIEFS MOTIVATING PARENTAL DISCUSSIONS OF MARIJUANA USE WITH

CHILDREN

Tenie Khachikian, MSW, Linda D. Cameron, Ph. D.

University of California, Merced, Merced, CA

Adolescent marijuana use remains a significant issue in the United States. Recent legalization
of marijuana use could lead parents to perceive marijuana use as less harmful and possibly
decrease concerns of marijuana use by their children. Parents play an important role in
adolescent substance use prevention and reduction, and their engagement in discussions of
substance use with their children can significantly influence their experimentation and use.
Little is known about what beliefs and motivations influence parents decisions to discuss
marijuana use with their children. Guided by the Prototype Willingness Model, we tested
hypotheses that higher parental perceived risk of harms, negative prototypes of youth who
use marijuana, and worry about the health and social harms of their child using marijuana
would predict higher intentions and willingness to have marijuana use discussions with
children and that, in turn, intentions and willingness would positively predict having these
discussions and behavior. We administered a national, longitudinal survey to American
parents of children ages 10 to 17 years old (N = 499) assessing risk perceptions, negative and
positive prototypes of youth marijuana users, worry, intentions, and willingness. Participants
(N = 409) completed a survey one month later assessing whether they had talked with their
child about marijuana use. Overall, 40% of the parents reported having marijuana use
discussions over the prior month. Structural equation modeling revealed that higher
perceived risks of use ( = .59, p < .001) and less positive prototypes ( = -.34, p < .001)
predicted higher worry of harms of marijuana use. Higher worry predicted greater intentions
( = .71, p < .001) and willingness ( = .27, p < .001) to discuss marijuana use with their
children. Higher intentions (= .26, p < .001) and higher willingness (= .19, p < .001) positively
predicted marijuana use discussions one month later. These findings contribute to our
understanding of how risk perceptions, prototypes, and worry might shape parent intentions
and willingness to discuss marijuana use with their children, with intentions and willingness
independently contributing to discussion behavior. Further research can explore how
communications targeting these measures might increase parent-child discussions of
substance use.
S842 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Tenie Khachikian, MSW, University of California, Merced, Merced,

CA, 95343; [email protected]

Paper Session 19: Substance Use: Patterns and Treatments 3:55 PM-4:12 PM

MEDICAL CANNABIS USE AND ALCOHOL USE DISORDER AMONG PATIENTS PRESCRIBED LONG-
TERM OPIOID THERAPY

Shannon Nugent, PhD1, Bobbi Jo H. Yarborough, PsyD2, Ning Smith, PhD2, Carla A. Green,
PHD, MPH3, S Dobscha, MD4, Richard Deyo, MD, MPH4, B Morasco, PhD1

1
VA Portland Health Care System, Portland, OR; 2Kaiser Permanente Center for Health
Research, Portland, OR; 3Kaiser Permanente Northwest, Portland, OR; 4Oregon Health &
Science University, Portland, OR

Purpose:

Debate exists about the effectiveness and safety of simultaneous medical cannabis (MC) use
and prescription long term opioid therapy (LTOT) among chronic pain patients. Yet, among
patients with simultaneous use, little is known about clinical characteristics that may impact
safety or suggest a need for additional services. We present baseline data from an ongoing
longitudinal study of patients prescribed LTOT to (1) compare clinical characteristics between
patients who endorsed using MC for pain and those who did not; and (2) examine associations
between alcohol use disorder (AUD) and use of MC.

Methods:

Three hundred seventy-one patients prescribed LTOT were recruited from two large
integrated health care systems. Data about opioid status were collected from participants
medical records and data about substance use (including MC), mental health conditions, and
pain were assessed with validated self-report measures.

Results:

Mean age was 60.0 (SD=11.0), 54% were male, 19% met criteria for a current AUD, and 18%
endorsed using MC for pain. Among those using MC, 66% rated it as at least moderately
helpful in reducing pain. Bivariate analyses comparing patients who do and do not use MC for
pain showed that compared to non-users, MC users had statistically significant higher rates of
nicotine use (42% vs 26%), higher mean scores of risk for prescription opioid misuse (16.8, SD
= 7.5 vs 12.5, SD = 7.3), and higher rates of AUD (34% vs 15%). No significant group differences
were detected on pain intensity or disability, depression, opioid dose, or anxiety scores.
Logistic regression analysis adjusting for clinical and demographic variables revealed that MC
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S843

use was associated with an increased likelihood of a current AUD, OR=2.21, 95% CI=1.14-4.31,
p=.02.

Conclusion:

While the majority who used MC reported it to be helpful, MC using patients also had
significantly higher rates of AUD. These findings highlight the need to screen for MC use
among patients prescribed LTOT and to screen for, and address AUD among patients using
MC. Future research could examine the long-term impact of MC use on pain-related outcomes
among patients prescribed LTOT.

CORRESPONDING AUTHOR: Shannon Nugent, PhD, VA Portland Health Care System, Portland,
OR, 97239; [email protected]
S844 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 19: Substance Use: Patterns and Treatments 4:13 PM-4:30 PM

PREDICTING OPIATE CRAVING WITH GPS DATA: COMPARISON OF MACHINE LEARNING

TECHNIQUES ON QUALITY AND USABILITY

William Kowalczyk, PhD, Matthew Tyburski, PhD, Karran Phillips, MD, David Epstein, PhD,
Kenzie Preston, PhD

National Institute on Drug Abuse, Intramural Research Program, Baltimore, MD

Some behaviors (e.g. manic episodes, suicide) are better to prevent than to treat in their
aftermath. In addiction we want to prevent relapse, a problem that is separate from
persistent ongoing drug use. Prior attempts to predict imminent lapse have focused on time
scales of weeks, which is problematic when attempting to intervene in a behavior that occurs
in a moment. Our goal is to predict behavior in real time, to open up a world of live, just-in-
time (JIT) mHealth interventions for drug lapses and other behaviors. The present study takes
a step towards that goal by comparing methods for making ambulatory predictions of craving.

In opiate-using, methadone-maintained outpatients (n = 81), we collected GPS data and

ratings of opiate craving over 16 weeks. We mapped participants GPS tracks for the 6.5 hours
before each assessment, coding each set of tracks with an independently obtained observer
rating of visible neighborhood disorder (NIfETy). We dropped the 90 minutes of NIfETy data
immediately preceding an assessment and used the remaining 5 hours to train randomForest
(rF) or Support Vector Machine (SVM) models to predict opiate craving (with and without
demographic factors, reserving 50% or 10% of the data for testing). We then applied the
models in a 24/7 manner, using all the GPS data to examine the utility of the posterior
probability scores for possible use in a JIT intervention.

In models without demographic input, the rF approach was clearly better, as SVM failed to
predict craving (kappa: rF = 0.59, SVM = 0.0). Models with demographic input predicted opiate
craving at levels of excellent agreement using either type of machine learning (kappa for
either SVM or rF = 0.63). However, the patterns of prediction differed. In 24/7 models, the
scores from rF models tended to form a jittery pattern around a horizontal centroid. In
contrast, the scores from SVM models showed very prominent spikes, indicating much more
clearly when an intervention should be delivered. This pattern was present for most
participants.

Our goal is to harness advances in environmental data to reduce the burdens of behavioral
disorders. Both models were able to predict a patients future state using GPS data, using
algorithms that could ultimately run on a smartphone. The SVM predictions had a more
desirable pattern with which to use in a JIT mHealth intervention to predict drug lapses. The
step represented here may be extended to other behavioral disorders.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S845

CORRESPONDING AUTHOR: William Kowalczyk, PhD, National Institute on Drug Abuse,

Intramural Research Program, Baltimore, MD, 21224; [email protected]
S846 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 20: Tobacco: An Ongoing Behavioral Medicine Challenge 3:00 PM-3:18 PM

CITATION AWARD WINNER

CROWDSOURCING DATA COLLECTION FOR PUBLIC HEALTH: A COMPARISON WITH

NATIONALLY REPRESENTATIVE TOBACCO DATA

Darren Mays, PhD, MPH1, Andrew Strasser, PhD2, Raymond Niaura, PhD3, Eric Lindblom, JD4,
John Kraemer, JD, MPH5

1
Georgetown University Medical Center, Lombardi Comprehensive Cancer Center,
Washington, DC; 2University of Pennsylvania School of Medicine, Philadelphia, PA; 3Truth
Initiative, Washington, DC; 4Georgetown University Law Center, Washington, DC;
5
Georgetown University School of Nursing & Health Studies, Washington, DC

Online crowdsourcing platforms such are increasingly used for behavioral research in public
health. Crowdsourcing platforms are advantageous because they provide accessibility to
existing online populations to gather data efficiently and at relatively low cost. However, the
representativeness of crowdsourced samples and generalizability are unclear. This study
sought to compare crowdsourced data on tobacco use to publicly available, nationally-
representative U.S. survey data, the 2012-2013 National Adult Tobacco Survey (NATS). A
sample of 3,999 U.S. young adults ages 18 to 30 years recruited through Amazon Mechanical
Turk completed measures taken from NATS assessing demographics, tobacco use behaviors
and perceptions, and exposure to tobacco warning labels. Post-hoc sample weights were
created for crowdsourced data using procedures similar to the NATS. Weighted analyses
compared the crowdsourced and NATS data on demographics, tobacco use behavior, and
other measures administered. The crowdsourced sample was comparable to the NATS on
variables used to create sample weights (gender, race/ethnicity, cigarette smoking status).
Those in the crowdsourced sample were less likely than young adults in the NATS to report a
household income of $50,000 per year or higher (25.1% vs. 45.4%, 95% CI diff. -23.0%, -
17.7%), and more likely to be married/living with a partner (20.9% vs. 14.8%, 95% CI diff.
3.9%, 8.4%). The prevalence of current use of e-cigarettes (13.9% vs. 2.6%, 95% CI diff. 9.7%,
13.1%), hookah (6.4% vs. 1.9%, 95% CI diff. 3.1%, 5.8%), and cigars (8.4% vs. 3.2%, 95% CI diff.
3.7%, 6.7%) was greater in the crowdsourced sample versus the NATS. High proportions of
both samples (ranging from 71.9% to 88.8%) endorsed responses indicating cigarette smoking
is very harmful and very addictive. Comparisons of the proportions of non-smokers and
smokers in the crowdsourced and NATS data showed similar patterns with respect to the
frequency of cigarette warning label exposure. However, the likelihood of reporting that
smoking is very harmful by frequency of warning label exposure was lower among smokers in
the crowdsourced sample. This study provides some of the first empirical insights into the
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S847

ways in which a large, crowdsourced sample compares to national data. Crowdsourced data
collection has important applications in public health research (e.g., correlational,
experimental studies) but the applicability to answer population-level questions may be
limited.

CORRESPONDING AUTHOR: Darren Mays, PhD, MPH, Georgetown University Medical Center,
Lombardi Comprehensive Cancer Center, Washington, DC, 20007; [email protected]
S848 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 20: Tobacco: An Ongoing Behavioral Medicine Challenge 3:19 PM-3:36 PM

DISPARITIES IN SMOKING AND E-CIGARETTE USE AMONG YOUNG SEXUAL MINORITY MEN

Mary A. Gerend, PhD1, Michael E. Newcomb, Ph.D.1, Brian Mustanski, Phd2

1
Northwestern University, Chicago, IL; 2Northwestern university, Chicago, IL

Background: Because sexual minorities (gay, lesbian, and bisexual individuals) are significantly
more likely to use tobacco than their heterosexual counterparts, identifying predictors of
tobacco use in these groups is essential. Little is known about which subgroups of men who
have sex with men (MSM) are more likely to smoke, and even less is known about MSMs use
of new and rapidly changing products like electronic cigarettes (e-cigarettes). This study
examined rates and predictors of smoking (combustible cigarette use) and e-cigarette use
among 16-29 year-old MSM.

Methods: Participants (N=773) were drawn from the baseline visit of an ongoing longitudinal
cohort study of young MSM aimed at identifying multilevel influences on HIV risk and
substance use. Sociodemographic correlates (e.g., age, race/ethnicity, education, gender
identity, sexual orientation, physical attraction) of smoking and e-cigarette use were
examined using logistic regression.

Results: Although only 21% identified as current cigarette smokers, 39% had smoked within
the past 30 days. Nearly 40% had ever tried an e-cigarette, but only 4% reported regular
current or past e-cigarette use. E-cigarette use was significantly more common among current
and former (vs. never) smokers. Predictors of smoking were largely consistent with those
observed in the general population (e.g., older age, less education), but also included factors
specific to MSM (e.g., HIV positive, more attracted to females than males). Black (vs. white)
MSM and participants aged 16-18 (vs. 19-24) were less likely to have ever used e-cigarettes.
Regular e-cigarette users were more likely to be transgender (male-to-female), white (vs.
black), more educated, and attracted to mostly females, but some males (vs. only males).

Conclusions: Findings highlight potential disparities in smoking and e-cigarette use among
young sexual minority men. Although some variables (e.g., gender identity, attraction)
demonstrated similar relationships with smoking and e-cigarette use, others (e.g., race,
education) demonstrated opposite patterns. Findings underscore the urgent need for tailored
tobacco prevention and cessation efforts for young MSM and suggest that different
frameworks may be required for different tobacco products.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S849

CORRESPONDING AUTHOR: Mary A. Gerend, PhD, Northwestern University, Chicago, IL,

60611; [email protected]
S850 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 20: Tobacco: An Ongoing Behavioral Medicine Challenge 3:55 PM-4:12 PM

UNDERSTANDING CIGARETTE SMOKING TRAJECTORIES THROUGH ADOLESCENCE AND

WEIGHT STATUS IN YOUNG ADULTHOOD IN THE UNITED STATES

Minal Patel, PhD, MPH1, Annette Kaufman, PhD, MPH2, Yvonne Hunt, PhD, MPH3, Linda
Nebeling, PH.D., MPH, RD, FAND3

1
National Cancer Institute, Rockville, MD; 2Tobacco Control Research Branch, Behavioral
Research Program, Division of Cancer Control and Population Sciences, National Cancer
Institute, National Institutes of Health, Rockville, MD; 3National Cancer Institute, Bethesda,
MD

Background: Adolescent cigarette smoking has steadily declined since 1999, while obesity
rates have quadrupled since the 1980s. Few studies have looked at the longitudinal
relationship of smoking and weight in youth and young adulthood. This study examines the
relationship of smoking trajectory beginning in adolescence on weight status in young
adulthood in a nationally representative US longitudinal sample.

Methods: The study sample was drawn from Add Health Waves I-IV (N=13,361). Four
trajectories were generated through repeated-measures latent class analyses using current
cigarette smoking status in Waves I-IV and age of initiation: nonsmokers (44%), early
establishers (23%), late establishers (21%), and former smokers (12%). Wave IV weight status
included self-reported body mass index (BMI) and measured waist circumference (WC).
Bivariate analyses addressed associations between smoking trajectory and BMI at each wave.
Multivariate linear regression models tested the relationship of smoking trajectory and weight
status at Wave IV. Future planned analyses include addressing related health behaviors as
potential moderators of the relationship between smoking trajectory and BMI.

Results: Weight status increased over time for all trajectories. In bivariate analyses, early and
late establishers had a significantly lower BMI at Wave III and all smoking trajectories had a
significantly lower BMI at Wave IV compared to nonsmokers. All smoking trajectories had a
significantly lower BMI than nonsmokers [early establishers: (B=-1.27, CI: -1.56, -0.98); late
establishers: (B=-0.84, CI: -1.16, -0.52); & former smokers: (B=-0.63, CI: -0.93, -0.34); (p <
0.05)] in an adjusted multivariate regression model predicting Wave IV BMI. Males reported a
lower BMI ( = -0.28; CI: -0.53, -0.03) than females. Results predicting Wave IV WC showed a
similar trend.

Discussion: Smoking during adolescence does not attenuate the risk of becoming obese in
young adulthood, even though smoking trajectories had a lower BMI and WC than
nonsmokers. These results have implications for interventions concerning cigarette smoking
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S851

and weight control in youth and young adults. Both tobacco use and obesity need to be
studied concurrently in order to better understand potential additive and/or synergistic
effects among youth and young adults in the US.

CORRESPONDING AUTHOR: Minal Patel, PhD, MPH, National Cancer Institute, Rockville, MD,
20850; [email protected]
S852 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 20: Tobacco: An Ongoing Behavioral Medicine Challenge 4:13 PM-4:30 PM

MISPERCEPTIONS OF PEER NORMS FOR TOBACCO USE AND PERSONAL USE: EVIDENCE FROM
STUDENTS IN GRADES 6-12 ACROSS 63 SCHOOLS IN THE US

Jessica Mayson. Perkins, PhD1, H. Wesley. Perkins, PhD2, David Craig, Ph.D.2

1
Massachusetts General Hospital, Cambridge, MA; 2Hobart and William Smith Colleges,
Geneva, NY

Actual peer social norms and students perceptions of those norms within the school context
may be an important determinant of tobacco use initiation and frequency among adolescents
as youth look to peers for social cues on what to believe and how to behave. This study
distinguishes between actual and perceived peer norms regarding tobacco use attitudes and
behavior, and examines the independent associations of both actual and perceived norms
with personal attitude and personal use among U.S. students in grades 6-12. Data were
collected from 27,545 students across 63 schools in 11 states from 1999 to 2014 via
anonymous surveys. Even though permissive attitudes about tobacco use and the actual
prevalence of tobacco use increased across grade and age levels, believing that tobacco use is
okay or actually using tobacco during the year was rarely the norm in any grade level in any
specific secondary school. However, dramatic misperception was found across a diversity of
schools and students. Although 77% of students said tobacco use was never good, 64% of
students thought that most others in their grade thought use was okay. Similarly, although
78% of students reported never using tobacco, 67% of students perceived that students in
their grade most typically used tobacco monthly or more often. Although the prevalence of
misperceptions of the attitudinal and behavioral norms regarding tobacco use was greater
and more inaccurate across increasing grade and age levels, these misperceptions were
pervasive across all racial categories of students, socio-economic status levels of schools, size
of grade cohort in ones school, and time periods within this study. In addition, perception of
the attitudinal norm was highly predictive of personal attitude about use. Moreover,
misperceiving tobacco use as the norm among same-grade peers strongly predicted personal
tobacco use even after adjusting for the actual same-grade prevalence of tobacco use,
personal attitude about use, and several other individual and school-level factors. Researchers
should give increased attention to designing experiments to assess causality while also testing
interventions to reduce misperceptions, and presumably reduce permissive attitudes and
actual use. Furthermore, practitioners could begin to explore comprehensive ways to
promote the awareness of positive actual norms related to tobacco use.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S853

CORRESPONDING AUTHOR: Jessica Mayson. Perkins, PhD, Massachusetts General Hospital,

Cambridge, MA, 02138; [email protected]
S854 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 21: Influence of Tobacco Advertising 3:00 PM-3:18 PM

PRO-TOBACCO ADVERTISEMENT EXPOSURE AMONG AFRICAN AMERICAN SMOKERS: AN

ECOLOGICAL MOMENTARY ASSESSMENT STUDY

Cendrine Robinson, PhD MPH1, Joanna R. Sells, M.S.2, Andrew Waters, PhD 2

1
National Cancer Institute, Rockville, MD; 2Uniformed Services University of the Health
Sciences, Bethesda, MD

Significance: African Americans live in communities with a disproportionately high density of

tobacco advertisements compared to Whites. Some research indicates that point-of-sale
advertising is associated with unplanned purchases of cigarettes, craving, and smoking. Most
previous research is limited by recall bias and did not focus on African American smokers.
Ecological Momentary Assessment (EMA) can be used to obtain prospective associations
between tobacco advertisement exposure and smoking in the natural environment. The
current study used EMA to assess self-reported advertisement exposure, craving, smoking,
purchases, and impulse purchases among African Americans.

Methods: Non-treatment seeking African American smokers (N = 64) were given a mobile
device for 2 weeks. They were prompted four times per day to respond to questions about
advertisement exposure, craving, smoking, and unplanned purchases. At each assessment,
participants reported their craving right now, the number of cigarettes smoked since the
last assessment, the number of advertisements seen since the last assessment, and
whether they made any unplanned purchases.

Results: Participants provided data on advertisement exposure in 2,298 assessments.

Participants reported seeing at least one advertisement in 33.54% of assessments. Among
assessments in which an advertisement was seen, 33.07% of assessments had one
advertisement, 29.31% had two advertisements, 17.51% had three advertisements, and
20.10% had four or more. Participants reported that 86.82% of advertisements seen were
menthol. Number of advertisements seen was positively associated with smoking, F (1,2110) =
112.85 p F (1,1417) = 16.54 p p p > .05.

Conclusions: African American smokers are frequently exposed to pro-tobacco marketing.

Advertisement exposure is associated with smoking and impulse purchases, but not craving.
Future research should assess advertisement exposure among smokers wishing to quit.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S855

CORRESPONDING AUTHOR: Cendrine Robinson, PhD MPH, National Cancer Institute,

Rockville, MD, 20850; [email protected]
S856 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 21: Influence of Tobacco Advertising 3:19 PM-3:36 PM

CITATION AND MERITORIOUS AWARD WINNER

STANDARDIZED PACKAGING MAY REDUCE THE PERCEPTION THAT AMERICAN SPIRIT

CIGARETTES ARE LESS HARMFUL

Eric Leas, MPH1, Claudiu Dimofte, PhD2, David Strong, PhD3

1
University of California, San Diego, San Diego, CA; 2San Diego State University, San Diego, CA;
3
Moores Cancer Center, UC San Diego, La Jolla, CA

Background: Although the Family Smoking Prevention and Tobacco Control Act of 2009
essentially banned the marketing of any cigarette as safer in the United States, more than
2.5 million U.S. consumers believe they are smoking a brand of cigarettes that might be less
harmful. This misconception of reduced harm appears to be most common among those who
smoke the brand American Spirit.

Methods: Using a between-subject fractional-factorial experiment (3-brands x 3-packaging/

labeling styles), smokers recruited through Amazon Mechanical Turk (N=909) rated their
perceptions of a randomly assigned cigarette package. We assessed whether smokers tended
to rate the standard packaging of American Spirit Full-Bodied cigarettes higher than the
standard packaging of Marlboro Red or Newport Green cigarettes (the most popular U.S.
brands) on an 18-point scale assessing whether the packaging conveyed that the cigarettes
were less harmful (Cronbachs = 0.92). Then, we estimated the effect of standardized
cigarette packaging (all branding replaced with a green color) and Australian-like packaging
(all branding replaced with a green color and a graphic image) on harm perceptions conveyed
through American Spirit packaging, using the same 18-point scale.

Results: Smokers ratings of the standard American Spirit package (mean=9.9; standard
deviation(SD)=4.4) were 1.5 times (t(df=165.5)=6.0; p < 0.001) higher than smokers ratings of
the Marlboro Red package (mean=6.6; SD=3.4) and 1.4 times (t(df=171.9)=4.9; p < 0.001)
higher than smokers ratings of the Newport Green package (mean=7.0; SD=3.6) on the
reduced harm perceptions scale. The standardized packaging had a medium-sized effect
(Cohens d=0.66; t(df=179.4)=4.5; p < 0.001) and the Australian-like packaging had a large
effect (Cohens d=1.56; t(df=146.3)=10.5; p < 0.001) on reducing the perception that American
Spirit packaging conveys that its cigarettes are less harmful. The standardized packaging had a
negligible effect on reducing harm perceptions conveyed through Marlboro Red (Cohens
d=0.02; t(df=207.3)=6.7; p=0.87) and Newport Green (Cohens d=0.16; t(df=176.5)=1.1;
p=0.29) packaging; however, the Australian-like packaging had a large effect on both
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S857

Marlboro Red (Cohens d=0.88; t(df=207.3)=; p < 0.001) and Newport Green (Cohens d=0.84;
t(df=167.8)=5.7; p < 0.001) packaging.

Conclusion: These data suggest that many smokers perceive that American Spirits cigarette
packaging conveys that its cigarettes are less harmful. The data also suggest that standardized
packaging, both with and without warning images, may reduce this perception.

Implication: Standardized packaging may be a viable option for preventing erroneous

inferences about harms of cigarettes conveyed through cigarette packaging.

CORRESPONDING AUTHOR: Eric Leas, MPH, University of California, San Diego, San Diego, CA,
92103; [email protected]
S858 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 21: Influence of Tobacco Advertising 3:37 PM-3:54 PM

ICONS OF HEALTH EFFECTS OF CIGARETTE SMOKE: A TEST OF SEMIOTIC TYPE

Allison Lazard, PhD1, Annie Schmidt, BA1, Huyen Vu, MS1, M. Justin Byron, PhD1, Ellen Peters,
PhD2, Marcella H. Boynton, PhD1, Noel T. Brewer, PhD3

1
University of North Carolina - Chapel Hill, Chapel Hill, NC; 2The Ohio State University,
Columbus, OH; 3University of North Carolina, Chapel Hill, Chapel Hill, NC

Purpose. U.S. law requires the government to inform the public about the harmful and
potentially harmful chemicals in cigarette smoke. Information about these chemicals and
their harmful effects must be understandable and not misleading. One way to achieve this
communication goal is through the use of simple visuals, such as icons. We sought to identify
icons to effectively communicate the health harms of cigarette smoke.

Methods. Participants were 701 U.S. adults. Within-subjects online experiments explored the
effects of icon semiotic type: symbolic (arbitrary, most abstract), indexical (illustrating
evidence of a physical connection), and iconic (representation of an object, most concrete).

Results. For the not-easy-to-visualize harms of cancer and addiction, symbolic icons received
higher ratings than other icon types on cognitive and affective reactions (perceived
representativeness of the icon, affect toward smoking, elaboration about the harms of
smoking, perceived severity of health effects of smoking, perceived effectiveness of the icon
for discouraging smoking, and icon preference) (all p < .001). For the easy-to-visualize
symptoms of heart attack/stroke, indexical icons received the highest ratings (all p < .001). For
the easy-to-visualize harm of reproductive organ damage, the iconic image did best (all p <
.001).

Conclusion. Our findings contribute new insights for icon development for health
communication a process that is often anecdotal rather than informed by theory. Using
semiotics as a framework to anticipate how people would interpret icons, we found
systematic preferences detectable through cognitive and affective responses, as well as
through simple preference selection. For communicating harms cause by cigarette smoke,
health effects not easily visualized were best represented with symbolic icons. Health effects
linked to a specific organ or observable symptom were best represented with iconic and
indexical icons, respectively.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S859

CORRESPONDING AUTHOR: Allison Lazard, PhD, University of North Carolina - Chapel Hill,
Chapel Hill, NC, 27514; [email protected]
S860 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 21: Influence of Tobacco Advertising 3:55 PM-4:12 PM

CITATION AWARD WINNER

MEDIATORS OF THE IMPACT OF PICTORIAL CIGARETTE PACK WARNINGS ON QUIT ATTEMPTS

Noel T. Brewer, PhD1, Humberto Parada, Jr., PhD, MPH2, Kurt M. Ribisl, PhD3

1
University of North Carolina, Chapel Hill, Chapel Hill, NC; 2University of North Carolina at
Chapel Hill, Chapel Hill, NC; 3UNC, Chapel Hill, NC

Background. Pictorial warnings on cigarette packs increase quit attempts, but the mechanism
of the effect remains uncertain. We examined several hypotheses about mediators of the
impact of pictorial cigarette pack warnings on quit attempts.

Methods. We conducted a 4-week randomized clinical trial in California and North Carolina.
We recruited a convenience sample of 2,149 adult cigarette smokers from the general
population from September 2014 through August 2015. We randomly assigned participants to
receive on their cigarette packs for 4 weeks either a text-only warning (one of the Surgeon
Generals warnings currently in use in the United States on the side of the cigarette packs) or
a pictorial warning (one of the Family Smoking Prevention and Tobacco Control Acts required
text warnings with pictures that showed harms of smoking on the top half of the front and
back of the cigarette packs). The primary trial outcome was attempting to quit smoking during
the study. Analyses were intent-to-treat: 1) warnings to quit attempts (c path), 2) warnings to
mediators (a path), and 3) multivariable analysis of the full mediational pathways from
warnings to mediators to quit attempts (a*b paths).

Results. First, smokers whose packs had pictorial warnings were more likely than those whose
packs had text-only warnings to attempt to quit smoking during the 4-week trial (40% vs 34%;
odds ratio=1.29; 95% CI, 1.09-1.54). Second, pictorial warnings increased several potential
mediators: attention to the warning message, 8 different warning reactions such as negative
affect and avoidance, and social interactions about the warnings (p. < 05). However, pictorial
warnings changed only 2 of 15 belief and attitude measures (p. < 05). Third, pictorial warnings
led to greater avoidance of the warnings which led to more quit attempts (mediation, p. < 05).
Other mediators were cognitive elaboration (thinking about the warning and harms of
smoking), negative pack attitudes, and quitting self-efficacy (all, p. < 05).

Discussion. Pictorial warnings increased quit attempts by both creating aversive reactions and
by keeping the message vividly in smokers minds. Contrary to predictions from several
theories of health behavior, the warnings exerted little of their influence through changes in
beliefs and attitudes and none of their influence through changes in risk perception.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S861

CORRESPONDING AUTHOR: Noel T. Brewer, PhD, University of North Carolina, Chapel Hill,
Chapel Hill, NC, 27599; [email protected]
S862 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 21: Influence of Tobacco Advertising 4:13 PM-4:30 PM

WHY DO PEOPLE USE ELECTRONIC NICOTINE DELIVERY SYSTEMS (ELECTRONIC CIGARETTES)?

A CONTENT ANALYSIS OF TWITTER, 2012-2015

John Ayers, PhD MA, Jon-Patrick Allem

http://www.johnwayers.com/, Chula Vista, CA

The reasons for using electronic nicotine delivery systems (ENDS) are poorly understood and
primarily documented by expensive cross-sectional surveys that force respondents to adhere
to investigator's preconceived notions rather than using their own words. We passively
identify the reasons for using ENDS longitudinally from a content analysis of public postings on
Twitter. All English language public tweets including several ENDS terms (e.g., e-cigarette or
vape) were captured from the Twitter datastream during 2012 and 2015. After excluding
spam, advertisements, and retweets, posts indicating a rationale for vaping were retained.
The specific reasons for vaping were then inferred based on a supervised content analysis
using annotators from Amazons Mechanical Turk. During 2012 quitting combustibles was the
most cited reason for using ENDS. Forty-three percent (95%CI 39-48) of all reason-related
tweets cited quitting combustibles, e.g., I couldnt quit till i tried ecigs, eclipsing the second
most cited reason by more than double. Other frequently cited reasons included social image
(21%; 95%CI 18-25), use indoors (14%; 95%CI 11-17), flavors (14%; 95%CI 11-17), safety
relative to combustibles (9%; 95%CI 7-11), cost (3%; 95%CI 2-5) and lack of odor (2%; 95%CI 1-
3). By 2015 the reasons for using ENDS cited on Twitter had shifted. Both quitting
combustibles and use indoors significantly declined in mentions to 29% (95%CI 24-33) and
12% (95%CI 9-16). At the same time, social image increased to 37% (95%CI 32-43) and lack of
odor increased to 5% (95%CI 2-5), the former leading all cited reasons in 2015. Our data
suggest the reasons people vape are shifting away from cessation and toward social image.
These data also show how the ENDS market is responsive to a changing policy landscape. For
instance, smoking indoors was less frequently cited in 2015 as indoor smoking restrictions
where applied to vaping in many places since 2012. Because the data and analytic approach
are scalable, adoption of our strategies in the field can inform follow-up survey-based
surveillance (so the right questions are asked), interventions, and policies for ENDS.

CORRESPONDING AUTHOR: John Ayers, PhD MA, http://www.johnwayers.com/, Chula Vista,

CA, 91914-5210; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S863

Paper Session 22: HIV in Low-Resource Settings 3:00 PM-3:18 PM

HIV VULNERABILITY AMONG ZAMBIAN MIGRANTS AND MOBILE POPULATIONS

Stefani Butts, BA1, Nicole Barylski, N/A2, Maria L. Alcaide, MD1, Violeta J. Rodriguez, MSED1,
Annette Kayukwa, MPH3, Stephen M. Weiss, PhD, MPH1, Jackson Thoya, Deborah L. Jones,
PhD, M.Ed.1

1
University of Miami Miller School of Medicine, Miami, FL; 2University of Miami, Elmwood
Park, NJ; 3University of Zambia/Maimi University, Lusaka, Lusaka, Zambia

HIV vulnerability among Zambian migrants and mobile populations

Introduction: Migrant and mobile populations (MMP) throughout Zambia face unique barriers
to accessing HIV counselling, testing, and treatment associated with employment-driven
mobility. These factors in combination with elevated alcohol consumption heighten MMPs
risk of HIV infection and transmission. This qualitative study assessed barriers to accessing HIV
prevention and treatment services among MMP within Zambia, and explored potential
strategies to reduce barriers and enhance HIV prevention.

Method: This studyutilizedCommunity Based Participatory Research methods to investigate

MMP risk factors and proposed community-based HIV prevention strategies in Zambian
districts with the largest communities of MMP (Chingola, Mazabuka, Mufulira). Male and
female urban and rural Zambian MMP were enrolled in focus group discussions (n =137) and
key informant interviews (n=156). MMPs were primarily miners, cane cutters, mobile
contractors, truck drivers, sex workers and fisher folk. Stem questions explored HIV risk
factors and strategies to address barriers and facilitators of condom use, VCT and HIV
treatment. Resulting content was coded for dominant themes.

Results: Mobility increased the opportunity for large sexual networks and networking with
multiple partners. Local bars, lodges and truck stops were the primary locations for sexual
solicitation of MMP. Cheap, easily accessible short term lodging were hot spots for HIV risk
behavior and provided convenient venues for sexual exchange. Difficulty accessing support
groups, condoms, HIV treatment and health information and counseling due to transience
placed MMP at elevated risk. A lack of clinics in less accessible rural areas such as truck stops
and country borders exacerbated HIV risk, particularly among female MMP. Proposed
initiatives to reduce risk included promotion of border-based health clinics and employer-
driven programs to provide medical support and health resources to MMP. Regulation of bars
and lodges to reduce their use by sex workers was proposed, as well as economic stimulus
programs for sex workers to reduce their reliance on survival-based sex exchange.
S864 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Discussion: Barriers to HIV prevention tools in combination with large sexual networks
increase HIV vulnerability among Zambian MMP. Stimulation of economic opportunities for
commercial sex workers serving MMP should be explored as an avenue to substantially
reduce HIV risk. Involvement of MMP employers, Zambian governmental agencies and district
health facilities are needed to implement educational and social programs tailored to the
MMP lifestyle and respond to urgent health needs.

CORRESPONDING AUTHOR: Stefani Butts, BA, University of Miami Miller School of Medicine,
Miami, FL, 33136; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S865

Paper Session 22: HIV in Low-Resource Settings 3:19 PM-3:36 PM

IMPACT OF DEPRESSION IN HIV-INFECTED WOMEN ON INFANT DEVELOPMENT IN RURAL

SOUTH AFRICA

Violeta J. Rodriguez, MSED1, Gladys Matseke, MS2, Audrey Mathebula, BA(Hons)2, Ryan Cook,
MSPH1, Maria L. Alcaide, MD1, Stephen M. Weiss, PhD, MPH1, Karl Peltzer, PhD2, Deborah L.
Jones, PhD, M.Ed.1

1
University of Miami Miller School of Medicine, Miami, FL; 2Human Sciences Research Council,
Pretoria, Gauteng, South Africa

Introduction: High rates of depression have been identified among pregnant women living
with HIV (WLHIV) in rural South Africa (49%). Pregnancy-related depressive episodes among
WLHIV are associated with decreased prenatal care, and increased alcohol use and
interpersonal disengagement. HIV-exposure has also been associated with developmental
delays in infants, and HIV-exposed infants born to depressed women may be at an even
greater risk for adverse developmental outcomes. This study examined the impact of
depression in HIV-infected women in rural South Africa (SA) on infant development at 12
months of age.

Method: Participants (n = 68) were mother-infant dyads recruited from antenatal clinics in
rural SA. All women had a male partner. Women were 18 5 weeks pregnant at enrollment
and completed assessments of demographics, HIV disclosure status, pre- and postnatal
depression, male involvement during pregnancy, and alcohol use at baseline and 12 months
postpartum. Infant HIV serostatus (PCR) and domains of developmental functioning
(cognitive, expressive and receptive communication, and fine and gross motor) were assessed
at 12 months of age.

Results: Women were an average of 29 5 years old. One-third had completed at least 12
years of education, and 32% had a monthly income of ~US$76. Two-thirds of the pregnancies
were unplanned, and 45% of the women were diagnosed with HIV during pregnancy (average
time since diagnosis: 32 43 months); 59% had disclosed their HIV status to their partners.
Nearly half (45%) were depressed and 16 were referred for treatment for antenatal
depression. Four infants (6%) tested positive for HIV at 12 months. Infant seropositive HIV
status was associated with delayed receptive communication in bivariate analyses. In
multivariate analyses, not being married to their partner (p = .028), nondisclosure of HIV
status (p = .016), and postnatal depression (p = .038) predicted infant cognitive delay at 12
months of age. There also was an association between decreased prenatal male involvement
S866 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

and delayed infant gross motor development at 12 months (p = .002) and a non-significant
trend between delayed receptive communication and postnatal depression (p = .090),

Discussion: Findings highlight the need for healthcare providers to assess postpartum
depression among pregnant WLHIV, given its negative impact on infant development.
Interventions to support the mental health of pregnant WLHIV and increase male partner
involvement may enhance antenatal care and neonatal development outcomes in rural SA.
Further research examining child development theories in the South African context may be
merited to inform pedagogical approaches to managing developmental delay among infants
born to HIV-infected mothers experiencing postpartum depression.

CORRESPONDING AUTHOR: Violeta J. Rodriguez, MSED, University of Miami Miller School of

Medicine, Miami, FL, 33134; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S867

Paper Session 22: HIV in Low-Resource Settings 3:37 PM-3:54 PM

CITATION AWARD WINNER

REDUCING HIV VULNERABILITY AMONG ZAMBIAN ADOLESCENT GIRLS AND YOUNG WOMEN

Stefani Butts, BA1, Lauren E. Parmley, BA2, Maria L. Alcaide, MD1, Violeta J. Rodriguez, MSED1,
Annette Kayukwa, MPH3, Stephen M. Weiss, PhD, MPH1, Jackson Thoya, Deborah L. Jones,
PhD, M.Ed.1

1
University of Miami Miller School of Medicine, Miami, FL; 2Johns Hopkins Bloomberg School
of Public Health, Baltimore, MD; 3University of Zambia/Maimi University, Lusaka, Lusaka,
Zambia

Reducing HIV vulnerability among Zambian adolescent girls and young women

Introduction: HIV/AIDS is the leading cause of death for adolescent girls and young women
(AGYW) aged 15 to 24 in Zambia. AGYW have the highest rates of HIV infection in Zambian risk
groups (e.g., 13% girls, 7% boys). Social and economic factors, including low social and
economic status, unemployment and early marriage, make AGYW especially vulnerable to
HIV. This qualitative study examined the sociocultural context and behavioral factors placing
AGYW at risk and explored community-based solutions to prevent HIV infection among
AGYW.

Method: Community Based Participatory Research methods were used to assess HIV AGYW
risk factors and proposed community-based prevention strategies in Zambian provinces with
high HIV prevalence (Lusaka, Copperbelt, Southern Provinces). Urban and rural participants (n
= 566 males and females aged 15 - 75 years) were enrolled in focus group discussions (73
groups) and key informant interviews (n = 128), and content was coded for dominant themes.

Results: Gender inequity underlay HIV vulnerability in AGYW. The potential to complete
school and options for employment among AGYW were diminished by traditional gender roles
and inequity, resulting in sexual exchange, sex work and intergenerational sex. Sexual
practices increasing HIV risk were sustained by mens sexual preferences and dominance in
sexual decision making, and often enforced by physical violence and sexual assault. Both
genders reported sexual risk associated with alcohol use; violence was more prevalent in rural
communities. Governmental agency and NGO programs were proposed to implement
economic stimulus strategies, e.g., loans for small businesses, financial aid for AGYW students,
particularly targeting those engaged in sex work. Community sensitization for both women
and men was encouraged, with a focus on empowering women, risky traditional practices,
youth friendly clubs and support groups.
S868 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Discussion: Gender equity appears critical for HIV prevention in AGYW. Current Zambian
government initiatives target equity and the reduction of gender-based violence, but
community-based and endorsed programs are needed to promote and facilitate education
and financial security among AGYW, and reduce dependence on survival-based sexual
exchange. Community buy-in for gender equity programs, particularly in rural areas, should be
sought to reduce resistance to initiatives to enhance the status of women. While womens
voices are increasingly heard, men play a key role in bridging the gap between gender
disparities. Both male and female opinion leaders are needed to champion equity between
genders, and future research should involve both genders to establish new social norms for
women.

CORRESPONDING AUTHOR: Stefani Butts, BA, University of Miami Miller School of Medicine,
Miami, FL, 33136; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S869

Paper Session 22: HIV in Low-Resource Settings 3:55 PM-4:12 PM

MERITORIOUS AWARD WINNER

IMPACT OF A FAMILY-BASED PREVENTION PROGRAM ON ECONOMIC COMMUNICATION AND

HOUSEHOLD ECONOMICS: RESULTS OF A RANDOMIZED TRIAL

Elsa Friis, MSc1, Eric Green, PhD2, Eve Puffer, Ph.D.1

1
Duke University, Durham, NC; 2Duke Global Health Institute, Durham, NC

Objective. Evaluate effects of a family- centered HIV intervention in rural Kenya on secondary
outcomes including family economic communication and household economic behavior
including budgeting, saving, and spending on youths basic needs in rural Kenya.

Method. The intervention was developed using community-based participatory methods and
focused on strengthening family relationships to improve future developmental outcomes for
youth and reduce HIV acquisition. Modules addressed economic empowerment, emotional
support, and HIV knowledge and prevention skills using evidence-based strategies informed
by social cognitive theory and culturally grounded content. The intervention was delivered in
four churches and evaluated using a stepped wedge cluster randomized trial. Participants
included 124 families (237 adolescents ages 10-16; 203 caregivers) and interviewer-
administered surveys were completed over 5 rounds. Primary outcomes included:
engagement in income generating activities, couples economic communication, having a
household budget, amount of household savings, spending on youths basic needs, and youth-
reported material resources. Intent-to-treat effects were estimated via ordinary least squares
regression with clustered standard errors.

Results. The majority of participants initiated new income generating activities following the
intervention. The intervention significantly increased couples economic communication 1
month post intervention compared with controls as reported by females and males (effect
size, female = 0.12, p < 0.05; effect size, male = 0.09, p < 0.05). There was no effect on
household savings, but the intervention had small effects on household spending on youths
clothing (effect size = 0.37, p < 0.05) and on youth-reported ownership of shoes (effect size =
0.19, p < 0.05). No effects were observed for other indicators of spending and youth material
resources.

Conclusions. Results suggest that a family-strengthening intervention can change family

interaction patterns related to finances, a major source of conflict and risk factor for HIV
acquisition, especially in very low-resource settings. However, limited treatment effects on
S870 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

youth provision suggest that some effects on actual spending behavior may take longer to
emerge or that more focused interventions, such as direct financial assistance, are needed in
this context to effect large changes.

CORRESPONDING AUTHOR: Elsa Friis, MSc, Duke University, Durham, NC, 27707;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S871

Paper Session 23: Complementary and Integrative Approaches to Health 3:00 PM-3:18 PM

PASSIVE SPIRITUAL HEALTH LOCUS OF CONTROL IS ASSOCIATED WITH LOWER PHYSICAL

ACTIVITY LEVELS IN AN URBAN, FAITH-BASED COMMUNITY

Alyssa T. Brooks, PhD1, Samantha Louis Thomas, BS2, Gwenyth R. Wallen, PhD, RN3, Colby
Ayers, MS4, Valerie Mitchell, BA5, Tiffany Powell-Wiley, MD, MPH, FAHA6

1
National Institutes of Health Clinical Center, Bethesda, MD; 2Office of Intramural Training and
Education of the National Institutes of Health (NIH), National Institutes of Health, Bethesda,
MD; 3Clinical Center, National Institutes of Health, Bethesda, MD; 4UT Southwestern, Dallas,
TX; 5NIH/National Heart, Lung and Blood Institute, Bethesda, MD; 6National Heart Lung and
Blood Institute, National Institutes of Health, Bethesda, MD

Background: Assessing ones spiritual health locus of control (SLOC) provides a way to
examine how spiritual beliefs may be related to health outcomes based on whether an
individual has an active or passive orientation to his or her health behaviors. While an
individual with an active SLOC believes that God empowers them to take personal healthy
actions, an individual with a passive SLOC relies on God to take care of his/her personal
health.There is some evidence of SLOC being associated with health behaviors in African-
American communities. Less is known about the relationship between SLOC, physical activity
(PA) behaviors, and the use of wearable mobile health (mHealth) PA-monitoring technology.
This analysis examines the relationship between SLOC, health behaviors, and the use of
mHealth technology in a predominantly African-American church population in at-risk
Washington, D.C. communities.

Methods: This analysis was conducted using data from a community-based participatory
research (CBPR) study designed to evaluate psychosocial and environmental factors
associated with cardiovascular health in a predominantly African-American church population
in at-risk Washington, D.C. communities (NCT#01927783). Participants (n=99, 78% female,
99% African American) also received a mHealth PA monitor and were instructed to wirelessly
upload PA data weekly to church-based data collection hubs. Participants viewed PA data on
home- or church-based computers. SLOC was captured using a 13-item validated and scored
SLOC scale. Associations between PA and mHealth usage and SLOC scores were evaluated by
Spearman correlation coefficients (rs).
Results: Among participants (age: 58 11.0 years; BMI: 337 kg/m2), median score for active
and passive SLOC was 46 (interquartile range [IQR]=41-51) and 3 (IQR=2-5) points
respectively. Increasing passive SLOC was associated with less leisure-time vigorous PA in
minutes/week (rs=-0.30, p=0.04), more hours watching TV/week (rs=0.26, p=0.01), and annual
household income under $60,000 (rs=0.34, p= 0.001). There was no association between
S872 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

mHealth device usage and passive SLOC (rs=-0.08, p=0.43). In contrast, active SLOC was not
associated with PA, sedentary time, or mHealth device usage (p >0.05 for all).

Conclusions: Among faith-based communties in at-risk Washington, D.C. areas, those with
passive SLOC may be less likely to engage in leisure-time PA. These findings highlight the
potential for SLOC to identify a target population for a CBPR-designed PA intervention in this
community. SLOC does not appear to be a barrier to mHealth PA-monitoring device usage in
this population. Thus, mobile health technology may serve as a useful method to help
engage at-risk, faith-based populations from urban communities in increased PA for improved
cardiovascular health.

CORRESPONDING AUTHOR: Alyssa T. Brooks, PhD, National Institutes of Health Clinical Center,
Bethesda, MD, 20892; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S873

Paper Session 23: Complementary and Integrative Approaches to Health 3:19 PM-3:36 PM

PRELIMINARY EFFECTS OF A MINDFULNESS-BASED INTERVENTION FOR LATINO YOUTH WITH

PSYCHOLOGICAL DISTRESS

Cara Young, PHD, RN, FNP-C, Rodolfo Aguilar, RN, Adama Brown, PhD

The University of Texas at Austin, Austin, TX

Nearly one third (~30%) of United States adolescents report disturbances in their usual
activities due to feelings of sadness and hopelessness, and Latino adolescents consistently
report higher levels of depressive symptoms and disorders than their non-Hispanic White,
African American and Asian American peers. A substantial body of literature with adults
supports the effectiveness of mindfulness-based interventions in the reduction of several
forms of psychological distress, including depression. Mindfulness based studies with
adolescents suggests similar effects, although limited evidence exists regarding the impact of
mindfulness interventions with diverse populations of youth. The purpose of this study,
therefore, was to test the feasibility and preliminary effects of a mindfulness-based training
program with at-risk adolescents from a predominantly Hispanic/Latino community. A single
group, pre- and post-test intervention design was used to test a five week, group-based
mindfulness intervention with young adolescents (12-15 years) reporting high levels of
perceived stress or brooding. Participants were recruited from a pediatric primary care clinic
serving un-insured and underinsured families in the southern United States. Measures of
perceived stress, brooding, self-esteem, depressive symptoms, and mindfulness were
completed by adolescents at baseline and immediately post-intervention. Participants are
currently completing a 1 month post-intervention assessment. Seven adolescents (57%
female, 85% Hispanic/Latino) of 15 interested met study inclusion and exclusion criteria and
were able to attend the weekend sessions. 100% participation in all sessions was achieved
with two participants completing one make-up session each. Wilcoxon-Signed Rank Tests
were used to evaluate pre- to post-intervention changes in key study variables. Statistically
significant improvements were detected in depressive symptoms (Z=-2.37, p =.018), perceived
stress (Z=-2.02, p =.043), and self-esteem (Z=-2.20, p =.028). No statistically significant
changes were identified with brooding or mindfulness. Preliminary results suggest brief
mindfulness interventions may be effective at reducing depressive symptoms and perceived
stress and enhancing self-esteem in high risk adolescents. The lack of change with brooding
and mindfulness warrants further study as both have been suggested as mechanisms through
which mindfulness-based interventions exert their effect. One month post-intervention
assessments will provide additional evidence as to the sustained effects of the intervention.
S874 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Cara Young, PHD, RN, FNP-C, The University of Texas at Austin,
Austin, TX, 78701; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S875

Paper Session 23: Complementary and Integrative Approaches to Health 3:37 PM-3:54 PM

CITATION AWARD WINNER

YOGA THERAPY FOR VETERANS WITH CHRONIC LOW BACK PAIN

Erik Groessl, PhD1, Lin Liu, Ph.D.2, Laura Schmalzl, PhD3

1
University of California San Diego, La Jolla, CA; 2University of California, San Diego, La Jolla,
CA; 3University of California San Diego, Encinitas, CA

Military veterans are more likely to have chronic low back pain (CLBP) than non-veterans, and
many current treatment options such as opioid medications have limited effectiveness and/or
significant side effects. Hatha yoga has been shown to improve CLBP in non-veteran
populations, but has not been well studied among military veterans with CLBP. We present
the main results of a randomized, controlled trial (RCT) of yoga for military veterans with
CLBP. Participants were recruited through primary care and specialty clinics at VA San Diego.
Eligible patients were randomized to either yoga or a delayed treatment comparison group.
Yoga consisted of 2x weekly yoga sessions for 12 weeks. Yoga home practice was encouraged
via a manual. Both groups received usual care and were asked to not change other CLBP
treatments unless medically necessary. The primary outcome was back-specific disability
(RMDQ). Secondary outcomes included pain severity, pain interference, fatigue, quality of life,
and depression. Assessments occurred at baseline, 6-weeks, 12-weeks, and 6-months.
Multivariable random effects models were used in intent-to-treat analyses of change in health
outcomes by treatment group over time. 150 VA patients with CLBP were enrolled in 6
cohorts. Participant had a mean age of 53.4 years; 26% were women, 51% were non-White or
Hispanic, 34% were employed, and 18% were homeless in last 5 years. At baseline, 20% were
taking opioid, 15.0 years was the mean duration of CLBP, and 61% had never tried yoga. For
the primary outcome, both groups reported less disability at 12-weeks (p < 0.05), but group
differences were not significant. However, at 6-months, yoga participants had significantly
greater reductions in disability (p = 0.003). Yoga participants also had greater reductions in
pain severity, pain interference, SF-12 Physical, and fatigue at 12-weeks (p < 0.05), and at 6-
months, greater reductions in pain severity and fatigue (p < 0.05). Attendance was lower than
expected, with 53% of yoga participants attending at least 12 of the 24 sessions. A main effect
of reduced use of opioid medications and other medical pain treatments was also found. In
conclusion, although some veterans receiving VA care faced numerous challenges with
attending the 2x weekly yoga sessions, intent-to-treat analyses indicate that yoga participants
reported better outcomes after 12 weeks and after 6 months. Efforts to increase accessibility
and facilitate attendance may result in stronger effects in the future.
S876 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Erik Groessl, PhD, University of California San Diego, La Jolla, CA,
92093; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S877

Paper Session 23: Complementary and Integrative Approaches to Health 3:55 PM-4:12 PM

CITATION AWARD AND MERITORIOUS WINNER

SELF-COMPASSION AND MINDFULNESS: THE RELATIONSHIP WITH DISORDERED EATING,

APPEARANCE-MOTIVATED EXERCISE, AND DEPRESSION

Mandrila Das, B.A., Steven Richards, PhD Clinical Psychology, Brandy Pia-Watson, Ph.D.

Texas Tech University, Lubbock, TX

Background: Disordered eating is a prevalent problem among college students in the United
States (Eisenberg, Nicklett, Roeder, & Kirz, 2011). Past research shows that the percentage of
students eating according to a special weight loss diet increased from 4.2% in 1995 to 22% in
2008 (White, Reynolds-Malear, & Cordero, 2011). Although concerns with weight and body
appearance have become readily accepted by society, they can often develop into serious
problems, which affect mental and physical health. The present study examined the role of
self-compassion and mindfulness in predicting disordered eating behaviors, depression, and
appearance-motivated exercise among college students. Methods: Participants were 732
college students, both male (46%) and female (54%), ages 18-25 (M=19). Hierarchical
regression analysis was conducted to test main and interaction effects of self-compassion and
mindfulness on disordered eating behaviors, appearance-motivated exercise, and depression.

Results: Self-compassion was a significant negative predictor of disordered eating behaviors,

appearance-motivated exercise, and depression. Mindfulness was a significant negative
predictor of disordered eating and depression, and a partial negative predictor of appearance-
motivated exercise.

Conclusion: Results of this study suggest that self-compassion and mindfulness may serve as
protective factors for disordered eating, depression, and appearance-motivated exercise. The
present study provides support for the advancement of self-compassion and mindfulness
research. These findings build on previous research suggesting that self-compassion and
mindfulness may reduce the risk of disordered eating behaviors and depressive
symptomology. Additionally, self-compassion may reduce the desire to exercise for
appearance-motivated reasons. One possible explanation, which builds on previous research,
is that self-compassion promotes body acceptance, which in turn reduces the need to exercise
to control ones appearance. While research on self-compassion in the exercise domain is
fairly limited, this study provides initial support for further research in this area.
S878 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Mandrila Das, B.A., Texas Tech University, Lubbock, TX, 79415;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S879

Paper Session 23: Complementary and Integrative Approaches to Health 4:13 PM-4:30 PM

RELIGIOUS COMMITMENT MODERATES THE RELATIONSHIP BETWEEN SYMPTOM BURDEN

AND POSITIVE AFFECT

Andrea L. Canada, Ph.D.1, Peter Hill, PhD2, Neal Krause, Ph.D. 3, Gail Ironson, MD, PhD4

1
Biola University, La Mirada, CA; 2Rosemead School of Psych., Biola University, La Mirada, CA;
3
University of Michigan, Bloomington, IN; 4University of Miami, Miami, FL

Religious Commitment Moderates the Relationship Between Symptom Burden and Positive
Affect

Background: Religion/spirituality has consistently been linked to positive affect; conversely,

physical symptom burden has demonstrated an association with negative emotion. It is not
clear, however, if religion/spirituality offers protection against the deleterious effects of
higher symptom burden levels on positive affect.

Methods: In the Landmark Spirituality and Health Survey study, a representative sample of
the U.S. population (N=3,010), completed a survey measuring physical and mental health
indices as well as a variety of religious/spiritual dimensions. Specific variables of interest
included physical health symptoms, religious commitment, and the positive emotions of
gratitude, hope, and meaning in life. Hierarchical regression analyses were used to test if
religious commitment moderated the associations between symptom burden and positive
affect.

Results: Controlling for covariates, religious commitment and symptom burden were
significantly associated with each positive affect (ps < .001). Entering the interaction term
between religious commitment and symptom burden further accounted for a significant
proportion of the variance in gratitude (R2 = .003, F(1, 2742) = 6.667, b = .017, t = 3.110, p =
.002), hope (R2 = .005, F(1, 2695) = 15.501, b = .027, t = 3.937, p < .001), and meaning in life
(R2 = .004, F(1, 2680) = 13.148, b = .035, t = 3.626, p < .001).

Conclusion: Participants with higher levels of religious commitment had greater gratitude,
hope, and meaning in life than those with less religious commitment in the context of high
symptom burden. As such, the religious commitment should be supported within the
healthcare setting.

CORRESPONDING AUTHOR: Andrea L. Canada, Ph.D., Biola University, La Mirada, CA, 90639;
[email protected]
S880 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 24: Integrated Primary Care as a Vehicle for Behavioral Medicine Excellence
3:00 PM-3:18 PM

BRIEF BEHAVIORAL INTERVENTIONS FOR BEHAVIORAL MEDICINE CONCERNS FOR USE IN

ADULT PRIMARY CARE SETTINGS: A SYSTEMATIC REVIEW

Jennifer S. Funderburk, PhD1, Robyn L. Shepardson, PhD1, Jennifer M. Wray, PhD2, John Acker,
M.S.3, Kyle Possemato, Ph.D.4, Gregory Beehler, PhD, MA5, Laura Wray, PhD6

1
VA Center for Integrated Healthcare, Syracuse, NY; 2Ralph H. Johnson VA Medical Center,
Buffalo, NY; 3Syracuse VA Medical Center, Syracuse, NY; 4Syracuse Veterans Affairs Medical
Center, Syracuse, NY; 5VA Western New York Healthcare System, Buffalo, NY; 6VA Center for
Integrated Healthcare, Buffalo, NY

Large healthcare organizations (e.g., Department of Defense, Veterans Health Administration)

are embracing the Primary Care Behavioral Health (PCBH) model of integrated primary care.
Although evidence-based behavioral interventions exist that target a variety of behavioral
medicine concerns, integrated behavioral health providers need interventions that are
sufficiently brief (i.e., 6 sessions) to be compatible with the PCBH model. Therefore, we
conducted a systematic review of published studies examining behavioral interventions
targeting behavioral medicine concerns that can be employed by PCBH providers in adult
primary care settings. Our literature search targeted behavioral medicine concerns identified
as prevalent within primary care or among patients seen in PCBH. A total of 67 published
articles representing 63 original studies met eligibility criteria. We extracted data on the
behavioral interventions employed, results comparing the active intervention to a comparison
group, general fit with the PCBH model, and methodological quality. The vast majority of
studies (n=47) examined brief interventions targeting either sleep difficulties or increasing
physical activity, with no studies targeting sexual health or sleep apnea, and 5 studies
targeting each of the remaining domains. The most commonly employed interventions
comprised techniques derived from cognitive behavioral therapy (CBT) and motivational
interviewing (MI). Outcomes were generally statistically significantly in favor of the active
intervention groups relative to comparison groups, with highly variable methodological
quality ratings (range=0-5; M=2.0). Results will be discussed in relation to the need for more
evidence for brief behavioral interventions targeting other behavioral medicine concerns, as
well as the need for more specificity regarding the compatibility of such interventions with the
PCBH model of care.

CORRESPONDING AUTHOR: Jennifer S. Funderburk, PhD, VA Center for Integrated Healthcare,

Syracuse, NY, 13210; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S881

Paper Session 24: Integrated Primary Care as a Vehicle for Behavioral Medicine Excellence
3:19 PM-3:36 PM

MEASURING THE NEED AND IMPACT OF BEHAVIORAL HEALTH CONSULTANTS IN PRIMARY

CARE

William Sieber, PhD1, Alexandra Lambeck, B.S.1, Zephon Lister, PhD, LMFT2, Katrin Seifert,
PsyD3

1
UC San Diego, La Jolla, CA; 2University of California, San Diego, Menifee, CA; 3University of
California San Diego, La Jolla, CA

One proposed benefit of collaborative care, or integrated behavioral health, is that warm
hand-offs from primary care physicians (PCPs) to behavioral health providers (BHPs) increase
the chance that the patient in need will follow through with a referral for specialty mental
health care. However a research gap exists estimating the actual increase in the rate of follow-
through that such a warm hand-off provides, as well as which factors moderate this impact.
In fact few if any research reports have documented what evidence-based activities the BHP
provides. A unique opportunity to address this research gap was presented when BHPs were
available only during certain times across 3 primary care outpatient clinics operating within a
larger university-based healthcare system. While a BHP is trained to provide evidence-based
brief interventions to patients, oftentimes they also support a referral from the PCP to
psychotherapy or health coaching, if warranted. However, there is a paucity of research
literature quantifying the various activities of BHPs, nor of the frequency of their use, or the
impact on patient completion of a referral for more extended care.

Data was extracted from an EHR on all patients presenting to any of 3 primary care clinics
between January 1, 2016 and May 31, 2016, with follow-up data on visits for specialty mental
health care extracted through July 31, 2016. A total of 14,748 patient visits resulted in a total
of 649 referrals for psychotherapy, of which 216 had involved a BHP warm hand-off at the
time of referral. One-way ANCOVAs showed that patients referred for therapy associated
with a BHP warm hand-off were more likely to be seen by a psychotherapist within 60 days of
referral than patients referred without such BHP involvement (p < .05). Regression analyses
suggest that the factors most predictive of patient follow-through with a referral were BHP
warm hand-off, female gender, and diagnosis of patient by PCP at time of referral (PTSD &
anxiety > depression & stress, p < .05).
S882 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Overall, BHPs were more often called upon by PCPs when the patient was non-white, male,
and on private insurance. Overall, we found a significant impact of a warm hand-off on
referral follow-through rates (p < .05) but more often when the patient was female, an ethnic
minority, or had a specific DSM diagnosis already in their medical record. Analyses also
revealed that when BHPs were available, they were most often used by PCPs to assess suicide
risk, provide information on sleep, use motivationla interviewing and SMART goals for weight
loss and physical activity, address medication adherence barriers, or increase support for
follow-through for specialty mental health care.

This demonstrates that the benefit of short behavioral medicine interventions is not that they
can help all patients equally, but rather that they can have a substantial impact on certain
subsets of the community. Further analsyes on the different activities of the BHPs and the
implications of this on further development of integrated behavioral health services will be
discussed.

CORRESPONDING AUTHOR: William Sieber, PhD, UC San Diego, La Jolla, CA, 92093-0807;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S883

Paper Session 24: Integrated Primary Care as a Vehicle for Behavioral Medicine Excellence
3:37 PM-3:54 PM

FACTORS INFLUENCING SUICIDE RISK ASSESSMENT AND MANAGEMENT IN VA PRIMARY CARE

SETTINGS

Megan Petrik, PhD, LP1, Benjamin F. Miller, PsyD2, Bridget Matarazzo, PsyD3

1
University of Minnesota Medical School, Minneapolis, MN; 2Eugene S. Farley, Jr. Health
Policy Center, Aurora, CO; 3Rocky Mountain MIRECC, Denver, CO

Background: Individuals who die by suicide are commonly seen in primary care in the months
prior to death (Ahmedani et al., 2014). One tool to improve primary care suicide prevention is
the Veterans Affairs/Department of Defense Clinical Practice Guideline for the Assessment
and Management of Patients at Risk for Suicide (2013; VA/DoD Suicide Risk CPG), as it offers
an algorithm for evidence-based assessment and management of suicide risk in primary care.
CPGs alone minimally change provider behavior. Thus, understanding the context of the
practice environment, which can be obtained through stakeholder viewpoints, is essential to
successful CPG implementation (Stetler et al., 2011).

Aim: Obtain VA primary care provider (PCP) perspectives regarding the barriers and
facilitators of suicide risk assessment and management and the impact of integrated mental
health services on PCP suicide risk assessment and management practices.

Methods: Eleven VA PCPs in four states (AK, CO, UT, VT) participated. PCPs completed a semi-
structured qualitative interview that took on average 26 minutes (SD = 8.6). Data was
analyzed using a hybrid inductive-deductive thematic analysis approach.

Results: Qualitative analysis identified the following themes: (1) suicide risk assessment and
management is most effective when conducted as a collaborative, team-based endeavor; (2)
integrated mental health providers are an outlet for suicide prevention training; (3) variability
in access to mental health follow-up care impacts management; and (4) VA-system suicide
prevention efforts supports generalists ability to assess and manage suicide risk.

Discussion: Results from this multi-site qualitative study produced cross-cutting themes
regarding the barriers and facilitators of suicide prevention in VA primary care settings.
Nuanced implementation implications that are informed by the results will be discussed.
Adherence to the VA/DoD Suicide Risk CPG would likely enhance PCPs delivery of evidence-
based care for suicide risk in primary care and may reduce Veteran deaths by suicide.
S884 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Megan Petrik, PhD, LP, University of Minnesota Medical School,
Minneapolis, MN, 55455; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S885

Paper Session 24: Integrated Primary Care as a Vehicle for Behavioral Medicine Excellence
3:55 PM-4:12 PM

PHYSICIAN MANAGEMENT OF BEHAVIORAL HEALTH CONCERNS IN PRIMARY CARE PATIENTS

AND ITS IMPLICATION FOR RESOURCE UTILIZATION

Marisa Sklar, PhD1, Cara Fuchs, PhD2, Daniel R. Evans, Ph.D.3, Lisa A. Uebelacker, PhD4, Justin
Nash, PhD5

1
Brown University, providence, RI; 2Brigham and Women's Hospital, Jamaica Plain, MA;
3
University of Washington Medical Center, Seattle, WA; 4Brown University and Butler
Hospital, Providence, RI; 5University of Connecticut, Storrs, CT

Although behavioral health services are increasingly offered in primary care, physicians
continue to manage behavioral health needs in most patients. Understanding physicians
ability to manage behavioral health concerns in patients can guide the design of behavioral
health services in primary care. We surveyed physicians perceived ability to diagnose and
manage depression, anxiety, and bipolar disorders in their patients. Physicians included
residents in training and attending faculty from a family medicine residency practice in an
underserved community.

Repeated measures General Linear Models tested main effects of disorder type and provider
type, and their interaction effects. Separate analyses compared responses by resident to
faculty providers (noted by FvsR), and by resident year (noted by Ryear). Special contrast follow-
ups tested differences by disorder type, and polynomial trend follow-ups tested differences
by resident year. Some significant results follow. Main effects of disorder type ((FFvsR(2,
42)=24.45, p < .001); (FRyear(2, 28)=20.38, p < .001)), provider type (F(1, 21)=5.86, p=.025), and
resident year (F(2, 14)=4.67, p=.028) were found for confidence diagnosing. Providers were
more comfortable diagnosing depression and anxiety than bipolar disorders ((FFvsR (1,
21)=29.84, p < .001); (FRyear (1, 14)=23.79, p < .001)). Faculty were more comfortable
diagnosing than residents (F(1, 21)=5.86, p=.025), and 2nd year residents were more
comfortable diagnosing than 1st and 3rd years (F(1, 14)=9.29, p=.009). A main effect of
disorder type was found for confidence personally providing therapy ((FFvsR (2, 42)=19.37, p <
.001); (FRyear (2, 28)=20.52, p < .001)). Providers were significantly more comfortable
providing therapy for depression and anxiety than bipolar disorders ((FFvsR (1, 21)=25.56, p <
.001); (FRyear (1, 14)=30.84, p < .001)).
S886 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Providers of all levels of experience express less comfort diagnosing and managing bipolar
than depression and anxiety disorders. Thus, behavioral health specialists in primary care
practices may be best used as consultants to providers in their management of depression
and anxiety, and as direct providers for more challenging bipolar and other mental health
concerns.

CORRESPONDING AUTHOR: Marisa Sklar, PhD, Brown University, providence, RI, 02906;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S887

Paper Session 24: Integrated Primary Care as a Vehicle for Behavioral Medicine Excellence
4:13 PM-4:30 PM

CITATION AWARD WINNER

CUSTOMIZING DEPRESSION CARE FOR PATIENT COMPLEXITY:: AN APPROACH TO EXPAND THE

EVIDENCE BASE FOR INTEGRATED PRIMARY CARE

Duncan Campbell, PhD1, Edmund Chaney, PhD2, Thomas Waltz, PhD3, Anayansi Lombardero,
PhD4, Laura Bonner, PhD5, Barbara F. Simon, MA6, Bolkan Cory, PhD7, Kara Zivin, PhD8, Teri
Davis, PhD6, Alissa R. Simon, MA9, Lisa V. Rubenstein, MD, MHS10

1
University of Montana, Missoula, MT; 2VA Puget Sound Health Care System, Bandon, OR;
3
Eastern Michigan University, Ypsilanti, MI; 4New Mexico VA Health Care System,
Albuquerque, NM; 5VA Puget Sound Healthcare System, Seattle, WA; 6VA Greater Los Angeles
Healthcare System, Los Angeles, CA; 7Washington State University Vancouver, Vancouver,
WA; 8VA Ann Arbor Healthcare System, Ann Arbor, MI; 9VA Greater Los Angeles Healthcare
System, North Hills, CA; 10VA Greater Los Angeles Healthcare System, UCLA and RAND, Los
Angeles, CA

Objectives: Complex biopsychosocial factors other than depression symptoms can complicate
depression care and impact patient outcomes. Yet depression guidelines provide little
direction on addressing these factors, particularly in the primary care settings where most
depressed patients are detected and treated. We aimed to identify complexity dimensions
or areas of focus beyond depression itselfand strategies for addressing these, while relying
on empirical data. We aimed to prioritize complexity dimensions and strategies for
incorporation in Veterans Affairs (VA) depression care guidelines through a key stakeholder
expert panel.

Methods: We reviewed the relevant literature and products from a multisite longitudinal
study of depression among Veterans in primary care to identify complexity dimensions for
depression management and to generate clinical management strategies. We identified
employment, social support, comorbidity, stigma, patient preferences, faith/spirituality,
women Veterans' needs, and ethnicity as key dimensions and identified 50 clinical strategies
for addressing them. We convened an expert panel of 13 VA stakeholders with diverse clinical,
administrative and research expertise who rated, discussed and re-rated the importance and
feasibility of each complexity dimension and strategy. Importance ratings ranged from 1=low
to 7=high.
S888 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Results: Patients treatment preferences (6.3, SD=.86), social support (6.2, SD=.90), and
employment (5.9, SD=.86) were rated as the most important complexity dimensions for
accommodation in primary care-based depression treatment. Ratings showed agreement on
some strategies for addressing complexity dimensions but uncertainty about others. Expert
panelists' commentary provided guidance on implementation barriers and feasibility of
specific strategies and reflected a tension between the need to address the dimensions and
uncertainty about which specific strategies were feasible and effective for doing so.

Implications: Our VA expert panel identified specific patient complexity dimensions as critical
for achieving optimal depression care outcomes in primary care. Consistent with recent work
in multimorbidity/patient complexity, our panelists ascribed high importance to incorporating
treatment preferences, social support, and employment considerations in depression
guidelines. While panelists agreed on some potential management strategies (e.g., educating
care managers about the impact of complexity on clinical status) our results suggest an urgent
need for further development of standard methods for addressing key complexity
dimensions.

CORRESPONDING AUTHOR: Duncan Campbell, PhD, University of Montana, Missoula, MT,

59812; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S889

Paper Session 25: Interventions for Pain Management 3:00 PM-3:18 PM

CITATION AWARD WINNER

ACCEPTANCE AND COMMITMENT THERAPY VS. MEDICAL TREATMENT AS USUAL WAIT-LIST

CONTROL GROUP FOR PRIMARY HEADACHES: THE ALGEA STUDY

Maria Karekla, Ph.D., Vasilis Vasileiou, M.A.

University of Cyprus, Nicosia, Nicosia, Cyprus

Headache sufferers commonly use experiential avoidance as a coping mechanism to manage

pain or other private events (e.g. thoughts) related to their headache difficulties. However,
avoidance of headache triggers or of other headache-related internal experiences are found
to instead increase trigger potency, restrict lifestyle, decrease internal locus of control, and
exacerbate and maintain pain perception. New treatment approaches, such as Acceptance
and Commitment Therapy (ACT; Hayes et al., 2012), emphasize acceptance and valued-living
as alternatives to avoidance. Though APA has included ACT as an empirically supported
treatment for chronic pain, very little evidence exists as to its effectiveness for head pain. This
paper presents the results of a Randomized control trial comparing an ACT-based intervention
(added to medical treatment as usual-MTAU) compared to MTAU only control for primary
headache. Secondly, it investigates whether the proposed ACT process of change mechanisms
(i.e. acceptance and values-based actions) mediate headache interference and treatment
outcomes (quality of life). 120 headache sufferers were randomly assigned to either 9
sessions of ACT or to MTAU, and completed questionnaire at three point assessments (pre,
post, follow up periods). Results demonstrate significant reductions in headache-related
interference, increases in emotional and physical functioning and improvements in quality of
life in the ACT group compared to control. Mediation analysis show ACT to work through its
proposed mechanisms of action (acceptance, defusion and values consistent living). This study
provides new evidence regarding the utility of ACT in the management of primary headaches.

CORRESPONDING AUTHOR: Maria Karekla, Ph.D., University of Cyprus, Nicosia, Nicosia, 1678;
[email protected]
S890 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 25: Interventions for Pain Management 3:19 PM-3:36 PM

TREATMENT OUTCOMES FOR AN INTERDISCIPLINARY PAIN REHABILITATION PROGRAM:

COMPARISON OF OUTCOMES BASED ON OPIOID USE STATUS

Wesley Gilliam, Ph.D., L.P., Eleshia J. Morrison, PhD, L.P., Jeannie Sperry, PhD, ABPP

Mayo Clinic, Rochester, MN

Introduction: Chronic pain is a major public health concern with profound negative effects on
the well-being of millions of Americans, and widespread use of prescription opioids for
chronic pain has contributed to the escalating problem of opioid use disorder. Medically-
directed opioid tapering is increasingly recognized as an unmet need for patients with chronic
pain who desire to discontinue long-term opioid therapy. Multidisciplinary pain rehabilitation
programs can be highly effective in discontinuing opioids in patients with chronic pain while
also improving patients functional status. This study sought to examine treatment outcomes
among two cohorts of patients enrolled in a three-week interdisciplinary pain rehabilitation
program: Patients engaged in interdisciplinary pain treatment + physician supervised opioid
taper vs. nonopioid users engaged in interdisciplinary treatment.

Methods: Mixed model ANOVAs 2 (Group: opioid use, no opioid use) x 2 (Time: admission,
discharge) were used to compare groups on pre- to post-treatment changes on self-report
(pain severity, interference, depression, catastrophizing, performance in ADLs) and observer-
rated outcomes (aerobic capacity, trunk strength) among 285 treated patients.

Results: All patients in the opioid cohort successfully completed their opioid taper (mean
MME = 79). For self-report measures, significant main effects for time were detected for all
self-report outcomes (Fs > 247.0; ps < .001; n2 >.46). No significant main effect of opioid use
or opioid use x time or interaction effects detected. Similarly, for observer-rated outcomes,
significant main effects for time were detected (Fs > 124.2; ps < .001; n 2 >.33), while no main
effect of opioid use or opioid use x time effects were detected. Self-report treatment gains
were maintained at 6 month follow-up (Fs >41.4; ps < .001; n2 >.44). Patients in the opioid
cohort also reported significant pre- to post-treatment reduction in opioid craving (F= 20.6; p2
=.16)

Conclusion: Interdisciplinary pain programs that incorporate opioid withdrawal demonstrate

significant improvements in functional and psychological sequelae of chronic pain,
irrespective of opioid use status. More programs utilizing evidence-based approaches are
needed to address this significant public health issue.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S891

CORRESPONDING AUTHOR: Wesley Gilliam, Ph.D., L.P., Mayo Clinic, Rochester, MN, 55905;
[email protected]
S892 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 25: Interventions for Pain Management 3:37 PM-3:54 PM

CO-MORBID ADDICTION AND PAIN: DEVELOPING A CBT TREATMENT TO ALTER PSYCHO-

PHYSIOLOGICAL PAIN RESPONSE

Amy Wachholtz, PhD, MDiv, MS

University of Colorado Denver, Denver, CO

Aims: The purpose of this study is to evaluate STOP (Self-regulation Therapy for Opioid
addiction and Pain) a novel rolling-entry group therapy intervention developed directly from a
previous study identifying psycho-physiological treatment needs of patients with comorbid
pain and opioid addiction. Method: 16 individuals with diagnosed long term chronic pain and
opioid addiction (2 + years) on stable doses of methadone or buprenorphine for opioid
addiction were recruited from local addiction treatment centers. Participants completed a
psychological assessment and a cold water task to assess pain sensitivity at 3 time points pre-
post- and at a 3 month followup. Participants completed the 12 session weekly 90-minute
group therapy intervention as part of their MMT/Bup treatment agreement with their local
treatment centers. Urine toxicology was given weekly. Repeat measure ANOVAs were
performed on pain, activity, and illicit drug cravings and use. Results: The study showed high
feasibility and acceptability among patients with an 80% attendance rate. There was no illicit
drug use, as confirmed by urine toxicology and self-report, after week 8. Pain tolerance
(seconds) to a laboratory pain trigger significantly improved from Pre (M=38.3, SD=7.10) to 3
month follow up which approached the national average (M=123.5, SE=11.09; pConclusion:
STOP, a novel 12-week, rolling entry group therapy based on the results of a
psychophysiological needs assessment of individuals with co-morbid pain and opioid
addiction, has early evidence of high feasibility, acceptability, and effectiveness. Focusing on
improving pain self-efficacy may be a key to successful treatment for patients with comorbid
pain and opioid addiction. Future research should examine if substance abuse clinicians
without experience in pain can be taught to use STOP in a community treatment setting.

CORRESPONDING AUTHOR: Amy Wachholtz, PhD, MDiv, MS, University of Colorado Denver,
Denver, CO, 80217-3364; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S893

Paper Session 25: Interventions for Pain Management 3:55 PM-4:12 PM

A PATIENT DECISION AID TO REDUCE DEMAND FOR OPIOIDS

Andrea Gielen, ScD ScM

Johns Hopkins Bloomberg School of Public Health, Baltimore, MD

Background We developed a patient decision aid (PDA), a web-based, patient-centered

educational intervention that helps patients have an informed discussion about pain
medication with their ED provider. We hypothesized that better informed patients would be
less likely to prefer an opioid pain reliever (OPR). This presentation will describe the formative
research used to develop the PDA, the content of the PDA, and the RCT pilot test protocol for
the two test sites (Maryland, West Virginia) with preliminary results.

Methods. In-depth interviews were conducted with 13 ED clinicians, 6 nationally recognized

pain experts, and 17 recently discharged patients. Interviews were transcribed and coded for
analysis. Two rounds of beta testing of the PDA were conducted with 10 ED patients. For the
pilot test, 144 participants are recruited in the ED waiting rooms.

Results Clinicians uniformly expressed the belief that an OPR PDA could be useful as a patient
centered educational tool and a supplement to their treatment planning. Patients reported
receiving little to no information on OPRs during their most recent ED visit, and they would
have liked more information. Beta test participants thought the tool was useful and would be
appreciated by patients waiting to be seen. They provided suggestions, which we addressed
through improving the readability of the printed information and the graphics. The pilot
testing protocol involves random assignment of eligible patients (adult, chief complaint of
pain, acute injury, pain score 7-10, no current use of OPR, English-speaking) to complete the
PDA or a health risk assessment (HRA), both completed on a tablet prior to seeing the treating
provider. The PDA provides tailored feedback for use by the patient and provider. Measures
taken before and after interacting with the PDA/HRA, after the ED visit, and 6 weeks later
include: decisional conflict; knowledge; pain management preference; pain scores; patient
satisfaction; and use, storage, and disposal of medication if prescribed an OPR. The trial will
run from September, 2016-December, 2016, with outcome data available at the conference.
S894 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions Creating a patient-centered decision aid benefits from formative research.

Computer technology can efficiently provide tailored patient education, which should
facilitate shared decision making for pain management.

CORRESPONDING AUTHOR: Andrea Gielen, ScD ScM, Johns Hopkins Bloomberg School of
Public Health, Baltimore, MD, 21205; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S895

Paper Session 25: Interventions for Pain Management 4:13 PM-4:30 PM

INTRODUCTION TO PROGRESSIVE MUSCLE RELAXATION THERAPY FOR MIGRAINE IN THE

EMERGENCY DEPARTMENT: A PILOT FEASIBILITY STUDY

Mia Minen, MD, MPH1, Alexandra Boubour, N/A2, Scott W. Powers, PhD3

1
NYU Langone Medical Center, New York, NY; 2Barnard College of Columbia University, New
York, NY; 3Cincinnati Children's Hospital and University of Cincinnati College of Medicine,
Cincinnati, OH

Background: There are 5 million visits to US emergency departments (ED)/year for

migraine/headache. Though many of these patients would benefit from behavioral or
pharmacologic preventive therapy these teachable moments are usually lost. Herein we
assess whether ED patients with migraine would agree to progressive muscle relaxation
(PMR), a proven therapy amenable to introduction in the ED and independent practice at
home. Objective: We sought to examine whether migraine patients would agree to try PMR
while in the ED, and for how long they would engage in an introductory session. Methods:
Research assistants in our urban academic ED introduced the study concept to eligible
migraine patients in the ED. Patients completed a survey about their headache history and
were asked to listen to a 20-minute audio recording of PMR. Results: We approached 11
patients and 8 agreed to participate (73%). Mean age of participants was 35.5 (13.17) years
and 63% were female. Headaches occurred on 10.7 days/month on average and were usually
severe (63%) or moderate (25%). No one had previously done PMR though two had
previously tried biofeedback. All had previously visited an ED for their headaches; mean
number of prior ED visits was 2.38 (range 1-5). Half had seen a neurologist for treatment of
their headaches. One person did not do the PMR session in the ED. Of the remaining seven
who did the PMR, the mean time spent doing it was 7.6 minutes (range 1-20). Conclusions:
Migraine patients presenting to the ED have severe and frequent migraines which make them
candidates for preventive behavioral treatments. PMR is a behavioral technique which can be
introduced in the ED setting. Follow-up studies should examine whether patients will continue
to use these behavioral techniques at home.
S896 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Mia Minen, MD, MPH, NYU Langone Medical Center, New York,
NY, 10016; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S897

B001 6:00 PM-7:00 PM

DETERMINANTS OF PARTNER SUPPORT FOR PHYSICAL ACTIVITY: A COUPLES-FOCUSED

ANALYSIS AMONG OLDER ADULTS WITH OSTEOARTHRITIS

Heather Wasser, PhD MPH RDN IBCLC1, Christine Rini, PhD2

1
University of North Carolina, Chapel Hill, NC; 2University of North Carolina at Chapel Hill,
Chapel Hill, NC

Objective: Physical activity (PA) can decrease pain and improve function among older adults
with osteoarthritis (OA), yet most people with OA are insufficiently active. While partner
support for PA is one of the most robust predictors of PA, research on the determinants of
partner support is limited. Using data from the Partners in Active Living Study (PALS), we
aimed to identify individual and household characteristics associated with the quantity of
positive and negative partner support for PA. Methods: Data are from baseline surveys of
main participants (mains) and support partners (partners) participating in PALS, a 12-month
longitudinal dyadic study of 170 couples in central North Carolina. Inclusion criteria for mains
include self-report of OA, ability to walk unaided, insufficient levels of PA (< 120 minutes of
PA/week), and marital or cohabiting status with a partner who is willing to participate.
Partners may or may not have osteoarthritis. Dependent variables are amount of positive and
negative enacted support for PA, dichotomized into high versus low using a median split.
Independent variables are individual and household characteristics and a global score for the
quality (effectiveness) of enacted support. Bivariate and multivariable models were fit using
logistic regression. Statistical significance was set at P < .05. Results: Mains and partners were
of similar age (~659 years) race/ethnicity (~80% Non-Hispanic white), and education (~55%
>4y college). Mains were significantly more likely than partners to be female (64.71% versus
36.47%, P<.001). In bivariate analyses, support effectiveness, the partner being female, and
the main being female were significantly associated with positive enacted support for PA
(odds ratio [OR]=3.80, [95% confidence interval (CI): 2.00-7.19]), OR=2.91, 95% CI: 1.51-5.58,
and OR=0.38, 95% CI: 0.20-0.73, respectively). These associations remained significant in
multivariable analyses. Partner stiffness and greater depression levels in couples were
significantly associated with negative enacted support for PA in bivariate analyses (OR=1.90,
95% CI: 1.03-3.52 and OR=2.48, 95% CI: 1.04-5.89, respectively), but became nonsignificant in
adjusted models. Conclusion: Identifying determinants of enacted social support, a consistent
predictor of PA, is important information for the targeting and tailoring of behavioral
interventions. Our findings demonstrate the importance of the quality (effectiveness) of
support provided and suggests women and men may differ in their ability to provide positive
support for PA, with men potentially needing more guidance. Future research among larger,
more demographically diverse samples is warranted.
S898 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Heather Wasser, PhD MPH RDN IBCLC, University of North
Carolina, Chapel Hill, NC, 27510; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S899

B002 6:00 PM-7:00 PM

DIFFERENCES AMONG US: UNDERSTANDING THE IMPACT OF BEHAVIORAL VARIATION ON

COACHING ENGAGEMENT

Anne Wallace, PhD1, Matt Bogard, M.S.2, Varun Kukreja, MBA3, Susan M. Zbikowski, PhD4

1
Humana, Cincinnati, OH; 2Humana, Woodburn, KY; 3Humana, Irving, TX; 4Humana, Seattle,
WA

BACKGROUND: Health coaching is a key intervention in combatting the prevalence of lifestyle-

driven conditions. Little research, however, has examined differences in how people progress
through such programs. The current purpose was to examine whether individuals with
different behavioral profiles (based on demographic and health-related behaviors), differ
regarding their activity and outcomes in a coaching program.

METHOD: Humana created the behavioral profiles to improve how it engages and intervenes
with members in clinical and wellness services. Exploratory analyses of participants enrolled in
health coaching between2012-2016 indicated 4 of 8 total behavioral profiles had different
levels of participation than the total population. These analyses examined the first 60 days of
coaching for the 4 segments (N=3761) Overwhelmed Guidance Seeker (OGS), Participating
Realist (PR), Selfless Support Seeker (SSS), and Simply Compliant (SC). Significance tests by
behavioral profile in number/type of interactions, number of goals set/complete, and
biometric outcomes were conducted, as well as survival analyses of probability/time to goal
completion.

RESULTS: Behavioral segments varied significantly by modality: SC were more likely to work
online than PR (p < 0.001) and SSS (p < 0.05); PR more likely to use telephonic/online mix than
OGS (p < 0.01) or SC (p < 0.0002). SSS were less likely to interact with a coach than other
profiles (ps < 0.05). The PR and OGS segments set (p < 0.001) and completed (p < 0.0001)
more goals than SC and SSS segments. Survival analysis indicated PR had higher probability of
goal completion/shorter time to completion than other segments; OGS had higher probability
of goal completion/shorter time to completion than SSS or SC (ps < 0.002). Participants in
coaching most commonly work on managing weight; analyses of weight loss by segment
revealed no group differences (p=0.579).

DISCUSSION: Understanding how different behavioral profiles respond to coaching enables

practitioners to implement tailored approaches that better support participation and progress
among individuals in groups (Selfless Support Seekers and Simply Compliant), who tend to
S900 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

participate less and achieve fewer goals. Next steps are to test which tailored interventions
are most effective.

CORRESPONDING AUTHOR: Anne Wallace, PhD, Humana, Cincinnati, OH, 45213;

[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S901

B003 6:00 PM-7:00 PM

DISMANTLING THE THEORY OF PLANNED BEHAVIOR: EVALUATING CHANGE IN ATTITUDES,

NORMS, AND PERCEIVED BEHAVIORAL CONTROL

Erika Montanaro, PhD

University of North Carolina Charlotte, Charlotte, NC

Background: It is still largely unknown which theoretical constructs in a particular theory (e.g.,
the Theory of Planned Behavior) are the active ingredients of change. The lack of a solid
understanding of how key theoretical constructs interact to motivate behavior change may be
partly to blame. Purpose: The current study examines the utility of each of the hypothesized
determinants of behavior in the Theory of Planned Behavior (i.e., attitudes, norms, perceived
behavioral control, and intentions) and explores the optimal combination (single construct
interventions vs. three constructs vs. four) of these constructs in an intervention to increase
condom use intentions and behavior among college students. Methods: 317 participants
(Mage = 19.31; SDage = 1.31; 53.3% female; 74.1% Caucasian) were randomly assigned to one
of seven computer-based interventions. Interventions were designed using the Theory of
Planned Behavior as the guiding theoretical framework. 196 (61.8%) completed behavioral
follow-up assessments three-months later. Results: In a meditational model, we found that
the four construct intervention was marginally better at changing intentions (estimate = -.06,
SE = .03, p = .06), but the single construct interventions were related to risky sexual behavior
at follow-up (estimate = .04, SE = .02, p = .05). Conclusions: This study suggests that many of
the constructs manipulated here are conceptually and empirically interconnected. Perhaps
these constructs synergistically work together in order to produce change. Further research is
necessary to develop the optimal combination of constructs in the Theory of Planned
Behavior and their ordering of presentation in intervention materials. (ClinicalTrials.gov
number NCT# 02855489).

CORRESPONDING AUTHOR: Erika Montanaro, PhD, University of North Carolina Charlotte,

Charlotte, NC, 28223; [email protected]
S902 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B004 6:00 PM-7:00 PM

EFFECTS OF CONTEXTUALLY-TAILORED SUGGESTIONS FOR PHYSICAL ACTIVITY: THE

HEARTSTEPS MICRO-RANDOMIZED TRIAL

Predrag Klasnja, PhD1, Shawna N. Smith, PhD1, Nicholas J. Seewald, MS1, Andy Jinseok Lee,
B.S.E2, Kelly Hall, N/A1, Susan A. Murphy, Ph.D.1

1
University of Michigan, Ann Arbor, MI; 2University of Michigan, Ann Arbor, Ann Arbor, MI

Effects of contextually-tailored suggestions for physical activity: The HeartSteps micro-

randomized trial

Predrag Klasnja, Shawna Smith, Nicholas Seewald, Andy Lee, Kelly Hall, Susan A. Murphy

University of Michigan

Often the easiest way to increase physical activity is by walking more. Yet, opportunities for
walking vary throughout the day, by contexts like location, weather and busyness.
Increasingly, mHealth technologies can sense such context and deliver behavior-change
interventions at moments of opportunity or risk. HeartSteps is an Android application that
delivers contextually-tailored activity suggestions to help individuals walk more. HeartSteps
contains a library of over 500 suggestions, each tagged with a set of contexts indicating
appropriate suggestion use. Suggestions are delivered up to five times a day (decision
times)morning commute, lunch, mid-afternoon, evening commute, and after dinnerand
are tailored to user location (work, home, other), time of day, weather, and
weekday/weekend. HeartSteps includes two types of suggestions: suggestions for walking or
for disrupting sedentary behavior. We evaluated HeartSteps in a 6-week micro-randomized
trial with 37 sedentary adults. Each participant was randomized with .6 probability to receive
a suggestion at each of the five decision times each day; if a suggestion was provided, a
second equal randomization determined whether to send a walking or sedentary suggestion.
We assessed if providing a suggestion impacts participants step count in the 30 minutes
following randomization, and if that effect changes over time. GEE models estimated the
effect of delivering a suggestion on 30-min step count (logged) post-randomization,
controlling for 30-minute step count prior to randomization (logged). Results: Averaging over
study days and suggestion type, the effect of providing a suggestion was positive and
marginally significant (eb=1.14 p = .059, a 14% 30-min step count increase). Separating
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S903

suggestion types showed a significant effect for walking suggestions, increasing 30-min step
count by 24% (p=.02) on average, but not for sedentary suggestions. For an individual with the
mean 30-minute step count of 253 steps, this equates to 61 extra steps in the 30 minutes
post-randomization. When an interaction with time is added, suggestions show a large initial
effect (eb=1.66, p < .01), a 66% increase in 30-minute step count on average, that diminishes
over time (eb=.98, p < .01). This effect is also driven by walking suggestions.

CORRESPONDING AUTHOR: Predrag Klasnja, PhD, University of Michigan, Ann Arbor, MI,
48109; [email protected]
S904 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B005 6:00 PM-7:00 PM

EFFECTS OF PHYSICAL ACTIVITY GOAL ATTAINMENT ON ENGAGEMENT AND OUTCOMES IN

THE NATIONAL DIABETES PREVENTION PROGRAM

Natalie Ritchie, PhD1, Jodi Holtrop, PhD2, Havranek Edward, MD3, Jennifer Carroll, MD MPH4

1
Denver Health, denver, CO; 2University of Colorado School of Medicine, Aurora, CO; 3Denver
Health, Denver, CO; 4University of Colorado Department of Family Medicine, Aurora, CO

The National Diabetes Prevention Program (DPP) is a widely-disseminated, evidence-based

intervention to prevent or delay diabetes onset. Greater engagement in the yearlong program
is expected to yield better outcomes, but high attrition is a major concern and improved
retention strategies are needed. The National DPP curriculum, published by the Centers for
Disease Control and Prevention, is highly prescriptive for weight loss including through a
weekly goal of 150 or more minutes of physical activity. However, previous evidence suggests
it may be overly challenging and unnecessary to require a high threshold of physical activity.
We assessed the effects of failing to achieve prescribed physical activity goals on outcomes in
the National DPP. We included 406 participants who attended the National DPP at a large
healthcare organization. Participants reported not meeting physical activity goals about half of
the time on average. Across all participants, each occurrence of goal failure predicted a 25%
lower likelihood of attending the next session (OR 0.75; P < .001; 95% CI 0.64-0.87). In
adjusted models, there was an inverse relationship between the frequency of failure to
achieve physical activity goals and attendance, =-0.21 (p < .01), and a positive linear
relationship with weight gain, = 0.14 (p < .05). Similarly, there was an inverse relationship
between failing to meet physical activity goals and completing the program (OR 0.47; P < .05;
95% CI 0.24-.90) and obtaining at least five percent weight loss (OR 0.39; P < .01; 95% CI 0.19-
0.76). In contrast, there was no significant effect of not meeting physical activity goals on the
amount of weight change per session attended. Findings suggest the experience of goal
failure may be a deterrent to retention in the National DPP, yet the program was equally
efficacious in terms of weight loss per session attended regardless of meeting physical activity
goals. Results are not fully generalizable as this was an observational study with no control
group. Implications are to consider adopting a more patient-centered, individualized approach
to setting physical activity goals in the National DPP.

CORRESPONDING AUTHOR: Natalie Ritchie, PhD, Denver Health, denver, CO, 80204;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S905

B006 6:00 PM-7:00 PM

EXAMINING THE BEHAVIORAL MEDICINE STANDARD OF CARE FOR ETOH CONSULTATIONS

WITHIN TRAUMA DEPARTMENT

Kelly Lynn. Gilrain, PhD1, Alexa Hays, BA1, Stephanie Goldstein, MS1, William M. Skelton, BA,
MS, MA1, Lisa Capano-Wehrle, BS, MPH1, Jennifer Petrongolo, MSW, MA2, Steven Ross, MD,
FACS, FCCM1, Cori E. McMahon, PsyD3

1
Cooper University Hospital, Camden, NJ; 2La Salle University, Lancaster, PA; 3Cooper
University Hospital/ MD Anderson Cancer Center at Cooper, Camden, NJ

The ACOS and Committee on Trauma require patients admitted to a trauma service with
elevated Blood Alcohol Concentration (BAC) be seen for brief ETOH screening and
intervention. Research indicates there is no current standard of care (SOC) across hospitals. At
Cooper University Hospital (CUH), the Behavioral Medicine Consultation-Liaison (CL) service
has completed screens and interventions since 2014. A protocol at CUH was created based
upon the Screening, Brief Intervention and Referral of Treatment (SBIRT), which provides only
recommendations for screening and intervention. At CUH, the Alcohol Use Disorders
Identification Test (AUDIT) is used to assess alcohol use/abuse. Psychosocial stressors, acute
trauma symptoms and stage of change are also assessed. Intervention focuses on
psychoeducation, the effects of alcohol use, and includes brief MI to determine desire for
behavior change. Treatment resources are provided to interested patients. The AUDIT is
repeated post-discharge. The current study aims to provide retrospective rationale for
effectiveness of CUHs SOC. Trauma patients (n = 289, 76.5% male, 56.4% White, mean age =
45) with elevated BAC (M = 178.96) were referred as part of the SOC. Of those, 287 patients
completed the intervention and 37 (12.8 %) followed-up post discharge. Baseline AUDIT
scores were significantly related to ETOH levels on admission (r = .19, p < .01). A paired
samples t-test indicated that those who completed SBIRT significantly reduced AUDIT scores
at follow-up (t (38) = 3.34, p < .01). A repeated measures ANOVA was conducted to compare
main effects of time and baseline ETOH and interaction effect of time and baseline ETOH on
AUDIT scores. Results revealed main effect of time (F (1, 33) = 14.24, p < .01) and baseline
ETOH (F (1, 33) = 11.11, p < .01). A trend-level interaction effect was observed in which higher
BAC on admission led to steeper improvements at follow-up (F (1, 33) = 2.12, p = .15). At
follow-up, 21.1% reported seeking treatment for alcohol and 71.1% reported modifying
drinking behavior. Eleven patients reported hospitalizations since intervention, 2 were
alcohol-related. A positive reduction in alcohol symptoms was found, which may decrease
future hospitalizations related to alcohol use. Results provide sufficient rationale to continue
this SOC and to consider it for implementation at other hospitals. Modifications may include
S906 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

strategies to increase follow-up, such as establishing a specific timeline for post-assessments

and collecting multiple types of contact information for patients at screening. Additional data
is anticipated and will be included in future analyses.

CORRESPONDING AUTHOR: Kelly Lynn. Gilrain, PhD, Cooper University Hospital, Camden, NJ,
08103; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S907

B007 6:00 PM-7:00 PM

EXERCISE MAY ENHANCE SMOKING CESSATION THROUGH IMPROVED SLEEP

Shira Dunsiger, PhD1, Beth C. Bock, PhD2, Jessica Whiteley, PhD3, Joseph T. Ciccolo, PhD4,
Ernestine Jennings, PhD5, Sarah Linke, PhD, MPH6, Rochelle K. Rosen, PhD7, Santina Horowitz,
BS8, Ryan Lantini, MA9, Bess H. Marcus, Ph.D.10

1
The Miriam Hospital and Brown University, Portsmouth, RI; 2Brown Medical School,
Providence, RI; 3UMass Boston, Boston, MA; 4Teachers College, Columbia University, New
York, NY; 5The Miriam Hospital/ Alpert Medical School, Providence, RI; 6UC San Diego, La Jolla,
CA; 7Brown University School of Public Health, Providence, RI; 8The Miriam Hospital, Newport,
RI; 9The Miriam Hospital, Providence, RI; 10University of California, San Diego, La Jolla, CA

The present study examined data from 166 women in the exercise arm of a randomized
controlled community-based exercise and smoking cessation trial. Female smokers (N=330,
Mage=44.1[SD=9.9], M=17.4 cig/day[SD=8]) were randomly assigned to a 12-week cognitive-
behavioral smoking cessation program plus either exercise or contact control. Dose of
exercise was defined as total days over 12 weeks of reported aerobic and/or resistance
exercise. Previous analyses have shown a dose-response relationship between exercise dose
and 7-day point prevalence abstinence (PPA) at 12 weeks. Using a series of generalized
models, we tested the association between exercise dose and sleep as measured by the
Insomnia Severity Index over the 12-week treatment. Logistic regression was used to test
associations between bio-verified 7-day PPA and changes in sleep over 12 weeks as well as the
associations between patterns of quitting behavior and changes in sleep. Patterns have
previously been identified using Latent class models, suggesting five distinct patterns of
quitting behavior: those who made few quit attempts, non-quitters, early attempters, late
attempters (quit at 12 weeks but didnt attempt until late) and quitters, with exercise
participants reporting a significantly greater proportion of late attempters than controls.
Results of the present study show significant associations between greater exercise dose and
better quality of sleep at 12 weeks controlling for baseline (b=-0.14, SE=0.07, p < .05). Exercise
dose was also significantly negatively associated with difficulty falling asleep (b=-0.03,
SE=0.01, p=0.04) and satisfaction with sleep (b=-0.03, SE=0.02, p=0.04) at 12 weeks. The odds
of bio-verified 7-day PPA at 12 weeks was significantly higher for those reporting better sleep
at 12 weeks (OR=0.73, p=.04), and a trend for higher odds of quitting for those whose sleep
scores improved from baseline to 12 weeks emerged (OR= 1.78, p=0.09). Finally, non-quitters
had more difficulty falling asleep compared to quitters, and late attempters had worsening
sleep scores over 12 weeks compared to quitters (ps < .05).The association between exercise
S908 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

and quitting seen in our prior study may be related to improved sleep. Additional mediational
analyses will explore whether improved sleep may be a mechanism underlying the efficacy of
exercise for helping smokers quit.

CORRESPONDING AUTHOR: Shira Dunsiger, PhD, The Miriam Hospital and Brown University,
Portsmouth, RI, 02871; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S909

B008 6:00 PM-7:00 PM

EXPLORING MODERATORS OF THE RELATIONSHIP BETWEEN EXERCISE AND ALCOHOL USE

Alena Kryvanos, A.A1, Bethany R. Shorey Fennell, B.S.1, Renee E. Magnan, PhD2

1
Washington State Univeristy Vancouver, Vancouver, WA; 2Washington State University
Vancouver, Vancouver, WA

Exercise has been proposed as an adjunct to alcohol prevention and treatment efforts, yet
research typically supports a positive correlation between alcohol use and exercise. This
relationship is not well understood and one specific area needing more focus is for whom this
relationship may be more or less pronounced. The current investigation explored
demographic, personality, motivational, and behavioral moderators of the relationship
between exercise and alcohol. In a cross-sectional investigation, participants (N=122)
provided assessments of alcohol use and exercise over the past three months. Variables
tested as moderators included age, sex (demographic), positive and negative urgency, positive
and negative affect (personality), alcohol dependence and drug use (behavioral), and
motivation to use exercise as a compensatory behavior (motivational). Alcohol use and
exercise over the past three months were positively correlated (r = .23, p = .01). A series of
hierarchical regressions tested the main effects of exercise, the proposed moderator variable
in Step 1, and the exercise x moderator interaction term in Step 2 on alcohol use. With the
exception of positive affect, all proposed moderator variables were associated with alcohol
use. However, only drug use (=.20, p=.03) and compensatory motivation ( =-.15, p = .08)
moderated the exercise-alcohol use relationship. Those who engaged in greater exercise and
drug use reported greater alcohol use. Those who engaged in less exercise and had lower
compensatory were less likely to consume alcohol. Although replication and extension of
these findings are needed, they do provide some insight into factors that could contribute to
the exercise-alcohol use relationship. Specifically, previous drug use and compensatory
motivation could be important factors to take into consideration when designing
interventions focused on alcohol reduction or problem-drinking prevention with exercise as
an adjunct.

CORRESPONDING AUTHOR: Alena Kryvanos, A.A, Washington State Univeristy Vancouver,

Vancouver, WA, 98686-9600; [email protected]
S910 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B009 6:00 PM-7:00 PM

EXPLORING THE EFFECT OF LONGSTANDING ACADEMIC-COMMUNITY PARTNERSHIPS ON

STUDY OUTCOMES:A CASE STUDY

Annette E. Maxwell, DrPH, Catherine M. Crespi, PhD, Roshan Bastani, PhD

University of California Los Angeles, Los Angeles, CA

Background: We conducted two large trials (CRC1 & CRC2) to promote colorectal cancer (CRC)
screening among Filipino Americans. By design, about half of the organizations that
participated in CRC2 had also participated in CRC1. While sustained academic-community
partnerships can improve working relationships between research partners, they could also
influence study outcomes, thereby impacting the generalizability of study findings. We
explored the effect of being a new versus a previous research partner on CRC2 study
outcomes.

Methods: CRC2 tested two strategies (basic and enhanced) for implementing an evidence-
based multi-component intervention to promote CRC screening at community organizations
(N=17) with the help of trained community health advisors (CHAs; N=63). In both the basic
and enhanced arms, CHAs recruited Filipino Americans (N=673) who were not adherent to
CRC screening guidelines. The main study outcome was CRC screening status of participants at
6-month follow-up. Previous versus new served as a stratification variable for randomization
of organizations. Using multilevel logistic regression, we obtained estimates of the odds ratio
(OR) for new versus previous partner controlling for several organizational characteristics
(study arm, faith-based versus social service organization, years in operation) and further
controlling for CHA and participant characteristics.

Results: Screening rates were substantially higher among participants of previous versus new
partner organizations in unadjusted analysis (68% versus 45%, OR 2.5, p=.21), after adjusting
only for variables at the organization level (73% versus 34%, OR 7.3, p=.01) and after
additionally adding CHA and participant level factors to the model (70% versus 40%, OR 4.8,
p=.10). Although power was low due to the limited number of organizations, absolute
differences in screening rates were large and consistent.

Discussion: These findings suggest that study outcomes that are achieved with long-term
community partners may not be generalizable to new partners. Inclusion of organizations that
do not have an ongoing research relationship with an academic partner may improve external
validity of trials in community settings.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S911

CORRESPONDING AUTHOR: Annette E. Maxwell, DrPH, University of California Los Angeles,

Los Angeles, CA, 90095-6900; [email protected]
S912 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B010 6:00 PM-7:00 PM

FAMILY-BASED LAY HEALTH WORKER INTERVENTION PROMOTES HEALTHY NUTRITION AND

PHYSICAL ACTIVITY IN VIETNAMESE AMERICANS: A RCT

Janice Y. Tsoh, PhD1, Tung T. Nguyen, MD1, Nancy Burke, PhD2, Ginny Gildengorin, PhD1, Jin
Kim, PhD1, Khanh Le, MD, MPH1, Ching Wong, BS1, Bang H. Nguyen, DrPH3, Stephen J.
McPhee, MD1

1
University of California San Francisco, San Francisco, CA; 2University of California, Merced,
Merced, CA; 3University of California San Francisco, Richmond, CA

Background: Many Asian Americans do not meet nutrition and physical activity (NPA)
recommendations, which increases their vulnerabilities to chronic diseases. A majority of
Vietnamese Americans, the 4th largest Asian American group, are foreign-born with limited
English proficiency. This study reports outcomes of a randomized controlled trial of a family-
based lay health worker (LHW) intervention in promoting healthy NPA among Vietnamese
Americans.

Methods: Participants were 107 dyads of Vietnamese male cigarette smokers and non-
smoking family members (total N = 214) who were randomized to one of the two groups: the
NPA intervention (n=106 or 53 dyads) or smoking cessation control (n=108 or 54 dyads). All
participants received 2 LHW-led small group sessions and 2 individual telephone calls on their
assigned topic within 6-8 weeks. Participants were assessed at baseline and 6-month follow-
up. Multivariable logistic regression models were used to compare between the study arms
on the changes in knowledge of USDAs MyPlate and CDCs physical activity (PA)
recommendations, self-reported vegetable/fruit intake and PA, and changes in body weight
from baseline to 6-month.

Results: The 214 participants included 107 male daily smokers and 107 family members
(92.5% female) with mean age = 55.0 years (range: 19-86), 93.4% spoke limited English, and
60.7% with BMI >23 (overweight or obese using recommended cut-off for Asians). At baseline,
a minority knew about MyPlate (29.0%) or PA recommendations (5.1%) or met the
recommended vegetable/fruit intake (39.3%) or PA level (41.6%). At 6-month, the NPA group
reported a greater increase in knowing MyPlate (NPA: 20.9% vs. control: 5.5% increase,
p=0.003) and PA recommendations (NPA: 28.3% vs. control: 0% increase, p < 0.001) as well as
meeting recommended vegetable/fruit intake (NPA: 39.3% vs. control: 10.6% increase, p <
0.001) and PA level (NPA: 12.3% vs. control: 2.8% increase, p=0.006). In multivariable models,
baseline BMI was not associated with any of the outcome measures. The NPA group was more
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S913

likely to know MyPlate (adjusted odds ratio [AOR] = 2.9; 95% CI: 1.6 5.3) or PA
recommendations (AOR: 8.0; 95% CI: 3.0 21.2). Although they were not more likely to meet
PA recommendation (p=0.09), the NPA group were more likely to meet vegetable/fruit intake
recommendation (AOR = 4.5; 95% CI: 2.2 9.1) and have had weight loss of 5 lbs. or more
(AOR: 1.8, 95% CI: 1.2 2.6) than the control group.

Conclusions: Family-based LHW outreach delivering NPA information was effective in

increasing NPA knowledge, self-reported vegetable/fruit intake and weight loss in Vietnamese
Americans. Findings underscore the continuing needs to promote NPA knowledge and
behaviors in Vietnamese Americans with limited English proficiency and the promise of
involving families and peers in addressing such needs.

CORRESPONDING AUTHOR: Janice Y. Tsoh, PhD, University of California San Francisco, San
Francisco, CA, 94143; [email protected]
S914 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B011 6:00 PM-7:00 PM

FEMININE HYGIENE PRODUCTS: ATTITUDES AND BELIEFS ABOUT SAFETY

Pragya Rai, MS, Drishti R. Shah, MS, B.Pharm, Kimberly M. Kelly, PhD, MS

West Virginia University, Morgantown, WV

Objective Despite being used by women all over the world for 80% of their lifetime, limited
research on the use and safety of various feminine hygiene products persists, particularly
among men. Our objectives were to assess 1) use (women) and awareness (men) of feminine
hygiene products, 2) differences in safety attitudes and norms between men and women, and
3) association of practice with safety attitudes and norms of these products in women.

Methods A cross-sectional, online survey was conducted among young adults 18 to 30 years
old. The Theory of Planned Behavior guided survey development. T-tests compared
differences in attitudes and norms between men and women. Associations between current
use of feminine hygiene products (tampons, maxi-pads) with safety attitudes and norms were
estimated using logistic regression.

Results Our sample (n=179) included men (n=43), was highly educated (>bachelors degree=
59%), employed (75%), had an annual family income of >$50,000 (58%), and resided in an
Appalachia county (82%). Common products used by women included tampons (82%), panty
liners (59%), and maxi pads (54%). Unconventional products used included menstrual cups
(8%) and sanitary panties (4%). More controversial products such as feminine wipes (22%) and
douches (3%) were also used. Some men were unaware of common feminine hygiene
products: tampons (21%), maxi pads (26%), panty liners (51%), douches (40%) and feminine
wipes (54%), as well as unconventional products: menstrual cups (79%) and sanitary
panties (71%). There was a trend such that women had more positive safety attitudes
(p=0.06) and norms (p=0.10) all products than men. Most women had high safety attitudes
about tampons (mean=2.77, SD=0.67) and maxi-pads (mean=2.8; SD=0.67). Women using
tampons were less likely to be Hispanic (p=0.01), more likely to be White (p=0.003), and had
positive safety attitudes (p=0.03) as compared to non-users. Women using maxi-pads were
more likely to have an annual family income of

Conclusion Our novel study included men while studying products used by women. Despite
being widely advertised and sold in markets, some men were unaware of common feminine
hygiene products. Safety attitudes guide selection of feminine hygiene products. Additional
research and education is needed about the actual safety of these products to help women in
feminine hygiene decision-making.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S915

CORRESPONDING AUTHOR: Pragya Rai, MS, West Virginia University, Morgantown, WV,
26505; [email protected]
S916 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B012 6:00 PM-7:00 PM

GENDER DIFFERENCES IN PEER SUPPORT IN TYPE 2 DIABETES SELF-MANAGEMENT: AN

INTERNATIONAL QUALITATIVE STUDY

Florence Okoro, PhD, RN, MScN 1, Debra J. Barksdale, PhD, FNP-BC, FAANP, FAAn2, Shawn
Kneipp, PhD, RN, ANP-BC, APHN-BC, FAANP3, Kathleen Knafl, PhD FAAN4, Jennifer Leeman,
DrPH, MDIV4

1
School of Nursing, University of North Carolina at Charlotte, Charlotte, NC; 2VIrginia
Commonwealth University, Richmond, VA; 3University of North carolina Chapel Hill School of
Nursing, Chapel Hill, NC; 4School of Nursing, University of North Carolina at Chapel Hill, Chapel
Hill, NC

Gender Differences in Peer Support in Type 2 Diabetes Self-Management: An International

Qualitative Study

Florence Okoro, PhD,1 Debra Barksdale, PhD,2 Shawn Kneipp, PhD,3 Kathleen Knafl, PhD,3
Jennifer Leeman, DrPH, 3 Edwin Fisher, PhD, 3

1
University of North Carolina at Charlotte, 2Virginia Common Wealth University, Richmond
Virginia, 3University of North Carolina Chapel Hill

Background: The effectiveness of peer support programs as a means to sustain self-

management in patients with type 2 diabetes has been well-documented. However little is
known of how gender differences in peer support might influence the provision and receipt of
support.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S917

Purpose: This study used the key functions of peer support as framework to explore gender
differences in peer support programs and identified gender-related issues and global cultural
contexts that influence peer support.

Method: Qualitative data were generated through telephone interviews using open-ended
questions to nine programs across the globe: Africa (Cameroon and Uganda), Asia (Cambodia,
Hong Kong, Thailand, and Vietnam), and the United States (Alabama, California, and Illinois).
Qualitative content analysis was done to achieve in-depth exploration of categories.

Results: Five themes emerged: 1) Differences in emotional support: female peer supporters
reportedly displayed more emotional support than any other form of support; 2) Differences
in instrumental support: male peer supporters gave information as their primary form of
support; 3) Who is a peer supporter: males dominated as peer supporters in two Asian
countries (Cambodia and Hong Kong) and females dominated in African American and Latino
peer support programs in the US; 4) Matching by gender: male peer supporters were assigned
only male recipients, but female peer supporters could be assigned to either sex; 5) Gender
differences in participant availability: there was considerable variety in how gender was
manifest in programs and both peer supporters and participants responses to them.

Conclusion: This study provided evidence that substantial gender differences and gender-
related issues were present in all the peer support programs and that the manifestation of
differences varied considerably across settings and cultures. Gender differences have to be
taken into consideration when planning peer support programs.

CORRESPONDING AUTHOR: Florence Okoro, PhD, School of Nursing, University of North

Carolina at Charlotte, NC 28223; [email protected]

CORRESPONDING AUTHOR: Florence Okoro, PhD, RN, MScN , School of Nursing, University of
North Carolina at Charlotte, Charlotte, NC, 28269; [email protected]
S918 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B013 6:00 PM-7:00 PM

MERITORIOUS AWARD WINNER

GET ACTIVE! EFFICACY OF A BRIEF ACCEPTANCE-BASED BEHAVIORAL INTERVENTION TO

PROMOTE EXERCISE ADOPTION AND MAINTENANCE

Courtney J. Stevens, MA1, Angela D. Bryan, PhD2

1
Alpert Medical School of Brown University, Providence, RI; 2University of Colorado Boulder,
Boulder, CO

Regular exercise confers tremendous transdiagnostic benefits for health, yet 54% of American
women are insufficiently active. Affective/experiential factors influence exercise adoption and
maintenance, but how best to address these constructs and maximize health outcomes is not
well understood. This study proposed that participants may be able to learn acceptance-based
skills for managing inherent/inevitable exercise discomfort. The central aims of this
investigation were to (1) compare exercise change across time and condition, and (2) assess
mechanisms of exercise behavior change over time. A sample of N = 119 insufficiently active
women aged 21 65 were randomly assigned to 1 of the 3 conditions: (1) acceptance-based
health coaching (ACT), (2) education-based health coaching (EDUC), or (3) no-health coaching
control (CTRL). The study was divided into: Phase 1 Adoption (baseline visit post
intervention), and Phase 2 Maintenance (3- and 6- months follow up). During Phase 1, all
participants completed a 30-day exercise program, and ACT and EDUC participants received
visit 1 and 2 workshops. All participants were contacted at 3- and 6-months post-intervention
to complete follow-up assessments. The overall trend across Phases 1 and 2 was for ACT
participants to score highest on exercise outcomes. Notably, a time X condition interaction for
PAR total minutes of exercise was observed such that ACT participants completed more
exercise minutes in the past week at visit 2 compared to EDUC and CTRL participants, F(2,103)
= 4.58, p = .012, 2 = .082. Additionally, only ACT participants completed 150-minutes of
exercise per week, on average, over the course of the 30-day intervention (as recorded
objectively by HR monitors). Repeated measures tests showed significant main effects of time
on exercise scores through Phase 2, F(3, 201) = 22.99, p < .001, 2 = .255; but not significant
main effects of condition or time X condition interactions. Regarding mechanisms of action,
assignment to ACT predicted higher experiential acceptance scores at visit 2, and visit 2
experiential acceptance scores predicted exercise scores at both 3- and 6-months follow-up as
tested via mediational analyses. These results demonstrate that a brief acceptance-based
intervention can improve metrics of psychological flexibility, which in turn influences exercise
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S919

behavior maintenance. Future work should assess the optimal intervention dose for
bolstering these associations across time.

CORRESPONDING AUTHOR: Courtney J. Stevens, MA, Alpert Medical School of Brown

University, Providence, RI, 02906; [email protected]
S920 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B014 6:00 PM-7:00 PM

HOW PEER MENTORSHIP CONVERSATIONS SUPPORT PEOPLE WITH SCI: A THEMATIC

ANALYSIS

Rhyann McKay, Bachelor of Human Kinetics1, Emily E. Giroux, Bachelor of Science

Kinesiology1, Christopher B. McBride, PhD2, Sheila Casemore, BASc3, Teren Clarke, Bachelor of
Nursing, Master of Management 4, Kristy L. Baxter, Human Kinetics1, Shane N. Sweet, PhD5,
Heather Gainforth, PhD6

1
University of British Columbia Okanagan, Kelowna, BC, Canada; 2Spinal Cord Injury BC,
Vancouver, BC, Canada; 3Spinal Cord Injury Ontario, Toronto, ON, Canada; 4Spinal Cord Injury
Alberta, Edmonton, AB, Canada; 5McGill University, Montreal, PQ, Canada; 6University of
British Columbia, Kelowna, BC, Canada

Background: Participation is defined as the extent and nature of a persons involvement in

society. Full participation in society is a U.N. human right. People with a spinal cord injury (SCI)
often have lower levels of participation compared to the general population. Peer mentorship
is an approach that shows promise for improving participation among people with SCI. Peer
mentors can be defined as individuals with lived experience of SCI who provide knowledge,
counsel, and encouragement to mentees with SCI. To date, research has not examined how
SCI peer mentorship conversations enhance participation among mentees with SCI, or the
nature of the topics discussed during these conversations.

Purpose: The purpose of this study is to gain an understanding of how peer mentorship
conversations enhance participation among mentees with SCI. Specifically, it aims to use
thematic analysis to understand topics discussed during SCI mentorship conversations.

Methods: Three provincial SCI community organizations in Canada recruited participants with
SCI. Participants completed a baseline survey to collect information about demographics and
prior experience with peer mentorship. Peer mentors and mentees were matched, and their
telephone conversations were audio recorded and transcribed verbatim. Two researchers
inductively coded each transcript independently to identify topics. The researchers then came
together to reach a consensus of the key themes.

Results: Similar to previous analyses of behavioural support conversations, 5 conversations

between 5 mentors and 5 mentees were recorded (mean age: 49.2 14.3 years, 40% female).
The mean duration of the conversations was 38.25 21.28 minutes. Identified themes
included: sports and recreation; perception of their injury; similarities between the mentor
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S921

and mentee; accessibility issues in the home and community; information about available
resources; the importance of support from peers; aging with SCI; and physiological changes
due to SCI, particularly with sensory and motor function and changes with the bladder and
bowel. Mentees primarily initiated conversation around these themes, with physiological
changes and aging being the most commonly discussed.

Discussion: Peer mentorship conversations address various topics related to participation

with topics related to physiological changes and aging being most commonly instigated by
mentees. This study is the first to record and inductively code SCI peer mentorship
conversations. Future research will apply these findings to develop coding frameworks that
will allow for in-depth conversation analyses of peer mentorship conversations and will
ultimately aim to improve SCI peer mentorship programs.

CORRESPONDING AUTHOR: Rhyann McKay, Bachelor of Human Kinetics, University of British

Columbia Okanagan, Kelowna, BC, V1Y 9Y9; [email protected]
S922 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B015 6:00 PM-7:00 PM

IDENTIFYING HEALTH AND WELLNESS PERCEPTIONS AND NEEDS AMONG WIC EMPLOYEES

Natasha Birchfield, B.S., M.S.1, Jennifer Huberty, PhD 2, Carrie Zavala, M.S., R.D.3, Jeanene
Fowler, B.S.3

1
Arizona State University, Chandler, AZ; 2Arizona State University School of Nutrition and
Health Promotion, Phoenix, AZ; 3Maricopa County Department of Health, Phoenix, AZ

Introduction: Social service employees are at risk for experiencing poor health due to the
burden associated with job stress (i.e. difficult clients, volatile situations). There is a need for
worksite interventions specifically designed for stress associated with social service
employment. The purpose of this qualitative research study was to identify Women, Infants,
and Childrens Nutrition Program (WIC) employee perceptions of health and wellness to
inform a worksite wellness program. Methods: Four focus groups were conducted with 38
WIC staff and 14 30-minute interviews with WIC supervisors. Data were analyzed by
WestGroup Research. Results: Data were organized by individual health and wellness (IHW)
and role of the employer in health and wellness (EIHW). IHW had three subthemes: (1) health
and wellness concerns, (2) ways staff stay healthy (3) barriers to health and wellness. EIHW
had five subthemes: (1) aspects employer should support; (2) ways employer supports; (3)
ways employer could support; (4) obstacles to participation; and (5) motivators for
participation. The major health and wellness concern was stress. Walking was most identified
as a way to stay healthy. Health and wellness barriers were time and work environment. In
relation to the role of the employer, WIC staff felt physical and mental health, and the
environment should be supported. Staff felt WIC provided health insurance, health coaching,
and mental health services but most services were inaccessible. Staff felt the employer could
provide support for stress reduction, fitness and diet, environmental improvements (i.e.,
cleanliness, standing desks), and staff appreciation. Obstacles to participation included
location (i.e., programs are off-site and too far away), time, policies, stress (i.e., demands of
the day leads to fatigue and low motivation), email challenges, lack of space in clinic, and cost.
Motivators included positive encouragement from leadership, fellowship (i.e., inclusion in
programs, support from others), offering programs in-clinic, allowing worktime (i.e. scheduled
breaks) to participate, participation incentives, and staff appreciation. Discussion:
Programming that is supported by leadership and is aligned with employee needs is critical for
the success of worksite health and wellness programs. The results of this study will inform the
implementation of an evidenced-based wellness program targeting stress reduction and
physical activity behaviors in WIC staff.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S923

CORRESPONDING AUTHOR: Natasha Birchfield, B.S., M.S., Arizona State University, Chandler,
AZ, 85225; [email protected]
S924 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B016 6:00 PM-7:00 PM

IMPACT OF A COMMUNITY-BASED DIABETES PREVENTION AND MANAGEMENT PROGRAM

FOR IMPROVING PHYSICAL ACTIVITY IN WEST VIRGINIA

Jonathan Stewart, M.S., Ranjita Misra, PhD, Samantha Shawley, MS, Spencer Ingels, M.A.,
Brandon Lucke-Wold, MS

West Virginia University, Morgantown, WV

Physical activity is critical for blood glucose management and overall health in individuals with
diabetes and pre-diabetes. Researchers have estimated that as much as 10% of type 2
diabetes cases worldwide are caused by physical inactivity. More specifically, West Virginia is
burdened by higher rates of type 2 diabetes and physical inactivity. This study explored the
relationship of physical activity with anthropometric and clinical factors among participants in
a community-based diabetes prevention and management (DPM) program.

Eighty-three adults with type 2 diabetes and pre-diabetes completed a 12-month (22-
session) DPM program in two churches in Morgantown and Charleston WV (completion rate
82%). The program was modeled after the Diabetes Prevention Program and modified to
include diabetes management sessions. The trained student health coaches (HCs) (included a
multidisciplinary group of undergraduate and graduate students) were used as change agents;
HCs had weekly communications with the participants to provide continuous feedback and
reinforcement of health education messages. Program assessments included step count
(pedometer), surveys, clinical (blood pressure, triglycerides, cholesterol) and anthropometric
measurements (weight, waist circumference (WC), body mass index, and % body fat).

Mean age was 58.5 years; 63% self-reported that they had diabetes; 38 participants (75%
females, 63% diabetics) completed the weekly activity log to track their physical activity for at
least 24 weeks (46% activity log completion rate). A significant improvement in physical
activity occurred from baseline (2nd and 3rd week) to mid-program (17th to 19th week)
(p=0.05). There was a significant (negative) bivariate association between average steps and
BMI and WC (p < .05). Controlling for Weight, WC and HDL in the multiple linear regression,
physical activity was significantly associated with change in A1c between baseline and 6-
months (p=0.042).

Results show that a lifestyle education program can have a positive impact on improving
physical activity. Partnerships with churches to deliver evidence-based diabetes programs can
serve as a prototype for future efforts to improve physical activity and health outcomes for
WV communities
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S925

CORRESPONDING AUTHOR: Jonathan Stewart, M.S., West Virginia University, Morgantown,

WV, 26505; [email protected]
S926 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B017 6:00 PM-7:00 PM

IMPACT OF A HEALTH PROMOTION INTERVENTION ON FEMALE ULTRA-ORTHODOX JEWISH

TEACHERS' HEALTH BEHAVIORS AND HEALTH SELF-EFFICACY

Elisheva Leiter, PhD1, Sima Wetzler, BS2, Keren Greenberg, MPH3, Osnat Keidar, MPH, PHD4,
Donna Zwas, MD, MPH5

1
inda Joy Pollin Cardiovascular Wellness Institute for Women, Hadassah University Medical
Center, Jerusalem, Yerushalayim, Israel; 2Linda Joy Pollin Cardiovascular Wellness Institute for
Women, 2, Yerushalayim, Israel; 3Linda Joy Pollin Cardiovascular Wellness Institute for
Women, 222, KS; 4Linda Joy Pollin Cardiovascular Wellness Institute for Women, ff, CO;
5
Cardiology Department, Hadassah University Medical Center, Jerusalem, Yerushalayim, Israel

Introduction: Women in low socio-economic, religious, cloistered communities often have

lower levels of health knowledge and preventive health behaviors. Ultra-Orthodox Jewish
(UOJ) women engage in lower levels of preventive health behaviors and have higher rates of
diabetes and obesity. Interventions aimed at improving engagement in preventive health
behaviors in this population are unknown. This study evaluated the impact of a health
promotion intervention with UOJ female high school teachers on health knowledge,
preventive health behaviors, general health self-efficacy, and self-rated health status.

Methods: 246 female UOJ teachers from two girls' high schools in Israel completed
questionnaires before and one year following a school-based health promotion intervention.
Questions included demographic, preventive health behaviors, health knowledge items, self-
rated health status, as well as a general health self-efficacy scale.

Results: Women's age range was 21-71, average number of children was 6.4 (range= 0-17),
36% were below poverty level and 55% were overweight or obese. Post intervention, women
had increased knowledge of heart attack symptoms p < .005) and cardiovascular disease risk
factors (p < .05). They had increased whole grain consumption (p < .05), decreased margarine
consumption (p < .05), and increased blood pressure testing (p < .005). Women also reported
increased accuracy in self-rated health status (p < .005) and increased general health self-
efficacy (p < .005).
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S927

Conclusions: This study is the first investigating an intervention aimed at improving

engagement in preventive health behaviors with UOJ women. School-based health promotion
interventions can improve health knowledge, preventive health behaviors, self-perception of
health status, and health self-efficacy in female UOJ high school teachers. Utilizing the school
setting in interventions aimed at improving engagement in preventive health behaviors may
provide an accessible, effective method for preventive health behaviors promotion in UOJ and
other women from low SES, religious, cloistered populations.

CORRESPONDING AUTHOR: Elisheva Leiter, PhD, inda Joy Pollin Cardiovascular Wellness
Institute for Women, Hadassah University Medical Center, Jerusalem, Yerushalayim, 91163;
[email protected]
S928 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B018 6:00 PM-7:00 PM

IMPACT OF A MODULAR SOCIAL COGNITIVE SMARTPHONE APPLICATION ON PHYSICAL

ACTIVITY IN ADULTS: A RANDOMIZED FACTORIAL TRIAL

Jason Fanning, PhD1, Sarah Roberts, M.S.2, Charles Hillman, PhD3, Sean P. Mullen, PhD4, Lee
M. Ritterband, PhD5, Edward McAuley, PhD6

1
Wake Forest University, Winston Salem, NC; 2University Of Illinois At Urbana - Champaign,
Urbana, IL; 3Northeastern University, Boston, MA; 4University of Illinois at Urbana-Champaign,
Urbana, IL; 5Center for Behavioral Health & Technology, Charlottesville, VA; 6University of
Illinois at Urbana Champaign, Urbana, IL

Smartphones have received considerable attention as vehicles for behavioral intervention, but
a focus on randomized controlled trials has provided mHealth developers with little guidance
in constructing their own health apps. The present study aimed to assess the individual and
combined impact of two social cognitive theory (SCT)-based intervention ingredients on
moderate to vigorous physical activity (MVPA) and important SCT outcomes. Adults (N = 116;
Mage = 41.38 7.57) were recruited to participate in a goal-driven 12-week MVPA intervention,
and were randomized to receive the base-level app containing 3 common eHealth tools (i.e.,
MVPA tracking, instant/bi-weekly feedback, weekly educational modules), base-app + guided
goal setting (i.e., GGS; participant behavior automatically generated weekly goal-setting
strategies and recommendations), base-app + points-based feedback (i.e., PBF: all app
features were associated with points, motivational content, levels, and badges), or the base-
app + both modules. MVPA was assessed via accelerometry at baseline and week 12. SCT
outcome measures included the exercise self-efficacy scale (EXSE); lifestyle self-efficacy scale
(LSE); barriers self-efficacy scale (BARSE); multidimensional outcome expectations (OE) for
exercise scale, which included physical (POE); social (SOE); and self-evaluative (SEOE)
subscales; the perceived barriers scale (PBE); and exercise goal-setting scale (EGS). Repeated
measures factorial analysis of variance (RM-ANOVA) revealed a significant effect for time
(pp=.05): those with PBF recorded an additional 5.94 minutes of daily MVPA. For BARSE, there
was a significant PBF*time interaction. Post-hoc analyses indicated a protective effect for
those with PBF (p=.20) and a decrease in those without PBF (p < .01). There were time effects
for EXSE and LSE (ps .02). Participants reported poorer efficacy following the intervention;
however a significant time*PBF*GGS interaction for EXSE (p = .04) indicated those with PBF
and GGS maintained efficacy (p = .98), while those without PBF significantly decreased (p <
.01). No significant effects were present for OE. For POE there was a trend toward significance
in the time*PBF interaction (p = .05): there was a nonsignificant increase for those with PBF
(p=.14), and a non-significant decrease for those without (p=.2). The time effect for EGS was
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S929

significant (p < .01): all individuals increased goal setting strategy usage, and those with PBF
engaged in still greater strategy usage (p = .01). For PBE, the main effect for time was
significant (p < .01). Participants perceived fewer barriers to exercise following the
intervention. These results are of immediate use to mHealth developers, suggesting that
although the base-level app may have been overbuilt, the use of a theory-based PBF system
may positively impact MVPA and important SCT outcomes.

CORRESPONDING AUTHOR: Jason Fanning, PhD, Wake Forest University, Winston Salem, NC,
27103; [email protected]
S930 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B019 6:00 PM-7:00 PM

INTER-RATER RELIABILITY OF TAXONOMY CODING OF THE BEHAVIOR CHANGE TECHNIQUES

USED IN A WEIGHT MANAGEMENT INTERVENTION

Clara Adegbite-Adeniyi, BA1, Oluwatomisin Olayinka, MPH, RN2, Shirley M. Moore, PhD1

1
Case Western Reserve University, Cleveland, OH; 2School of Nursing, Case Western Reserve
University, Cleveland, OH

Background: The Behavior Change Technique Taxonomy (BCTTv1) was developed as a

common language for describing behavior change interventions across disciplines (Michie et.
al, 2013). This taxonomy tool can be used to identify the behavior change techniques that are
the most effective and most frequently used in multi-component interventions. Requisite to
using a taxonomy to identify and classify behavior change techniques, however, is the need to
establish inter-rater reliability among taxonomy coders.

Purpose: The purpose of this study was to determine the feasibility and reliability of novice
coders in coding the behavior change techniques (BCTs) used in two theoretically-different
interventions of a randomized clinical trial to reduce childhood obesity.

Methods: The process involved a pair of expert coders and a pair of novice coders. Novice
coders received 120 hours of online training over 3 months, including phone and e-mail
correspondence with expert coders. Each person of the two pairs (expert and novice coders)
independently coded the materials from two interventions. The materials consisted of 884
pages of the interventionist protocol manual and participant materials used in a 1-year family-
oriented behavior change intervention that met face-to face every 2 weeks. The materials for
both interventions were coded separately by each coder with adjudication by coder pairs in
early phases of coding. The number of BCTs identified by each set of coders were recorded,
and inter-rater reliability was calculated using prevalence- and bias-adjusted kappa (PABAK)
and percent agreement.

Results: Training for each novice coder took approximately 125 hours. It took 120 hours for
each novice coder to independently code the intervention materials. Inter-rater reliability
between the novice coder pairs were PABAK scores of 0.80 and 0.81 and agreement scores of
88% and 90% for the two interventions, respectively. Reliability between novice coder pairs
and expert coder pairs showed PABAK scores of 0.46 and 0.59 for the two interventions
(acceptable PABAK 0.67 (Sim and Wright, 2005)). Percent agreement was 73% and 80%,
respectively.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S931

Conclusions: The results indicate good taxonomy coding inter-rater reliability among novice
coders of two different interventions. Although the PABAK scores were not as high between
novice and expert coders, the agreement was still good. Our results indicate that with
adequate training, reliable taxonomy coding of behavior change techniques in behavior
change interventions can be successfully achieved with newly trained individuals.

CORRESPONDING AUTHOR: Clara Adegbite-Adeniyi, BA, Case Western Reserve University,

Cleveland, OH, 44106-4904; [email protected]
S932 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B020 6:00 PM-7:00 PM

OBJECTIVELY MEASURED E-NEWSLETTER ENGAGEMENT AND WORKPLACE SEDENTARY

BEHAVIOR

Miranda L. Larouche, B.S.1, Sarah L. Mullane, Ph.D.2, Meynard J. Toledo, B.S., M.S.2, Sarah A.
Rydell, MPH3, Linda Feltes, M.S.4, Brenna Vuong, M.P.H.5, Aaron Bryant, M.B.A.6, Noe C.
Crespo, Ph.D., M.P.H.7, Mark A. Pereira, Ph.D.8, Matthew P. Buman, Ph.D.2

1
School of Nutrition and Health Promotion, Arizona State University, Phoenix, AZ; 2Arizona
State University, Phoenix, AZ; 3Division of Epidemiology & Community Health, University of
Minnesota School of Public Health, Minneapolis, MN; 4Minnesota Department of Health, Saint
Paul, MN; 5Fairview Health Services, Saint Paul, MN; 6Arizona State University, Tempe, AZ;
7
San Diego State University, San Diego, CA; 8Division of Epidemiology and Community Health,
Minneapolis, MN

Workplace behavior change interventions often incorporate educational electronic messages

(i.e., E-newsletters). However, there is limited objective data on whether engagement with E-
newsletters actually drives behavior change. We examined whether E-newsletter engagement
was associated with changes in sedentary and active behaviors during the first three months
of a sedentary behavior workplace intervention. Participants (N = 413, 75.6% female, 72.4 %
White, 43.6 11.1 years of age) were recruited from 14 worksites in the Phoenix and
Minneapolis regions to participate in a multi-component intervention to reduce sitting and
increase light-intensity physical activity (LPA). E-newsletters were sent weekly for the first
month and biweekly thereafter and were monitored using email tracking software
(Salesforce.com, Inc.). E-newsletter engagement was assessed by open rates (i.e, number of
emails opened) and click rates (i.e., number of embedded hyperlinks clicked). Participants
wore an activPAL accelerometer to assess sitting, standing, LPA (< 100 steps/min), moderate-
vigorous physical activity (MVPA; 100 steps per/min), number of sit-stand transitions, and
long bouts of sitting (30 minutes) during work hours. Outcomes were standardized to an 8h
workday. Mixed-effects regression models (clustered by worksite) were adjusted for age,
gender, race, BMI, and job type. Participants spent 323.0 86.1 minutes sitting, 106.8 74.5
minutes standing, 27.4 13.5 minutes in LPA, 5.7 5.2 minutes in MVPA, 6.3 4.0 sit-stand
transitions, and 146.5 92.1 minutes in long bouts of sitting at work. Open rates were
associated with reductions in sitting (b[se]=-5.4 [2.6] min/workday, p=.04), but not with
changes in standing, LPA, MVPA, sit-stand transitions, or long bouts of sitting. Click rates were
not associated with any of the outcomes. E-newsletter engagement - as measured objectively
by e-mail open rates - was associated with reductions in sitting time at work. However, the
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S933

effect did not translate to significant increases in standing or more active behaviors. Future
interventions should continue to focus on E-newsletters as a viable means of intervention
delivery; however, more robust methods are likely necessary for long-term and substantive
decreases in workplace sedentary time.

CORRESPONDING AUTHOR: Miranda L. Larouche, B.S., School of Nutrition and Health

Promotion, Arizona State University, Phoenix, AZ, 85004; [email protected]
S934 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B021 6:00 PM-7:00 PM

ACCULTURATION MODERATES AMBIVALENCE OVER EMOTIONAL EXPRESSIONS AND WELL-

BEING AMONG CHINESE BREAST CANCER SURVIVORS

William Tsai, Ph.D.1, Qian Lu, M.D., Ph.D.2

1
California State University, San Marcos, Vista, CA; 2University of Houston, Houston, TX

Ambivalence over emotional expression (AEE) is defined as individuals internal conflict

between wanting to express their emotions but fearing the consequences of such expressions.
In one study among European American gastrointestinal cancer patients, AEE was associated
with greater pain and lower quality of life. Whereas low levels of AEE are regarded as
essential to healthy interpersonal and intrapersonal functioning among European Americans,
the ability to restrain or control emotions to maintain social harmony is a sign of maturity and
social awareness among Asian Americans. Hence, the relations between AEE and well-being
may be different among Asian Americans. Few studies to date have examined these
emotional processes using a sample with chronic illness that prioritizes emotional self-control.
The present study examines the relations between AEE and well-being (viz., depressive
symptoms and quality of life) as a function of acculturation among a sample of 93 Chinese
breast cancer survivors. Specifically, we predicted that AEE would not be predictive of quality
of life and depressive symptoms among less acculturated Chinese breast cancer survivors,
whereas AEE would be associated with lower quality of life and greater depressive symptoms
among highly acculturated Chinese breast cancer survivors. Consistent with our hypotheses,
we found that acculturation moderated the relations between AEE and depressive symptoms,
F(1, 82) = 6.18, p < .05, R2 = .06, p2 = .07, and marginally moderated the relations between
AEE and quality of life, F(1, 84) = 3.36, p = .07, R2 = .03, p2 = .04. Specifically, AEE was
positively associated with higher depressive symptoms (B = .20, p < .001) and lower quality of
life (B = -.45, p < .001) for women with high acculturation, but not for women with low
acculturation (B = .04 and -.15, p > .05 for depressive symptoms and quality of life,
respectively). As Chinese women acculturate to mainstream American culture and obtain
Western values of individuality, self-assertion, and emotional expression, AEE becomes a
detrimental emotional process associated with maladjustment. Future research would benefit
from exploring the directionality of these processes and explore specific cultural values that
were obtained or lost as a function of acculturation.

CORRESPONDING AUTHOR: William Tsai, Ph.D., California State University, San Marcos, Vista,
CA, 92081; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S935

B022 6:00 PM-7:00 PM

ADVOCACY AND STIGMA AMONG PATIENTS WITH LUNG CANCER

Sarah N. Price, B.A.1, Jamie S. Ostroff, Ph.D.2, Megan Shen, PhD3, Anna J. Thomas, Ph.D.4,
Simon J. Craddock. Lee, PhD, MPH5, Heidi Hamann, PhD6

1
The University of Arizona, Tucson, AZ; 2Memorial Sloan Kettering Cancer Center, New York,
NY; 3Weill Cornell Medicine, New York, NY; 4Baylor University Medical Center, Dallas, TX; 5UT
Southwestern Harold C. Simmons Comprehensive Cancer Center, Dallas, TX; 6University of
Arizona, Tucson, AZ

Introduction: Stigma is pervasive among patients with lung cancer, and is associated with
negative psychosocial and behavioral outcomes. To develop stigma-focused interventions,
there is a need to understand potential avenues of adaptive coping and stigma resilience
among lung cancer patients. Preliminary investigations (e.g., Brown & Cataldo, 2013),
including our previous qualitative work (Hamann et al., 2014), indicate that patient
engagement in lung cancer advocacy (supporting a cause, idea, and/or policy through
knowledge, personal contact, fundraising and other organized activities) may mitigate
negative effects of stigma for patients. The present quantitative analysis builds on this work to
examine lung cancer advocacy and possible relationships to stigma and other characteristics
among patients with lung cancer.

Methods: The sample included 231 patients (147 women; 84 men) diagnosed with lung
cancer and recruited from outpatient clinics associated with two NCI-designated cancer
centers. Lung cancer advocacy was assessed through four items endorsed on a 1-5 scale (I
have tried to learn as much as I can about my lung cancer; I have been involved in a group
that promotes lung cancer awareness; I have wanted to help others who have lung cancer;
I have tried to share with others what I know about lung cancer). Lung cancer stigma was
assessed using the Lung Cancer Stigma Inventory (LCSI; Hamann et al., In press), a 25-item
measure reflecting factors of Perceived Stigma (negative appraisal and devaluation from
others), Internalized Stigma (self-blame, guilt, and regret), and Constrained Disclosure (social
avoidance, disclosure discomfort, and disengagement).

Results: We found a statistically significant difference in advocacy based on smoking status, F

(2, 216) =3.35, p=.04), with never smokers reporting higher scores (M=12.28) on the advocacy
measure than current smokers (M=9.79; p=.02) and marginally higher scores than former
smokers (M=11.37; p=.10). Although not significantly associated with factors of perceived or
S936 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

internalized stigma, advocacy was negatively correlated with the factor of constrained
disclosure (r=-0.19; p < .01).

Conclusions: Involvement in advocacy may buffer lung cancer patients against the social
avoidance and disengagement associated with lung cancer stigma. Advocacy involvement may
enhance social support and feelings of shared purpose that help guard against constrained
disclosure. Facilitating community involvement and advocacy may be useful components of
multifaceted interventions to mitigate lung cancer stigma. There may, however, be barriers to
advocacy engagement for current smokers that require increased focus and understanding.

CORRESPONDING AUTHOR: Sarah N. Price, B.A., The University of Arizona, Tucson, AZ, 85719;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S937

B023 6:00 PM-7:00 PM

CHALLENGING THERAPEUTIC DISCUSSIONS WITH LATINO CANCER PATIENTS: PROVIDER

PERSPECTIVES FROM LATIN AMERICA AND THE USA

Rosario Costas-Muiz, PhD1, Olga Garduo-Ortega, B.A.1, Xiomara Rocha-Cadman, MD2,

Javier Gonzlez, MFA3, Francesca Gany, MD, MS4

1
Memorial Sloan Kettering Cancer Center, New York, NY; 2MSKCC, Ny, NY; 3Memorial Sloan-
Kettering Cancer Center, New York, NY; 4Immigrant Health and Cancer Disparities Center,
Memorial Sloan Kettering Cancer Center, New York, NY

BACKGROUND: Latino cancer patients (LCP) are at risk of poor psychological outcomes.
Managing distress in LCP requires therapeutic skill and cultural competency. In this study, we
explored mental health providers(MHP) perceptions about the challenging or difficult
aspects of discussing the cancer diagnosis (CD), cancer prognosis (CP) and death and dying
(DD) in therapeutic discussions.

METHODS: A survey with open-ended questions was emailed to MHP practicing in the USA
and Latin America. Three qualitative questions explored difficult aspects of counseling LCP
when discussing the CD, CP and DD issues. Fifty five MHP providing mental health services in
Spanish participated. Four independent coders coded responses and inductive content
analysis was utilized to analyze the data.

RESULTS: Ten percent of MHP reported that they find challenging having discussions about
the CD, 21% about CP, and 28% about DD. Thematic categories across the three areas (CD, CP,
and DD) were: 1) emotional impact and coping, 2) cultural issues, and 3) patient-provider
related themes. Providers perceived difficulty addressing the emotional impact and coping
strategies of patients; particularly, it was difficult to address distress and coping mechanisms
such as denial and avoidance when talking about CD, CP, and DD. There was a need for
increased cultural sensitivity to address LCP specific cultural beliefs. There was an expressed
need to build therapeutic trust and to receive specialized training to treat LCP.

A significant theme was the lack of information, cancer-related education, and/or

communication about the CD and CP. Stigma was challenging in conversations about CP and
DD, specifically, fatalistic perceptions about cancer and death taboo. Family impact and the
desire to protect the family were challenging themes during DD conversations.

DISCUSSION: MHP describe encountering many challenges in their therapeutic discussions

with LCP about CD, CP and DD. Addressing the emotional impact, the educational needs, the
S938 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

cultural context, stigma, and family impact, were challenging aspects of having these
conversations. The need for specialized training to improve therapeutic effectiveness during
therapeutic discussions was emphasized. Psycho-oncology educators and training programs
should incorporate additional training on how to address these challenging topics and the
cultural context of patients in a sensitive manner.

CORRESPONDING AUTHOR: Rosario Costas-Muiz, PhD, Memorial Sloan Kettering Cancer

Center, New York, NY, 10065; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S939

B024 6:00 PM-7:00 PM

CHANGES IN OBJECTIVELY MEASURED ACTIVITY OVER 6 MONTHS AMONG BREAST CANCER

SURVIVORS ON ADJUVANT THERAPY

Cynthia C. Forbes, PhD1, Melanie Keats, PhD1, Daniel Rainham, PhD2, Tallal Younis, MD3,
Cynthia Blanchard, PhD1

1
Dalhousie University, Halifax, NS, Canada; 2Dalhousie Univeristy, Halifax, NS, Canada; 3Nova
Scotia Health Authority, Halifax, NS, Canada

Objective: The objective of this study was to assess the prevalence of objectively measured
moderate-to-vigorous PA (MVPA), light and sedentary time among breast cancer survivors
(BCS) currently undergoing adjuvant therapy. We also wanted to determine if there were any
changes over time.

Methods: Participants completed baseline surveys and wore an accelerometer to measure

MVPA during waking hours for nine consecutive days at baseline and 6 months. Seven valid
days were used by deleting the first and last day if nine were completed. 5-second epochs
were used to provide a more sensitive estimate of MVPA levels. Repeated Measures Analyses
of Variances (R-ANOVAs) were used to determine changes over time.

Results: BCS were recruited from the QEII Health Science Centre in Halifax, Nova Scotia. Of
the 114 approached 77 (68%) agreed to participate. Participants were aged 57 years, 68%
married, 51% employed, and BMI of 28kg/m2. 70 (91%) and 56 (72%) provided PA data for 3
valid days at T1 and T2 respectively. At baseline, the average duration of MVPA bouts was 119
minutes and 27% were meeting guidelines. At 6 months, the average duration was 138
minutes and 38% were meeting guidelines. Sedentary time did not change from T1 to T2 (597
vs 601 minutes). Significant increases from T1 to T2 were found in daily light minutes (+6.9;
p=.048), daily moderate minutes (+5.6; p=.021), weekly moderate minutes (+42.7; p=.007),
and weekly MVPA minutes (+36.5; p=.045).

Conclusion: This study found that BCS increased their objectively measured light and MVPA
over six months. This would generally be when most survivors have completed the majority of
their treatments. They may be feeling better and are returning to their previous activity levels.
However, the smaller number of T2 measures may bias the results and artificially inflate them.
More research needs to be conducted with larger sample sizes to gain a more accurate
representation of activity levels of BCS undergoing adjuvant therapy.
S940 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Cynthia C. Forbes, PhD, Dalhousie University, Halifax, NS, B3H
1V7; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S941

B025 6:00 PM-7:00 PM

DEVELOPMENT OF A PATIENT NAVIGATION INTERVENTION TO IMPROVE COLORECTAL

CANCER CARE

Tatianna Clark, NA1, Claudia Carrizosa, MD2, Samir Gupta, MD3, Balambal Bharti, Staff Res
Assoc3, Jesse Nodora, DR. PH3, Elva M. Arredondo, PhD2, Christian Ramers, MD4, Yessenia
Magana, n/a5, Eden Sosa, MBA4, Felipe Garcia-Bigley, MHA4, Jessica Haughton, MPH, MA6,
Jamie Moody, MS7, Kristen Wells, Ph.D., MPH2

1
San Diego State University, Chula Vista, CA; 2San Diego State University, San Diego, CA;
3
University of California, San Diego, La Jolla, CA; 4Family Health Centers of San Diego, San
Diego, CA; 5Family Health Centers of San Diego, Sasn diego, CA; 6San Diego State
University/UC San Diego Moores Cancer Center Partnership Institute for Behavioral and
Community Health, San Diego, CA; 7Institute for Behavioral and Community Health, San
Diego, CA

Purpose: Compared to non-Hispanic whites, Hispanics/Latinos are at an increased risk for

cancer disparities. This intervention development study outlines the creation of a patient
navigation (PN) intervention designed to improve colorectal cancer (CRC) diagnostic care and
screening at a Federally Qualified Health Center (FQHC). Methods: Evidence-based
environmental and behavioral determinants from CRC literature and data from semi-
structured interviews with staff (n=17) at a FQHC provided the needs and context for the
intervention. Intervention Mapping (IM) steps 1-4, encompassing themes generated from the
Social Ecological Model (SEM), were then used to develop an outcome-oriented intervention.
Results: During step 1 of IM, researchers collected data from literature searches and
interviews conducted with the FQHCs staff to identify the specific assets and barriers
contributing to completion of recommended CRC screening and diagnostic care. In step 2,
researchers created a logic model of the problem based on the SEM model; this model
outlined the behavioral and environmental determinants and factors contributing to the
health problem (premature death from CRC and poor health-related quality of life).
Researchers then decided on specific behavioral and environmental outcomes for the
intervention (e.g., completing diagnostic/screening colonoscopy within 3 months of provider
order [behavioral], and navigator explaining a health care provider's recommendation for
screening/diagnostic colonoscopy in the patients preferred language [environmental]).To
accompany intervention outcomes, performance objectives describing how the outcomes
could be accomplished were included. To complete steps 3-4, researchers are using the
outcomes and performance objectives to create the programs structure, materials, and
S942 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

protocols using health behavior theory. Conclusion:This study describes the development of a
CRC PN intervention for Hispanic/Latinos served by a FQHC that has the potential of lessening
the CRC burden among this community.

CORRESPONDING AUTHOR: Tatianna Clark, NA, San Diego State University, Chula Vista, CA,
91911; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S943

B026 6:00 PM-7:00 PM

DEVELOPMENT OF AN EDUCATIONAL VIDEO TO PROMOTE CERVICAL CANCER SCREENING

AMONG BURMESE AND BHUTANESE REFUGEE WOMEN IN THE US

India Ornelas, PhD1, Bouapanh Lor, MPH1, Maya Magarati, PhD2, H. Hoai Do, Ying Zhang, MD,
MPH1, J. Carey Jackson, Victoria M. Taylor, MD, MPH3

1
University of Washington, Seattle, WA; 2Indigenous Wellness Research Institute, School of
Social Work, University of Washington, Seattle, WA; 3Fred Hutchinson Cancer Research
Center, Seattle, WA

Background: Immigrant and refugee women in the United States (US) have high cervical
cancer incidence rates coupled with low cervical cancer screening rates. Women from
Southeast Asian and other US communities with a high proportion of refugees are among the
most likely to have never been screened. Refugees from Burma and Bhutan were the two
largest refugee groups arriving in the US in 2011, making up 56 percent of all resettled
refugees. These refugee communities come from countries that experience marked cervical
cancer disparities. However, little is known about cervical cancer screening among these
refugee groups and few health education interventions have targeted this population.
Methods: Using the Behavioral Model for Vulnerable Populations as a conceptual framework,
we conducted eight focus groups in King County, Washington to gather information about
womens perceptions and knowledge of cervical cancer (need for care factors) as well as their
cervical cancer screening barriers and facilitators (predisposing and enabling factors). A total
of 58 women participated (Burmese n=31; Bhutanese n=27). Results: Only 28% reported
being screened for cervical cancer before coming to the US and 55% reported being screened
after resettling in the US. The focus groups revealed limited knowledge about cervical cancer
and the need for screening; barriers such as competing priorities and cost; and facilitators
such as trusted providers, interpreters, and support from social networks. We used our
findings to develop two 15 minute narrative educational videos (one in Nepali for Bhutanese
women and one in Karen for Burmese women) that provide basic information about womens
anatomy and reproductive health, guidelines for cervical cancer screening, a description of
screening procedures, and how to request female providers and interpreters. We are
currently pilot testing the videos to assess changes in knowledge and cervical cancer screening
intentions. Conclusions: Bhutanese and Burmese refugee women are in need of culturally
relevant health education about cervical cancer screening. If effective, educational videos
may serve as a useful tool for reaching a population at high risk for cervical cancer.
S944 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: India Ornelas, PhD, University of Washington, Seattle, WA, 98195;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S945

B027 6:00 PM-7:00 PM

DEVELOPMENT OF AN MHEALTH BEHAVIORAL CANCER PAIN PROTOCOL FOR MEDICALLY

UNDERSERVED COMMUNITIES

Sarah Kelleher, Ph.D.1, Caroline Dorfman, Ph.D.1, Rebecca Shelby, Ph.D.1, Beverly Thorn,
Ph.D., ABPP2, Francis Keefe, Ph.D.1, Tamara Somers, Ph.D.1

1
Duke University Medical Center, Durham, NC; 2The University of Alabama, Tuscaloosa, AL

Cancer patients in medically underserved areas, characterized by a shortage of healthcare

providers and services and economic or cultural barriers to healthcare, report higher pain,
more negative pain-related outcomes, and are less likely to have their pain adequately
treated. Women with breast cancer in these areas are particularly vulnerable to pain; they are
often diagnosed at a later stage, increasing their risk of disease- and treatment-related pain.
Behavioral cancer pain interventions can help with pain management. Persistent barriers (e.g.,
transportation, lack of experts) inhibit the use of these interventions with medically
underserved populations. To address this need, we developed a mobile-health (mHealth) Pain
Coping Skills Training protocol (mPCST) specifically tailored for women with breast cancer in
medically underserved areas. A nationally recognized expert in applying behavioral pain
protocols to patients with low-literacy in rural areas provided extensive input into tailoring
the intervention. Patient focus groups were conducted to aid in intervention development.
Participants (N=19) had a primary diagnosis of breast cancer, reported clinical pain at a 3/10
or greater (0=no pain, 10=worst pain imaginable), and received cancer treatments at a clinic in
a medically underserved area. Three 90-minute focus groups were held. Demographic,
medical, and self-reports of pain were collected. Participants provided verbal information
about their pain experiences and input on the intervention. Group feedback was used
iteratively to update the intervention. Field notes and audio recordings were reviewed, and
themes were identified. Women were M=60 years old, married (63%), and the majority were
African-American (53%); 25% had a college degree or more and 33% were working full or part
time. Women reported their pain to be M=3.2/10 in the last week. Participants reported pain
and fatigue were significant unaddressed concerns. Participants reported limited behavioral
intervention opportunities and were enthusiastic about mHealth resources. Participants
endorsed willingness to travel to their home clinic to participate in tele-video conferencing
and to use mobile phones as an adjunct intervention component. They suggested explicit
direction be given for use of mHealth technologies. Results suggest that an mPCST
intervention designed to meet the needs of breast cancer patients in medically underserved
areas is likely to be feasible and acceptable. Key protocol modifications based on focus group
data include addressing fatigue, modifying activity suggestions, and creating detailed mHealth
S946 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

technology instructions. mHealth technologies provide new opportunities to implement

efficacious behavioral cancer pain interventions and have the potential to benefit
underserved populations. We are currently testing the mPCST protocol in a small trial.

CORRESPONDING AUTHOR: Sarah Kelleher, Ph.D., Duke University Medical Center, Durham,
NC, 27705; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S947

B028 6:00 PM-7:00 PM

DEVELOPMENT OF THE TRUENTH PROSTATE CANCER PEER NAVIGATION CORE COMPETENCY

FRAMEWORK

Parminder K. Flora, PhD1, Jacqueline L. Bender, BSc, MSc, PhD1, Aaron Miller, MHI, MSc2, Lisa
Parvin, MSc3, Shimae Soheilipour, PhD4, Nandini Maharaj, PhD candidate4, Andrew G.
Matthew, PhD., C.Psych5, Arminee Kazanjian, DrSoc 4

1
Princess Margaret Cancer Centre, University Health Network, Toronto, ON, Canada; 2Prince
Margaret Cancer Centre, Toronto, ON, Canada; 3School of Population and Public Health,
University of British Columbia, Vancouver, BC, Canada; 4University of British Columbia,
Vancouver, BC, Canada; 5Princess Margaret Cancer Centre, Toronto, ON, Canada

Peer navigation (PN) is an emerging concept in cancer care where individuals who have
experienced cancer volunteer valuable informational, emotional and practical support to new
patients and their caregivers (CG; Freeman, 1995; CPAC, 2010). PN in prostate cancer (PC) is in
line with reports from men with PC about their preference for one-to-one support from peer
cancer survivors (King et al, 2015). While some existing literature is available to guide the
definition, role, and scope of peer navigators, no existing framework defined core
competencies specific for PC peer navigators. Furthermore, no known sources included the
CGs of men with PC as peer navigators. The study aim was to develop a set of core
competencies for PC peer navigators to guide the definition, role, and scope of peer
navigators, and a subsequent PN training program. Drawing on existing literature and
stakeholders, we developed the PN Core Competency Framework in 3 phases. Phase 1:
Identified PN knowledge, skills, abilities, or attributesfrom formative research including (a) a
systematic review, (b) an environmental scan of cancer navigation programs, and (c) semi-
structured qualitative interviews exploring PC navigational support needs among patients,
CGs, and healthcare providers. Phase 2: Drafted initial set of competency statements and
mapped into domains. Phase 3: Verified competency statements with PN stakeholders via an
online survey in which agreement with inclusion (yes/no) and priority (low/medium/high)
were rated. In the verification phase, 27 participants (19 patients, 3 CGs, and 5 healthcare
providers) completed the online survey. The majority of participants rated the competencies
as medium and high priority. Most statements had an affirmative consensus for inclusion,
with comments about exclusion pertaining to concerns about overlapping role with
healthcare providers, time required to execute, and limited knowledge/information.
Following revisions, the resulting consensus-based framework consisted of 6 core competency
domains for PC peer navigators: (1) navigator attributes (e.g., professionalism, cultural
S948 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

competency, responsibility, self-awareness, empowerment, problem-solving, character,

personal and professional development); (2) communication; (3) knowledge/information; (4)
facilitate patient-centered care; (5) eHealth/technology; and (6) CG needs. This framework
can be used to develop a PN training program to prepare PC survivors for a peer navigator
role.

CORRESPONDING AUTHOR: Parminder K. Flora, PhD, Princess Margaret Cancer Centre,

University Health Network, Toronto, ON, M5G 2C4; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S949

B029 6:00 PM-7:00 PM

DIFFERENT METHODS OF LOCALIZATION FOR NON-PALPABLE BREAST TUMORS AND PATIENT

REPORTED ANXIETY

Cindy Matsen, MD1, Yue Zhang, PhD2, Matthew Stein, MD3, Kimberly Kaphingst, ScD4

1
Huntsman Cancer Institute at the University of Utah, Salt Lake City, UT; 2University of Utah,
Salt Lake City, UT; 3University of Utah School of Medicine, Salt Lake City, UT; 4University of
Utah and Huntsman Cancer Institute, Salt Lake City, UT

Background: Women with breast tumors that cannot be felt on exam need to have a
localization procedure done in radiology to help the surgeon find the right area to take out.
The most common way this is done is with a wire (WL) that is placed into the breast on the
day of surgery. An alternative method uses a radioactive seed (SL) that gives off a signal that
can be detected with a probe in the operating room. The seed is contained entirely within the
breast and can be placed days before surgery. There are several advantages to using seed
localization with the scheduling advantages for the surgeon and radiologist being the most
widely recognized, but there is very little data on the patient experience and no study has
specifically looked at patient reported anxiety at the time of the procedure As we
implemented a seed localization program in our institution, we aimed to explore the patient
experience with a specific focus on peri-procedure anxiety. We hypothesized that WL would
be associated with greater anxiety because it must be done on the day of surgery and because
WL can be associated with delaying the surgery start time which may be anxiety provoking.

Methods: Patients with non-palpable, invasive or pre-invasive breast cancer with no history of
anxiety disorder, who had not undergone a localization procedure in the past were recruited.
The WL group consisted of patients prior to implementation of the SL program and patients
in the SL group were recruited after program implementation. All patients completed a
validated, single-item anxiety assessment (5 point Likert scale with a 0.78 correlation to the
State Trait Anxiety Inventory) before and after the procedure. Scheduled surgery time and the
time the patient arrived in the OR were recorded for all patients. Logistic regression models
were used to compare the likelihood of increased anxiety and time delay in the start time in
OR between the two groups, controlling for age and family history of cancer.

Results: The 29 patients in the WL group and 31 patients in the SL group were well matched
for all demographic and disease characteristics. There were no significant differences in pre-
procedure, post-procedure, or change between pre and post-procedure anxiety between the
two groups. The majority of patients reported low pre and post-procedure anxiety. Patients
S950 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

with high levels of post-procedure anxiety were seen more often in the WL group, but this
was not statistically significant. There were significant differences in delay time to the OR with
seed localization associated with lower likelihood of delays (OR for seed 0.024 (95% CI 0.003-
0.207), p < 0.001).

Conclusions: Implementation of a seed localization program reduced delays to the operating

room. Delays to the operating room decrease patient satisfaction and impact negatively on
the patient experience, but this is not associated with anxiety levels around the time of the
localization procedure.

CORRESPONDING AUTHOR: Cindy Matsen, MD, Huntsman Cancer Institute at the University of
Utah, Salt Lake City, UT, 84112; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S951

B030 6:00 PM-7:00 PM

DO SPOUSAL CANCER CAREGIVERS RECEIVE ADEQUATE DEPRESSION TREATMENT? EVIDENCE

FROM A POPULATION-BASED STUDY

Kristin Litzelman, PhD1, Abiola Keller, PA-C, MPH, PhD2

1
University of Wisconsin-Madison, Madison, WI; 2Marquette University, Milwaukee, WI

Background: Recent research suggests that mental health problems in spouses are associated
with worse mental health in cancer survivors. Adequately treating spousal mental health
problems therefore represents an opportunity to improve outcomes for both cancer survivors
and their co-surviving family members. Using nationally representative data, this study sought
to determine how depression treatment differs between spousal cancer caregivers with
depression compared to the general married population and assess rural/urban disparities in
treatment.

Methods: Data are from the Medical Expenditures Panel Survey, a household-based survey of
US adults; we concatenated data from 2004-2013. We identified spouses of cancer survivors
and a comparison group of married adults. Key measures included depression (ICD-9 codes
generated from participant interviews), adequacy of depression treatment (at least four
prescriptions related to depression treatment, or at least eight psychotherapy or counseling
visits), and sociodemographics. We calculated the prevalence of depression in spousal
caregivers and comparisons. Among caregivers and comparisons with depression, logistic
regression evaluated the association between caregiver status and receipt of adequate
treatment, controlling for sociodemographics; secondary analyses included rurality as a
moderator. Analyses were weighted to account for the complex sampling design.

Results: Twelve percent of spousal caregivers were depressed, compared with 9% of the
general married population. Most depressed caregivers received guideline concordant
treatment (65%), with no difference from comparisons (60%; p=0.50). Urban caregivers were
less likely to receive adequate treatment compared to urban comparisons (Odds Ratio [OR]:
0.39, 95% Confidence Interval [CI]: 0.17-0.91). Rural caregivers did not differ significantly from
urban comparisons (OR: 0.88, 95% CI: 0.59-1.31) or rural comparisons (OR: 1.05, 95% CI: 0.72-
1.52). However, rural comparisons were less likely to receive treatment than urban
comparisons (OR: OR: 0.84, 95% CI: 0.72-0.98).

Conclusions: Although spousal caregivers overall receive adequate depression treatment at

rates roughly equal to their non-caregiving counterparts, these data suggest that urban
caregivers may be at increased risk of inadequate treatment. Future research should examine
S952 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

and seek to mitigate clinical- and community-level barriers to depression care for urban
caregivers.

CORRESPONDING AUTHOR: Kristin Litzelman, PhD, University of Wisconsin-Madison,

Madison, WI, 53706; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S953

B031 6:00 PM-7:00 PM

EFFECTS OF SOCIAL MEDIA ADVERTISEMENT CONTENT ON THE RECRUITMENT OF YOUNG GAY

AND BISEXUAL MEN FOR AN HPV VACCINE INTERVENTION

Paul L. Reiter, PhD1, Annie-Laurie McRee, DrPH2

1
The Ohio State University, Columbus, OH; 2University of Minnesota, Minneapolis, MN

Background. Social media is a promising approach for reaching and recruiting participants for
research studies. We sought to evaluate the effects of different images and text included in
social media advertisements (ads) used to recruit young gay and bisexual men for an online
human papillomavirus (HPV) vaccine intervention.

Methods.In July and September 2016, we used paid Facebook advertisements to recruit men
who were ages 18-25, self-identified as gay or bisexual, lived in the US, and had not received
any doses of HPV vaccine. A 4x2x2 factorial experiment varied the ad image (a single young
man, a male couple, a group of young men, or a young man talking to a doctor), content focus
(text about HPV infection or HPV vaccine), and disease framing (text about a sexually
transmitted disease [STD] or cancer). Poisson regression determined if these experimental
factors affected ad performance.

Results. The recruitment campaign reached a total of 35,646 Facebook users and resulted in
an overall unique click-through rate of 2.01% and an overall conversion rate of 0.66%. Ads
that included an image of a male couple had a higher reach (IRR=4.91, 95% CI: 2.68-8.97) and
conversion rate (IRR=2.56, 95% CI: 1.13-5.77) compared to ads that included an image of a
single young man. The unique click-through rate was higher for ads with text about an STD
compared to ads with text about cancer (IRR=1.34, 95% CI: 1.06-1.69). The campaign cost a
total of $413.72 and resulted in 150 eligible and enrolled individuals ($2.76 per enrolled
participant).

Conclusions. Facebook was an efficient and cost-effective approach for reaching and recruiting
a national sample of young gay and bisexual men for an HPV vaccine intervention. The
success of individual ads varied based on their image and text. Researchers should consider
the importance of ad content in order to maximize the recruitment of participants into health-
related studies via social media.

CORRESPONDING AUTHOR: Paul L. Reiter, PhD, The Ohio State University, Columbus, OH,
43201; [email protected]
S954 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B032 6:00 PM-7:00 PM

EFFECTS OF THE DOCTOR-PATIENT RELATIONSHIP ON LUNG CANCER PATIENTS TREATMENT

EXPECTATION AND QUALITY OF LIFE

Huynh Nguyen, BS1, Laurie E. Steffen, PhD2

1
University of New Mexico, Rio Rancho, NM; 2Wake Forest University School of Medicine,
Winston-Salem, NC

The doctor-patient relationship is a key component of patient quality of life in terminal illness.
Patients present with different backgrounds and belief systems about illness, health, and
death that can make effective communication with their physician challenging and shape
treatment expectations. The extent to which a patient feels their physician is culturally
competent may impair or enhance communication and care; therefore, understanding the
relationship between patient perceptions of their physician and quality of life is important to
improving patient care. Our study used a 21-day longitudinal design in which 20 lung cancer
patients (75% stage IIIB/IV disease, MN age = 73 years; MN education = 13 years; 60% female,
80% white) completed questionnaires regarding their daily quality of life, treatment
expectations, and perception of their physicians cultural competency, communication, trust,
and partnership building. Higher ratings of physicians patient-centered cultural competence
corresponded with higher ratings of trust (r = .67, p = .001), effective communication (r = .57,
p = .009), and partnership building (r = 0.69, p = .001). Patient information index, a subscale of
communication, was a significant predictor of treatment expectation (b = 1.48, 95% CI [.71,
2.3]). Better physician communication predicted higher daily social and role functioning (b =
4.04, 95% CI [1.12, 6.96]), higher daily physical functioning (b = 4.51, 95% CI [1.28, 7.74]), and
fewer physical symptoms (b = -.11, 95% CI [-.18, -.04]. To our knowledge, this is the first study
to relate aspects of physician communication to daily quality of life during treatment for
advanced lung cancer. Results underscore that the doctor-patient relationship is multifaceted
and that physician communication does play an important role in patients quality of life.
Results highlight the need for increased attention to the quality of physician interactions with
patients and cultural factors, as these interactions impact patients daily lives during illness.

CORRESPONDING AUTHOR: Huynh Nguyen, BS, University of New Mexico, Rio Rancho, NM,
87124; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S955

B033 6:00 PM-7:00 PM

EMPOWERING BREAST CANCER PATIENTS WITH WEARABLE TECHNOLOGY: HEALTH LITERACY,

COMMUNICATION, AND CONNECTION: A PILOT STUDY

Cori McMahon, Psy.D.1, Tina Harralson, Ph.D.2, Kristen Scott, RN,BSN3, Kelly Lynn. Gilrain,
PhD4, Philip Fizur, MA5, Mark Redlus, BS3

1
Cooper University Hospital, Department of Medicine, Woodbury, NJ; 2Polaris Health
DIrections, Inc., Wayne, PA; 3Polaris Health Directions, Wayne, PA; 4Cooper University
Hospital, Camden, NJ; 5MD Anderson Cancer Center at Cooper, Cooper University Hospital,
Camden, NJ

According to the NCI, it was estimated that in 2013, there were over 3 million women living
with breast cancer in the US; that there will be more than 249,000 new cases of breast cancer
in 2016, and that 89.7% of those women will survive more than 5 years. With increasing
survival rates and well-established benefits of patient-centered communication, such as
improved quality of life, adherence to medical treatment regimen, and health literacy, there is
a need for more efficient communication and monitoring of psychosocial and physical
functioning. Wearable technology allows patients to consistently monitor symptoms and to
quickly share that information with the medical team. The current study aims to assess the
utility of a novel app used via Apple watch and iPhone in improving communication,
improving health literacy, and tracking patient data, including passive data (heart rate, activity
level), self-report data (mood, pain, and fatigue levels), and engagement data (use of social
media, daily challenges, ability to communicate with study cohort). Twenty women with
various stages of breast cancer were enrolled in the study: mean age 50 years (range: 27-66
yrs.); 80% White, 20% African American; 15% having completed some high school and 85%
completing at least some college (20% graduate school). 70% of subjects reside with spouse
and/or children with only 5% live alone. Intake assessment completed at enrollment
indicated: 60% moderate to severe distress, 15% moderate to severe anxiety (GAD7), and 20%
moderate to severe depression (PHQ9). Overall, subjects reported very good social and
emotional well-being (FACT-B) and showed a range of difficulty with functional well-being
(mean 18.6, SD 6.7, range: 6-27) on a scale from 28 (excellent) to 0 (very poor). Subjects
reported very good satisfaction with treatment team, including communication with medical
staff, confidence and trust, and willingness to recommend providers. Qualitative feedback has
allowed researchers to tailor the app and includes requests for afternoon (as opposed to
morning) check-in about mood, pain, and fatigue; inclusion of a sufficient number of daily
challenges (i.e. take a hike, talk to a friend); and for data to be incorporated into medical chart
for monitoring. Subjects reported overall satisfaction with the app and using wearable
S956 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

technology for self-report and monitoring, appreciated being electronically connected, and
learning about their patterns of physical and emotional functioning during treatment, which,
many noted, improved targeted communication with physicians during medical exams.

CORRESPONDING AUTHOR: Cori McMahon, Psy.D., Cooper University Hospital, Department of

Medicine, Woodbury, NJ, 08096; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S957

B034 6:00 PM-7:00 PM

ENHANCING BREAST CANCER SURVIVORSHIP FOR KOREAN WOMEN USING A RANDOMIZED

CONTROLLED TRIAL TO IMPROVE QOL AND DECREASE STRESS

Sunmin Lee, ScD1, Mary Y. Jung, MPH2, Min Qi Wang, PHD3

1
University of Maryland School of Public Health, College Park, MD; 2University of Maryland
School of Public Health, Kensington, MD; 3UNIVERSITY OF MARYLAND, COLLEGE PARK, MD

Background: Existing research on cancer survivorship among Asian subgroups is highly sparse.
Specifically, Korean American women have experienced a surge of breast cancer incidence in
recent years and report low levels of quality of life (QOL) and high levels of stress, partly due
to linguistic and cultural barriers, which diminish their ability to cope with cancer. However,
few studies have examined breast cancer survivorship in this population underscoring the
strong need for linguistically and culturally appropriate resources for Korean American breast
cancer survivors (KABCS). We developed and assessed the effectiveness of one of the first
linguistically and culturally tailored randomized controlled trial (RCT) for KABCS.

Methods: This RCT aimed to improve QOL and decrease stress using cognitive-behavioral
stress management. Participants attended seven 2-hour weekly sessions and one follow-up
session seven weeks after the intervention. The analytic sample consisted of 77 KABCS in the
Washington, D.C. metropolitan area. The standard intervention (SI) group (n=37) were
provided linguistically and culturally appropriate information related to breast cancer
survivorship, whereas the enhanced intervention (EI) group (n=40) were provided the SI
content in addition to information on practicing relaxation, enhancing coping, reducing stress,
using social support, managing anger, and being assertive.

QOL and stress were measured at baseline (T1), end of the program (T2), and follow-up (T3).
The Functional Assessment of Cancer Therapy-Breast (FACT-B), which consists of five
subscales [physical well-being (PWB), social well-being (SWB), emotional well-being (EWB),
functional well-being (FWB), and breast cancer subscale (BCS)] was used to assess QOL, and
the Perceived Stress Scale (PSS) was used to measure stress. Generalized estimating equations
(GEE) models were used to examine program effects on QOL and PSS separately.
S958 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Results: The EI group had a 5.2 unit greater increase in mean FACT-B scores, 1.7 unit greater
increase in mean SWB scores, 1.1 unit greater increase in mean EWB scores, and 2.2 unit
greater increase in mean FWB scores from T1 to T2 as compared to the SI group.

For mean PSS scores, EI group participants had a 2.4 unit greater decrease in mean PSS scores
as compared to those in the SI group from T1 to T2. Among participants who resided in the
U.S. for less than 10 years, the EI group had a significantly greater decrease of 6.8 units in
mean PSS scores from T1 to T2 as compared to the SI group (p= 0.03). Likewise, among
participants who resided in the U.S. for 10 or more years, EI group participants had a
marginally greater decrease of 2.2 units in mean PSS scores (p= 0.05).

Conclusions: Findings support the use of a linguistically and culturally tailored support
program to improve QOL and decrease stress among KABCS.

CORRESPONDING AUTHOR: Sunmin Lee, ScD, University of Maryland School of Public Health,
College Park, MD, 20742; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S959

B035 6:00 PM-7:00 PM

EVALUATION OF BLACK/AFRICAN AMERICAN CANCER SURVIVORS HEALTH BEHAVIORS

Matthew Asare, PhD, MBA, CHES1, Karen Mustian, PHD, MPH, MS2, Luke J. Peppone, PhD,
MPH3, Joseph J. Guido, MS2, Gary Morrow, PhD, MS4, Calvin L. Cole, PhD1, Ian Kleckner, PhD2,
Charles Kamen, PhD, MPH2

1
University of Rochester, Rochester, NY; 2University of Rochester Medical Center, Rochester,
NY; 3University of Rochester, Wilmot Cancer Institute, Rochester, NY; 4university of Rochester,
pittsford, NY

Introduction: Healthy eating, physical activity, tobacco cessation and maintaining healthy
weight are healthy lifestyles can help cancer survivors control their risk for cancer recurrence,
secondary cancers, and improve health outcomes and overall quality of life. However, the
health behaviors of Black/African American cancer survivors have been understudied. We
examined young, middle-aged, and older black/African American cancer survivors health
behaviors.
Methods: 2,878 Black/African American cancer survivors (age 63.79 12.59 years)
participated in the post-diagnosis survey and it was mixed cancer types. Survivors self-
reported their body mass index (BMI), and health behaviors (Yes/No for exercising, smoking,
alcohol use, and fruit/vegetable consumption). Chi-squared analyses and analyses of variance
(ANOVA) were used to compare young (18-39 years), middle-aged (40-64 years) and older
adults (>65 years) health behaviors.
Results: Middle-aged and older survivors were more likely to be obese/overweight than
young survivors (young 64%, middle-aged 78%, older 75%). Older survivors were more likely
to be current drinkers than middle-aged and young survivors (young 49%, middle-aged 68%,
older age 78%). Young survivors were more likely to be binge drinkers (young 18%, middle-
aged 7%, and older 3%); and current smokers (young 31%, middle-aged 23%, and older 11%).
Older adult and young survivors were more likely than middle-aged survivors to lead
sedentary lifestyle (young 42%, middle-aged 28%, older adult 43%). Middle-aged survivors
were less likely to eat one or more fruit servings per day (young 60%, middle-aged 58%, and
older 65%). All comparisons p< 0.05.
Conclusion: Differences were noted across the age spectrum for unhealthy behaviors. While
young African American survivors reported more unhealthy behaviors like binge drinking,
tobacco use, and sedentary behaviors, middle-aged survivors reported low fruits and
vegetables consumption, and older adults reported high alcohol consumption. Clinical efforts
should be tailored to African American survivors to address their specific unhealthy behaviors
including life stage considerations. Health behavior interventions at community and individual
S960 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

levels should be implemented to educate Black/African cancer survivors, especially the young
survivors about healthy behaviors and its impact on health outcomes and quality of life.

CORRESPONDING AUTHOR: Matthew Asare, PhD, MBA, CHES, University of Rochester,

Rochester, NY, 14642; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S961

B036 6:00 PM-7:00 PM

EXPERIENCE OF CANCER TREATMENT IN THE US AMONG PATIENTS FROM SAUDI ARABIA

Salah Alshagrawi, MS Health Care Adminstration

TAMU, College Station, TX

Statement of the problem: An increasing number of Saudi cancer patients seek treatment in
the United States. To understand this groups experience, we examined the perspectives from
a group of Saudi cancer patients in a large U.S. metropolitian cancer center.

Study design: A qualitative grounded theory study.

Sample size and composition: Twelve cancer patients from Saudi Arabia; six females and six
males, who varied in ages, diagnoses, therapies and treatment durations.

Measures utilized: In-depth face-to-face interviews

Analysis method: Interviewes were audio recorded and transcribed. A grounded theory was
employed to collect and analyze the data through constructing analytical codes and
categories.

Results: Our sample recognized the professionalism and kindness of their U.S. healthcare
providers; especially the knowledge that their physicians displayed and their skills, expertise,
and attitudes. Moreover, they were successful in building a trustful relationship with their
providers. Although interviewees encountered difficulties of living abroad, the trust with they
had on their providers and the support of other Saudi cancer patients enhanced their faith in
cancer treatment internationally.

Conclusion: Our Saudi cancer patient participants had a positive experience in receiving
cancer treatment in the United States.

CORRESPONDING AUTHOR: Salah Alshagrawi, MS Health Care Adminstration, TAMU, College

Station, TX, 77845; [email protected]
S962 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B037 6:00 PM-7:00 PM

EXPRESSED AFFECT AND PERCEPTIONS OF CLINICAL INTERACTIONS

Nicole Senft, PhD1, Lauren M. Hamel, PhD2, Louis A. Penner, PhD2, Terrance L. Albrecht, PhD2,
Felicity Harper, PhD3, Tanina Foster, PhD2, Robert Chapman, MD4, Eggly Susan, PhD2

1
Karmanos Cancer Institute/Wayne State University, Ann Arbor, MI; 2Wayne State
University/Karmanos Cancer Institute, Detroit, MI; 3Karmanos Cancer Institute/Wayne State
University, Detroit, MI; 4Josephine Ford Cancer Center, Detroit, MI

Physicians and patients convey information via emotional behavior that is used in forming
judgments about each other and their interactions. The current research is a preliminary
investigation of the role of emotional expressivity in clinical interactions between Black cancer
patients and their non-Black physicians. This is a secondary data analysis examining patient-
physician communication during discussions of cancer treatment for expressions of emotion.
Nave coders rated the emotional expressivity of patients and physicians, as well as the
rapport between them, using five thin slices of 74 interactions. Using generalized estimating
equation (GEE) regression models to control for patients nested within physicians, we tested
whether observer ratings of patient and physician positive affect, negative affect, or the
combination of both predicted (a) observer ratings of patient-physician rapport and (b)
patients post-visit perceptions of physician patient-centeredness. The interaction between
patients positive and negative affect predicted perceptions of both rapport (=.12, SE=.03, p
< .001) and patient-centeredness (=.10, SE=.03, p=.004). Observers perceived higher rapport
among patients displaying positive affect, especially when patients also expressed relatively
high levels of negative affect.Patients perceived their physician to be less patient-centered if
they had high levels of expressed negative affect and low levels of expressed positive affect,
whereas expressed negative affect was unrelated to patient-centeredness among those who
had high levels of expressed positive affect. Regarding physician affect, physicians expressed
negative affect was inversely related to ratings of rapport (=-0.31, SE=.05, p < .001) and
patient-centeredness (=-.36, SE=.14, p=.011). Physicians positive affect was unrelated to
either outcome (ps>.10). These findings suggest patients positive and negative emotional
expressivity jointly contribute to positive perceptions of physicians and interactions, perhaps
by eliciting empathic responses from physicians. However, in reverse, physicians high
expressed negative affect may have negative consequences for patient perceptions and the
quality of interactions. Future work will examine the contribution of pre-interaction attitudes
and experiences that may influence patient and physician emotional expressivity.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S963

CORRESPONDING AUTHOR: Nicole Senft, PhD, Karmanos Cancer Institute/Wayne State

University, Ann Arbor, MI, 48103; [email protected]
S964 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B038 6:00 PM-7:00 PM

FACTORS ASSOCIATED WITH INTENTIONS OF HPV INITIATION AMONG VACCINE NAVE

FEMALES WITH AND WITHOUT A HISTORY OF CHILDHOOD CANCER

Robin Hardin, M.A.1, Kathryn Russell, Ph.D. 2, Jessica Simmons, M.S.2, Jasmine Eddinger, M.S.
3
, Lauren Dahlke, B.A. 1, James Klosky, Ph.D.2

1
St. Jude Children's Research Hospital and University of Memphis, Memphis, TN; 2St. Jude
Children's Research Hospital, Memphis, TN; 3University of Missouri-Kansas City, Kansas City,
MO

Purpose: Effective vaccination is available to prevent many high-risk types of human

papillomavirus (HPV), the most common sexually transmitted infection and cause of cervical
cancer. This study aimed to examine vaccine intentions among vaccine nave young adults
with and without a cancer history in addition to identifying factors associated with HPV
vaccination intentions.

Methods: Vaccine nave females (aged 18-26 years, n = 120) completed surveys querying HPV
vaccination intentions, HPV knowledge and communication, along with sociodemographic,
medical, and health belief factors specific to HPV vaccine intent. Multivariate logistic
regression was utilized to calculate odds ratios (OR) and 95% confidence intervals (CI) for HPV
vaccination intentions.

Results: Univariate analyses indicated no differences between participants with/without

history of cancer on intention to vaccinate or considered risk factors. Multivariate logistic
regression analyses for those who self-identified as non-white revealed that those who
reported their physician recommended HPV vaccination were more likely to report having
intentions to vaccinate (ORs = 4.25 8.08, ps p =.014).

Conclusion: There are various factors that affect vaccine intentions among vaccine nave
young adult women. Future interventions aimed at increasing HPV vaccination intentions
among childhood cancer survivors should include training providers in recommending the
vaccine as a mechanism to increase vaccine intentions, and subsequently, vaccine initiation.
Additionally, future research should further examine the relationship between HPV-related
knowledge and intentions to vaccinate to elucidate this inverse relationship in
underrepresented racial groups.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S965

CORRESPONDING AUTHOR: Robin Hardin, M.A., St. Jude Children's Research Hospital and
University of Memphis, Memphis, TN, 38106; [email protected]
S966 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B039 6:00 PM-7:00 PM

FACTORS THAT INFLUENCE CERVICAL CANCER SCREENING UPTAKE AMONG LESBIAN AND
BISEXUAL WOMEN AND TRANSGENDER MEN

Michael Johnson, Ph.D., RN

University of Nevada, Las Vegas, Henderson, NV

The purposes of this study were to measure the prevalence of, and identify factors associated
with, cervical cancer screening uptake among a sample of lesbian, bisexual and queer (LBQ)
women, and transgender men. Past research has found that LBQ women underutilize cervical
screening services, and because deficient screening remains the most significant risk factor for
cervical cancer, it is essential to understand the differences between routine (Pap test < 3
years) and non-routine (Pap test > 3 years) screeners. The widely used health care utilization
framework, Andersens Behavioral Model of Health Services Use (BMHSU), was used as a
structure to mix the data sets and contextualize the results. It was chosen because it accounts
for the different levels of determinants. Participants for this convergent-parallel mixed
methods study were conveniently recruited from the Internet and community events from
August to December 2014. To be included, participants had to be female or transgender male,
age 21-65 years and identify with a sexual orientation other than heterosexual. Quantitative
data were collected via a 48-item Internet questionnaire, and qualitative data were collected
through in-depth interviews. The quantitative and qualitative data were analyzed separately
and then mixed. Quantitative data analyses entailed four steps: 1) screening the data and
reporting descriptive statistics for the observed cases, 2) multiple imputations, 3) bivariate
analyses, and 4) multivariate logistic regression. Qualitative data were analyzed using
deductive-inductive content analysis. Finally, a matrix approach was used to compare the
quantitative and qualitative results. Seventy-three percent of the participants received
routine cervical cancer screening. Twenty factors were found to be statistically significantly
different between routine and non-routine screeners. Eighteen themes/factors were found in
the qualitative data. Factors triangulated during the mixed analyses if they were statistically
significant and identified in the qualitative data. Routine screeners reported feeling more
welcome in the health care setting, while non-routine screeners reported more discrimination
related to their sexual orientation and gender expression. Routine screeners were also more
likely to be 'out' to their provider. Many of the factors identified in this study related to the
healthcare environment and to interactions between the patient and provider. The factors
found in the quantitative and qualitative data have implications for research and practice
around cancer prevention for LBQ women and transgender men. The factors that contrasted
between the quantitative and qualitative data sets are signals for additional research.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S967

CORRESPONDING AUTHOR: Michael Johnson, Ph.D., RN, University of Nevada, Las Vegas,
Henderson, NV, 89052; [email protected]
S968 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B040 6:00 PM-7:00 PM

FEAR OF PALLIATIVE CARE IN YOUNGER ADULTS WITH CANCER: AN INDIRECT EFFECT OF

DEPRESSION SYMPTOM SEVERITY

Ashley Beggin, N/A1, Laura M. Perry, BS2, Michael Hoerger, PhD, MSCR1

1
Tulane University, New Orleans, LA; 2Tulane University, New Orlenas, LA

Title: Fear of Palliative Care in Younger Adults with Cancer: An Indirect Effect of Depression
Symptom Severity

Objectives: Drawing upon socioemotional selectivity theory, this study examined differences
in fear of palliative care consultations based on age and depression symptom severity.
Increased knowledge about how different age groups view palliative care could help inform
tailored interventions in the future.

Methods: The sample of 598 participants was drawn from Project EMPOWER, an internet
based palliative care educational intervention study. Participants had no prior palliative care
consultations and were adults with cancer in current oncological care. The sample was
heterogeneous in terms of caner type, had an age range of 27 to 93 years old, and was 42.8%
female. Participants completed self-report measures of their health history, depression
symptom severity (PROMIS), and fear of palliative care (Palliative Care Preference Scale
emotion subscale). Hierarchical regression was used to determine how baseline palliative care
preferences varied based on age.

Results: Results indicated that younger age was associated with greater fear of palliative care
consultations (p < .01). This association was mediated by depression symptom severity.

Discussion: Younger participants indicated greater fear of palliative care consultations,

potentially due to elevated levels of depression symptoms. Future studies should examine
whether educating patients about the impact of mood on preferences reduces barriers to
palliative care utilization.

Funding sources: U54GM104940

CORRESPONDING AUTHOR: Ashley Beggin, N/A, Tulane University, New Orleans, LA, 70118;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S969

B041 6:00 PM-7:00 PM

FEASIBILITY OF A PHYSICAL ACTIVITY INDEX IN CLINICAL PRACTICE: PERSPECTIVES OF

PROVIDERS AND CANCER SURVIVORS

Shirley M. Bluethmann, PhD, MPH1, Sarah Keadle, PhD, MPH2, Charles Matthews, PhD1, James
McClain, PhD, MPH1, Frank Perna, EdD, PhD1

1
National Cancer Institute, Rockville, MD; 2Cal Poly, San Luis Obispo, CA

Background: Physical activity (PA) is a well established behavioral recommendation for older
adults and has specific benefits for cancer survivors. However, most older adults and survivors
do not meet PA recommendations. Clinical tools are being sought to help counsel these
increasingly intersecting groups, but the feasibility of these tools in clinical practice is not well
understood. We developed an integrated Physical Activity Index (PAI) that synthesizes
physical and behavioral components to assist with this task. The purpose of this project was to
conduct interviews to assess the feasibility and clinical utility of the PAI from both provider
and patient perspectives.

Methods: We recruited 18 participants from the Washington, DC area (9 providers, 9 post-

treatment cancer survivors) for one-on-one, in-person interviews. Providers were purposively
sampled to include a mix of practice settings (e.g., primary care and oncology). Eligible
providers reported seeing survivors regularly; eligible survivors had completed treatment
within the last 5 years. Interviews were recorded and transcribed verbatim. An interview
guide and interviewer training ensured consistently in interview procedures. Segmented
coding procedures tracked responses to specific PAI elements (i.e., cognitive assessment).
Thematic coding synthesized qualitative data to assess barriers and feasibility of implementing
the PAI in clinical practice.

Results: Five oncology providers (including 2 nurses) and 4 primary care providers (including 2
nurses) were interviewed. Patient participants included 5 prostate and 4 breast cancer
survivors with a mean age of 69 yrs (range 55-79 yrs) and mean time of 4 years post
treatment. Most providers were receptive to the PAI as a brief screener and liked the idea of
report-card style output to track and counsel patients on lifestyle changes, especially if this
could be linked to electronic medical records (EMR). Most survivors had received a provider
recommendation to exercise, but liked the idea of creating customized exercise goals and
partnering with providers to achieve them. For barriers, providers worried about having time
to explain the tool and output to patients, a concern also shared by many survivors. Some
survivors mentioned symptoms and chronic conditions as potential barriers to lifestyle
S970 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

changes.

Conclusion: The PAI could be a feasible tool in clinical practice, for providers and survivors,
especially if the tool could be integrated with the EMR to assist providers and patients in
pursuing PA recommendations. Supplementary provider education resources on patient
counseling would facilitate use of the PAI in primary care and oncology practices. Survivors
may need additional support from the medical team to adapt PA goals based on individual
health concerns, especially older survivors or those with multi-morbidities.

CORRESPONDING AUTHOR: Shirley M. Bluethmann, PhD, MPH, National Cancer Institute,

Rockville, MD, 20850; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S971

B042 6:00 PM-7:00 PM

FORGIVENESS AS A BUFFER TO LUNG CANCER STIGMA

Kristen E. Riley, Ph.D. 1, Heidi Hamann, PhD2, Megan Shen, PhD3, Anna J. Thomas, Ph.D.4,
Jamie S. Ostroff, Ph.D.1

1
Memorial Sloan Kettering Cancer Center, New York, NY; 2University of Arizona, Tucson, AZ;
3
Weill Cornell Medicine, New York, NY; 4Baylor University Medical Center, Dallas, TX

The majority of individuals diagnosed with lung cancer report experiencing stigma, such
as perceiving that they are responsible and/or deserving of lung cancer. Stigma is associated
with a number of deleterious psychosocial and medical outcomes in lung cancer patients,
including delayed medical help-seeking, increased psychosocial distress, lower quality of life,
and poorer provider communication. As such, a better understanding of modifiable targets for
development of stigma-reducing interventions are needed. Forgiveness, or resolving
emotional and cognitive distress through an intraindividual, prosocial change process, has
been shown to decrease stigma and buffer individuals from the deleterious effects of stigma
broadly. However, forgiveness has never been studied in the context of lung cancer stigma.
We aim to examine whether forgiveness provides a buffer to stigma in lung cancer patients in
order to guide the development of interventions for better mental and physical wellbeing
outcomes in this vulnerable population of cancer survivors.

Method

Lung cancer patients (N = 231) completed a battery of questionnaires assessing

demographic information, lung cancer stigma (Lung Cancer Stigma Inventory; Hamann, Shen,
Thomas, Craddock Lee, & Ostroff, 2016) and forgiveness (Forgiveness subscale of the Fetzer
Multidimensional Measure; Fetzer Institute, 1999; Neff, 2006), as part of a larger study.

Results & Discussion

Internal reliability for the forgiveness subscale in this sample was good ( = .758).
Bivariate correlations between forgiveness and lung cancer stigma indicated a statistically
significant negative relationship (r = -.211 p < .001). Stepwise regression analyses indicated
that forgiveness was a statistically significant correlate of lung cancer stigma, even when
accounting for age, gender, race, ethnicity, and education (B = -2.27, p = .008).

It appears as though forgiveness may mitigate lung cancer stigma such that more
S972 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

reported forgiveness of self, God, and others is related to less reported lung cancer-related
stigma. Forgiveness may serve as a buffer against experiencing lung cancer stigma, or may
protect against the deleterious effects of lung cancer stigma, and may serve as a target for
interventions. These preliminary findings must be confirmed in future longitudinal studies.

CORRESPONDING AUTHOR: Kristen E. Riley, Ph.D. , Memorial Sloan Kettering Cancer Center,
New York, NY, 10022; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S973

B043 6:00 PM-7:00 PM

GENDER ROLE NORMS AND SPIRITUAL PEACE AMONG CHINESE HUSBANDS OF BREAST
CANCER SURVIVORS: THE MEDIATING ROLE OF COPING PROCESSES

Nelson Yeung, Ph.D.1, Lu Guohua, Ph.D2, Qian Lu, MD, PhD3

1
The Chinese University of Hong Kong, Hong Kong, N/A, Hong Kong; 2Weifang Medical
University, Weifang, Shandong, China (People's Republic); 3University of Houston, Houston, TX

Studies have found that stress/coping processes are associated with well-being among
husbands of breast cancer survivors (BCS). However, little is known about how gender role
norms associate with these stress/coping processes, and subsequently affect the husband
caregivers well-being. This study applied the Stress and Coping Model to examine the
psychosocial correlates of spiritual peace among Chinese husbands of BCS. Participants
(n=176, mean age =54.4) were recruited in Weifang, China. Their individual characteristics
(including male gender role norms), caregiving burden, self-efficacy, denial coping, and
spiritual peace were measured in a cross-sectional survey.

Path analysis results supported the proposed mediation model, with satisfactory fit indices ( 2
(5) = 6.79, p = .24; CFI = .99, TLI = .99, RMSEA = .045). The masculinity strength norm was
associated with lower caregiving burden, higher self-efficacy, less denial coping; whereas the
avoid being feminine norm was associated with higher caregiving burden. Lower caregiving
burden was associated with less denial coping and higher self-efficacy, which in turn
associated with higher spiritual peace. The model explained 43.2% of the variance in spiritual
peace.

Bootstrapping results indicated that both the masculinity strength (= 0.39, 95% CI =0.34,
0.45) and avoid being feminine (= -0.11, 95% CI =-0.16, -0.07) norms had significant
indirect effects on spiritual peace, supporting that stress/coping variables (caregiving burden,
self-efficacy, and denial coping) mediated between the gender role norms and spiritual peace.

This was the first study exploring how male gender role norms were associated with
stress/coping processes and spiritual peace among husbands of Chinese BCS. Interventions
that help caregivers reduce caregiving burden, enhance self-efficacy, and encourage the use
of adaptive coping strategies may increase their spiritual peace. Regarding gender role norms,
supporting men by appreciating their commitment to take good care of their spouses
(masculinity strength norm) and by reducing their barriers to express tenderness (avoid
being feminine norm) may help to reduce caregiving burden and improve their well-being.
S974 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Nelson Yeung, Ph.D., The Chinese University of Hong Kong, Hong
Kong, N/A, 000000; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S975

B044 6:00 PM-7:00 PM

HEALTH BEHAVIORS AND SUPPORTIVE CARE NEEDS OF LOW-INCOME LATINA BREAST CANCER
SURVIVORS

Alix Sleight, OTD, OTR/L

University of Southern California, Los Angeles, CA

Background: Latina breast cancer survivors are disproportionately at risk for poor health
outcomes. Supportive care services can promote well-being during cancer survivorship and
reduce health disparities. However, little is known about the supportive care needs of low-
income Latina breast cancer survivors, particularly as they relate to health behaviors. This
pilot study aimed to identify the unique supportive care needs and health behaviors of low-
income Latina breast cancer survivors. Design: This descriptive study utilized a battery of
three self-report questionnaires. Setting: Subjects were recruited from the oncology clinic of a
major public hospital. Participants were breast cancer survivors (n=102) with no metastasis
who had completed surgery, chemotherapy, and radiation. Outcome Measures: Three self-
report questionnaires: 1) The Supportive Care Needs Survey (SCNS-SF34), a standardized, 34-
item survey designed to capture the supportive care needs of cancer survivors. 2) The Health
Behaviors and Logistical Challenges Survey, a 31-item measure developed for the purposes of
this study to gather supplemental data about lifestyle-related needs. 3) A standard
demographics questionnaire. Results: All participants self-identified as Latina, with an average
age of 54. 77% disclosed an annual income of $15,000 or less. 72% of participants indicated
having at least one co-morbidity, most commonly hypertension (38% of participants) and
diabetes (31%). Respondents reported, on average, 16 unmet supportive care needs. They
indicated the highest level of unmet need in the domain of health systems and information.
Items most commonly rated as 5/5 (high need) in this domain included "being informed
about things you can do to help yourself to get well" and "being given information about
aspects of managing your illness and side-effects at home. The most frequently reported
needs related to health behaviors and logistical challenges were help with everyday
management of another health condition (45% of respondents) and paying my bills/rent
(48% of participants). Other commonly-reported health behavior challenges included
obtaining adequate nighttime sleep and addressing weight loss/maintenance. Conclusions:
These findings suggest that low-income Latina breast cancer survivors may require more
and/or better-tailored health-related information and increased support in self-management
of co-morbidities, along with guidance in health behavior change. These findings carry
implications for optimizing multidisciplinary supportive care for this population.
S976 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Alix Sleight, OTD, OTR/L, University of Southern California, Los
Angeles, CA, 90065; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S977

B045 6:00 PM-7:00 PM

HEALTHCARE USE AND COST AMONG DEPRESSED AND NON-DEPRESSED CANCER PATIENTS

Gabrielle Decastro, N/A1, Brent Mausbach, PhD2

1
San Diego State University, La Jolla, CA; 2University of California San Diego, La Jolla, CA

OBJECTIVE: We examined differences in healthcare costs between depressed and non-

depressed cancer patients.

BACKGROUND: Cancer is a disease for which treatment may be both lengthy and costly. The
prevalence of depression is significantly higher among individuals with cancer relative to those
without. In non-cancer patients, depression has been associated with increased annual
healthcare use and cost. Among those with cancer, depression has been associated with
increase in annual healthcare visits, increased risk for emergency department visits, and
inpatient hospitalizations. However, the link between depression and healthcare costs in
those with cancer is less clear. This study investigated whether annual healthcare costs were
higher among depressed cancer patients relative to non-depressed cancer patients.

METHODS: A retrospective analysis of UCSD healthcare system administrative data was

conducted on 13,194 patients who had an ICD-9 diagnosis of cancer. Of these, 1,624 (12.3%)
had an ICD-9 diagnosis consistent with a depressive disorder. GLM with gamma distribution
and log link was used to estimate annual healthcare costs for depressed vs non-depressed
participants after controlling for participant age, gender, race/ethnicity, comorbid medical
conditions, metastasis status, and number of mental health visits. Secondary analyses
examined healthcare charges for outpatient visits in depressed and non-depressed
participants.

RESULTS: Depressed cancer patients had annual healthcare charges (M SE = $15,761

$497) that were 61.1% higher than non-depressed cancer patients (M SE = $9,690 $112).
Similarly, outpatient charges for depressed patients were 57.4% higher than non-depressed
patients (mean SE = $3,307 $78 vs $1,915 $18, respectively). When mental health
charges were excluded from total outpatient charges, costs remained almost identical to
those found when mental health costs were included (depressed = $3,102 $73; non-
depressed = 1904 $18), suggesting that increased cost of depression was not due to costs
associated with mental health treatment.

CONCLUSIONS: Depressed cancer patients had higher average annual healthcare costs, as
well as higher outpatient charges, than non-depressed cancer patients. These data suggest
S978 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

that depression comorbidity be taken into account when assessing healthcare and treatment
options for those with cancer. Future studies should replicate results using standardized
diagnostic criteria for depression, determine the impact of severity of depressive symptoms
on healthcare costs, and evaluate if change in depression status via psychiatric treatment
reduces healthcare costs in depressed cancer patients.

CORRESPONDING AUTHOR: Gabrielle Decastro, N/A, San Diego State University, La Jolla, CA,
92093; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S979

B046 6:00 PM-7:00 PM

HIGH PREVALENCE OF COMBINATION TANNING AMONG UNDERGRADUATES: SURVEY OF A

SOUTHEASTERN U.S. UNIVERSITY

Marcus Tan, MBBS1, Natalie R. Gassman, PhD1, Alyssa Fernandez, MPH2, Sejong Bae, Ph.D.3,
Casey Daniel, PhD, MPH1

1
University of South Alabama Mitchell Cancer Institute, Mobile, AL; 2USA Mitchell Cancer
Institute, Mobile, AL; 3University of Alabama at Birmingham, Birmingham, AL

Background: Tanning is a common behavior among young adults. While outdoor tanning (OT)
is common, indoor tanning (IT) and sunless tanning products (STP) are also utilized. However,
little is known about how young adults combine multiple tanning modalities. The objective of
this study was to define the prevalence of combination tanning among undergraduates and to
identify characteristics distinguishing combination tanners from single-modality
tanners. Methods: Undergraduate students at a public university in the southeastern United
States were invited to complete an online survey about their participation, knowledge, and
attitudes regarding OT, IT and STP. Combination tanning was defined as engaging in two or
more tanning behaviors. Uni- and multivariate analyses were performed to identify factors
associated with combination vs. single-modality tanning. Results: 2,668 of 11,015
undergraduates completed the survey (24% response rate). Of these, 1,942 (73%) engaged in
tanning behavior(s). There were 726 single modality tanners (38%) and 1,216 combination
tanners (62%). Among combination tanners, 772 used two tanning modalities (63%), most
commonly OT and IT, whereas 444 used three tanning modalities (OT+IT+STP,
37%). Combination tanners, compared to single-modality tanners, were more likely to be
aware of the risks of tanning beds (OR 1.6, 95% CI 1.1-2.4), but also more likely to agree or
strongly agree that using a tanning bed is fine if not used too frequently (OR 2.5 (1.7-3.6), 6.4
(3.0-13.9), respectively). Furthermore, combination tanners had greater intention to use IT
within the next 12 months (OR 23.7, 95% CI 11.9-47.3). Conclusions: The majority of those
engaged in tanning behaviors used more than one modality. Combination tanning appears to
be an under-recognized phenomenon, with the beliefs and perceptions of this population not
well-studied. Understanding concurrent tanning behaviors may be important in order to
develop effective interventions.

CORRESPONDING AUTHOR: Marcus Tan, MBBS, University of South Alabama Mitchell Cancer
Institute, Mobile, AL, 36604; [email protected]
S980 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B047 6:00 PM-7:00 PM

HOW ARE MULTIFACTORIAL BELIEFS ABOUT CANCER CAUSATION ASSOCIATED WITH RISK
PERCEPTIONS, CANCER COGNITIONS, AND WORRY IN THE US?

Jada G. Hamilton, Ph.D., M.P.H.1, Erika A. Waters, Ph.D., M.P.H.2

1
Memorial Sloan Kettering Cancer Center, New York, NY; 2Washington University in St. Louis,
St. Louis, MO

Background: Cancer, like many common health conditions, is caused by a combination of

genetic, behavioral, and environmental factors. Understanding this multifactorial nature of
cancer etiology is a critical component of genomic health literacy. Our past work has
demonstrated that multifactorial beliefs about cancer causation are common in the U.S.
(endorsed by 64.3%), and that endorsement of multifactorial beliefs is associated with some
cancer screening behaviors. However, it is unclear whether these multifactorial causal beliefs
are associated with other cognitive and affective variables that theory and research have
identified as contributors to the adoption of health behaviors.

Methods: Using nationally representative cross-sectional data from the U.S. Health
Information National Trends Survey (N=2,756), we examined whether endorsing a
multifactorial model of cancer causation (operationalized as simultaneous endorsement of
genetic and behavioral causal beliefs) was associated with perceptions of risk, cancer
cognitions, and worry about toxic environmental exposures.

Results: Multivariable linear regression analyses adjusted for demographics and health history
indicated that multifactorial beliefs were associated with higher absolute (b = 0.18, p = 0.01)
and affective perceived cancer risk (b = 0.10, p = 0.05), more worry about ingested (b = 0.21, p
= 0.002) and emanating (b = 0.17, p = 0.01) environmental exposures, and with more
agreement that cancer can be prevented (b = 0.24, p < 0.001). Multifactorial beliefs were not
associated with comparative perceived cancer risk or with believing that there are too many
cancer recommendations to follow and that everything causes cancer.

Conclusions: Multifactorial causal beliefs are associated with cognitive and affective factors
that drive healthy behaviors. Specifically, multifactorial beliefs about cancer etiology are
related to personal understanding of disease risk, the possibility of preventing cancer, and
worries about toxic environmental exposures. These findings highlight possible mechanisms
by which multifactorial beliefs may support healthy behaviors. They can also inform the
development of health communication efforts that aim to translate basic genomics research
into clinical and public health practice.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S981

CORRESPONDING AUTHOR: Jada G. Hamilton, Ph.D., M.P.H., Memorial Sloan Kettering Cancer
Center, New York, NY, 10022; [email protected]
S982 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B048 6:00 PM-7:00 PM

IMPACT OF A MULTIMEDIA INTERVENTION ON TREATMENT DECISION MAKING AMONG

PROSTATE CANCER PATIENTS: A NATIONWIDE TRIAL

Michael A. Diefenbach, Ph.D.1, Catherine Benedict, PhD2, Suzanne M. Miller, PhD3, Linda
Fleisher, PhD, MPH4, Annette L. Stanton, Ph.D.5, Sara Fleszar, N/A1, Gianna M. Torre, BA1,
Mary Ropka, Ph.D.6, Kuang-Yi Wen, PhD7

1
Northwell Health, Manhasset, NY; 2Hofstra Northwell School of Medicine, Manhassat, NY;
3
Fox Chase Cancer Center/Temple University Health System, Philadelphia, PA; 4Children's
Hospital of Philadelphia, Philadelphia, PA; 5UCLA, Los Angeles, CA; 6UVA School of Medicine,
Charlottesville, VA; 7Fox Chase Cancer Center, Cheltenham, PA

Introduction: Men diagnosed with early stage prostate cancer (PC) face multiple treatment
options, each with distinctive side effects that have significant implications for post-treatment
quality of life. Healing Choices for Men with Prostate Cancer is a multimedia educational
and decision aid program designed to support patients personalized treatment decision
making. This nation-wide randomized controlled trial evaluated the impact of the Healing
Choices program on decision variables.

Methods: Eligible cancer patients who called the National Cancer Institutes Cancer
Information Service (CIS) were invited to participate. Participants completed a baseline
interview and were randomized to either a usual service call with a CIS information specialist
(comparison condition), or CIS standard consultation plus the Healing Choices program
(intervention condition). Follow-up interviews were conducted at 2-months post-enrollment.
Decisional conflict and psychological distress about PC treatment were assessed using the
Decision Conflict Scale and the Impact of Event Scale respectively. Group differences at 2
months post-randomization were tested using the intent-to-treat principle. Hypothesized
moderation effects of demographic and clinical characteristics were evaluated.

Results: 440 participants were randomized to the intervention (n=221) or comparison (n=219)
conditions. The sample was on average 64 years old, primarily White and well-educated.
Intervention effects were not significant on total decisional conflict (t[312]=1.13, ns) or
psychological distress (t[338]=-.48, ns). However, evaluation of specific aspects of decision
processes indicated a significant effect on levels of perceived decisional support (t[339]=2.01,
p < .05. The intervention effect was greates for non-White minority participants (b=10.19,
SE=4.36) and those with lower educational attainment (b-3.87, SE=2.21).
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S983

Conclusions: This interactive, comprehensive education and decision aid program may be
most effective for a subset of patients at risk for unmet needs and decisional conflicts. It is
therefore critical to be able to identify at-risk patients in need of decision support services.
Future work should determine the best ways to leverage this technology to provide tailored
interventions with the greatest reach.

CORRESPONDING AUTHOR: Michael A. Diefenbach, Ph.D., Northwell Health, Manhasset, NY,

11030; [email protected]
S984 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B049 6:00 PM-7:00 PM

IMPACT OF DIAGNOSIS ON HEALTH BEHAVIORS OF BREAST CANCER SURVIVORS AND THEIR

PARTNERS: A QUALITATIVE STUDY

Caroline Dorfman, Ph.D.1, Tamara Somers, Ph.D.1, Rebecca Shelby, Ph.D.1, Gretchen Kimmick,
MD, MS1, Linda Craighead, PhD2, Francis Keefe, Ph.D.1

1
Duke University Medical Center, Durham, NC; 2Emory University, Atlanta, GA

Obese breast cancer survivors are at increased risk of disease progression, recurrence, and
death. The stress and responsibility faced by caregiving partners may also result in weight
gain, unhealthy eating, and inactivity putting them at risk for poor health outcomes. Given the
strong concordance of health practices of couples, we conducted a qualitative study of breast
cancer survivors and their partners to better understand weight-related behaviors among
dyads. Focus groups were held with obese breast cancer survivors (n=14) and their
overweight/obese partners (n=14). Participants provided information about changes in weight
and factors impacting health behaviors following diagnosis and treatment. Field notes and
audio recordings were reviewed; relevant themes were identified. Couples were married
(93%) and had been together for M=27 years (SD=14.6). The average BMI was 33.1 (SD=2.7)
for survivors and 30.6 (SD=3.5) for partners. The majority of survivors were diagnosed with 2
other medical comorbidities, and 57% of partners reported 2 or more health conditions;
hypertension and pain-related conditions (e.g., osteoarthritis) were most common. Weight
gain following diagnosis was common, and weight loss was difficult. Couples endorsed fatigue,
change in taste making high fat/sugar foods more palatable, and increased eating outside the
home as challenges to weight management. Survivors and partners reported a link between
emotions and eating behaviors, describing food as a source of comfort. Couples shared in the
experience of eating for comfort with one partner describing him and his wife as partners in
crime. Couples provided information about barriers to activity. Survivors noted that their
physical activity was limited by joint stiffness, pain, and fatigue. For some partners, side
effects from medical conditions affected their activity level. Behavioral weight loss
interventions for cancer survivors have not typically involved partners and rarely address
physical and psychological symptoms related to poor eating and inactivity. Qualitative data
suggest that a dyadic behavioral weight loss intervention that includes symptom management
strategies may improve weight management and health behaviors for couples during cancer
survivorship.

CORRESPONDING AUTHOR: Caroline Dorfman, Ph.D., Duke University Medical Center,

Durham, NC, 27705; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S985

B050 6:00 PM-7:00 PM

IMPACT OF HEALTH LITERACY AND PATIENT ACTIVATION ON USE OF PROSTATE CANCER AND
COPING WEBSITE INTERVENTION

Alicja Bator, MPH1, Erin K. Tagai, PhD, MPH2, Suzanne M. Miller, PhD3, Michael A. Diefenbach,
Ph.D.4, Sara Fleszar, N/A4, Stefanie Garner, B.S.5, Gianna M. Torre, BA4, Alexander Adia, N/A6,
Alexander Kutikov, MD2, David Y.T. Chen, MD2, Curtis Miyamoto, MD7, Adam Reese, MD8,
Shawna Hudson, PhD9

1
Rutgers University - Robert Wood Johnson Medical School, New Brunswick, NJ; 2Fox Chase
Cancer Center, Philadelphia, PA; 3Fox Chase Cancer Center/Temple University Health System,
Philadelphia, PA; 4Northwell Health, Manhasset, NY; 5Fox Chase Cancer Center/ Temple
University Health Center, Philadelphia, PA; 6Northwell Health, Providence, RI; 7Fox Chase
Cancer Center at Temple University Hospital, Philadelphia, PA; 8Lewis Katz School of
Medicine/Temple University Hospital, Philadelphia, PA; 9Rutgers Robert Wood Johnson
Medical School, New Brunswick, NJ, New Brunswick, NJ

Background: Patient activation and health literacy are important indicators of self-
management and health seeking behaviors. We examined their impact on patient use of an
interactive web-based multimedia program (PROGRESS) that is theoretically guided by the
Cognitive-Social Health Information Processing Model (C-SHIP) to help prostate cancer
patients with self-management after treatment.

Methods: The PROGRESS website intervention for prostate cancer patients focused on
common concerns (e.g., communication, emotional and body-image issues, managing
treatment side effects) and provided training tools and tips to promote coping and reduce
disruption of daily activities. Prostate cancer patients (n=186) within six months of treatment
were given access to the PROGRESS website. Health literacy was measured using a three-item
health literacy screening tool designed by Chew et al. Patient activation was measured using
the patient activation measure (PAM-13) scale. Descriptive statistics were used to summarize
website usage, health literacy, patient activation and demographics such as race, age,
household income, marital status, and education. Chi-square and ANOVA were used to
determine bivariate associations. Logistic regression was used to analyze multiple predictors
of website use.
S986 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Results: High health literacy levels (mean=4.36, SD=0.76) and high patient activation
(55%=high compared with 6%=low) were observed. Bivariate analyses reported patients with
higher levels of health literacy were more likely to access and use the website (r=0.22, p

Discussion: Data suggests that patients with high literacy are more likely to find the website
intervention useful. Findings also suggest participants with greater health literacy scores have
greater patient activation. Future research is needed to better understand the relationships
between health literacy, patient activation, and other predictors of use of interactive
educational websites tools for prostate cancer and other patient populations.

CORRESPONDING AUTHOR: Alicja Bator, MPH, Rutgers University - Robert Wood Johnson
Medical School, New Brunswick, NJ, 08901; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S987

B051 6:00 PM-7:00 PM

INTRA- AND INTERPERSONAL FACTORS THAT MAY INFLUENCE GENETIC COUNSELING AMONG
HIGH RISK BREAST CANCER SURVIVORS

Maija Reblin, PhD1, Monica Kasting, PhD1, Kelli Nam, MPH1, Courtney L. Scherr, PhD2, Jongphil
Kim, PhD1, Ram Thapa, PhD1, Cathy Meade, PhD, RN, FAAN1, M. Catherine Lee, MD FACS1,
Tuya Pal, MD1, Gwendolyn Quinn, Ph.D. 3, Susan Vadaparampil, PhD MPH1

1
Moffitt Cancer Center, Tampa, FL; 2Northwestern University, Chicago, IL; 3Moffitt Cancer
Center - USF, Tampa, FL

Purpose: Many survivors at increased risk for hereditary breast cancer (BC) do not receive
genetic counseling (GC) and/or testing. We evaluated intra- and interpersonal factors
predicting survivor readiness to engage in GC.

Methods: BC survivors meeting National Comprehensive Cancer Network criteria for referral
to GC were recruited using clinic and community based approaches. After consent,
participants responded to items based on the Health Belief Model assessing intrapersonal
(perceived risk, susceptibility, severity, worry and impact of BC, benefits and barriers to GC,
self-efficacy, and decisional conflict), interpersonal-provider (provider encouragement of GC),
and interpersonal-family (cancer family history, family encouragement of testing) factors and
demographic items. Participants also reported whether they planned to make a GC
appointment, and drawing from the Transtheoretical Model of Behavior Change, were
categorized by pre-contemplation (did not plan to make an appointment) and contemplation
(did plan to make an appointment in the next 1-6 months) status. We used X2 and t-tests to
compare pre-contemplators and contemplators on intra- and interpersonal factors.

Results: Of 111 participants analyzed, 48 (43%) were contemplators and 63 (67%) were pre-
contemplators. Compared to pre-contemplators, contemplators were younger (M=61 vs 65,
p=.04), but there were no other significant demographic differences. Contemplators reported
higher perceived susceptibility and severity, and more worry and intrusiveness related to BC
(ps

Conclusions: Our results suggest that intrapersonal factors may be more important than
interpersonal factors in predicting whether at-risk BC survivors seek GC. However, families of
pre-contemplative survivors may recognize their reticence and offer more explicit
encouragement for testing. Findings suggest potentially modifiable targets to improve rates of
GC in high risk BC survivors.
S988 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Maija Reblin, PhD, Moffitt Cancer Center, Tampa, FL, 33612;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S989

B052 6:00 PM-7:00 PM

PREDICTORS OF ADHERENCE TO AEROBIC EXERCISE IN RECTAL CANCER PATIENTS DURING

AND AFTER NEOADJUVANT CHEMORADIOTHERAPY

Andria Morielli, MSc1, Normand Boul, PhD1, Nawaid Usmani, MD2, Kurian Joseph, MD2, Keith
Tankel, MD2, Diane Severin, MD2, Kerry S. Courneya, PhD1

1
University of Alberta, Edmonton, AB, Canada; 2Cross Cancer Institute, Edmonton, AB, Canada

Background: Treatment for locally advanced rectal cancer includes 5-6 weeks of neoadjuvant
chemoradiotherapy (NACRT) followed by surgery 6-8 weeks later. Unfortunately, NACRT is
associated with side effects including fatigue, diarrhea, and physical deconditioning.
Preliminary evidence suggests that aerobic exercise may improve cardiorespiratory fitness in
rectal cancer patients during and after neoadjuvant chemoradiotherapy (NACRT), however,
adherence to exercise is a challenge in this clinical setting. Objective: To explore the
predictors of adherence to supervised aerobic exercise during NACRT and unsupervised
aerobic exercise after NACRT in rectal cancer patients. Methods: Rectal cancer patients
(N=18) scheduled to receive NACRT followed by definitive surgery were recruited from the
Cross Cancer Institute in Edmonton, Alberta. The exercise program consisted of 3 supervised
moderate-intensity aerobic exercise sessions/week during NACRT followed by 150
minutes/week of unsupervised moderate-intensity aerobic exercise after NACRT.
Demographic, health, psychosocial, and motivational variables were assessed pre- and post-
NACRT.. Results: For the patients included in the current analysis, adherence to the
supervised exercise sessions during NACRT was 77% (SD = 19%, n = 17) and the mean number
of aerobic exercise minutes per week post-NACRT was 160 (SD = 67 minutes, n = 13). Although
not statistically significant, adherence to supervised exercise during NACRT was meaningfully
better for patients who were women (d= 0.82; p=0.12), younger (d=-0.62; p=0.30), married
(d=0.62; p=0.42), with better mental health (r=0.32; p=0.21), fewer diarrhea symptoms
(r=0.48; p=0.052), and higher anticipated enjoyment (r=0.31; p=0.23), support (r=0.32;
p=0.22), and motivation (r=0.31; p=0.23). After NACRT, adherence to unsupervised exercise
was significantly better for patients who reported worse mental health (r= -0.56, p=0.046) and
meaningfully better for patients who were women (d= 0.54; p=0.38), better educated (d=
0.77; p=0.22), had no comorbidities (d=-0.63; p=0.17), and were aerobic exercisers at baseline
(d=1.05; p=0.12). Conclusions: This pilot study suggests that demographics, tumor side
effects, and motivational variables may predict adherence to supervised exercise during
NACRT. Predictors of adherence to unsupervised exercise after NACRT may differ and be
influenced by treatment response.
S990 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Andria Morielli, MSc, University of Alberta, Edmonton, AB, T6G
2H9; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S991

B053 6:00 PM-7:00 PM

SMOKERS ATTITUDES TOWARDS CONTINUED SMOKING AFTER A NEGATIVE CT RESULT &

LATER TOBACCO USE IN A LUNG CANCER SCREENING POPULATION

Sarah Lillie, PhD, MPH1, Anne Melzer, MD2, Angela Fabbrini, MPH2, Kathryn Rice, MD2,
Barbara Clothier, MA1, Melissa Partin, PhD1, Steven Fu, MD1

1
Minneapolis VA Center for Chronic Disease Outcomes Research, Minneapolis, MN;
2
Minneapolis VA Division of Pulmonology, Minneapolis, MN

Background: Low-dose computed tomography (LDCT) for lung cancer screening decreases
lung cancer mortality and offers a teachable moment to deliver smoking cessation
treatment. However, there is concern that patients given negative results may believe they
are no longer at risk of cancer and will continue to smoke - the health certificate effect.
Methods: We surveyed Veterans 3 mo. after screening invitation as part of a larger lung
cancer screening evaluation. Survey respondents were asked how strongly they agreed or
disagreed with the statement If I had a normal CT scan result, I could continue to smoke
without worry. We dichotomized responses strongly disagree/disagree (i.e., worried about
continued smoking) vs neutral/agree/strongly agree (i.e., not worried about continued
smoking). Current smokers at invitation (45% of original sample) completing the item were
included in analyses (N=263).We assessed smoking status 6 mo. later (N=217) and collected
LDCT completion and results (negative result; abnormal result) from medical records. We
evaluated bivariate associations between the reported worry of continued smoking post-
negative LDCT, sociodemographics, and subsequent smoking status.
Results: Respondents were mostly male (94%) and white (87%). Overall, 46% were not
worried about continued smoking after a negative LDCT result. Not worrying about continued
smoking varied by age group, with older smokers (age 65) reporting this view more often
than younger smokers (56% vs 37%, X2=9.16; P < 0.01). Compared to those who worried
about smoking post-negative LDCT, a higher proportion of respondents who were not worried
about continued smoking later reported smoking everyday, and a lower proportion smoked
some days or not at all (X2=7.57; P < 0.05). Forty-eight Veterans completed LDCT; 13
received normal and 35 abnormal results, of which 2 required further diagnostic evaluation.
Though the sample was too small for formal analyses, 100% of Veterans who were not
worried about continued smoking on the survey and actually received a negative LDCT result
later reported smoking daily at follow up.
Discussion: Nearly half of respondents reported that they would not be worried about
continued smoking after a negative LDCT result, and this attitude was associated with actual
tobacco use. This suggests that the health certificate effect is present in lung cancer screening,
S992 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

particularly among older smokers. Successful lung cancer screening implementation will
require evidence-based smoking cessation treatment and addressing patients views towards
continued smoking post-negative LDCT.

CORRESPONDING AUTHOR: Sarah Lillie, PhD, MPH, Minneapolis VA Center for Chronic Disease
Outcomes Research, Minneapolis, MN, 55417; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S993

B054 6:00 PM-7:00 PM

SYMPTOMS AND QUALITY OF LIFE IN PATIENTS WITH DIVERSE MALIGNANCIES PRESCRIBED

ORAL CHEMOTHERAPY

Jamie M. Jacobs, PhD1, Molly E. Ream, B.A.1, Charn-Xin C. Fuh, B.S.1, Nicole A. Pensak, PhD2,
Joel N. Fishbein, B.A.3, Lauren E. Nisotel, B.A.4, James MacDonald, B.A.5, J Buzaglo, PhD6, Inga
T. Lennes, M.D1, Steven A. Safren, PhD7, William F. Pirl, M.D., MPH8, Jennifer S. Temel, MD1,
Joseph A. Greer, PhD1

1
Massachusetts General Hospital Cancer Center/Harvard Medical School, Boston, MA;
2
University of Colorado Anschutz Medical Campus, Aurora, CO; 3University of Colorado,
Boulder, Department of Psychology, Boulder, CO; 4Harvard T.H. Chan School of Public Health,
Boston, MA; 5University of California, Los Angeles, Department of Psychology, Los Angeles,
CA; 6Cancer Support Community, Washington, DC; 7Department of Psychology, University of
Miami, Coral Gables, FL; 8Sylvester Cancer Center, University of Miami Miller School of
Medicine, Miami, FL

Background: Oral chemotherapies are increasingly utilized in oncology care. However,

patients on these regimens lack close monitoring with regard to adherence, side effects, and
toxicities. Data are lacking regarding the types of physical and psychological symptoms
patients experience. Little is known about patients quality of life (QOL) and factors associated
with QOL during treatment.

Methods: From 12/20148/2016, adult patients with solid and hematologic malignancies
prescribed oral chemotherapy were enrolled in a randomized controlled study of adherence
and symptom management at the Massachusetts General Hospital Cancer Center. Patients
reported sociodemographic information and completed baseline assessments of QOL (FACT-
G), symptom severity (MDASI), depression and anxiety (HADS), fatigue (BFI), satisfaction with
social support (MSPSS), and satisfaction with providers and treatment (FACIT-TS-PS).
Prevalence of symptoms and relationships among these constructs were examined using
measures of central tendency, proportions, Pearson product-moment correlations, and
multivariate linear regression.

Results: Patients (N=174) commonly reported fatigue (88.7%), pain (44.6%), and nausea
(37.6%). Higher QOL (M=81.66, SD=15.13) was associated with lower depressive and anxiety
symptoms, lower fatigue, better social support, and higher satisfaction with providers and
treatment (ps < .001). Having children was associated with lower QOL (r=-0.19, p=.013). In a
S994 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

multivariate regression model, each factor was associated with QOL above the effects of
others; (F[6,157]=67.13, p < .001, adj. R2=.71).

Conclusions: Findings elucidate common symptoms and correlates of QOL in patients

prescribed oral chemotherapy. Data suggest the potential value of addressing symptoms and
psychosocial targets for interventions to enhance patients QOL.

CORRESPONDING AUTHOR: Jamie M. Jacobs, PhD, Massachusetts General Hospital Cancer

Center/Harvard Medical School, Boston, MA, 02114; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S995

B055 6:00 PM-7:00 PM

TRAJECTORIES OF CANCER-SPECIFIC DISTRESS OF INDIVIDUALS WITH PROSTATE CANCER AND

THEIR PARTNERS

Kevin R. Criswell, PhD1, Sharon L. Manne, PhD2

1
Rutgers Cancer Institute of New Jersey, Edison, NJ; 2Rutgers Cancer Institute of New Jersey,
New Brunswick, NJ

Introduction: The prostate cancer experience affects both prostate cancer (PCA) survivors and
their partners. Although existing observational longitudinal studies of cancer-specific distress
differ regarding the trajectories of distress they report, there have been no studies to date
that have reported whether different trajectories of distress may emerge from the same
sample of PCA survivors and their partners. The current study sought to characterize cancer-
specific distress trajectories for PCA survivors and their partners and to determine predictors
of trajectories.

Method: The current study analyzed data spanning 26 weeks across four time points from
PCA survivors (n = 79) and their partners (n = 79) within the usual care arm of a clinical trial for
an intimacy-enhancing intervention (NIH CA140297). Growth mixture modeling was utilized so
as to extract trajectory classes of cancer-specific distress for survivors and their partners,
separately. Logistic regression models were run to determine baseline univariate and
multivariate predictors of class membership.

Results: Two-class linear models of cancer-specific distress emerged for both PCA survivors
(high-high, low-low) and their partners (high-low, low-low). The high-high and low-low classes
began at high and low levels, respectively, and remained at those levels. The high-low class
began at a high level that decreased over time. Univariate predictors included age, bowel
function, psychological distress, psychological wellbeing, intimacy, concern, and masculine
self-esteem for survivors (ps < .05). Univariate predictors included mutual avoidance, holding
back concern, concern, general psychological distress, and psychological wellbeing for
partners (ps < .05). Significant multivariate predictors of class membership included bowel
function and general psychological distress for survivors and concern and general
psychological distress for partners (ps < .05). Multivariate models correctly classified 83.3%
and 78.7% of survivors and partners, respectively.

Discussion: The data suggest that discreet trajectories of cancer-specific distress exist for both
PCA survivors and their partners. Future studies should demonstrate whether the extracted
trajectories and their predictors are generalizable to other samples. Healthcare professionals
S996 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

should consider cancer-specific distress screening for PCA survivors with poor bowel function
as they may benefit from psychopharmacological or psychosocial interventions.

CORRESPONDING AUTHOR: Kevin R. Criswell, PhD, Rutgers Cancer Institute of New Jersey,
Edison, NJ, 08837; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S997

B056 6:00 PM-7:00 PM

ANXIETY SENSITIVITY IN SMOKERS WITH INDICATORS OF CARDIOVASCULAR DISEASE

Samantha Farris, Ph.D.1, Ana Abrantes, Ph.D.2

1
Alpert Medical School of Brown University, Providence, RI; 2Butler Hospital/Alpert Medical
School of Brown University, Providence, RI

There is growing recognition of the importance of understanding the nature of the

associations between anxiety and cardiovascular disease (CVD), although limited research has
examined mechanisms that may explain the anxiety-CVD link. Anxiety sensitivity (fear of
anxiety-relevant somatic sensations) is a cognitive-affective risk factor implicated in the
development of anxiety psychopathology and various behavioral risk factors for CVD, although
has not been examined among individuals with CVD. Adult daily smokers (n = 619; 50.9%
female; Mage = 44.0, SD = 13.67) completed an online survey that included the Anxiety
Sensitivity Index-3 (ASI-3) and the Patient Health Questionnaire (PHQ). The presence of CVD
indicators was assessed via the presence of 1 of the follow: heart attack, heart murmur,
positive stress test, heart valve abnormality, angina, and heart failure. Smokers with CVD
indicators (n = 66, 10.7%) had significantly higher scores on the ASI-3 (M = 33.5, SD = 22.15),
relative to smokers without CVD (M = 22.0, SD = 17.92; Cohen s d = .57). Physical and social
concerns about the meaning of somatic sensations were common among smokers with CVD.
Mediational analyses indicated that anxiety sensitivity indirectly accounted for the association
between (a) anxiety psychopathology and likelihood of CVD, and (b) CVD and likelihood of
anxiety psychopathology. While these findings are preliminary and warrant replication in
future prospective studies, anxiety sensitivity is a potentially novel psychological mechanism
that plays a facilitatory role in bidirectional associations between anxiety psychopathology
and CVD.

CORRESPONDING AUTHOR: Samantha Farris, Ph.D., Alpert Medical School of Brown

University, Providence, RI, 02906; [email protected]
S998 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B057 6:00 PM-7:00 PM

HOW MANY US CITIZENS ARE AT RISK FOR HEART DISEASE? FINDINGS FROM THE NATIONAL
HEALTH AND NUTRITION EXAMINATION SURVEY, 1999-2014

Haley Moss, MPH Candidate1, Kaigang Li, PhD2

1
Colorado School of Public Health, Fort Collins, CO; 2Colorado State University, Fort Collins, CO

Background: Cardiovascular Disease (CVD) is the leading cause of death worldwide, costs
billions of dollars in healthcare spending each year, and affects people of all ethnicities, ages,
and geographic locations. With the increasing age of the U.S. population and the obesity
epidemic, CVD related morbidity and mortality is likely to increase in prevalence in the future.
Early prediction of individual risk may aid in reversing this trend.
Purpose: Assess the 10-year atherosclerotic CVD (ASCVD) risk of the U.S. population aged 40-
79 years utilizing the newly developed, race-and sex-specific Pooled Cohort Equations for non-
Hispanic White and African-American men and women from the American College of
Cardiology (ACC) and the American Heart Association (AHA).
Methods: Using data from the 1999-2014 National Health and Nutritional Examination
Survey, 10-year ASCVD risk scores were calculated using stepwise addition of the provided
equation parameters for 11,553 participants. After calculating the total, the percent risk was
calculated using 1-(baseline survival rate)^e^(total-overall mean). Both baseline survival rate
and overall mean were provided by the ACC/AHA. We then sorted scores into categories of
less than 5%, 5-10%, and greater than 10% risk. The analyses were conducted using SAS 9.4
program and accounted for complex survey design.
Results: Of all participants, 57.66% (weighted % hereafter, 95% confidence interval [CI]: 55.89
59.43%) had a 10-year ASCVD risk of 5%. In addition, 20.29% (CI: 18.92-21.66%) had a risk
of 5 - 10%, and 22.06% (CI: 20.99-23.12%) had a risk of < 10%. Sex and age disparities were
found, with 28.02% males (vs. 15.48% females; Rao-Scott chi-square=1,251.43, df=2, p < .001)
and 41.83% of those 55 years and older (vs. 3.28% 50-55, 1.03% 45-50, 0.33% 40-45 year; Rao-
Scott chi-square = 11,437.59, df=6, p < .001) had a < 10% risk.
Conclusion and Future Directions: Given the high percentage of individuals with a 10-year
ASCVD risk < 10%, steps must be taken to lower individual risk. Previous studies suggest that
providing specific diet and exercise guidelines as well as individual and group counseling
about healthy eating and exercise behaviors in gender and race specific contexts can aid in
creating lasting behavioral changes that lead to lower risk. In addition to the efforts to
improve individual CVD risk lowering behaviors, environmental improvements must be
emphasized to support individuals in making risk-lowering decisions.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S999

CORRESPONDING AUTHOR: Haley Moss, MPH Candidate, Colorado School of Public Health,
Fort Collins, CO, 80526; [email protected]
S1000 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B058 6:00 PM-7:00 PM

LONELINESS, CARDIOVASCULAR DISEASE, AND DIABETES PREVALENCE IN THE HISPANIC

COMMUNITY HEALTH STUDY/STUDY OF LATINOS

Samantha A. Reina, MS1, Tasneem Khambaty, PhD1, Orit Birnbaum-Weitzman, PhD1, William
Arguelles, PhD2, Frank Penedo, PhD3, Rebeca Espinoza Giacinto, MA/MPH, PhD4, Angela P.
Gutierrez, MS5, Linda C. Gallo, PhD6, Aida L. Giachello, PhD7, Neil Schneiderman, Ph.D.8, Maria
M. Llabre, PhD1

1
University of Miami, Coral Gables, FL; 2Center for Research and Grants, Baptist Health South
Florida, Coral Gables, FL; 3Northwestern University, Chicago, IL; 4San Diego State
University/UC San Diego, Auburn, CA; 5SDSU/UC San Diego Joint Doctoral Program in Clinical
Psychology, La Jolla, CA; 6San Diego State University, Chula Vista, CA; 7University of Illinois at
Chicago, Chicago, IL; 8University of Miami, Miami, FL

Background: Cardiovascular, immune, and nervous system changes have been identified in
chronically lonely people. Yet, the relationship between loneliness and prevalence of
cardiovascular disease (CVD) and diabetes mellitus (DM) has been understudied in U.S.
Hispanics/Latinos, a group at high risk for DM. Purpose: We examined whether loneliness was
associated with CVD and DM, and whether sex, marital status, and years in U.S moderated
these associations. Methods: Participants were 5,313 adults ages 18 -74 (M (SD) age = 42.39
(15.01), 54.5% female) enrolled in the Hispanic Community Health Study/Study of Latinos
Sociocultural Ancillary Study. Loneliness was assessed via the 3-item Revised UCLA Loneliness
Scale. CVD was assessed via self-reported medical history and DM from serum glucose levels,
A1C percentage, OGTT findings, and diabetes medication use. Results: Although mean level
of reported loneliness was low, loneliness adjusted for demographic, disease risk factors
(body mass index, smoking) and depression was significantly associated with CVD: OR = 1.10
(CI: 1.01-1.20), p = .024 and DM: OR = 1.08 (CI: 1.00 1.16), p = .048. Each unit increase in the
loneliness total score was associated with 10% greater odds of having CVD and 8% greater
odds of having DM. Results did not vary among groups defined by sex, marital status or years
in the US. Conclusions: Results from this large epidemiological study indicate that increased
loneliness is associated with higher cardiometabolic disease prevalence beyond depressive
symptoms, regardless of sex, marital status, or years living in the U.S. These results raise the
possibility that Hispanic/Latino adults experiencing high levels of loneliness may be a
subgroup at particularly elevated risk of CVD and DM.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1001

CORRESPONDING AUTHOR: Samantha A. Reina, MS, University of Miami, Coral Gables, FL,
33146; [email protected]
S1002 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B059 6:00 PM-7:00 PM

A NEEDS ASSESSMENT OF MIDWESTERN MOTHERS TO BETTER UNDERSTAND MATERNAL

MENTAL AND PHYSICAL HEALTH

Kailey Snyder, M.S., Danae Dinkel, Ph.D., Sofia Jawed-Wessel, Ph.D.

University of Nebraska at Omaha, Omaha, NE

Background: Up to 23% of women struggle with depression during pregnancy and few meet
nutrition and physical activity recommendations during pregnancy and the postpartum
period. Specifically, up to 60% of pregnant women engage in little to no physical activity in a
typical week and these numbers do not improve during the postpartum period. In regards to
nutrition, few women prior to pregnancy are meeting nutrition recommendations and
minimal improvements to nutrition occur during pregnancy. There is a vital need to improve
the mental and physical health of mothers. Purpose: The purpose of this study was to conduct
a needs assessment to assess womens perceptions of their own mental and physical health as
well as their desire for additional services. Methods: This non-experimental cross-sectional
study utilized a 35-question survey completed by pregnant and postpartum women. Data
were analyzed using descriptive statistics. Results: A total of 527 mothers living within a large
Midwestern metropolitan city completed the survey. The majority of mothers were Caucasian
(94.2%), employed full-time (54.6%) and had a household income of $50,000-$75,000 (25.6%).
When asked about causes of stress, the most common causes reported were exhaustion
(53.5%) and financial distress (38.2%). Mothers indicated these components had been often
to very bothersome in the past year. Additionally, 47.3% of mothers felt at least somewhat
bothered by a lack of access to physical activity and healthy food options. When asked if they
felt they were receiving empathy, trust, and love in their life 15.7% reported no. When asked
if they would like more information regarding their own mental and physical health 82.2% of
mothers said yes primarily through the format of online support groups (58.4%) or in-person
workshops/events (48.5%). Conclusion: Mothers need greater access to resources related to
mental and physical health to improve overall health thus improving the well-being of their
family. Future interventions could focus on utilizing free workshops/events that included an
education component as well as online resources related to financial planning, mental health,
and health behaviors (i.e., healthy eating, physical activity).

CORRESPONDING AUTHOR: Kailey Snyder, M.S., University of Nebraska at Omaha, Omaha, NE,
68106; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1003

B060 6:00 PM-7:00 PM

EAT MOVE WIN: PILOT RESULTS OF AN ONLINE NUTRITION EDUCATION PROGRAM FOR HIGH
SCHOOL STUDENTS

Trina Robertson, M.S., R.D.N1, Genevieve Dunton, PhD, MPH2, Jaclyn Maher, PhD3, Janel
Alberts, PhD4

1
Dairy Council of California, Irvine, CA; 2University of Southern California, Institute for Health
Promotion & Disease Prevention, Pasadena, CA; 3Data Analysis Consultant, Pasadena, CA;
4
Research Consultant, Mission Viejo, CA

Background: Childhood overweight and obesity are major health problems. School-based
nutrition education programs can be implemented with relative ease and cost-effectiveness,
and disseminate health information to large groups of adolescents through in class activities.
Eat Move Win delivers information about nutrition and its impact on personal health through
issues relevant to adolescents. The program consists of 5 interactive online lessons that
inform and empower students to take small steps to improve their eating habits. Students are
encouraged to critically reflect on the content, which aligns with the USDA MyPlate, and make
SMART goals for behavior change.
Objective: The goal of this research was to conduct a pilot test of Eat Move Win to assess
preliminary efficacy at changing students eating behaviors.
Methods: Pilot testing of the five lessons took place in ninth and tenth California classrooms
(N = 4) in early Spring 2016. One hundred and sixty students completed a pre-survey before
the curriculum was taught and a post-survey immediately after the five lessons. The survey,
from the California Department of Public Health, utilizes the School Physical Activity and
Nutrition Survey (SPAN) instrument. It measures food consumption, physical activity, and
psychosocial factors influencing fruit and vegetable consumption with high school students.
Results: Pilot data indicated a significant positive change in fruit, carrot, milk, yogurt, whole
grain bread intake from pre-to post survey (ps < .05). Changes in salad and vegetable
consumption were also positive and approached statistical significance (ps < .10). Chi-square
tests indicated that participants did not significantly change their breakfast eating habits
between the pre- and post-survey, although a small increase was observed. Approximately
three-quarters (73.6%) of participants reported consuming breakfast on the pre-assessment,
and 79.2% of participants reported consuming breakfast on the post-survey. Physical activity
(past 7 days) significantly increased from pre-to post survey (p < .05). Moreover, pilot data
indicated small but not statistically significant reductions in the consumption of sports drinks,
soda, French fries/chips, and pastries from pre to the post survey.
Conclusion: Eat Move Win is an online nutrition curriculum that shows promise for increasing
S1004 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

healthy eating habits and physical activity, and decreasing consumption of unhealthy
foods/drinks for high school students.

CORRESPONDING AUTHOR: Trina Robertson, M.S., R.D.N, Dairy Council of California, Irvine,
CA, 92612; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1005

B061 6:00 PM-7:00 PM

EFFECTS OF AEROBIC EXERCISE ON COGNITION AND MENTAL HEALTH SYMPTOMS IN

CHILDREN WITH ATTENTION-DEFICIT HYPERACTIVITY DISORDER

Catherine L. Davis, PhD1, Sarah Premji, B.S. 2, Yoon Ji Ahn, BS3, Celestine Williams, MS2, Jacob
Looney, BS4, Christopher F. Drescher, PhD2, Eduardo E. Bustamante, Ph.D.5

1
Georgia Prevention Institute, Pediatrics, Medical College of Georgia, Augusta Univ., Augusta,
GA; 2Augusta University, Augusta, GA; 3Medical College of Georgia, Augusta University,
Augusta, GA; 4Augusta University, North Augusta, SC; 5University of Illinois at Chicago,
Chicago, IL

PURPOSE: To test the effects of aerobic exercise on cognition and mental health symptoms in
children with Attention-Deficit/Hyperactivity Disorder (ADHD); to bring a new level of rigor to
this field (diagnosis; randomization; attention to CONSORT); and to calculate effect sizes to
inform future trials.

METHODS: 8 children were randomized into exercise (n = 4) or wait list control group (n = 4).
Participants were selected based on diagnosis of ADHD (Parent/Teacher Disruptive Behavior
Disorders (DBD) Rating Scale, the Impairment Rating Scale (IRS), and Diagnostic Interview
Schedule-IV Parent Interview by a child psychologist). The exercise group attended a 4-week,
40-minute daily aerobic exercise program. Independent samples t-tests measured changes
over time by group (Hearts & Flowers, Flanker Fish, DBD, IRS, and the Strengths and
Difficulties Questionnaire (SDQ)). Cohens d assessed effect size.

RESULTS: No significant differences between groups were found. Effect size in Flanker Fish
and each of the symptom ratings (DBD, IRS, SDQ) showed large differences between groups.

CONCLUSIONS: Results did not show significant group effects, but large effect sizes suggest
we might encounter significant benefits of exercise in a larger study. Limitations include a
small sample size, some baseline group differences and the short duration of the intervention.

CORRESPONDING AUTHOR: Catherine L. Davis, PhD, Georgia Prevention Institute, Pediatrics,

Medical College of Georgia, Augusta Univ., Augusta, GA, 30907; [email protected]
S1006 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B062 6:00 PM-7:00 PM

EXPANDING THE LIVING HEALTHY SUMMER CAMP TO LATINO FAMILIES

Brook E. Harmon, PhD, RD, FAND1, Brennan K. Berg, PhD1, Idia B. Thurston, PhD1, Cardella
Leak, MPH1, Claudio R. Nigg, PhD, FSBM2, Laura L. Hayman, PhD, RN, FAAN, FAHA3

1
University of Memphis, Memphis, TN; 2University of Hawaii, Honolulu, HI; 3University of
Massachusetts Boston, Boston, MA

The prevalence of childhood obesity in the US remains high with little recent change. While
parent-focused interventions are recommended for modifying obesity-related behaviors
among children, participation and retention of parents is a major barrier. This study used a
community-based participatory approach to translate a child-focused summer program to
families in Memphis, TN. The Living Healthy Summer Camp began in 2010 as a 4-session
culinary camp offered through a church in South Carolina, which served a primarily low-
resource African-American community. Participants found the camp acceptable as a way of
providing nutrition and cooking skills to families. In translating the camp, the study team met
with representatives from 10 churches with multi-ethnic congregations. These meetings
expanded the camp into a 4 month curriculum that included: weekly mailings of family-
focused diet and physical activity (PA) education activities; a week-long camp offering child-
focused cooking and PA skill building; 2.5 hour sessions offered to parents 4 times and
focused on diet and PA education and skill building. Thirty youth enrolled along with 17
parents. Parents and children completed diet and PA behavior questionnaires and had
anthropometrics measured. Parents also completed home environment questionnaires.
Baseline data are reported, with measurements being repeated mid and post program. All
participants identified as Latino. Children were between 7 and 12 years old (9.4 1.5),
primarily male (70%), and 47% were classified as obese, using age and sex specific growth
charts. Parents were on average 32.25.1 years old, obese (BMI=31.26.5 kg/m2), female
(77%), and received public assistance (65%). Almost half (47%) of children reported eating at a
fast food restaurant 1-2 times a week. They also reported eating on average 3.42.0
servings/day of fruits and vegetables and spending 3.63.0 hours/week in moderate-to-
vigorous physical activity. Most children (53%) reported not being involved in meal
preparation during the past week and their average cooking ability was 3.82.9 on scale of 1
to 10 (highest). Baseline data indicates a population with a high need, and recruitment
indicated a high interest, for this type of intervention.

CORRESPONDING AUTHOR: Brook E. Harmon, PhD, RD, FAND, University of Memphis,

Memphis, TN, 38152; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1007

B063 6:00 PM-7:00 PM

FAMILY STRESS AND PAIN-RELATED DISTRESS IN YOUTH WITH IBD: TARGETS FOR
TREATMENT?

Bonney Reed-Knight, PhD1, Miranda A.L. van Tilburg, PhD2, Shelby L. Langer, PhD3, Joan M.
Romano, PhD4, Rona L. Levy, MSW, PhD, MPH4

1
Emory University, Atlanta, GA; 2University of North Carolina at Chapel Hill, Chapel Hill, NC;
3
Arizona State University, Phoenix, AZ; 4University of Washington, Seattle, WA

PURPOSE: Abdominal pain is common in pediatric inflammatory bowel diseases (IBD) but
physiological causes have not adequately accounted for the pain experience. Pain is a
subjective experience and predicts unfavorable outcomes, but it is unclear who is at risk, and
who experiences typical adjustment in the face of chronic illness. The present study aimed to
examine the role of several factors prior pain research has found to be associated with poor
outcomes - life stressors, coping, and depressive symptoms - in pain-related distress in youth
with IBD.

METHODS: 183 parents and their children ages 8-18 years old (Mean age=13.75; 47.5%
female, 4.3% Hispanic, 88.0% Caucasian) diagnosed with IBD were recruited (68.3% Crohns
Disease and 31.7% Ulcerative Colitis). Parents completed measures on child pain related
distress, family stressors, and child coping, and children completed the Childrens Depression
Inventory.

RESULTS: Mediation analysis was conducted using ordinary least squares path analysis.
Higher family stress was positively associated with passive coping. Passive coping was
positively associated with depressive symptoms (p < .001) and pain-related distress (p < .001).
Depressive symptoms were positively associated with pain-related distress (p < .05). Mediated
effects were demonstrated by significant indirect effects in the relationship between family
stress and pain-related distress through passive coping (point estimate=.03, 95%CI=.01-.04),
depressive symptoms (point estimate=.001, 95%CI= .0003-004) and through both passive
coping and depressive symptoms sequentially (point estimate=.004, 95%CI=.00-.01). The full
model including mediators accounted for 66% of the variance in pain-related distress; family
stress alone account for 27% of the variance in pain-related distress.

CONCLUSIONS: Children whose families have experienced greater family stressors reported
greater passive coping and depressive symptoms, which each independently were associated
S1008 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

with pain-related distress. These findings suggest that addressing coping and depressive
symptoms during treatment may reduce the burden of pain in young IBD patients.

CORRESPONDING AUTHOR: Bonney Reed-Knight, PhD, Emory University, Atlanta, GA, 30322;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1009

B064 6:00 PM-7:00 PM

FOOD CRAVING MEDIATES THE RELATIONSHIP BETWEEN EXTERNAL EATING AND EXCESS
GESTATIONAL WEIGHT GAIN

Lauren Blau, B.S.1, Natalia Orloff, M.S.1, Renee Samuelson, M.D.2, Julia M. Hormes, Ph.D.1

1
University at Albany, SUNY, Albany, NY; 2Albany Medical Center, Albany, NY

Excess gestational weight gain (GWG) is a major public health concern and contributes to
medical complications in both mother and child. Craving frequency in pregnant women
accounts for a considerable portion of the variance in excess GWG (Orloff et al., 2016). Food
craving has also been shown to mediate the effect of external eating (i.e., eating in response
to environmental cues) on body mass index (BMI) in university students (Burton et al., 2007).
This study tested the hypothesis that craving frequency similarly mediates the relationship
between externality and excess GWG.

68 pregnant women (mean age=30.47, SD=4.23; average weeks of gestation=24.01, SD=8.91)

completed the Dutch Eating Behavior Questionnaire, which measures emotional
(Cronbachs a=.95), external (a=.86), and restrained (a=.90) eating styles, and the Food
Craving Inventory, which measures frequency of cravings for high fat foods (a=.70), sweets
(a=.86), carbohydrates/starches (a=.82), and fast food fats (a=.66). Excess GWG was calculated
based on pre-pregnancy BMI and stage of gestation at time of survey completion.

36.9% of participants had gained in excess of the recommended GWG for their stage of
gestation. Excess GWG was significantly and positively correlated with emotional (r=.45,
pr=.30, p=.02), and restrained eating (r=.31, p=02), as well as with reported frequency of
cravings for high fat foods (r=.42, p=.001) and fast food fats (r=.29, p=.02). Eating styles
accounted for 27% of the variance in frequency of cravings for high fat foods [F(3,52)=6.44,
p=.001], with external eating emerging as the single significant predictor (=.41, pb=6.06,
p=.01, 95% C.I.: 1.59, 10.53), as was the regression of external eating on frequency of
craving for high fat foods (b=.37, pb=3.67, p=.14, 95% C.I.: -1.30, 8.63), suggesting full
mediation. A Sobel test approached statistical significance (z=1.76, p=.08).

Findings are consistent with past research showing that food craving is a mediator of the
relationship between externality and BMI, and suggest that interventions focused on coping
with external cues and cravings may help pregnant women more effectively manage
gestational weight gain.
S1010 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Lauren Blau, B.S., University at Albany, SUNY, Albany, NY, 12222;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1011

B065 6:00 PM-7:00 PM

GARDASIL FOR BOYS AND YOUNG MEN: PARENTS' AND PHYSICIANS' REPORTED BARRIERS TO
VACCINE UPTAKE

Kate E. Dibble, M.A.1, Jessica L. Maksut, M.A.1, Elizabeth J. Siembida, Ph.D.2, Keith M. Bellizzi,
Ph.D., MPH1

1
University of Connecticut, Storrs, CT; 2University of Connecticut, Bethesda, MD

Background: The human papillomavirus (HPV) can cause penile, anal, and oropharynx cancers,
as well as genital warts among boys and men who are sexually active. Within the United
States, the Human Papillomavirus Quadrivalent vaccine, Gardasil(R), was approved for use in
boys and young men in 2009; however, despite its documented effectiveness, uptake of the
vaccine among this population remains limited.

Methods: In order to better understand the discrepancy between the vaccines documented
effectiveness and its lack of uptake among boys and young men, we conducted a systematic
search of the literature and qualitatively reviewed 10 published studies that focused on
barriers to Gardasil(R) uptake and reasons for uptake among boys and young men, as reported
by both parents (n=5 studies) and physicians (n=5 studies).

Results: Parents reported a number of barriers to vaccine uptake, including low awareness of
the existence of Gardasil(R) (n=4, 80%), its applicability to boys and young men (n=4, 80%), and
the vaccines cost (n=2, 40%). Parents were most likely to vaccinate their sons to protect them
from cancer (n=4, 80%) and protect their sons future sexual and romantic partners (n=4,
80%). Physicians also reported concerns regarding the cost of the vaccine, specifically in
regards to the patients healthcare coverage (n=3, 60%), but also reported additional barriers
to vaccine uptake, including the lack of physician/parent communication (n=4, 80%) and a
concern for vaccine efficacy (n=3, 60%). Similar to parents, physicians primary reasons for
promoting the Gardasil(R) vaccine to boys and young men included cancer prevention (n=3,
60%) and promotion of sexual health (n=3, 60%).

Conclusion: Parents and physicians aware of the Gardasil(R) vaccine reported similar reasons
for uptake of the vaccine in boys and young men. However, parents and physicians differed on
the main barriers to vaccine uptake. Particularly, many parents were unaware of the
Gardasil(R) vaccine and how it might be useful for their sons. Physicians, in contrast, struggled
with the doctor-parent relationship and how to engage in conversations regarding the
S1012 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Gardasil(R) vaccine. Future research needs to establish best practice promotional strategies for
Gardasil(R) uptake, and should focus on both parental and physician audiences in order to
maximize vaccination uptake.

CORRESPONDING AUTHOR: Kate E. Dibble, M.A., University of Connecticut, Storrs, CT, 06269-
1058; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1013

B066 6:00 PM-7:00 PM

IMPACT OF PROVIDER SUPPORT SERVICES ON BREASTFEEDING DURATION

Emily D. Fry, Undergraduate Student, Sarah Rhoades-Kerswill, Ph.D., Misty AW Hawkins, Ph.D.

Oklahoma State University, Stillwater, OK

Background: The benefits of breastfeeding for both infants and mothers are well-
documented. However, many women struggle to successfully breastfeed resulting in a call
to action to promote and aid in breastfeeding by providing support and services to increase
the success of breastfeeding mothers. This support may include lactation consultants inside
and outside of the hospital, nurses, doulas, midwives and medical providers (MD/CNP).

Objective: This study examined factors relating to breastfeeding duration, including types of
provider support and services. Specifically, the use of lactation consultants, nurses, doulas,
midwives, and medical providers on breastfeeding success were explored. It was hypothesized
that these support services have a positive effect on breastfeeding duration.

Methods: Snowball sampling was used to recruit participants for the Breastfeeding Opinions,
Outcomes, Behaviors, and Services (BOOBS) online survey. The current sample consisted of
1,964 mothers whose first child was born between 2004 and 2014, who had attempted to
breastfeed this child, and who answered questions about using providers to aid with
breastfeeding efforts.

Results: Of the sample, 61% breastfed for at least 12 months. It was found that breastfeeding
support from a lactation consultant outside of a hospital setting (t(1860)= -4.275, p< .05), a
doula (t(1699)= -7.581, p< .05) or a midwife (t(1698)= -8.572, p< .05) was associated with
longer breastfeeding durations. The greatest difference was seen among women who had
support from a doula. These women breastfed for 17 months, almost 4 months longer than
those who did not have a doula. Women who received breastfeeding support from a nurse or
doctor/nurse practitioner reported shorter breastfeeding durations (t(1,774) = 4.06, p =.01,
t(1752) = -8.57, p < .05 respectively). This was most notable for women who had seen a
doctor/CNP, who breastfed for 12 months, 2 months less than women who did not.

Conclusion: Results indicate that the use of breastfeeding support services, specifically
lactation consultants outside of the hospital, doulas, and midwives were associated with
longer breastfeeding durations. These findings indicate that implementation and utilization of
these breastfeeding support services may be important factors in increasing the likelihood of
successful breastfeeding experiences.
S1014 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Emily D. Fry, Undergraduate Student, Oklahoma State University,

Stillwater, OK, 74075; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1015

B067 6:00 PM-7:00 PM

INTERVENTIONS TARGETING MENS ALCOHOL USE AND FAMILY RELATIONSHIPS IN LOW AND
MIDDLE-INCOME COUNTRIES: A SYSTEMATIC REVIEW

Ali Giusto, MA, Eve Puffer, Ph.D.

Duke University, Durham, NC

Problem drinking disproportionally affects men and has disabling psychosocial, behavioral,
and physical consequences. These can lead to a cascade of negative effects on mens families,
with documented ties to intimate partner violence (IPV) and child maltreatment. These multi-
level problems are often exacerbated in contexts with high poverty rates, including low and
middle-income countries (LMICs). In societies where strong patriarchal norms place men in
positions of power, family-level consequences are often even more pronounced. To
understand how these intertwined consequences have been addressed, we conducted a
systematic review of the empirical literature on interventions conducted in LMICs targeting
mens problem drinking and any family-related outcomes (i.e., parenting, IPV). Cochrane and
PRISMA procedures guided the review, and the search was conducted in PsychInfo, PubMed,
and Web of science databases. The initial search yielded 1261 publications that were
evaluated, with eight studies from four different countries meeting inclusion criteria. Of
those, none had the primary goal of improving both drinking and related family-level
outcomes; rather these were secondary outcomes for the majority of the interventions, and
none directly targeted parent-child relationships. However, half of the studies documented
modest improvements on both drinking and couples or family relationship outcomes,
supporting the value of combined alcohol and family interventions. Specific strategies
emerged as promising across outcomes including behavioral, communication, narrative, and
participatory learning; gender-transformative approaches were associated with reduced IPV
and more equitable gender norms, and motivational interviewing and behavioral approaches
were beneficial for reducing alcohol use. Overall findings highlight the scarcity of
interventions addressing mens drinking and its effects on families, particularly in the domain
of parenting. However, results point to strategies that, combined with the broader evidence-
base on family-based interventions, can guide the development and rigorous evaluation of
new combined approaches.

CORRESPONDING AUTHOR: Ali Giusto, MA, Duke University, Durham, NC, 27705;
[email protected]
S1016 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B068 6:00 PM-7:00 PM

SELF-EFFICACY PREDICTS CHANGE IN HEALTH KNOWLEDGE IN MIDDLE SCHOOL AGE GIRLS

Mary Kate Clennan, B.S.1, Daniella Carucci, B.A., M.S.2, Shannon Chiles, M.S3, Marissa D. Alert,
M.S.1, Qingyun Liu, N/A1, Sarah Conboy, MA 4, Ayesha Ware, BA5, Maria Isabel. Leeder, M.A.6,
Sylvia Perez, M.A.7, Maria M. Llabre, PhD1, Judy Brown, Emerita8, Patrice G. Saab, Ph.D.1

1
University of Miami, Coral Gables, FL; 2University of Miami, Miami, FL; 3University of Miami,
Louisville, KY; 4NYSCI, Corona, NY; 5New York Hall of Science, Corona, NY; 6Patricia and Phillip
Frost Museum of Science, -, FL; 7New York Hall of Science, Queens, NY; 8Patricia and Phillip
Frost Museum of Science, Miami, FL

Introduction:
Health knowledge and self-efficacy are common outcome variables measured in health
promotion studies. While motivational theory suggests children with higher general self-
efficacy are more motivated to learn information, and, thus, produce higher classroom
achievement scores, research on the relationship between health self-efficacy and health
knowledge is virtually nonexistent. The purpose of the current study was to investigate
whether baseline self-efficacy predicts change in health knowledge scores following a health
promotion intervention.

Methods:
This project investigated the relationship between self-efficacy and health knowledge in 288
middle school age girls of diverse ethnic backgrounds who were enrolled in a 3-week summer
camp that emphasized the importance of a healthy lifestyle. Nutrition and physical activity
knowledge were quantified as the percent of items correct on a 7-item nutrition knowledge
assessment and on a 6-item physical activity knowledge assessment, respectively. The
participants also completed two questionnaires assessing their nutrition and physical activity
self-efficacy. All measures were taken at baseline and at the end of camp.

Results:
Two multilple regression models were run, a dietary and physical activity model, controlling
for baseline knowledge, age, maternal education, ethnicity, and treatment condition. Dietary
results suggest that baseline nutrition self-efficacy and the covariates accounted for
significant variability in nutritional knowledge score at the end of the program, adjusted R2 =
.286, F (10, 220) = 10.31, p < .001. Results indicate that, controlling for covariates, for every
one-point increase in baseline nutrition self-efficacy, week 3 nutritional knowledge score
increased by 5.5%, 95% CI [.12, 10.93]. Similarly, the physical activity model results suggest
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1017

that baseline physical activity self-efficacy and the covariates accounted for significant
variability in physical activity knowledge score at the end of the program, adjusted R2 = .305,
F (10, 221) = 11.23, p< .001. Results indicate that, controlling for covariates, for every one-
point increase in average baseline physical activity self-efficacy item score, week 3 physical
activity knowledge score increased by 19.03%, 95% CI [4.38, 33.67].

Conclusion:
The results of the present study show that greater gain in health knowledge scores were
predicted by higher baseline health self-efficacy. Heath knowledge is important to possess as
increased knowledge may lead to positive attitudes toward healthy behaviors, instill fear of
the risks of being unhealthy and motivate healthy weight behaviors in children. The findings
suggest that, in order to assure that children can maximize the benefits of health
interventions, it may be advantageous to direct intervention efforts at increasing self-efficacy
before targeting other health variables.

CORRESPONDING AUTHOR: Mary Kate Clennan, B.S., University of Miami, Coral Gables, FL,
33146; [email protected]
S1018 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B069 6:00 PM-7:00 PM

WHAT DOES PHYSICAL LITERACY MEAN TO YOU? AN EXPLORATION OF PARENTS

UNDERSTANDING AND PERCEPTIONS OF PHYSICAL LITERACY

Kristen C. Reilly, B.HSc, MPH, Shauna M. Burke, Ph.D.

Faculty of Health Sciences, Western University, London, ON, Canada

Purpose: Physical literacy is defined as the knowledge, confidence, motivation, and physical
competence to engage in physical activity for life (Whitehead, 2010). The purpose of this
study is to explore parents understanding and perceptions of physical literacy and its
importance in their family.

Methods: Parents of children aged 16 years and younger, living in Ontario, Canada, were
recruited to participate in an online questionnaire via social media (i.e., Facebook and Twitter)
and word-of-mouth. Questionnaires contained a combination of Likert-scale and open-ended
questions to examine parents: (1) understanding of physical literacy (e.g., What does
physical literacy mean to you?), and (2) perceived importance of physical literacy in their
family (e.g., How much do you agree or disagree with this statement: physical literacy is
important to me for my family?). Quantitative data were analyzed with descriptive statistics
and qualitative data underwent thematic analysis assisted by QSR NVivo 10.

Results: Participants (n = 81, MAge= 39.2 years) were predominantly female (93.8%) and
married (79.0%). More than half of respondents (62.5%) were familiar with the term physical
literacy. Participants provided a wide variety of meanings and definitions for physical literacy
including knowledge of physical activitys impact on health, awareness of physical activity
guidelines, and ability to perform fundamental physical movements. After reading a definition
and several examples of physical literacy provided by the researchers, all participants
(100.0%) agreed that it was important to them for their family. Overall, participants expressed
that they valued physical literacy because it promotes health and wellbeing, strengthens
family bonds through exercise, and enhances childrens capacity to engage in physical activity
for life. The majority of respondents (77.6%) believed that it could be improved by a moderate
or great extent in their family.

Conclusions: Preliminary findings demonstrate the variability in parents understandings of

physical literacy. Though, generally speaking, parents perceived physical literacy to be
important, they also believed that it could be improved within their families. Thus,
researchers should be encouraged to explore and investigate the feasibility of home and
community-based interventions and resources to enhance physical literacy in families.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1019

CORRESPONDING AUTHOR: Kristen C. Reilly, B.HSc, MPH, Faculty of Health Sciences, Western
University, London, ON, N6G 1H1; [email protected]
S1020 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B070 6:00 PM-7:00 PM

BODY IMAGE DISSATISFACTION, DEPRESSION, AND DISORDERED EATING IN COLLEGE-AGED

WOMEN SEEKING A YOGA INTERVENTION

Aviva H. Ariel-Donges, M.S., Eliza L. Gordon, M.S., Vivian Bauman, B.S., Masyn McMillin, N/A,
Oshin Rai, N/A, Anaiz Rodriguez, N/A, Daniela Sanchez, N/A, Brittney Dixon, MPh, Abraham
Eastman, B.S., Kathryn Ross, PhD, MPh, Michael Perri, PhD, ABPP

University of Florida, Gainesville, FL

Introduction: Previous studies have shown that 75% of college-aged women report body
image dissatisfaction, and negative body image has been shown to be an independent
predictor of disordered eating in young women. This study examined baseline characteristics
and correlates of young women interested in a yoga intervention for body image
dissatisfaction.

Methods: The preliminary study sample included 118 women (M SD; age = 21.0 2.7 years,
White = 64.4%, Asian = 14.4%, Black = 13.6%, More than one race/Not reported = 7.6%,
attending undergraduate/graduate school = 96.6%, BMI = 24.5 5.4 kg/m2) screened for a
yoga intervention for body image dissatisfaction. The relationships between baseline body
image dissatisfaction (Multidimensional Body-Self Relations Questionnaire Appearance
Scales; MBSRQ-AS), depressive symptoms (Beck Depression Inventory Second Edition; BDI-
2), disordered eating symptoms (Eating Attitudes Test 26), and demographic variables were
assessed using Pearson correlations and independent samples t-tests. An MBSRQ-AS score
< 3.0 was utilized as the clinical cut-off for significant body image dissatisfaction, and a BDI-2
score > 14 was utilized as the clinical cut-off for significant depressive symptoms.

Results: Clinically significant body image dissatisfaction was endorsed by 72% of participants.
Although prior studies suggest a 10% prevalence rate of depression in college-aged women
(e.g., Michael, 2006), 33.1% of women in this sample screened positive for clinically significant
depressive symptoms. Moreover, while only 28.8% of the women had BMIs in the
overweight/obese range, 55.1% reported trying to lose weight over the prior three months
and 70.3% reported being > 5 lbs. heavier than their ideal weight. More severe body image
dissatisfaction was associated with greater depressive symptoms, r = -.324, p < .001, and
greater disordered eating symptoms, r = -.264, p = .004. Moreover, those with negative body
image reported significantly greater depressive symptoms, t(97.8) = -3.63, p < .001, and
greater disordered eating symptoms, t(116) = -2.15, p = .034, than those with positive body
image. Mean body image dissatisfaction was not significantly different across ethnicities or
races (e.g., Caucasian vs. African American, p = .115).
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1021

Discussion: In this sample, clinically significant depressive symptoms were three times more
prevalent than in the general population of college-aged women, and a substantial proportion
of participants reported wanting to lose weight despite being at a healthy BMI. Notably,
worse body image dissatisfaction was associated with greater depressive and disordered
eating symptoms. These findings highlight the interrelation of mood, appearance evaluation,
and health behaviors in young women interested in a yoga intervention for body image
dissatisfaction.

CORRESPONDING AUTHOR: Aviva H. Ariel-Donges, M.S., University of Florida, Gainesville, FL,

32609; [email protected]
S1022 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B071 6:00 PM-7:00 PM

A RANDOMIZED CONTROLLED TRIAL COMPARING TWO CONSENT INTERVENTIONS FOR

ENROLLMENT INTO A SEQUENCING STUDY

Erin Turbitt, Ph.D.1, Paola Chrysostomou, B.S. 2, Alexis Heidlebaugh, BS1, Barbara B. Biesecker,
PhD, MS3

1
National Human Genome Research Institute, Bethesda, MD; 2National Institutes of Health/
National Institute of Child Health and Human Development, Bethesda, MD; 3National Human
Genome Research institute, National Institutes of Health, Bethesda, MD

Despite the widespread use of genome sequencing in research, evidence-based procedures

for obtaining participants consent to enter such studies are lacking. It is unknown how
successfully participants make an informed choice (defined as having sufficient knowledge
and attitudes consistent with ones decision) to enroll in a sequencing study. Increasingly,
such studies include return of secondary findings that should be made clear to participants.
Secondary findings are results unrelated to the primary reason (such as the health condition
under investigation) the sequencing was obtained.

This study compared an evidence-based shorter consent to a standard consent among

women affected with primary ovarian insufficiency and eligible to participate in an NIH
sequencing study. Participants were randomized to receive the shorter or the standard
consent document. A mixed methods approach yielded data from surveys at baseline,
immediate and six month follow up. Quantitative data were analyzed using descriptive
statistics and bivariate analyses. Thematic content analysis was used to analyze qualitative
data. Of the 387 women contacted, 212 were recruited and randomized (rr=55%), with
complete data available from 188 participants. There were no differences between the two
consent type groups in genome sequencing knowledge, perceived benefits or decisional
conflict (p>0.05). Participants had high expectations to learn information of personal benefit,
including secondary findings. Most (n=171) decided to participate in the sequencing study
prior to the consent process. Intentions to receive secondary findings were high (mean, 10.9;
SD, 1.6; maximum possible score 12) as were attitudes (mean, 6.5; SD, 0.9; maximum possible
score 7). Based on the data from the knowledge, attitudes and intentions scores, 90 (47.9%)
participants made an informed choice to receive secondary findings.

Participants have high expectations for receiving sequencing results of personal benefit. This
study demonstrates a lack of differnce in outcomes between the longer, standard consent and
the shorter consent suggesting that a more concise, evidence based consent is as effective at
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1023

consenting individuals to genome sequencing studies. However, it may be challenging to

achieve informed choice, as defined by current measures, for receiving secondary variants.
Future research should focus on the level of knowledge participants require to make an
informed choice about entering a sequencing study.

CORRESPONDING AUTHOR: Erin Turbitt, Ph.D., National Human Genome Research Institute,
Bethesda, MD, 20894; [email protected]
S1024 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B072 6:00 PM-7:00 PM

A SYSTEMS-THINKING, PROBLEM-SOLVING APPROACH TO STUDY MIGRAINE SELF-

MANAGEMENT AND EMPOWERMENT IN SOCIALLY VULNERABLE WOMEN

Deanna Befus, BSN, RN1, Kristen Hassmiller Lich, PhD2, Shawn Kneipp, PhD, RN, ANP-BC,
APHN-BC, FAANP3, Remy Coeytaux, MD, PhD4

1
Duke University School of Nursing, Durham, NC; 2University of North Carolina-Chapel Hill,
Gillings School of Global Public Health, Chapel Hill, NC; 3University of North carolina Chapel
Hill School of Nursing, Chapel Hill, NC; 4Community and Family Medicine, Duke University
Health System, Durham, NC

Migraine affects over 20% of women and is the sixth most disabling condition worldwide.
Both in incidence and severity, migraine disproportionately disables women with less
education, living in poverty, un- or underemployed, and/or with high levels of chronic stress
due to the unequal distribution of health-damaging experiences. Self-management is a crucial
part of chronic illness management, with direct correlations to health outcomes. Yet,
effectiveness of self-management behaviors is inextricably linked to social determinants of
health. Socially vulnerable women are often ignored in highly medicalized migraine research,
making their experiences and choices in managing migraine relatively unknown.

We developed a systems-thinking data collection approach that can feasibly be applied to

study migraine self-management behaviors using a complexity/systems science framework
with women in diverse social locations. We crafted a two-part activity for use in focus groups.
First, using systems dynamics concept mapping, women tracked a migraine episode, from
triggers through self-management strategies and desired and actual outcomes, including
behavior needs, facilitators, and supports. Next was a causal loop diagramming activity to
delve deeper into the relationships between triggers, self-management behaviors, and
outcomes. Part two--while often enlightening for affluent participantsleft more vulnerable
women feeling overwhelmed and defeated, with a solution to one challenge becoming the
cause of another. Through constant comparison analysis, we recalibrated the approach, using
a combination of theory, clinical experience, and participant feedback to replace the causal-
loop diagram with an individually-tailored, strengths-based problem-solving activity
highlighting relationships and trade-offs in a more empowering, actionable way. Bringing a
critical lens and strengths-based approaches for vulnerable populations can replace
traditional deficit thinking in healthcare, developing options for leveraging resources and
understanding complex health behaviors without losing sight of systemic, distributional justice
issues. These tools can provide a way to extrapolate the complexities of actual self-
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1025

management behaviors and challenges faced by vulnerable women with migraine, versus
what they may be instructed to do by a medical model that does not always account for the
social determinants of health.

CORRESPONDING AUTHOR: Deanna Befus, BSN, RN, Duke University School of Nursing,
Durham, NC, 27701; [email protected]
S1026 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B073 6:00 PM-7:00 PM

ASSESSING FIDELITY FOR A COMPUTER TAILORED INTERVENTION WITH TELEPHONIC

COACHING FOR DECISION-MAKING IN KIDNEY TRANSPLANT

Zoe Mushkat, BS1, Brian Tapscott, MA2, Alexandra Murray, MA2, Katherine Ashley, BS3,
Andrea Paiva, PhD4, Crystal K. Amerson, MPH5, Mark Robbins, PhD4, Amy Waterman, PhD5

1
University of Rhode Island, Narragansett, RI; 2University of Rhode Island, Kingston, RI;
3
University of Rhode Island, Portsmouth, RI; 4Cancer Prevention Research Center, University
of Rhode Island, Kingston, RI; 5David Geffen School of Medicine at UCLA, Department of
Nephrology, Los Angeles, CA

Your Path to Transplant (YPT) is an ongoing longitudinal study that aims to enhance
decision to pursue living donor kidney transplant (LDKT) in adults with end-stage renal disease
utilizing a telephonic coaching intervention grounded in the Transtheoretical Model (TTM) of
behavior change. This study aimed to evaluate reliability by assessing treatment fidelity to the
intervention protocol. YPT coaches complete standardized surveys and deliver tailored
feedback at three time-points over eight months, examining whether transplant education
and individualized, computer tailored TTM feedback increases readiness, self-efficacy and
decisional balance to pursue LDKT compared to usual care transplant education.
Five raters evaluated a random sample of audio recorded coaching sessions. Rater pairs
assessed at least 10% of sessions at each study time-point. Coaching fidelity was assessed
using a standardized rubric that assigns a scale value (1-3) to the coachs standardized
interviewing capacity, quality of TTM feedback, and relevant clinical skills (e.g. reflective
listening). Mean scores of all rubric items served as fidelity adherence scores for the coachs
survey. Rater agreement was evaluated using the two-way random effects Intraclass
Correlation Coefficient (ICC (2, k)) for each rater pair.
All surveys (n=777) were recorded for fidelity review, with 13% (101) of all intervention
condition surveys rated by 2+ raters; scores ranged from 1.40 to 3 (M=2.58, SD=0.27). ICC
revealed that rater agreement ranged from 0.76 to 0.92 where sample size was sufficient
(n10 ratings between rater pairs), (M=0.83, SD=0.07). Raters demonstrated good (ICC=0.75
0.90) to excellent (ICC>0.90) reliability.
Significant ICC results indicate that fidelity raters are evaluating TTM telephonic
feedback similarly and consistently. Monitoring intervention fidelity supports treatment
efficacy and provides coaches with consistent evaluation and feedback for telephonic expert-
system delivery. Future research will aim to evaluate whether rater agreement, which
supports reliability, is also providing effective feedback for telephone coaches that supports
patient decision-making for kidney transplant.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1027

CORRESPONDING AUTHOR: Zoe Mushkat, BS, University of Rhode Island, Narragansett, RI,
02882; [email protected]
S1028 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B074 6:00 PM-7:00 PM

DEVELOPING AN INTERNATIONAL MEASURE OF BELIEFS ABOUT CAUSES OF MENTAL ILLNESS:

THE MENTAL ILLNESS ATTRIBUTION QUESTIONNAIRE

Brandon A. Knettel, Ph.D.

Duke Global Health Institute, Raleigh, NC

The study of mental illness attributions, or beliefs about the causes of mental illness, is well-
documented and ongoing in the academic literature. Attributions have traditionally been
dichotomized along four dimensions based on their locus, controllability, specificity, and
stability and have been associated with a wide variety of thoughts, emotions, beliefs, and
actions related to the stigma of mental illness. However, more recent studies have introduced
a new cross-cultural model incorporating lay beliefs about the specific biological, emotional,
social, and spiritual causes of mental health problems. The present research outlines the
design and initial validation of a comprehensive, international measure of causal beliefs using
this new cross-cultural model, the Mental Illness Attribution Questionnaire (MIAQ). The
research project included item formulation, piloting, qualitative rating, and preliminary
validation with a sample of 680 international students representing 94 nations. Factors
captured causes related to supernatural forces, social/stress, lifestyle, physical health,
substance use, heredity/biology, and personal weakness. This structure was tested for model
fit using confirmatory factor analysis across three vignette conditions one each describing a
man with schizophrenia, depression, or alcoholism with further examination yielding strong
test-retest reliability and promising convergent, discriminant, and cultural validity data. Taken
together, the results support the MIAQ as a valid and reliable measure of mental illness
attribution with strong potential for use in future studies, particularly those examining
barriers to help-seeking and stigmatizing behavior across cultures.

CORRESPONDING AUTHOR: Brandon A. Knettel, Ph.D., Duke Global Health Institute, Raleigh,
NC, 27613; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1029

B075 6:00 PM-7:00 PM

DOES COGNITIVE STATUS MODIFY THE RELATIONSHIP BETWEEN SELF-RATED HEALTH AND
MORTALITY?

M. Scott DeBerard, Ph.D.1, Gail Rattinger, PharmD, PhD2, Mindy Dokos, B.A.1, Bryce Cooley,
N/A1, Chelsea Sanders, M.S.1, Stephanie Behrens, N/A3, Elizabeth K. Vernon, BA4, JoAnn
Tschanz, Ph.D.1

1
Utah State University, Logan, UT; 2Binghamton University, Binghamton, NY; 3Utah State
University, Lewiston, UT; 4Utah State University, River Heights, UT

Self-rated health status (SRHS) predicts mortality in older hospitalized patients and
community-dwelling persons, however its predictive validity in persons with dementia has not
been established. We examined: 1) the correlation between SRHS and nurse ratings of health
status (NRHS); 2) whether SRHS/NRHS predicted 12-year mortality; and 3) whether cognitive
status modified the association between SRHS/NRHS and mortality in a population-based
sample of older adults.

3340 participants [58% female; mean(SD) age = 77.3(6.47) years] had SRHS; a subset (686;
56% female) had both SRHS and NRHS. SRHS was assessed by asking how the participant felt
over the past week (excellent, good, fair or poor), and NRHS was based on the General
Medical Health Rating, which considers stable/unstable health conditions, frailty, and results
of a physical exam. We used Spearmans rank correlation to examine the association between
SRHS and NRHS and Cox proportional hazards regression models to examine time to mortality
for each factor. Covariates tested included cognitive impairment [none, cognitive impairment-
no dementia, dementia], age, sex, depression, and body mass index.

SRHS and NRHS were modestly correlated ( = .31), but not in persons with dementia. NRHS
and SRHS predicted mortality: hazard ratio (95% CI) for NRHS fair/poor was 3.49 (2.59-4.70);
SRHS poor was 2.65 (2.18-3.21), compared to those with excellent health. Cognitive status
did not modify the associations between SRHS/NRHS and mortality. BMI and depression did
not further predict mortality.

Self-rated health predicted mortality regardless of cognitive status. A simple self-rating may
provide a quick and valid way to monitor health in older adults, even among persons with
mild dementia.
S1030 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: M. Scott DeBerard, Ph.D., Utah State University, Logan, UT,
84322-2810; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1031

B076 6:00 PM-7:00 PM

EFFECT OF SERVICE QUALITY AND PERCEIVED RISK ON SERVICE VALUE, CONSUMER

SATISFACTION AND CHURN INTENTION IN KOREAN NURSING HOMES

Jiyoung An, PhD, MPH, Claudio R. Nigg, PhD, FSBM

University of Hawaii, Honolulu, HI

Background: Korea has a rapidly increasing population of elders with 11.0% aged 65 and older
in 2010, growing to 20.8% in 2026. Accordingly, service quality provided by long-term care
(LTC) facilities and consumer retention (vs. new consumers) is paramount.

Purpose: Therefore, the effect of service quality (SQ) and perceived risk (PR) by consumers on
perceived service value (PSV), consumer satisfaction (CS), and churn intention/intention to
leave (CI) in LTC facilities was investigated.

Methods: Three dimensions of SQ are modelled(technical quality (TQ), functional quality (FQ),
and service environment (SE)) and PR, assessing their effects on CS and PSV. A cross-sectional
survey using convenience sampling of families or guardians of elders living in LTC facilities was
utilized because these individuals are likely to make substantial elder-care decisions (n=235).

Results: The model fit statistics were 2 = 340.202 (d.f. = 152, p = 0.000), GFI = 0.862, AGFI =
0.810, NFI = 0.879, CFI = 0.928, RMR = 0.041, and RMSEA = 0.073, indicating acceptable fit. In
the model, SQ consisting of TQ, FQ, and SE had a statistically significant impact on PSV. PR had
a negative impact on PSV ( = -0.122, p = 0.004). Only TQ ( = 0.492, p = 0.016) and PSV ( =
0.397, p = 0.008) had statistically significant impacts on CS. Finally, CS ( = -0.327, p = 0.037)
and PR ( = 0.510, p = 0.001) had statistically significant impacts on CI. For direct effects, CS
was positively related to TQ (standardized coefficient = 0.474) and PSV (standardized
coefficient = 0.328). CI was positively related to PR (standardized coefficient = 0.453).

Conclusion: PR related to nursing service had a direct impact on CI. Therefore, it is necessary
to reduce uncertainty and risks related to the service. It also seems necessary to improve SQ
perceived by consumers if the goal is to reduce CIof families or guardians.

CORRESPONDING AUTHOR: Jiyoung An, PhD, MPH, University of Hawaii, Honolulu, HI, 96822;
[email protected]
S1032 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B077 6:00 PM-7:00 PM

ENHANCING PROCEDURES FOR INFORMED CONSENT (EPIC) STUDY

Jordan Skan, MS1, Mariah Knox, BA2

1
Alaska Native Tribal Health Consortium, Anchorage, AK; 2Alaska Native Medical Hospital,
Anchorage, AK

Informed consent is an important part of research with human participants. Historically,

deception was common practice. Through extensive litigation, the rights of human
participants in research have grown. This study is evaluating a technology-mediated method
of informed consent. The project is part of the Healing and Empowering Alaskan Lives
Towards Healthy-Hearts (HEALTHH) study; a video telemedicine intervention to improve heart
health for Alaska Native (AN) People in rural Alaska. AN People have endured attempted
cultural, physical, and social genocide. In addition, unethical research practices with AN
People are well documented. Due to the problematic history of unethical research, historical
trauma, and intrusion of a western worldview, it is a reasonable concern that AN People may
be a vulnerable research population. This, coupled with additional concerns presented by
telehealth delivery, highlights the need for researchers to ensure that truly informed consent
is obtained from AN People. A multimedia enhanced form of the consent process may provide
a better understanding of the elements of the study and allow participants to be more
thoroughly informed when choosing if they wish to take part in studies. This project created a
9-minute animated video version of the consent form to enhance the informed consent
process. The information presented as a hypothetical patient joining the study and going
through study activities. Half of the participants were randomized to the standard paper-
based consent while the other received the addition of the video. Immediately after, a brief
survey and the MacArthur Competence Assessment Tool for Clinical Research (MacCAT-CR)
were administered. The MacCAT-CR is an adaptable capacity assessment that measures
understanding, appreciation, reasoning, and the ability to express a choice. It provides an
overall score which can be used for comparisons. Preliminary data were gathered from 17 AN
People (8 men, 9 women) with an age range of 32 to 64 (M=52 SD=10.5). The MacCAT-CR
scores for the intervention group (M=27, SD=4) were significantly higher (better) than the
control group (M=18, SD=9); t(15) = 2.66, p = 0.018, demonstrating better comprehension and
ability to express a choice. Our video enhancement to the informed consent process is a
useful prototype with large effects and applicability to a variety of health-related research
projects.

Study funding: NHLBI Diversity Supplement Award to Grant #3R01HL117736

ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1033

Study funding: NHLBI Diversity Supplemnt Award to Grant #3R01HL117736

CORRESPONDING AUTHOR: Jordan Skan, MS, Alaska Native Tribal Health Consortium,
Anchorage, AK, 99501; [email protected]
S1034 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B078 6:00 PM-7:00 PM

HOW DO HEALTH EXPECTANCIES & VALUES COMBINE TO DETERMINE DIETARY INTENTIONS?

APPLYING FUNCTIONAL MEASUREMENT TO FOOD INTAKE

Lauren A. Fowler, BS1, Paige Clarke, MA2, Steffi Renninger, MPhil3, Zeljka Macura, B.S.1, Philip
Moore, PhD4, Tonya Dodge, PhD1

1
George Washington University, Washington, DC; 2George Washington University, Bedford,
NY; 3The George Washington University, Germantown Rd, MD; 4The George Washington
University, Washington, DC

Background. The expectancy-value (EV) framework is one of the most prominent theoretical
approaches to understanding and predicting health-related outcomes. By and large, EV
models have predicted behavioral intentions (and, by extension, behaviors) by aggregating
expectancies and values multiplicatively, positing that this is how these factors are cognitively
integrated. However, this assumption has yet to be tested vis--vis other cognitive integration
rules. This research used Information Integration Theory (IIT) to test empirically how health-
related expectancies and values are cognitively integrated to determine dietary intentions.

Method. Ninety participantswith a mean age of 19.03 years (SD=0.92) and 73 females
were presented with a scenario depicting physical health consequences of junk food
consumption (weight gain, disease risk), which varied in terms of objective probability (low,
medium, high, and no probability information), and corresponding subjective participant
expectancies. The values participants assigned to these consequences were trichotomized
into low, medium, and high levels, and participants indicated their likelihood of consuming
junk food for each of the probability conditions.

Results. Participants combined their health-related expectancies and values using an

additiverather than a multiplicative or averagingcognitive integration rule to determine
their intentions to consume junk food.

Conclusions. Dietary research and treatmentparticularly for junk food consumptionmay

benefit by aggregating expectancies and values additively, and by addressing both in
interventions designed to promote behavior change.

CORRESPONDING AUTHOR: Lauren A. Fowler, BS, George Washington University, Washington,

DC, 20001; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1035

B079 6:00 PM-7:00 PM

IMPACT OF SNAP ON HEALTH DISPARITIES AMONG ELDERLY PATIENTS WITH NUTRITIONAL

DISEASE

Layla Booshehri, PhD, Jerome Dugan, PhD

Drexel University, Philadelphia, PA

This study investigates the impact of households altering their financial behavior to qualify for
the nutritional safety net programs (i.e., Supplemental Nutritional Assistance Program, SNAP)
on the incidence of nutritional diseases. Our study exploits the fact that individuals have the
option to modify their financial behavior to deduct medical expenses at age 60 to qualify for
SNAP (instead of being subject to a gross income test) and we use a difference-in-differences
(DD) model to examine the impact of the SNAP eligibility on the morbidity of nutritionally
sensitive diseases. Our primary study population consists of Medical Expenditure Panel Survey
(MEPS) respondents aged 56 to 64. For the overall sample, the DD estimates show that SNAP
has no impact on nutritional disease morbidity, with an exception for diabetes, where SNAP is
associated with a statistically significant 10.4% (P < 0.10) decline in the incidence of diabetes.
However, when narrower racial subgroups are considered, the results vary widely across race.
Whites only report a 9.5% (P < 0.05) decline in the incidence of diabetes. On the other hand,
blacks report that SNAP is associated with a 9.1% (P < 0.10) decline in the incidence of cancer,
29.8% (P < 0.01) decline in the incidence of hypertension, and 24.6% (P < 0.05) decline in
diabetes. These results are significant, as they demonstrate that low-income households
exploit changes in the SNAP eligibility rules by altering their financial decision-making (i.e.,
itemizing deductions on tax returns) to receive a direct cash-transfer from the government.
Moreover, the results demonstrate that the altered financial decision-making leads to the
adoption of health promoting behaviors that directly lead to a lower incidence of nutritional
disease, and that minorities, blacks in particular, benefit the most from the program, which
closes a significant baseline nutritional disease disparity across racial groups.

CORRESPONDING AUTHOR: Layla Booshehri, PhD, Drexel University, Philadelphia, PA, 19102;
[email protected]
S1036 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B080 6:00 PM-7:00 PM

LIVING KIDNEY DONATION: ASSESSING POTENTIAL BARRIERS AND FACILITATORS THROUGH

NOMINAL GROUP TECHNIQUE

Lindsey Dorflinger, PhD1, Sanjay Kulkarni, MD2, Carrie Thiessen, MD, PhD2, Liana Fraenkel,
M.D.C.M; MPH2

1
Walter Reed National Military Medical Center, Bethesda, MD; 2Yale School of Medicine, New
Haven, CT

The number of individuals in need of donated kidneys far surpasses the number of donors,
and, despite the literature showing improved outcomes for recipients of living compared to
deceased kidney donation, the number of living donors has decreased over the past decade.
The vast majority of individuals are aware of the need for donation, support the idea of living
donation, and would consider donating to a known recipient, and many even report that they
would consider donating to an unknown recipient. However, only a tiny proportion of these
individuals ever go on to donate. This study utilized nominal group technique, a method
designed to generate ideas around a specific topic and determine which ideas are most salient
among group members, to determine what factors would make one more or less likely to
serve as a living donor, as well as which factors are most important in the decision-making
process. Thirty individuals, recruited from a large academic medical center, participated in
four groups (two groups of seven women and two groups of eight men) in which they were
asked to brainstorm factors that would impact willingness to be a living donor and vote on
which factors they deemed most important. Responses were analyzed and categorized into
themes. Participants identified and ranked the following factors (from most to least
important): altruism, relationship to recipient, knowledge, personal risk/impact,
convenience/access, cost, support, personal benefit, and religion. Ranking lists were similar
for men and women, though women were more likely to report knowledge as a key factor in
decision making, while men were more likely to identify altruism as a key factor. While much
of the literature on living donation tends to debate the ethics of providing incentives for
donation and eliminating disincentives, cost appears to play a smaller role in decision making
compared to factors such as altruism and knowledge about the process and the personal
impact that donation would have on the donor. Future studies should explore whether public
campaigns that specifically address these factors might help to reduce the significant gap in
supply and demand of kidneys.

x
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1037

CORRESPONDING AUTHOR: Lindsey Dorflinger, PhD, Walter Reed National Military Medical
Center, Bethesda, MD, 20889; [email protected]
S1038 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B081 6:00 PM-7:00 PM

LUNG CANCER SCREENING: A QUALITATIVE STUDY EXPLORING THE DECISION TO OPT OUT OF
SCREENING

Lisa Carter-Harris, PhD, RN, ANP-C1, Susan Brandzel, MPH2, Karen J. Wiley. Wernli, PhD, RN,
ANP-C2, Joshua Roth, PhD, MHA3, Diana S.M. Buist, PhD, MPH2

1
Indiana University-Purdue University Indianapolis (IUPUI), Indianapolis, IN; 2Group Health
Research Institute, Seattle, WA; 3Fred Hutchinson Cancer Research Center, Seattle, WA

Aims. Lung cancer screening with annual low-dose computed tomography is relatively new for
long-term smokers in the U.S. supported by a US Preventive Services Task Force Grade B
recommendation. As lung cancer screening programs are more widely implemented and
providers engage patients about lung cancer screening, it is critical to understand what
influences a patients decision to screen, or not, for lung cancer. Understanding lung cancer
screening behavior among high-risk smokers who opt out provides insight, from the patient
perspective, about the shared decision-making process for lung cancer screening. The purpose
was to explore screening-eligible patients decision to opt out of lung cancer screening after
receiving a provider recommendation. This knowledge will inform intervention development
to enhance shared decision-making processes between long-term smokers and their providers
and decrease decisional conflict about lung cancer screening.

Methods. Semi-structured qualitative telephone interviews were performed with 18 lung

screening-eligible men and women who were members of an integrated, mixed model
healthcare system in Seattle about their decision to opt out of lung cancer screening.
Participants met lung cancer screening criteria for age, smoking and pack-year history. Audio-
recorded interviews were transcribed verbatim. Two researchers with cancer screening and
qualitative methodology expertise conducted data analysis using thematic content analytic
procedures.

Results. Participant mean age was 66 years (SD 6.5). Majority were female (61%), Caucasian
(83%), current smokers (61%). Five themes emerged: 1) Knowledge Avoidance; 2) Perceived
Low Value; 3) False Positive Worry; 4) Practical Barriers; and 5) Patient Misunderstanding.

Conclusions. Many screening-eligible, smokers opt out of lung cancer screening. The
participants in our study provided new insights into why some patients make this choice.
While there are known drawbacks to lung cancer screening, and it is not necessarily the best
option for everyone who is eligible, it is known to be effective in early lung cancer detection
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1039

among high-risk patients. Understanding why people decide not to screen will enhance future
efforts to improve knowledge transfer from providers to patients about the risks and benefits
of lung cancer screening and ultimately enhance shared decision-making about lung cancer
screening.

CORRESPONDING AUTHOR: Lisa Carter-Harris, PhD, RN, ANP-C, Indiana University-Purdue

University Indianapolis (IUPUI), Indianapolis, IN, 46202; [email protected]
S1040 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B082 6:00 PM-7:00 PM

PREDICTING COLLEGE STUDENTS HUMAN PAPILLOMAVIRUS VACCINATION INTENTION

Apurva Barve, M.A.1, Scott Roesch, Ph.D2, Eric Walsh-Buhi, Ph.D., MPH2, Kristen Wells, Ph.D.,
MPH2

1
University of California, Los Angeles, Los Angeles, CA; 2San Diego State University, San Diego,
CA

A catch-up Human papillomavirus (HPV) vaccine is recommended for 13 to 26 years-old

females and 13 to 21 years-old males who have not been vaccinated previously. Most
Americans reach adulthood without receiving the recommended HPV vaccine doses. HPV
vaccine uptake is especially lower in key priority populations, indicating significant racial
disparities and potential for enhancing preventive healthcare. Increasing the HPV vaccination
series uptake requires theory-driven empirical research on factors influencing HPV-related
decisions among young adults. Based on Social Cognitive Theory, this study assessed the
interaction between personal and environmental determinants of vaccine uptake and its
effect on behavioral intention to obtain the HPV vaccine. The lead authors Psychology
Department participant pool was utilized to anonymously collect data from 209 participants,
aged 18-26 years who had not yet obtained any dose of HPV vaccine. Data were analyzed
using a series of eight Pearson or point biserial correlation analyses, one mediation analysis
using PROCESS macro, and one path analysis. Findings demonstrated that: (1) social influence
(i.e., parent and peer endorsement of HPV vaccine), HPV vaccine-related self-efficacy,
attitudes and knowledge regarding the vaccine, and physicians recommendation to obtain
the HPV vaccine were correlated with intentions to obtain the HPV vaccine (all ps < .05); (2)
controlling for gender, self-efficacy mediated the relationship between social influence and
intentions to obtain HPV vaccine, p were related to physicians recommendation, and this
relationship, in turn, predicted intentions to obtain the HPV vaccine (both ps< .05). The
present study offers an understanding of personal and environmental factors associated with
HPV vaccine decisions among college students, and also highlights the need for studying these
mechanisms longitudinally.

CORRESPONDING AUTHOR: Apurva Barve, M.A., University of California, Los Angeles, Los
Angeles, CA, 90066; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1041

B083 6:00 PM-7:00 PM

REDUCED ALTRUISM DURING DECISION MAKING IN VACCINE HESITANT INDIVIDUALS

Alex Francisco, M.A.1, Jared Bruce, PhD1, David Jarmolowicz, PhD2, Seung-Lark Lim, PhD1,
Amanda S. Bruce, Ph.D.3

1
University of Missouri - Kansas City, Kansas City, MO; 2University of Kansas, Kansas City, MO;
3
University of Kansas Medical Center, Kansas City, KS

The United States has seen a recent resurgence of vaccine preventable diseases. Cases of
measles have doubled in some instances when vaccine hesitant individuals intentionally
withheld immunizations. Rates of personal beliefs exemptions for school vaccinations have
increased as well. Under herd immunity, people who receive vaccinations provide a health
benefit to people who are not vaccinated (Vietri et al., 2012). For instance, vulnerable
populations who cannot safely be vaccinated receive protection from high levels of population
vaccination uptake. This suggests that altruism may influence vaccine uptake. Our hypothesis
was that individuals who are hesitant in the efficacy of vaccines would show lower altruism,
thereby discounting the benefits to individuals outside their immediate social networks.
mTurk was used to recruit vaccine confident (n = 50) and vaccine hesitant (n = 70) individuals
for the present study. Participants completed a hypothetical treatment choice paradigm that
presents a range of vaccine risks and benefits. This paradigm asks participants to consider
benefits to socially distant individuals (using a scale from 1-100) as well as benefits to their
children, and examines how their willingness to risk side effects changes as benefits to others
are considered. Area under the curve (AUC) was used to assess altruistic motivation. Larger
AUC indicates more willingness to risk side effects to prevent disease at each social distance.
A 2x5 (hesitancy by social distance) repeated measure ANOVA assessed the main effects of
hesitancy and social distance, as well as the interaction of vaccine hesitancy and social
distance. Social distance proved a significant main effect (f(4,472) = 3.05, p < .0001) as did
vaccine hesitancy (f(1,118) = 10.57, p = .001). Moreover, we uncovered a significant social
distance by vaccine hesitancy interaction (f(4,472) = 3.05, p = .04). Results suggest that
vaccine hesitant individuals are less altruistic when making their decisions and heavily
discount vaccine benefits for their children as well as individuals outside their immediate
family. Previous models have suggested that altruism may account for up to 25% of decisions
to vaccinate (Shim et al., 2012), and this therefore could be used as a method to promote
vaccine uptake and health outcomes. Our research quantifies the amount that vaccine
hesitant individuals de-value potential benefits to their children and others during vaccine
decisions. Future research may wish to examine means of providing psychoeducation to
increase altruism among vaccine hesitant individuals.
S1042 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Alex Francisco, M.A., University of Missouri - Kansas City, Kansas
City, MO, 64109; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1043

B084 6:00 PM-7:00 PM

ASSOCIATION BETWEEN PURPOSE IN LIFE AND GLUCOSE CONTROL AMONG MIDDLE-AGED

AND OLDER ADULTS

Dina Griauzde, MD1, Michele Heisler, MD, MPA1, Hwajung Choi, PhD1, Claire Ankuda, MD,
MPH1, Tyler Winkelman, MD1, Jeffrey Kullgren, MD, MS, MPH2

1
University of Michigan, Ann Arbor, MI; 2University of Michgan, Ann Arbor, MI

Background: Greater purpose in life is associated with lower rates of certain chronic diseases.
Whether purpose can protect against development of prediabetes or type 2 diabetes is
unknown.

Purpose: To examine the association between purpose in life and change in blood glucose
among 3,914 adults age 50 and without a history of type 2 diabetes or prediabetes in the
Health and Retirement Study.

Methods: Baselinepurpose in life was measured using a 7-item, validated adaptation of Ryff
and Keyes' Scales of Psychological Well-Being. Participants were then grouped into tertiles of
purpose in life (high, medium, or low). We used multivariable linear regression to examine
the association between change in HbA1c from baseline to 4-year follow-up and purpose in
life. Multivariable logistic regression was used to examine the association between purpose in
life and incident prediabetes or type 2 diabetes over the same time period.

Results: After adjusting for sociodemographic factors, body mass index, physical activity, and
physical and mental health factors, change in HbA1c was 0.07 percentage points lower among
participants with high purpose compared to participants with low purpose (95% CI, -0.12 to -
0.02; p=0.011). Participants with high purpose had lower odds of developing prediabetes or
type 2 diabetes compared to participants with low purpose (adjusted odds ratio 0.78 [95% CI,
0.62 to 0.98]; p=0.037).

Conclusions: Among older adults, greater levels of purpose in life are associated with a lower
incidence of prediabetes or type 2 diabetes. Strategies to promote greater purpose in life
should be tested as part of type 2 diabetes prevention efforts.

CORRESPONDING AUTHOR: Dina Griauzde, MD, University of Michigan, Ann Arbor, MI, 48105;
[email protected]
S1044 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B085 6:00 PM-7:00 PM

DEVELOPING A CULTURALLY TAILORED DIABETES SELF-MANAGEMENT PROGRAM FOR

CHINESE AMERICANS WITH TYPE 2 DIABETES

Bin Xie, PhD1, Yawen Li, PhD2, Wei-Chin Hwang, PhD3, Zhongzheng Niu, MS, MD1, Xiaomeng
Lei, BS1, Ruizhi Yu, MD1, Jamie Q. Felicitas-Perkins, MPH1, Shabana Masood, MS4, Kelsey
Lawler, MPH5, Diana Zhao, BS6

1
Claremont Graduate University, Claremont, CA; 2San Diego State University, San Diego, CA;
3
Claremont McKenna College, Claremont, CA; 4Claremont Graduate University, Clarmeont,
CA; 5Scripps College, Claremont, CA; 6Stanford University, Stanford, CA

Background/Aims: Despite having lower age- and sex-adjusted body mass index, Chinese
Americans (CA) evidence higher risk for developing Type 2 diabetes than non-Hispanic Whites.
Unfortunately, few attempts have been made to culturally adapt evidence-based diabetes
interventions to this at risk group. The goal of this study is to develop a culturally adapt
evidence-based intervention program for CA patients with Type 2 diabetes. A series of
formative research under an integrative top-down and bottom-up community-participatory
approach was conducted to identify areas for cultural adaptation.

Methods: A systematic review of diabetes self-management programs was conducted to

evaluate strengths and weaknesses of extant evidence-based interventions. Five focus groups
were conducted in the San Gabriel Valley region of Southern California with 23 CA participants
(14 females and 9 males). Participants also completed a questionnaire regarding their
diabetes self-management. On average, participants had a diabetes history of 18 years, and
they were all first-generation of immigrants with 17 (73.9%) originally from mainland China
and the rest from Taiwan, Hong Kong and Vietnam.

Results: Review of the literature indicated a number of areas where cultural adaptations can
potentially take place such as communications with health professionals, medication
adherence, family support and stress coping, dietary control and physical activity. Descriptive
analyses with the questionnaire survey revealed that the average score was 9.96 out of 14 for
diabetes knowledge, 2.93 out of 6 for distress, 2.61 out of 5 for diabetes quality of life, and
6.73 out of 10 for diabetes self-efficacy. Focus group results showed that most participants
felt depressed when they were first diagnosed with Type 2 diabetes. Problems they
encountered in their management of disease included lack of adequate communications with
physicians or other health practitioners (e.g. clinical dieticians) on diabetes treatment and
control, lack of culturally appropriate health education and materials, lack of access to
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1045

training on how to optimize diabetes self-management. The biggest challenges they faced are
dietary control and glucose monitoring.

Conclusion: Our formative research highlighted needs and directions for developing a
culturally tailored diabetes self-management program for CAs.

CORRESPONDING AUTHOR: Bin Xie, PhD, Claremont Graduate University, Claremont, CA,
91709; [email protected]
S1046 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B086 6:00 PM-7:00 PM

DIABETES-RELATED DISTRESS AND INSUFFICIENT SLEEP FOR PERSONS WITH TYPE 2 DIABETES
MELLITUS EXPERIENCING HEALTH DISPARITIES

Kerry Littlewood, Ph.D., MSW1, Kerry Littlewood, Ph.D., MSW1, Abhishek Pandey, MD2

1
AAJ Research and Evaluation, Seminole, FL; 2University of Arizona, Seminole, FL

Background: Diabetes-related distress, referring to patient concerns about disease

management, support, emotional burden, and access to care, is substantively related to
symptoms of depression, but a separate and distinct construct(Fisher, Skaff, Mullan, Arean,
Mohr, Masharani, Glasgow, & Laurencin, 2007). Similar to depression, distress is significantly
and positively associated with HbA1c(Fisher, Skaff, Mullan, Arean, Glasgow, & Masharani,
2008), and is highly responsive to intervention (Fisher, Hessler, Glasgow, Arean, Masharani,
Naranjo, & Strycker, 2013). Fisher, Gonzalez, & Polonsky (2014) suggest studies clarify
diabetes related distress so that appropriate and targeted patient-centered interventions can
occur. Additionally, insufficient sleep has been associated emotional distress related to
chronic diseases, such as T2DM (Liu, Croft, Wheaton, Perry, Chapman, Strine, McKnight-Eily, &
Presley-Cantrell, 2013). Furthermore, few studies have examined insufficient sleep as it
relates to disease-related distress (DD) in persons with T2DM.

Purpose: This study uses data from a Diabetes Self-Management Education Program to
explore the relationship between insufficient sleep and DD. Descriptive, anovas and logistic
regression are used to examine these relationships.

Results: Ninety-seven persons with T2DM (mean HbA1c = 8.79 2.42) participating in a DSME
Program at a southern urban community non profit hospital enrolled in this study. The
majority of these patients were obese (mean BMI=38.568.20). Only 11% reported normal
sleep, with 41% reporting short sleep ( < 6 hours). 28% (n=22) reported distress meeting
clinical thresholds. 73% (n=16) were also short sleepers [x2(1, N=78)= 8.71, p=.003]. A logistic
regression was performed to ascertain the effects of short sleep on the likelihood that
participants would have clinical distress. The regression model was statistically significant,
x2(1)=1.56, p < .05

Conclusions: Weve found that people with T2DM who have insufficient sleep are five times
more likely to experience clinical levels of distress. Interventions promoting healthy sleep may
improve distress and ultimately improve disease management for those in need. Future
research could more closely examine the relationship between sleep and distress.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1047

CORRESPONDING AUTHOR: Kerry Littlewood, Ph.D., MSW, AAJ Research and Evaluation,
Seminole, FL, 33774; [email protected]
S1048 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B087 6:00 PM-7:00 PM

DISORDERED EATING, NEGATIVE EMOTIONS, AND TYPE 1 DIABETES MANAGEMENT DURING

LATE ADOLESCENCE

Ashley C. Baker, B.S.1, Deborah Wiebe, MPH, PhD2, Eunjin Lee, Ph.D.3, Michelle Litchman,
PhD, FNP-BC, FAANP3, Cynthia A. Berg, Ph.D.3

1
University of California Merced, Merced, CA; 2University of California, Merced, Merced, CA;
3
University of Utah, Salt Lake City, UT

Late adolescence is a difficult time for those with type 1 diabetes (T1D), in part due to
increased risk behaviors such as disordered eating (e.g. binge eating and insulin restriction).
Disordered eating behaviors are more common among those who have high negative affect
(i.e. depressive symptoms), which itself is associated with poorer glycemic control. These
associations have generally been studied as a function of individual differences in negative
affect and disordered eating rather than considering within-person daily processes. The
present study examined whether negative affect was associated with individual differences
and within-person day-to-day changes in disordered eating, and whether disordered eating
was associated with glycemic control independent of negative affect. Adolescents recruited
during their senior year in high school (N = 219; 59.6% female; 77% non-Latino Caucasian, 12%
Latino, 5% African American) completed surveys evaluating their depression symptomology
and insulin restriction with the intent to control weight; glycemic control was measured with
HbA1c assay kits. Participants then completed a 2-week daily diary indicating daily binge-
eating and daily blood glucose levels. At the individual difference (between-subjects) level,
adolescents who restricted insulin to lose weight had higher depressive symptoms (r=.36), and
poorer HbA1c (r = .15, ps < .05). Adolescents who restricted insulin also reported higher daily
depressive symptoms (r=.23) and higher daily binge eating (r=.27) on average over the
subsequent two weeks (ps < .05). Regression analyses revealed that associations of insulin
restriction with HbA1c was marginally significant when depressive symptoms were covaried
(=.18, p=.07). At the daily (within-subjects) level, adolescents were more likely to report
binge eating on days that they experienced heightened depressive symptoms, and this
association occurred independent of a tendency to restrict insulin (coefficient (SE) = .30 (.15),
p = .04). Finally, at the daily level, binge eating was associated with higher blood glucose levels
independent of depressive symptoms that day, coefficient = 17.46 (SE = 6.98), p = .013.
Findings demonstrate that risky disordered eating behaviors are more common in the context
of negative affect, and that there are both individual and daily contextual factors that may be
targeted to promote better diabetes management in this vulnerable population.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1049

CORRESPONDING AUTHOR: Ashley C. Baker, B.S., University of California Merced, Merced, CA,
95340; [email protected]
S1050 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B088 6:00 PM-7:00 PM

ECOLOGICAL MOMENTARY ASSESSMENT TO MEASURE MOTIVATORS AND BARRIERS TO

BLOOD GLUCOSE MONITORING IN YOUTH WITH TYPE 1 DIABETES

Jennifer L. Warnick, BA, Sarah Westen, MS, David Janicke, PhD, ABPP

University of Florida, Gainesville, FL

Background: Type 1 diabetes (T1D) is the second most common pediatric chronic illness;
adequate disease management requires daily monitoring of childrens blood glucose (BG)
levels. BG levels are often poorly controlled during adolescence, a time when youth are
becoming more autonomous. Ecological momentary assessment (EMA) is the repeated
sampling of behaviors and experiences in real time in the natural environment. The current
study used EMA via text-message delivered surveys to evaluate the immediate motivators and
barriers to BG monitoring in adolescents to better understand T1D treatment adherence in
youth.

Methods: Adolescents between the ages of 11-21 with T1D were recruited during routine
outpatient endocrinology appointments at the University of Florida Shands Hospital.
Adolescents received 3 text messages per day for 10 consecutive days with a link to a brief
survey which asked if the adolescent checked his/her BG the last time he/she ate, and why or
why not. Descriptive and frequency analyses were conducted to assess the barriers and
motivators to monitoring BG.

Results: 36 adolescents (mean age(SD)=16.6(2.7) years, 20 males, 78% Caucasian) with T1D
answered a total of 592 surveys over a 10-day period (53.62% surveys answered). Of those
who responded to the surveys, 71.5% indicated they had checked their blood glucose the last
time they ate (10.3%=did not; 18.2%=hadnt eaten yet). The greatest motivator for checking
BG was acknowledging it was important for health (97.1%), followed by being afraid of
highs/lows (31.7%), and not wanting to upset doctors (26.9%) or parents (21.4%). Barriers to
checking BG included forgetting (55%), ignoring (33.3%), not having a meter with them (30%),
thinking it was not necessary (8.3%), and not wanting to check in front of others (6.7%).

Conclusion: This study provides preliminary evidence on the barriers and motivators of BG
monitoring adherence. Given the greatest barrier to BG monitoring in this sample was
forgetfulness, future research may examine the use of mobile health interventions to remind
adolescents with T1D to monitor.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1051

CORRESPONDING AUTHOR: Jennifer L. Warnick, BA, University of Florida, Gainesville, FL,

32610-5285; [email protected]
S1052 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B089 6:00 PM-7:00 PM

EVALUATION OF AN INTEGRATED RESEARCH PRACTICE PARTNERSHIP TO ADDRESS DIABETES

DISPARITIES IN LATINO FAMILIES

Erica G. Soltero, PhD1, Crystal Ramos, BA1, Allison Williams, MSW1, Yolanda Konopken,
Registered Dietician2, Elva Hooker, RD2, Jennifer Mendez, Certified Group Fitness Instructor3,
Heidi Wildy, BA3, Karen Davis, BA3, Addey Rascon, RD3, Matthew Sandoval, M.Ed.3, Valentina
Hernandez, BS4, Scott Leischow, PhD5, Omar Contreras, MPH6, Gabriel Shaibi, PhD7

1
Center for Health Promotion and Disease Prevention, Arizona State University, Phoenix, AZ;
2
St. Vincent de Paul Medical and Dental Clinic, Phoenix, AZ; 3Valley of the Sun YMCA, Phoenix,
AZ; 4Mountain Park Health Center, Phoenix, AZ; 5Mayo Clinic, Scottsdale, AZ; 6Arizona
Department of Health Services, Phoenix, AZ; 7Arizona State University, Phoenix, AZ

Multi-sector, participatory approaches are needed to address the complex challenge of

preventing type 2 diabetes and reducing health disparities in high-risk populations. Integrated
research practice partnerships bring academic, community, and clinical partners together to
develop and implement innovative strategies to accelerate the translation of research to
practice and to improve the health of communities. This study aimed to evaluate the impact
of one such partnership on building capacity for developing, implementing and assessing a
community-based diabetes prevention program for Latino families.

The collaboration included a Federally Qualified Health Center, a family wellness program, a
YMCA, and an academic research institution. The partnership was created to foster resource
sharing, bidirectional exchanges of knowledge and skills, communication, and capacity
building to deliver a diabetes prevention program to Latino families. Administrators and staff
(N=12) from each institution completed the Community Impact of Research Oriented
Partnerships tool, which uses a 7 point Likert scale (7=great impact-1=no impact) to assess
perceptions of the personal and organizational impact of the partnership.

After the first year of planning and implementation, Partners reported that the partnership
had a moderate to high impact on their personal knowledge development (M= 6.4.2),
personal research skill development (M=4.6.5), organizational access to and use of
information (M= 5.3.2), and organizational development (M= 4.9.4). Areas within the
partnership that had the most impact included the collective ability to develop a shared
mission, partner with a medical home, and improve capacity to engage in research. Areas of
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1053

least impact included advocacy, securing additional funding, and engaging a broader range of
collaborators.

These findings demonstrate that partnership activities are impacting personal and
organizational capacity to engage in research. Capacity building has significant implications for
the sustainability and expansion of the diabetes prevention program and will empower
Partners to engage in future research opportunities. Similarly, improving personal and
organizational knowledge through the use of health information can benefit the current
program and improve other services within each institution. Evaluation of collaborative
partnerships is essential to identify strengths within the partnership that can be leveraged and
weaknesses that should be addressed.

CORRESPONDING AUTHOR: Erica G. Soltero, PhD, Center for Health Promotion and Disease
Prevention, Arizona State University, Phoenix, AZ, 85004; [email protected]
S1054 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B090 6:00 PM-7:00 PM

EXPLORING TYPE II DIABETES AND HYPERTENSION: THE LIVED EXPERIENCE

Hollie Pellosmaa, PhD1, Angela Liegey Dougall, PhD2

1
The University of Texas at Arlington, Knoxville, TN; 2The University of Texas at Arlington,
Arlington, TX

BACKGROUND: Helping to reduce the occurrence, and improve the management, of Type II
diabetes and hypertension is a major goal for primary care providers. To achieve these goals,
it is important to elucidate how patients understand multiple factors of their conditions; such
as what caused the condition, its associated symptoms, complications, and treatments. The
overall aim of the current study was to elicit qualitative responses about these factors, in
order to understand illness beliefs among diabetic and hypertensive patients. METHODS:
Older diabetic (N = 27, mean age = 63 years) and hypertensive (N = 50, mean age = 72 years)
adults recruited from the community took part in a 60-minute semi-structured interview,
which covered the entire course of their disease. Four themes associated with each condition
were analyzed using NVivo: (1) causes, (2) complications, (3) symptoms, and (4) treatments.
RESULTS: Participants spent most of the interviews discussing their medical treatments;
diabetics mentioned treatments 277 times, and hypertensive patients 580 times. Subjects
focused on their prescribed medications as well as their diet and exercise regimes (or lack
thereof), and to a lesser extent complementary and alternative medications. Diabetic
participants were more likely to discuss the difficulty of adhering to diet/exercise plans, and
they were less specific than hypertensives in acknowledging types of foods/exercises which
were healthy. Interviews also focused on the complications (diabetic = 97, hypertensive = 269)
and symptoms (diabetic = 115, hypertensive = 195) that the participants experienced or had
heard about. Diabetic participants were more likely to discuss extreme complications and
symptoms which they had experienced firsthand, whereas hypertensive participants often
stated that they were asymptomatic but explained symptoms and complications which they
had heard about from friends and family. A small portion of the interviews focused on the
causes of the conditions (diabetics = 53, hypertensives = 145). CONCLUSIONS: These results
highlight the lived experience of the participants, and reveal that there are differences in the
participants beliefs depending on which condition they had. Overall, participants focused on
their daily actions and consequences of having a chronic illness. This research may lead to
advances in the ways medical care is presented to specific elderly populations, thereby
enhancing the lives of individuals.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1055

CORRESPONDING AUTHOR: Hollie Pellosmaa, PhD, The University of Texas at Arlington,

Knoxville, TN, 37923; [email protected]
S1056 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B091 6:00 PM-7:00 PM

GOAL-SETTING FOR TYPE 1 DIABETES MANAGEMENT

Tammy Stump, PhD1, Deborah Wiebe, MPH, PhD2, Cynthia A. Berg, Ph.D.3, Michelle Litchman,
PhD, FNP-BC, FAANP3

1
Feinberg School of Medicine, Northwestern University, Chicago, IL; 2University of California,
Merced, Merced, CA; 3University of Utah, Salt Lake City, UT

Diabetes management is a complex self-regulatory task, which requires consistent monitoring

and adjustments. Such tasks may be aided by setting domain-relevant goals. However, not all
types of goals may be equally beneficial. According to current research on goal-setting, the
more specific a goal is, the better it will serve as a guide to behavior. For instance, whereas
higher-level goals (e.g., get sugars down) only specify a general target and not the means to
achieve that target, a low-level goal (e.g., test blood sugar before every meal) specifies the
precise behavior to perform, potentially making it more automatic to perform that behavior.
Thus, diabetes-related goals that deal with low-level behaviors, such as performing blood
glucose checks, may be more effective than higher-level goals, such as improving blood
glucose or HbA1c levels.

In the present study, we examined predictors of the domains of goals as well as how these
goals were related to diabetes-related management over the next two weeks among young
adults with type 1 diabetes (N=239). At baseline, participants were asked to set a diabetes-
related goal that they intended to work on over the next two weeks. Two coders categorized
goals as belonging to one of six domains (Blood glucose checks, diet, insulin, blood glucose
levels, HbA1c, Other). A subset of 20% of the goals were double-coded; agreement was 92%.

Baseline predictors of goal content were analyzed in a series of logistic regression models in
which predictors (IQ, HbA1c, self-reported adherence, gender, and time since diagnosis) were
entered simultaneously. Participants exhibited responsiveness to their current weaknesses;
those with lower reported adherence were more likely to set a goal related to blood glucose
testing, and those with higher HbA1c were more likely to set a goal related to their HbA1c
level.

Next, in a series of multilevel models, goal domain was used to predict diabetes-related
management reported in daily diaries (mean blood glucose, self-reported adherence) over the
next two weeks. Overall, compared to the other domains, setting a goal related to blood
glucose testing predicted worse management. However, during the two-week diary period, a
significant decrease in mean daily blood glucose over time was observed only among those
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1057

who set a goal related to blood glucose testing, suggesting that this type of goal, which
contains the specific behavior to be performed, assisted young adults to begin to change their
behavior. Young adults with type 1 diabetes are more likely to benefit from setting realistic,
short-term goals targeting specific behaviors.

CORRESPONDING AUTHOR: Tammy Stump, PhD, Feinberg School of Medicine, Northwestern

University, Chicago, IL, 60613; [email protected]
S1058 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B092 6:00 PM-7:00 PM

INFLUENCE OF NEIGHBORHOOD DISORDER AND LEVEL OF EDUCATION ON DIABETES DISTRESS

AND SELF-CARE

Naomi Kane, M.A.1, Elizabeth A. Walker, PhD, RN2, Clyde Schechter, M.A., M.D.3, Elizabeth
Seng, Ph.D.4, Jeffrey Gonzalez, Ph.D.4

1
Yeshiva University, New York, NY; 2Albert Einstein College of Medicine, lrchmont, NY; 3Albert
Einstein College of Medicine, Cypress, CA; 4Yeshiva University, Bronx, NY

Type 2 diabetes (T2DM) affects a greater proportion of socially and economically

disadvantaged individuals. The conditions in which individuals live may have important
consequences for diabetes self-care. This study examined the relationships between
indicators of socioeconomic status, neighborhood disorder, and diabetes distress alongside
diabetes self-care behaviors.

N= 941 adults with T2DM in suboptimal glycemic control were recruited for a larger
telephonic intervention study in the South Bronx. N= 480 completed self-report measures
assessing neighborhood disorder, diabetes distress, and diabetes self-care behavior.
Regression examined main effects of individual and neighborhood socioeconomic factors on
health behaviors; and moderation effects of individual and neighborhood level disadvantage
on the association between diabetes distress and health behaviors.

Participants (N= 480) were predominately middle-aged adults (mean[SD]=56.4[11.7] years

old), female (67%), Latino (68%), Black (28%), had incomes below $20,000 (77%), and
completed the 11th grade or less (50%). Diabetes distress demonstrated stronger and more
consistent relationships with self-care behaviors (R2= .11, p< .001 for diet, R2= .08, p< .001 for
medication) than neighborhood disorder (R2= .02, p= .131 for diet, R2= .02, p= .021 for
medication). Neighborhood disorder was independently associated with diabetes distress (B=
.04, p< .001). The relationship between diabetes distress and both diet and exercise was
significantly stronger among individuals with higher educational attainment (ps< .001). On
the other hand, the relationship between diabetes distress and exercise was significantly
stronger at lower levels of neighborhood disorder (p< .001).

Diabetes distress, unlike neighborhood disorder, was consistently associated with health
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1059

behaviors. Moderation results suggest the strength of the relationship between distress and
self-care behavior is influenced by level of education and perceptions of neighborhood
disorder. At lower levels of education and higher levels of neighborhood disorder, other
factors besides diabetes distress may play a greater role in influencing diabetes self-care. It
may be important for health professionals to factor in the neighborhood context when
evaluating and addressing diabetes-specific distress.

CORRESPONDING AUTHOR: Naomi Kane, M.A., Yeshiva University, New York, NY, 10003;
[email protected]
S1060 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B093 6:00 PM-7:00 PM

INFORMATION, MOTIVATION, & BEHAVIORAL SKILLS BARRIERS TO ADHERENCE AMONG

ADULTS WITH DIABETES IN SAFETY-NET CLINICS

Lindsay S. Mayberry, MS, PhD1, Lyndsay A. Nelson, PhD1, Chandra Osborn, PhD, MPH2

1
Vanderbilt University Medical Center, Nashville, TN; 2One Drop, Nashville, TN

Nonadherence to diabetes medication is common and associated with suboptimal glycemic

control and a cascade of health problems. According to the Information-Motivation-
Behavioral Skills (IMB) model of diabetes medication adherence, adherence to diabetes
medications is determined by adherence-related information, personal and social motivation,
and behavioral skills. This model has explained over 40% of adherence behavior among adults
with type 2 diabetes (T2D) and low socioeconomic status using validated measures of the IMB
constructs. However, barriers to diabetes medication adherence have not been identified
using the IMB model as a guiding framework.

We conducted a thorough review of published studies reporting medication adherence

barriers among adults with T2D and identified 36 unique barriers (31 general and 5 insulin-
specific). We mapped each barrier onto Information (IN), Personal Motivation (PM), Social
Motivation (SM), or Behavioral Skills (BS). Adults with T2D (N=102) receiving care at Federally
Qualified Health Centers in Nashville, TN rated how much each barrier got in the way of taking
their diabetes medicine(s) (1=not at all to 10=a lot). For each IMB model construct, we
identified the most frequently endorsed (scored>1) barriers and averaged scores to examine
bivariate relationships with participant characteristics (i.e., age, gender, race/ethnicity,
education, income, health literacy [Brief Health Literacy Screen, BHLS], diabetes duration,
insulin status, number of diabetes medications), medication adherence (Adherence to Refills
and Medications Scale for Diabetes), and glycemic control (A1c) using nonparametric tests
(Spearmans /Mann-Whitney U).

Participants endorsed BS and PM barriers most frequently and ranked IN and PM as creating
the largest (highest scored) barriers. Younger participants reported more IN (=-.21), PM (=-
.25) and SM (=-.21) barriers (all pppppp

Identification of specific IMB barriers can improve patient education and by informing
intervention content and priorities. Among adults with T2D and low socioeconomic status,
adults IMB barriers were associated with younger age, limited health literacy, and worse
medication adherence and A1c. These findings, combined with prior validation of the IMB
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1061

model of diabetes medication adherence, suggest targeting these barriers will improve
adherence and, in turn, glycemic control.

CORRESPONDING AUTHOR: Lindsay S. Mayberry, MS, PhD, Vanderbilt University Medical

Center, Nashville, TN, 37203; [email protected]
S1062 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B094 6:00 PM-7:00 PM

MOTIVATING DIABETES SELF-CARE: PATIENTS PERSPECTIVES ON SOCIAL AND GOAL-BASED

COMPARISONS OF DIABETES HEALTH STATUS

William Martinez, MD, MS1, Kenneth A. Wallston, PhD2, Gerald Hickson, MD1

1
Vanderbilt University Medical Center, Nashville, TN; 2Vanderbilt University, Pisgah Forest, NC

Background
Social comparisons, self-evaluation in comparison to others, can influence motivation and
behavior. However, little is known about patients perspectives toward social comparison
information regarding their diabetes health status. This study aims to understand patients
perspectives on social-comparison information regarding their diabetes health status and how
receiving such information might compare to goal-based comparisons (i.e., self-evaluation in
comparison to a set standard such as hemoglobin A1C < 7).

Methods
We conducted semi-structured interviews with 25 patients with type 2 diabetes regarding
social comparisons involving their diabetes health status (e.g., self-evaluation of their own
A1C in comparison to others A1Cs). Using thematic content analysis, 2 independent
researchers coded interview transcripts, reached consensus, and identified themes.

Results
Participants were 48% female, 68% white, 52% college graduates, mean age was 65 years
(range: 45 81), mean A1C was 7.5% (range: 5.6 - 12.6%). Participants reported comparing
their diabetes health status to others they knew with diabetes to varying degrees. Participants
who compared themselves to others commonly reported comparing to others who were
socio-demographically similar (e.g., same race/ethnicity, age, gender, neighborhood), and that
these similarities made the comparisons more relevant to their self-evaluation. Themes
included desire for information about ones diabetes health status compared to similar
others, belief that this information would be motivating either to improve (if doing worse than
most others) or maintain their self-care behaviors (if doing better than most others), and that
this information was more personalized and motivating than comparisons to a standard
goal (e.g., A1C < 7). If faced with social comparison information that indicated that their
diabetes health status was worse than the majority of similar others, eleven participants
described a desire for information about what the others were doing differently (e.g., taking a
different medication, eating differently), and said they would use this information to improve
their own diabetes health status.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1063

Conclusions
Providing patients with social comparison information about their diabetes health status may
increase motivation for self-care and promote information seeking among some patients.

CORRESPONDING AUTHOR: William Martinez, MD, MS, Vanderbilt University Medical Center,
Nashville, TN, 37212; [email protected]
S1064 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B095 6:00 PM-7:00 PM

PROGRAM ACTIVE II: TREATING MAJOR DEPRESSION IN T2DM

Mary de Groot, PhD1, Guy Hornsby, Ph.D.2, Chandan K. Saha, PhD1, Yegan Pillay, Ph.D.3, Karen
Fitzpatrick, MD2, Kieren Mather, MD1, Jay Shubrook, DO4

1
Indiana University School of Medicine, Indianapolis, IN; 2West Virginia University,
Morgantown, WV; 3Ohio University, Athens, OH; 4Touro University California, Vallejo, CA

Program ACTIVE II: Treating Major Depression in T2DM

Objective: Program ACTIVE II (R18DK092765) is a multi-state comparative effectiveness study

of cognitive behavioral therapy (CBT) and community-based exercise on depression and
glycemic outcomes in adults with major depressive disorder (MDD) and type 2 diabetes
(T2DM).

Research Design and Methods: The study utilized a 2X2 factorial repeated measures RCT
design in which participants were randomized to: CBT (10 individual sessions), EXER (12 weeks
of community-based exercise), CBT+EXER (10 individual CBT sessions and 12 weeks of
concurrent exercise) or usual care (UC). Participants were assessed at baseline and post-
intervention (POST). Participants were recruited from Indiana, Ohio, and West Virginia using a
community-engaged research (CEnR) approach. Community partners in mental health and
exercise provided interventions to participants.

Results: The sample (N=140) was mean age 56 years (SD=10.7), 77% female, 71% White, 52%
married and 34% completed high school or trade school. At POST, participants assigned to
CBT, EXER, or CBT+EXER reported fewer depressive symptoms (p < .01), improved automatic
thoughts (p < .04), improved physical quality of life (QOL; p=.01), and improved diabetes
distress (p < .01). Participants in EXER and CBT+EXER reported improved diabetes-specific QOL
(p < .01) compared to UC. Pairwise comparisons showed the odds of achieving at least partial
remission from MDD for CBT and EXER groups were 12.6 and 5.6 times the odds for UC,
respectively (p < .03). Odds of full remission for CBT, EXER, and CBT+EXER were 5.3 to 6.5
times the odds for the UC group (p < .01) as assessed by psychiatric interview. For participants
with a baseline HbA1c7.0, EXER resulted in a clinically meaningful 0.7% improvement in
HbA1c (p < .03).

Conclusions: Program ACTIVE is a set of manualized interventions that has demonstrated

clinically meaningful improvements in depression and glycemic outcomes in adults with
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1065

T2DM. These interventions are beneficial to behavioral health and exercise professionals to
increase access to depression treatment for patients with T2DM.

CORRESPONDING AUTHOR: Mary de Groot, PhD, Indiana University School of Medicine,

Indianapolis, IN, 46202; [email protected]
S1066 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B096 6:00 PM-7:00 PM

PSYCHOLOGICAL STRESS, SOCIAL SUPPORT, AND GLUCOSE REGULATION IN A LOW-INCOME

MIGRANT COMMUNITY ON THE U.S.-MEXICO BORDER

Daniela G. Vital, B.S.1, Selena Lopez2, Jessica McCurley, M.S.3, Linda C. Gallo, PhD4

1
University of California, San Diego, San Diego, CA; 2UC San Diego, San Diego, CA; 3SDSU/UC
San Diego, San Diego, CA; 4San Diego State University, Chula Vista, CA

The global increase in type 2 diabetes (T2DM) prevalence has become an urgent public health
concern. Migrants and other low-income individuals living near the U.S-Mexico border are at
elevated risk for T2DM due to poor access to medical care, disrupted employment,
psychological stress, and lack of social support. Prior research has shown a relationship
between psychological stress and the development of glucose dysregulation and T2DM. Social
support is associated with better health and may influence the relationship between stress
and glucose regulation. This cross-sectional study explored the association of self-reported
psychological stress with clinically measured glucose regulation in a convenience sample of
individuals seeking free medical care in Tijuana, Mexico in 2016, and explored whether social
support moderated this relationship. Participants were 165 Hispanic/Latino adults >18 years
old who spoke Spanish or English. Stress was measured using the 4-item Perceived Stress
Scale (PSS); social support was assessed via the 12-item Interpersonal Support Evaluation List
(ISEL). Glucose regulation was determined by point-of-care hemoglobin A1c (HbA1c)
percentage. Participant mean age was 46.9 years (SD = 11.7) and 76.4% were male. The
majority (74.5%) reported history of migration to the U.S., 78.9% reported at least one
experience of deportation out of the U.S., and 35.8% reported intent to migrate to the U.S. in
the next year. Mean HbA1c percentage was 6.1% (SD = 1.8), which falls in the range of
prediabetes. Mean PSS score was 7.1 (SD = 3.0, range = 0-16); mean ISEL-12 score was 19.5
(SD = 6.8, range = 0-36). Stress was significantly associated with glucose regulation (r = .24, p <
0.01). Adjusting for age, sex, and education, individuals who reported higher stress had
significantly higher (poorer) HbA1c values ( = 0.14, p < .01). This relationship remained
significant after controlling for previous diabetes diagnosis. Social support was negatively
associated with stress (r = -.42, p < .01), but not associated with glucose regulation, and did
not moderate the relationship between stress and glucose regulation. Psychological stress is a
prevalent and potentially modifiable determinant of cardiometabolic risk in low-income
individuals and migrants living along the U.S.-Mexico border. Interventions for stress
reduction and management may reduce risk for diabetes and decrease individual and public
heath costs in both the U.S. and Mexico.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1067

CORRESPONDING AUTHOR: Daniela G. Vital, B.S., University of California, San Diego, San
Diego, CA, 92117; [email protected]
S1068 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B097 6:00 PM-7:00 PM

RACIAL DIFFERENCES IN PREVALENCE OF DIABETES AND DIABETES SELF-MANAGEMENT USING

CALIFORNIA HEALTH INTERVIEW SURVEY (CHIS)

Tashi Lhamo, NA, Hee-Soon Juon, PhD

Thomas Jefferson University, Philadelphia, PA

Background: Diabetes affected 9.3% of the U.S population in 2012 and the prevalence
continues to rise rapidly. Minority groups are affected by diabetes significantly at a greater
rate compared to White Americans. Despite having lower BMI, Asian Americans are more
likely to have diabetes compared to White. Among Asians, Asian Indians and Filipinos are
found to have the highest prevalence of diabetes. Scholars have stated that 95% of the
diabetes care is self-care. Therefore, the integral part of effectively managing diabetes is to
practice good diet, physical activity, blood sugar monitoring, medication adherence and self-
efficacy. Although diabetes management practices among different ethnic groups have been
studied, little is known about diabetes management among Asian.

Methods: We used data from 2009 CHIS to study racial differences in the prevalence of
diabetes and diabetes self-management. We performed a univariate analysis to describe the
prevalence of diabetes among racial and ethnic groups. A multivariate logistic regression
model (including age, gender, level of education, insurance, BMI, poverty level and regularity
of exercise) was conducted with weighted analyses using svy from STATA version 13.

Results: Among 47,614 California adults, the prevalence of diabetes was 8.46% (95% CI 0.08-
0.10). American Indians/Alaska Natives (26%, 95% CI 0.16-0.40) had the highest prevalence of
diabetes among all race groups. White Americans had the lowest (6.3%, 95% CI 0.06-0.07)
prevalence of diabetes followed by Asian Americans (7.75%, 95% CI 0.06-0.10). However, the
disaggregation of Asian data indicates that Filipinos (12.4%, 95% CI 1.23-3.57) are significantly
more likely to have diabetes compared to White. In multivariate analysis, being a male, older,
having lower education, health insurance, low poverty level, being obese, and not exercising
were associated with having diabetes. Among ethnic groups, Koreans are worst at self-
managing diabetes.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1069

Conclusion: Finding indicates that diabetes prevalence is increasing rapidly in Asians

compared to Whites. It also highlights the importance of disaggregating data to understand
diabetes prevalence among racial and ethnic groups. The difference in diabetes management
demonstrates the importance of culturally integrated interventions to target these
populations.

CORRESPONDING AUTHOR: Tashi Lhamo, NA, Thomas Jefferson University, Philadelphia, PA,
19104; [email protected]
S1070 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B098 6:00 PM-7:00 PM

REPRODUCTIVE HEALTH AWARENESS AND COMMUNICATION IN DIABETIC ADOLESCENT

LATINAS AND THEIR MOTHERS

Frances Peterson-Burch, RN BSN BA1, Denise Charron-Prochownik, PHD, RN, CPNP, FAAN1,
Keirsten Montgomery, FNP-C2, Dara H. Sorkin, PhD3, Hiba Abujaradeh, MSN, RN1, Patricia
Schmitt, MPM1, Ellen Olshansky, PhD, RN, WHNP-BC, FAAN4

1
University of Pittsburgh, Pittsburgh, PA; 2CHOC Children's, Orange, CA; 3University of
California Irvine, Irvine, CA; 4University of Southern California, Los Angeles, CA

Funding: Eli Lilly USA H60-US-X002

Introduction: Diabetic adolescent Latinas are at increased risk for reproductive health (RH)
complications, including poor maternal/child outcomes. Current recommendations require
that preconception counseling (PC) start at puberty.

Purpose: To describe understanding of diabetes, sex, family planning, pregnancy, and PC

among diabetic young Latinas (13-21 yrs.) and their mothers and to explore cultural
modifications of a validated PC program for diabetic teens (READY-Girls).

Methods: Qualitative, descriptive study; used open-ended questions and markups of the
READY-Girls book. Latina mothers and daughters reviewed pages from the book and wrote
comments on pages. Analysis was guided by qualitative descriptive methods. Responses to
questions and markups were transcribed for theme analysis. Three members of the research
team discussed themes and consensus was reached.

Sample: 19 daughters and 13 mothers were recruited from a childrens hospital. Most
daughters (n=11) were 16-18 yrs.; 11 had type 2 diabetes (T2D) and 8 T1D.

Results: Major themes: 1) Lack of Latina mother-daughter (M-D) communication about sex
and pregnancy (e.g., most M-Ds had not talked about sexuality or pregnancy; mothers want
daughters to initiate conversations but daughters may be uncomfortable), 2) importance of
diabetes management (most M-Ds reported healthcare providers (HCPs) discussed diabetes
but did not provide PC or RH information), 3) diabetes complications on pregnancy (e.g.,
infant mortality, maternal death, infertility), 4) importance of safe sex, while difficult to
discuss (e.g., both M-Ds expressed that talking about RH was important but awkward,
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1071

embarrassing; mothers were concerned about teen pregnancy; daughters stated that teen
sex/pregnancy is seen as disappointing; most M-Ds stated their culture encourages delay of
sex/pregnancy until marriage), and 5) lack of communication between girls HCPs about RH
(e.g. only 5 daughters and 1 mother reported having heard of PC).

Conclusions: While diabetes information was provided by HCPs, deficits were reported in RH.
Many M-Ds do not appear to be discussing RH amongst themselves. These data can help to
modify a culturally-relevant PC program for young Latinas and support further research.

CORRESPONDING AUTHOR: Frances Peterson-Burch, RN BSN BA, University of Pittsburgh,

Pittsburgh, PA, 15261; [email protected]
S1072 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B099 6:00 PM-7:00 PM

SELF- AND SOCIAL REGULATION IN EMERGING ADULTS WITH TYPE 1 DIABETES

Ascher Munion, BS in Psychology1, Jonathan Butner, Ph.D.2, Deborah Wiebe, MPH, PhD3,
Caitlin Kelly, M.A.1, Sara L. Turner, M.S.1, Cynthia Munion, PhD1

1
University of Utah, Salt Lake City, UT; 2University of UTah, Salt Lake City, UT; 3University of
California, Merced, Merced, CA

Self-management of type 1 diabetes (T1D) poses significant regulatory challenges where

patients must simultaneously engage in self-regulation (regulating their thoughts, behaviors,
and emotions) together with regulating their social context to facilitate optimal diabetes
management and outcomes. Previous dynamic systems modeling with this sample during late
adolescence revealed multiple facets of self-regulation were coordinated together and that
facets of social regulation (disclosing to parents so that parents are knowledgeable and can
help) were coordinated separately for mothers and fathers. Further, mothers helpfulness
specifically served to support adolescents self-regulatory efforts. Due to normative
developmental changes in relationships during late adolescence, with parents potentially
becoming less involved, it is important to examine how self- and social regulation function in
diabetes management in the face of these changes. The goal of the study was to investigate
whether this same self- and social-regulatory structure characterized the self- and social
regulation of these individuals one year later as these emerging adults managed their
diabetes in the context of friend relationships. Emerging adults with T1D (N=203, Mage = 18.7)
reported for both mothers and fathers - how much their parents knew about their diabetes
management, whether they disclosed to their parents, how helpful their parents and friends
were for their diabetes management, as well as multiple facets of their own self-regulation
surrounding diabetes care (negative affect, self-regulatory failures, and adherence behaviors)
for 14 days in an online diary. Using a dynamical systems approach, a multi-level latent
coordination model was used to first evaluate the coordinative structure of self- and social
regulation and then identify which of the variables drove the changes in other variables in the
diabetes management system. Consistent with patterns observed in late adolescents, the
best-fitting model included separate factors for social regulation involving mother and father
and for self-regulation. However, a shift emerged in the relations between these coordinated
structures from the previous time point. Mothers no longer drove day-to-day changes in
regulation, and fathers were much more distally connected. Neither mothers nor fathers
significantly influenced changes in daily self-regulation surrounding diabetes. The results
suggest greater independence between self and social regulation in the diabetes
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1073

management system during emerging adulthood than during adolescence, which is consistent
with broader normative changes in the emerging adults relationships with their parents.

CORRESPONDING AUTHOR: Ascher Munion, BS in Psychology, University of Utah, Salt Lake

City, UT, 84102; [email protected]
S1074 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B100 6:00 PM-7:00 PM

SELF-CONTROL IS ASSOCIATED WITH ADHERENCE TO CERTAIN SELF-CARE BEHAVIORS AMONG

ADULTS WITH TYPE 2 DIABETES

Lyndsay A. Nelson, PhD1, Kenneth A. Wallston, PhD2, Chandra Y. Osborn, PhD, MPH3, Lindsay
S. Mayberry, MS, PhD1

1
Vanderbilt University Medical Center, Nashville, TN; 2Vanderbilt University, Pisgah Forest,
NC; 3Informed Data Systems Inc., Nashville, TN

The initiation and maintenance of diabetes self-care behaviors (i.e., taking medications, eating
healthily, exercising, and self-monitoring blood glucose) are critical for achieving optimal
glycemic control and preventing diabetes complications. However, rates of self-care
adherence remain low among adults with type 2 diabetes (T2D). Self-control is the ability to
alter ones emotions, desires, and behaviors to meet long-term goals, and may be a
determinant of self-care initiation and maintenance. Associations between self-control and
healthier eating and more exercise have been reported among people without diabetes and
people with type 1 diabetes, but evidence of these relationships among people with T2D is
limited. Therefore, we examined cross-sectional associations between self-control and
adherence to diabetes self-care behaviors using baseline data from a diabetes self-care
promotion intervention.

Adults with T2D who were receiving care at Federally Qualified Health Centers in Nashville, TN
consented, provided patient characteristics (i.e., age, gender, race/ethnicity, education,
income, insulin status, and diabetes duration), completed the Brief Self-Control Scale (BSCS),
Adherence to Refills and Medications Scale for Diabetes (ARMS-D), Personal Diabetes
Questionnaire (PDQ) to assess eating behaviors, the International Physical Activity
Questionnaire (IPAQ), and the Summary of Diabetes Self-Care Activities (SDSCA) blood glucose
monitoring subscale, and received a blood-drawn A1c test. We regressed each self-care
behavior onto self-control, separately, first unadjusted and then adjusted for patient
characteristics and A1c.

Participants (n=64) were 51.510.5 years old; 60% female; 61% non-White; 38% had a high
school degree, 61% had annual incomes < $35K; 53% were on insulin and reported being
diagnosed with T2D for 8.66.5 years with an A1c of 8.72.0. Participants scored 30.35.2 on
the BSCS (possible range 8-40). Higher self-control was associated with healthier eating in
unadjusted (=.38, p < .01) and adjusted ( =.34, p < .05) analyses, and with better
medication adherence in unadjusted (=.28, p < .05), but not adjusted ( =.24, p=.12)
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1075

analyses.

Among adults with T2D, self-control was associated with both better adherence to
medications and healthier eating, but was a stronger determinant of eating than taking
medication. Self-control may be an overlooked reason some patients with T2D adhere to
certain self-care behaviors while others do not. Future studies should examine these
associations longitudinally with larger samples and different self-care measures. Finally,
examining modifiable predictors of self-care (e.g., self-efficacy) as mediators of associations
between self-control and self-care may identify targets for interventions among adults with
low self-control.

CORRESPONDING AUTHOR: Lyndsay A. Nelson, PhD, Vanderbilt University Medical Center,

Nashville, TN, 37203; [email protected]
S1076 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B101 6:00 PM-7:00 PM

STRESSFUL LIFE EVENTS AND GLYCEMIC CONTROL IN YOUNG ADULTS WITH T1D

Kevin L. Joiner, PhD, APRN, ANP-BC1, Robin Whittemore, PhD, APRN, FAAN2, Margaret L.
Holland, PhD, MPH, MS1, Margaret Grey, DrPH, RN, FAAN3

1
Yale University, Orange, CT; 2Yale University, School of Nursing, Orange, CT; 3Yale University,
West Haven, CT

Background: For young adults with type 1 diabetes (T1D), stressful life events may impact
their ability to achieve glycemic goals, possibly by altering their ability to perform self-
management.

Methods: Data from 4,745 participants (age 18 to < 31 years old) in the T1D Exchange Registry
were analysed. Report of a stressful life event was defined as one or more positive responses
on a 15-item stressful life events index created for use in the registry, and defined as a
dichotomous variable (yes-no). Comparisons using Wilcoxon and 2 tests were conducted
between participants who were classified as having a stressful life event vs. those who did
not. Linear regression was used to test associations between stressful life events and A1c
levels. Ordered logistic regression was used to test associations between stressful life events
and frequency of missing insulin doses. Frequency of missing insulin doses was treated as a
categorical variable with 3 values, never/rarely missing a dose ( < 1x/week), occasionally
missing a dose (1-2x/week), or often/freqently missing a dose (>2x/week). Regression models
were adjusted for age, sex, and race/ethnicity.

Results: Nearly half (48.0%) of respondents reported having a stressful life event within the
preceding 12-months. The most frequently reported stressful life events were problems at
work or school (15.8%), financial problems in the family (14.5%), serious arguments with
family members or a close friend (13.9%), and moved to a new home (12.1%). Compared to
the group not reporting stressful life events, those that reported stressful life events were
more likely to be female, have a lower household income, and have a higher level of
educational attainment (all p < 0.001). The stressful life event group had a higher mean A1c
8.51.8% vs 8.21.7% (p < 0.001). Compared to a person not reporting stressful life events, an
equivalent person reporting a stressful life event was more likely to miss insulin doses
>2x/week (13.3% vs 20.0%), and less likely to miss insulin doses < 1x/week (64.5% vs. 54.7%).

Conclusion: These findings may indicate that for young adults with T1D the experience of a
stressful life event may increase their risk of poorer glycemic control, possibly by disrupting
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1077

performance of self-management. Further exploration of these relationships may allow for

the potential for future identifying those at risk and assisting them with more positive
approaches to managing stress.

The source of the data is the T1D Exchange, but the analyses, content and conclusions
presented herein are solely the responsibility of the authors and have not been reviewed or
approved by the T1D Exchange.

CORRESPONDING AUTHOR: Kevin L. Joiner, PhD, APRN, ANP-BC, Yale University, Orange, CT,
06477; [email protected]
S1078 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B102 6:00 PM-7:00 PM

ACCULTURATION IS ASSOCIATED WITH RESTRAINED EATING AMONG LATINO ADOLESCENTS

Christopher Johansen, MPH, Kim Reynolds, PhD

Claremont Graduate University, Claremont, CA

Background: Dietary behavior shifts among adolescents as they acculturate from their
heritage culture to the receiving culture in the United States. Eating habits such as restrained
eating have been understudied in Latino adolescents. As acculturation to the United States
increases, exposure to social norms to maintain or lose weight might lead to an increase in
levels of restrained eating as a means to control diet and improve weight status. The objective
of this study was to explore the influence of acculturation on restrained eating among Latino
adolescents in Southern California. It was hypothesized that greater acculturation would be
positively associated with greater restrained eating.

Methods: Latino adolescents (n = 131) in Grades 9-12, aged 14 to 17 years, at public schools
in Southern California were recruited and completed a single self-report survey using mini
laptops. Acculturation was assessed using Ungers Acculturation, Habits, Interests
Multicultural Scale for Adolescents, 5-item family support for dietary habits scale, and a 10-
item subscale from the Dutch Eating Behavior Questionnaire to measure restrained eating.
Multiple linear regression tested acculturation on restrained eating controlling for age,
gender, BMI, SES and family support for dietary habits.

Results: As acculturation increased, (B=0.254, p < .01) restrained eating increased as

predicted. BMI (B=0.464, p < .01) and family support (B=.228, p < .01) were positively
associated with restrained eating, while SES (B=-.199, p < .05) was negatively associated with
restrained eating.

Conclusions: Acculturation was positively associated with restrained eating among Latino
adolescents. Future research should continue to examine acculturation as a predictor of
restrained eating to delineate if this relationship holds in more ethnically diverse populations.

CORRESPONDING AUTHOR: Christopher Johansen, MPH, Claremont Graduate University,

Claremont, CA, 91711; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1079

B103 6:00 PM-7:00 PM

COMPARING MINDFULNESS-BASED WEIGHT MANAGEMENT TO CURRENT STANDARD

PRACTICES

Teresa Lee, RD, LD, Kelly H. Webber, PhD, RD, LD

University of Kentucky, Lexington, KY

Background: Comprehensive approaches to weight loss that incorporate behavioral

components are needed to address barriers to long-term weight loss maintenance.
Investigating the practice of mindfulness as a behavioral tool to enrich weight loss programs is
an emerging field of research.

Objectives: The purpose of this study was to compare the effectiveness of a 12-week
mindfulness-enriched weight management program to a 12-week standard behavioral weight
loss program on changes in weight.

Methods: This was a two-group randomized experimental design study. Adults (n=53)
between the ages of 25 and 65 with body mass index between 28 and 45 kg/m 2 were
recruited. The pool of participants was randomized into either the standard group or the
mindfulness group, and then further subdivided into two smaller groups of 8 to 13
participants each. Each group met once per week for 60 minute sessions with a registered
dietitian for twelve weeks. Paired t-tests were used to assess changes over time among
participants. Independent t-tests were used to compare changes over time between the two
groups. Pearsons correlation coefficient was used to determine associations between
variables.

Results: 40 people completed the 12-week intervention. The sample was 90% white, 90%
female, 48 11.5 years of age on average, with an average BMI of 33.9 4.3
kg/m2. At 12 weeks, the mean difference in weight loss between the two groups was not
significant (p=.45), with the intervention group losing 8.75 7.8 lbs. and the standard group
losing 6.35 6.8 lbs. Mindful eating scores improved by 14% on average in the mindful group
versus 6.6% in the standard group (p=0.18). Levels of general mindfulness improved by 6.9%
in the mindful group versus 0.8% in the standard group (p=0.326). A change in mindful eating
was correlated with weight loss in women (R=0.444, p=0.008), but not men (R=-0.833,
p=0.167) in the entire sample.

Conclusions: Both groups produced significant weight loss, however, the two groups were not
different at 12 weeks. The impact of change in mindful eating on weight was evident in
S1080 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

women, but not men; this suggests different causal pathways for overeating in women and
men.

CORRESPONDING AUTHOR: Teresa Lee, RD, LD, University of Kentucky, Lexington, KY, 40504;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1081

B104 6:00 PM-7:00 PM

PERCEPTION VS. REALITY: IS IT MORE EXPENSIVE TO GROCERY SHOP ON A

VEGAN/VEGETARIAN DIET?

Jaime Coffino, M.P.H.1, Sydney Heiss, B.A.2, Julia M. Hormes, Ph.D.1

1
University at Albany, SUNY, Albany, NY; 2University at Albany, SUNY, Delmar, NY

Introduction: Adhering to a vegetarian or vegan diet is associated with a range of health

benefits, including lower risk of cardiovascular disease, hypertension, and type II diabetes.
Despite these health benefits, Americans are predominately omnivores. A common argument
against adopting a vegetarian/vegan diet is that it is more expensive than an omnivorous diet.
This study aims to examine if there is an empirical basis for this concern by contrasting
perceptions of cost of vegan and vegetarian diets with the actual amount of money spent
grocery shopping by omnivores, vegetarians, and vegans. Methods: Community members
(n=540, mean Age= 31.12, SD= 12.72, 79.7% female, mean BMI=24.22, SD=5.14) completed a
computer-based battery of measures assessing food preferences and eating behaviors.
Results: Using a paired samples t-test, it was found that omnivores view vegans as richer than
themselves (t(158) = 3.31, p = .001), vegetarians view omnivores as richer than themselves
(t(76) = -2.16, p = .034), and vegans view both omnivores and vegetarians to be richer than
themselves (t(274)= -4.64, p < .001; t(274)= -4.07, p < .001). 52.3% of omnivores (n= 58),
13.3% of vegetarians (n=6), and 5.5% of vegans (n=10) agreed that a vegetarian diet is more
expensive than an omnivorous diet [ = 90.83, p = < .001, =.52]. Similarly, 61.6% of
omnivores (n=98) and 41.8% of vegetarians (n=33) thought a vegan diet is more expensive
than a vegetarian diet [ = 8.42, p = .01, =.19]. Despite these perceptions, there were no
significant group differences in reported amount of money spent on groceries in an average
week (M = 103.86, SD = 78.42 in omnivores vs. M = 110.29, SD = 114.41 in vegetarians vs. M =
103.14, SD = 80.20 in vegans; F(2, 296) =.12, p = .89, = .001). Discussion: Although vegan and
vegetarian diets are often perceived as more expensive, this study provides evidence that
there is no difference in the amount of money spent on groceries per week between
omnivores, vegetarians, and vegans. In fact, many countries around the world eat less meat
because these diets are more affordable. It is important to educate people on the health
benefits associated with adopting a vegan/ vegetarian diet and inform them that the cost
barrier is a mere perceived, as opposed to a real, obstacle of adopting this diet.

CORRESPONDING AUTHOR: Jaime Coffino, M.P.H., University at Albany, SUNY, Albany, NY,
12210; [email protected]
S1082 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B105 6:00 PM-7:00 PM

PROMOTING HEALTHIER CHILDRENS MEALS AT QUICK-SERVICE AND FULL-SERVICE

RESTAURANTS: RESULTS FROM A FEASIBILITY STUDY

Nanette Lopez, PhD1, Stephanie Anzman-Frasca, PhD2, Sara Folta, PhD3, Vanessa Lynskey,
MPH4, Anjali A. Patel, MPH, RD5, Meaghan Glenn

1
University of Southern California, San Diego, CA; 2University at Buffalo, Buffalo, NY; 3Tufts
University, Boston, MA; 4ChildObesity180 at Tufts University, Boston, MA; 5Accents on Health,
Inc. (dba Healthy Dining), San Diego, CA

Background: Energy dense restaurant foods contribute to childhood overweight. Promoting

healthier kids meals in restaurants may lead to increased sales of these items, thus improving
childrens nutrition and demonstrating to industry a demand for these items.

Methods: An 8-week intervention using four strategies (toy incentive, placemats, server
prompts, signage) to promote healthier kids meals was implemented at two locations each of
a quick-service (QSR) and full-service (FSR) restaurant chain. Convenience samples of 1st-4th
grade children and their parents (n=28) were surveyed regarding parent and child perceptions
of intervention components. Restaurants sales of healthier kids meals were examined before
and during the intervention.

Results: Parent and child-reported awareness of intervention components varied, with

greatest awareness of the sign and placemat. At the FSRs, the percentage of monthly kids
meals sales from healthier kids meals was 4.9% at baseline, 8.1% during implementation, and
6.3% at study end. The percentage of monthly sales from healthier kids entrees at the QSRs
increased from 61.7% at baseline to 61.8% during implementation, and ended at 63.1%;
however, the percentage of monthly sales from healthier kids side dishes decreased from
25.0% at baseline to 23.0% at implementation, and ended at 22.9%.

Conclusions: Initial study findings indicate promising strategies for future restaurant
interventions, while also uncovering barriers that can inform future work. Given the mixed
results for parent and child awareness of intervention components, future studies can build
upon these findings and maximize the feasibility, effectiveness, and sustainability of efforts to
promote healthier eating in restaurants.

CORRESPONDING AUTHOR: Nanette Lopez, PhD, University of Southern California, San Diego,
CA, 92106; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1083

B106 6:00 PM-7:00 PM

SEVERITY OF DEPRESSIVE SYMPTOMS AFFECTS WEIGHT LOSS GOALS, EATING PATHOLOGY,

AND DIETING BEHAVIORS IN COLLEGE WOMEN

Leila Azarbad, Ph.D.1, Jacey Keeney, M.S.2

1
North Central College, Naperville, IL; 2Rosalind Franklin University of Medicine & Science,
Marseilles, IL

College women are at particularly high risk for dieting and unrealistic weight loss goals,
disordered eating, and depressive symptoms. A number of studies have explored how
disordered eating behaviors and weight loss goals differ between depressed and non-
depressed college women. However, few studies have investigated whether such behaviors
and goals differ across various degrees of depressive symptomatology. The purpose of this
study was to explore differences in weight loss goals, eating pathology, and dieting behaviors
among college women screened as minimally depressed, moderately depressed, and severely
depressed.

Methods: Participants consisted of 166 undergraduate women who completed the Center for
Epidemiologic Studies Depression Scale (CES-D), Binge Eating Scale, and Three Factor Eating
Questionnaire. Participants also self-reported their dieting histories and self-efficacy, weight
loss goals, and eating behaviors. Participants were divided into three groups using CES-D
cutoff criteria previously demonstrated to correspond with minimal, moderate, and severe
depressive symptomatology. ANOVA and post-hoc testing were used to explore differences
among the three groups.

Results: Although there were no significant differences in BMI across the three depression
groups, minimally depressed women endorsed a higher percentage of desired weight loss
compared to moderately depressed women, F(2,162) = 4.48, p=.013, as well as greater dieting
self-efficacy compared to severely depressed women, F(2,59) = 2.16, p=.043. Significant
differences also emerged with respect to eating pathology, with minimally depressed women
reporting lower levels of eating disinhibition than severely depressed women, F(2,159) = 3.06,
p=.049, as well as lower levels of binge eating compared to both other groups, F(2,157) = 6.65,
p=.002. With respect to dieting behaviors, severely depressed women initiated dieting
behavior at an earlier age compared to moderately depressed women, F(2,74) = 3.24, p=.044,
whereas moderately depressed women endorsed less weight loss via diet and exercise
compared to minimally depressed women, F(2,74) = 3.21, p=.046.

Discussion: Results suggest that severity of depressive symptoms may affect eating behaviors
S1084 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

and weight loss goals differently in college women. Specifically, those with lower levels of
depression may be at the highest risk for unrealistic weight loss goals, whereas those with
higher levels of depression may be at the highest risk for eating pathology. Future studies
should continue to explore how severity of depressive symptoms affects eating behavior and
weight loss in college women, as doing so can allow for enhanced outreach and screening
efforts aimed at this population.

CORRESPONDING AUTHOR: Leila Azarbad, Ph.D., North Central College, Naperville, IL, 60563;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1085

B107 6:00 PM-7:00 PM

THE ROLE OF LATINO CHILDRENS ACCULTURATION ON THEIR MOTHERS DIETARY INTAKE

AND DIETARY BEHAVIORS

Sandra Soto, MPH, BSN1, Elva M. Arredondo, PhD2, Scott Roesch, Ph.D2, Bess H. Marcus,
Ph.D.3, Holly Shakya, PhD4, Guadalupe X. Ayala, PhD, MPH2

1
SDSU/UCSD, San Diego, CA; 2San Diego State University, San Diego, CA; 3University of
California, San Diego, La Jolla, CA; 4UCSD, La Jolla, CO

Purpose. Though theories posit the importance of family-level factors on health behaviors,
the role of children in mothers dietary intake has received little attention. This study
examined the relation between Latino childrens acculturation and mothers dietary
intake/behaviors. We also examined the mother-child acculturation gap to identify the types
of dyads related to mothers lower diet quality. Methods. Baseline surveys were collected
from 361 Latino mother-child dyads enrolled in a family-based dietary intervention. Dyads
resided in Imperial County, CA, located on the US-Mexico border. Children were between 7-13
years old. Daily intake of fruits, vegetables, and sugary beverages, percent of calories from fat,
weekly away-from-home foods, percent of weekly grocery dollars spent on fruits and
vegetables, and summary scores of healthy and unhealthy dietary practices were examined.
Bidimensional acculturation was assessed using continuous measures of Hispanic and non-
Hispanic dimensions and acculturation groups (assimilated, bicultural, traditional, and
marginalized). Three computation methods determined the acculturation gap: a) difference
score in acculturation; b) match/mismatch in acculturation groups; and c) interaction between
mothers and childrens acculturation scores. Separate regression models with child
acculturation or mother-child acculturation gap were tested for each dietary outcome.
Results. Having an assimilated vs. a bicultural child was inversely related to mothers
vegetable intake and positively associated with the likelihood of engaging in unhealthy dietary
practices. The match/mismatch computation method demonstrated the most associations
with unhealthy dietary outcomes (sugary beverage servings, percent of calories from fat,
away-from-home eating, and engaging in unhealthy dietary practices). Overall, traditional
mothers of assimilated children reported lower diet quality than others. Discussion. Results
suggest that assimilated children may be important to their mothers dietary
intake/behaviors. Researchers should examine the potential mechanisms for how and under
which circumstances assimilated children affect their traditional mothers dietary
intake/behaviors.
S1086 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Sandra Soto, MPH, BSN, SDSU/UCSD, San Diego, CA, 92123;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1087

B108 6:00 PM-7:00 PM

THE ROLE OF STRESS AND NEGATIVE AFFECT IN FOOD CONSUMPTION FOR INDIVIDUALS WITH
COMORBID ADDICTION TO FOOD AND ILLICIT SUBSTANCES

Hannah Tuttle, BA1, Nisha C. Gottfredson, PhD2

1
University of North Carolina School of Global Public Health, Chapel Hill, NC; 2UNC Gillings
School of Global Public Health, Chapel Hill, NC

Individuals recovering from a substance use disorder (SUD) tend to experience excessive
weight gain in the first year of recovery (Cowan et al, 2008). There has been little or no
research into the nature of this phenomenon, or its consequences. Koob et als (1997) theory
of spiraling dysregulation suggests that people with food addiction (FA) during the early
stages of recovery from a SUD might be prolonging a cycle of negative reinforcement,
replacing an illicit substance with unhealthy food, in response to negative affect (NA) or daily
stressors. In the long term, continuing the negative reinforcement cycle may lead to
relapse.We examine eating patterns of N=108 people in their first 12 months of recovery from
alcohol and drugs with an emphasis on self-medicating with food (i.e., non-nutritive eating in
response to daily stress or negative affect). We describe who in the sample meets criteria for
FA, which was measured using the modified Yale Food Addiction Scale. We then test the
hypothesis that who meet the threshold for FA will eat more calories and a diet characterized
by a higher glycemic load (GL) on days when they report feeling higher levels of stress and
NA.Participants received brief ecological momentary assessment surveys delivered to their
personal mobile phones up to 4 times per day for 7 consecutive days. These surveys assessed
affect and stress levels in the moment. There were up to 3 of these intensive measurement
bursts, each spaced 6 weeks apart. Additionally, participants received a phone call from a
trained interviewer to conduct a 24-hour dietary recall each day during the bursts. Nutritional
information was coded using NDSR software.Because our outcomes were counts and daily
data were nested within individuals, we applied generalized linear mixed models to assess the
effect of daily stress, daily NA, and the interaction between daily stress and daily NA with FA
on total calorie consumption (controlling for gender and weight) and total GL (controlling for
total calories). People with FA, but not those without, consume more calories on days
characterized by higher levels of stress. Whereas all people tend to consume a higher GL diet
on days characterized by higher levels of NA, this was particularly true for people who met
criteria for FA. We characterize participants with comorbid SUD and FA and those with SUD
only.
S1088 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Hannah Tuttle, BA, University of North Carolina School of Global
Public Health, Chapel Hill, NC, 27514; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1089

B109 6:00 PM-7:00 PM

USAGE AND PERCEPTIONS ASSOCIATED WITH HEALTHFULNESS OF VENDING MACHINES IN

PUBLIC INSTITUTIONS

Teresa Smith, PhD1, Holly Dingman, MS, RD2, Courtney Pinard, PhD1

1
Gretchen Swanson Center for Nutrition, Omaha, NE; 2Childrens Hospital & Medical Center,
Lincoln, NE

Energy-dense snack foods are ubiquitous among vending machines, and though school
settings have become a recent focus for intervention, little work has been done to assess
vending machines at public institutions, worksites, and other key locations such as hospitals.
The purpose of this study was to assess the healthfulness, usage and perception, and
nutritional standards of vending machines in a sample of 13 public institutions, including
hospitals, across a Midwest state. Data were collected using the Nutrition Environment
Measures Survey-Vending (NEMS-V) to assess the healthfulness of foods and beverage
options, a survey to assess employee perceptions and usage of vending machines at work, and
interviews with designated representatives about practices and policies regarding nutritional
standards of vending machines at their institutions. Of the 44 vending machines assessed,
seven (each at different institutions) received a NEMS-V award because at least 30% of foods
or 55% of beverages met nutritional criteria aligning with the 2010 Dietary Guidelines for
Americans. Four institutions reportedly engaged in practices or policies with regard to
purchasing requirements; placement, promotion, pricing policies; and/or comprehensive
nutrient standard(s) for their vending machines, all of which had one award-level vending
machine. Of the 520 employees surveyed, 12% reported purchasing food at least once per
week and 20% reported purchasing a beverage at least once per week from the vending
machine, though 40% reported that they would be more likely to use the vending machine if
there were a greater variety of healthful options. While few (20%) reported that they
perceived most foods and beverages in the vending machines at work to be healthy,
employees of institutions with an award-level vending machines were significantly more likely
to perceive the foods and beverages to be healthy (X2(1, N = 471) = 66.8, p < 0.001). Given the
prevalence of less healthful options available through vending machines, perception of
employees that vending machines are relatively unhealthy, and lack of nutritional standards
with regard to vending machines, these findings underscore a need for implementing
evidence-based nutrition standards for vending machines in public institutions, especially
among institutions where vending may be an important source for meals.
S1090 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Teresa Smith, PhD, Gretchen Swanson Center for Nutrition,
Omaha, NE, 68114; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1091

B110 6:00 PM-7:00 PM

ADVOCACY AND INFLUENCE IN A NEW MEDIA LANDSCAPE: WHAT CAN WE LEARN FROM BIG
MEDIA DATA?

Eric Leas, MPH1, John Ayers, PhD MA2

1
University of California, San Diego, San Diego, CA; 2http://www.johnwayers.com/, Chula
Vista, CA

The strategies that experts have used to share information about social causes have
historically been top-down, meaning the most influential messages are believed to come from
planned events and campaigns. However, more people are independently engaging with
social causes today than ever before, in part because online platforms allow them to
instantaneously seek, create, and share information. Big data analytics make it possible to
rapidly detect public engagement with social causes by analyzing the same platforms from
which organic advocacy spreads. To demonstrate these claim we evaluated several high-
profile cases covering a range of subjects, including Leonardo DiCaprios 2016 Oscar
acceptance speech citing climate change, Charlie Sheens HIV status disclosure, Brazilian
President Lulas throat cancer disclosure which he attributed to smoking, and Awareness Days
such as the Greater American Smokeout and World No Tobacco Day. Using news trends
(Bloomberg Terminal news archives), social media (Twitter postings) and information-seeking
(Google searches and Wikipedia page views) about the related health issues we can assess the
impact of these events. Because these data are freely available in real-time our analytical
strategy provides substantial lead time for experts to detect and participate in organic
advocacy while an issue is salient. By amending our communication frameworks, building
infrastructure for monitoring, and reacting to organic advocacy in near real time, we open
novel possibilities for communication advocacy and research. we believe that ignoring organic
advocacy and bottom-up communication approaches imperils the effectiveness of media
advocacy and, ultimately, public health. The scientific community must adapt to the 21st
century dynamic communication landscape and ready itself for the next opportunity to
harness the agents of change.

CORRESPONDING AUTHOR: Eric Leas, MPH, University of California, San Diego, San Diego, CA,
92103; [email protected]
S1092 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B111 6:00 PM-7:00 PM

DESIGN CONSIDERATIONS FOR MHEALTH TOOLS FOR SMOKERS NOT READY TO QUIT

Jennifer McClure, PhD1, Predrag Klasnja, PhD1, Jaimee Heffner, PhD2, Sheryl Catz, PhD3

1
Group Health Research Institute, Seattle, WA; 2Fred Hutchinson Cancer Research Center,
Seattle, WA; 3University of California - Davis, Sacramento, CA

Most smoking interventionsincluding appsare designed for smokers who are ready to
quit. In contrast, most smokers want to quit someday, but are not yet ready, so existing
cessation apps do not fully address their needs. If apps were designed for smokers not ready
to quit, they could encourage and assist more people in quitting. But little is known about how
to design these programs to be appealing and effective. To address this gap, we surveyed
smokers online from across the U.S. (n=116) to assess whether they 1) were interested in
mHealth tools to help them decide if, when, or how to quit smoking and 2) their preferred app
features and functionality. Participants were adults who wanted to quit smoking someday,
but were not yet ready to quit, and owned a smartphone. The sample was 72% female, 7%
Hispanic, 69% white, and 41% only had a high school degree. Only 3% had downloaded a
cessation app, but most were interested in an app to help them reduce smoking (88%) or
decide if, when or how to quit (91%). Participants rated the importance and appeal of 35
specific app features, functions, and topical content. We also assessed the maximum they
would be willing to pay for this app and what reputational metrics they thought were most
important (e.g., app source, recommendation from doctor, user ratings). Features viewed as
very important included: password protection, a trusted source, and that the app does not
access other personal information. Functions rated as very appealing included: ability to
track smoking and spending on cigarettes, earn points/rewards, and redeem these for gifts.
Content rated as very appealing included guidance how to quit and help managing
withdrawal. Participants generally were not interested in video chat with experts or peers,
sharing their progress with family/friends, or connecting via social media. These and other
findings will be discussed. Overall, smokers not ready to quit are interested in smoking-
focused apps, but their design and marketing may need specific tailoring. Results will be
contrasted against findings from similar research with smokers who are ready to quit. Results
are relevant to addiction treatment experts, design experts, and software developers
interested in mHealth.

CORRESPONDING AUTHOR: Jennifer McClure, PhD, Group Health Research Institute, Seattle,
WA, 98101; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1093

B112 6:00 PM-7:00 PM

DEVELOPMENT OF A MOBILE APP TO ENGAGE AND EMPOWER BREAST CANCER PATIENTS

Tina L. Harralson, PhD1, Kristen Scott, RN,BSN1, Cori E. McMahon, PsyD2, Mark Redlus, BS1

1
Polaris Health Directions, Wayne, PA; 2Cooper University Hospital/ MD Anderson Cancer
Center at Cooper, Camden, NJ

Background: The emPower mobile application (App) was developed to engage and support
breast cancer (BC) patients by increasing awareness of emotions, physical symptoms, and
physical activity. The App offers a mobile connection between the patient, care team and
other BC patients via the Apple Watch and iPhone. Procedure: To test the feasibility and
acceptability of a native iOS App in a clinical setting, patients in active treatment for BC were
recruited between Dec 2015-May 2016. Participants were given an Apple Watch and an
iPhone 5s with a service plan, if they did not own an iPhone. Group training sessions were
held at the Cancer Center lasting about 3 hours: approximately 1 hour of training/questions
and 2 hours of socialization. Participants were alerted daily to complete a brief check-in on
the Watch or Phone regarding subjective wellbeing, distress, pain, fatigue, sleep quality. They
also received a daily inspiration and choice of physical, social, or mindfulness challenge.
Optional questions regarding daily physical symptoms could be completed on the Phone if
desired. Participants were also prompted to complete weekly and bi-monthly study questions
on the Phone: 1) a weekly social support check-in, and 2) questions pertaining to depression
(PHQ2) and anxiety (GAD2) every 2 weeks. The App interfaces with the Apple healthkit,
which passively obtains active calories and steps from the Watch. Distress, pain, fatigue and
active calories were presented in an interactive graph on the Phone revealing to the
participant the interrelationships between these variables. This data was also available to the
care team. Results: Twenty women with various stages of BC were enrolled in the study.
Mean age 50 years (range= 27-66 yrs.); 80% White, 20% African American. Intake assessment
completed at enrollment indicated 60% moderate to severe distress, 15% moderate to severe
anxiety (GAD7), 20% moderate to severe depression (PHQ9). Data analyzed during the first 4
weeks of use indicated a trend toward increased number of steps taken (Week 1: Mean= 4262
steps, s.d.=3766; Week 4: Mean=5116 steps, s.d.=4293; p=.06) and a slight increase in active
calories (Week 1: Mean= 215 calories, s.d.=128; Week 4: Mean=241 calories, s.d.=150; p= ns).
Since the participants were in various stages of cancer and treatment, the small changes in
activity were expected. Issues encountered during the study mainly pertained to use of Watch
and Phone. Participants were evenly split regarding who completed check-ins on the Watch or
Phone. Conclusion: Overall, the use of the emPower App proved feasible in a clinical setting
and feedback from participants has been positive. One of the most advantageous features
S1094 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

noted by participants was support and knowledge gained from the cohort. This App can be
adapted to all types of cancer as well to other chronic diseases.

CORRESPONDING AUTHOR: Tina L. Harralson, PhD, Polaris Health Directions, Wayne, PA,
19087; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1095

B113 6:00 PM-7:00 PM

HEALTH IT TO INCREASE PATIENT ACTIVATION: DEVELOPMENT AND USABILITY OF A PATIENT-

FACING, DIABETES DASHBOARD

William Martinez, MD, MS

Vanderbilt University Medical Center, Nashville, TN

Background
This study aimed to develop and evaluate a patient-facing, diabetes dashboard embedded
within an existing patient portal tethered to an electronic health record system to display
social and goal-based comparisons of users diabetes health data and increase patient
activation and improve self-management.

Methods
We performed a qualitative evaluation consisting of semi-structured interviews with 6
stakeholders and key informants (e.g., patient, endocrinologist, primary care physicians,
nurse, diabetes educator, health psychologist) to discern content, support needs, and
preferences regarding dashboard features and tools. We used grounded theory to analyze
these interviews for themes. After prototype design, we conducted usability testing under
controlled conditions with 8 patients with type 2 diabetes mellitus. Each patient participated
in a single usability session using a think aloud process. Patients preformed standardized
tasks within a dashboard prototype displaying fictitious patient data. Sessions were audio
recorded and computer screen activity was captured using QuickTime Player. Transcripts and
recordings were evaluated for common themes. Following testing, patients completed the
Computer Usability Satisfaction Questionnaire which assessed their perceptions of the
dashboards ease of use and the likability of the interface using a 7-point Likert scale, with
seven indicating the highest possible satisfaction.

Results
Results showed that stakeholders and key informants desired a dashboard that: (1) displays
diabetes health data in ways that facilitate patient understanding and promotes patient
activation, (2) facilitates patient action in response to the data they see (e.g., request a test or
vaccine, seek advice from their doctor), (3) provides diabetes education relevant to the data
being displayed, and (4) facilitates goal setting, tracks progress, and celebrates
accomplishments. Usability testing revealed problems with navigation, intuitiveness, design,
and terminology. Satisfaction was greatest for "easy to learn to use" (mean 6.6) and lowest for
"organization of information is clear" (mean 5.0). Overall mean satisfaction scores improved
between prototype iteration one and two (means, 5.9 vs. 6.6).
S1096 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions
This study demonstrated that diabetes stakeholders desire a patient-facing, diabetes
dashboard that facilitates patients understanding of their diabetes-related health data and
facilitates and track action. Task-based usability testing demonstrated difficulties in
navigation, medical language complexity, and organizational schema which led to
improvements. The current study also demonstrates that usability testing can assist in making
healthcare application interfaces for patients more useable and potentially more usefull.

CORRESPONDING AUTHOR: William Martinez, MD, MS, Vanderbilt University Medical Center,
Nashville, TN, 37212; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1097

B114 6:00 PM-7:00 PM

ONE DROP MOBILE APP USERS REPORT IMPROVED GLYCEMIC CONTROL

Jeff Dachis, MA1, Chandra Osborn, PhD, MPH2, David Rodbard, MD3, Brian Huddleston, JD4

1
Informed Data Systems Inc., New York, NY; 2One Drop, Informed Data Systems Inc.,
Nashville, TN; 3Biomedical Informatics Consultants LLC, Potomac, MD; 4Informed Data
Systems Inc., Austin, TX

Objective: Most mobile health apps lack scientific evidence. The One Drop mobile app offers
tracking of diabetes self-care behaviors and blood glucose (BG) values manually or
automatically via Bluetooth-enabled meters, CGMs, and other health apps. The app includes a
robust food library, medication scheduler, recipes, health tips, user polls, and community
support (likes, stickers, and data sharing). Users specify BG and self-care goals, and receive
data-driven insights to draw connections between behaviors, goals, self-reported hemoglobin
A1c and weight. In July 2016, we assessed A1c change among One Drop users self-reporting a
diabetes diagnosis.

Method: We queried data on ~50,000 users, identifying those users with two A1c values at
least 60 and no more than 365 days apart. A two-sided paired t-test assessed changes from
first to last A1c. Two-sided independent t-tests assessed gender and diabetes type differences
on the A1c change score, and Spearmans correlation coefficients assessed relationships
between diabetes duration, tracking data, and A1c change.

Results: The sample included 133 users (63 males), including 41 with type 1 (T1D) and 88 with
type 2 diabetes (T2D). Average diabetes duration was 12.4 10.8 years with 44.2% using
insulin and 50.4% using the One Drop medication scheduler. The average time interval
between A1c values was 146.1 76.7 days, with 3,225 5,780 (range: 2-37,838) tracking
entries during that period. A1c improved significatly from initial (7.8% 2.1%) to final period
(7.1% 1.3%), two-sided t=4.2, p < .001, and the change in mean A1c was similar irrespective
of gender or diabetes type. However, subgroup analyses by diabetes type revealed substantial
A1c improvement among people with T2D (7.7% 2.2% vs. 6.9% 1.2%; t=4.0, p < .001), with
a trend of improvement among people with T1D (8.0% 2.0% vs. 7.5% 1.3%; t=1.7, p=0.09).
Changes in A1c were not associated with duration of diabetes or tracking intensity.

Conclusion: People with diabetes using the One Drop health app reported a nearly 0.70%
reduction in A1c during 2-12 months of using it. Mean A1c change did not differ by gender or
S1098 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

diabetes type. Ongoing analyses of user data combined with results from a third-party trial
will provide further evidence of One Drops self-care and clinical value.

CORRESPONDING AUTHOR: Jeff Dachis, MA, Informed Data Systems Inc., New York, NY,
10002; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1099

B115 6:00 PM-7:00 PM

PERCEPTIONS OF SOCIAL MEDIA ENGAGEMENT AND SOCIAL NORMS FOR WEIGHT LOSS

Melissa Napolitano, PhD1, Meghan Mavredes, MPH1, Laura L. Hayman, PhD, RN, FAAN,
FAHA2, Samuel J. Simmens, PhD3, Jessica Whiteley, PhD4

1
The George Washington University, Washington, DC; 2University of Massachusetts Boston,
Boston, MA; 3George Washington University, Washington, DC; 4UMass Boston, Boston, MA

Background: Examining social media engagement and encouragement as it relates to body

weight can provide a channel for understanding the relative influence and impact of social
contacts. Purpose: Assess and examine social media engagement, social norms for weight
loss, and determine influences and differences by weight status or sex. Method: University
students (N=165; mean age=21.8+3.1 years; mean BMI=31.7+3.6 kg/m2; %female=80%; %
non-white=42.4) completed: demographics, social norms for weight loss (i.e., social contacts
approve or disapprove of weight loss/provide weight loss information or tools); social media
engagement (6-items: Facebook is part of my every day activity, Im proud to tell people
Im on Facebook, I feel out of touch when I havent logged into Facebook, I feel I am part
of the Facebook Community, I would be sorry if Facebook shutdown; Facebook has
become part of my daily routine Cronbachs alpha=.836). Results: Almost all (98.5%) reported
close social contacts would somewhat or strongly approve if they were to lose weight, yet
79.1% reported close social contacts rarely or never provided weight loss information or tools.
There were no differences in social norms for weight loss by weight status (overweight vs.
obese) or sex. Most endorsed high levels of social media engagement, agreeing or strongly
agreeing with Facebook being a routine (78.7%) and daily activity (82.9%), feeling Facebook
proud (82.9%) and part of a community (57.7%), and disappointed if Facebook no longer
existed (53.8%); yet, only 46.2% reported they would feel out of touch without Facebook.
There were no differences in social network engagement by weight status or sex. Chi-square
analyses examined the relationship between perceptions of social norms and social
networking engagement. When examining the social norm of contacts approving of weight
loss, there were significant results for Facebook being a part of ones routine, a daily activity,
and feeling proud to be part of Facebook (ps < .05). No significant differences were found for
the social norm of contacts providing weight loss assistance and social networking
engagement. Discussion: Although most reported social contacts approved of weight loss,
few reported contacts provided tangible weight loss support. Future studies are needed to
examine the directionality of the relationship between approval of weight loss by social
contacts and level of engagement in social networks.
S1100 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Melissa Napolitano, PhD, The George Washington University,

Washington, DC, 20052; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1101

B116 6:00 PM-7:00 PM

PREDICTORS OF ENGAGEMENT WITH A SMOKING CESSATION APP

Kisha I. Coa, PhD, MPH1, Ellen Beckjord, PhD, MPH2, Maria Asencio, MS3, Jason Choi, MS in
Data Science 3, Brian Keefe, MA1, Augustson Erik, PhD, MPH4

1
ICF International, Rockville, MD; 2UPMC Health Plan, Pittsburgh, PA; 3ICF International,
Fairfax, VA; 4National Cancer Institute, Rockville, MD

Background: Mobile phones are increasingly being used to deliver smoking cessation
interventions. One factor that limits the potential effectiveness of these programs is lack of
maintained use of mobile cessation apps (engagement). There is a need to better
understand factors that are associated with higher engagement with smoking cessation
mobile health interventions.

Objective: To examine patterns of use and functionality that are associated with engagement
among users of an evidence-based, publicly available smoking cessation app, QuitGuide.

Methods: Individual-level Google Analytics event data tracking and time stamping captured
each user action in the app between August 2015 and June 2016. Classification and regression
tree analysis (CART) was used to construct a model predicting the number of days a user was
active in the app.

Results: 9,348 unique users downloaded the app. Most users were daily smokers (90.5%),
reported their cravings were highest in the morning (51.0%), and smoked a mean of 16.2
cigarettes per day. Users were active in the app between 1 and 181 days with more than
half of users only being active 1 day (40.6%) or 2 days (18.9%). In the app, users can report
being smoke free, a slip, a craving, or their mood. Tracking cravings and reporting mood were
the most used features (53.0% of users reported each). CART analysis identified nine unique
groups with varying average number of days in the app. Those who were active in the app for
more days more frequently: (1) returned to the app after receiving local notifications (i.e.,
reminders encouraging to return to the app for support or tips); (2) reported being smoke
free; and (3) tracked their cravings.

Conclusion: These findings suggest that specific features local notifications and self-
monitoring (days smoke free, cravings) may encourage users to maintain use of a smoking
cessation app. Also, particular support may be needed for individuals to maintain use of a
cessation app beyond one or two days and before they achieve initial abstinence.
S1102 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Kisha I. Coa, PhD, MPH, ICF International, Rockville, MD, 20850;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1103

B117 6:00 PM-7:00 PM

PROSTATE CANCER PATIENTS INFORMATION TECHNOLOGY USE, PREFERENCES AND NEEDS: A

CROSS-SECTIONAL POPULATION BASED SURVEY

Jacqueline L. Bender, BSc, MSc, PhD1, Deb Feldman-Stewart, PhD2, Christine Tong, BScH, BEd,
MA 2, Howard Pai, MD3, Jacky WY. Au, BSc1, Michael D. Brundage, MD, MSc4, John W.
Robinson, PhD5, B. Joyce Davison, BN, MN, PhD6, Arminee Kazanjian, DrSoc 7

1
Princess Margaret Cancer Centre, University Health Network, Toronto, ON, Canada; 2Queen's
University, Kingston, ON, Canada; 3BC Cancer Agency Vancouver Island Centre, Victoria, BC,
Canada; 4Queens University, Kingston, ON, Canada; 5Tom Baker Cancer Centre and University
of Calgary, Calgary, AB, Canada; 6University of Saskatchewan, Saskatoon, SA, Canada;
7
University of British Columbia, Vancouver, BC, Canada

Objective: To determine how prostate cancer patients want to use the Internet for
information and support, and barriers and facilitators to use.

Methods: Surveys were conducted in three provinces in Canada (BC,AB,SK) in 2014-15, using a
modified Dillman mail-survey methodology. A random sample of ~55% of the population of
men in each provincial registry diagnosed with prostate cancer in the last half of 2012 was
surveyed.

Results: Response rates ranged from 46%-55% (N=1007). Mean age was 69 years old. 76%
used the Internet, and 63% used the Internet daily. Younger age, higher education, higher
income, urban residence and broadband Internet access were associated with Internet use.
74% accessed the Internet with a desktop/laptop 32% with a tablet, and 28% with a
smartphone. 53% found it easy to get information from the Internet. 37% agreed that they
felt confident using information from the Internet to help make health decisions (eHealth
literacy). 31% felt that their confidence in using the Internet to help make health decisions
had increased since being diagnosed with prostate cancer. Greater eHealth literacy was
associated with better self-reported health status. From a list of 10, top barriers to using the
Internet relate to low eHealth literacy inluding: not knowing what to trust (46%); not knowing
if it was personally relevant (41%); and not knowing how/ where to search (20%). If access
were easy, 58% would want to use the Internet for prostate cancer information. If there was a
website designed specifically for men with prostate cancer, 20% would not use it. From a list
of 13, top 5 features wanted in such a website were: library (64%); decision support tools
(57%); navigation support tools (49%); links to trusted websites (46%); and health assessment
S1104 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

tools (43%). 28% would want it to include an online peer support forum. In this website, 50%
or more would want information on: prostate cancer treatments (65%); disease progression
(64%); how to manage treatment side effects (61%); personally tailored information (54%);
and latest research (53%). Top facilitators for using such a website were: telephone support
(40%); using the website with a family/friend (28%); or a personal/group training session (25-
24%).

Conclusions: The majority of prostate cancer patients want to use the Internet as a source of
information about prostate cancer, but many do not know what information to trust or how
to find personally relevant information. To be optimal, prostate cancer websites should offer
tailored health information focusing on treatment options and side effects, and include
decision support tools and telephone technical support.

CORRESPONDING AUTHOR: Jacqueline L. Bender, BSc, MSc, PhD, Princess Margaret Cancer
Centre, University Health Network, Toronto, ON, M5G 2C4; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1105

B118 6:00 PM-7:00 PM

USE AND ACCEPTANCE OF A MOBILE APPLICATION TO PREVENT ALCOHOL MISUSE AMONG

MILITARY PERSONNEL

Cynthia Simon-Arndt, MBA, MS1, Suzanne Hurtado, MPH2, Mary Jennifer McAnany, PhD1,
Emily Schmied, PhD, MPH1

1
Naval Health Research Center/LEIDOS, San Diego, CA; 2Naval Health Research Center, San
Diego, CA

Background: Alcohol misuse is a serious and growing issue among active duty military
servicemembers. There are many barriers to implementing military alcohol interventions,
such as the organizational drinking culture, stigma against substance abuse treatment, and
the transient lifestyles of servicemembers. The accessibility of mobile health applications
(apps) offers a unique approach to reaching servicemembers, yet no prior research has
examined the feasibility of employing this tool for alcohol misuse prevention in a military
setting. The objective of this study was to assess the use and acceptance of a mobile app for
alcohol misuse among servicemembers.

Methods: An original app, Alcohol Determinator, was developed specifically for Marines. App
content was based on the principles of brief intervention and included five components: self-
assessment with feedback, drinking log with goal-setting, blood-alcohol content (BAC)
calculator, information, and resources. One-hundred fifty-eight active duty Marines were
randomly assigned to an intervention group (n=65; 89.1% male, 49.2% White) or an attention
control group (n=93; 86.8% male, 50.0% White). The intervention group received Alcohol
Determinator and the control group received a popular fitness app. After one month, user
ratings and app usage data were collected for both study groups via in-app tracking and
electronic surveys.

Results: A majority of intervention participants agreed the app was easy to use (82.7%), was
appropriate for Marines (74.5%), and would recommend it to another Marine (75.0%). In
terms of content, intervention participants were most satisfied with the BAC calculator,
feedback on money spent on alcohol and the drinking log charts. Group comparison tests
showed intervention participants used the app less often than control participants during the
observation period (p.05).

Conclusions: Servicemembers used and favorably reviewed the Alcohol Determinator. Mobile
apps to prevent alcohol misuse built on evidence-based principles may be an appropriate
intervention for this unique, at-risk population.
S1106 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Cynthia Simon-Arndt, MBA, MS, Naval Health Research

Center/LEIDOS, San Diego, CA, 92106; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1107

B119 6:00 PM-7:00 PM

USING MHEALTH APPS IN SELF-MANAGEMENT INTERVENTIONS TO PROMOTE PHYSICAL

ACTIVITY: RANDOMIZED CONTROLLED PILOT STUDY

Matthew Plow, PhD, Meghan Golding, BS

Case Western Reserve University, Cleveland, OH

Purpose: Examine the effectiveness of using commercially available mHealth apps in a self-
management intervention to promote physical activity in 46 adults with mobility-impairing
musculoskeletal or neurological conditions, such as arthritis, multiple sclerosis, and
fibromyalgia.

Methods: Participants were randomized to one of three groups: (1) mHealth-based self-
management intervention, (2) paper-based self-management intervention, and (3) contact-
control intervention. All three interventions consisted of one in-person visit plus three phone
calls over a six week period. Participants in the mHealth and paper-based intervention groups
received a Google Nexus Tablet or a paper diary, respectively, to facilitate goal-setting, self-
monitoring, and action planning. Participants in the control intervention received information
on healthy behaviors without being taught self-management skills. Outcomes were
administered at baseline and seven weeks. Three MANOVAs were conducted: (1) self-report
physical activity (planned exercise and leisure-time physical activity, general physical activity,
and employment physical activity), (2) psychosocial constructs (self-efficacy, self-regulation,
and social support), and (3) physical function (PROMIS, 6-Minute Walking Test, chair stands,
and arm curls).

Results: Both the mHealth and paper-based interventions had significant (Wilks' = 0.71, F6,
76 = 2.34, P = .04) and large effect size increases in planned exercise and leisure-time physical
activity compared to the control intervention (Cohens d = 1.20 and d = 0.82, respectively).
The mHealth-based intervention had non-significant but small to moderate effect size
increases in planned exercise and leisure-time physical activity compared to the paper-based
intervention (Cohens d = 0.47). Both the mHealth and paper-based interventions had non-
significant (Wilks' = 0.85, F6, 76 = 1.10, P = .37) differences but moderate effect size
improvements in self-efficacy (d = 0.48 and d = 0.75; respectively) and self-regulation (d = 0.59
and d = 0.43, respectively) compared to the control intervention. There were no significant
changes in physical function within and between the three interventions. Participants were
significantly (P = .02) more likely to track physical activity behaviors using the paper diary
compared to the Google Nexus Tablet.
S1108 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions: This study demonstrates the feasibility of conducting a larger randomized

controlled trial to examine the use of mHealth apps in self-management interventions. We
found that using commercially available mHealth apps in self-management interventions
shows promise in promoting physical activity in adults with neurological and musculoskeletal
conditions. Further research is needed to improve the usability of mHealth apps for adults
with neurological and musculoskeletal conditions.

CORRESPONDING AUTHOR: Matthew Plow, PhD, Case Western Reserve University, Cleveland,
OH, 44106; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1109

B120 6:00 PM-7:00 PM

VIRTUAL GARDENING PILOT INTERVENTION INCREASES SELF-EFFICACY TO COOK AND EAT

FRUITS AND VEGETABLES IN MINORITY YOUTH

Brooke Bell, BA1, Lauren C. Martinez, PhD2, Marientina Gotsis, MFA1, H. Chad Lane, PhD3,
Jaimie Davis, PhD4, Luz Antunez Castillo, MBA1, Giselle Ragusa, PhD1, Donna Spruijt-Metz,
MFA, PhD1

1
University of Southern California, Los Angeles, CA; 2University of Southern California,
Pasadena, CA; 3University of Illinois, Urbana-Champaign, Champaign, IL; 4University of Texas
at Austin, Austin, TX

African American and Hispanic youth in Los Angeles have significantly higher rates of
overweight and obesity than is reported for these children nationwide. Intake of fruits and
vegetables (FV) has been inversely associated with body weight. School gardening programs
have shown success in increasing preference of and improved attitudes towards FV, however
implementation in urban settings has several limitations. Virtual Sprouts (VS) is a virtual
gardening game developed for minority youth and their families. VS teaches users about FV,
nutrition, gardening, and cooking through meaningful play.

In this quasi-experimental pilot intervention, elementary school students were enrolled in a 3-

week program that included three VS gaming sessions, three in-school lessons, and three in-
home activities, using a nutrition- and gardening-focused curriculum. Pre- and post-
intervention questionnaires were used to assess psychosocial determinants of dietary
behavior, including knowledge about and self-efficacy to eat FV. Data was collected on FV,
whole grains, fiber, total sugar, added sugar, and energy from sugar beverages intake via the
Block Kids Food Screener (last week version). Repeated-measures analysis of covariance
models was used for continuous outcomes, controlling for age, sex, ethnicity, school, and free
school lunch. Ordinal logistic regression was used for binary outcomes with the same
covariates.

180 students from 3rd, 4th, and 5th grade classrooms in two Los Angeles Unified School
District schools were randomized into intervention and control groups. After the intervention,
the intervention group (n=116) had an improved self-efficacy to eat FV score (p=0.01), an
improved self-efficacy to cook score (p=0.05), but not a significantly improved self-efficacy to
garden score. There were no significant differences in dietary intake between intervention
and control groups.
S1110 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

The results from this 3-week pilot study suggest that a virtual gardening game with a
nutrition- and gardening-focused curriculum can improve psychosocial determinants of
dietary behavior in minority youth.

CORRESPONDING AUTHOR: Brooke Bell, BA, University of Southern California, Los Angeles,
CA, 90017; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1111

B121 6:00 PM-7:00 PM

COMMUNITY KNOWLEDGE OF SOCIAL DETERMINANTS OF HEALTH

Rachel Barth, MSW, Nikole Lobb Dougherty, MA, Emily Kryzer, MSW, MPH, Jasmine L.
Williams, MPH, Jason Purnell, PhD, MPH

Washington University in St. Louis, St. Louis, MO

Background: Following the publication of a reporton the health and well-being of African
Americans in St. Louis in May 2014, the For the Sake of All initiative has focused on engaging
and disseminating information to community to increase knowledge of social determinants of
health (SDH). Engagement approaches included community feedback meetings, discussion
guides, action toolkits, and Community Action Forums (forums).

Methods: Six forums were held at community sites throughout St. Louis from October 2014
through June 2016. Each forum engaged community in a discussion of the impact of SDH on
regional health disparities. Forum attendees self-reported attitudes toward SDH were
assessed upon forum registration (pre-event) and approximately 12-weeks following (post-
event). Dissemination reach as well as reported changes in attitude were aggregated and
assessed.

Results: A total of 426 people have attended the first four forums. Attendees have come from
36 unique zip codes and have been predominantly female (63%-84%), highly educated (70%-
86% with at least a bachelors degree), and affiliated with a community organization (23%-
45%) and with higher education (14%-21%). In terms of race, between 41% and 53% of
attendees have been white, 25%-34% black, 5%-7% other, and 15%-20% unreported. Each
forum draws between 48% and 57% new attendees.

Preliminary results from paired samples t-tests comparing mean differences in pre- and post-
event responses indicate small but significant changes in attitude regarding the extent to
which health is affected by level of income (Pre M = 8.71, SD = 1.84, Post M = 9.45, SD = 1.00;
t(72) = -4.21, p < .001), level of education (Pre M = 8.03, SD = 1.78, Post M = 8.95, SD = 1.33; t
(72) = -4.99, p < .001), community safety (Pre M = 8.49, SD = 1.71, Post M = 9.27, SD = 1.06; t
(70) = -3.80, p < .001), and having health insurance (Pre M = 8.85, SD = 1.68, Post M = 9.25, SD
= 1.16; t (72) = -2.37, p = .02).There were no significant changes with respect to the influence
of personal health practices, access to affordable care, or genetic makeup on health (all ps >
.05).

Discussion/Conclusion: Public discourse about improving health has long focused on

S1112 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

improving access to health care and on individual health behaviors. Engaging community
members in targeted education and discussion about the impact of SDH on health is one
method by which public health researchers can shift this discourse to focus on advancing
social and economic policy.

CORRESPONDING AUTHOR: Rachel Barth, MSW, Washington University in St. Louis, St. Louis,
MO, 63130; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1113

B122 6:00 PM-7:00 PM

COMMUNITY TO CLINIC NAVIGATION TO INCREASE COLON CANCER SCREENING: A

DISSEMINATION/IMPLEMENTATION MODEL

Usha Menon, PhD, RN, FAAN1, Linda K. Larkey, PhD2, Laura Szalacha, EdD1

1
University of Arizona, Tucson, AZ; 2Arizona State University, Phoenix, AZ

Community to Clinic Navigation to Increase Colon Cancer Screening: A

Dissemination/Implementation Model

Usha Menon, PhD, RN, FAAN1; Linda Larkey, PhD2; Laura Szalacha, EdD1

1
The University of Arizona

2
Arizona State University

Background: Regular screening facilitates early diagnosis of colorectal cancer (CRC) and
contributes to reduction of CRC morbidity and mortality. Recent increases in CRC screening
utilization are not mirrored in poor and minority populations. CRC screening rates are
particularly low for those without a primary care provider or clinic and who also have lower
levels of education, income, and insurance. A large proportion of ethnically diverse and poor
patients do not regularly visit primary care clinics, further lowering their likelihood for getting
screened and underscoring the need to reach this population with effective interventions in
community settings.

Purpose: We developed and tested a community to clinic tailored navigation intervention

using a dissemination (randomized phase one) and implementation (non-randomized phase
two) framework targeting low-income, diverse communities. The final outcomes were: a)
clinic attendance, b) CRC screening, and tracking RE-AIM factors of dissemination.
S1114 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Methods: We randomized 239 community sites to Education + Tailored Navigation (TN) and
116 to Education only. In phase one, all participants received education on CRC risk and
screening. Those in the Education only group received up to 5 calls to inquire about
appointments/screening. Those in the Education + TN group received up to 5 calls from
navigators who assisted them with barriers using a tailored message bank, guiding them to
schedule a clinic appointment. In phase two, all those who attended a clinic, regardless of
previous group assignment, received TN from a trained study navigator to guide them to
complete screening.

Results: Of the 416 participants who completed the baseline survey (53% Female, 47% Male,
Mean age 61.5; SD=7.4; 52.7% Latino, 36% Uninsured); 25%; (n = 104) made clinic
appointments, and 80% of these participants (n=65) completed a screening test. Overall, the
Education + TN participants were 6 times more likely to get screened (54/209, 25.9%, OR =
6.28) than were Education alone participants (11/207, 5.3%).

Conclusions: Adding TN to community education was significantly more efficacious than

Education alone, indicating TN was successfully translated into the community setting, with
continuing implementation in the clinics to increase CRC screening rates. Reach, effectiveness,
adoption, implementation and maintenance data will be discussed.

CORRESPONDING AUTHOR: Usha Menon, PhD, RN, FAAN, University of Arizona, Tucson, AZ,
85721; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1115

B123 6:00 PM-7:00 PM

EXAMINING PARENT ENGAGEMENT IN A FAMILY-BASED CHILDHOOD OBESITY INTERVENTION:

A QUALITATIVE STUDY

Emily Schmied, PhD1, Emmeline Chuang, PhD2, Hala Madanat, PhD1, Jamie Moody, MS3,
Leticia Ibarra, MPH4, Guadalupe X. Ayala, PhD, MPH1

1
San Diego State University, San Diego, CA; 2University of California, Los Angeles, Los Angeles,
CA; 3Institute for Behavioral and Community Health, San Diego, CA; 4Clinicas de Salud del
Pueblo, Inc, Brawley, CA

Background: Low parent engagement, or limited attendance at or active participation in

planned program activities, is often cited as a barrier to the implementation and ultimately,
the efficacy of family-based childhood obesity prevention and control programs. This
qualitative study examined factors influencing parent engagement in a family-based
childhood obesity prevention and control program.

Methods: Participants included 22 mothers (mean age=34.8 years) enrolled in a family-based

childhood obesity prevention and control program. Semi-structured interviews guided by the
Health Belief Model and Transtheoretical Model assessed parents motivation for
participating, facilitators and barriers to participation, and satisfaction with the program.
Interviews were conducted in participants language of choice (English=13.6%;
Spanish=86.4%) in person (31.8%) or via phone (62.2%) following the conclusion of program
activities. Interviews were transcribed and summarized. Data were analyzed for differences
between parents who attended at least two-thirds of program activities (high engagement)
and those who did not (low engagement).

Results: There were no significant demographic differences between parents who did and did
not complete two-thirds of program activities. Parent engagement was influenced by
childrens level of support and enthusiasm for the program, and by parental expectations of
program benefits to the family. Parents were highly satisfied with the programs content and
the community health workers who delivered the program.

Conclusions: This study adds to emergent literature regarding parents experiences in family-
based childhood obesity prevention and control programs. Findings identify potential targets
for improving engagement, such as strategies to engage additional family members to
increase the support received by the participating parent.
S1116 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Emily Schmied, PhD, San Diego State University, San Diego, CA,
92120; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1117

B124 6:00 PM-7:00 PM

FEASIBILITY AND ACCEPTABILITY OF GROUP VISIT FOR DERPRESSION IN PRIMARY CARE

Sona Hovsepian, MSW, ASW1, Megan Johnson, MD2, Sonya Gabrielian, MD, MPH1, Alexander
Young, MD, MSHS1

1
Veterans Health Administration, Los Angeles, CA; 2Kaiser Permanente, Los Angeles, CA

Background: Depression is common in primary care settings and is often chronic, recurring,
and comorbid with chronic medical illness. Improving outcomes for depression and chronic
medical illnesses requires patients to become educated, active partners in their care.
Outpatient psychiatric care is frequently characterized by brief, infrequent individual
medication management visits that often fall short of providing guideline-concordant care.
Such visits are also insufficient for helping patients with chronic illnesses gain the knowledge,
skills, and self-efficacy needed to promote treatment adherence and self-care.
Objective: To assess the feasibility and acceptability of a group visit for depression
(Medication and Self-Management to Empower Recovery (MASTER)) within primary care
settings.
Methods: Within the VA Greater Los Angeles patient-centered medical home, psychiatrists
referred patients 18 years old with depression. Participants were scheduled for a series (16
weeks) of group appointments for depression in VA primary care. Each group was 90 minutes,
split between psychoeducation, mindfulness, and a round-robin for medication adjustment
and discussion of side effects for each patient. PHQ-9 scores were performed at baseline and
group completion. Semi-structured questionnaires at the end of the 16 weeks and qualitative
interviews at 3 months follow-up assessed patient experience of the groups, barriers and
facilitators, and recommendations.
Results: Of the 9 veterans enrolled, a subsample (n = 6) ultimately completed an average of
7.8 support groups. 5 of the veterans completed the 3 month qualitative interviews. Veterans
who completed the interviews would recommend the group to a peer. Some of the Veterans
discussed additional topics they would like in the support groups; these were, veterans of
different war eras, introduction to relaxation music, pain management for specific conditions.
Veterans saw and average of 10.2 points improvement in their PHQ-9 score during the 16
weeks, identified enjoyable activities and coping skills at each session.
Discussion: Group visits for depression were feasible and acceptable in primary care. Veterans
with a diagnosis of depression reported that they benefitted from depression support groups.
This pilot experience suggests the need for a larger trial to assess the effectiveness of this
hybrid treatment model medication management and group self-management support in
primary care setting.
S1118 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Sona Hovsepian, MSW, ASW, Veterans Health Administration, Los
Angeles, CA, 90073; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1119

B125 6:00 PM-7:00 PM

HEALTHY EATING ACTIVE LIVING: PILOT STUDY OF A MULTI-COMPONENT COMMUNITY-BASED

PROGRAM AMONG LOWER-INCOME PREGNANT WOMEN

Shreela Sharma, PhD RD LD1, Ru-Jye Chuang, DrPH2, Courtney Byrd-Williams, PhD3, Gregory
Bounds, MPH1, Sania Durrani, B.S., B.A.4, Deanna M. Hoelscher, PhD, RD, LD5

1
University of Texas School of Public Health, Houston, TX; 2UTHealth School of Public Health,
Houston, TX; 3University of Texas School of Public Health, Austin, TX; 4The University of Texas
Health Science Center at Houston, School of Public Health, Friendswood, TX; 5University of
Texas School oF Public Health, Austin, TX

Healthy Eating Active Living (HEAL) is a multi-component, community-based, community

health worker(CHW)-led six-week program designed to promote healthy eating, physical
activity, and intention to breastfeed among Medicaid-eligible pregnant women in Houston,
TX. Pregnant women are recruited through clinics and participate in weekly 90-minute
facilitated and experiential, group-education sessions. The purpose of this pilot study is to
evaluate the feasibility and effectiveness of HEAL in improving participants dietary habits,
home nutrition environment, and time spent in physical activity (PA), and intention to
breastfeed.

This was a one-group pre-post evaluation design. Availability of fruits and vegetables (FV) (23
items), daily consumption of FV (24 items) and sugar-sweetened beverages (SSB) (1 item),
perception (1 item) and frequency (2 items) of PA, and intention to breastfeed (4 items) were
measured using self-administered surveys. Ninety-two Medicaid-eligible pregnant women
completed both pre-and post- surveys from October 2014-September 2015. Wilcoxin signed-
rank test and McNemars test were conducted to evaluate pre- and post-intervention
changes. Process evaluation was conducted weekly using participant surveys to assess
program acceptability, and from CHWs through post-session facilitator survey to evaluate
program fidelity.

Results showed that participation in HEAL significantly increased home availability of 11 FV (p

< 0.05), daily FV consumption (p < 0.05), and decreased SSB consumption (p < 0.05). At post-
intervention, significantly more women reported being physically active (p < 0.05), and being
physically active for over 30 minutes per day (p=0.03), and at least 10 minutes at a time (p <
0.05). Significant increases in attitude and perceived benefit of breastfeeding (p < 0.05) were
found at post-intervention, but not in intention to breastfeed. Among women intending to
breastfeed, more women planned to continue after returning to work/school (p < 0.05).
S1120 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Process evaluation showed that participants found HEAL sessions very helpful in improving
their health behaviors (mean= 3.83, possible score range 0-4). Post-session facilitator survey
showed high fidelity of program components (discussion topics 98.9%, physical activity 95.8%,
and recipe demonstration 100%).

This pilot study demonstrates the initial feasibility and success of creating successful clinic-
community linkage and using a CHW-led intervention to improve health behaviors among
low-income pregnant women.

CORRESPONDING AUTHOR: Shreela Sharma, PhD RD LD, University of Texas School of Public
Health, Houston, TX, 77030; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1121

B126 6:00 PM-7:00 PM

SUSTAINED IMPLEMENTATION OF A BIOBEHAVIORAL INTERVENTION BY COMMUNITY

PROVIDERS

Marlena Ryba, PhD1, Brittany Brothers, PhD2, Stephen Lo, M.A. 1, Barbara L. Andersen, PhD1

1
The Ohio State University, Columbus, OH; 2Indiana University, Bloomington, IN

Background. There is accumulating research on dissemination and implementation (DI) of

evidence based treatments (EBTs) for mental health. However, there is a paucity of research
regarding the sustainability of these interventions. In the context of a DI effort, mental health
professionals completed a 3-day training institute in an evidence-based biobehavioral
intervention (BBI) for cancer patients aimed at reducing stress and enhancing mental health.
Training was effective (Brothers et al., 2015) and clinicians were provided 6 months of
implementation support (e.g. conference calls, web resources). Purpose. This study aimed to
examine the 12 month sustainability of the providers usage of the BBI. Methods. Providers in
social work, psychology, and other licensed mental health professions from large medical
centers, community facilities, and private practice all were BBI trained and then reported BBI
usage (i.e., the percentage of the patients seen in the prior month that were treated with the
BBI). During months 1-6 they were provided with implementation support but none
thereafter. Usage rates at 2, 4, 6, and 12 months were reported. A generalized estimating
equation with a logistic link function was used to evaluate the level of BBI usage over
time. Results. The model estimated an average BBI usage of 60.1% (95% CI=53.4-66.4%) at 2
months, 66.3% (95% CI=59.8-72.4%) at 4 months, 68.4% (95% CI=62.2-73.9%) at 6 months,
and 70.9% (95% CI=63.6-77.3%) at 12 months. There was a significant difference in BBI usage
from month 2 to 6 (p=0.010) and 12 (p=0.005). Conclusions. This first dissemination cancer-
specific psychological intervention shows success at all stages. There were early, high usage
rates which steadily increased during a 6 month period of implementation support. Moreover,
usage rates even climbed to rates exceeding 70% at 12 months. The paradigm used here for
the successful training of mental health professionals provides guidance for the dissemination
of other behavioral medicine psychosocial interventions to community providers.

CORRESPONDING AUTHOR: Marlena Ryba, PhD, The Ohio State University, Columbus, OH,
43210; [email protected]
S1122 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B127 6:00 PM-7:00 PM

THE INFLUENCE OF RISK FACTOR FEEDBACK ON PARTICIPATION IN A DIABETES PREVENTION

PROGRAM FOR AFRICAN AMERICAN WOMEN

Heather Kitzman-Ulrich, PhD1, Abdullah Mamun, MS2, Leilani Dodgen, MPH3, Kisa Gant, BS3,
Surendra Reddy Mandapati, BDS3, Ike Eke, BA2

1
Baylor Scott & White Health, Dallas, TX; 2University of North Texas Health Science Center,
fort Worth, TX; 3University of North Texas Health Science Center, Fort Worth, TX

Background: Previous research has found that providing health risk data, such as diabetes
(fasting glucose or HbA1c) or cardiovascular risk (total cholesterol, high density lipoprotein,
low density lipoprotein), can serve as a call to action and increase participation in prevention
programs. However, little is known on how provision of health risk data increases
participation in prevention programs for African American (AA) women, who are
disproportionally affected by diabetes and cardiovascular disease (CVD). Methods: Data
collected during baseline measures of the Better Me Within randomized study were used to
evaluate whether risk status provided at baseline was related to greater participation in a 16-
week church based diabetes prevention program. Physiological data was calculated from a
fasting blood sample collected through finger stick by trained research staff using a point of
care test (Alere Cholestech LDX Analyzer) including total cholesterol (TC), high density
lipoprotein (HDL), low density lipoprotein (LDL), and triglycerides (TRG). Fasting glucose (FG)
and glycated hemoglobin (HbA1c) were calculated with a fasting blood sample using a point of
care test (Bayer A1cNow). Attendance was tracked for each individual by week and a mean
attendance score was calculated. Results: 263 AA women participated in baseline measures
(MBMI= 36.5+8.1; MAge = 51.6+11.9), 43.3% had an annual household income < $50,000, 26%
had a high school education or less, 77.2% had at least one risk factor for CVD, and 74.1% had
at least one risk factor for diabetes. Of the 5 CVD risk factors evaluated (TC, HDL, LDL, TRG,
BMI), only borderline high or high TRG (>150mg/dL) predicted greater participation (>70%
attendance of 16 sessions). Neither diabetes risk factor (FG or HbA1c) predicted greater
participation. The strongest predictors of participation were household income, with 66% of
higher income as compared to 51% of lower income participants (p=.01); and education, with
65% of participants with a higher education as compared to 44% of participants with a lower
education (p < .01), attending at least 70% of the 16-week program. Conclusions: Greater
participation in health promotion programs have been associated with better participant
outcomes such as weight loss or reduced diabetes risk. These findings suggest that real time
physiological feedback, for the exception of TRG, does not predict greater participation, and
that education and income may be key drivers of individual participation rates. Future
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1123

research should explore mechanisms for increasing participation in prevention programs by

individuals with lower incomes and less educational attainment as they are at higher risk for
chronic disease.

CORRESPONDING AUTHOR: Heather Kitzman-Ulrich, PhD, Baylor Scott & White Health, Dallas,
TX, 75210; [email protected]
S1124 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B128 6:00 PM-7:00 PM

THE ROLE OF MIDDLE MANAGERS IN INNOVATION IMPLEMENTATION IN HEALTH CARE

Robin Urquhart, PhD1, Cynthia Kendell, MSc1, Amy Folkes, MAHSR1, Tony Reiman, MD2, Eva
Grunfeld, MD, DPhil3, Geoff Porter, MD1

1
Dalhousie University, Halifax, NS, Canada; 2Dalhousie University, Saint John, NB, Canada;
3
University of Toronto, Toronto, ON, Canada

Objective: To explore middle managers role in and experiences with innovation

implementation in the cancer system.

Methods: We conducted a grounded theory study using in-depth, semi-structured qualitative

interviews of middle managers in two Canadian provinces (Nova Scotia, New Brunswick).
Participants were purposively sampled, based on their involvement in implementation
initiatives and to obtain variation in manager characteristics (location, years of managerial
experience). Interviews were audiotaped and transcribed verbatim. Data were collected and
analyzed concurrently, using an inductive constant comparative approach.

Results: Seventeen middle managers participated in this study. Collectively, they were
involved in the implementation of a range of innovations in cancer screening, diagnosis, and
treatment. The analysis revealed four main categories. Middle managers see themselves as
being (1) responsible for making implementation happen in their programs/services. As a
result, they take on varied tasks, such as leading committees/working groups, navigating
bureaucratic processes, gathering information, and developing new operational processes. At
the same time, they have (2) limited decision-making authority with respect to the
implementation decision, with middle managers working within parameters set by upper
levels of the organization. Their role during implementation is akin to that of (3) team captain,
executing leadership tasks but also very much a part of the team effort. This is often
challenging, since middle managers (4) perform many roles, both clinical and managerial in
nature, with innovation implementation added on top of their regular duties. Consequently,
most engage in much self-directed learning to help them succeed in these multiple roles.

Conclusions: Middle managers are given scant attention in the implementation literature in
health care, where the focus is on senior leaders and frontline staff. As a result, their
contribution is often seen as negligible. This study suggests middle managers play a leading
role in implementation processes. Optimizing their capacity to fulfill this role may be key to
improve innovation implementation in healthcare organizations.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1125

CORRESPONDING AUTHOR: Robin Urquhart, PhD, Dalhousie University, Halifax, NS, B3H 2Y9;
[email protected]
S1126 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B129 6:00 PM-7:00 PM

THEMATIC CONTENT ANALYSIS OF PRACTITIONERS' INQUIRIES ABOUT INTERVENTIONS:

IMPLICATIONS FOR RESEARCH IMPLEMENTATION

Margaret M. Farrell, MPH RD1, Jordan N. Tompkins, BA, MAA1, Wynne Norton, Kaelin
Rapport, N/A2

1
National Cancer Institute, Bethesda, MD; 2National Cancer Institute, College Park, MD

Background: Improving communication and strengthening collaboration between researchers

and public health practitioners is critical for effectively disseminating and implementing
evidence-based interventions. To this end, NCIs Research to Reality (R2R) online community
of practice convenes researchers and practitioners around a shared commitment to
implement evidence-based, cancer-related (e.g., physical activity, tobacco control, screening,
prevention) interventions in public health settings. Several dynamic approaches are used to
facilitate dynamic communication and collaboration, including monthly webinars, question
and answer (Q&A) sessions, and discussion posts.

Methods: This study explored the programmatic considerations most important to R2R
community members through a content analysis of webinar Q&A sessions and related
discussions. We used deductive coding for segments of text related to implementation
strategies (Powell et al 2015; Wang 2016) and intervention adaptation (Wiltsey-Stirman et al,
2012), and inductive coding to explore thematic patterns from participants questions. This
coding schema was applied to 62 transcripts of webinar presentations (Q&A section only) and
the related online discussions over a 6-year period (2010-2016).

Findings: Several themes emerged from the content analysis, including queries around
institutional support, sustainability, tools and resources. Community members requested
additional information from presenters about contextual variables of research studies (e.g.,
the role of partnerships in implementing disease prevention and public health initiatives) as
well as evidence-based programs. Practitioners consistently sought to better understand
variables that made an intervention effective and sought guidance on the most appropriate
implementation strategies for their projects.

Implications: While a number of resources are designed to implement evidence-based health

programs, the strategies and programmatic considerations most salient to practitioners are
difficult to elicit but essential to address. These data highlight additional information
researchers should include in their intervention presentations, summaries and publications
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1127

when communicating with practitioners. By addressing the priority concerns of practitioners,

researchers can to improve communication and enhance the actual, widespread
implementation of promising public health interventions across jurisdictions.

CORRESPONDING AUTHOR: Margaret M. Farrell, MPH RD, National Cancer Institute,

Bethesda, MD, 20852; [email protected]
S1128 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B130 6:00 PM-7:00 PM

TRAINING STUDENTS IN UNDERSERVED HIGH SCHOOLS ALONG THE US-MEXICO BORDER TO

IMPLEMENT PROJECT STUDENTS ARE SUN SAFE (SASS)

Loescher J. Lois, PhD, FAAN1, Sarah Rawdin, MPH, BSN, RN1, Gail Emrick, MPH2, Tashina M.
Machain, BA2, Denise A. Spartonos, BA1, Alice Pasvogel, PhD1, Riley E. Johnson, CITI3, David
Campas, B.S. Public Health 4

1
The University of Arizona, Tucson, AZ; 2Southeast Arizona Area Health Education Center,
Nogales, AZ; 3University of Arizona Cancer Center Skin Cancer Institute, Tucson, AZ;
4
University of Arizona Health Promotion Research Assistant, Tucson, AZ

Background: Skin cancer is a major public health issue in the US and Arizona with
overexposure to ultraviolet radiation (UVR) a known cause. Most people living near the rural
Arizona-Mexico border are Hispanic; skin cancer incidence is rising within Hispanics. Youth are
prone to UVR overexposure, yet few skin cancer prevention programs target rural and
underserved youth. Schools are ideal intervention settings, but teachers in border-area high
schools have limited time and resources for skin cancer prevention education. Educators and
researchers at the University of Arizona (UA) have successfully implemented Project Students
are Sun Safe (SASS) in urban schools. In the current model, UA health sciences students (peer
leaders) take a one-semester skin cancer prevention academic course: online modules on skin
cancer epidemiology, types, prevention and communication, followed by in-person skills
evaluation. Trained peer leaders implement, in community classrooms, a brief SASS lesson
(basic epidemiology, skin structure, skin cancer, UVR and protection strategies, and tanning
consequences), reinforced by three interactive activities. Adapting this model for high school
students will allow further dissemination and implementation of SASS, especially into rural
and underserved schools.

Purpose: To report preliminary findings on the training of border-area high school students to
deliver SASS, specifically a) adaptation process of the college course; b) strategies to engage
the students in training; and c) evaluation of the feasibility and efficacy of the training.

Methods: Using a CBPR framework, our research team and community partner with a
longstanding relationship with border-area high schools, jointly adapted the SASS training
modules and classroom lesson. Eighteen students from 3 border-area high schools were
offered the online training. They completed the online pretest and posttest at 3 weeks,
measuring skin cancer prevention risk, knowledge, attitudes, and behaviors. We monitored
successes and challenges of our strategies.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1129

Results: The adaptation process of several months reduced the modules from 5 to 3. The
students had no experience in online learning, which required adjustments in training. The
majority of trainees were female, Hispanic, and had skin cancer risk factors: 83.3% were
raised in Arizona; over 70% reported skin susceptible to sunburn, with 35% having 2+
sunburns the past year. Following training, sun-safety knowledge improved (p = .002),
perceived seriousness (p = .000) and risk (p = .02) were more favorable, and self-reported sun
safety behaviors improved (p < .005).

Conclusions: We successfully adapted SASS sun-safety training to youth living along the US-
Mexico border. These students had stronger skin cancer risk factors than we anticipated. In
January 2017, these students will take a final posttest to assess longer-term efficacy of
training.

CORRESPONDING AUTHOR: Loescher J. Lois, PhD, FAAN, The University of Arizona, Tucson, AZ,
85721-0203; [email protected]
S1130 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B131 6:00 PM-7:00 PM

USE OF CONTENT CURATION, TAILORED CAMPAIGNS & SOCIAL MEDIA OUTREACH TO

SUPPORT LATINO HEALTH

Amelie Ramirez, DrPH1, Rosalie P. Aguilar, MS2, Lisa Ellis, BA3, Amanda Merck, MPH4, Eric
Moreno, MA Communications5, Cliff Despres, BJ1, Kip Gallion, MA1

1
UT Health San Antonio, San Antonio, TX; 2Salud America! /UT Health Science Center at San
Antonio, San Antonio, TX; 3Salud America, San Antonio, TX; 4Salud America!, San Antonio, TX;
5
University of Texas Health Science Center at San Antonio, San Antonio, TX

Use of Content Curation, Tailored Campaigns & Social Media Outreach to Support Latino
Health

BACKGROUND: By 2023, it is expected that 30% of public school students will be Latino.
Despite rapid population growth and a longer life expectancy, Latinos face disparities in
childhood obesity, diabetes, asthma, and certain cancers, which greatly reduces their quality
of life and leads to increased costs of health care. In 2014, approximately 73% of Latinos
reported using their smartphones to access health information from the Internet. Salud
America! engages its 50,000+ member network through multi media content, research and e-
communications aimed at empowering local level leaders and helping build their case for
healthier, more resilient communities.

METHODS: Between Jan-Aug 2016, 3 SA! curators developed an average of 41 geo-tagged

blog posts, 33 resource posts and 3 full-length feature role model stories per month. Tenets
from social cognitive theory were used to craft campaigns and disseminate messages to
support the growth of the SA! network and to raise intentions to advocate for policy and
systems change. Research, infographics, and videos were used to highlight key issues affecting
Latino childhood obesity rates. Materials were disseminated via @SaludTodays social media
channels in addition to monthly e-news blasts and weekly #SaludTues Tweetchats. Facebook
ads, and Google adwords were also used boost content.

RESULTS: In 8 months, SaludTodays Twitter account generated 2,500 tweets.The network

gained 2,000 new followers and produced 17,000 engagements (all organically grown) on
Twitter. On Facebook, SaludToday gained over 3,700 new fans and produced over 27,000
engagements. Twitter and Facebook are currently the most popular tools of engagement,
however SaludTodays Instagram was the fastest-growing social media channel, with 14.08%
follower growth (July-Sept. 2016). Facebook ads and boosted posts have increased user
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1131

engagement rates.

CONCLUSION: Social media engagement is a powerful way to reach audiences. Since

implementing the use of paid ads, Facebook became the SA! webpages primary source of
social traffic. The role that SA! curators have in the SA! network has become central to the SA!
hub. Public health officials should consider allocating resources to content curation and social
media marketing.

CORRESPONDING AUTHOR: Amelie Ramirez, DrPH, UT Health San Antonio, San Antonio, TX,
78229; [email protected]
S1132 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B132 6:00 PM-7:00 PM

E-MENTORING VIA TEXT MESSAGE: A MODEL FOR MAINTAINING PROFESSIONAL

RELATIONSHIPS AND PROVIDING MENTORSHIP FROM ACROSS THE SEA

Zachary Jacobs, MD, Amy Kennedy, MD, Brian Primack, MD, PhD, Thuy Bui, MD

University of Pittsburgh Medical Center, Pittsburgh, PA

Background: International medical service affords the opportunity to work with and provide
mentorship for local physicians-in-training. However, these relationships tend to rapidly
dissipate after a brief stint abroad. Similarly, local mentor-mentee relationships may also
quickly diffuse once participants transition to a new roles and/or institutions. For this reason,
electronic mentoring (E-mentoring) may be a useful way to complement traditional face-to-
face mentoring. The aim of this study was to assess the feasibility of using a mobile
application (app) to sustain pre-existing mentoring relationships between Global Health
residents at the University of Pittsburgh Medical Center (UPMC) with local Malawian medical
students.

Methods: In 2015, we developed a mentoring group involving UPMC Global Health residents
and local medical students at the University of Malawi College of Medicine. The group met
weekly for eight weeks during our stay in Malawi. After our return to our local institution we
maintained these relationships using the text-messaging program WhatsApp. Each week we
sent inspirational quotations and clinical pearls relevant to practice in the resource-limited
setting. These quotations and pearls were tailored by each resident based on her/his in-
person experiences with the Malawi residents.

Results: There are currently a total of 31 members in the WhatsApp mentoring group. From
March 2016 until the present, there have been a total of 76 unique unsolicited student
responses to the weekly messages. This represented an average of 10 responses per month
and 3 responses to each individual teaching post. One hundred percent of the 76 student
responses were positively valenced, with responses expressing gratitude, positivity, and
interest in the information provided.

Conclusions: These findings suggest that it is feasible to leverage the text-messaging program
Whatsapp to maintain mentor-mentee relationships that end due to distance. While we used
this platform to provide weekly inspirational messages and clinical pearls with local medical
students at the University of Malawi College of Medicine, the specific nature and frequency of
communications can be tailored based on the particular needs of mentors and mentees. In
the future, we hope to provide prompts to encourage students to pose their own clinical
questions and experiences, and also to share strategies for coping with stressful situations.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1133

CORRESPONDING AUTHOR: Zachary Jacobs, MD, University of Pittsburgh Medical Center,

Pittsburgh, PA, 15213; [email protected]
S1134 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B133 6:00 PM-7:00 PM

A TWITTER ANALYSIS OF MEASLES AND VACCINATION CONVERSATION DURING THE 2015

MEASLES OUTBREAK

Loren Saulsberry, PhD 1, Vish Viswanath, PhD2

1
Harvard T.H. Chan School of Public Health, Cambridge, MA; 2Harvard T.H. Chan School of
Public Health, Boston, MA

Background: An evolving communications landscape has made social media applications like
Twitter a popular platform for conveying opinions to a broad audience. Given its increased
utility as an information exchange tool, Twitter provides insight into public views and
perceptions that could impact health-related behaviors. We evaluated Twitter discourse on
measles and vaccines during the 2015 measles outbreak to determine the evolution of
dominant themes of the online conversation and identify how Twitter was used for
communication during this critical public health event.

Methods: We analyzed over 745,000 posts on Twitter related to measles and vaccines during
the 2015 measles outbreak (December 7, 2015-May 7, 2015) using social media analysis
software provided by Crimson Hexagon. We captured relevant conversation from Crimson
Hexagons data library of all public Tweets. We established the key messages driving online
conversation, traced the evolution of those themes, and determined the online authors with
significant influence.

Results: The online conversation on measles and vaccines evolved over time with some
messages gaining popularity. Twitter discourse included vaccination advocacy, criticism of
vaccines and support for anti-vaccination movements, descriptions of personal experiences
with measles/vaccines, and news reporting. Twitter users also shared stories from news
organizations, locations of newly confirmed measles diagnoses, and information on persistent
measles cases. News and public health organizations were some of the most influential
authors on Twitter.

Conclusions: Posts on Twitter discussing measles and vaccinations became increasingly

polarized (supportive or oppositional) towards vaccination during the outbreak. Twitter was
an online news source and a sentinel for the spread of measles for its users. Twitters
capability to reach large audiences presents an enormous potential to impact public views
that influence behaviors. Since news and health organizations were influential Twitter authors
during the measles outbreak, Twitter may be an impactful channel for disseminating vital
public health information during critical, fast-moving public health emergencies. While
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1135

traditional research approaches often focus either on how messages are transmitted or
received, Twitter analysis enables real-time assessment of media messages about the
outbreak, the publics reactions to unfolding events, and the evolution of both over time.

CORRESPONDING AUTHOR: Loren Saulsberry, PhD , Harvard T.H. Chan School of Public Health,
Cambridge, MA, 02138; [email protected]
S1136 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B134 6:00 PM-7:00 PM

ADVERSE EFFECTS OF COMORBID MENTAL ILLNESS ON OVERALL HEALTH

Jennifer A. Lueck, Ph.D.

Texas A&M University, College Station, TX

Comorbidities, the bi-directional relationship of multiple disorders, have been well

documented especially for patients who suffer from mental illness. Co-occuring depression
and heart disease are the major causes of disability worldwide. Unfortunately, mental illness
is rarely detected and largely undertreated among patients who are treated for chronic
disease, increasing the likelihood of poor overall health. Recently, efforts have been made to
better assess depression and anxiety symptoms in the primary care setting in order to
contribute to a holistic treatment approach, detect underlying mental illness early, and adjust
treatment plans accordingly. Although evidence points to the complex relationship of mental
illness and chronic disease, less is known about how mental illness might contribute to the
presence of multiple disorders and vice versa. Using 2014 data from the nationally
representative Health Information National Trends Survey, the aim of the current study was
to test whether a diagnosis for depression or anxiety (Has a doctor or other health
professional ever told you that you had depression or anxiety disorder?) predict other
disorders and how symptoms of mental illness contribute to poor overall health. Preliminary
results from binary logistic regression analyses indicate a positive relationship between
mental illness and the simultaneous presence of heart disease (OR=1.58, 95%CI=1.24-2.01),
diabetes (OR=1.89, 95%CI=1.58-2.28), lung disease (OR=2.97, 95%CI=2.43-3.64), and arthritis
(OR=2.34, 95%CI=1.99-2.75). Results further suggest that symptoms of anxiety (over the past
two weeks, how often have you been bothered by not being able to stop or control
worrying?) in particular decreased a patients perceived ability to take care of ones own
health (OR=.41, 95%CI=.33-.51), reducing the likelihood of engaging in healthy lifestyle choices
such as exercise and abstaining from smoking reinforcing the cycle of poor health. Despite
these vulnerability factors, an encouraging finding emerged. Comorbid patients are not less
likely to seek professional help for medical issues highlighting the urgent need for physicians
to equip comorbid patients who seek treatment with tools to manage symptoms of mental
illness alongside other disorders. Opportunities for improving patient-provider
communication for patients managing both chronic disease and mental illness will be
discussed.

CORRESPONDING AUTHOR: Jennifer A. Lueck, Ph.D., Texas A&M University, College Station,
TX, 77845; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1137

B135 6:00 PM-7:00 PM

CONFLICTING PRIMARY LANGUAGE AFFECTS PHYSICIAN-PATIENT RELATIONSHIP IN THE

EMERGENCY DEPARTMENT

Brendan Swan, B.S.1, Melinda Chang, M.S.1, Chanel Rojas, AA1, Beatrice Konrad, BA1, Emily
Cea, BS1, Donald Edmondson, PhD, MPH2

1
Center for Behavioral Cardiovascular Health, Columbia University Medical Center, New York,
NY; 2Columbia University Medical Center, New York, NY

Communication between physicians and patients during evaluation for acute coronary
syndrome (ACS) in the emergency department (ED) is often challenging due to a number of
factors. This communication is further complicated if the primary language of the physician
does not match that of the patient. It has been shown that effective physician-patient
communication is associated with decreased patient anxiety, fewer subsequent PTSD
symptoms, and increased patient satisfaction. However, no study has tested whether
discrepancy in primary languages of physician and patient negatively impact patient
perceptions of communication quality.

We hypothesized that patients who report a different primary language than their ED
physician perceive worse communication from that ED physician.

851 participants (mean age, 60.74 (SD, 13.1); gender, 54% female; ethnicity, 57% Hispanic)
were enrolled as a part of the Reactions to Acute Care and Hospitalization (REACH) study,
which recruits ED patients being evaluated for potential ACS. Participants then reported on
the communication received from their ED physician using the physician sub-scale of the
Interpersonal Processes of Care Survey (IPCS; 14 items, e.g., How often did the doctor use
words that were hard to understand?), as well as if their primary ED physician appeared to
be a native speaker of their primary language. Age, sex, Global Registry of Acute Coronary
Events (GRACE) cardiac risk score, and Charlson comorbidity score were included as
covariates.

In a linear regression model predicting IPCS physician score [F(5,845)= 2.24, p< .05], physician-
patient first language discrepancy was associated with significantly worse perceptions of
communication quality (= .11, p= .002). No other covariates were significantly associated
with patient perceptions of communication quality.

Patients who report a different primary language than their ED physician report poorer
S1138 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

perception of physician-patient communication. Prior research has shown that poor

physician-patient communication is associated with risk for adverse psychological outcomes.
Extra effort may be required to ensure that non-native speakers feel informed and cared for
by their ED physicians.

CORRESPONDING AUTHOR: Brendan Swan, B.S., Center for Behavioral Cardiovascular Health,
Columbia University Medical Center, New York, NY, 10032; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1139

B136 6:00 PM-7:00 PM

DIGITAL STORYTELLING AS A TOOL FOR HEALTH MESSAGING ON AN AMERICAN INDIAN

RESERVATION: A DEVELOPMENT PROCESS

Maja Pedersen, M.S.1, Julia Malich, B.A.1, Kari Harris, PhD, MPH1, Brenda Bodnar, BS, RD,
CDE2, Mike Tryon, MS3, Gyda Swaney, PhD1, Sonja France, N/A1, Blakely Brown, PhD, RD1

1
University of Montana, Missoula, MT; 2Confederated Salish and Kootenai Tribes, St. Ignatius,
MT; 3Salish Kootenai College, Pablo, MT

Introduction:
Digital stories offer a collaborative approach to health messaging, and can raise awareness
among viewers on issues presented in stories. This study used an iterative, participatory
process to produce and vet digital stories for increasing community awareness about
strategies to improve child and family health on an American Indian (AI) reservation.

Methods:
We used a collaborative approach to develop digital story topics and content. This included
input from the study leadership team, with members representing a tribal college, a tribal
health organization, and a university. Story attributes, such as likability and cultural
embeddedness of characters, identification with the storyteller, and language were included
to increase influence on viewers attitudes, beliefs, and behavior change. The advisory board
and the tribal health and wellness committee were then engaged in a four-step, iterative
process (view, evaluate, prioritize, and revise) to evaluate the story. The survey was adapted
from a health promotion digital story evaluation study completed in rural Alaska. The twelve-
item, closed ended, Likert-style survey explored community members perspective on digital
storytelling likability, intent to communicate with family and friends about the topics
presented, cultural acceptability, and intent to change family health behavior. Viewers were
asked to complete a written evaluation, then to share feedback and recommendations on
how to make it more likeable, informative, and culturally relevant.

Results/Discussion:
A total of 22 community members viewed and evaluated the digital story. The viewers were
mostly female (63%), mostly AI or Alaska Native (63%), and 50% were between the ages of 30-
39. They reported liking the story (86%), feeling that it was a good way to learn about
improving the health of their families (90%), and 90% felt digital stories are a culturally
respectful way to receive health messages. Changes suggested by viewers included different
S1140 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

music, including kids voices, and adding text to the screen to emphasize the message.
Dissemination plans include social media outlets for the tribal college, tribal health
organization, and a local hospital. Study findings suggest the four-step, iterative digital story
development process may produce an effective health messaging tool for improving
community awareness of child health, and that digital storytelling can be a likeable and
culturally acceptable tool for health promotion.

CORRESPONDING AUTHOR: Maja Pedersen, M.S., University of Montana, Missoula, MT,

59812; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1141

B137 6:00 PM-7:00 PM

ENHANCING SMOKING RISK COMMUNICATIONS: THE INFLUENCE OF HEALTH LITERACY AND

MESSAGE CONTENT

Diana S. Hoover, PhD1, David W. Wetter, Ph.D.2, Damon J. Vidrine, PhD3, Nga Nguyen, M.S.1,
Summer Frank, PhD3, Yisheng Li, PhD1, Andrew Waters, PhD 4, Cathy Meade, PhD, RN, FAAN5,
Jennifer I. Vidrine, PhD3

1
University of Texas MD Anderson Cancer Center, Houston, TX; 2University of Utah, Salt Lake
City, UT; 3Stephenson Cancer Center, University of Oklahoma Health Sciences Center,
Oklahoma CIty, OK; 4Uniformed Services University of the Health Sciences, Bethesda, MD;
5
Moffitt Cancer Center, Tampa, FL

Background: For nearly half of United States adults, difficulty with health literacy interferes
with the capacity to find, process, and understand health information. Low health literacy is
associated with higher smoking prevalence, established predictors of smoking, and relapse
following smoking treatment. It has been suggested that risk communications and
interventions intended to prevent smoking and promote cessation may not reach those with
limited health literacy. Thus, it is critical that smokers with limited health literacy are provided
with understandable information about the health consequences of smoking and the benefits
of quitting. This study examined the influence of health literacy on smokers responses to
smoking health risk messages manipulated on framing (gain- vs. loss-framed) and
emotionality (factual vs. emotional).
Method: Participants (N=402; 66% male; 70% Black; 27% <high school education; 70% <
$10,000 total annual household income) were randomized to evaluate one of four sets of
smoking health risk messages (factual gain-framed, factual loss-framed, emotional gain-
framed, or emotional loss-framed). Multiple linear regression analysis was used to examine
main effects of health literacy, message emotionality, and message framing on: 1) risk
perceptions, 2) behavioral expectations (i.e., cut down, limit, quit), and 3) risk knowledge.
Two-way interactions between health literacy and message emotionality and between health
literacy and message framing were examined for these same outcomes.
Results: Smokers with higher health literacy reported significantly stronger risk perceptions
and better knowledge retention regardless of message type. A significant main effect
emerged for emotionality such that emotional (vs. factual) messages were associated with
lower risk perceptions, regardless of level of health literacy. As hypothesized, significant two-
way interactions emerged between health literacy and emotionality. Among smokers with
lower health literacy, emotion-based messages produced both higher levels of perceived risk
S1142 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

and stronger expectations for quitting, whereas among smokers with higher health literacy,
factual messages produced higher levels of these two outcomes.
Discussion: This study provides initial evidence that health literacy plays an important role in
influencing how smokers respond to different types of messages. Findings suggest that a
one-size-fits-all approach to communicating about the health risks of smoking may have
limited impact. Thus, ones health literacy should be considered when determining whether
smoking risk communications emphasize factual or emotional content.

CORRESPONDING AUTHOR: Diana S. Hoover, PhD, University of Texas MD Anderson Cancer

Center, Houston, TX, 77230; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1143

B138 6:00 PM-7:00 PM

LUNG SCREENING UNCERTAINTY AND NUMERACY PREFERENCES AMONG CURRENT AND

FORMER SMOKERS

Daniel L. Hall, Ph.D.1, Inga T. Lennes, MD2, Alaina Carr, BA3, Justin Eusebio, BA, MA4, Elyse R.
Park, PhD, MPH5

1
Harvard Medical School/MGH/BIDMC, Wellesley, MA; 2Massachusetts General Hospital
Cancer Center, Newton, MA; 3University of Colorado at Denver, Denvwr, CO; 4Massachusetts
General Hospital, Worcester, MA; 5Massachusetts General Hospital, Boston, MA

Background: Lung cancer is the leading cause of cancer death in the U.S. and is highly
attributable to smoking. Recent guidelines from the National Comprehensive Cancer Network
and American Cancer Society recommend ongoing lung screening for current and former
smokers (CS and FS) over the age of 55. Recently, Medicare began covering lung screening,
but concerns remain about its accuracy and costs. Lack of clarity about screening intent and
interpretation could undermine uptake of screening guidelines. This study thus aimed to
characterize potential sources of lung screening uncertainty (LSU) and numeracy preferences
among CS and FS.

Method: This observational, cross-sectional study examined self-report data obtained from CS
and FS undergoing lung screening at a large academic medical hospital. Two items assessed
LSU: clarity about purpose for referral (1=Not at all clear, 4=Extremely clear) and perceived
accuracy of screening (1=Not at all, 5=Extremely). Numeracy preference was assessed with
one item. Bivariate correlations examined demographic predictors of main variables. Group
differences were analyzed by independent samples t-tests, Wilcoxon Z-tests, and odds ratios
with 2-tailed Fisher exact tests.

Results: Respondents included 169 patients with a smoking history (88, 52.1% CS). CS and FS
perceived lung screening to be moderately accurate [Mtotal=3.76, SD=.72]. Mean levels were
similar between groups, p=.93. Clarity about the purpose for referral was high on average
(Mtotal =3.88, SD=.66). Greater clarity was associated with more education (r=.31, p=.001) and
receiving Medicare (r=.19, p=.048), but not smoking status (Z=1.72, p=.09). Preference for
numbers in health-related information was associated with years of education (r=.25, p=.002)
and smoking status. Specifically, FS were 2.5 times as likely to prefer numbers in health
information compared to CS (PFS=.56, OR=2.51, p=.005).
S1144 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Discussion: We present preliminary evidence that lung screening patients tend to have
favorable perceptions of the accuracy of the test and feel they understand the reason for their
referral. Values were similar between CS and FS, but varied by insurance and education
history. Notably, CS were significantly less likely than FS to favor numeric data. Taken
together, findings offer potential targets for interventions aiming to reduce LSU and to tailor
communication to suit numeracy preferences in the context of lung screening.

CORRESPONDING AUTHOR: Daniel L. Hall, Ph.D., Harvard Medical School/MGH/BIDMC,

Wellesley, MA, 02481; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1145

B139 6:00 PM-7:00 PM

UNDERSTANDING AFRICAN AMERICAN FAMILY HEALTH HISTORY COMMUNICATION: YOUNG

ADULT PERSPECTIVES

Sula Hood, PhD1, Amy Irby-Shasnmi, PhD2, Hadya Sow, BA (in progress)1, Kyle Benbow, BA1,
Jasmine Deskins, BA1, Elizabeth Adams, MPH1, Lisa Robison, MPH1

1
Indiana University-Purdue University Indianapolis Richard M. Fairbanks School of Public
Health, Indianapolis, IN; 2Indiana University-Purdue University Indianapolis Richard M.
Fairbanks School of Public Health, Indianapolils, IN

Background: Family health history (FHH) is the strongest known risk factor for many chronic
diseases. Ethnic minorities have been found to have a low awareness of their FHH, which may
pose a contributing factor to health disparities. To date, most research seeking to understand
FHH sharing within African American families has been limited to the perspective of older
adults. However, it is imperative to also gain African American young adults perspectives, as
they function as the ideal recipients of FHH information from older generations.

Purpose: This poster will present findings from a qualitative study designed to understand
African American young adults perspectives regarding FHH sharing.

Methods: We conducted semi-structured individual interviews with African American males

and females, ages 21-35. Interview questions assessed each participant's 1) perceptions about
the importance of FHH, 2) ability to define the phrase FHH, 3) knowledge of his or her
maternal and paternal family health histories, and 4) perceptions about health
communication practices in his or her immediate family, including frequency of health-related
conversations and health topics most and least discussed. We also assessed participants
perceived quality of their immediate familial relationships, as well as their interest in learning
more about FHH. Demographic characteristics were collected via written survey.

Results: A total of 38 (16 males, 16 females) African American young adults completed
interviews. Mean participant age was 28.5. Most participants were single (63.9%), and had a
college degree (41.7%) or higher (38.9%). During the semi-structured interviews, most
participants indicated that they perceived FHH to be very important, and wanted to learn
more about FHH. Most were able to correctly define the phrase FHH. Most participants
discussed having knowledge of their FHH, though many seemed to know more about their
maternal FHH than their paternal, particularly when they had a strained relationship with
their father. Mothers and grandmothers were the most commonly mentioned sources for
providing FHH information. Relatives who worked in healthcare were frequently identified as
S1146 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

FHH information disseminators. We observed mixed responses regarding the frequency of

health-related discussions within families, as some participants indicated that their families
discuss health often, while others indicated that health is rarely discussed. Nutrition was
described as a frequently discussed health topic within families, while mental illness was
described as health topic that is avoided.

Conclusion: African American young adults value FHH and want to learn more about it. Future
efforts should focus on assisting young adults with strategies to collect both their maternal
and paternal FHH, and should also provide them with strategies to initiate FHH discussions.

CORRESPONDING AUTHOR: Sula Hood, PhD, Indiana University-Purdue University Indianapolis

Richard M. Fairbanks School of Public Health, Indianapolis, IN, 46202; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1147

B140 6:00 PM-7:00 PM

ARE YOU COVERED? EXAMINING HOW KNOWLEDGE OF THE AFFORDABLE CARE ACT
INFLUENCES USE OF PREVENTIVE REPRODUCTIVE HEALTH SERVICES

Ashlee Sawyer, MS1, Eric Benotsch, PhD2

1
Virginia Commonwealth University, RIchmond, VA; 2Virginia Commonwealth University,
Richmond, VA

The Patient Protection and Affordable Care Act (PPACA) expanded access to insurance
coverage and health care services for many citizens, and has increased access for women in
particular by including preventive reproductive health services as essential health benefits.
The current national rates of sexually transmitted infections (STIs) and reproductive cancer
diagnoses serve as major areas of concern for womens health and public health. The present
study examined how knowledge of the PPACA influences receipt of preventive reproductive
health services among women (e.g., STI screening, breast exams). Women aged 18-44
(N=1083) from across the United States completed online questionnaires assessing
demographics, insurance status, preventive service use, and knowledge of the provisions of
the PPACA. A substantial portion of participants reported using preventive reproductive
health services (73.8%, N=799), including pelvic exams, breast exams, well womans visits, and
STI testing. A hierarchical logistic regression was performed to determine the independent
relationship between knowledge of the PPACA and use of preventive reproductive health
services after controlling for demographic factors (age, racial/ethnic status, relationship
status, education level, and income) and insurance status. Knowledge of the provisions of the
PPACA focused on preventive health services was associated with use of preventive
reproductive health services after controlling for these factors (OR=1.155, 95% CI = 1.021,
1.306, B=.144, SE=.063, p < .05). Results indicate that higher levels of knowledge of the PPACA
are associated with a greater likelihood of receiving cancer and STI screenings. The current
findings provide support for an increase in funds and resources to outreach and education of
the general population about the provisions and benefits of the PPACA, as well as individuals
personal health coverage plans.

CORRESPONDING AUTHOR: Ashlee Sawyer, MS, Virginia Commonwealth University,

RIchmond, VA, 23220; [email protected]
S1148 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B141 6:00 PM-7:00 PM

CAN BIG MEDIA DATA REVOLUTIONIZE GUN VIOLENCE PREVENTION?

John Ayers, PhD MA

http://www.johnwayers.com/, Chula Vista, CA

The scientific method drives improvements in public health, but a strategy of obstructionism
has impeded scientists from gathering even a minimal amount of information to address
Americas gun violence epidemic. We argue that in spite of a lack of federal investment, large
amounts of publicly available data offer scientists an opportunity to measure a range of
firearm-related behaviors. Given the diversity of available data including news coverage,
social media, web forums, online advertisements, and Internet searches (to name a few)
there are ample opportunities for scientists to study everything from trends in particular types
of gun violence to gunrelated behaviors (such as purchases and safety practices) to public
understanding of and sentiment towards various gun violence reduction measures. Science
has been sidelined in the gun violence debate for too long. Scientists must tap the big media
datastream and help resolve this crisis. To demonstrate this value we will present rich
analyses of web searches and Twitter posting for one year describing data driven conclusions
from these data.

CORRESPONDING AUTHOR: John Ayers, PhD MA, http://www.johnwayers.com/, Chula Vista,

CA, 91914-5210; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1149

B142 6:00 PM-7:00 PM

INTERVIEWS OF SUNLESS-ONLY TANNING BUSINESS OWNERS

Jessica Oleski, MA1, Alexa R. Choquette, N/A2, Joel Hillhouse, PhD3, Sherry Pagoto, PhD1

1
University of Massachusetts Medical School, Worcester, MA; 2UMass Medical School, Bolton,
MA; 3East Tennessee State University, Johnson City, TN

Indoor tanning is a group 1 carcinogen along with tobacco, radon, and arsenic. Fourteen
states have passed a ban on indoor tanning in minors and the FDA is considering a ban on
indoor tanning at the federal level. A frequently occurring argument against banning indoor
tanning is that bans will negatively impact small business owners. A counterargument is that
tanning businesses could remain viable by providing non-UV alternatives (i.e., spray/airbrush
tanning) to clientele who desire a tan. The purpose of our study was to interview sunless-only
business owners about their clientele, reasons for offering only sunless tanning, what their
clients like about sunless tanning, and about perceived threats to their business.

Participants were 15 sunless tanning business owners in the New England area including brick
and mortar salons (n=10; 5 also offered mobile services) and mobile-only businesses (n=5). On
average, owners reported being in business for 5 years (sd=3). None reported ever offering or
intentions to offer indoor tanning as a service. The cost of a service ranged from $20-$75,
discounts offered for packages. Owners said clients visit them on average 2 times per month
(sd=1) and that clients are mostly female ages 20-55. Owners estimated that on average 72%
(sd=23, range 20-99) are former indoor tanners and switched to sunless and 26% (sd=29,
range 0-90) currently use indoor tanning, which doesnt include all clients since some may not
have used indoor tanning at all. When asked what their clients like about sunless tanning, 86%
said clients liked the quality of the product (e.g., color of tan), 73% said clients liked that it
took less time than indoor tanning to achieve a tan, and 53% said clients liked that it was a
healthy alternative to tanning beds. Threats to business included other sunless-only
competition (40%), negative experience with sunless products (e.g., orange color) (20%), UV-
based competition (13%), and overhead costs (13%). All owners interviewed supported bans
on indoor tanning for minors.

Sunless tanning services appear to attract indoor tanners, former indoor tanners, and never
indoor tanners which would seem to be a potentially larger customer base than indoor
tanners. Product quality, quicker results than indoor tanning, and lack of health risks appear
to be selling points for sunless tanning over indoor tanning.
S1150 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Jessica Oleski, MA, University of Massachusetts Medical School,

Worcester, MA, 01655; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1151

B143 6:00 PM-7:00 PM

THE EFFECTS OF AGE AND COMORBIDITY ON THE LIKELIHOOD OF TRAUMATIC BRAIN INJURY
PATIENTS HIRING A HEALTH CARE ADVOCATE

Lauren E. McKinley, Student1, Laura Y. Alcaraz, N/A1, Symone McKinnon, Bachelors of Arts in
Psychology2, Terry A. Cronan, Ph.D.1

1
San Diego State University, San Diego, CA; 2San Diego State University, La Mesa, CA

Approximately 1.7 million people experience a traumatic brain injury (TBI) in the United States
each year and about 53,000 will die from TBI-related causes. The treatment options for a TBI
vary depending on severity. Patients often have complex and continuous treatment plans,
especially those who are older and have a comorbid health condition. A health care advocate
(HCA) is a trained professional who can help ensure patients optimize their treatment
outcomes by providing advisory and supportive services that simplify the navigation of the
healthcare system. The present study was conducted to determine whether age and
comorbidity (type 2 diabetes) predicted the perceived likelihood of hiring an HCA for those
with a TBI. It was hypothesized that being older and having a comorbid health condition
would predict a greater perceived likelihood of hiring an HCA. Participants were 441 randomly
selected community members, asked to complete a self-report questionnaire. A two
(comorbidity) by three (age) between subjects analysis of variance was performed on the
mean perceived likelihood of hiring an HCA. A significant main effect was found for age, F(2,
435) = 5.188, p = 0.006. A post-hoc Gabrielles test revealed that the perceived likelihood of
hiring an HCA was greater for 90 years old than 30 years old (p = 0.006) but not 60 years old.
No significant differences were detected between 30 years old and 60 years old in the
perceived likelihood of hiring an HCA. Comorbidity did not significantly predict the perceived
likelihood of hiring an HCA. These results indicate that older individuals who have sustained a
TBI report a greater perceived likelihood of hiring an HCA than do younger individuals who
have sustained a TBI. These findings may help various health care advocacy organizations
identify the specific populations most interested in using their services. Additional research is
necessary to determine the ways in which HCAs are able to help individuals who have
sustained TBIs receive optimal care and improve their overall wellbeing.

CORRESPONDING AUTHOR: Lauren E. McKinley, Student, San Diego State University, San
Diego, CA, 92120; [email protected]
S1152 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B144 6:00 PM-7:00 PM

AN EXPLORATION OF SOUTHERN FAITH LEADERS' PERSPECTIVES OF SEXUALITY AND SEXUAL

HEALTH TEACHINGS

Latrice Pichon, PhD, MPH1, Terrinieka W. Powell, PhD2, Idia B. Thurston, PhD3, Trisha Arnold,
BS4, Siri A. Ogg, BA, MS1, Andrea Williams Stubbs, MPA5, Nicole Becton-Odum, N/A6

1
University of Memphis School of Public Health Division of Social and Behavioral Sciences,
Memphis, TN; 2Johns Hopkins University Bloomberg School of Public Health, Baltimore, MD;
3
University of Memphis, Memphis, TN; 4University of Memphis Department of Psychology,
Memphis, TN; 5St. Jude Childrens Research Hospital, Memphis, TN; 6Restoring Lives through
Christ, LLC, Meridian, MS

A critical objective of the National HIV/AIDS Strategy is reducing HIV disparities in the South.
This paper describes findings from a qualitative study of southern Black faith leaders regarding
sexuality and sexual health teachings in their predominantly Black congregations in Memphis,
TN. An HIV coalition, faith-based non-profit, and university partnered to conduct four focus
groups using a semi-structured discussion guide. Faith leaders, aged 18 or older, who minister
to predominantly Black congregations were eligible to participate. Each focus group was
transcribed and then two investigators independently coded the data. The final sample
included 26 faith leaders with a mean age of 54. A broad range of denominations was
represented in the sample with over half being Baptist. Three themes emerged to describe
perspectives of sexuality and sexual health teachings among southern Black faith leaders: (1)
restricted by scripture, (2) problematic silence, and (3) tackling multiple stigmas. Churches
have been involved in HIV prevention efforts for three decades; however, few community-
academic partnerships have explored sexuality and sexual health teachings in southern
churches using a participatory approach. These results will benefit researchers when
developing faith-based interventions and help facilitate HIV prevention efforts in southern
churches.

CORRESPONDING AUTHOR: Latrice Pichon, PhD, MPH, University of Memphis School of Public
Health Division of Social and Behavioral Sciences, Memphis, TN, 38152;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1153

B145 6:00 PM-7:00 PM

COGNITIVE BEHAVIORAL THERAPY FOR BODY IMAGE AND SELF-CARE IN SEXUAL MINORITY
MEN LIVING WITH HIV: A RANDOMIZED CONTROLLED TRIAL

Aaron Blashill, Ph.D. 1, Steven A. Safren, PhD2, Sabine Wilhelm, PhD3, Jonathan Jampel, N/A4,
Wade Taylor, PHD5, Conall O'Cleirigh, PhD6, Kenneth H. Mayer, MD7

1
San Diego State University, San Diego, CA; 2Department of Psychology, University of Miami,
Coral Gables, FL; 3MGH/Harvard Medical School, Boston, MA; 4Clark University, Worcester,
MA; 5Wheelock College, Boston, MA; 6Massachusetts General Hospital/Harvard Medical
School, Boston, MA; 7Fenway Health, Boston, MA

Objective: Body image disturbance is a distressing and interfering problem among many
sexual minority (e.g., gay and bisexual) men living with HIV, and is associated with elevated
depressive symptoms, and poor HIV self-care (e.g., antiretroviral therapy (ART) non-
adherence). However, no known interventions exist that address body image disturbance and
ART adherence among this population. Thus, the current study tested the preliminary efficacy
of a newly created intervention: cognitive behavioral therapy body image and self-care (CBT-
BISC) for this population.

Methods: The current study was a two-arm randomized controlled trial (N = 44) comparing
CBT-BISC to an enhanced treatment as usual (ETAU) condition. Analyses were conducted at
post-treatment (3 months post-baseline) and follow-up (6 months post-baseline). The primary
outcome was body image disturbance, and secondary outcomes were ART adherence
(electronically monitored via Wisepill), depressive symptoms, and global functioning.

Results: At post-treatment, the CBT-BISC condition showed substantial improvement in body

image disturbance (B = -13.6, SE = 2.7, 95% CI: -19.0, -8.3, p < .001; d= 2.39), depressive
symptoms (B = -4.9, SE = 2.8, 95% CI: -10.6, .70, p = .086; d= .87), ART adherence (B = 14.7, SE
= 4.9, 95% CI: 5.2, 24.3, p = .003; d= .94), and global functioning (B = 12.3, SE = 3.2, 95% CI:
6.1, 18.6, p < .001; d= 2.91). Results were generally maintained, or improved, at 6-month
follow-up.

Conclusions: CBT-BISC shows preliminary efficacy in the integrated treatment of body image
disturbance and HIV self-care behaviors among sexual minority men living with HIV.
S1154 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Aaron Blashill, Ph.D. , San Diego State University, San Diego, CA,
92120; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1155

B146 6:00 PM-7:00 PM

CURVILINEAR ASSOCIATION OF DEPRESSION WITH SEXUAL RISK BEHAVIOR AMONG HIV-

INFECTED MEN WHO HAVE SEX WITH MEN

Jacklyn Babowitch, B.S.1, Peter A. Vanable, PhD1, Shannon M. Sweeney, PhD2

1
Syracuse University, Syracuse, NY; 2Le Moyne College, Kirkville, NY

Depression is highly prevalent among HIV+ men who have sex with men (MSM) and may
contribute to risky health behaviors. However, research linking depression to sexual risk
among HIV+ MSM has yielded mixed findings, and has focused primarily on documenting the
linear association of depressive symptoms to elevated rates of unprotected anal intercourse
(UAI). In this study, we test the hypothesis that depressive symptoms will show a curvilinear
association to UAI, such that depression will be more strongly linked to risky sex in HIV+ MSM
with moderate levels of depression. HIV+ MSM recruited during outpatient medical care (N =
96, M age = 44, 57% Caucasian) completed the Center for Epidemiological Studies
Depression Scale (CES-D) and self-report measures of sexual risk behaviors. The primary study
hypothesis concerning the nonlinear association of depression to sexual risk was tested using
logistic regression models in which CES-D scores were entered into Model 1, and both CES-D
and squared CES-D terms were included in Model 2. Participants reported high levels of
depressive symptoms (M = 20.15, SD = 12.94), and 34% reported at least one instance of UAI
in the last three months. Neither model explained any variance in UAI with a primary partner.
For encounters involving a casual partner, CES-D scores were not associated with UAI at the
last occasion of sex in model 1. However, CES-D (b = .28, SE = .12, p = .02) and squared CES-D
terms (b = -.005, SE = .002, p = .04) in model 2 were significantly associated with UAI at the
most recent sexual episode. Similarly, CES-D scores were not associated with any instances of
UAI for the previous three-months in model 1. However, when CES-D (b = .21, SE = .09, p =
.02) and squared CES-D terms (b = -.004, SE = .002, p = .03) were entered into model 2, both
were significantly associated with instances of UAI. Further, a comparison of model chi-square
values indicated that model 2 fit the data better than model 1 with regard to UAI at the most
recent sexual episode (c2= 2.32, p = .13 versus c2= 7.98, p = .02) and for the last three months
(c2= .04, p = .53 versus c2= 7.30, p = .03). In this sample of HIV+ MSM, findings confirm a
curvilinear relationship between depression and sexual risk-taking. Treating depressed mood
in HIV+ MSM may not only improve mental health functioning, but also lead to reductions in
sexual transmission risk.

CORRESPONDING AUTHOR: Jacklyn Babowitch, B.S., Syracuse University, Syracuse, NY, 13244;
[email protected]
S1156 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B147 6:00 PM-7:00 PM

DEVELOPING AND EVALUATING EHEALTH VIDEO TO INCREASE HIV TESTING AMONG HARLEM
AREA PATIENTS WHO INITIALLY DECLINE

Ian David. Aronson, Ph.D.1, Alex S. Bennett, PhD; MPA; MS2, Honoria Guarino, PhD3, Lisa
Marsch, PhD4, Theodore Bania, MD,MS5

1
NDRI, Inc., New York, NY; 2National Development and Research Institutes, Inc, New York, NY;
3
National Development and Research Institutes, Inc., New York, NY; 4Dartmouth College,
Lebanon, NH; 5Icahn School of Medicine at Mt Sinai, NY, NY

Undiagnosed HIV remains a significant problem in the United States. Many of those most at
risk may have limited access to primary care, testing, and related health education. Despite
federal recommendations of routine HIV testing in all healthcare settings, many emergency
department (ED) patients are not offered HIV testing, and among those who are offered
testing far more decline compared to those who accept.

In a recent trial of a tablet-based eHealth intervention designed to increase HIV testing among
ED patients who declined HIV tests offered by hospital staff (N=300), roughly 30 percent
agreed to an HIV test after watching a brief ( < 2 minute) video. The videos were developed in
accordance with the Information, Motivation, Behavioral Skills model, and the intervention
randomized participants into groups shown different videos to examine which would be more
efficacious. Videos depicted either a young, male African American doctor describing the
importance of HIV testing to an African American male patient, or a young male African
American community member who discloses that he is HIV positive on camera and urges
participants to test. Although test rates did not differ by treatment group or participant race,
qualitative interviews with patients (n=40) in the Harlem-are hospital where data were
collected indicate differing responses to the videos. Most interviewed participants responded
strongly in favor of the community member, yet some White and Asian patients used words
like stereotypical to describe the community members appearance (he wore his own
clothing in the video and styled himself). African American and Latino participants spoke
highly of the community member and largely did not mention his physical appearance. Results
indicate patient populations may have very different reactions, by demographic group, to the
same intervention video. Further research is warranted to optimize videos to further increase
patient acceptance and to better facilitate behavior change (e.g. HIV testing among patients
who initially decline) among those most at risk.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1157

CORRESPONDING AUTHOR: Ian David. Aronson, Ph.D., NDRI, Inc., New York, NY, 10010;
[email protected]
S1158 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B148 6:00 PM-7:00 PM

IMPAIRED PERFORMANCE ON THE HIV-DEMENTIA SCALE PREDICTS SURVIVAL OVER UP TO

THIRTEEN YEARS

Nikhil Banerjee, B.A.1, Gail Ironson, M.D., Ph.D1, Bonnie Levin, Ph.D2

1
University of Miami, Coral Gables, FL; 2University of Miami Neuropsychology Division, Miami,
FL

Objective: There is a lack of empirical evidence linking HIV-associated neurocognitive

impairment (NCI) with mortality in the era of widespread antiretroviral (ARV) medication
availability. This study sought to examine whether global NCI, as defined by the HIV-Dementia
Scale (HDS), was associated with mortality among patients predominantly in the midrange of
their illness.

Methods: Between 1997 and 2001,210 HIV-positive men and women (MCD4 = 323.02
cells/mm3; Mage = 37.69; 72% male; 67% attending some college or more; 31% non-Hispanic
White, 35% non-Hispanic African American, and 30% Hispanic) were recruited, administered
the HDS, and followed for up to 13 years (1997-2010). Subjects were excluded if they had ever
experienced an AIDS-defining illness or had baseline CD4 cell counts outside the range of 150-
700 cells/mm3, in order to capture individuals predominantly in the mid-range of their illness
(92% had CD4 cell counts between 150 and 500 cells/mm3). Subjects were dichotomously
classified as either impaired (HDS<10) or unimpaired (HDS > 10). Cox proportional hazard
regression analysis was used to test whether global NCI predicted survival.

Results: Thirty-six (17%) subjects scored in the impaired range at baseline, and 59 subjects
(28%) died during the studys follow-up period. Participants scoring in the impaired range at
baseline were significantly more likely to die during the study period compared with their
unimpaired counterparts (HR: 2.06; 95% confidence interval, 1.09, 3.90) controlling for age,
sex, race, education, and baseline CD4 cell count, viral load, and antiretroviral medication
status.

Conclusion: These findings demonstrate that the HDS can be utilized as brief screening
measure that offers potential prognostic utility for assessing mortality risk in HIV-positive
adults during the era of AVR availability. Subjects scoring in the impaired range on the HDS
faced two times the risk of dying compared with non-impaired subjects during the 13-year
study period. This was independent of demographics and key illness variables previously
associated with mortality (i.e., disease status and ARV treatment status at study entry). These
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1159

data also add to the literature linking NCI with mortality in HIV, especially among patients in
the mid-range of their illness, during the era of ARV medication availability.

CORRESPONDING AUTHOR: Nikhil Banerjee, B.A., University of Miami, Coral Gables, FL,
33124; [email protected]
S1160 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B149 6:00 PM-7:00 PM

INCARCERATION, SEXUAL BEHAVIORS, AND HIV INFECTION AMONG WOMEN AT INCREASED

RISK OF HIV INFECTION, 20 UNITED STATES CITIES

Akilah Wise, PhD, Teresa Finlayson, PhD, Lina Nerlander, BMBCh, Catlainn Sionean, PhD,
Gabriela Paz-Bailey, PhD

Centers for Disease Control and Prevention, Atlanta, GA

Background: Women who are involved in the criminal justice system experience multiple risk
factors that increase the likelihood of acquiring HIV infection. We evaluated the prevalence of
incarceration and compared behaviors among women with and without a history of
incarceration.

Methods: We used data from the 2013 National HIV Behavioral Surveillance system in 20
cities among heterosexuals at increased risk for HIV infection. Participants completed a survey
and were tested for HIV infection. We evaluated the association between incarceration and
the following past 12 months outcomes: exchange sex, multiple (3) casual sex partners, and
multiple (3) casual condomless sex partners. We also evaluated associations of incarceration
with HIV prevalence and with last partner characteristics (ever injected drugs, ever used
crack/cocaine, ever incarcerated). Log-linked Poisson regression models, adjusted for
demographics and clustered on recruitment chain, with generalized estimating equations
were used to estimate adjusted prevalence ratios (aPR) and 95% confidence intervals.

Results: Of 5,154 women, 11% were incarcerated within the previous year, 36% were ever
incarcerated but not in the past 12 months, and 53% were never incarcerated. Prevalence of
exchange sex (aPR 1.32, 1.20 1.46), multiple casual partners (aPR 1.59, 1.2 2.1), multiple
casual condomless partners (aPR 1.47, 1.07 2.03), and HIV infection (aPR 1.66, 1.09 2.52)
were all higher among recently incarcerated women compared to those never incarcerated.
Recently incarcerated women were more likely to have a last partner who injected drugs (aPR
2.05, 1.66 2.53), used crack/cocaine (aPR 2.08, 1.79 2.41) and had himself been
incarcerated (aPR 1.39, 1.29 1.50).

Conclusions: Nearly half of women in our study had been incarcerated. Recent incarceration
was associated with several sexual risk behaviors and partner characteristics that increase the
risk of HIV. Incarcerated women were also more likely to be HIV-positive. HIV prevention,
testing, and early treatment among women with a history of incarceration can maximize the
effectiveness of the public health response to the HIV epidemic.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1161

CORRESPONDING AUTHOR: Akilah Wise, PhD, Centers for Disease Control and Prevention,
Atlanta, GA, 30329; [email protected]
S1162 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B150 6:00 PM-7:00 PM

LIPODYSTROPHY AND BODY CHANGE DISTRESS IN OLDER WOMEN WITH HIV

Greer A. Raggio, PhD, MPH1, Gregory Robbins, MD1, Elyse R. Park, PhD, MPH1, Sara Looby,
PhD1, Steven A. Safren, PhD2, Christina Psaros, PhD3

1
Massachusetts General Hospital, Boston, MA; 2Department of Psychology, University of
Miami, Coral Gables, FL; 3Massachusetts General Hospital/Harvard Medical School, Boston,
MA

Background: HIV and some antiretroviral drugs are associated with lipodystrophy, an
abnormal redistribution of body fat marked by increased abdominal adiposity and/or loss of
subcutaneous fat stores. Other risk factors include older age, genetic factors, and longer
treatment duration, making this issue salient to those living with HIV for long periods.
Lipodystrophy is associated with body image dysphoria, decreased quality of life, and lower
adherence to HIV treatment.

Objectives: (1) To describe body change and related distress, and (2) to examine associations
among body change distress, depression, and health among older women living with HIV.

Methods: Data were collected as part of a cross-sectional study examining psychosocial

concerns of older women with HIV. Participants (N=64) completed a demographic survey and
self-report measures of HIV-related body change, depression, and health status. Independent
samples t-tests and Pearsons r correlations were calculated.

Results: Average age was 52.4 years (SD=6.3), 58.7% were African American/Black, and 34.9%
had < high school education. Mean time since diagnosis was 13.6 years (SD=7.3). Among those
reporting current HIV treatment (n=60), common body changes included increased belt/waist
size (57.6%), decreased fat in the buttocks (35.6%) and arms/legs (30.5%), sunken cheeks
(29.3%), and fat deposits in the neck (23.7%). Women reporting more body change distress
reported significantly more depressive symptoms (r=.581, pr=.427, p=.001). Increases in
belt/waist size were significantly associated with higher depression scores (t[57]=2.33,
p=.023).

Conclusions: Fat redistribution, particularly increased abdominal adiposity, was common in

this sample of older women living with HIV, and was associated with more severe depressive
symptoms. Women endorsing body change distress also noted worse physical pain. It is
imperative to develop interventions for this population given the growing number of aging
women with HIV and the relation between lipodystrophy and health outcomes. Interventions
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1163

that combine aspects of depression/chronic pain management and physical activity

promotion may hold promise.

CORRESPONDING AUTHOR: Greer A. Raggio, PhD, MPH, Massachusetts General Hospital,

Boston, MA, 02114; [email protected]
S1164 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B151 6:00 PM-7:00 PM

NATIONAL HIV BEHAVIORAL SURVEILLANCE AMONG MEN WHO HAVE SEX WITH MEN IN THE
US:A COMPARISON OF WEB-BASED AND VENUE-BASED SAMPLES

Yen-Tyng Chen, PhD1, Kristina Bowles, MPH2, Qian An, PhD2, Elizabeth DiNenno, PhD2, Teresa
Finlayson, PhD2, Brooke Hoots, PhD2, Gabriela Paz-Bailey, PhD2, Cyprian Wejnert, PhD2

1
Centers for Disease Control and Prevention, Decatur, GA; 2Centers for Disease Control and
Prevention, Atlanta, GA

Objectives: To describe the difference in sociodemographic characteristics and risk behaviors

between men who have sex with men (MSM) recruited using two different approaches: web-
based (WB) and venue-based (VB) sampling.
Methods: We compared the 2012 Web-based HIV Behavioral Survey (WHBS), which recruited
MSM online (e.g., direct marketing on Facebook), and the 2011 National HIV Behavioral
Surveillance (NHBS), which sampled MSM from physical MSM venues. Chi-square tests were
used to compare sociodemographic characteristics. Poisson regression with generalized
estimating equations were used to estimate prevalence ratios for each behavioral outcome
adjusting for sociodemographic characteristics.
Results: There were 3221 WB and 9256 VB sexually active MSM included in analyses.
Compared to the VB sample, the WB sample was slightly older, less likely to be black or
Hispanic and from the South, and more likely to have higher income, more education, and
health insurance. WB MSM were more likely than VB MSM to self-identify as gay/homosexual
(aPR=1.05, CI: 1.03-1.07), ever socialize on the Internet (aPR=1.19, CI: 1.15-1.24), have met
their last partner online (aPR=2.41, CI: 2.24-2.59), have anal intercourse without a condom
with last male sex partner (aPR=1.21, CI: 1.15-1.29). The WB MSM were significantly less likely
to have multiple male sex partners (aPR=0.92, CI: 0.88-0.96), inject drugs (aPR=0.60, CI: 0.47-
0.78), test for HIV in the past 12 months (aPR=0.82, CI: 0.79-0.85), report HIV-negative status
(aPR=0.95, CI: 0.93-0.97), receive free condoms (aPR=0.86, CI: 0.81-0.92), and experience
physical assault (aPR=0.86, CI: 0.81-0.92) than VB MSM.
Conclusions: WB MSM were less likely to be minority MSM and had higher SES status. We also
found differences in sexual and drug using behaviors and HIV testing. The method for
recruiting MSM into a behavioral survey should consider the segment of MSM to be reached.

CORRESPONDING AUTHOR: Yen-Tyng Chen, PhD, Centers for Disease Control and Prevention,
Decatur, GA, 30033; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1165

B152 6:00 PM-7:00 PM

PATIENT-LEVEL BARRIERS AND FACILITATORS OF CERVICAL CANCER SCREENING AMONG HIV-

POSITIVE WOMEN IN INDIA

JK Kosambiya, M.D.1, Vipul Chaudhari, M.D.1, Apurva Barve, M.A.2, Janna Gordon, M.S. 3,
Ambuj Kumar, M.D., MPH4, Sukesha Gamit, M.D.1, Kristen Wells, Ph.D., MPH5

1
Department of Community Medicine (PSM), at the Government Medical College, Surat,
Surat, Gujarat, India; 2University of California, Los Angeles, Los Angeles, CA; 3SDSU/UC San
Diego Joint Doctoral Program in Clinical Psychology, San Diego, CA; 4Morsani College of
Medicine, University of South Florida, Tampa, FL; 5San Diego State University, San Diego, CA

Although the risk of cervical cancer (CC) is higher in women living with human
immunodeficiency virus (HIV), there is no national policy for CC screening in India. Thus, many
women living with HIV in India do not get regular CC screening. Based on the Social Ecological
Model, this novel qualitative study evaluated patients and health care providers perceptions
of patient-level factors associated with receipt of CC screening among women living with HIV.
Using a standardized guide, in-depth interviews were conducted with 25 women living with
HIV (mean age: 37.2 years) and 15 health care providers (mean age: 36.1 years; 66.6% female)
recruited from New Civil Hospital in Surat, India. Interviews were audio recorded, transcribed
verbatim, translated if necessary, and analyzed using content analysis. Participants noted that
women with HIV were generally motivated to stay healthy but had significant gaps in
understanding of CC and CC screening. Fear of obtaining the Pap test and its results were also
noted by some participants as barriers. Some participants indicated that it was difficult for
women to take time away from family, work, or household responsibilities, and this made it
difficult to get CC screening. Finally, most interview participants stated that women were very
interested in learning more about CC screening. Results obtained from this study, combined
with data assessing organizational, interpersonal, community, and societal barriers and
facilitators to CC care, will be used to develop intervention strategies to assist women living
with HIV in getting CC screening

CORRESPONDING AUTHOR: JK Kosambiya, M.D., Department of Community Medicine (PSM),

at the Government Medical College, Surat, Surat, Gujarat, n/a; [email protected]
S1166 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B153 6:00 PM-7:00 PM

BEHAVIORAL HEALTH INTEGRATION IN 3 PRIMARY CARE CLINICS: QUALITATIVE EVALUATION

& STORIES FROM THE FRONT LINES

Julie E. Richards, MPH1, Amy K. Lee, MPH2, Paula Lozano, MD, MPH2, Carol Achtmeyer, MN,
ARNP2, Emily Williams, PhD, MPH3, Santino Telles, MA4, Jessica Ridpath, BA2, Caldeiro M.
Ryan, MD5, Ludman J. Evette, PhD2, Rebecca L. Parrish, MSW, LICSW6, Kathy A. Bradley, MD,
MPH2

1
Group Health Research Institute and University of Washington Department of Health
Services, Seattle, WA; 2Group Health Research Institute, Seattle, WA; 3VA Puget Sound Health
Services Research & Development and University of Washington Department of Health
Services, Seattle, WA; 4Group Health, Northgate Medical Center, Seattle, WA; 5Group Health
Behavioral Health Services, Seattle, WA; 6Group Health Behavioral Health Support Services,
Seattle, WA

Background
Health care systems nationwide are implementing routine behavioral health screening,
including screening for depression and substance use, in response to mounting evidence
supporting benefits of integrating behavioral health in primary care. Between 3/2015 and
3/2016, 3 pilot primary care clinics in a large Pacific Northwest health care system
sequentially implemented behavioral health integration [BHI]. This included routine screening
and diagnostic assessment for depression, alcohol, marijuana and drug use, and same-day
referrals to social workers [LICSWs] for patients with acute care needs (i.e. suicide risk,
substance-use issues). Screening rates increased from 14-35% for depression and 5-13% for
substance use, prior to BHI, to 60-82% for both, following BHI. This qualitative study
evaluated the multi-pronged implementation strategy.

Methods
We used the Greenhalgh Model for diffusion of innovations to assess barriers to and
facilitators to behavioral health screening and assessment, two key elements of BHI. The
implementation strategy included: use of automated electronic health record [EHR] systems
to prompt and document care, partnership between front line clinical operations and a
leadership team, and tools that explicitly addressed frontline staff attitudes and skills.
Research team members assessed barriers and facilitators to implementation in weekly
formative evaluation meetings.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1167

Results
Facilitators of successful implementation included: perceived value of the
screening, diagnostic assessments and EHR tools/prompts by clinicians (Innovation), hands-on
participation in implementation by frontline staff (Linkage), and use of LICSWs to provide
same-day behavioral health care (Resources). Barriers included: fragmented clinic
communication and work flow, clinic quality improvement process limitations (Infrastructure),
and competing priorities for clinicians time during the patient visit (Readiness).

A key facilitator was sharing stories from frontline staff adopters with new sites. For example,
one MA described, When we were first approached about BHI, we all thought oh, more
paperwork, more work for us it turned into this very powerful role, making us feel like its
not just about paperwork, its us being allowed to get closer to the patient... I am able to be a
big part of somebodys betterment. Its rewarding on such a deep level. Other stories
included identification of heavy alcohol use by a diabetic man at risk of a second foot
amputation, and acutely suicidal patients being treated for other common conditions (e.g.
influenza, wart removal).

Conclusion
This study identified key facilitators and barriers of BHI implementation in 3 clinics in which
these strategies were piloted. Findings were used to refine the implementation strategy for
the health care system.

CORRESPONDING AUTHOR: Julie E. Richards, MPH, Group Health Research Institute and
University of Washington Department of Health Services, Seattle, WA, 98101;
[email protected]
S1168 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B154 6:00 PM-7:00 PM

SLEEP PROBLEMS ARE RELATED TO POOR MEDICATION ADHERENCE AMONG YOUNG ADULTS

Lauren Strainge, M.A., Dean Cruess, PhD

University of Connecticut, Storrs, CT

Suboptimal medication adherence is common among individuals with chronic medical

conditions, and is also associated with poorer health outcomes, decreased quality of life, and
increased morbidity and mortality. For emerging young adults, transition to college and
corresponding increases in independence are associated with declines in medication
adherence. College students are also likely to report insufficient and significantly disrupted
sleep patterns, which has been shown to increase risk for medication non-adherence in
middle-aged and older adults. The purpose of this study was to examine the relationship
between sleep quality and daily medication adherence among a representative sample of
college students. We sampled 785 undergraduates at a large state university (mean age =
18.89; 84% female; 28% non-white) who were prescribed a daily medication (including oral
contraceptives) and had them complete an online cross-sectional survey assessing self-
reported sleep patterns and medication adherence. Sleep was assessed using the Pittsburgh
Sleep Quality Index, and medication adherence was assessed using the Morisky Medication
Adherence Scale. Results showed that 78.4% of participants reported fairly bad or very bad
sleep quality over the past month, and 72.6% reported problems with medication adherence.
Pearson's correlation coefficients were calculated to determine the relationships between
sleep quality, hours of sleep, and daytime functioning with medication adherence. We also
conducted a series of hierarchical linear regression analyses predicting medication adherence
from each of our sleep variables. Adherence scores were positively correlated with sleep
duration (r = 0.419, p < 0.001), sleep quality (r = 0.461, p < 0.001), and daytime functioning
scores (r = 0.434, p < 0.001). Longer sleep duration (b = 0.143, p < 0.001), and greater daytime
functioning (b = 0.099, p = 0.007) significantly predicted better medication adherence, even
after controlling for participant age, race, and gender. These results indicate that insufficient
sleep and related daytime dysfunction were significant risk factors for poor medication
adherence in our college sample. These findings also suggest that interventions that address
poor sleep among college students may also have a positive effect on their medication
adherence.

CORRESPONDING AUTHOR: Lauren Strainge, M.A., University of Connecticut, Storrs, CT,

06269; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1169

B155 6:00 PM-7:00 PM

ACADEMIC ACHIEVEMENT AND DEPRESSIVE-LIKE SYMPTOMS AMONG ADOLESCENT

CAREGIVERS FOR GRANDPARENTS WITH AND WITHOUT DEMENTIA

Gabriel Holguin, BA1, Ruby Ceullar, BA2

1
California State University San Marcos, san marcos, CA; 2California State University San
Marcos, San Marcos, CA

Alzheimers disease remains the sixth leading cause of death in the United States, and rates
are predicted to increase after age 65. Common symptoms include loss of interest in activities,
trouble overcoming difficulties, self-harm, feelings of shame, and cognitive decline. However,
little research has explored dementia as a developmental disease and examined the academic
abilities and cognitive function of younger individuals who may be genetically predisposed by
a family history of dementia. This study seeks to elucidate a link between a family history of
dementia and cognitive functioning such as academic performance and depressive-like
symptoms among adolescents by comparing those who report caregiving for a grandparent
with and without dementia. Adolescents who reported caregiving for a grandparent with
(n=19) and without (n=19) dementia, grades 9 through 12, were matched on demographics
and were compared for differences in self-reported GPA, perceived stress, shame, and
depressive symptoms. A paired samples t-test revealed no academic grade differences
between the dementia and control group, p=0.27, nor was there a difference in overall stress
or depressive symptoms, p=0.12. Given previous literature linking dementia with trouble
overcoming difficulties, loss of interest in activities, and increased self-harm, an analysis of
individual items concerning these topics were made. The dementia group (M=3.19, SD=1.28)
reported more trouble overcoming monthly difficulties compared to the control group
(M=2.13, SD=1.26), p=0.02. They also reported increased loss of interest in usual activities
(M=0.94, SD=1.16) compared to the control group (M=0.22, SD=0.55), p=0.01. The dementia
group (M=1.72, SD=0.75) showed a trend of increased shame compared to the control group
(M=1.33, SD=0.41), p=0.05. Lastly, there was a trend such that the dementia group (M=0.76,
SD=1.35) reported more frequent thoughts of self-harm compared to the control group
(M=0.06, SD=0.24), p=0.05. These findings suggest that a family genetic predisposition of
dementia may increase dementia related symptoms, such as experiencing trouble overcoming
daily difficulties and a loss of interest in usual activities, at a younger age. Limitations of the
present study notwithstanding, further research to increase our understanding of dementia as
a developmental disease that may influence cognition and behavior as early as adolescence is
recommended.
S1170 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Gabriel Holguin, BA, California State University San Marcos, san
marcos, CA, 92078; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1171

B156 6:00 PM-7:00 PM

CAREGIVER RELATIONSHIPS AND GENERAL SUPPORT RELATED TO DEPRESSIVE SYMPTOMS

AND BMI

Monique Griffiths, B.S.1, Jessica Chiang, Ph.D2, Shu-Sha Angie Guan, Ph.D3

1
California State University, Northridge, Sunrise, FL; 2Northwestern University, Evanston, IL;
3
California State University, Northridge, Northridge, CA

Supportive relationships have been associated with decreased risk for poor mental and
physical health outcomes (Holt-Lunstad et al., 2010; Oliveira et al., 2013). However, the
relationship between specific sources of social support at different developmental time
periods and health-related outcomes is less clear. Thus, we examined the linkages between
support from caregivers early in life, current social support from friends, and current general
support, and makers of physical health (i.e., total-body adiposity as a risk factory for a number
of chronic diseases) and mental health (i.e., depressive symptoms) among young adults.

Undergraduate students (N = 103, Mage=19.91, SD=1.91) from diverse backgrounds

(Asian American = 59, Latino = 21, Other ethnicity = 23) reported on their early caregiver
relationships (e.g., During childhood and adolescence, my primary caregiver was
affectionate to me), current friend social support (e.g., I could count on my friend when I
needed to talk), and current general support (I have someone who understands my
problems). They also reported on their depressive symptoms (e.g., you felt lonely), and
height/weight, which was used to calculated body mass index (BMI), a validated indicator of
total-body adiposity.

Multiple regression analyses controlling for sociodemographic variables (i.e., ethnicity,

parent education, and age) revealed that higher current friend support and general support
were associated with lower depressive symptoms (bs = -.21, -.26, SE = .06, p < .001). By
contrast, support from primary caregivers early in life was not associated with depressive
symptoms (b = -.10, SE = .08, p = .24). When all three forms of relationship were entered into
the same model, only general support (i.e, having someone), but not support from specific
source, was associated with depressive symptoms (b = -.20, SE = .07, p = .01). In terms of
physical health, friend and general support were only marginally predictive of BMI (bs = -.77, -
89, -.26, SE = .46, .49, p < .10).

These results suggest that having a friend or someone else to rely on from support may
decrease risk of depression and perhaps adiposity. Supportive relationships during childhood
and adolescence may be less consequential. This may be due to the protective role of social
S1172 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

support in the effects of stress. Stressful events often precede depressive episodes, and
support is known to be an effective coping mechanism. Current support may be immediately
helpful for dealing with current stress whereas support in the past may be less relevant.

CORRESPONDING AUTHOR: Monique Griffiths, B.S., California State University, Northridge,

Sunrise, FL, 33313; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1173

B157 6:00 PM-7:00 PM

COGNITIVE BEHAVIORAL THERAPY COMBINED WITH PHYSICAL EXERCISE FOR DEPRESSION IN

ADULTS WITH CHRONIC ILLNESSES: A META-ANALYSIS

Paquito Bernard, PhD1, Gregory Moullec, PhD2

1
Universit du Qubec Montral, Montreal, PQ, Canada; 2Public Health School, University of
Montreal, Montreal, PQ, Canada

Interventions combining cognitive behavioral therapy and physical exercise (CBTEx) have been
shown to be efficacious at reducing depressive symptoms in various chronic illnesses. The
objectives of this review were to systematically examine the absolute efficacy of CBTEx
interventions, and then relative to usual care, CBT alone or exercise alone on depressive
symptoms reduction in adults with chronic illnesses. Database searches identified randomized
controlled trials of people with chronic illnesses (e.g., cancer, heart failure, chronic fatigue)
participating in CBTEx interventions compared with control (usual care, waiting list) or active
comparators (e.g., exercise, CBT alone). Relevant studies, published before July 2016, were
traced through a systematic search of Pubmed, Embase, PsyInfo, Cinahl, SportDiscus and
Cochrane Library databases. Standardized mean differences (SMD) were calculated for each
intervention arm and control comparison. The pooled SMD associated to CBTEx (19 trials,
2190 participants) was -0.24 (95%CI: -0.43 to -0.06; I2= 56%). When compared with usual care
(11 trials, 1256 participants), the magnitude of effect was greater (SMD = -0.37, [CI: -0.58 to -
0.17]); I2 = 56%). The reduction in depressive symptoms was not statistically different between
CBTEx and exercise (SMD = -0.26, [CI: -0.74 to 0.23], 7 trials, 623 participants) or between
CBTEx and CBT (SMD = -0.08, [CI: -0.24 to 0.07], 10 trials, 845 participants). Findings suggest a
short-term effect of CBTEx on depression in adults with chronic illness. However, CBTEx does
not seem to be superior to exercise or CBT alone for decreasing depressive symptoms. Further
research should investigate the potential moderators of interventional efficacy, such as
participants adherence, the methodological quality and selected sample characteristics.

CORRESPONDING AUTHOR: Paquito Bernard, PhD, Universit du Qubec Montral,

Montreal, PQ, H2X 1YK; [email protected]
S1174 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B158 6:00 PM-7:00 PM

DISTRESS INTOLERANCE AS A MODERATOR OF THE BENEFIT OF PHYSICAL ACTIVITY:

OUTCOMES FOR DEPRESSION

Kristin Szuhany, M.A.1, Michael Otto, Ph.D.2

1
Boston University, New York, NY; 2Boston University, Boston, MA

Negative affect and distress intolerance (DI) are linked to a wide variety of negative health
behaviors. For example, DI has been linked to overeating, underactivity, smoking, and illicit
drug use behaviors (Otto et al., 2016). There is strong evidence from both laboratory and
clinical studies that physical activity (PA) interventions are efficacious for both improving
mood and rapidly reducing DI (e.g., Smits et al., 2008), and that PA may have particular
efficacy for those with high DI. For example, Smits found that the benefits of a program of PA
for smoking cessation were specific to smokers with high DI (measured by anxiety sensitivity;
Smits et al., 2016). In the present study, we examine the role of PA as an adjunctive
treatment for major depression, and likewise examine whether mood benefits were
moderated by baseline level of DI (measured with the Distress Intolerance Index; DII, McHugh
& Otto, 2012).

A total of 32 sedentary depressed outpatients (M age=33.713.4, 78% female) were

randomized to receive 9 sessions of a behavioral activation treatment in combination with
either PA (n=15) or a control treatment (stretching; n=17) over the course of 12 weeks. These
augmentation strategies involved 6 weekly 30-minute sessions in which participants planned
exercise/stretching for the week and received motivational interventions (Otto & Smits, 2011)
to improve adherence to their respective regimens.

Analyses indicated the total sample was in the moderate range of depression (27.010.2) and
distress intolerance (DI; 19.18.8) at baseline. Patients reported engaging in 359 metabolic
equivalent (MET) minutes of vigorous and moderate intensity exercise at baseline (about 81
minutes of moderate intensity exercise). In a mixed effects linear regression, results indicated
that regardless of PA condition, patients improved significantly with treatment (p < .001).
Also, those in the adjunctive PA condition achieved a significantly greater reduction in DI (p <
.02). In addition, the degree of benefit was moderated by baseline DII score; those with high
DI who were treated with adjunctive PA achieved significantly more mood benefit over time
(p < .01). These results are consistent with recent findings identifying DI as a transdiagnostic
factor that may help explain comorbidity and may moderate the benefits of treatment for
conditions ranging from depression and smoking to Type II Diabetes, HIV, and obesity.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1175

CORRESPONDING AUTHOR: Kristin Szuhany, M.A., Boston University, New York, NY, 10128;
[email protected]
S1176 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B159 6:00 PM-7:00 PM

GENERAL AND MENTAL HEALTH SERVICE CONNECTION IN POST 9/11 VETERANS

Shadi Gholizadeh, M.Sc., M.S., M.P.H. 1, Caroline A. Thompson, M.P.H., Ph.D. 2, Vanessa L.
Malcarne, Ph.D.3, Autumn Backhaus, Ph.D.4, Niloofar Afari, PhD5

1
SDSU/UC San Diego Joint Doctoral Program in Clinical Psychology, San Diego, CA; 2Graduate
School of Public Health/San Diego State University, San Diego, CA; 3San Diego State University,
San Diego, CA; 4Veteran's Administration San Diego Healthcare System and University of
California, San Diego, San Diego, CA; 5University of CA, San Diego & VA San Diego Healthcare
System, La Jolla, CA

Despite the relatively high numbers of service-connected injuries among Post 9/11 veterans
returning from Iraq and Afghanistan (OEF/OIF), predictors of service connection have only
been minimally studied in this group of veterans. The present study aimed to explore
differences among service connected and non-service connected OEF/OIF veterans and
examine sociodemographic and military predictors of general service connection. Prevalence
and predictors of service connection were assessed in 1193 OEF/OIF veterans registering with
the VA San Diego Healthcare System. Veterans completed a self-report questionnaire that
recorded service history, exposures while in the military, number of deployments, and combat
experience. Descriptive statistics and chi-square and t-test analyses were conducted to
explore differences in key variables between service-connected and non-service connected
veterans. Multivariable logistic regression was used to identify predictors of service
connection status. Over one-half of veterans were service connected for at least one
condition (n = 641; 53.7%). A smaller subset of veterans (n = 251; 21.0%) were service-
connected for at least one mental health condition. Modeling general service connection
revealed that career branch was associated with general service connection but concerns
about exposures, combat experience, number of deployments, and rank were not. Models for
mental health service connection identified that perceived harmful exposures (OR: 1.67; 95%
CI: 1.13, 2.45) and career branch (Army compared to Marines OR: 2.32; 95% CI: 1.22, 4.44)
were statistically significant predictors. Improved understanding of predictors of service
connected disability can inform strategies for prevention of such disabilities and interventions
to improve support for returning veterans.

CORRESPONDING AUTHOR: Shadi Gholizadeh, M.Sc., M.S., M.P.H. , SDSU/UC San Diego Joint
Doctoral Program in Clinical Psychology, San Diego, CA, 92120; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1177

B160 6:00 PM-7:00 PM

LATINO VETERAN MENTAL HEALTH SERVICE UTILIZATION

James O. E. Pittman, PhD1, Judith Baer, MSW, PhD2, Joyce Everett, MSW, PhD2, Mary Dozier,
MA3, Niloofar Afari, PhD4

1
VA San Diego Healthcare System, San Diego, CA; 2Smith College, School for Social Work,
Northampton, MA; 3University of California, San Diego, San Diego, CA; 4University of CA, San
Diego & VA San Diego Healthcare System, La Jolla, CA

Latinos represent 18% of the 2.1 million US troops deployed to combat Post 9/11. Since 2001,
56.9 percent of the 1.1 million veterans who used VA healthcare received a mental health
(MH) diagnosis. Research has identified disparities in health service utilization between
Latinos and non-Latino White individuals. Given these disparities in non-veterans and the
need for MH care among Post-9/11 veterans, the purpose of this study was to better
understand the MH treatment seeking behavior of Post 9/11 Latino veterans. Aims were to: 1)
examine differences in MH service utilization between Latino and non-Latino White veterans;
and 2) examine the relationship between potential predictors of MH service utilization by
using the Behavioral Model for Health Service Utilization (BMHSU). Previously collected data
at VA San Diego Healthcare System were used to examine patterns of MH service utilization in
a 740 Post-9/11 veterans with 35% Latino (n = 259) and 65% non-Latino White (n = 481). The
sample was primarily male (84.9%) and aged 21 to 67 years (M = 30.48; SD = 7.31).
Predisposing, enabling, and need factors from BMHSU were tested as predictors of MH
service utilization. Stepwise logistic regression analyses were performed using significant
variables from preceding univariate analyses to determine predictors of service utilization. No
significant differences in MH symptoms (all ps > .05) or rate of MH service utilization (30.5%
vs 29.0%, p = 0.68) between the groups emerged. All BMHSU factors were significantly related
to utilization in both groups, but only need predicted utilization in Latino veterans. Further, a
positive pain screen made MH service utilization 2.97 times more likely in Latino veterans
only. Except for pain, patterns of MH service utilization were similar in both groups of
veterans in San Diego. The BMHSU does not seem to fully apply to Latino or non-Latino White
Post 9/11 MH care utilization. Findings reinforce VA policy to screen veterans regularly for
symptoms of depression, PTSD, TBI, and pain. The relationship between pain and MH
utilization in Latino veterans warrants further study.

CORRESPONDING AUTHOR: James O. E. Pittman, PhD, VA San Diego Healthcare System, San
Diego, CA, 92161; [email protected]
S1178 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B161 6:00 PM-7:00 PM

NEIGHBORHOOD CHARACTERISTICS AND ADOLESCENT SUICIDAL BEHAVIOR

Kristina E. Allen, BA1, Sidra Goldman-Mellor, PhD, MPH2

1
University of California Merced, Mariposa, CA; 2University of California Merced, Merced, CA

Prior research suggests that neighborhood characteristics may be related to rates of suicidal
ideation and suicide attempt in adolescents, but no study has examined these associations
using a comprehensive array of both objective and subjective neighborhood factors. In the
current study, we used a population-based sample of adolescents (N=4,615 12-17-year-olds)
from the California Health Interview Survey (CHIS 2011-2014) Adolescents residential
addresses were geocoded to the census tract level and linked to objective measures of local
violent crime risk, property crime risk, and socioeconomic disadvantage. We also assessed
adolescents subjective perceptions of their neighborhoods safety and neighborhood
cohesion. Our outcome variables were any report of past-year suicidal ideation, suicide
ideation in the past two months, and past-year suicide attempt, reported by 8.4% of the
sample. We found that adolescents who lived in neighborhoods objectively characterized by
high levels of violent crime, property crime, or by socioeconomic disadvantage were no more
likely than their peers to engage in suicidal behavior. In contrast, adolescents who perceived
their neighborhoods to be less safe were 45% more likely to have suicidal ideation in the past
year, 42% more likely to have suicidal ideation in the past two months and 40% more likely to
attempt suicide in the past year compared to their non-suicidal peers. Further, adolescents
who perceived their neighborhoods to be less cohesive were 21% more likely to have suicidal
ideation in the past year and 24% more likely to have suicidal ideation in the past two months
compared to their non-suicidal peers. In our population-representative adolescent sample
from California, only subjective perceptions of neighborhood characteristics appear to be
related to risk of suicidal behavior. Individual-level risk factors may exert a more powerful
influence on psychologically vulnerable youths

CORRESPONDING AUTHOR: Kristina E. Allen, BA, University of California Merced, Mariposa,

CA, 95338; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1179

B162 6:00 PM-7:00 PM

PHYSICAL ACTIVITY AMONG ADULTS WITH PSYCHOTIC DISORDERS: PREFERENCES, BARRIERS

AND PERCEPTIONS OF ADVICES

Ahmed Jerome Romain, PhD, Christophe Longpr-Poirier, Medical science, Marc Tannous,
Medical science, Amal Abdel-Baki, MD, Msc, FRCPC

University of Montreal Hospital Research Centre, Montreal, PQ, Canada

Introduction: People with serious mental illness are more likely to have a poor lifestyle
including high sedentary behaviours and low physical activity (PA). Recent finding
acknowledge that PA can be beneficial on several parameters and it has been suggested that
considering PA preferences would facilitate its initiation or maintenance. However, few
studies assessed the PA preferences of people with serious mental illness. The objective of the
present study was to describe the PA preferences of people with serious mental illness.

Methods: Cross-sectional study. Participants with serious mental illnesses were asked to
complete several questionnaires about their PA preferences and their preferred context for
PA practice.

Results: Among the 114 included participants (men: 60.5%; mean age: 31.94 10.86), 55.3%
were physically inactive and 49.5% were overweight or obese. The preferred PA were walking
(55.3%), cycling (46.5%) and running (33.3%). Participants did not have preferences on the
mode of supervision (37.8%) or preferred supervision by a PA specialist (31.9%). Close to half
of participants preferred outdoor PA (46.8%) or did not have preferences (20.7%). The main
obstacle to PA was lack of motivation (45.5%). Most of participants highlighted that advices
from different mental health professionals were important for them (nutrition: 69.2%;
tobacco cessation: 56.1%; PA: 73%) and considered their physical health as important as the
improvement of their mental health (64.3%).

Conclusion: These results can be useful to conduct future PA interventions aiming at

maintaining PA over the long-term and underline the importance of involving mental health
professionals in this process.

CORRESPONDING AUTHOR: Ahmed Jerome Romain, PhD, University of Montreal Hospital

Research Centre, Montreal, PQ, H2X0A9; [email protected]
S1180 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B163 6:00 PM-7:00 PM

PHYSICAL ACTIVITY AND SUICIDE IDEATION AMONG YOUTH: THE ROLE OF PHYSICAL ACTIVITY
ENJOYMENT

Chaelin Karen Ra, MPH1, Adam Leventhal, PHD2, Genevieve F. Dunton, PhD, MPH3

1
Health Behavior Research, Pasadena, CA; 2USC, Pasadena, CA; 3University of Southern
California, Los Angeles, CA

Physical activity and suicide ideation among youth: The role of physical activity enjoyment

Background: The suicide rate among children and adolescents in the United States has
increased dramatically in recent years. Hence, a growing number of studies have been
conducted to understand the underpinnings of suicidal ideation (SI) in youth. A few studies
have suggested that engaging in physical activity (PA) may be protective against SI in youth.
Yet, inconsistencies in the results of these studies indicate that there may be unexplored
factors such as physical activity enjoyment (PAE) that can help us to better understand the
mental health benefits of PA in relation to SI. Thus, this study aimed to investigate whether
PAE is associated with SI, and whether PAE moderates the association between PA and SI in
children.

Methods: This study used baseline questionnaire data from 185 mothers and their children
enrolled in the Mothers and Their Childrens Health (MATCH) study. SI was assessed through
mothers reports on the Child Behavior Checklist (CBCL). Minutes per day of moderate to
vigorous physical activity (MVPA) was assessed through the child-reported 3-Day Physical
Activity Recall (3DPAR), and dichotomized by whether the child met the recommendation of
at least 60 minutes per day or not. PAE was assessed through the child-reported Physical
Activity Enjoyment Scale (PACES). Multiple logistic regression models were conducted to
estimate the likelihood of SI as a function of PAE and engaging in at least 60 minutes of daily
MVPA. The interaction of 60+ minutes of MVPA and PAE was included in a subsequent model
controlling for age, gender, ethnicity, single parent status and receiving free lunch at school.

Results: MVPA was not associated with SI (OR= 1.782, 0.582-5.452), and PAE was associated
with a reduced likelihood of SI (OR=0.485, 0.267-0.883). In the model with the interaction
term, the interaction between MVPA and PAE was not significant (OR=0.388, 0.111-1.359).

Conclusion: The enjoyment of physical activity instead of physical activity behavior itself may
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1181

be more closely tied to lowered SI among children. Interventions that seek to boost childrens
enjoyment of physical activity in addition to increasing levels of physical activity behavior may
be critical to children receiving the mental health benefits of physical activity, especially
among those at risk of suicide.

CORRESPONDING AUTHOR: Chaelin Karen Ra, MPH, Health Behavior Research, Pasadena, CA,
91106; [email protected]
S1182 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B164 6:00 PM-7:00 PM

SEXUAL SATISFACTION, ANXIETY, AND ESTEEM WHILE PRESCRIBED PRE-EXPOSURE

PROPHYLAXIS (PREP) AMONG GAY AND BISEXUAL MEN.

Thomas H.F. Whitfield, MA1, H. Jonathon Rendina, PhD, MPH2, Christian Grov, PhD, MPH3,
Jeffrey T. Parsons, PhD4

1
The CUNY Graduate Center / The Center for HIV Educational Studies and Training, New York,
NY; 2Hunter College of the City University of New York (CUNY), New York, NY; 3CUNY
Graduate School of Public Health and Health Policy, New York, NY; 4Hunter College and The
Center for HIV Educational Studies and Training, New York, NY

Background: Once-daily PrEP is the most effective biomedical HIV prevention tool currently
available. In addition to its protective effects against HIV, PrEP may have other benefits for
gay and bisexual men (GBM), including improved mental health. The present study sought to
investigate the association between PrEP use and levels of sexual anxiety, sexual esteem, and
sexual satisfaction among PrEP users.

Methods: Within a larger U.S. national sample of 1,013 HIV-negative GBM, 580 (57.3%)
reported recently engaging in sex with a casual partner. These men provided data on whether
they were prescribed PrEP and completed the sexual satisfaction, sexual anxiety, and sexual
esteem (multidimensional sexual self-concept) questionnaire.

Results: In total, 70 (12.1%) reported currently being prescribed PrEP and 510 (87.9%) were
not currently prescribed. Adjusting for demographic differences (i.e. age, race, income), men
currently prescribed PrEP reported more sexual satisfaction ( = 0.30, SE = 0.14, p = 0.04) and
less sexual anxiety ( = -0.28, SE = 0.12, p = 0.02), but were not statistically different with
regard to sexual esteem ( = 0.25, SE = 0.14, p = 0.07).

Conclusions: These results indicates that there may be benefits to taking PrEP for some GBM
that go beyond (but may be due in part to) its protection against HIV acquisition. These
findings suggest that being in PrEP may lead to reduced sexual anxiety associated and greater
satisfaction though the directionality of these associations may also be reversed. Future
longitudinal research is needed to better examine whether the health-protective effects of
PrEP may also lead to psychological benefits.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1183

CORRESPONDING AUTHOR: Thomas H.F. Whitfield, MA, The CUNY Graduate Center / The
Center for HIV Educational Studies and Training, New York, NY, 10018;
[email protected]
S1184 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B165 6:00 PM-7:00 PM

SUPPORT TO KINSHIP CAREGIVERS IMPROVES CHILD BEHAVIOR: RESULTS FROM THE KIN TECH
RCT

Kerry Littlewood, Ph.D., MSW1, Lawrence Cooper, LCSW, MSW2, Michelle Rosenthal, Ph.D.3,
Liliana Hernandez, MSW, MPP4, Anne Strozier, Ph.D., MSW5, Abhishek Pandey, MD6

1
AAJ Research and Evaluation, Seminole, FL; 2The Children's Home, Inc., Tampa, FL; 3Data
With Purpose, Oakland, CA; 4US Administration on Children and Families, Washington, DC;
5
University of South Florida, Tampa, FL; 6University of Arizona, Staten Island, NY

Background: One in 11 of all children and 1 in 5 African American children live with a
grandparent at some point before the age of 18. While children raised by grandparents
experience more risks of psychological difficulties than children in the general population,
especially experiencing trauma at a young age, few studies have examined which
interventions can improve psychosocial impairment for children raised by grandparents.

Purpose: This study identified psychosocial impairment using baseline, six month post test,
and 12 month follow up self-report data on the Pediatric Symptom Checklist obtained from
kinship caregivers enrolled in the KIN-Tech randomized controlled trial funded by the US
Childrens Bureau. Descriptives and repeated measures ANOVAS were used to examine mean
differences between and within treatment groups (Navigation, Extended Family Support, and
Usual Care).

Results: Participants in this study included 1100 female kinship caregivers from low income
households (m=$23,000), middle-aged (m=48), unmarried (70%), African American (47%)
caring for multiple relative children (65% caring for more than one child). 48% (n=528) of
caregivers earned at least a high school diploma. Participants in the navigator and extended
family support programs reported a statistically significant reduction in their childrens
conduct, anxiety and attention problems [Navigator Program, F(252,3)=3.58, p = .023]
[Extended Family Support, F(849,3)=5.14, p = .000]. Participants in the usual care group
reported an increase in conduct and anxiety problems, but a decrease in attention problems.

Conclusions: Participants randomly assigned to a kinship navigator program and an extended

family support program showed reduction in the conduct, anxiety and attention problems
that was significantly different than those assigned to the usual care group. Future studies
should explore which specific program components are most beneficial for kinship caregivers
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1185

and help promote responsive and competent caregiving for the children living in non-parental
care.

CORRESPONDING AUTHOR: Kerry Littlewood, Ph.D., MSW, AAJ Research and Evaluation,
Seminole, FL, 33776; [email protected]
S1186 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B166 6:00 PM-7:00 PM

THE MENTAL HEALTH SYSTEM IN NORTHERN TANZANIA: PROVIDER PERCEPTIONS OF THE

PRESENTATION AND TREATMENT OF MENTAL ILLNESS

Brandon A. Knettel, Ph.D.1, Janvier Rugira, Ph.D.2, Florence Tesha, N/A3

1
Duke Global Health Institute, Raleigh, NC; 2University of Johannesburg, Johannesburg,
Gauteng, South Africa; 3Duke University, Durham, NC

The steady globalization of western models of mental health treatment is causing an influx of
practitioners, ideas and resources to developing nations. One such country, the United
Republic of Tanzania, is at a crucial stage in the development of its mental health system. As
of 2009, there were 18 psychiatrists serving this nation of 51 million people and a gross lack of
resources being dedicated to mental health treatment. However, as international
organizations shift focus to noncommunicable diseases, there have been some signs of
growth in the Tanzanian mental health system, including the establishment of new training
programs and professional associations in psychology. As services continue to develop, there
is a need for solid foundational research aimed at understanding the unique challenges and
strengths of providing mental health care in this context.

We conducted interviews with 29 mental health providers in a variety of treatment settings in

northern Tanzania to examine their perceptions of the treatment currently being offered, the
characteristics of the patient population, and future directions in mental health treatment.
Responses were analyzed using consensual qualitative research (CQR) methodology and offer
a nuanced perspective of the characteristics, strengths, and challenges of the Tanzanian
mental health system. Participants acknowledged that there is a widespread lack of mental
health providers in the region, inadequate resources dedicated to mental health, very little
coordination among providers, and major social, logistical, and cultural barriers that prevent
all but the most severe patients from seeking or receiving care.

The research provides an important baseline of understanding to inform the ongoing

development and improvement of mental health services in the region. In response to the
concerns identified, we propose a three-part improvement plan consisting of (1) improving
the availability and quality of care via investment in training programs and facilities, (2)
Maximize existing resources via task shifting programs aimed at improved coordination of
various types of care, and (3) Address barriers to seeking care via public education, stigma
reduction, and improved cultural compatibility of care.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1187

CORRESPONDING AUTHOR: Brandon A. Knettel, Ph.D., Duke Global Health Institute, Raleigh,
NC, 27613; [email protected]
S1188 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B167 6:00 PM-7:00 PM

USING ECOLOGICAL MOMENTARY ASSESSMENT TO EXAMINE RELATIONS BETWEEN PHYSICAL

ACTIVITY AND AFFECT IN ADULTS WITH MENTAL ILLNESS

Chun Nok Lam, PhD Canditate1, Brian W. Redline, BA1, Harmony Rhoades, PhD1, Genevieve F.
Dunton, PhD, MPH1, Benjamin F. Henwood, PhD, MSW2

1
University of Southern California, Los Angeles, CA; 2University of Southern California, los
angeles, CA

Background: Adults with serious mental illness (SMI) engage in limited physical activity, which
contributes to significant health disparities and premature mortality in this population. The
need for basic research on physical activity mechanisms that can be translated into life style
interventions for individuals with SMI is urgent. Individuals with SMI may experience different
patterning of emotional experiences than individuals without SMI. This study aims to examine
the interrelations between affective states and physical activity among the SMI population.

Methods: This study used ecological momentary assessment (EMA), a real-time data capture
strategy to assess subjective positive and negative affective states in free-living setting up to 7
times per day across 4 days from 21 Latino adults diagnosed with SMI. Participants were
equipped with an accelerometer which provided an objective measure of their moderate-to-
vigorous physical activity (MVPA) during the study period. Multilevel modeling was used to
test the bi-directional associations between affective states and MVPA across the day.

Results: Participants (49.013.1 years, 76.2% female) engaged in limited MVPA, with only
30.2% engaging in 1 minute of MVPA in the 30-minute prior to any given EMA prompt
(median MVPA: 2.5 minutes, range 1-28 minutes). Participants who on average reported more
positive affect than others in the study had a higher probability in engaging in some MVPA
versus no MVPA during the study period. However, among participants who engaged in some
MVPA during the study period, reporting more positive affect on average than others was
associated with less MVPA on average. In addition, engaging more MVPA than ones usual in
the 30-minutes prior to an EMA prompt was associated with more negative affect at that
prompt.

Conclusions: Engagement in more MVPA was generally associated with feeling less positive
and more negative affect among individuals with SMI. This finding suggests that the types of
physical activity performed in this population (such as active transportation) may not be
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1189

perceived enjoyable or pleasant. The results provide new insights into the unmet needs of
individuals with SMI and offers new directions for future research.

CORRESPONDING AUTHOR: Chun Nok Lam, PhD Canditate, University of Southern California,
Los Angeles, CA, 90033; [email protected]
S1190 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B168 6:00 PM-7:00 PM

WERE GRAY, GAY AND HERE TO STAY: TRAUMA HISTORY, SOCIAL SUPPORT AS CORRELATES
OF RESILIENCY IN LESBIANS AND GAY MEN OVER AGE 50

Bunyong Dejanipont, (BSBME), Kinsie J. Dunham, BA, Mark Vosvick, PhD

University of North Texas, Denton, TX

Traumatic experiences have long-lasting effects, such as the development of psychopathology

(Hatzenbuhler, 2009). Lesbians and gay men are more likely to experience trauma (Brown et
al., 2011). Further, older cohorts of lesbians and gay men (LG+50) likely have larger
cumulative trauma histories (TH; Hash et al., 2013). The dearth of research on TH in LG+50
requires examination of protective factors and trauma outcomes (Brown et al., 2011).

Social support (SS), a protective factor, may increase Resilience among sexual minorities, as it
lessens the impact of discriminatory events (Kwon, 2013). Resiliency may be a preferential
trauma outcome, as Resiliency is associated with brief dysregulation (Westphal et al., 2007).
Additionally, Resiliency is more prevalent in populations that face adversity (Tomac, 2011). As
LG+50 face dual discrimination, Resiliency is the ideal outcome in response to a traumatic
event. Despite the prevalence of traumatic events for LG50+, little research explores SS and
Resiliency. As both are relevant to LG+50, the aim of the current study is to examine the
mediating effect of SS (UCLA SSI; Dunkel-Schetter et al., 1986) on TH (THQ; Green, 1996) and
resilience (CD-RISC; Connor et al., 2003).

We recruited LG+50 (N = 100; n =71; Mage = 59, SD = 5.9), and excluded those who had not
experienced a traumatic event. To test for mediation we entered covariates (gender,
religiosity and HIV-positive status) into the first block of each analysis. HIV-positive status was
the only significant covariate in indirect path a, showing that people living with HIV scored
significantly lower on SS ( = -.42, p = .003). In the first step we examined direct relationship c
by regressing TH on Resilience ( = .03, p = .806). Though insignificant, we tested for full
mediation. Indirect path a between TH and SS ( = -.23, p = .039) and indirect path b, SS and
Resilience ( = .44, p = .001) were significant. Additional Sobel analysis indicated full
mediation (TS = 1.83, p = .003). This suggests SS is a mechanism that may help LG+50 develop
Resilience. LG+50 with greater TH did not achieve resiliency, except though SS, reinforcing the
importance of this developmentally and socially important mechanism.

CORRESPONDING AUTHOR: Bunyong Dejanipont, (BSBME), University of North Texas, Denton,

TX, 76201; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1191

B169 6:00 PM-7:00 PM

ARE SELF-EFFICACY MEASURES CONFOUNDED WITH MOTIVATION? AN EXPERIMENTAL TEST

David Williams, Ph.D.1, Shira Dunsiger, PhD2, Jessica Emerson, MS3, Laura Dionne, B.A.4, Ryan
E. Rhodes, PhD5, Mark R. Beauchamp, PhD6

1
Brown University School of Public Health, Providence, RI; 2The Miriam Hospital and Brown
University, Portsmouth, RI; 3Brown University, East Providence, RI; 4Brown University,
Charlestown, RI; 5University of Victoria, Victoria, BC, Canada; 6University of British Columbia,
Vancouver, BC, Canada

Self-efficacy (SE) is defined as confidence that one can perform a target behavior in the face of
various obstacles. SE is highly predictive of a range of health behaviors and is a central
construct in many theories of behavior.

Previous authors have argued that standard SE measures may actually tap respondents
motivation rather than their confidence that they can perform the behavior, particularly when
the target behavior is under volitional control (e.g., physical exercise). This occurs because
respondents tend to interpret the word can as willingness when asked whether they, for
example, can exercise even if they are not feeling well. If this is true, then priming
respondents regarding the definition of the word can or to the contrast between can and
will, should increase SE ratings in the absence of any intervention.

In this study, 134 university students (65% female; 55% White) responded to a questionnaire
assessing SE for exercise using a standard SE measure, behavioral intentions (BI) to exercise,
and their definition of the word can, as in I can exercise today even though I am not feeling
well. One month later participants were randomly assigned to complete either (a) the same
SE and BI questionnaire as at baseline (i.e., control), (b) the same questionnaire but with
presentation of each participants definition of can prior to responding to the SE measure
(i.e., definition priming), or (c) the same questionnaire but with SE items (I can exercise . . .)
placed side-by-side with BI items (I will exercise . . .).

Generalized linear models were used to compare groups. Results showed increases in SE
responses, relative to controls, for side-by-side (b=12.08, SE=2.70, p < .01), but not definition
priming (ns). As expected, there were no effects of side-by-side priming, relative to control, on
BI (ns), and the effects of side-by-side priming on SE ratings remained significant even when
controlling for BI at times 1 and 2 (b=8.87, SE=2.11, p < .01), thus illustrating that SE ratings
were manipulated by the side-by-side priming independent of BI ratings.
S1192 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

The results show that responses to SE items can be increased simply by placing them next to
BI items. This suggests that responses to standard assessments of SE do not simply reflect
respondents confidence that they can perform the target behavior, but are confounded by
their motivation to perform the behavior. Studies using standard assessments of SE should
perhaps be reinterpreted in light of these findings.

CORRESPONDING AUTHOR: David Williams, Ph.D., Brown University School of Public Health,
Providence, RI, 02912; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1193

B170 6:00 PM-7:00 PM

COMPARISON OF THE DIAGNOSTIC ACCURACY OF FOUR BRIEF DEPRESSION SCREENING

INSTRUMENTS IN OVARIAN CANCER PATIENTS

Shinn Shinn, Ph.D. 1, Cindy Carmack, Ph.D.1, Kelly Kilgore, B.A. 1, Alan D. Valentine, M.D. 1,
Karen Basen-Engquist, PhD, MPH2

1
University of Texas M. D. Anderson Cancer Center, Houston, TX; 2The University of Texas MD
Anderson Cancer Center, Houston, TX

Objectives: We compared the diagnostic accuracy of 4 depression screening scales, using

traditional and alternative scoring methods, to the gold standard Structured Clinical Interview-
DSM IV major depressive episode (MDE) in ovarian cancer patients on active treatment.

Methods: At the beginning of a new chemotherapy regimen, ovarian cancer patients

completed the following on the same day: the Center for Epidemiological Studies Depression
Scale (CES-D), the Beck Depression Inventory Fast-Screen for Primary Care (BDI-FastScreen),
the Patient Health Questionnaire-9 (PHQ-9), and a 1-item screener (Are you depressed?).
Each instruments sensitivity, specificity, positive predictive value (PPV) and negative
predictive value were calculated with respect to major depression.

Results: One hundred and fifty-three ovarian cancer patients were enrolled into the study.
Only fourteen participants met SCID criteria for current MDE (9%). When evaluating all
patients regardless of whether they were already being treated with antidepressants or not,
using a two-phase scoring approach with an alternate cutpoint of 6 on the PHQ-9 had the best
positive predictive value (PPV=32%). Using a traditional cutpoint of 16 on the CES-D resulted
in the lowest PPV (5%); using a more stringent cutpoint of 22 resulted in a slightly improved
but still poor PPV, 7%.

Conclusions:

Using a two-phase scoring method for the PHQ-9, with an alternate cutpoint of 8 resulted in
the best diagnostic performance among the four screening instruments. The traditional
cutpoint of 16 on the CES-D and the one-item screener were among the worst methods and
are not recommended as first-line screening methods in oncological settings.

CORRESPONDING AUTHOR: Shinn Shinn, Ph.D. , University of Texas M. D. Anderson Cancer

Center, Houston, TX, 77030-3721; [email protected]
S1194 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B171 6:00 PM-7:00 PM

DEVELOPMENT AND VALIDATION OF AN AGENT BASED MODEL TO EXAMINE NEUROTICISM

AND OBESITY

Hilary C. Tanenbaum, MPH, MS, RD1, Brian Hilton, Ph.D.2, Kim Reynolds, PhD2, Bin Xie, PhD2

1
School of Community & Global Health, Claremont Graduate University, Claremont, CA;
2
Claremont Graduate University, Claremont, CA

The numerous, interacting factors involved in the etiological web of the obesity epidemic
warrant examination with modeling methods that can realistically reflect these dynamics.
Agent-Based Modelling (ABMs) is one of several complex systems techniques garnering
interest among researchers in obesity research. ABMs use computer-generated algorithms to
simulate the behavior of individual entities by assigning specific behavior rules, incorporating
elements of stochasticity, heterogeneity, and non-linearity. They can also provide a blank
canvas to explore potential interventions. Despite recent advances in computer processing
and software that have broadened their accessibility, the novelty of this approach in the
health field, combined with the wide range of possible techniques, makes navigating ABM
construction challenging. Few studies exist that comprehensively explain the development
process. A new approach for designing an ABM to examine the linkage between Neuroticism
and obesity is presented in this study, providing a detailed description of the process for
populating an ABM with real-world data, defining the behavior rules of the agents, structuring
the model for future interventions, and validating the model. AnyLogic software (v. 7.6) was
used to construct the model with data from two collection points (spanning 13 years) from
the National Longitudinal Study of Adolescent to Adult Health (Add Health). The dataset
(N=9176) was segmented by gender and cross-tabulated to define the starting or ending
probabilities of weight classes (Under/Normal, Overweight, Obese) by Neuroticism score from
the first wave of data collection. The data was randomly split into two equal groups, using half
to construct the model, and reserving the rest to verify that no coding or structural errors
existed. To validate that the simulated population did not differ significantly from the Add
Health dataset, 1000 randomly-seeded Monte Carlo (MC) runs were performed, and a one-
sample t-test was used to compare the mean counts in each weight class from the MC runs to
the actual dataset means. No significant differences were found for starting weight
(Under/Normal: p=0.16, Overweight: p=0.20, Obese: p=0.68), transitioning from starting
weight (Under/Normal: p=0.32, Overweight: p=0.44, Obese: p=0.72), or ending weight
(Under/Normal: p=0.95, Overweight: p=0.76, Obese: p=0.87), suggesting that the agents
accurately reflect the real-world data. With the simulated population established, potential
interventions to change Neuroticism scores can be implemented to identify the best feasible
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1195

solution for reducing obesity. The ABM construction process presented in this study serves as
a modifiable template that may be used to explore relationships among other variables of
interest in obesity research.

CORRESPONDING AUTHOR: Hilary C. Tanenbaum, MPH, MS, RD, School of Community &
Global Health, Claremont Graduate University, Claremont, CA, 91711;
[email protected]
S1196 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B172 6:00 PM-7:00 PM

EDUCATION, HEALTH LITERACY, AND NUMERACY: IMPLICATIONS FOR UNCERTAINTY ABOUT

RISK

Erika A. Waters, PhD, MPH1, Elizabeth Schofield, MPH2, Caitlin Biddle, MA, MS3, Jennifer L.
Hay, Ph.D.2, Marc T. Kiviniemi, BA, PhD3, Yuelin Li, PhD.2, Kimberly Kaphingst, ScD4, Heather
Orom, PhD3

1
Washington University in St. Louis, St. Louis, MO; 2Memorial Sloan Kettering Cancer Center,
New York, NY; 3University at Buffalo, Buffalo, NY; 4University of Utah and Huntsman Cancer
Institute, Salt Lake City, UT

Background: Uncertainty about risk, as operationalized by marking the dont know response
option for survey items assessing perceived risk of colon cancer, is related to
sociodemographic characteristics (e.g., education), that have been associated with limited
knowledge about colon cancer, limited cancer information seeking, and lower engagement in
some cancer prevention and detection behaviors. Whether risk uncertainty is also associated
with limited health literacy and/or numeracy has been hypothesized but not confirmed.

Objective: To examine whether uncertainty about colon cancer risk and diabetes risk are
related to objective and/or subjective health literacy and numeracy.

Methods: Participants were recruited from the GfK Knowledge Networks Internet survey
panel and answered questions regarding demographics, objective and subjective numeracy,
and objective and subjective health literacy. They also completed items assessing perceived
risk of colon cancer (N = 999) and diabetes (N = 893). Responses to risk perception items were
dichotomized according to whether the participant did or did not respond dont know.

Results: Unadjusted analyses indicated that dont know responding (DKR) for perceived
diabetes risk was lower among people with higher objective and subjective health literacy and
numeracy (ps < .05). DKR for perceived colon cancer risk was lower among people with higher
subjective health literacy and numeracy (ps < .05), but not higher objective literacy or
numeracy. However, after adjusting for education, DKR for diabetes was only associated with
subjective health literacy, and DKR for colon cancer was only associated with subjective health
literacy and subjective numeracy (ps < .05).

Conclusion: Objective measures of health literacy and numeracy do not appear to account for
significant variance in DKR above and beyond that accounted for by education. However,
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1197

subjective measures of health literacy and numeracy remain unique predictors of DKR. It is
possible that these subjective self-report measures assess not only ability, but also confidence
in understanding, engaging with, and using health information in a way that relates to
peoples levels of uncertainty about their risk of disease.

CORRESPONDING AUTHOR: Erika A. Waters, PhD, MPH, Washington University in St. Louis, St.
Louis, MO, 63110; [email protected]
S1198 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B173 6:00 PM-7:00 PM

ESTABLISHING CLINICAL CUTOFFS FOR THE SCREENING OF AVOIDANT/RESTRICTIVE FOOD

INTAKE DISORDER

Jordan Ellis, MA1, Hana F. Zickgraf, MA2, Matthew C. Whited, PhD, Licensed Psychologist1,
Amy T. Galloway, Ph.D.3

1
East Carolina University, Greenville, NC; 2University of Pennsylvania, Philadelphia, PA;
3
Appalachian State University, Boone, NC

Avoidant/Restrictive Food Intake Disorder (ARFID) is a new DSM-5 diagnosis intended to

identify individuals who exhibit clinically significant restrictive eating problems (not primarily
accounted for by shape or weight concerns), persistently fail to meet nutritional and/or
energy needs, and experience psychosocial difficulties. Recent studies have indicated picky
eating (PE; one manifestation of ARFID) can persist into adulthood and is associated with
anxiety, depression, and OCD symptomology.

Two brief measures for ARFID symptoms and PE behaviors were recently developed to
address a lack of measurement tools. The Nine Item ARFID Screener (NIAS) is a measure of
rigid eating behaviors associated with ARFID, and the Adult Picky Eating Questionnaire (APEQ;
Ellis et al., 2016) is a multidimensional measure of PE behaviors. The purpose of the current
study was to assess the sensitivity and specificity of these measures to detect ARFID and
establish a clinically useful cut score.

The current study included two samples: 1.) A general sample of 1219 college students who
completed an online survey that included the NIAS and APEQ, and 2.) A sample of 450 adults
selected for eating difficulties who completed the NIAS. Participants also completed the ARFID
Symptoms Checklist (ARFID-cl). The ARFID-cl assesses cause, duration, and impairment and
conforms to DSM-5 diagnostic criteria. It is meant to be used with people who meet a
screening criterion. In the general sample 31 (2.5%) college students met criteria for ARFID; 65
(14.4%) in the selective sample.

Receiver Operating Characteristic (ROC) curve analyses were employed to establish cutoff
scores for identifying a positive ARFID diagnosis. Area under curve (AUC) indicated that the
NIAS (AUCgeneral= .87, p < .001; AUCselected= .80, p < .001) and APEQ (AUC = .83, p < .001) were
good screeners for ARFID. In the general sample, a cutoff of 44 on the APEQ and 24 on the
NIAS both demonstrated good sensitivity (0.74) and specificity (0.84; PPV = 0.11; NPV = 0.99).
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1199

In the selective sample, a cutoff 28 on the NIAS demonstrated adequate sensitivity (0.71) and
specificity (0.68; PPV = 0.69; NPV = 0.70).

The NIAS and APEQ are useful ARFID screeners but higher cutoffs need to be used in selective
samples. This finding is likely due to medical and symptom complexity of the participants in
the selective sample. Future research should investigate the utility of these measures in
detecting symptom changes during treatment.

CORRESPONDING AUTHOR: Jordan Ellis, MA, East Carolina University, Greenville, NC, 27858;
[email protected]
S1200 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B174 6:00 PM-7:00 PM

FEASIBILITY OF USING ECOLOGICAL MOMENTARY ASSESSMENT WITH LOW-INCOME LATINO

ADULTS EXPERIENCING SERIOUS MENTAL ILLNESS

Brian W. Redline, BA1, Chun Nok Lam, MPH1, Benjamin F. Henwood, PhD, MSW2, Harmony
Rhoades, PhD1, Stephen Intille, Ph.D.3, Eldin Dzubur, MS1, Genevieve F. Dunton, PhD, MPH1

1
University of Southern California, Los Angeles, CA; 2University of Southern California, los
angeles, CA; 3Northeastern University, Boston, MA

Background: People with serious mental illness (SMI) are more physically inactive and have
higher rates of obesity, cardiovascular disease, and premature death compared to the general
population. Emerging mobile technologies offer opportunities to examine how unique
affective and environmental factors influence weight-related behaviors in this population and
develop targeted real-time interventions. However, concerns about poor compliance,
unreliable responses, and participant burden present potential barriers to implementation.
The aim of this exploratory mixed-methods study was to test the feasibility of using
smartphone-based ecological momentary assessment (EMA) methods to examine weight-
related behaviors in low-income Latino adults with SMI.

Methods: Twenty-one adults (ages 21-63 years, 76% female, 95% Latino, 48% Spanish-
speaking) were recruited from an integrated physical and behavioral healthcare clinic for low-
income adults with SMI. After receiving instruction on how to use the study phone, EMA
application, and waist-worn accelerometer, participants completed four days of activity, GPS,
and signal-contingent EMA monitoring with up to 7 survey prompts per day. EMA surveys
measured mood, weight-related behaviors, social and physical context, SMI symptoms, and
substance use. Participants returned one week later to complete in-depth interviews that
focused on study experience. An interactive GPS map of EMA responses was used to elicit
information about how mood and context impacted physical activity using qualitative
methods.

Results: Participants completed 82.9% of 520 EMA surveys overall, and provided an average of
3.67 valid days (>10 hrs/day) of accelerometer data. Participants reported being alone
immediately before the prompt in 39.7% of surveys, physical activity in the past hour in 15.9%
of surveys, psychotic symptoms in the past hour in 4.8% of surveys, and crying spells, poor
focus or poor appetite in 10.6% of surveys. Qualitative analysis of follow-up interviews
revealed almost half of participants experienced feelings of anxiety or stress using the EMA
app. Reasons for reporting feelings of anxiety or stress included feeling a need to stay vigilant
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1201

to not miss surveys, lose the phone, or let the battery die, and from limited experience using
smartphones.

Discussion: This pilot study demonstrates that although EMA can be effectively used to collect
real-time physical activity data in low-income populations with SMI, measures should be
taken to minimize undue participant burden due to low technological literacy and pressure to
comply that may result in exacerbation of SMI symptoms (e.g., anxiety).

CORRESPONDING AUTHOR: Brian W. Redline, BA, University of Southern California, Los

Angeles, CA, 90015; [email protected]
S1202 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B175 6:00 PM-7:00 PM

IDENTIFYING BEHAVIORAL PATHWAYS TO SMOKING CESSATION: DECIDE2QUIT RANDOMIZED

TRIAL

Mickayla Aboujaoude, B.A. Economics Expected May 20171, Arlene S. Ash, PhD1, Daniel
Amante, MPH1, Rajani Shankar Sadasivam, PhD2, Thomas Houston, MD, MPH1

1
University of Massachusetts Medical School, Worcester, MA; 2Univ of Massachusetts Medical
School, Worcester, MA

Background: Understanding behavioral pathways through which technology-assisted tobacco

interventions support smokers can inform improvements.
Methods: A randomized trial with complete follow-up on 369 smokers had previously
demonstrated more quitting (OR = 1.7) among users of an interactive website with
motivational messages (enhanced) than for those assigned to the website only. Here we
explore patient-reported influences of the intervention on subsequent cessation. After 6
months of website use, we asked if users had quit smoking in the last 30 days and if they
agreed using the intervention had influenced 8 smoking cessation behaviors (cutting down;
quitting; setting a quit date; using nicotine replacement therapy; using behavioral strategies;
making a list of reasons to quit; talking to a doctor; getting support for quitting). We compare
perceived influence between randomized groups, and assess six-month cessation rates by
each influencing variable, then examine 9 logistic regressions to predict quitting, controlling
for age, sex, and baseline readiness-to-quit: 1 regression for each influence item alone and 1
with all 8.
Results: Most were female (63%) over age 45 (65%) and had been thinking of quitting (81.5%)
at baseline. Most agreed that the intervention influenced them to cut down on smoking
(61%), talk to a doctor (56%), use behavioral strategies (55%), and make a list of reasons to
quit (53%). More smokers in the enhanced group endorsed the value of the intervention for
each of the 8 questions (differences ranged from 2.8 to 11.8%; only 3 were significant at the P
< 0.05 level). In the 8 individual regressions, influence on quitting smoking was positively
associated with quitting (OR 3.01 (95% CI 1.76-5.15)), while influences on cutting down on
smoking and using nicotine replacement therapy were negatively associated (ORs = 0.58
(95% CI, 0.34-0.98) and 0.44 (0.24-0.83), respectively); no other influences were associated
with quitting (all p > 0.20). Findings from the combined regression were qualitatively similar.
Conclusion: Cutting down to quit is controversial, and was negatively associated with
cessation. Our unexpected finding (re: NRT) could be due to residual confounding by severity
of addiction, needs further exploration.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1203

CORRESPONDING AUTHOR: Mickayla Aboujaoude, B.A. Economics Expected May 2017,

University of Massachusetts Medical School, Worcester, MA, 01610;
[email protected]
S1204 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B176 6:00 PM-7:00 PM

MONITORING DAILY EXERCISE FOR 1 YEAR USING SMARTPHONES: VALIDITY OF ECOLOGICAL

MOMENTARY ASSESSMENT

Kevin Sundquist, B.S.1, Ipek Ensari, PhD1, Jacob Julian, B.A.1, Karina Davidson, PhD, MASc2,
Matthew Burg, PhD3, Joseph Schwartz, PhD1, Keith M. Diaz, PhD1

1
Center for Behavioral Cardiovascular Health, Columbia University Medical Center, New York,
NY; 2Columbia University Medical Center, New York, NY; 3Department of Internal Medicine,
Yale School of Medicine, New Haven, CT

Background: Accelerometers have been used in research to objectively quantify exercise for
over two decades. However, these devices are expensive, are burdensome to participants,
and the derivation of exercise bouts from their data is arduous. The purpose of this study was
to calculate the agreement between smartphone-based ecological momentary assessment
(EMA) and objective accelerometer defined exercise to determine the feasibility of using EMA
in lieu of accelerometers in the long term assessment of exercise. Methods: Data were
collected from an observational study of intermittently exercising, otherwise healthy adults
(N=79; 43% male, mean age: 329 years) who were studied for 365 days. Smartphone-based
EMA self-report of exercise entailed daily end-of-day response to the question: Did you
exercise today for 30 minutes or more at a moderate or vigorous level? Participants also
wore a Fitbit Flex to objectively determine exercise bouts. Within-subject Kappa statistics of
agreement were calculated between objectively-determined exercise (24 mins of moderate to
vigorous physical activity [MVPA] within 30 minutes) and EMA self-reported exercise.
Predictors of agreement (i.e., sex, age) were also examined. Results: On average, participants
provided 21191 days of data. The average within-subject Kappa statistic for agreement
between objective and self-reported exercise was =0.270.19 (range: -0.03-0.71). The
average within-subject Kappa ranged from 0.16 to 0.27 when the definition of an
accelerometer-based exercise bout was varied from 15 to 30 mins (of 30 mins) MVPA. In a
multivariable adjusted model, gender was significantly associated with the level of agreement,
being higher for women (=33.60.19) than for men (=19.20.17). Discussion: The
agreement between objective (i.e., accelerometer) and self-reported (EMA) measures of
exercise was fair. These findings suggest that smartphone-based EMA for long term exercise
monitoring in lieu of an accelerometer may be feasible, but does not provide high reliability to
an objectively assessed exercise bout.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1205

CORRESPONDING AUTHOR: Kevin Sundquist, B.S., Center for Behavioral Cardiovascular

Health, Columbia University Medical Center, New York, NY, 10032;
[email protected]
S1206 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B177 6:00 PM-7:00 PM

QUANTIFYING SUN PROTECTION MAINTENANCE AND ENVIRONMENTAL PROMOTERS IN

INDIVIDUALS AT RISK FOR MELANOMA

Yuelin Li, PhD., Elizabeth Schofield, MPH, Kristen E. Riley, Ph.D. , Jennifer L. Hay, Ph.D.

Memorial Sloan Kettering Cancer Center, New York, NY

Achieving health behavior maintenance, while a gold standard outcome of behavioral

interventions, is difficult to measure accurately and reliably, especially in quantifying the
influences of habit and environmental promoters on behavior. We propose a Bayesian
Hierarchical Linear Model (HLM) that clearly quantifies both influences.

Ecological Momentary Assessment (EMA) data were collected from 59 first-degree relatives of
melanoma patients regarding sun protection behaviors (sunscreen use, shade seeking, hat
and protective clothing) and the circumstances (e.g., weather and convenient availability of
sunscreen), twice daily (1 pm; 5 pm) over 14 summer days (1,304 total assessments). A
Bayesian HLM used random effects to represent habit and environmental promoters for
determining whether each persons behaviors were primarily determined by habit or by
environmental promoters.

Sunscreen use was influenced by environmental promoters, such as sunny and hot weather
(OR=6.44, CI: 3.4,13.1; p< 0.0001), and also by person-specific habit (0.286; CI: 0.10,0.73; p <
0.01). Each persons habit was independent of propensity to use sunscreen in hot and sunny
weather, shown in a non-significant correlation (rho=0.14, CI: 0.82,+0.70). A gender
moderating effect was observed. Sunny and hot weather exerted a greater promoting
influence on mens than womens use of sunscreen (OR=3.80, p=0.040), but not when the
weather was cool (OR=0.205, p=0.072). Men were less likely than women to wear long sleeve
clothing (OR=0.18, p=0.03), more likely to wear a hat (OR=3.6, p=0.05), but equally likely to
seek shade. The gender moderating effect was only present in sunscreen use, not the other
behaviors. Bayesian HLM yielded a propensity of each persons sun protection behaviors
being dictated more by environment than by habit, which showed that women were affected
more by weather than habit (73% women vs. 50% men, p=0.096).

In conclusion, sun protection behaviors are mainly a function of weather and other
environmental promoters rather than habit. The Bayesian approach to EMA data makes it
possible to compare habit and environment in one coherent framework, a option not
available by conventional methods. It informs the gender-specific strategies in future
interventions.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1207

CORRESPONDING AUTHOR: Yuelin Li, PhD., Memorial Sloan Kettering Cancer Center, New
York, NY, 10022; [email protected]
S1208 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B179 6:00 PM-7:00 PM

USING ECOLOGICAL MOMENTARY ASSESSMENT METHODS TO MEASURE STRESS AND

WEIGHT-RELATED BEHAVIORS IN HISPANIC POST-PARTUM WOMEN

Sydney G. O'Connor, BA, Jane Wai-Quan Cabison, BS, Wangjing Ke, MA, Carrie V. Breton, ScD,
Tracy Bastain, PhD, Genevieve F. Dunton, PhD, MPH

University of Southern California, Los Angeles, CA

Background Excessive gestational weight-gain and post-partum weight retention are

associated with an increased risk of lifelong obesity and chronic disease, which
disproportionately affects minorities and women of low socioeconomic status. Ecological
Momentary Assessment (EMA) methods may provide dynamic and real-time insight on the
social and contextual factors influencing weight-related behaviors and obesity risk during this
critical period of womens lives. However, barriers to EMA in this population, including
concerns over time demands and participant burden, remain. Thus, this study examined the
feasibility and challenges of using EMA methods in low-income, Hispanic post-partum women
with young infants.

Methods Twelve low-income Hispanic mothers (mean age: 31 yrs, SD: 7 yrs; 58% Spanish-
speaking), of young infants (age range: 6-31 weeks) completed the study procedures.
Participants completed up to 5 randomly prompted EMA surveys per day by smartphone
across four days. EMA prompting was customized to mothers sleep and wake times. EMA
surveys assessed stressors, diet, physical activity and sedentary screen behavior. Exit
interviews were conducted with each participant to further assess barriers and burden
associated with the EMA procedures.

Results Overall, participants responded to 82.05% (range: 0 100%) of EMA surveys. Older
mothers had higher compliance. Participants reported experiencing one or more stressors in
43.21% of prompts, and tension with children (25.6%) and work at home (15.4%) were
the most frequently reported sources. Across all prompts, the percentages reporting specific
weight-related behaviors were as follows: Intake of fruits and vegetables (7.8%), intake of
high-calorie, low nutrient foods (25.63%), sedentary screen behavior (44.4%), and
engagement in exercise or sports (1.8%). In exit interviews, mothers reported concern over
the compatibility of EMA prompting schedule with their variable sleep times.

Discussion This pilot study provides preliminary evidence that EMA methods may be
reasonable for use in low-income, Hispanic post-partum women with young infants and
multiple competing time demands. Future studies may use EMA to examine time-varying
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1209

influences on weight-related behaviors in this population to understand factors contributing

to post-partum weight retention.

CORRESPONDING AUTHOR: Sydney G. O'Connor, BA, University of Southern California, Los

Angeles, CA, 90032; [email protected]
S1210 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B180 6:00 PM-7:00 PM

WHICH CONTROLS VARIABLES DID YOU INCLUDE? DOES IT MATTER? THE CASE FOR UNIFORM
CONTROL LISTS IN PUBLIC HEALTH STUDIES

Theodore L. Caputi, N/A

University of Pennsylvania, Washington Crossing, PA

One of the greatest threats to the validity and reliability of public health research is the
selective inclusion (or exclusion) of control variables. Most studies include an in-depth
explanation of both the model's independent and dependent variables, with only a cursory
review -- sometimes as short as half a sentence -- of what other factors were controlled for in
the model. However, controls represent major assumptions of the model, and without
agreement on those assumptions, even high-quality research can easily be dismissed.

In this session, based on a 2016 publication in the International Journal of Drug Policy, I will
show how researchers can find two contradictory conclusions simply by selecting a different
vector of controls. For example, a recent study connecting strict alcohol policies to
methamphetamine use captured media attention from prominent news outlets as the
Economist and the Washington Post. And yet, with the addition of just one innocuous and
reasonable control, I show that virtually all of the author's significant results disappear.

So how do researchers address this systemic issue? I recommend that expert panels in
different disciplines of public health convene to create a list of uniform controls that should
be considered in all relevant studies. Peer reviewers should make a point to question authors
of the reasoning behind the inclusion or exclusion of those listed controls. This proposal
would not coerce authors into including controls they feel are irrelevant -- rather, it would
demand that researchers explicate the reasoning for their model's assumptions. In the
meantime, I will discuss best practices for choosing appropriate controls in public health
studies, based on research from Dr. Gary King.

CORRESPONDING AUTHOR: Theodore L. Caputi, N/A, University of Pennsylvania, Washington

Crossing, PA, 18977; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1211

B181 6:00 PM-7:00 PM

DEMOGRAPHIC PREDICTORS OF INDIVIDUAL DIFFERENCES IN COPING BEHAVIORS OF

SOLDIERS

Ellen P. Merrill, MA1, Adam J. DiChiara, BS2, Alan O. Wright, MS, CFS3, Adrienne Hatch, MS,
RD4, Renee E. Cole, PhD, RDN, LD5, Susan McGraw, BS6, Scott J. Mountain, PhD7, Harris R.
Lieberman, Ph.D. 7, Asma Bukhari, PhD, RDN8

1
Cognitive Science and Applications Team, US Army Natick Soldier Research Development &
Engineering Center, Sudbury, MA; 2U.S. Army Natick Soldier Research, Development &
Engineering Center, Holliston, MA; 3Natick Soldier Research, Development, and Engineering
Center, North Oxford, MA; 4Military Nutrition Division, U.S. Army Research Institute of
Environmental Medicine, Boston, MA; 5U.S. Army Research Institute of Environmental
Medicine, Natick, MA; 6US Army Research Institute of Environmental Medicine, Natick, MA;
7
USARIEM, Natick, MA; 8U.S Army, Bethesda, MD

Background: Most military personnel use dietary supplements (DS) and they are much more
likely to use dangerous DS than their civilian counterparts. Given the prevalence and
associated risks of DS use, it is important to understand specific trait characteristics of Soldiers
when considering these health-related behaviors. Purpose: To assess coping behaviors
associated with various demographic factors in a sample of active duty U.S. Army Soldiers and
how they relate to DS usage. Methods: Demographic and behavioral questionnaires were
administered to 289 Soldiers (83% male and 17% female, mean age SD 286y) at 3
installations. The validated Brief Cope Scale (dimensions: self-distraction, active coping,
denial, substance use, emotional support, instrumental support, behavioral disengagement,
venting, positive reframing, planning, humor, acceptance, religion and self-blame) was used to
assess coping behaviors. One-way ANOVAs compared effects of gender, age, education level
and marital status on various coping dimensions. A subset of Soldiers (n=128) participated in
focus groups to better understand the motivation for use or non-use of DS. Results: Seventy-
five percent of those sampled were DS users. No significant differences were found between
DS users and non-users on coping scale dimensions. Males were more likely than females to
use humor (p=.02) and acceptance (p=.03) while females were more likely to use emotional
support (p=.02), religion (p=.03) and self-distraction (pp=.00) and self-distraction (p=.03) while
older Soldiers were more likely to use religion (p=.00). Soldiers with higher levels of education
were more likely to use active coping (p=.00), emotional support (p=.04) and planning
behavior (p=.00); whereas Soldiers with lower levels of education were more likely to use
more behavioral disengagement, self-distraction and denial (p=.00) strategies. Married
S1212 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Soldiers were more likely to use religion (p=.00) while single Soldiers were more likely to use
self-distraction (p=.01). In focus groups, Soldiers stated they use DS to address job pressures
and meet the performance demands in their military roles to reach short-term goals, despite
the known and unknown long-term consequences of DS consumption. Conclusion: Younger,
less educated and single Soldiers use negative coping styles more frequently and may be more
vulnerable to questionable DS claims. Negative coping styles reportedly can increase risk in
stressful situations while it has been shown adaptive coping styles, such as emotional support,
are protective in such circumstances. Training to encourage positive coping behaviors early in
an individuals military career may improve resiliency, facilitate healthful behaviors and
provide short and long term benefits.

Supported by USAMRMC.

The views expressed in this abstract are those of the author(s) and do not reflect the official

policy of the Department of Army/Navy/Air Force, Department of Defense, or U.S.

Government.

CORRESPONDING AUTHOR: Ellen P. Merrill, MA, Cognitive Science and Applications Team, US
Army Natick Soldier Research Development & Engineering Center, Sudbury, MA, 01776;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1213

B182 6:00 PM-7:00 PM

POSTTRAUMATIC GROWTH AND BIOMARKERS OF CHRONIC STRESS AMONG COMBAT

VETERANS OF OPERATIONS ENDURING AND IRAQI FREEDOM

Cara L. Blevins, MA1, Richard G. Tedeschi, PhD1, Jeanette M. Bennett, PhD1, Amy Canevello,
PhD2, Christine Elnitsky, PhD, RN1

1
The University of North Carolina at Charlotte, Charlotte, NC; 2University of North Carolina,
Charlotte, Charlotte, NC

Combat Veterans are at increased risk for cardiovascular disease (CVD) and chronic illness due
to in-theater stressors. Growing evidence suggests that experiences of posttraumatic growth
(PTG), or positive psychological change in the aftermath of trauma, may convey salutary
health benefits; however, the relationship between PTG and Veteran health is currently
unknown. The present pilot study sought to elucidate this relationship by examining the
influence PTG on levels of inflammation and diurnal cortisol slope in Veterans of Operations
Enduring and Iraqi Freedom. We hypothesized that greater levels of PTG would buffer against
the adverse effects of stress as indicated by a steeper and more negative cortisol slope and
lower levels of C-reactive protein (CRP; mg/L). Participants (N=33) focused on their combat
experiences when completing self-report measures and provided saliva samples over the
course of 2 days to assess CRP and cortisol. Separate hierarchical multiple regression analyses
were run to predict CRP and the diurnal cortisol slopes from Day 1 and Day 2 of the study.
Consistent with existing theory, significant positive relations were found between PTG and
PCL scores (r(33)=.83, p < .001). The regression models predicting CRP and Day 1 cortisol slope
were not significant. The regression model predicting Day 2 cortisol slope revealed a
significant positive interaction between PTG and PTSD symptomology (=.001, p < .05).

Examination of the simple slopes revealed a disordinal interaction between PCL and PTG
scores such that the cortisol slope of veterans reporting a greater degree of PTG became
increasingly positive as PTSD symptoms increased. Relatedly, the cortisol slope of veterans
reporting a low degree of PTG became increasingly negative as PTSD symptoms increased.
Results of this pilot study provide preliminary evidence that experiences of PTG may influence
how Veterans respond to stress at a physiological level. The positive cortisol slope
demonstrated by those Veterans reporting high levels of PTG with greater trauma symptoms
may be a reflection of the processes by which one seeks to reconstruct their assumptive
worldview in the aftermath of trauma. That those Veterans reporting lower levels of PTG with
higher levels of trauma symptoms did NOT demonstrate higher levels of cortisol may indicate
a sign of dysregulation or stress desensitization. Future work is warranted before causal
assumptions can be made.
S1214 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Cara L. Blevins, MA, The University of North Carolina at Charlotte,
Charlotte, NC, 28223; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1215

B183 6:00 PM-7:00 PM

PROGRAM OF COMPREHENSIVE ASSISTANCE FOR FAMILY CAREGIVERS: CAREGIVERS

PERSPECTIVE ON PROGRAM VALUE

Nina Sperber, PhD1, Sara M. Andrews, MAT, MPH, RD2, Jennifer Henius, MSW, LCSW3

1
Veterans Affairs, Durham, NC; 2VA, Durham, NC; 3VA, Washington, DC

The Caregivers and Veterans Omnibus Health Services Act of 2010 (P.L. 111-163) directed VA
to provide sweeping supports for family caregivers of all Veterans within the Veterans Affairs
(VA) health system. This legislation is an innovation in service delivery that acknowledges the
important role of family caregivers in health care delivery. Stemming from this legislation, VA
caregiver support greatly expanded, with formation of the Program of Comprehensive
Assistance for Family Caregivers (PCAFC) and the Program of General Caregiver Support
Services, and funding of a Caregiver Support Coordinator (CSC) at every VA medical center
(VAMC). PCAFC provides services for caregivers of post-9/11 Veterans seriously injured in the
line of duty, including a monthly stipend and required training, and optional services such as
health insurance, mental health care, and respite care. We used mixed-methods to assess
how caregivers use and value PCAFC services to improve the programs ability to refine and
optimize services. We conducted a national web-based survey of caregivers (N=1,407) and
semi-structured phone interviews with caregivers from 10 VAMCs (N=50). Caregivers rated all
services as helpful, 7.9 to 8.8 on a scale of 1 to 10 (most helpful). They rated the stipend
highest, valuing it for providing flexibility to be physically present for their Veterans;
caregivers also valued trainings for instructing them on how to manage TBI/PTSD symptoms.
However, caregivers reported low use of optional services; e.g., only 7% used respite care,
with reports that Veterans with TBI/PTSD sometimes have difficulty with strangers in the
home. Seventy-five percent of Caregivers reported interacting with their CSCs since joining
the program, though for most (57%) less than once a month. A majority (59%) reported
obtaining referrals and direct support from their CSC, for example to help navigate the VA
health system. Almost two-thirds agreed/strongly agreed that the PCAFC has helped them to
feel more confident in caregiving, know about resources to help their caregiving, and support
their Veterans progress and health care engagement. Overall, caregivers valued receiving VA
support with particular gratitude for home support, PTSD/TBI symptom management, and
assistance with navigating the VA healthcare system. Future work could better evaluate and
guide use of optional services to support caregivers and Veterans, especially those with
conditions that impose barriers to use.
S1216 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Nina Sperber, PhD, Veterans Affairs, Durham, NC, 27705;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1217

B184 6:00 PM-7:00 PM

SMOKING IN U.S. MILITARY PERSONNEL DURING TRANSITION FROM ACTIVE DUTY TO

VETERAN STATUS

Chiping Nieh, PhD, CPH1, James Mancuso, MD, DrPH, MPH2, Gary D. Gackstetter, DVM, MPH,
PhD3, Marleen Welsh, PhD1, Teresa Powell, MS4, Tomoko I. Hooper, MD, MPH5

1
Uniformed Service University of the Health Sciences, Bethesda, MD; 2USUHS, DPO AE, N/A;
3
Uniformed Services Univ of the health Sciences, Fairfax, VA; 4Henry Jackson Foundation, San
Diego, CA; 5Uniformed Services University of the Health Sciences, Bethesda, MD

The prevalence of cigarette smoking has been consistently higher in the US military than in
civilian populations. Although risk factors for smoking behavior among Service members has
been studied, there is a gap in the literature on smoking behavior during transition from
active duty military to civilian life. We sought to investigate any change in smoking among
those transitioning from military to Veteran status.

Participants in the Millennium Cohort Study (longitudinal study of US military personnel) who
separated from service prior to first follow-up or remained in service for at least two survey
cycles were included (n=30,996). Using generalized estimating equations, we examined
associations between smoking and separation status as well as demographic, military, health
behavioral characteristics, and life stressors. Sixteen percent of participants reported smoking
at baseline, and 32% subsequently separated from service. Multivariable analysis showed that
Service members who separated were significantly less likely to be a current smoker (adjusted
odds ratio: 0.89, 95% confidence interval: 0.83-0.95) than those who remained in service.
Adjusted odds of smoking increased with higher number of life stressors (1 stressor: 1.05,
1.01-1.10, 2 stressors: 1.12, 1.06-1.19, 3 or more stressors: 1.25, 1.16-1.34 vs. 0 stressor).
Higher odds of smoking were also found among those with greater alcohol consumption,
being black non-Hispanic, not married, normal weight or underweight, Army, junior enlisted,
and screening positive for post-traumatic stress disorder or depression.

In our study, separation from service was protective for current smoking. This finding is
interesting as transition out of service might be stressful for some and result in maladaptive
coping mechanisms. Although results seem to indicate that staying in the military may
S1218 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

facilitate smoking behaviors, since the models were adjusted for life stressors, alcohol
consumption, and other risk factors, the effect of separation from service may have been
lessened. More research is needed to delineate modifiable risk factors and specific military
and Veteran subpopulations for targeted smoking interventions.

Disclaimer:

This work is supported by MOMRP under work unit no. 60002. The views expressed in this
abstract are those of the authors and do not necessarily reflect the official policy or position
of the Department of the Navy, Department of the Army, Department of the Air Force,
Department of Veterans Affairs, Department of Defense, or the U.S. Government. Approved
for public release; distribution unlimited. Human subjects participated in this study after
giving their free and informed consent. This research has been conducted in compliance with
all applicable federal regulations governing the protection of human subjects in research
(NHRC.2000.0007).

CORRESPONDING AUTHOR: Chiping Nieh, PhD, CPH, Uniformed Service University of the
Health Sciences, Bethesda, MD, 20814-4712; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1219

B185 6:00 PM-7:00 PM

SOCIAL SUPPORT AND MASTERY IN THE LINK BETWEEN FUNCTIONAL LIMITATIONS AND NEED
FOR HELP AMONG ILL AND INJURED MILITARY MEMBERS

Kimberley Watkins, MA, Jennifer Lee, PhD, Alla Skomorovsky, PhD

Department of National Defence, Ottawa, ON, Canada

Functional limitations (i.e., difficulties in conducting daily tasks) due to illness or injury tend to
be associated with a high perceived need for the help of others in performing activities, such
as meal preparation or personal hygienic care. However, other variables, such as ones
perceived social support and/or feelings of self-efficacy or mastery, may influence this
association between functional limitations and perceived need for help.

The present study investigated the moderating roles of social support and mastery in the
association between functional limitations and perceived need for help among 164 ill or
injured Canadian military members. Participants completed a paper survey measuring various
aspects of health, as well as psychosocial variables, including social support and mastery. To
assess functional limitations, participants were asked to report the general degree of
impairment and frequency of limitations experienced in various tasks. For illness- or injury-
related need for help, participants reported the frequency of requiring aid with various
activities.

Two moderation analyses were conducted using Hayes PROCESS macro for SPSS. Each
analysis included functional limitations as the independent variable and perceived need for
help as the dependent variable, while social support and mastery were included as the
hypothesized moderating variables in their respective analyses. Both analyses were
significant, and both showed the same effect on the positive association between functional
limitations and perceived need for help. Specifically, when social support/mastery was low,
this association was much stronger, while at high levels of social support/mastery, there was
much less of an association between functional limitations and perceived need. In other
words, at low levels of functional limitations, participants at all levels of social
support/mastery reported relatively low need for aid. However, when functional limitations
were high, those with low levels of social support/mastery reported high need for help while
those with high levels of social support/mastery reported low need for help.

Though the patterns were similar, the findings for social support and for mastery may be
interpreted differently. The results for social support were as expected: when social support is
high, the perceived need for help was low, regardless of functional limitations. For mastery,
S1220 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

however, results are potentially concerning, in that participants with high perceived mastery
reported a relatively low need for help, even when their functional limitations were high. To
the extent that their actual need for help exceeds their perceived need, such participants may
be less likely to seek and receive help. The implications of these findings for ill and injured
military personnel will be discussed.

CORRESPONDING AUTHOR: Kimberley Watkins, MA, Department of National Defence,

Ottawa, ON, K2H0B7; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1221

B186 6:00 PM-7:00 PM

THE ASSOCIATION BETWEEN COMBAT AND LATENT TRAJECTORIES OF PHYSICAL WELL-BEING

Ben Porter, PhD1, Geroge A. Bonanno, PhD2, Paul D. Bliese, PhD3, Susan P. Proctor, ScD4

1
Naval Health Research Center, San Diego, CA; 2Columbia University, New York, NY;
3
University of South Carolina, Columbia, SC; 4US Army Research Institute of Environmental
Medicine, Natick, MA

Combat exposure is associated with an increased risk of adverse mental and physical health
conditions. Little research has examined whether trajectories of physical health over time
within deployed service members can be separated into distinct classes or patterns and, if so,
whether these classes differ by deployment combat exposure experienced. Using the
Millennium Cohort Study, this investigation examined physical well-being assessed at four
timepoints between 2001 and 2012 among 11 950 (6013 combat-exposed) military service
members who deployed between their first and second assessment. Physical well-being was
measured using the physical component scale score calculated from the Veterans RAND-36.
Latent growth mixture modeling was used to estimate heterogeneous trajectories of physical
well-being. Results indicated that five distinct trajectories of physical well-being existed (high-
stable, low-stable, improving, delayed declining, worsening). Combat exposure was entered
into subsequent models as a known class to compare trajectories of combat-exposed and
unexposed service members. The shapes of trajectories between these two groups were
highly similar. The high-stable class was the modal class among combat-exposed and
unexposed participants (75.6% and 82.2%, respectively). However, chi-squared analyses
indicated a larger proportion of service members unexposed to combat were in the high-
stable class (ppsp=.45). Results indicate that physical well-being surrounding deployment has
several distinct trajectories. However, the shape of these trajectories does not vary much
between combat-exposed and unexposed service members. Physical health differences
between these groups are likely to be reflected by the larger proportion of participants with
poorer physical well-being trajectories.

CORRESPONDING AUTHOR: Ben Porter, PhD, Naval Health Research Center, San Diego, CA,
92106; [email protected]
S1222 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B187 6:00 PM-7:00 PM

TRANSLATING EVIDENCE-BASED PRACTICE GUIDELINES INTO PATIENT-CENTERED OBESITY

CARE

Jessica A. Lohnberg, Ph.D.1, Lauren M. Greenberg, Ph.D.2, Cheryl Bates, M.D.1, Jessica Y.
Breland, Ph.D.3, Susan Frayne, M.D.1, Dan Eisenberg, M.D., M.S.1

1
VA Palo Alto Health Care System, Palo Alto, CA; 2War Related Illness and Injury Study Center,
VA New Jersey Health Care System, East Orange, NJ; 3VA Palo Alto & Stanford University
School of Medicine, Menlo Park, CA

Background. Obesity, a leading cause of mortality, morbidity, disability, and rising healthcare
costs, affects roughly one-third of the millions of Veterans seeking treatment at VA medical
facilities. As such, expanded access to multidisciplinary specialty obesity care, including
bariatric surgery, is of paramount importance. New obesity treatment guidelines suggest
healthcare providers develop individualized weight loss plans incorporating diet, exercise, and
behavioral interventions for all patients, and bariatric surgery for individuals with body mass
index (BMI)>35 kg/m. For patients undergoing bariatric surgery, evidence-based clinical
practice guidelines recommend multidisciplinary pre- and post-op management. However,
little is known about how to implement such programs and whether patients are satisfied
with the care. We developed an innovative approach to weight management that provides
access to bariatric surgery care for Veterans from multiple VA facilities across the Western US.
Objective. We had two goals: 1) To describe the interdisciplinary MOVE TIME Clinic, which
offers a patient-centered approach to weight management before and indefinitely after
bariatric surgery in accordance with evidence-based guidelines. 2) To assess patient
satisfaction with this clinic.
Methods. As a program evaluation activity, clinic patients (N=54 sequential Veterans)
completed questionnaires assessing satisfaction with the clinic, satisfaction with functional
status as a result of care received, and satisfaction with access to care.
Results. The sample was primarily male (76%), 50-70 years of age (81%), and white (66%).
Almost all patients (98%) reported overall satisfaction with the clinic, with a majority (94%)
reporting that MOVE TIME improved their quality of life, and with two-thirds (65% of pre-
surgery patients and 74% of post-bariatric surgery patients) indicating that their health was
better than the previous year. Most said they would recommend the clinic to others (94%).
Conclusion. Given increasing obesity rates and demand for bariatric surgery, access to
interdisciplinary specialty care is essential in addressing the rising obesity prevalence among
Veterans. MOVE TIME is a novel clinic that provides intensive and accessible patient-centered
obesity care and is well-received by Veterans, with high rates of satisfaction and reports of
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1223

improved quality of life. This program can serve as a model for how to translate evidence-
based obesity and bariatric surgery guidelines recommending multidisciplinary pre- and post-
op management into patient-centered clinical practice. Such models are critical given the
epidemic of obesity among Veterans and VAs increasing efforts to provide expanded access
to multidisciplinary specialty obesity care.

CORRESPONDING AUTHOR: Jessica A. Lohnberg, Ph.D., VA Palo Alto Health Care System, Palo
Alto, CA, 94304; [email protected]
S1224 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B188 6:00 PM-7:00 PM

UNDERSTANDING HOW OEF/OIF VETERANS CONCEPTUALIZE HEALTH DURING AND AFTER

MILITARY SERVICE

Alicia Sparks, PhD1, Mira Brancu, PhD2, Jennifer M. Gierisch, PhD, MPH3, Meredith Stewart,
MPH, CPH4, Suzanne Maman, MHS, PhD5, Kristy Straits-Troster, PhD, ABPP6

1
Abt Associates, Silver Spring, MD; 2Mid-Atlantic VA Mental Illness Research, Education and
Clinical Center (VISN 6 MIRECC); Duke University Psychiatry & Behavioral Sciences
Department, Durham, NC; 3Duke Univeristy, Durham, NC; 4Duke University, Durham, NC;
5
University of North Carolina at Chapel Hill, Chapel Hill, NC; 6Dept. of Veterans Affairs,
Phoenix, AZ

Objectives: Combat veterans returning from Operation Iraqi Freedom and Operation Enduring
Freedom in Afghanistan have shown a high prevalence of readjustment problems that impact
health behaviors. Approximately 40% of OEF/OIF veterans reported difficulty acclimating to
civilian life on returning from deployment (5). Civilian life readjustment issues often manifest
in difficulty with social relationships (56%), problems finding or keeping a job (25%), elevated
drug and alcohol use (31%), divorce or separation (35%), and increased anger control
problems (57%) .This study sought to understand how veterans conceptualize health from
active duty to veteran status.

Methods: We conducted in-depth interviews with Operation Iraqi Freedom and Operation
Enduring Freedom veterans. Fourteen participants were recruited through the Department of
Veterans Affairs Mid-Atlantic Mental Illness Research, Education, & Clinical Center (VISN 6
MIRECC) Database for the Study of Post-Deployment Mental Health. We conducted a
qualitative content analysis of telephone interviews that were recorded and transcribed.
Codes were reviewed by a cross-disciplinary study team and themes across constructs were
analyzed.

Results: We found that health was defined in four categories: Physical and mental health,
appearance, and conduct. Participants enumerated challenges in the transition from military
to civilian life, including maintaining healthy habits and resisting unhealthy behaviors less
prevalent in the military. For the participants in this study, being healthy required either
addressing mental health symptoms or being devoid of them, fitting the military appearance
expectations of both weight and grooming, and following the conduct expected of them in the
military, even while in the civilian world. These four constructs seemed to remain from active-
duty service to becoming a veteran, even though participants had more difficulty maintaining
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1225

their health as civilians.

Conclusions: Findings from this study have policy implications for improving post-deployment
health as well as providing important context for primary care providers working with
veterans, their families and peers supporting veterans in efforts to retain healthy habits as
civilians. The most important implications are that preventive measures can be taken during
the transition from active duty to veteran status that will help maintain or improve a service
members health. Providers working with veterans need to understand what is considered
healthy in the military setting and that a veterans self-perception of their own health might
not reflect clinical definitions of health. Providing guidance on maintaining structure in a way
that can encourage the continuing of healthy behaviors might be valuable in creating a
smoother transition from military to civilian life.

CORRESPONDING AUTHOR: Alicia Sparks, PhD, Abt Associates, Silver Spring, MD, 20902;
[email protected]
S1226 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B189 6:00 PM-7:00 PM

VA TRUST AMONG US ARMY RESERVE/NATIONAL GUARD SOLDIERS WITH SUBSTANCE USE OR

MENTAL HEALTH DISORDERS

Jennifer Garcia-Cano, BA & BS1, D. Lynn Homish, BS2, Sarah Cercone. Cercone Heavey, MPH3,
Gregory G Homish, PhD2

1
University at Buffalo, San Bernardino, CA; 2University at Buffalo, Buffalo, NY; 3State
University of New York at Buffalo, Buffalo, NY

VA Trust among US Army Reserve/National Guard Soldiers with Substance Use or Mental
Health Disorders

Community Health & Health Behavior, University at Buffalo

Background: Trust in the healthcare system is an important factor in deciding to use

healthcare. Long wait times are functional barriers that can negatively impact trust in the
healthcare system. This can discourage individuals from seeking treatment. The objective of
this study is to examine trust in the VA Healthcare System and to determine if substance use
or mental health conditions impact soldiers trust. Additionally, this work will examine the
functional barrier of wait times on the basis of mental vs. physical health role limitations.

Methods: Data are from Operation: SAFETY (Soldiers And Families Excelling Through the
Years), an ongoing longitudinal study of US Army Reserve/National Guard Soldiers and their
partners. All soldiers were asked about their trust level in the VA Healthcare System. Mental
health disorders (PTSD, depression, anger, anxiety) and substance use were also assessed.
Soldiers who have had an appointment at the VA were asked about the length of wait time, a
factor which could impact trust. We examined differences in wait time on the basis of
emotional and physical health role limitations.

Results: At a trend level, trust scores were found to be lower among female soldiers who
were currently using drugs or drinking at hazardous levels. Among male soldiers, trust levels
were lower among those who were hazardous drinkers. Additionally, male soldiers who had
higher anger scores were more untrusting of the VA healthcare system. However, anxiety,
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1227

PTSD, and depression were unrelated to trust among both male and female soldiers. In terms
of wait times, female soldiers findings are as one would expect: as functional limitations
occurred for either physical or mental health conditions, wait times decreased. However, for
male soldiers, functional impairment for either physical or mental health conditions was not
related to wait times.

Conclusions: The current results suggest that mental health and substance use issues appear
related to less trust in the VA healthcare system. This fact, coupled with structural barriers
such as long wait times, could further impact trust and put our soldiers at risk for not receiving
treatment.

Supported by R01DA034072 (GGH) & R01DA0340724S1 (GGH).

CORRESPONDING AUTHOR: Jennifer Garcia-Cano, BA & BS, University at Buffalo, San

Bernardino, CA, 92410; [email protected]
S1228 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B190 6:00 PM-7:00 PM

WOMEN VETERANS PERCEPTIONS OF VA MENTAL HEALTH CARE

Joya G. Chrystal, MSW, LCSW1, Cynthia E. Gammage, BA2, Bevanne Bean-Mayberry, MD,
MHS3, Fay Saechao, MPH4, Susan Frayne, M.D.5, Elizabeth M. Yano, PhD, MSPH2, Alison B.
Hamilton, PhD, MPH6, John E. Zeber, PhD7, Pamela Smith-Beatty

1
Department of Veterans Affairs, Health Services Research & Development Center for the
Study of Healthcare Innovation, Implementation & Policy, Los Angeles, CA; 2Department of
Veterans Affairs, North Hills, CA; 3VA Greater Los Angeles, Noth Hills, CA; 4Center for
Innovation to Implementation (Ci2i), VA Palo Alto Health Care System, Menlo Park, CA; 5VA
Palo Alto Health Care System, Palo Alto, CA; 6VA Greater Los Angeles Healthcare System,
Culver City, CA; 7VA / Baylor Scott & White, Temple, TX

Background: Women Veterans are the fastest growing sector of eligible Veterans
Administration (VA) users and are increasingly accessing mental health services.

Objective: To assess women Veterans perceptions of their VA healthcare and decision

making processes related to healthcare coverage, as well as healthcare experiences within VA.

Design: Nine discussion groups in California, Texas, and Florida across five facilities.

Participants: Women Veteran participants (n=49) currently utilizing VA services.

Approach: Interviews were audio recorded, transcribed and summarized for salient themes.
Perceptions of women Veterans mental healthcare needs and suggestions to improve related
services within VA were assessed.

Key Results: Participantswere high utilizers of mental health services (68% saw a mental
health provider in past year). The majority had service-connected disability (70%) and used VA
exclusively for care (69%), though some dual-coverage participants sought care outside VA for
specialty care, such as mental health. Mental healthcare was identified as an area needing
improvement, e.g., a supportive infrastructure to better deliver services, including a service
climate sensitive to womens needs and stronger patient-provider relationships. Women
discussed desired improvements in mental health resources and program areas, including
additional individual therapy options, increased appointment duration and frequency, gender-
specific trauma-focused therapies, bereavement support, and childcare options during mental
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1229

health appointments. Desired improvements in service climate factors include streamlined

phone triaging, clarity on patient advocacy support, extended service hours, and provision of
transportation. Improvements were also desired in patient-provider relationships such as
enhanced communication skills, opportunities for building trust with rotating providers
(interns, residents), and ways to address privacy/confidentiality concerns especially for
women Veteran employees.

Conclusions: Women Veterans first-hand perspectives on desired improvements in mental

health care provide stakeholders with targeted service areas for infrastructure improvement
and greater healthcare engagement. These improvements may improve women Veterans
access to and sustained engagement in care. In an era where Veterans have expanded options
for their healthcare coverage, it is incumbent upon the VA to continue to respond to patient-
centered feedback on quality of care.

CORRESPONDING AUTHOR: Joya G. Chrystal, MSW, LCSW, Department of Veterans Affairs,

Health Services Research & Development Center for the Study of Healthcare Innovation,
Implementation & Policy, Los Angeles, CA, 90073; [email protected]
S1230 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B191 6:00 PM-7:00 PM

"THE FRESHMAN FIFTEEN" AND BEYOND: A META-ANALYSIS

Hannah Lammert, MA1, Rick LaCaille, PhD2, Lara LaCaille, PhD2

1
Pacific University, Beaverton, OR; 2University of Minnesota Duluth, Duluth, MN

The rising rates of overweight and obesity are concerning due to the increased risk for
negative health consequences. Not surprisingly, the transition to college is associated with
various changes (e.g., increased freedom to food choices, access to unhealthy foods,
decreased physical activity, altered life schedules) that may contribute to additional weight
gain in students compared to their non-student peers. The freshman fifteen is a colloquial
term used to specify typical weight gain experienced by students during their first year of
college and is concerning given that approximately 30% of college students are considered
overweight or obese. The present meta-analysis examined body weight and adiposity changes
over the first year as well as to the end of the senior year of college. A search of seven
electronic databases resulted in 55 peer-reviewed studies (from 1985 to 2015) for inclusion in
the meta-analysis, although only a subset of these were included because of missing data.
Using random-effects models to assess change in weight, an overall mean weight gain of 0.74
kg (CI = 0.56 0.92, k = 27) was found for freshman year across all students. However, the
average weight gain for only those students gaining weight was 3.43 kg (CI = 2.32 4.53, k =
4). Comparisons of five studies examining long-term weight change revealed a gain of 2.13 kg
(CI = 1.02 3.25) by the end of the senior year. Weight change differences between males
(1.47 kg; 0.90 2.05, k = 10) and females (1.31 kg, CI = 0.86 1.76, k = 19) were found during
the freshman year and continued through the senior year with gains of 2.89 kg (CI = 0.78
4.99) for males and 1.42 kg (CI = 0.69 2.14) for females. When sufficient data were available
for analysis, the findings with adiposity measurements generally paralleled the weight gain
results. Although the research does not support the notion of the freshman fifteen, a
considerable number of students gain weight in the first year. Cumulative weight gain
throughout college is notable and of continued concern because of the potential harmful
health behaviors and patterns that may carry into adulthood. These findings highlight the
need for greater focus being paid to developing targeted strategies to prevent weight gain
throughout students experiences with college.

CORRESPONDING AUTHOR: Hannah Lammert, MA, Pacific University, Beaverton, OR, 97006;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1231

B192 6:00 PM-7:00 PM

A DYNAMICAL SYSTEMS MODEL FOR UNDERSTANDING HOW THE INTRAUTERINE

ENVIRONMENT AFFECTS FETAL GROWTH IN AN INTERVENTION CONTEXT

Mohammad T. Freigoun, BS in Chemical Engineering1, Penghong Guo, Master in Chemical

Engineering2, Daniel E. Rivera, PhD2, Emily Hohman, PhD3, Abigail Pauley, MS Exercise
Science4, Danielle Symons Downs, Ph.D.5, Jennifer S. Savage, PhD6

1
Arizona State University, Mesa, AZ; 2Arizona State University, Tempe, AZ; 3Center for
Childhood Obesity Research, University Park, PA; 4Penn State University, Bellefonte, PA; 5The
Pennsylvania State University, University Park, PA; 6Penn State University, University Park, PA

BACKGROUND: The underlying mechanism for how maternal perinatal obesity and
intrauterine environment influence fetal development remains largely unknown. Dynamical
systems modeling is used to develop a fetal growth model within the context of an
intervention to manage gestational weight gain (GWG) that illustrates how maternal GWG
and the intrauterine environment influences fetal growth and ultimately, infant birth weight.

METHODS: The Healthy Mom Zone (HMZ) intervention is an individually-tailored, intensively

adaptive intervention aiming to manage weight gain among overweight/obese pregnant
women (OW/OBPW). Intensive longitudinal data (ILD) is collected daily (weight, physical
activity, self-reported dietary intake) and bi-weekly/monthly (sleep, eating/physical activity
self-regulatory behaviors) using mHealth tools. ILD is also collected on indicators of fetal
growth (six ultrasound scans per participant) and stress (bi-weekly urine). OW/OBPW (N=30)
will be randomized to treatment and control groups from approximately 8-36 weeks
gestation. Relying on this data and a previously developed maternal weight gain model, the
parameters of a system of differential and algebraic equations based on the First and Second
Laws of Thermodynamics are estimated to obtain a predictive model for fetal weight gain.

RESULTS: Estimated models from four completed participants allow contrasting model
predictions to available measurements of fetal weight and placental volume. These models
point to an acceptable goodness-of-fit (within the standard deviation of the measurements)
obtained from a semi-empirical approach relying on fixed parameter values available from the
literature (energy densities and efficiencies) with adjustable parameters estimated from the
data (fetal growth rate, back-calculated energy intake, and selected conversion factors).
Model enhancements (an improved equation for placental volume) and simplifications (a
single first-order differential equation for fetal weight) are obtained from this analysis, in
contrast to earlier models. The overall result is a parsimonious system of equations that
S1232 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

reliably predicts maternal GWG and fetal weight gain based on a sensible number of
assessments.

CONCLUSION: Dynamical systems modeling developed from ILD on fetal growth is feasible.
Findings from our dynamical models can be used in future studies to explore the underlying
mechanisms for how maternal perinatal obesity programs fetal development, related
metabolic disorders, and later obesity. Use of these models can inform the development of
intensively adaptive interventions for preventing childhood obesity.

CORRESPONDING AUTHOR: Mohammad T. Freigoun, BS in Chemical Engineering, Arizona

State University, Mesa, AZ, 85201; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1233

B193 6:00 PM-7:00 PM

A NOVEL STEPPED-CARE APPROACH TO WEIGHT LOSS: THE ROLE OF SELF-MONITORING AND

HEALTH LITERACY IN TREATMENT OUTCOMES

Jennifer Selensky, B.S.1, Robert Carels, Ph.D., MBA1, James J. Rossi, MA1, Chelsey Solar,
M.S./Ed.S1, Reid Hlavka, B.A.2, Emily Hefner, Student3

1
East Carolina University, Greenville, NC; 2East Carolina University, Winterville, NC; 3East
Carolina Universtiy, greenville, NC

Objectives: The aims of the current study were twofold: 1) examine the effectiveness of an
innovative three-step, stepped-care behavioral weight loss treatment, and 2) examine factors
that contribute to poor weight loss outcomes and the need for more intensive treatment.

Methods: The total sample for the study consisted of 53 individuals (87% female; 62%
Caucasian, 34% African American, 4% Asian/Pacific Islander) with MBMI=35.6, SDBMI=6.4.
Participants were recruited through a faculty and staff listserv email at a large southeastern
university and through a community newspaper with the advertisement of a free weight loss
program. A three-step, stepped-care treatment approach was implemented over six months.
Step 1 included the Diabetes Prevention Program manual adapted for self-administration and
the use of a Fitbit Zip device with the associated application/website to self-monitor their
diet, physical activity, and weight loss progress. Participants who were unsuccessful at
achieving predetermined weight loss goals received stepped-up treatments in 2-month
increments beginning at month 2. The stepped progression included the addition of meal
replacement at Step 2 and individual counseling concurrent with meal replacement at Step 3.
We assessed for contributing factors to poor weight loss outcomes using the following
measures: The Newest Vital Sign, Center for Epidemiological Study of Depression, Universal
Measure of Bias FAT, Weight Bias Internalization Scale, and the Binge Eating Scale.

Results: Non-stepped and once-stepped participants lost a clinically significant amount of

weight (8.7% and 6.0% weight loss over 6 months, respectively), while twice-stepped
participants lost an insignificant amount of weight (1.0% weight loss over 6 months). Twice-
stepped participants were significantly lower in health literacy (p = .03) and self-monitoring
frequency (p = .02).

Conclusions: In this investigation, approximately 60% of the participants were able to lose a
clinically significant amount of weight utilizing a minimally intensive intervention with little
additional support. Regular self-monitoring and high health literacy proved to be significant
correlates of weight loss success.
S1234 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Jennifer Selensky, B.S., East Carolina University, Greenville, NC,
27858; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1235

B194 6:00 PM-7:00 PM

A SCHOOL-BASED HEALTH CENTER OBESITY PREVENTION PROGRAM

Robin Whittemore, PhD, APRN, FAAN1, David Collett, MA, MAT, RN2, Sangchoon Jeon, PhD3

1
Yale University, School of Nursing, Orange, CT; 2Yale University School of Nursing, New
Haven, CT; 3Yale University, West Haven, CT

Background: Approximately 34% of youth are overweight or obese. School-based health

centers represent a system of care for providing health education and obesity prevention
programs. The purpose of this study was to describe the obesity-risk behaviors of youth 5-12
years of age and the feasibility of a school-based health center obesity prevention program
that was initiated by the school.

Methods: A one-group pre-post design was used to evaluate an obesity prevention program
promoting 9 hours of sleep, 5 fruits and vegetables, less than 2 hours of screen time, 1 hour of
physical activity, and 0 sugar-sweetened beverages/day. The program consisted of interactive
health education provided by nurses during lunch; lesson plans provided to teachers; weekly
health messages in morning announcements; and posters, bulletin boards, and newsletters
for parents. Data were collected on obesity-risk behaviors with established measures by
parents of youth 5-9 years and by students 10-12 years of age. BMI was calculated from
height and weight measures. Descriptive statistics and t-tests were used.

Results: There were a total of 345 youth whose parents did not opt-out of the program
evaluation. At the time of enrollment, youth in the school were 52% male, with 56% white
non-Hispanic, and 48% of students eligible for free lunch. At baseline, 44% of youth were
overweight or obese, 80% of youth ate < 5 servings of fruits/vegetables per day, with a mean
of 2.7 days/week of moderate-vigorous exercise. The program was challenging to implement
with competing responsibilities of nurses and teachers. The program did not improve obesity-
risk behaviors over time and BMI increased significantly over time (p < .01)

Conclusions: High rates of overweight/obesity and poor health behaviors continue to be

evident in youth of diverse race/ethnicity. School-based obesity prevention programs are
challenging to implement. Designated program staff, family involvement, and school health
policies may be necessary to ensure success with implementation and improvement in youth
outcomes.
S1236 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Robin Whittemore, PhD, APRN, FAAN, Yale University, School of
Nursing, Orange, CT, 06477; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1237

B195 6:00 PM-7:00 PM

AN ECOLOGICAL MOMENTARY ASSESSMENT STUDY OF INTERNAL AND EXTERNAL WEIGHT

STIGMA EVENTS IN A WEIGHT LOSS TREATMENT SEEKING SAMPLE

James J. Rossi, MA1, Chelsey Solar, M.S./Ed.S1, Jennifer C. Selensky, Bachelor of Science1, Reid
Hlavka, B.A.2, Robert Carels, Ph.D., MBA1

1
East Carolina University, Greenville, NC; 2East Carolina University, Winterville, NC

Objectives: This study sought to explore temporal relationships between experiences of

internal and external weight stigma and mood and coping in real time using ecological
momentary assessment (EMA). Furthermore, we sought to explore cross sectional differences
in binge eating, levels of internalized weight bias and lifetime history of experienced stigma
between those individuals who reported instances of weight stigma during a 2-week pre-
weight loss intervention EMA study and those who did not.

Methods: 51 BWL participants completed EMA assessments of weight stigma events as well
as their emotional and coping responses. Participants were asked to report external instances
of weight stigma, including hearing inappropriate/rude comments about their weight,
encountering structural or physical barriers in their environment, or internal experiences or
personal reminders that made [you] feel bad about your weight (eg. trying on clothes,
looking in the mirror, etc.). Participants also completed random prompt entries assessing
mood at various times not related to weight stigma events. In addition participants completed
measures at baseline including the Binge Eating Scale (BES), Weight Bias Internalization Scale
(WBIS) and the Stigmatizing Situations Inventory (SSI).

Results: 30 of 51 participants reported 80 instances of experienced weight stigma over the

course of two weeks. The majority of stigma experiences were personal reminders of ones
weight (74%). Given the limited amount of observations per participant, responses were
aggregated and between subjects differences were analyzed between those who reported
weight stigma over the two week study period and those who did not. In those that
experienced weight stigma, stigmatizing experiences were associated with lower levels of
positive emotion (less happy t(27)=6.75, p < .01, less inspired t(27)=4.51, p < .01, less relaxed
t(28)=4.07, p<.01) and higher levels of negative emotion (more depressed t(27)=3.04, p=.01,
more ashamed t(27)=2.61, p=.02 and more frustrated t(26)=-2.92, p=.01). Participant coping
style was also related to mood. Isolation and avoidance was significantly associated with
higher levels of feeling depressed (r=.40, p<.05), feeling ashamed (r=.52, p < .01), and was
inversely associated with feeling happy (r=-.38, p < .05). Furthermore, there was an inverse
association between positive self talk and feeling numb (r=-.37, p < .05). Participants
S1238 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

reporting any weight stigma event over the two week study period were more likely than non-
reporters to evidence higher levels of experienced lifetime weight stigma t(34.8)=2.54, p=.02,
internalized weight bias t(53)=2.87, p < .01 and binge eating t(50)=2.71, p < .01.

Discussion: This study suggests that stigmatizing experiences have momentary impacts on
psychological well-being and coping among participants with overweight and obesity.

CORRESPONDING AUTHOR: James J. Rossi, MA, East Carolina University, Greenville, NC,
27858; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1239

B196 6:00 PM-7:00 PM

AUTONOMY SUPPORT BY ROMANTIC PARTNERS IS ASSOCIATED WITH DECREASED

CONTROLLED REGULATION FOR WEIGHT LOSS AMONG HIGH BMI WOMEN

Katelyn M. Gettens, M.A.1, Noemie Carbonneau, Ph.D.2, Theodore A. Powers, PhD3, Richard F.
Koestner, PhD4, Amy A. Gorin, Ph.D.1

1
University of Connecticut; Institution for Collaboration on Health, Intervention, and Policy,
Storrs, CT; 2University of Quebec in Trois-Rivieres, Trois-Rivieres, PQ, Canada; 3University of
Massachusetts Dartmouth, N Dartmouth, MA; 4McGill University, Montreal, PQ, Canada

Background: Self-Determination Theory and ecological models strongly suggest that provision
of autonomy support (AS) by existing social networks engenders successful health behavior
change. This dual-study analysis examined the impact of romantic partners' autonomy
support on women's reasons for weight control, particularly the extent to which women felt
pressured or controlled to lose weight. Methods: Study 1 participants (n=159) were couples
from the general population (MageM= 44.113.1; MageF = 43.6312.1 years; MBMI =
26.55.3kg/m2; 97.8% Caucasian; 61% married). Men self-reported their provision of
autonomy support (AS). Controlled regulation (i.e., more externally motivated reasons for
weight loss) was measured using the Regulation of Eating Behaviors Scale and was regressed
onto: Step 1, women's mean-centered BMI and men's mean centered provision of AS; and
Step 2, the product term for the women's BMIxMen's AS interaction. Study 2
participants (N=62; 79% female, 86% Caucasian; MageF = 47.311.3 years; MBMI= 36.76.2
kg/m2) were females in a behavioral weight loss intervention that included modifications of
the home environment. Participants completed perceived AS and Sallis Support Scales in
reference to their partners' support. Controlled regulation (measured using the 12-item
Treatment Self Regulation Questionnaire) was regressed onto: Step 1, women's mean-
centered baseline BMI and increase in men's AS from baseline to 18 months; and Step 2, the
product term for the women's baseline BMIMen's increase in AS interaction. Results: Study
1: Womens BMIpartners provision of AS interaction was significant (p = .0001). For high BMI
women, mens provision of AS was negatively and significantly related to womens controlled
regulation (p = .0002). Study 2: Mens increase in AS was marginally related to a decrease in
womens controlled regulation ( = -.241, p = .064). Addition of the two-way interaction term
(womens baseline BMI x partners increase in AS) significantly increased explained variance
(F(1, 57) = 8.551, p = .005, R2 = .123)). Simple slopes analyses indicated that when womens
baseline BMI is high, mens increase in AS is significantly related to decreased controlled
regulation among women over 18 months (t(57) = -3.3361, p = .0015)). Conclusion: Greater
S1240 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

provision of autonomy support from male romantic partners, as rated from both male and
female perspectives, is associated with decreased controlled regulation, and is particularly
influential among the highest BMI women. Lower controlled motivation is expected to be
associated with better emotional regulation and superior outcomes. Our results indicate that
this pattern is reflected both in the general population and within a weight loss intervention.

CORRESPONDING AUTHOR: Katelyn M. Gettens, M.A., University of Connecticut; Institution

for Collaboration on Health, Intervention, and Policy, Storrs, CT, 06269;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1241

B197 6:00 PM-7:00 PM

COLD FEET AT THE THRESHOLD OF WEIGHT LOSS TREATMENT: PRE-RANDOMIZATION

DROPOUT AMONG THOSE ELIGIBLE AND CONSENTING TO TREATMENT

Sara A. Hoffman, B.A.1, Christine Pellegrini, PhD2, Angela Fidler Pfammatter, PhD1, Amanda
Paluch, PhD3, Linda M. Collins, PhD4, Bonnie Spring, PhD5

1
Northwestern University Feinberg School of Medicine, Chicago, IL; 2University of South
Carolina, Columbia, SC; 3Northwestern University, Feinberg School of Medicine, Chicago, IL;
4
Penn State, University Park, PA; 5Northwestern University, Chicago, IL

A minority (10-20%) of those screened for weight loss treatment research reach
randomization to a study condition. Among those eligible and consenting to participate, we
examined differences between individuals who dropped out prior to and those who
proceeded on to randomization. Our aim was to understand bias and limitations to
generalizability arising from late-stage pre-randomization study withdrawal. Participants were
Chicago-area residents with a BMI of 25-40 who indicated interest in a weight loss study,
attended an orientation session, gave informed consent, and were deemed eligible. All were
offered 6 months of treatment with a smartphone app and calls from a health coach. Each
participant was randomized to a combination of other helpful strategies to augment
treatment; there was no inactive group. These analyses compared consented eligible
individuals who proceeded to be randomized and begin treatment (Randomized), versus
those who chose not to proceed (Non-Randomized). Individuals were compared on age,
BMI, smartphone type (Android vs. iPhone) and self-reported exercise coded from 0 (do not
exercise or plan to) to 5 (exercise 4+ days a week). We hypothesized that Randomized
participants would have more motivation and fewer barriers to weight loss (i.e., older, lower
BMI, more likely to report regular exercise) than Non-Randomized participants. We also
examined differences in smartphone ownership (iPhone vs. Android). Study candidates who
attended an orientation session and consented as of December 2016 [n=931, mage=36.51 yrs,
mBMI=31.88 kg/m2] were examined. Of these, 519 were Randomized and 412 were Non-
Randomized. Independent-samples t-tests and Chi-square showed Non-Randomized
participants were younger (M=33.96, SD=10.75) than Randomized (M=38.55, SD=10.81; p <
.001). Non-Randomized (M=2.74, SD=.97) versus Randomized (M=2.96, SD=1.07) participants
were less likely to exercise (p=.001). The groups did not differ on BMI (p=0.7) or smartphone
type (p=.21). Treatment-seeking eligible and consenting weight loss volunteers who drop out
prior to randomization tend to be younger and report less exercise than those who persist
through randomization, but do not differ on BMI. Future studies should be aware of selective
potential for dropout among these participants during screening and engage in participant-
S1242 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

centered retention methods. By further exploring these characteristics, researchers may

enhance generalizability of weight loss study findings.

CORRESPONDING AUTHOR: Sara A. Hoffman, B.A., Northwestern University Feinberg School

of Medicine, Chicago, IL, 60611; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1243

B198 6:00 PM-7:00 PM

EFFECTIVE RECRUITMENT CHANNELS IN AN ONGOING SOCIAL MEDIA WEIGHT LOSS TRIAL FOR
UNIVERSITY STUDENTS

Jessica Whiteley, PhD1, Meghan Mavredes, MPH2, Jamie Faro, MS3, Laura L. Hayman, PhD,
RN, FAAN, FAHA4, Melissa Napolitano, PhD5

1
UMass Boston, Boston, MA; 2The George Washington University, Washington, DC;
3
University of Massachusetts Boston, Newton, MA; 4University of Massachusetts Boston,
Boston, MA; 5The George Washington University, Washington DC, DC

Examining the effectiveness of recruitment strategies utilized in intervention studies may lead
to more efficient and cost effective recruitment strategies for future trials. University
students initiated or completed an online screener to determine eligibility for a weight loss
trial. Following initial online consent, students were asked to indicate how they heard about
the study: email, flyer/poster, Facebook post, Facebook advertisement, in-person event,
friend, other. A total of 898 (23.254.01 years old, 42% non-white, BMI 28.66.4 kg.m-2)
students indicated how they heard about the study. Most participants (92.5%, n=831)
indicated they heard about the study from only one source with the following rank order
among the choices: email (77%, n=690), flyer/poster (16%, n=142), other (4.3%, n=39), a
friend (4.0%, n=36), in-person event (3.7%, n=33), Facebook post (3.5%, n =31), and Facebook
ad (0.7%, n=6). Approximately 13% (n=116) were randomized into the trial (22.263.5 years
old, 54% non-white, BMI 31.73.87 kg.m-2). Among those randomized, most participants
continued to indicate that they heard about the study from only one source (87%). The ways
in which participants heard about the study were more distributed but retained their rank
order: email (64.7%, n=75), flyer/poster (28.4%, n=33), other (8.6%, n=10), a friend (6.0%,
n=7), in-person event (3.4%, n=4), Facebook post (2.6%, n =3), followed by a Facebook ad
(0.9%, n=1). Results indicated that emails, either in the form of blast emails or emails
distributed by professors, clubs, or organizations, were the most successful as well as cost
effective method for reaching university students. It is recommended that researchers track
and monitor recruitment efforts in real time so that they may realize increased cost effective
participant yields in ongoing trials while also collecting useful data to inform recruitment
efforts and cost estimates for future trials.

CORRESPONDING AUTHOR: Jessica Whiteley, PhD, UMass Boston, Boston, MA, 02125;
[email protected]
S1244 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B199 6:00 PM-7:00 PM

EFFECTS OF HIGH-INTENSITY INTERVAL TRAINING ON WEIGHT LOSS AND BODY COMPOSITION

IN OBESE POST-MENOPAUSAL WOMEN

Joan A. Grossman, Registered Dietitian1, Danielle Arigo, Ph.D., Licensed Psychologist1, Jessica
L. Bachman, PhD, RD2

1
The University of Scranton, Scranton, PA; 2University of Scranton, Scranton, PA

Many women experience weight gain as a result of menopause. Exercise can mitigate weight
gain and associated cardiometabolic risks. This study examined the impact of high-intensity
interval training (HIIT) among obese, sedentary, post-menopausal women in a 16-week
behavioral weight loss program. Participants (n=11, MAGE=59, MBMI=32.0 kg/m2) were
assigned to one of two exercise groups (HIIT vs. Endurance). The HIIT group (n=6) engaged in
15 minutes of daily resistance exercises (DVD guided), progressing to a total of 75 minutes
weekly. The E group (E; n=5) engaged in 15-50 minutes of aerobic exercise (e.g., brisk walking)
at 65% of age-predicted HRMax, progressing to a weekly total of 250 minutes. All participants
wore a wristband physical activity monitor and were prescribed a restricted diet (1200-1500
calories). Measurements completed at baseline, mid- and end program included body
composition (BC), body weight (BW), % body fat (%BF), fat mass (FM), fat-free mass (FFM),
and anthropometric measures (size of waist, abdomen, hips, thigh, biceps). All participants
received weekly behavior change information through an online platform and attended both
weekly weigh-ins and monthly treatment group meetings. Across groups, average weight loss
was 11.9 pounds (6.7% of baseline BW); average total inches lost was 22.4. Both groups
significantly decreased waist circumference (psF[2,8]=33.83, pF[2,8]=85.44, pF[2,8]=5.07,
p=0.04). The HIIT group also showed a slight decrease in mean calories consumed and W/H
ratio. Meanwhile, the E group showed an overall decrease in BMI; FFM and W/H ratio did not
show significant change. Findings suggest that, relative to endurance exercise, HIIT may be
more effective for weight loss and body composition change among obese, sedentary, post-
menopausal women. Consequently, there is need for large-scale tests of the benefit of HIIT for
reducing disease risk in this population.

CORRESPONDING AUTHOR: Joan A. Grossman, Registered Dietitian, The University of

Scranton, Scranton, PA, 18510; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1245

B200 6:00 PM-7:00 PM

EXAMINING THE DOSE-RESPONSE RELATIONSHIP IN THE VETERANS HEALTH

ADMINISTRATIONS MOVE! WEIGHT MANAGEMENT PROGRAM

Stephanie H. Chan, MPH1, Susan D. Raffa, PhD2

1
Department of Veterans Affairs, Bedford, MA; 2Department of Veterans Affairs, Durham, NC

BACKGROUND: Clinical practice guidelines recommend intensive, multicomponent behavioral

interventions for individuals who are obese or overweight with an obesity-associated
condition. For more than a decade, the Veterans Health Administration (VHA) has offered
MOVE!, an evidence-based comprehensive lifestyle intervention designed as a series of
sessions.

OBJECTIVE: To inform VHA policy and planning, this VHA-wide study examines the relationship
between MOVE! participation and weight outcomes.

DESIGN: A longitudinal observational study of Veterans across VHA who participated in the
MOVE! Weight Management Program.

PARTICIPANTS: Veterans who initiated their most recent episode of weight management with
MOVE! between 2004 and 2014 were included in this study.

MAIN MEASURES: Weight measurements were abstracted from VHA electronic health
records. The primary outcome was the proportion of Veterans with clinically relevant weight
loss. The predictor of interest was number of MOVE! contacts during the 12 months following
MOVE! initiation.

KEY RESULTS: The cohort consisted of 237,577 Veterans (87.4% male; mean age 54.4 years),
who had, on average, 5.3 contacts in the 12 months following initiation. Veterans with
between 2 and 5 contacts had the same odds of achieving clinically relevant weight loss as
Veterans with only one contact (Adjusted OR:1.00, 95%CI:0.98,1.03). Veterans with between 6
and 9, 10 to 13, 14 to 17, and 18 or more contacts had significantly higher odds of clinically
relevant weight loss (Adjusted OR:1.17, 95%CI:1.13,1.20; Adjusted OR:1.53, 95%CI:1.47,1.59;
Adjusted OR:1.84, 95%CI:1.74,1.94; Adjusted OR:2.21, 95%CI: 2.12, 2.31, respectively)

CONCLUSIONS: Veterans with 6 or more MOVE! contacts in the year following treatment
initiation were significantly more likely to achieve clinically relevant weight loss than those
with 1 contact, with participation increasing odds of clinically relevant weight loss. Further
S1246 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

characterization of key predictors of weight loss is needed to understand how contact

frequency may modify weight change and associated implications for increasing engagement.

CORRESPONDING AUTHOR: Stephanie H. Chan, MPH, Department of Veterans Affairs,

Bedford, MA, 02155; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1247

B201 6:00 PM-7:00 PM

KNOWLEDGE AND AWARENESS OF METABOLIC RISK IN UNIVERSITY STUDENTS

Meghan Mavredes, MPH1, Loretta DiPietro, PhD, MPH1, Laura L. Hayman, PhD, RN, FAAN,
FAHA2, Samuel J. Simmens, PhD3, Jessica Whiteley, PhD4, Melissa Napolitano, PhD5

1
Milken Institute School of Public Health, The George Washington University, Washington, DC;
2
University of Massachusetts Boston, Boston, MA; 3George Washington University,
Washington, DC; 4UMass Boston, Boston, MA; 5The George Washington University,
Washington DC, DC

Background. The prevalence of young adults having one or more components of metabolic
syndrome, (MetS); i.e., increased abdominal adiposity, high blood pressure, high fasting blood
glucose, high triglycerides, body mass index [BMI] overweight/obese, is increasing.
Understanding knowledge and awareness of risk is likely an important first step in preventing
the numerous chronic conditions associated with MetS, including cardiovascular disease,
diabetes, and stroke. Purpose. We examined the knowledge of metabolic risk in a sample of
university students with BMIs in the overweight or obese range. Methods. Students (N=177;
79.7% female; age 22.13.6 years; BMI 31.73.6 kg/m2) completed a 5-item measure
assessing knowledge of metabolic risk, and self-perception or awareness of metabolic risk
including knowledge of blood pressure, cholesterol, and blood glucose. Results. The majority
(68.4%, n=121) of participants had never heard of the term metabolic risk, with 66% of obese
(n=68) and 71.6% overweight (n=53) reporting being unfamiliar with the term. There were no
significant differences in knowledge of metabolic risk by BMI status, sex, age, or academic
year. Of those individuals who had prior knowledge of metabolic risk, 53.6% reported they
were not concerned with their own metabolic risk factors. The majority of all participants
reported knowing their blood pressure (66.1%) and having blood work done to measure
cholesterol (52.0%), but only 41.7% reported ever having blood work to measure glucose
levels. Conclusions. Although participants reported having components of MetS evaluated, a
large proportion of students did not recognize the term metabolic risk, despite having
elevated BMIs in the overweight and obese range. This disconnect is an important area of risk
communication and information dissemination. Implications of these findings will be
discussed such as leveraging university healthy lifestyle programming to educate young adults
about metabolic risk and the benefits of forming long-term healthy eating and physical activity
behaviors.
S1248 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Meghan Mavredes, MPH, Milken Institute School of Public Health,
The George Washington University, Washington, DC, 20052; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1249

B202 6:00 PM-7:00 PM

METABOLIC SYNDROME MARKERS AND PHYSICAL ACTIVITY IN OBESE WITH AND WITHOUT
PRADER-WILLI SYNDROME

Kelsey McAlister, Student1, Daniela A. Rubin, Ph.D.2, Koren L. Fisher, Assistant Professor2,
Marilyn C. Dumont-Driscoll, M.D., Ph.D.3

1
California State University, Fullerton, Pasadena, CA; 2California State University, Fullerton,
Fullerton, CA; 3University of Florida Gainesville, Gainesville, FL

Introduction: Metabolic syndrome (MS) is defined as a group of cardiovascular and diabetes

risk factors that occur simultaneously. Prader-Willi Syndrome (PWS) is a genetic disorder
associated with obesity. To date, only one study showed that MS could be a health risk in
PWS. In youth, physical activity (PA) is negatively associated to MS; however, youth with PWS
appear to participate in less PA than other youth. Thus, the purpose of this study is to
examine the relationship between MS and moderate-to-vigorous PA (MVPA) in youth with
PWS compared to youth with non-syndromic obesity (OB).

Methods: Participants were 21 youth with PWS (ages 10.7 2.6; 12 males, 9 females; body
fat= 46.0 8.9%) and 34 youth with OB (ages 9.6 1.0; 17 males, 17 females; body fat= 45.4
6.4%). MS markers (body mass index (BMI), blood pressure, triglycerides, high-density
lipoprotein, and glucose) were obtained in each child and evaluated using the adolescent
adaptation of the National Cholesterol Education Program. Accelerometry was used to
measure MVPA for eight consecutive days. Cut-points formed by Evenson et al. (2008) defined
MVPA. Chi-square analyses and independent t-tests compared frequency of meeting the MS
criteria and differences in MS severity between PWS and OB youth, respectively. A median-
split separated participants into four groups (PWS low PA, PWS high PA, OB low PA, OB high
PA). An ANOVA determined differences in MS severity in relation to MVPA with a Bonferroni
post hoc. Significance was set at p < 0.05.

Results The frequency of presenting BMI and glucose as a risk factor was lower in PWS
(61.9% and 14.3%, respectively) than in OB (91.2% and 44.1%, respectively) (BMI 2=6.952, p <
0.01; glucose 2=5.247, p < 0.05). MS severity was lower in PWS (0.10 0.81), than in OB (0.77
0.63, p < 0.01). The PWS high PA youth had lower MS severity scores (-0.01 0.71) than the
OB low PA youth (0.83 0.52, p < 0.05). There were no other significant differences.

Conclusion: The results indicate that the presence of PWS may have protective effects against
MS risk factors in youth with obesity. The evidence also suggests PA may have a role in MS
S1250 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

severity in youth with PWS and non-syndromic obesity. Further research should evaluate the
role of exercise interventions on MS risk factors in youth with PWS.

CORRESPONDING AUTHOR: Kelsey McAlister, Student, California State University, Fullerton,

Pasadena, CA, 91105; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1251

B203 6:00 PM-7:00 PM

PARENTING PROFILES ACROSS OBESITY-RELATED BEHAVIOURS THAT SUPPORT ADOLESCENTS'

HEALTH BEHAVIOURS

Louise C. Masse, PhD1, Nicole S. Carbert, MSc Candidate1, Kevin Lu, MPH1, Patti-Jean Naylor,
PhD2, Heather McKay, PhD3, Lucy Le Mare, PhD4, Rhona Hanning, PhD5

1
School of Population and Public Health/University of British Columbia/BC Children's Hospital
Research Institute, Vancouver, BC, Canada; 2School of Exercise, Physical and Health
Education/University of Victoria, Victoria, BC, Canada; 3Department of Orthopaedics and
Family Practice/Centre for Hip Health and Mobility/University of British Columbia, Vancouver,
BC, Canada; 4Faculty of Education/ Centre for Research on Early Child and Health Education
(CRECHE)/Simon Fraser University, Burnaby, BC, Canada; 5School of Public Health and Health
Systems/University of Waterloo, Waterloo, ON, Canada

Objective: The family environment is important in shaping adolescent health behaviours.

Parents can influence these behaviours directly through specific parenting practices (e.g.,
rules or routines). Further research is necessary to explore the extent to which parents use
more supportive practices with respect to influencing health behaviours. We used qualitative
methods to explore parenting practices in relation to adolescent health behaviours (physical
activity (PA), nutrition, and screen time behaviours).

Methods: We conducted 28 semi-structured interviews with parents of grade 7 students

(March-May 2016) in Surrey, British Columbia, Canada [68% mothers; ethnicity diverse: 11%
Chinese, 32% South Asian, 14% Southeast Asian, 25% White; 54% had income < $60,000]. We
asked open-ended questions regarding the environmental and behavioural factors that
influence their child's health behaviours. Interviews were recorded and transcribed
verbatim.Two staff and the principal investigator transcribed and coded the first five
interviews, while the remaining interviews were coded and reconciled by two staff. A
conceptual framework that grouped parenting practices under structure, control, and
responsiveness guided our analyses. We used Nvivo 11 for all analyses.

Results: Parenting practices predominantly related to structure and control were observed
across all health behaviours; differences also emerged across health behaviours. For PA,
adolescent behaviours were structured using practices such as co-participation, modeling, and
access/availability. In some cases, parents controlled their childs PA by pressuring them to
enroll in activities that they were not interested in. Other parents were responsive to their
child and used practices such as encouragement, involvement, and autonomy support to
S1252 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

influence their childs activity. In contrast, screen time and nutrition behaviours were
structured through expectations (e.g., rules). As parents experienced difficulty in achieving
certain behaviour for screen time, they sometimes resorted to controlling practices such as
bribes or discipline. With nutrition, we identified more permissive practices. Parents
acknowledged certain expectations around eating, but noted difficulty in maintaining these
expectations both for their child and themselves.

Conclusions: Our findings suggest that parenting practices differ across adolescent health
behaviours. They provide insight into the complexity of ensuring that parents utilize
supportive practicesan aspect that needs to be emphasized in current interventions to
encourage healthy behaviours.

CORRESPONDING AUTHOR: Louise C. Masse, PhD, School of Population and Public

Health/University of British Columbia/BC Children's Hospital Research Institute, Vancouver,
BC, V6H 3V4; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1253

B204 6:00 PM-7:00 PM

POSITIVE EMOTIONS AND EATING IN OVERWEIGHT ADULTS

Abby Braden, PhD, Tanya Watford, MS, Elizabeth Emley, BS

Bowling Green State University, Bowling Green, OH

Background: Recently, there has been increased interest in examining positive emotions as
triggers for eating. Studies have concluded that eating in response to positive emotions may
be a greater problem in normal or underweight samples than overweight samples. However,
eating in response to positive emotions is frequently reported by overweight adults. The
current study aimed to identify the positive emotions that trigger eating in this sample, and to
examine differential associations of eating in response to positive vs. negative emotions with
demographic and health outcomes.

Methods: Overweight adults (n = 188; BMI = 33.17 + 6.98; age = 41.78 + 13.61; 64.9% female)
were recruited using Amazon Mechanical Turk. Eating due to positive and negative emotions
was assessed with the Emotional Appetite Questionnaire (EAQ). Body mass index (BMI) was
calculated from self-reported height and weight. Mental health symptoms (i.e., depression,
anxiety, and somatization) were measured with the SCL-90R. Mean scores were calculated to
identify positive emotions that most commonly trigger eating. Next, correlational analyses
examined relationships between positive and negative emotional eating and outcomes.

Results: Of the positive emotions assessed, happiness was the eating trigger most commonly
reported (M = 5.2, SD = 1.7). Of the 15 total emotions assessed, happiness was the fourth
most common emotion reported to trigger eating, after boredom, loneliness, and depression.
Eating due to positive emotions was unrelated to BMI, gender, and depressive symptoms, but
significantly related to symptoms of anxiety (r = .15, p < .05) and somatization (r = .15, p <
.05). Eating due to negative emotions was significantly associated with greater BMI (r = .19, p
=.01), female gender (r = .19, p =.01), and symptoms of depression (r = .32, p < .001), anxiety,
(r = .22, p < .001), and somatization (r = .19, p =.01).

Conclusions: Happiness frequently triggers eating in overweight adults. Despite this, eating
due to positive emotions was unrelated to BMI, gender, and depressive symptoms, whereas
eating due to negative emotions was significantly related to BMI, female gender, and a range
of mental health symptoms. Among overweight/obese adults, eating due to positive
emotions may be less problematic, as it is less closely related to weight and depressive
symptoms than eating due to negative emotions.
S1254 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Abby Braden, PhD, Bowling Green State University, Bowling
Green, OH, 43403; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1255

B205 6:00 PM-7:00 PM

RACIAL/ETHNIC REPRESENTATION IN LIFESTYLE WEIGHT LOSS INTERVENTION STUDIES IN THE

UNITED STATES: A SYSTEMATIC REVIEW

Christina Haughton, MPH1, Valerie J. Silfee, PhD1, Monica Wang, PhD2, Andrea Lopez-Cepero,
BS, MHSN1, David Estabrook, MPH1, Christine Frisard, MS1, Milagros C. Rosal, PhD1, Sherry
Pagoto, PhD1, Stephenie Lemon, PhD1

1
University of Massachusetts Medical School, Worcester, MA; 2Boston Univeristy, Boston, MA

Objective: Despite efforts to enhance inclusion, underrepresentation of minorities in research

has been documented. The primary aim of this review was to evaluate representation of
racial/ethnic sub-group members in behavioral weight loss interventions conducted among
adults in the United States. The secondary aims were to assess recruitment and study design
approaches to include racial/ethnic groups and the extent of racial/ethnic sub-group analyses
conducted in these studies.

Methods: PubMed, PsycInfo, and Medline were searched for behavioral weight loss
intervention trials conducted in 2009-2015 using keywords: weight, loss, overweight, obese,
intervention and trial.

Results: The majority of the 87 studies reviewed included a majority White sample. Across the
included studies, 61% of participants were White, 18% were Black/African American, 9% were
Latino/Hispanic, 2% were Asian and 1% were American Indians. An additional 7.8% were
categorized as other. Nine of the 87 studies enrolled exclusively minority samples. More
than half (59.8%) of the studies did not report an intention, approach or specific site/location
to recruit a sample that was racially or ethnically diverse. Of the 54 studies that included more
than one racial/ethnic group, 8 included sub-group analyses of weight loss outcomes by
race/ethnicity.

Conclusions: Lack of adequate representation of racial and ethnic minority populations in

behavioral trials limits the generalizability and potential public health impact of these
interventions. Given persistent racial/ethnic disparities in obesity in the U.S., the high
morbidity, mortality, and economic costs associated with obesity and obesity-related
conditions among racial/ethnic minority groups, findings from this review emphasize the need
to maximize representation of some underrepresented racial/ethnic groups in behavioral
lifestyle weight loss trials.
S1256 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Christina Haughton, MPH, University of Massachusetts Medical

School, Worcester, MA, 01655; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1257

B206 6:00 PM-7:00 PM

REWARD-RELATED EATING IN OBESE AND LEAN ADULTS: A MULTIMETHOD

CHARACTERIZATION

Angela Wackerly, BFA1, Kirstin Aschabacher, PhD1, Kinnari (Nina) Jhaveri, BA, MBA2, Carlos
Almeida, BS1, Alison R. Hartman, BA3, Samantha Schleicher, BA1, Cindy W Leung, ScD MPH1,
Suneil Koliwad, MD PhD4, Elissa S. Epel, PhD3, Ashley E. Mason, PhD1

1
University of California San Francisco, San Francisco, CA; 2University of Southern California,
San Francisco, CA; 3University of California, San Francisco, San Francisco, CA; 4University of
California San Francisco - School of Medicine, San Francisco, CA

Reward-Related Eating in Obese and Lean Adults: A Multimethod Characterization

Background: An obesity phenotype characterized by high reward-related eating (RRE) may be

optimally assessed by combining self-report, neurocognitive, and biological measures.
Consciously experienced symptoms of enhanced drive to eat can be assessed using the
Reward-based Eating Drive (RED) scale. Less conscious dimensions of RRE may be
characterized using neurocognitive assessments of selective attention to food cues.
Moreover, selective attention to food cues may differ in a fasted state versus after glucose
ingestion, because glucose (a primary source of fuel for the brain) can facilitate cognitive
function.

Hypotheses: Compared to lean controls, participants with obesity will exhibit higher self-
reported symptoms of RED (H1), and greater glucose-induced enhancement of selective
attention to food cues (H2). Across all participants, greater blood glucose increases will
correlate positively with glucose facilitation of selective attention (H3).

Methods: Participants with obesity (n=11) and lean controls (n=12) completed the RED scale
before completing the N-Back neurocognitive assessment. The N-Back employed food and
non-food stimuli, and participants completed it both in a fasted state, and 50 minutes after
drinking 75g of glucose (Glutol). We assayed blood glucose before and 50 minutes after Glutol
consumption. We calculated selective attention to food cues as the difference between food
and non-food stimuli in 2-back correct identifications on the N-Back. We defined glucose
facilitation of selective attention as the increase in correct food relative to non-food
identifications from pre- to post-glucose ingestion.

Results: Participants with obesity had significantly higher RED scores (H1) and greater glucose
facilitation of selective attention to food cues compared to lean controls (H2; all ps < .05).
S1258 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Greater increases in blood glucose correlated with greater increases in selective attention to
food cues in the whole sample (H3; r=.51, p=.01).

Conclusions: People with obesity are more likely to have higher RED scores and greater
attentional biases toward food, which characterize reward-related eating (RRE). Used in
concert, self-report, neurocognitive, and blood glucose measures may comprise a brief
battery that can identify individuals with high RRE. The RRE phenotype may contribute to a
vicious cycle of overeating and resultant obesity.

CORRESPONDING AUTHOR: Angela Wackerly, BFA, University of California San Francisco, San
Francisco, CA, 94118; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1259

B207 6:00 PM-7:00 PM

STRESS IS ASSOCIATED WITH WEIGHT CIRCUMFERENCE IN HISPANIC ADULTS

Mary Kate Clennan, B.S.1, Erin N. Etzel, Ph.D.2, Judith R. McCalla, PH.D.3, Patrice G. Saab,
Ph.D.1

1
University of Miami, Coral Gables, FL; 2Presbyterian Medical Group, Albuquerque, NM;
3
University of Miami, Department of Psychology, Coral Gables, FL

Introduction:
Waist circumference is a commonly used indicator of abdominal obesity. Individuals with
abdominal obesity are at a greater risk for metabolic syndrome, cardiovascular disease and
diabetes. Similarly, evidence also shows that greater perceived stress is related to poorer
health outcomes. Moreover, perceived stress and waist circumference have been shown to be
related as individuals often engage in unhealthy weight behaviors in response to stress.
Minority groups may be particularly vulnerable because they are at an increased risk for
abdominal obesity and higher stress levels. As there is little research on the link between
perceived stress and waist circumference in minorities, the current study investigated this
relationship in a sample of Hispanic adults controlling for health behaviors.

Methods:
The Heart Smart exhibition at the Miami Science Museum educated museum visitors about
the importance of a heart healthy lifestyle. The exhibition invited visitors to take simple
personal measurements (e.g., waist size and self-reported habits) using the exhibitions
interactives. Visitors received personalized feedback, and, after consenting, could
anonymously contribute their information to a research database. The exhibition reached a
large sample of Hispanic adults in South Florida. Data from 12,272 consenting Hispanic
visitors (18-90 years old; 59.6 % female) were analyzed in the cross-sectional analysis.

Results:
A multiple regression analysis showed that gender, age, perceived stress, fruit consumption,
vegetable consumption, soda consumption, and physical activity accounted for significant
variance in waist circumference (M = 32.85, SD = 6.94), adjusted R2 = .133, F (2, 12,264) =
270.69, p < .001. Results indicate that, controlling for covariates, for every one-point increase
in stress level, waist circumference increased by .17 inches 95% CI [.05, .28], p = .005.
Interestingly, squared semi-partial correlations indicate that perceived stress explained just as
much variance in waist circumference as health behaviors.
S1260 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Discussion:
The current studys findings show that higher stress levels are related to greater waist
circumference in Hispanic adults, controlling for age, gender, and health behaviors. This is
consistent with the view that stress hormones (i.e., cortisol) contribute to the accumulation of
abdominal fat. These findings have important clinical implications. Stress reduction
techniques are often briefly introduced in most evidence-based weight loss protocols
designed to improve nutrition and increase activity. The present findings underscore the value
of devoting additional attention to stress management strategies when implementing
behavioral weight loss interventions.

CORRESPONDING AUTHOR: Mary Kate Clennan, B.S., University of Miami, Coral Gables, FL,
33146; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1261

B208 6:00 PM-7:00 PM

THE IMPACT OF WEIGHT ON PSYCHOLOGICAL FUNCTIONING: THE ROLE OF WEIGHT

DISCRIMINATION AND INTERNALIZED WEIGHT BIAS

Krystal Waldo, Ph.D.1, MacKenzie R. Peltier, MA2, Melanie Roys, MS3, Aaron F. Waters, B.S.4,
Renee Laurent, B.S.5, Robert Troups, B.S.6, Amy Copeland, PhD3

1
Pittsburgh VA Healthcare System, Pittsburgh, PA; 2Louisiana State University, New Haven, CT;
3
Louisiana State University, Baton Rouge, LA; 4Louisiana State University, San Anselmo, CA;
5
Louisiana State University, Destrehan, LA; 6LSU, Baton Rouge, LA

Obesity rates in the United States (U.S.) continue to rise with approximately 34.9% of adults
considered obese and an additional 34% that are overweight. In addition to several health
problems, obese individuals report greater psychological distress and are at a higher risk for
experiencing negative psychological effects. With the increased prevalence of obesity, there
has also been a rise in weight-based discrimination and weight bias in the U.S. The present
study examined if weight bias internalization mediated the relationship between weight-
based stigmatizing experiences and psychological functioning. Non-treatment seeking
overweight and obese participants (N = 112) completed several questionnaires including the
Weight Bias Internalization Scale (WBIS), Stigmatizing Situations Inventory (SSI), and Brief
Symptom Inventory (BSI). Participants height and weight was also measured to determine
body mass index (BMI). As predicted, weight-based stigmatizing experiences was found to be
a significant predictor of internalized weight bias, and weight-based stigmatizing experiences
was a significant predictor of psychological functioning. Results indicated that the direct effect
of weight-based stigmatizing experiences on psychological functioning was significant,
indicating that weight-based stigmatizing experiences also affects psychological functioning in
ways independent of internalization. Consistent with expectations, the indirect effect of
weight-based stigmatizing experiences on psychological functioning was significantly greater
than zero, indicating that individuals who have weight-based stigmatizing experiences were,
on average, 0.040 units higher in their likelihood of experiencing psychological distress as a
result of the effect of internalized weight bias. These results support the predicted hypothesis
that weight bias internalization would mediate the relationship between weight-based
stigmatizing experiences and psychological functioning. Results provide evidence for
significant relationships between internalized weight bias and weight-based stigmatizing
experiences and psychological functioning, as well as support the predicted hypothesis that
weight bias internalization mediates the relationship between weight-based stigmatizing
experiences and psychological functioning. Recommendations for future research based on
these findings will be discussed.
S1262 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Krystal Waldo, Ph.D., Pittsburgh VA Healthcare System,

Pittsburgh, PA, 15232; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1263

B209 6:00 PM-7:00 PM

THE USE OF THE LIFESTYLE QUESTIONNAIRE FOR WEIGHT MANAGEMENT TO IMPROVE

WEIGHT MANAGEMENT COUNSELING.

Christopher Anderson, PhD1, Ryan Roberts, Undergraduate Student2

1
Utah Valley University, orem, UT; 2Utah Valley University, Alpine, UT

Abstract

Background: Productive counseling sessions for weight management in clinical settings can be
challenging due to time-restraints and health-care providers lack of comfort with counseling.
Patients often feel frustrated due to a history of unsuccessful weight loss attempts which
leads to hopelessness in future attempts. An instrument is needed to facilitate short and
effective counseling sessions and to provide specific feedback to increase patients likelihood
of success. The Lifestyle QuestionnaireWeight Management (LQ-WM) is a brief instrument
(27 items) regarding behavioral and emotional patterns and also estimates patients likelihood
of success quantitatively based upon responses by research participants. The instrument was
recently piloted and showed concurrent validity in predicting weight loss success. The current
investigation expanded the sample size, adjusted several items, and re-validated its ability to
concurrently predict weight loss success. Method: 529 participants engaged in a weight loss
attempt completed an online questionnaire covering their demographics, behavioral patterns,
and weight management outcomes. Participants were recruited from college classes and a
popular social networking site. Results: A Lifestyle Score was computed for each participant
based upon their responses to healthy and unhealthy behavioral patterns. The average
Lifestyle Scores between participants reporting successful weight management (n=300) versus
unsuccessful weight management (n=228) were significantly different (m=28.6 & m=15.4, p <
.001). Participants whose lifestyle score was in the top 10% had a significantly higher rate of
reporting weight loss success versus those in the bottom 10% (83.8% & 35.7%, respectively. p
< .001). Discussion and Conclusion: The Lifestyle Score from the LQ-WM showed concurrent
validity in predicting weight loss success but with a larger and more diverse sample than the
pilot study. The LQ-WM may provide utility in clinical scenarios to facilitate brief, yet
effective, counseling interactions which provide the patient with specific feedback on their
behavioral patterns and their likelihood of success. However, future research should
investigate the performance of the LQ-WM longitudinally, use clinical samples, and employ
more objective weight management measures besides self-report. The LQ-WM has the
potential to facilitate more effective counseling interactions between providers and patients.
S1264 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Christopher Anderson, PhD, Utah Valley University, orem, UT,
84058; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1265

B210 6:00 PM-7:00 PM

WEIGHT STIGMA AND WEIGHT CYCLING IN YOUNG ADULTS WITH OVERWEIGHT/OBESITY

Samantha N. Martinez, M.A. , Gabriela Bolivar, M.A. , Sylvia Herbozo, Ph.D., Serena D.
Stevens, MA

Loma Linda University, Loma Linda, CA

The negative effects of weight stigma on psychological and physical health in individuals with
overweight/obesity are well documented. Less is known about weight stigma and weight
cycling, especially among young adults who report high levels of weight stigma. The current
study examined the influence of weight stigma on weight cycling in an ethnically diverse
sample of young adult women with overweight/obesity. Participants were 91 young adult
women (41.7% ethnic minority) aged 18 to 30 (M = 20, SD = 5.62), with an average body mass
of 29.03 (SD = 3.80). Within the sample, 46.4% of participants indicated weight cycling at least
one time. A hierarchical multiple regression analysis was used to determine the influence of
weight stigma on weight cycling, controlling for age and ethnicity. The regression model
accounted for a significant proportion of the variance in weight cycling frequency, R2 = .13,
F(7, 83) = 1.81, p < .05. Ethnicity was a significant independent predictor of frequency of
weight cycling for Asian Americans, such that on average, Asian Americans reported 0.69
fewer incidence of weight cycling than Caucasians, p < .05. Weight stigma significantly
predicted weight cycling, such that as the frequency of weight stigma increased by one-unit,
there was a 0.17 increase in frequency of weight cycling, p < .05. These results suggest that
young adult women with overweight/obesity who experience higher rates of weight stigma
have an increased incidence of weight cycling. In addition, Asian Americans appear to be less
susceptible to weight cycling compared to Caucasians. Future studies should examine weight
stigma and factors associated with weight cycling to further inform interventions for obesity.

CORRESPONDING AUTHOR: Samantha N. Martinez, M.A. , Loma Linda University, Loma Linda,
CA, 92354; [email protected]
S1266 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B211 6:00 PM-7:00 PM

A FIRST LOOK AT MILD TRAUMATIC BRAIN INJURY (MTBI) SYMPTOM FREQUENCY AMONG
U.S. FIREFIGHTERS

Victoria Torres, B.A.1, Michelle Pennington, B.A.2, Marc Kruse, Ph.D.3, Nathan Kimbrel, Ph.D.4,
Sara Dolan, PhD5, Suzy B. Gulliver, Ph.D.2

1
University of Mississippi, Oxford, MS; 2Baylor Scott and White Health, Waco, TX; 3Austin Fire
Department, Austin, TX; 4Durham VAMC, Waco, TX; 5Baylor University, Waco, TX

Rushing into burning buildings and performing daring acts of bravery, fighting fire is a
dangerous profession. Yet, over 1 million firefighters serve communities across the United
States. The National Fire Protection Agency (NFPA) estimates that 65,880 U.S. firefighters
were injured in the line of duty in 2013 alone. Often described as a silent epidemic,
traumatic brain injury (TBI) has been observed in military, athletic and pediatric populations.
However, gathering epidemiological data has been challenging due to debate on an adequate
definition, differences in study methods and comorbidities. A recent statewide population-
based survey by Whiteneck and colleagues (2016) revealed that out of 2701 adult Colorado
residents, 37.7% reported a head injury but no TBI, 36.4% reported mild traumatic brain injury
(mTBI) and 6.0% reported moderate-severe TBI. Further, an estimated 5.3 million Americans
are currently living with a TBI-related disability (Langlois, Rutland-Brown, & Wald, 2006).
However, the frequency of TBI among firefighters is completely unknown. Our anonymous
Qualtrics survey of career firefighters (N = 60) reveals a first look at firefighter mild traumatic
brain injury (mTBI) frequency. Due to the risky nature of their profession and their likely
former athletic and military participation, we hypothesized that prevalence of post-concussive
syndrome (PCS) symptoms and mTBI are inflated among firefighters. Basic descriptive
statistics were analyzed using SPSS statistical software package. In conjunction with fire
service, the majority of participants were Caucasian males between the ages of 27 and 67
(90.0% male; 80.0% Caucasian; M = 41.64; SD = 9.36). Though much debate has centered on
an adequate definition of traumatic brain injury, we classified mTBI symptoms using the
World Health Organizations (WHO) preliminary screening criteria: a) loss of consciousness
(LOC) for 30 minutes or less, b) posttraumatic amnesia for less than 24 hours and c) confusion
or disorientation after sustaining an injury to the head resulting from external physical forces.
Our initial review of lifetime head injury events revealed that 36.8% endorsed LOC for 30
minutes or less, 44.1% endorsed loss of memory, and 57.4% endorsed feeling dazed and
confused after sustaining an injury to the head. Altogether 61.8% met lifetime criteria for one
of the aforementioned mTBI symptoms. 25.0% of participants endorsed currently
experiencing post-concussive symptoms that they thought might be related to injuries
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1267

experienced either in their career as a firefighter, athletic or military participation. Though no

cure for TBI is available, useful treatments for TBI-related symptoms are readily
available.Results from this study suggest further research in this area is warranted to inform
development and improvement of firefighter policies and procedures.

CORRESPONDING AUTHOR: Victoria Torres, B.A., University of Mississippi, Oxford, MS, 38655;
[email protected]
S1268 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B212 6:00 PM-7:00 PM

SUN PROTECTION BEHAVIORS AMONG MALE HISPANIC OUTDOOR LABORERS

Adrienne Viola, MPH1, Jerod Stapleton, Ph.D.2, Michele Oschner, Ph.D.3, Germania
Hernandez, N/A4, Louis Kimmel, BA4, Daniel Giovenco, Ph.D., MPH5, Shawna Hudson, Ph.D.6,
Denalee O'Malley, MD, MPH7, Carolina I. Lozada, MPH8, Elliot Coups, PhD9

1
Rutgers Robert Wood Johnson Medical School, Lawrenceville, NJ; 2Rutgers Cancer Institute of
New Jersey, East Brunswick, NJ; 3Rutgers School of Management and Labor Relations, New
Brunswick, NJ; 4New Labor, New Brunswick, NJ; 5Columbia University- Mailman School of
Public Health, New York, NY; 6Department of Family Medicine, Research Division; Director of
Community Research, New Brunswick, NJ; 7Department of Family Medicine and Community
Health, Research Division Rutgers Robert Wood Johnson Medical School, New Brunswick, NJ;
8
Rutgers Cancer Institute of New Jersey, New Brunswick, NJ; 9Rutgers, The State University of
New Jersey, New Brunswick, NJ

Introduction: Individuals who work outside are at increased risk for skin cancer due to
exposure to the suns ultraviolet (UV) radiation. Hispanic individuals in the United States are
over-represented in several outdoor occupations that have high UV exposure. Little is known
about the sun protection practices of Hispanic individuals who work outdoors. In this project,
we examined Hispanic day laborers sun protection knowledge, beliefs, behaviors, and
barriers.

Methods: 175 Hispanic men (47% Mexican, 26% Honduran; M age 35.1) were recruited at
local outdoor locations where day laborers commonly congregate in New Brunswick, NJ.
Participants were eligible if they were male, self-identified as Hispanic or Latino, and reported
working outside in the past 3 months or the previous summer. Participants completed a 15-20
minute pencil and paper or interviewer administered survey in Spanish or English. Questions
focused on demographic factors as well as sun protection knowledge, beliefs, behaviors, and
barriers.

Results: Overall, knowledge of skin protection and skin cancer was modest (M = 3.9 correct
items out of 7, SD = 1.11). Participants were most frequently incorrect (97%) about the
recommended sun protection factor (SPF) of sunscreen. More than 90% of participants
somewhat/strongly agreed that it was important to protect my skin from the sun while
working outdoors (91%) and if I dont protect myself from the sun while working outdoors,
my skin will be damaged (93%). Only 11% of the sample felt that theres not much you can
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1269

do to lower your chances of skin cancer. The reported rate of often or always engaging in sun
protection behaviors when working outside was as follows: sunscreen, 17%; staying in the
shade, 10%; wearing a wide-brimmed hat, 13%; wearing a long-sleeved shirt, 44%; wearing
long pants, 89%. The top barriers to wearing sunscreen were forgetting to apply it and not
knowing what kind to use. The top barriers for not wearing long pants and a long-sleeved shirt
as well as wearing a wide brimmed hat were that they were too hot and uncomfortable to
wear.

Conclusions: Hispanic day laborers do not sufficiently engage in many sun protection
behaviors while working outdoors, despite believing in the importance of skin protection.
Participants reported several factors that hinder their ability to engage in sun protection
behaviors during outdoor work. Future research is needed to facilitate engagement in sun
protection behaviors in this at-risk population.

CORRESPONDING AUTHOR: Adrienne Viola, MPH, Rutgers Robert Wood Johnson Medical
School, Lawrenceville, NJ, 08648; [email protected]
S1270 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B213 6:00 PM-7:00 PM

THE PROVISION OF CLERGY HEALTH RESOURCES BY FAITH-BASED ORGANIZATIONS IN THE

UNITED STATES

Benjamin Webb, PhD, CSCS1, Scherezade Mama, DrPH2, Jonathon Kirby, BS1

1
Southern Illinois University Edwardsville, Edwardsville, IL; 2The Pennsylvania State University,
University Park, PA

The percentage of clergy classified as obese is reported to be as high as 41%. Additionally,

clergy are disproportionately affected by several chronic diseases. Provision of health
resources should be a priority for faith-based organizations (FBOs) that employ clergy, but
little is known about the extent to which clergy health is prioritized and addressed among
FBOs. The purpose of this study was to describe practices related to promoting clergy health
among a national sample of FBOs. A total of 154 administrators from the largest religious
denominations in the U.S. completed questionnaires on their beliefs and practices regarding
clergy health promotion. Stress (64.7%), obesity (45%), and lack of exercise (41.2%) were
identified as the top health-related issues facing clergy. The most commonly offered health-
related resources were health insurance (86.7%), an employee wellness benefit provided by
health insurance provider (61.3%), a health-related newsletter (61.3%) and a website with
general health information (61.3%). Lack of financial resources (61.3%), lack of trained staff
(46.7%) and lack of interest from clergy (47.4%) were cited as the most common barriers to
providing health resources for clergy. Respondents indicated that their FBOs encourage clergy
to lead healthy lives (91.2%), be physically active (65.4%), eat healthy (62.3%), and avoid
smoking (66.9%). Although a majority of respondents thought that clergy must adhere to a
healthy lifestyle to effectively encourage their congregation to do the same (88.5%), very few
(29.4%) respondents thought clergy were prepared to be good role models for health to their
parishioners. Additionally, few respondents (37.3%) thought their clergy were ready to face
the challenges of health-related self-care while in ministry, with 78.0% indicating that clergy
need more training related to health and prevention of chronic disease. Respondents
indicated that the most appropriate ways for clergy to receive health education were through
online training (41.9%) and special events (30.1%). The results of this study highlight potential
priorities (e.g., stress, obesity, exercise, and education) and modes of delivery (e.g., online
training) for FBOs interested in providing or enhancing resources aimed at promoting the
health of clergy. Further research is needed to determine whether the provision of health-
related resources is associated with more positive health and behaviors among professional
clergy.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1271

CORRESPONDING AUTHOR: Benjamin Webb, PhD, CSCS, Southern Illinois University

Edwardsville, Edwardsville, IL, 62026; [email protected]
S1272 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B214 6:00 PM-7:00 PM

A SYSTEMATIC REVIEW OF SOCIAL SUPPORT AND PAIN IN PATIENTS WITH RHEUMATOLOGIC

DISEASE

Emily Rose San Diego, BA1, Karina R. Aguilar, BA2, Rina S. Fox, PhD, MPH3, Erin L. Merz, PhD,
MPH4

1
California State University, Dominguez Hills, Costa Mesa, CA; 2California State University,
Dominguez Hills, Wilmington, CA; 3Northwestern University Feinberg School of Medicine,
Chicago, IL; 4California State University, Dominguez Hills, Carson, CA

Background: Pain is a significant quality of life concern in patients with rheumatologic disease.
A biopsychosocial conceptualization of pain is widely accepted, with perceived social support
being recognized as a potentially important factor in understanding pain in this population.
The aim of this study was to summarize the empirical evidence of the link between social
support and pain in rheumatology patients.

Methods: PRISMA guidelines were used to conduct the systematic review. Keyword searches
of pre-defined terms of the PsycINFO and PubMed/MEDLINE databases were conducted. A
standardized tool was used to extract and record study-level data. Social support was
classified into four taxonomies (overall, emotional, instrumental/tangible, informational).
Weighted correlations were aggregated using random effects models across each social
support type.

Results: Of the 411 screened articles, 27 studies met the inclusion/exclusion criteria (N =
4,069). The majority was comprised of patients with rheumatoid arthritis, although
osteoarthritis, lupus, and fibromyalgia were also represented. The pooled effect sizes from
cross-sectional analyses suggested a small inverse association with pain for overall social
support (r = -.15, 95% CI: -.09, -.21), emotional social support (r = -.18, 95% CI: -.09, -.18), and
instrumental/tangible support (r = -.14, 95% CI: -.08, -.19). There were insufficient studies to
evaluate the relationship between pain and informational social support, or any longitudinal
relationships. Studies were most limited by substantial heterogeneity in the measurement of
social support and pain, ambiguities with regard to how measures were used and reported,
cross-sectional designs, and representativeness of the samples.

Conclusions: The results suggest that rheumatology patients with greater social support
report lower pain, although the effect is small. These findings are consistent with research on
the health benefits of social support in numerous populations.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1273

CORRESPONDING AUTHOR: Emily Rose San Diego, BA, California State University, Dominguez
Hills, Costa Mesa, CA, 92627; [email protected]
S1274 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B215 6:00 PM-7:00 PM

PAIN EXPERIENCES AND PAIN MEDICATION USE AFTER SPINAL CORD INJURY: RELATIONSHIP
WITH RISK OF MORTALITY

Jillian Clark, M.A.1, James S. Krause, PhD2, Yue Cao, PhD2

1
University of Missouri-Kansas City, San Diego, CA; 2Medical University of South Carolina,
Charleston, SC

Spinal cord injury (SCI) has long been associated with diminished longevity, particularly among
those with more severe SCI. Lower income, less education, tobacco use, binge drinking,
prescription medication use, physical health conditions (e.g., heart disease), and secondary
conditions (e.g., pressure ulcer surgeries, depressive symptoms) have been associated with
increased risk of mortality in SCI. Pain is a common secondary condition after spinal cord
injury; however, there is no prior literature that examines pain and pain-related factors in
relation to mortality in SCI. The current study examined the roles of pain intensity, pain
interference, and pain medication use with risk of mortality, while controlling for biographic,
injury, socioeconomic, and health factors.

Participants (n = 2535) with spinal cord injury, identified through records at a specialty
hospital in the southeastern United States, completed self-report questionnaires. Predictor
variables assessed included: patient characteristics (e.g., demographics and injury), health
indicators (e.g., pressure sores, hospitalizations, broken bones, amputations, depressive
symptoms, pain intensity, and interference of pain on daily functioning), and frequency of
pain medication use. The outcome variable, mortality, was identified using the National Death
Index and the Social Security Death Index.

There were 335 deceased participants who had an average of approximately 38 survival
months, compared to approximately 73 months for those alive. The final stage of a three-
stage hierarchical multivariate model (i.e., stage 1 = patient characteristics, stage 2 = health
indicators, and stage 3 = pain medication use) using Cox proportional hazards modeling
indicated that being male (HR = 1.42), and having greater injury severity (HR = 3.06) were risk
factors for mortality, whereas, having a four-year college degree or higher was a protective
factor (HR = .64). With the exception of pain interference, pain intensity, and having a broken
bone, all other health indicators were predictive of increased risk of mortality. Those who
reported daily use of pain medications were 61% more likely to be deceased compared to
those who dont use pain medication (HR=1.61).
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1275

Frequency of prescription pain medication use was significantly related to risk of mortality,
even after controlling for all other predictors. It is important for providers to be cognizant of
the potential contributory effects of pain medication use on mortality in individuals with SCI.

CORRESPONDING AUTHOR: Jillian Clark, M.A., University of Missouri-Kansas City, San Diego,
CA, 92122; [email protected]
S1276 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B216 6:00 PM-7:00 PM

THE EFFECTS OF AN ANTI-INFLAMMATORY DIET IN THE REDUCTION OF PAIN SUBJECTIVITY

AMONG VETERANS WITH CHRONIC PAIN

Jay T. Wiles, M.A. in Clinical Psychology1, Melody Mickens, Ph.D.2, Ka Wong, Registered
Dietician2, Meghan Rooney, Psy.D.2

1
Regent University, Richmond, VA; 2Hunter Holmes McGuire VA Medical Center, Richmond,
VA

Background:
Research has suggested a direct link between poor diet, systemic inflammation, and increased
reports of subjective pain (Minihane et al., 2015). Newer approaches to pain management
have incorporated behavioral treatments and a focus on managing biopsychosocial aspects of
pain. Participants included in this study were veterans at VA medical Center who were actively
involved in a multimodal, self-management integrative pain clinic. One aspect of this
integrative clinic included following an anti-inflammatory diet, which targets the elimination
of pro-inflammatory foods (i.e., pain inducing) and introduction of anti-inflammatory foods
(i.e., pain reducing).

Methods: Participants (N=78) were predominately African-American (58%)male (70%)

Veterans with a mean age of 56 years (SD = 10.66) who completed an 8-week anti-
inflammatory diet group led by a registered dietician and licensed psychologist. One third of
participants (N = 30) had completed data for pre-and post measures. Veterans completed the
Pain Catastrophizing Scale (PCS) and Pain Outcomes Questionnaire (POQ) at intake and during
the final group. Paired T-tests were conducted to assess differences between pre and post
measures. Outcomes included total PCS scores, PCS subscale scores (Rumination,
Magnification, and Helplessness), POQ total score, and POQ subscales (Pain, Mobility,
Activities of Daily Living, Vitality, Negative Affect, and Fear).

Findings:When comparing pre-and post-group data, participants who completed the anti-
inflammatory group had lower scores on the following subscales: PCS Helplessness, t(29) =
2.08, p < .05, POQ Mobility, t(29) = 3.08, p.01, POQ Vitality, t(29) = 2.14, pp < .05. No
significant racial or gender differences were identified on outcome variables.

Conclusions: Results of the study suggest that multiple areas of pain subjectivity were
significantly influenced after an 8-week anti-inflammatory diet group. While most veteran
participants were involved in multiple treatment modalities for chronic pain management
during the time of the group, this research suggests that diet modification, specifically the
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1277

incorporation of an anti-inflammatory diet, may play a significant role in pain reduction

amongst veterans with chronic pain. Use of an anti-inflammatory diet may prove as an
effective adjunctive treatment to pharmacological and other non-pharmacological
management of pain, as well as help to improve overall quality of life and well-being.

CORRESPONDING AUTHOR: Jay T. Wiles, M.A. in Clinical Psychology, Regent University,

Richmond, VA, 23233; [email protected]
S1278 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B217 6:00 PM-7:00 PM

YOGA TREATMENT FOR CHRONIC LOWER BACK PAIN IN THE MILITARY: WHO BENEFITS THE
MOST?

Krista Highland, PhD1, Audrey Schoomaker, BSN2, Winifred Rojas, BSN3, Joshua P. Suen, N/A4,
Ambareen Ahmed, MS5, Zhiwei Zhang, BA6, Sarah Fink Carlin, MPH7, MIchael Kent, MD1,
Christian Calilung, BS3, Chester Buckenmaier III, MD3

1
Defense and Veterans Center for Integrative Pain Management, Rockville, MD; 2Uniformed
Services of the Health Sciences, Rockville, MD; 3Henry M. Jackson Foundation, Rockville, MD;
4
Henry M Jackson Foundation, Rockville, MD; 5Henry M. Jackson Foundation, Kansas City, MO;
6
Henry M. Jackson Foundaion, Rockville, MD; 7YogaMedics, Huntington Woods, MI

Background: To optimize treatment selection for patients with chronic pain, data must show
why and for whom particular modalities are most effective. The present study examines the
role of pre-treatment psychosocial characteristics on the effects of an 8-week yoga program
for chronic lower back pain (cLBP) in a military health setting.

Methods: Participants (N=68) were randomized to a control group or the Restorative Exercise
and Strength Training for Operational Resilience and Excellence (RESTORE) yoga group. The
Patient Reported Outcomes Measurement Information System (PROMIS) was administered at
baseline (BL), mid-treatment, and post-treatment (PT). Participants were stratified into risk
groups (e.g. 1 SD from the norm on PROMIS scales) for Sleep Disturbances, Social Functioning,
and Pain Interference. The independent and interactive effects of time, treatment, and risk
groups on Physical Functioning were examined with generalized estimating equations.
Sequential Bonferroni post-hoc tests compared group differences in BL to PT Physical
Functioning changes for each treatment x risk group.

Results: From BL to PT, the control group retained BL Physical Functioning levels, across all
risk groups. Participants in the RESTORE group with above average BL Sleep Disturbances and
Pain Interference scores reported significant increases in Physical Functioning, while those in
the RESTORE group with average scores did not change. Participants in the RESTORE group
with below average Social Functioning retained BL Physical Functioning levels, while those
with average Social Functioning reported increases from BL to PT.

Conclusions: For military service members with cLBP, yoga may be most helpful for individuals
with elevated sleep difficulties and pain interference, but adequate social support. Given the
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1279

high prevalence of sleep disorders among service members, coupled with the need to combat
the opioid epidemic with integrative pain management strategies, yoga is a promising avenue
of treatment for cLBP. Future research is needed to determine whether preparatory
interventions targeting social support can bolster functioning improvements in yoga-based
therapies, as well as compare treatment moderators across other integrative pain treatments.
By identifying and leveraging research regarding treatment moderators, providers can match
the best-fitting treatments with each patient, increasing the likelihood of treatment success.

Disclaimer: The views expressed in this presentation are those of the authors and do not
reflect the official policy of the Department of Army/Navy/Air Force, Department of Defense,
or U.S. Government.

CORRESPONDING AUTHOR: Krista Highland, PhD, Defense and Veterans Center for Integrative
Pain Management, Rockville, MD, 20852; [email protected]
S1280 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B218 6:00 PM-7:00 PM

A SYSTEMATIC REVIEW OF ADHERENCE TO RCTS INVOLVING PHYSICAL ACTIVITY

INTERVENTIONS TARGETING THREE CHRONIC DISEASES

Sean P. Mullen, PhD, Tiffany Bullard, B.S.

University of Illinois at Urbana-Champaign, Urbana, IL

The medical community has accepted physical activity as an effective behavioral medicine for
the prevention and treatment of a variety of diseases and chronic conditions. Chronic
conditions may vary in the extent to which they impact self-regulatory skills necessary for
initiating and maintaining a physical activity program. However, insufficient evidence is
available to make comparisons across chronic conditions regarding adherence to structured
physical activity interventions overall, or across different settings. To explore potential
differences, we conducted a systematic review and quantitative summary of randomized
controlled trials using physical activity or exercise as a treatment for one of three chronic
diseases. A total of 33 peer-reviewed scientific articles published between 2000 and 2016 (via
PsychInfo, PubMed, CINAHL, Clinical Key, SCOPUS, and hand-picked articles) met our inclusion
criteria (i.e., randomized controlled trial, at least one trial arm consisted primarily of aerobic
and/or strength training, lasted three or more months, reported % of sessions or made data
available). We targeted three clinical populations, i.e., cardiovascular disease [CVD] (n=10),
diabetes (n=10), and cancer (n=13), and data from all studies were included in our analyses.
Authors from 54.5% of identified studies confirmed data. Adherence rates varied considerably
across samples: 85.03% (SD=13.77%), 77.76% (SD=14.56%), and 66.94% (SD=24.52%),
respectively. Dropout rates showed a similar pattern (10.60%, 12.15%, and 12.42%,
respectively). Interestingly, overall, adherence for center-based trials (75.00%[SD=21.77%])
did not differ from home-based trials (77.42%[SD=17.89%]). Although few differences
emerged, substantial variability exists across studies and within each of the targeted
conditions. Moreover, these findings imply that the average patient is not receiving 25% of
their prescribed physical activity treatment. These findings underscore the need for novel
approaches to maximize the reach and treatment uptake of structured physical activity
interventions. It is important to note that many of these interventions preceded the "tech
boom" and it is possible that mobile health technologies could be incorporated in an effort to
increase participation motivation. Additionally, screening patients self-regulatory capacity at
the onset of intervention may facilitate touch points for mini-interventions designed to
enhance their trial experience and optimize treatment adherence.

CORRESPONDING AUTHOR: Sean P. Mullen, PhD, University of Illinois at Urbana-Champaign,

Urbana, IL, 61822; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1281

B219 6:00 PM-7:00 PM

ACCULTURATION, PHYSICAL ACTIVITY AND SEDENTARY: ASSOCIATION OF ANTHROPOMETRIC

VALUES IN A BINATIONAL MEXICAN POPULATION.

Katia Gallegos Carrillo, PhD MSc1, Lourdes Baezconde-Garbanati, PhD, MPH, MA2, britni R.
belcher, PhD, MPH3, Genevieve F. Dunton, PhD, MPH3, Jonathan Samet, MD3

1
University of Southern California, Glendale, CA; 2University of Southern California,
Department of Preventive Medicine, Institute for Health Promotion and Disease Prevention
Research, Los Angeles, CA; 3University of Southern California, Los Angeles, CA

Introduction: Increases in acculturation influence several health behaviors and can be a

challengeamong individuals of Mexican origin. This pattern has been shown in large
population studies based on people living in Mexico when compared to those living in the U.S.
The current study examined the influence of acculturation on the association between
physical activity, sedentary behaviors, and anthropometric measures of Mexicans and
Mexican-Americans living in the U.S. and Mexicans living in Mexico.

Methods: We examined data from two national surveys, the National Health and Nutrition
Examination Survey (NHANES) 2011-2012 and the National Health and Nutrition Survey of
Mexico (ENSANUT) 2012. Using the self-reported moderate to vigorous physical activity
(MVPA) and sedentary behavior (SB), we calculated 4 exclusive categories: a) physically active
and low sedentary, b) physically active and high sedentary, c) physically inactive and low
sedentary, and d) physically inactive and high sedentary. Anthropometric measures were
body mass index (BMI) and waist circumference (WC). Acculturation was assessed through a
combination of ethnicity, birth place, and country of residence: Mexican born & living in
Mexico, Mexican born & living in the US and Mexican-American born & living in the US.
Multiple linear regression models were conducted independently for BMI and WC, for each of
the four exclusive categories of MVPA and SB, with acculturation as the independent variable.
The models were adjusted by sociodemographic variables and health risk conditions.

Results: Among Mexicans living in the U.S., those physically active and low sedentary had a
BMI 1.47 kg/m2 higher, compared with Mexicans living in Mexico. The association remained
significant after adjusting for sociodemographic factors and health risk conditions (p= 0.028).
Mexican-Americans living in the U.S. who were categorized as physically inactive and high
sedentary had higher BMI values 3.96 kg/m2 (p< 0.001), than their Mexican counterparts
living in Mexico or in the US, statistical significant in adjusted models. Even among those with
the lower risk physically active and low sedentary group, the Mexicans and Mexican-
S1282 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Americans living in the U.S. had higher WC values compared to Mexicans living in Mexico
(p=0.01 and p=0.008 respectively).

Conclusion: According to these results, acculturation plays a role in the relationship between
MVPA and sedentary behaviors and adiposity measures in adults, participants more
acculturated have higher values of BMI and WC, even in categories of lower risk as physically
active and less sedentary. However, more research is needed to understand the underlying
mechanisms (e.g. dietary changes, occupation type) driving these relationships.

CORRESPONDING AUTHOR: Katia Gallegos Carrillo, PhD MSc, University of Southern

California, Glendale, CA, 91204; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1283

B220 6:00 PM-7:00 PM

ACCURACY OF LED-BASED HEART RATE MONITORING ON WRIST-WORN ACTIVITY TRACKERS

Lisa Cadmus-Bertram, Ph.D.1, Emily Wirkus, MPH1, Keith Thraen-Borowski, Ph.D.2, Ronald
Gangnon, Ph.D.1, Jessica Gorzelitz-Liebhauser, MS3

1
University of Wisconsin - Madison, Madison, WI; 2Loras College, Dubuque, IA; 3University of
Wisconsin-Madison, Madison, WI

BACKGROUND: Some wearable activity trackers now use light-emitting diode (LED) sensors to
monitor HR during both rest and exercise. Unlike traditional monitors requiring a separate
chest strap, wrist-worn models are unobtrusive and are appropriate for continuous long-term
wear. By providing objective feedback on exertion, they may help individuals achieve federal
physical activity guidelines. The objective of this study was to determine the validity of the HR
monitoring function on wrist-worn activity trackers at rest and during moderate-intensity
activity.

METHODS:This study used four wrist-worn trackers (Fitbit Charge, Fitbit Surge, Mio Fuse, and
Basis Peak) with LED-based HR monitoring capability. Participants were 40 healthy adults aged
30-65 years. Participants were connected to an electrocardiogram (ECG) and the four trackers
(two per wrist; location randomized). While sitting, participants resting HR was recorded from
each of the devices at 1-min intervals for 10 min. Next, participants exercised on a treadmill
until their HR reached 65% of maximum HR. Once the targeted HR was achieved, participants
exercised for an additional 10 min, with HR measured from each device at one-minute
intervals. Analyses were performed in R.

RESULTS: Participants were 49.3 years of age (SD=9.5) with a body mass index of 25.1 kg/m 2
(SD=3.9); 50% were female. At both rest and 65% MHR, bias was minimal for all four trackers
(range: -2.5, 2.8). At rest, the Fitbit Surge was most precise [device variance: 0.3; repeatability
over 10 measurements: 1.3(LoA: -2.0,1.4)], the Fitbit Charge HR and the Mio Fuse were
somewhat less accurate, and the Basis Peak was the least accurate [device variance: 7.0;
repeatability: 6.1(LoA: -11.9,17.4)]. At 65% MHR, bias remained minimal but all devices
exhibited substantially more error than at rest. The Mio Fuse performed the best [device
variance: 7.4; repeatability: 7.5(LoA: -14.0,17.5)]. Position on the wrist (first vs. second
position nearest to hand) did not affect accuracy.

CONCLUSIONS: At rest, all trackers produced relatively unbiased measures with high precision
for some models. During moderate-intensity activity, the trackers still produced
measurements with low bias, albeit with less precision. The data suggest that while
S1284 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

commercial activity trackers are not error-free, they may be accurate enough for general
health promotion purposes, particularly if they are effective in motivating participants to
engage in more moderate-vigorous vs. light-intensity activity. These devices also have the
potential to provide individuals with a useful, unobtrusive tool that could be a feasible and
convenient method of at-home continuous HR monitoring. More data are needed to address
the accuracy of consumer-based trackers during vigorous-intensity activities.

CORRESPONDING AUTHOR: Lisa Cadmus-Bertram, Ph.D., University of Wisconsin - Madison,

Madison, WI, 53562; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1285

B222 6:00 PM-7:00 PM

BEING ACTIVE IN PREGNANCY: THEORY-BASED PREDICTORS OF PHYSICAL ACTIVITY AMONG

PREGNANT WOMEN

Kyra Hamilton, PhD1, Martin S. Hagger, PhD2

1
Griffith University and Menzies Health Institute Queensland, Brisbane, Queensland, Australia;
2
Curtin University, Perth, Western Australia, Australia

Objective: Although regular physical activity is recommended for pregnant women and has
been associated with a range of physical and psychological benefits, research indicates that
compared to pre-pregnancy, antenatal physical activity often reduces or ceases completely.
The aim of the current study was to apply an integrative theoretical model to identify the
motivational and social cognitive factors that predict physical activity participation in pregnant
women. Methods: Participants were pregnant women aged 18 years and older (N=207, Time
1; Mage = 30.03, SDage = 4.49). The study used a prospective-correlational design with a one
week follow-up. At Time 1 (T1), participants completed an initial questionnaire assessing
intrinsic motivation from self-determination theory, social cognitive constructs from theory of
planned behaviour, and self-control from self-control theory. Demographic variables were
also collected. At Time 2 (N=117, T2), participants completed a follow-up questionnaire
assessing their self-reported physical activity behaviour in the previous week. Results: A well-
fitting structural equation model, controlling for baseline age, BMI, and gestation age,
accounted for 73% and 42% of the variance in intention and physical activity behaviour,
respectively. There was a significant indirect effect of intrinsic motivation on intention
mediated by attitude and perceived behavioural control, and significant direct effects of
intention, perceived behavioural control, and self-control on physical activity behaviour.
Conclusions: Current findings fill a knowledge gap in the empirical literature on the multiple
processes that guide physical activity behaviour in pregnant women and add to the
cumulative evidence advocating the use of integrative models in explaining health behaviour.
The findings of the current study indicate important potential routes to behavioural
performance that researchers and practitioners can use to ensure the design of future
physical activity interventions for pregnant women that are efficacious in eliciting behaviour
change. Future behavioural interventions aiming to promote physical activity during
pregnancy, a period when physical activity levels typically decline, should consider the
multiple processes advocated in the integrative model as necessary for motivated action.

CORRESPONDING AUTHOR: Kyra Hamilton, PhD, Griffith University and Menzies Health
Institute Queensland, Brisbane, Queensland, 4122; [email protected]
S1286 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B223 6:00 PM-7:00 PM

CHANGES IN PSYCHOSOCIAL FACTORS WITH PARTICIPATION IN A COLLEGE PHYSICAL ACTIVITY

CLASS: AN EXAMINATION BY RACE/ETHNICITY

Melissa Bopp, PhD, Zack Bopp, MPH, Michele Duffey, MS

Pennsylvania State University, University Park, PA

Participation in physical activity declines with the transition to college but can be combated
with participation in college physical activity (PA) classes. The impact of these classes on PA-
related psychosocial factors could help with long-term maintenance of PA. Limited research
has addressed how PA classes impact different student groups, therefore the purpose of this
study was to examine how participation in a college PA class could impact PA-related
psychosocial factors by race/ethnicity. Methods: Students enrolled in a college PA class were
recruited to participate in an online survey at the beginning and end of their class. The survey
included: demographics, PA participation, PA enjoyment, self-efficacy and exercise goal
setting. Change from pre to post was calculated. T-tests examined differences in the change in
psychosocial factors by race/ethnicity. Pearson correlations examined the relationship
between a change in PA and psychosocial factors separately by race/ethnicity. Regression
analyses predicted change in vigorous PA. Results: Students (n=838) were predominately Non-
Hispanic White (NHW; n=634, 81.6%) and female (n=451, 57.7%). At baseline NHW students
were more vigorously active than ethnic minority group (EMG) students (p=.004), had higher
PA enjoyment (p < .001), PA self-efficacy (p < .001) and greater exercise goal-setting behavior
(p < .001). At follow up EMG students had a greater change in moderate (p=.02) PA, PA
enjoyment (p < .001) PA self-efficacy (p < .001) and exercise goal setting (p < .001) than NHW
students. Among NHW, a change in vigorous PA was related to a change in PA enjoyment
(r=0.16, p < .001), self-efficacy (p=0.15, p < .001) and exercise goal-setting (r=0.19, p < .001).
For EMG students, a change in vigorous PA was associated with a change in PA enjoyment
(r=0.67, p < .001), PA self-efficacy (r=0.42, p < .001) and exercise goal setting (r=0.37, p <
.001). Among NHW students, psychosocial factor changes predicted 4.5% of the change in
vigorous PA (p < .001) with PA enjoyment changes as a significant predictor (B=1.01, p=.04)
while the regression model predicted 18.8% of the change in vigorous PA for EMG students
with PA enjoyment (B=0.92, p=.05) and self-efficacy (B=1.97, p=.03) changes as significant
predictors. Conclusion: College PA classes offer an opportunity to instill lifetime skills for
active lifestyles, though may be received differently by race/ethnicity. Further research should
examine how tailoring could impact this relationship.

CORRESPONDING AUTHOR: Melissa Bopp, PhD, Pennsylvania State University, University Park,
PA, 16802; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1287

B224 6:00 PM-7:00 PM

DAILY EXERCISE PRESCRIPTIONS ARE MORE INTUITIVE THAN WEEKLY EXERCISE

PRESCRIPTIONS: IMPLICATIONS FOR GUIDELINE DEVELOPMENT

Zachary Zenko, PhD, Dan Ariely, PhD

Duke University, Durham, NC

Exercise guidelines recommended by scientific and governing bodies are typically framed as a
weekly amount. For example, the guidelines from the American College of Sports Medicine,
US Department of Health and Human Services, and National Health Service each include
recommended minimum durations of aerobic activity per week, not per day. However, recent
evidence suggests that reframing exercise requirements into daily amounts may be beneficial
(Peetz et al., 2011). Understanding the most effective framing of exercise guidelines is
important for creating exercise guidelines that will lead to superior behavioral outcomes (e.g.,
increased adherence). Purpose: This study was designed to determine if people intuitively
create daily, rather than weekly, exercise prescriptions. Methods: Participants were recruited
at the North Carolina State Fair; 381 people (57% woman, mean age: 34.37 15.05 years)
followed instructions and were included in the analysis. First, each participant described the
person they felt closest to (friend, 58.5% woman, mean age: 38.42 17.40 years, 65.1%
normal weight) and then created an exercise prescription that their friend would be (1) most
likely to adhere to, (2) most likely to keep track of, and (3) most likely to benefit from. To
create the exercise prescriptions, each participant entered the number of minutes of aerobic
exercise that the friend should do per [frequency] and then chose a frequency (day, week, or
month). The frequently chosen was hypothesized to represent the intuitive framing of the
exercise prescription (i.e., daily, weekly, or monthly). Results: Most participants
recommended an amount of exercise per day (86.9%) rather than per week (12.6%),
suggesting that daily exercise is more intuitive than weekly exercise (p < .001). This result was
replicated on the Internet in a sample of 124 participants (50.8% woman, mean age: 38.30
12.41 years), with the majority (76.6%) of participants recommending daily exercise (p < .001).
Conclusion: These data suggest that recommending minimum durations of exercise per day
may be more intuitive than recommending minimum durations of exercise per week. Future
investigators should determine if daily exercise prescriptions are (1) more memorable and (2)
result in better adherence than weekly exercise prescriptions. These data can inform the
development of future exercise guidelines that may be more intuitive and more likely to be
completed.

CORRESPONDING AUTHOR: Zachary Zenko, PhD, Duke University, Durham, NC, 27701;
[email protected]
S1288 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B225 6:00 PM-7:00 PM

DEVELOPMENT AND RELIABILITY OF A STREETSCAPE OBSERVATION INSTRUMENT FOR

INTERNATIONAL USE: MAPS-GLOBAL

Kelli L. Cain, MA1, Carrie M. Geremia, BA2, Terry L. Conway, PhD1, Lawrence D. Frank, PhD3,
James E. Chapman, MA4, Eric Fox, BA4, Anna Timperio, PhD5, Jenny Veitch, PhD5, Delfien Van
Dyck, PhD6, Hannah Verhoeven, MS6, Rodrigo Reis, PhD7, Alexandre Augusto, BA8, Ester Cerin,
PhD9, Robin Mellecker, PhD9, Ana Queralt, PhD10, Javier Molina-Garca, PhD10, James F. Sallis,
PhD11

1
University California San Diego, San Diego, CA; 2UCSD, San Diego, CA; 3University British
Columbia, Seattle, WA; 4UD4H, Seattle, WA; 5Deakin University, Burwood, Victoria, Australia;
6
Ghent University, Ghent, N/A, Belgium; 7Washington University in St. Louis, Saint Louis, MO;
8
PONTIFCIA UNIVERSIDADE CATLICA DO PARAN, Curitiba, Parana, Brazil; 9The University of
Hong Kong, Pokfulam, N/A, Hong Kong; 10Universitat de Valncia, Valencia, N/A, Spain;
11
University of California, San Diego, San Diego, CA

Objective: Relationships between several built environment factors and physical activity and
walking behavior are well established. MAPS-Global is an observational streetscape audit
instrument developed for international use with inter-observer reliability evaluated in several
countries.

Methods: MAPS-Global was developed by compiling common concepts across 8 pedestrian

and biking audit tools from numerous countries, including the original MAPS instrument which
was developed in the US. Inter-rater reliability data were collected in neighborhoods selected
to vary on GIS-derived macro-level walkability in 5 countries (Australia, Belgium, Brazil, Hong
Kong-China, and Spain). MAPS-Global assessments (n=407) were completed along a 0.25 mile
route from a residence toward the nearest non-residential destination, and a commercial
block was also rated for each residence. Two raters in each country rated each route
independently. A tiered scoring system was created that summarized items at multiple levels
of aggregation, and positive and negative valence scores were created based on the expected
affect on physical activity. Intraclass correlations (ICC's) were computed for scales using
oneway random models.

Results: Overall, 89.2% of individual items, subscales and overall summary scores showed
excellent agreement (ICC> .75); 7.2% showed good agreement (ICC=0.60-0.74); and 3.6%
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1289

showed fair agreement (ICC=0.40 to 0.59). Two items were too rare to compute ICC's. The
median ICC was .94 with a range of .54-.99. Aesthetics and social characteristics showed lower
ICC values than other constructs but the scales were still in the excellent range.

Conclusion: Evaluation of inter-observer reliability of MAPS-Global in 5 countries indicated

virtually all items and scales had "good" or "excellent" reliability. The results demonstrate that
trained observers from multiple countries were able to reliably conduct observations of both
residential and commercial areas with the new MAPS-Global instrument. This instrument can
now be used to collect comparable microscale data across countries to identify strengths and
weaknesses of activity-supportive environments, identify microscale correlates of physical
activity, and evaluate changes in built environments, especially those designed to improve
physical activity and health.

CORRESPONDING AUTHOR: Kelli L. Cain, MA, University California San Diego, San Diego, CA,
92103; [email protected]
S1290 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B226 6:00 PM-7:00 PM

DEVELOPMENTAL TRAJECTORY PATTERNS OF VIGOROUS-INTENSITY PHYSICAL ACTIVITY FROM

AGE 5 TO 19 YEARS: IOWA BONE DEVELOPMENT STUDY

Soyang Kwon, PhD1, Kathleen F. Janz, EdD2, Elena M. Letuchy, MS2, Steven M. Levy, DDS2

1
Lurie Children's Hospital of Chicago, Chicago, IL; 2University of Iowa, Iowa City, IA

Background: At a population-level, physical activity decreases during childhood and

adolescence; however, the developmental trajectory of physical activity is not homogeneous.
In addition, researchers have hypothesized that at the same absolute volume, vigorous-
intensity physical activity (VPA) may effect health outcomes differently than moderate- and
vigorous-intensity physical activity (MVPA). A group-based trajectory analytic approach
enables data to identify district trajectory patterns of an outcome of interest. This approach is
a critical step in understanding differential effects of VPA and MVPA across the developmental
years. Using the group-based trajectory approach, the authors have previously identified
distinct developmental patterns of MVPA and sport participation from age 5 to 19 years.

Objective: We aimed to identify diverse trajectory patterns of daily minutes spent in VPA from
age 5 to 19 years and examine the associations of these patterns with sex and sport
participation trajectory.

Methods: Participants were Iowa Bone Development Study (IBDS) cohort members. They
were asked to wear an ActiGraph accelerometer for four or five days at approximate ages 5, 8,
11, 13, 15, 17, and 19 years. Vigorous-intensity was defined as 4,012 ActiGraph counts per
minute. VPA minutes from multiple assessment days were averaged. Sports participation was
assessed using a questionnaire every six months. Group-based trajectory analysis was used to
identify diverse patterns in VPA and sport participation.

Results: Based on the data from 537 IBDS participants (50% male, 95% White), four VPA
trajectory patterns were identified: Group 1 (7.2%) = increasing from approximately 10
minutes/day at age 5 to approximately 35 minutes at age 19; Group 2 (75.3%) = consistently
low below 10 minutes/day (unhealthy VPA pattern); Group 3 (13.7%) = increasing during
preadolescence and decreasing during adolescence; Group 4 (3.8%) = consistently high above
20 minutes/day. The sex distribution was significantly different across the four groups
(proportions of males: 66% in Group 1, 40% in Group 2; 90% in Group 3, and 85% in Group 4;
p < 0.01). Compared to youth who dropped out of sport participation, youth who maintained
sport participation were more likely to be in Group 1 or Group 4, as opposed to Group 2 (p <
0.01).
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1291

Conclusion: In our 14-year longitudinal follow-up, three-fourths of the participants followed

an unhealthy VPA pattern. This study suggests that females may be missing distinct health
benefits associated with VPA during the growing years. Research on health consequences,
such as obesity, related to these VPA patterns should follow.

CORRESPONDING AUTHOR: Soyang Kwon, PhD, Lurie Children's Hospital of Chicago, Chicago,
IL, 60611; [email protected]
S1292 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B227 6:00 PM-7:00 PM

MERITORIOUS AWARD WINNER

DIFFERENTIAL EFFECTS OF THE COMPANY OF MOTHERS VERSUS FATHERS ON PHYSICAL

ACTIVITY AND SEDENTARY BEHAVIOR IN BOYS AND GIRLS

Eldin Dzubur, MS, Sydney G. O'Connor, BA, Gayla Margolin, Ph.D., Genevieve F. Dunton, PhD,
MPH

University of Southern California, Los Angeles, CA

Background: The company of parents is thought to increase childrens level of physical

activity and decrease sedentary behavior. These effects, however, may differ for boys and
girls. For example, compared to mothers, fathers have a stronger influence on their
daughters physical activity. However, prior studies examining the role of parental influence
on childrens physical activity and sedentary behavior have been limited by retrospective
recall, and the failure to concurrently measure activity and social context. The aim of this
study was to determine whether children's physical activity and sedentary behavior levels
varied across different types of parental contexts (i.e., mothers, fathers, or both parents).
Methods: Children (N=192, ages 8-12, 53% female) wore accelerometers and answered
Ecological Momentary Assessment (EMA) survey prompts by smartphone up to 7 times a day
for 8 days. Each EMA survey prompt assessed whether one, both, or neither parents were
present. Accelerometer data were used to summarized minutes of moderate-to-vigorous
physical activity (MVPA) and sedentary behavior (SB) across the 30-minute window prior to
each EMA prompt. Separate multilevel models were used to examine main effects of type of
parental company girls and boys physical activity, disaggregating variance (between-subject
[BS] and within-subject [WS]) and adjusting for age and ethnicity.
Results: WS findings show that girls engaged in more SB when neither parent was present
than when they reported being with their mother or both parents (WS, z's=2.60 and 2.38,
respectively; p's < 0.05). Overall, BS findings show that girls who reported being with their
father more often than being with their mother, with neither parent, or with both parents,
engaged in more MVPA (BS, z's=-3.63, -2.92, and -2.78, respectively; p's < 0.01). Additionally,
boys who reported being with their mother more often than with neither parent or with both
parents, engaged in less SB (BS, z's=-2.60 and -2.25; respectively, p's < 0.05).
Discussion: The findings are consistent with prior research suggesting that spending more
time with fathers may promote physical activity to a greater extent in girls than boys. The
results indicate that reporting being with mothers more frequently may reduce sedentary
behavior more in boys. However, girls were less sedentary when concurrently with mothers or
both parents. Future interventions targetting sedentary behavior and physical activity may
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1293

seek to vary or prioritize the amount of time spent with each parent differently for boys and
girls.

CORRESPONDING AUTHOR: Eldin Dzubur, MS, University of Southern California, Los Angeles,
CA, 90032; [email protected]
S1294 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B228 6:00 PM-7:00 PM

EFFECTS OF A COMMUNITY-BASED EXERCISE PROGRAM ON PHYSICAL FUNCTION AND

EXERCISE-RELATED OUTCOMES IN OLDER ADULTS

Susan S. Levy, Ph.D.1, Katie J. Thralls, M.S., C.S.C.S.2, Kari Carmody, MPH3, Taylor B. Krippes,
MS1

1
San Diego State University, San Diego, CA; 2San Diego State University/University of
California, San Diego; Joint Doctoral Program, San Diego, CA; 3Aging and Independence
Services of San Diego County, San Diego, CA

Fostering healthy aging reduces risks of diminished function and loss of independence.
Exercise programs for older adults have led to improved function, yet few have been
integrated into services provided by state and related agencies serving older adults.
Community exercise programs have confirmed feasibility in their delivery and acceptance by
older adults, yet many lack controlled studies examining their efficacy on desired outcomes.
This study examined the efficacy of a community-based exercise program, using a controlled
design.

Participants (N=169) were recruited from senior centers served by the County Area Agency on
Aging, who offered exercise classes in 21 centers. Intervention participants (n=87,
Mage=73.38.5) were recruited from new classes starting at 7 centers. Controls (n=82,
Mage=74.97.7) were recruited from matched sites not offering classes. Classes emphasized
physical function related to activities of daily living (ADL), met twice weekly, and were taught
by trained instructors. Exercises were low-impact and focused on proper execution of
movement, core stabilization, strength, balance, flexibility, and aerobic fitness. Validated
measures of

physical function, exercise self-efficacy, and exercise behavior (outside of class) were
administered at baseline and 6 months.

Mixed design ANOVAs revealed significant improvements in upper body strength (p < .001) as
measured by the arm curl (24% increase), lower body strength (p < .001) as measured by the
30-second chair rise (21% increase), aerobic endurance (p < .001) as measured by the 6-
minute walk (8% increase in distance), and mobility (p < .004) as measured by the 8-ft up and
go (8% decrease in time) in the exercise, but not control group over time. Significant increases
in exercise self-efficacy (p < .002) and moderate/vigorous intensity activity (p < .05), and days
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1295

of muscular-based training (p < .005) were found in the exercise group but not controls. No
significant changes (p > .05) were found in flexibility training in either group.

These findings support the efficacy of a County-wide disseminated exercise program in

improving ADL-related function, and exercise-related outcomes in older adults. Building an
evidence base for community delivered programs should provide impetus for increasing
dissemination through involvement of appropriate state and national agencies. This will
increase the impact of programs on a growing population of older adults, and foster healthy
aging initiatives.

CORRESPONDING AUTHOR: Susan S. Levy, Ph.D., San Diego State University, San Diego, CA,
92115; [email protected]
S1296 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B229 6:00 PM-7:00 PM

INTRINSIC MOTIVATION FOR EXERCISE AND PATTERNS OF HEALTH BEHAVIORS

Elise L. Rice, PhD1, Chan Thai, PhD, MPH2, Rebecca A. Ferrer, PhD3, Annette Kaufman, PhD,
MPH4, William Klein, PhD5

1
National Cancer Institute, Bethesda, MD; 2Santa Clara University, Santa Clara, CA; 3National
Cancer Institute, Rockville, MD; 4Tobacco Control Research Branch, Behavioral Research
Program, Division of Cancer Control and Population Sciences, National Cancer Institute,
National Institutes of Health, Rockville, MD; 5NCI/NIH, Rockville, MD

The value of intrinsic motivation (IM; being driven by enjoyment to engage in a behavior) for
sustainable exercise has been documented in prior research. Given that most Americans fall
short of physical activity (PA) guidelines, it is crucial to understand the prevalence of IM for
PA. These analyses extend previous work by examining the relationship between IM and
exercise in a nationally representative sample, identifying moderators of that relationship,
and exploring whether IM for exercise is related to other health behaviors. The current study
utilized the Health Information National Trends Survey (HINTS; 2012-2014). Respondents
(N=7,307) indicated how much getting enjoyment from exercise influenced why they would
start or continue exercising regularly. They also reported on their patterns of PA, sedentary
behavior, fruit and vegetable (F&V) consumption, and sunscreen use, as well as their height,
weight, and smoking history. As expected, respondents IM was associated with more PA in a
typical week (=62.07, p < .01). However, that association was moderated by gender and BMI,
such that the relationship between IM and PA was weaker among women (=-46.65, p=.02)
and those with higher BMI (=-4.01, p=.02). Additionally, IM was associated with less
sedentary behavior (=-0.33, p < .01), although that relationship was moderated by self-
reported physical health, such that the link was stronger among those who reported that their
general health was poor (=0.30, p=.01) and those who reported having been diagnosed with
a medical condition (=0.30, p=.02). Finally, IM for exercise was also associated with other
health behaviors, even controlling for PA, including: F&V consumption (=0.55, p < .01),
sunscreen use (=0.13, p < .01), and non-smoker status (=-0.16, p=.03). Collectively, these
findings suggest that although IM for exercise is a robust statistical predictor of self-reported
heath behavior even beyond the domain of PA its predictive power is variable. Although
further study is needed, these results suggest opportunities to tailor IM-based interventions
to participants, which could potentially impact the broader lifestyle beyond the immediate
target behavior.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1297

CORRESPONDING AUTHOR: Elise L. Rice, PhD, National Cancer Institute, Bethesda, MD, 20892;
[email protected]
S1298 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B230 6:00 PM-7:00 PM

LATENT CLASS ANALYSIS OF FACTORS ASSOCIATED WITH EXERCISE DEPENDENCE AND EATING
PATHOLOGY

Brian Cook, PhD, Steven Kim, PhD

California State University Monterey Bay, Seaside, CA

Background: Exercise dependence is a multifaceted maladaptive pattern of exercise attitudes,

cognitions, and behaviors that result in significant health detriments. Specifically, it has been
identified as a key factor that may mediate the relationship among exercise and related
psychopathology (e.g., eating disorders, behavioral addiction, etc.). Several factors have been
associated with exercise dependence with and without comorbid psychopathology. However,
few studies have included multiple factors that may clarify individual differences among
exercise dependence. Classifying profiles of factors associated with exercise dependence and
related eating pathology may identify individual differences that could allow for tailored
interventions to maximize exercise-related benefits. Thus, the purpose of this study was to
examine profiles of factors associated with exercise dependence and eating pathology.
Method: Participants (N=1374; mean age = 28.24, SD = 10.20, 65.4% female) completed an
online survey that included the Leisure-time Exercise Questionnaire (LTEQ), Exercise
Dependence Scale (EDS), Positive and Negative Affect Schedule, Behavioral
Inhibition/Behavioral Activation Scales (BIS/BAS), and the Eating Disorders Examination-
Questionnaire (EDEQ). Latent class analysis was used to classify latent groups. We used
multiple imputations to account for uncertainty for missing data then selected a final model
which minimizes Bayesian Information Criterion to balance between model fit and simplicity.
Results: Analysis identified four distinct profiles: group 1 (36%) displayed high levels of EDEQ,
LTEQ, negative affect (NA) and EDS; group 2 (17%) displayed high EDS, LTEQ, BAS, and positive
affect (PA), and low EDEQ scores; group 3 (22%) displayed high EDEQ, BIS, and NA, and low
EDS scores; group 4 (25%) displayed low scores for all variables. Discussion: Results of this
study suggest that relationships among exercise dependence, eating pathology,
temperament, and affect may be more heterogeneous than previously conceptualized. This
study may inform researchers and clinicians on how to identify who may benefit from exercise
interventions, factors that should be considered when using exercise as behavioral medicine,
and identifying risk for exercise-associated pathology.

CORRESPONDING AUTHOR: Brian Cook, PhD, California State University Monterey Bay,
Seaside, CA, 93955; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1299

B231 6:00 PM-7:00 PM

OVERALL EFFECTIVENESS OF INTERVENTIONS TO INCREASE PHYSICAL ACTIVITY IN AFRICAN

AMERICAN WOMEN: A META-ANALYSIS

Brittany Miller, BA1, Anne Moyer, Ph.D.2

1
Stony Brook University, East Setauket, NY; 2Stony Brook University, Stony Brook, NY

African American women have one of the lowest physical activity rates and highest obesity
rates in the United States. Consequently, they are at increased risk for and suffer
disproportionately from related illnesses, such as cardiovascular disease and diabetes.
Researchers have developed and tested interventions to increase physical activity in African
American women, but their effectiveness lacks consistency, particularly over time. Two meta-
analyses, utilizing pre-intervention and post-intervention data, were conducted to examine
the overall effectiveness of physical activity interventions in African American women. Three
bibliographic databases were searched, and nine studies were included in the final meta-
analysis. The samples of the nine studies ranged from N = 13 to N = 447 African American
women. The mean age ranged from M = 21.2 to M = 60.1 years. Intervention length ranged
from 1.5 to 6 months. Five of the nine studies had data extracted from self-report measures,
while the other four had data extracted from a monitoring device. The interventions included
a range of components, including exercise training, motivational interviews, group exercise
sessions, problem-solving to overcome exercise barriers, group sessions addressing exercise
benefits, and pedometer use. The first meta-analysis extracted the baseline physical activity
measurement as pre data and the end of the study physical activity measurement as post
data. The second meta-analysis extracted the baseline physical activity measurement as pre
data and the last physical activity measurement taken as post data. The meta-analyses yielded
moderate aggregate effect sizes of r = 0.52 and r = 0.49, respectively. The results suggest that
interventions significantly increase engagement in physical activity, but those effects may be
slightly less strong in the long term. Additionally, whether the intervention design was
individual-based, group/community-based, or a mix of the two was found to be a significant
moderator in the second meta-analysis. Interventions targeting African American women are
a promising way to increase the rate of physical activity in this population, however, it may be
beneficial for researchers to focus on how to maintain their effects over time. Improving the
long-term effects of these physical activity interventions may be helpful in addressing related
health disparities negatively impacting African American women.

CORRESPONDING AUTHOR: Brittany Miller, BA, Stony Brook University, East Setauket, NY,
11733; [email protected]
S1300 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B232 6:00 PM-7:00 PM

PERCEIVED NEIGHBORHOOD ENVIRONMENT AND PHYSICAL ACTIVITY AMONG AFRICAN

AMERICAN WOMEN IN RURAL NORTH CAROLINA:

Mary L. Greaney, PhD, MPH1, Alyssa Platt, MA2, Dori Steinberg, PhD, RD2, Erica Levine, MPH2,
Perry Foley, MPH, MSW3, Gary G. Bennett, PhD4

1
University of Rhode Island, Kingston, RI; 2Duke Global Health Institute, Durham, NC; 3OCHIN,
Inc., Milwaukie, OR; 4Duke University, Durham, NC

Background/purpose: More than 75% of Black women living in the US are overweight or have
obesity, and more than 50% are not meeting current guidelines for physical activity. The
potential effects that the built environment and community have on physical activity
behaviors of Black women living in rural areas have not been thoroughly researched. To
address this gap, we conducted a cross-sectional secondary analysis of baseline data from The
Shape Program to explore how participants perception of their environment influenced
physical activity and sedentary time.

Methods: The Shape Program, a randomized control trial of a weight gain prevention
intervention implemented in community health centers in rural North Carolina, was designed
for socioeconomically disadvantaged Black women who are overweight or have obesity.
Participants perceptions of their environment were assessed using item from the Boehmer &
Brownson instrument (e.g., seeing billboards encouraging physical activity, seeing trees along
community streets, presence of sidewalks). The sample for the current study included
participants with 10+ valid days of accelerometer data and valid survey data for perceived
environment variables (n=159; 81.2% of Shape participants). Accelerometer data were
aggregated into 1- minute bouts of moderate to vigorous physical activity (MVPA) and
minutes of sedentary time. Assessed Boehmer & Brownson items were entered into an
exploratory factor analyses, a Scree test was conducted, and identified factor scores were
entered into log-linear ordinary least squares regression models examining MVPA and
sedentary time.

Results: Participants had an average of 52 minutes (median = 34) of MVPA and 663 minutes
(SD=173.2)/day of sedentary time. The exploratory factor analyses and Scree test identified 4
distinct perceived environmental factors: 1) facility of travel; 2) community aesthetic; 3)
physical activity messaging, and 4) existing physical activity facilities. After adjusting for
covariates, a 1 SD change in factor score for community aesthetic was associated with a
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1301

26.5% (95% CI: 7.9% - 45.1%) higher level of MVPA when measured. None of the examined
factors correlated with sedentary activity.

Conclusion: Among this sample of socioeconomically disadvantaged Black women, none of

the examined factors were associated with sedentary time and only community aesthetic
was associated with physical activity. Our results suggest that no other environmental or
community variables influenced physical activity.

CORRESPONDING AUTHOR: Mary L. Greaney, PhD, MPH, University of Rhode Island, Kingston,
RI, 02881; [email protected]
S1302 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B233 6:00 PM-7:00 PM

POST-EXERCISE AFFECTIVE RESPONSE: EXAMINING DIFFERENCES BETWEEN REGULAR AND

INFREQUENT EXERCISERS

Margarita Sala, BS1, David Rosenfield, PhD2, Austin S. Baldwin, PhD2

1
Southern Methodist University, Dalas, TX; 2Southern Methodist University, Dallas, TX

Background: Exercise-related affect is a critical factor in understanding why some people

exercise regularly and others do not. The vast majority of people report an improvement in
core affect after exercise (i.e., feeling better), and although it is reasonable to expect that
feeling better after exercising would reinforce exercise behavior, fewer than half of U.S.
adults exercise regularly (Centers for Disease Control Prevention, 2013). It is important to
examine the influence of post-exercise affect on exercise because post-exercise affect may
play a unique role in influencing future exercise. The purpose of this study was to determine
the extent to which specific post-exercise affective states (i.e., positive affect, negative affect,
fatigue, calmness, relief) differ between regular and infrequent exercisers, thereby elucidating
the feeling better phenomenon.

Methods: Regular (N = 32; > 75 minutes per week of vigorous exercise) and infrequent
exercisers (N = 25; < 30 minutes per week of vigorous exercise) completed a 10-minute bout
of vigorous exercise on a treadmill and reported their affective states at various time points
during the bout and during a 30-minute post-exercise period.

Results: Consistent with previous findings, both regular (p = .004) and infrequent exercisers (p
< .001) reported an improvement in core affective valence after the exercise bout (i.e.,
feeling better). However, infrequent exercisers reported a higher decrease in negative
affect (p = .01) as well as a greater sense of relief (p = .003) immediately post-exercise than
regular exercisers. Relief was more strongly correlated with changes in negative affect (r = .38)
than with changes in positive affect (r = -.12). Regular exercisers reported a higher increase in
positive affect (p = .01) immediately post-exercise than infrequent exercisers. Finally, there
were no differences in post-exercise fatigue and calmness.

Implications: Both regular and infrequent exercisers experienced increases in core affective
valence immediately post-exercise. However, for infrequent exercisers, feeling better
appears to be due to a decrease in negative affect and a greater sense of relief immediately
post-exercise. This study demonstrates how post-exercise affective states systematically differ
between regular and infrequent exercisers, thereby elucidating the feeling better
phenomenon
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1303

CORRESPONDING AUTHOR: Margarita Sala, BS, Southern Methodist University, Dalas, TX,
75204; [email protected]
S1304 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B234 6:00 PM-7:00 PM

PSYCHOSOCIAL FACTORS AS MODERATORS OF ASSOCIATIONS BETWEEN BUILT

ENVIRONMENT AND ADOLESCENTS ACTIVE TRANSPORTATION

Terry L. Conway, PhD1, Xiaobo Wang, PhD2, Kelli L. Cain, MA3, Lawrence D. Frank, PhD4, Brian
E. Saelens, PhD5, Carrie M. Geremia, BA6, Jacqueline Kerr, PhD, MSc7, Karen Glanz, PhD,
MPH8, Jordan A. Carlson, PhD, MA9, James F. Sallis, PhD1

1
University of California, San Diego, San Diego, CA; 2Zhengzhou University of Light Industry,
Zhengzhou, Henan, China (People's Republic); 3University California San Diego, San Diego, CA;
4
University of British Columbia, Vancouver, BC, Canada; 5University of Washington, Seattle,
WA; 6UCSD, San Diego, CA; 7Department of Family Medicine and Public Health, UC San Diego,
La Jolla, CA, USA, La Jolla, CA; 8University of Pennsylvania, Philadelphia, PA; 9Center for
Children's Healthy Lifestyles & Nutrition, Kansas City, MO

Objective: Factors across different levels of influence (e.g., individual and environmental) may
relate to active transport physical activity (PA) of adolescents, but the multiple levels are
seldom studied concurrently. The primary aim here was to examine individual-level
psychosocial variables as moderators of associations between objective built environment
measures and adolescents reported active transport PA.

Methods: Mixed-effects regression models were examined in the Teen Environment and
Neighborhood (TEAN) observational study (N=928, 12-16 year old adolescents). An index of
active transportation to/from school and neighborhood destinations was the dependent
variable. Explanatory constructs explored were from psychosocial and objective built
environment domains. Seven psychosocial measures related to PA (self-efficacy, enjoyment,
athletic ability, decisional balance, barriers to PA in the neighborhood, social support from
peers and family) were self-reported by adolescents. Objective built environment measures
included a neighborhood walkability index and count of nearby parks and recreation facilities
derived from Geographic Information System (GIS) procedures, and an environmental audit of
streetscape quality (MAPS tool). Covariates included demographic characteristics and study
design factors.

Results: Three of 21 interactions were related to the active transportation index (ps < 0.10):
self-efficacy X GIS walkability index, barriers to PA X MAPS streetscape score, and self-efficacy
X GIS count of parks and recreation facilities. Each interaction showed a synergistic pattern of
higher active transport scores among adolescents with the combination of activity-supportive
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1305

built environment and pro-physical activity psychosocial characteristics. All 3 objective

environment measures and 3 psychosocial variables (self-efficacy, social support from peers,
and enjoyment of PA) also had significant positive main effects with active transportation (ps
< 0.056).

Conclusion: Results provided modest support for the ecological model principle of interactions
across levels of influence, highlighted the importance of built environment and psychosocial
factors in facilitating or hindering adolescents active transportation, and suggested that
strategies for improving adolescents active transportation may be most effective when
targeting multiple levels of influence in ecological models.

CORRESPONDING AUTHOR: Terry L. Conway, PhD, University of California, San Diego, San
Diego, CA, 92103; [email protected]
S1306 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B235 6:00 PM-7:00 PM

PSYCHOSOCIAL PREDICTORS OF PHYSICAL ACTIVITY CHANGE AMONG COLLEGE STUDENTS IN

AN OBESITY PREVENTION TRIAL

Danielle Arigo, Ph.D., Licensed Psychologist1, Paul Rohde, Ph.D.2, Heather Shaw, Ph.D.2, Eric
Stice, Ph.D.2

1
The University of Scranton, Scranton, PA; 2Oregon Research Institute, Eugene, OR

Accumulating evidence shows the importance of moderate-to-vigorous physical activity

(MVPA) for maintaining a healthy body weight, particularly among those at risk for weight
gain. At present, however, little is known about psychological barriers to maintaining MVPA in
at-risk groups. As obesity prevention programs involve multiple behavior changes, identifying
individuals who specifically are unlikely to adopt or maintain MVPA in this context could
improve weight control outcomes. In the present study, we examined predictors of MVPA in
an obesity prevention trial targeted toward college students at risk for weight gain (n = 364;
72% female, MBMI = 23.4 kg/m2). Based on their proposed relations to self-regulatory abilities,
predictors of interest were impulsiveness, cognitive dissonance regarding unhealthy
behaviors, and perceived sociocultural pressure to be thin. Participants were randomly
assigned to one of three intervention conditions and completed assessments at pre- and post-
intervention, and at 6- and 12-month follow-ups. These assessments included measured
height and weight, self-report measures of psychosocial experiences, and four days of
waistband accelerometer wear. Of the total variability in MVPA minutes per day over 12
months, only 33% was attributable to between-person differences; across conditions,
participants decreased their total MVPA minutes per day during this time (B = -5.48, p < 0.01).
Baseline self-report scores for both impulsiveness and cognitive dissonance regarding
unhealthy behaviors negatively predicted MVPA over time. Participants with elevated baseline
impulsiveness scores (B = -7.36, p = 0.04) and dissonance scores (B = -3.89, p = 0.05) began the
study with more MVPA minutes, but showed sharper declines over time. These differences
(~5-30 minutes over four days) represent 3-22% of recommended MVPA per week.
Sociocultural pressure was not related to MVPA change (B = -1.56, p = 0.36). Targeted MVPA-
focused intervention for students who show elevated impulsiveness and cognitive dissonance
may improve both MVPA and weight control outcomes for these individuals over the long
term.

CORRESPONDING AUTHOR: Danielle Arigo, Ph.D., Licensed Psychologist, The University of

Scranton, Scranton, PA, 18510; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1307

B236 6:00 PM-7:00 PM

RELIABILITY OF ACTIVITY MONITORS FOR TRACKING STEPS AND ESTIMATING ENERGY

EXPENDITURE DURING A GRADED MAXIMAL TREADMILL TEST

Bradley J. Kendall, B.A., M.S.1, Bryanne Bellovary, M.S.2, Neha Gothe, MA, MS, PhD3

1
Wayne State University, Grosse Pointe Park, MI; 2University of New Mexico, Albuquerque,
NM; 3Wayne State University, Detroit, MI

Background: Due to the increased awareness of the importance of exercise and physical
activity (PA), individuals are more interested in their ability to track steps and energy
expenditure (EE). It is expected that individuals as well as employers, insurers, and clinicians
will begin to use activity monitors (AMs) more frequently in order to address health
behaviors. To date, there have been mixed findings on the accuracy of various AMs during
exercise and PA. Therefore, the purpose of this study was to (1) assess the accuracy of EE
estimation and step tracking abilities of six AMs in relation to indirect calorimetry and hand
counted steps and (2) assess the accuracy between high and low fit individuals.
Methods: Fifty participants (mean age: 25.848.10 years) wore the Basis watch, Fitbit Flex,
Polar FT7, Jawbone, Omron pedometer, and Actigraph during a maximal graded treadmill test.
Correlations, intra-class correlations (ICCs), and one-sample t-tests determined accuracy and
agreement between AMs and criterions.
Results: All AMs were significantly correlated with the hand counted steps (overall ICC=0.994,
p < 0.001, CI 95%=0.991-0.996). One sample t-tests revealed that the Omron, Fitbit, and
Actigraph showed agreement with the steps standard (p>0.05), while the Basis and Jawbone
significantly underestimated steps (p < .05). All fitness trackers were significantly correlated
with the metabolic cart (overall ICC=0.934, p < 0.001, CI 95%=0.856-0.968). One sample t-tests
revealed that no device showed agreement with the metabolic cart (p>0.05). When split by
fitness categories, the low fit group showed better correlations compared with the high fit
group. One sample t-tests revealed the Actigraph and Omron showed agreement to indirect
calorimetry in the low fit group while the Polar FT7 and Basis watch showed agreement in the
high fit group.
Conclusions: The Omron, Fitbit, and Actigraph were reliable for measuring steps. AMs were
more reliable for measuring EE at low intensity exercise compared to high intensity. Future
research on reliability of these devices as well as manufacturing and development of new AMs
should account for the greater metabolic demand of higher intensity activities in order to
improve EE estimates.
S1308 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Bradley J. Kendall, B.A., M.S., Wayne State University, Grosse
Pointe Park, MI, 48230; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1309

B237 6:00 PM-7:00 PM

RESPONSE PATTERNS AND PREDICTORS OF COMPLIANCE TO ECOLOGICAL MOMENTARY

ASSESSMENT: RESULTS FROM THE MOVINGU STUDY

Matthew Kwan, PhD1, Chloe Bedard, B.HSc.1, Scott Veldhuizen, MSc2, John Cairney, PhD3

1
McMaster University, Hamilton, ON, Canada; 2Brock University, Hamilton, ON, Canada;
3
University of Toronto, Hamilton, ON, Canada

Background: The transition out of high school represents a major life transition often
associated with significant declines in physical activity behaviors (Bray & Born 2004; Kwan, et
al., 2012). To date, the extant literature is largely based on cross-sectional or infrequent
longitudinal designs, and are limited in our understanding of contextual and intra-individual
factors impacting physical activity during this transitory time. Ecological momentary
assessment (EMA) is an intensive data capture method that can address these limitations
(Shiffman et al., 2008). The purpose of the current study was to describe the utility of EMA
data in a sample of first-year university students, and to examine several time invariant and
time-varying predictors of EMA compliance. Method: Current sample included 96 recent high
school graduates (Mage=18.30 +.54;males=45%), whom were asked to wear a wrist-worn
accelerometer and to complete EMA prompts 7x/day for a period of five days (Wed through
Sun). Each prompt included a very-brief questionnaire assessing context, acute feeling and
mood states, and social cognitions. Results: Overall, 68% of the sample met the minimal
acceptable compliance (answered >3 prompts on >3 days), of which only 26% were
considered to have excellent compliance (answered >5 prompts for >4 days). Results of
logistic regression did not find any time-invariant factors being related to compliance;
however, findings from the mixed effects logistic model indicates that there was a significant
time of day (Coefficient= -5.10 SE=.75, pDiscussions: Overall compliance to EMA prompts
were modest in our current sample, thus it was important to investigate for potential non-
response biases. While socio-demographic or behavioural factors did not predict response
compliance, our results found prompts being sent in the afternoons/early evenings were
more likely to be answered compared to mornings and later evening. Findings also showed
linear response pattern decline, suggesting that participants either became less inclined to
respond over time or less likely to respond on weekend days. Understanding factors related to
EMA compliance is an important step to recognize the potential biases for this data-intensive
sampling strategy, and current findings suggest that future EMA studies may require a smaller
sampling frame for the emerging adulthood population.
S1310 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Matthew Kwan, PhD, McMaster University, Hamilton, ON,

L8P1H6; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1311

B238 6:00 PM-7:00 PM

RESULTS OF THE GUIDELINES FOR EXERCISE IN MULTIPLE SCLEROSIS (GEMS) PROJECT: A

FEASIBILITY RANDOMIZED CONTROLLED TRIAL

Brynn Adamson, MS1, Yvonne C. Learmonth, PhD2, Dominique Kinnett-Hopkins, B.S. 1, Maria
Bohri, BS3, Robert W. Motl, PhD4

1
University of Illinois at Urbana-Champaign, Urbana, IL; 2University of Illinois, Urbana, IL;
3
University of Illinois at Urbana-Champaign, Washington, DC; 4University of Alabama at
Birmingham, Birmingham, AL

There is increasing recognition that exercise is an efficacious strategy for managing many
consequences of multiple sclerosis (MS), yet this population engages in less exercise than the
general population, and other chronic disease populations. The poor uptake of exercise
among persons with MS might be associated with the lack of emphasis on translational and
feasibility research. Additionally, most exercise studies completed in this population are
center-based and supervised, and not easily translated into practice and widespread
adoption. We conducted a randomized controlled trial in order to examine the feasibility of a
4-month, home-based exercise training program based on recent, MS-specific physical activity
guidelines. Feasibility was assessed based on the four domains of process (e.g., recruitment),
resource (e.g., monetary costs), management (e.g., personnel time requirements), and
scientific outcomes (e.g., treatment effect). We recruited persons with mild-to-moderate MS
who were randomized into an intervention or a wait-list control condition. Participants in the
intervention condition received exercise equipment (i.e., a pedometer, resistance bands, DVD,
training manual, calendars, and logbook), video coaching calls via Skype TM, and newsletters
based on principles of Social Cognitive Theory. Participants in both conditions completed
home-based assessments before and after the 4-month period. Ninety-nine persons with MS
were assessed for eligibility, and 57 were randomized. Fifty-one persons completed the study
(90%). Total costs of the study were US$5331.03. Personnel time to conduct the study totaled
263 hours. Participants in the intervention group complied fully with 71% of all exercise
sessions. There was a statistically significant time by group interaction on Godin Leisure-Time
Exercise Questionnaire scores (F=6.94, P p2 = 0.12); there was a moderate increase (d0.5) in
self-reported exercise behavior for the intervention participants. Qualitative surveys and
interviews indicated overall satisfaction with all aspects of the intervention. These results
support the feasibility, efficacy and acceptability of a home-based exercise intervention based
on the physical activity guidelines and supplemented with behavioral strategies for adults
with mild-to-moderate MS. This is an important step towards providing adequate and
S1312 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

appropriate resources that can aid in the adoption of exercise behavior among adults with
MS.

CORRESPONDING AUTHOR: Brynn Adamson, MS, University of Illinois at Urbana-Champaign,

Urbana, IL, 61801; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1313

B239 6:00 PM-7:00 PM

REVISED MODEL FOR PREDICTING ATTRITION FROM EXERCISE

Jason D. Cohen, B.S.1, Tiffany Bullard, B.S. 1, Daniel Palac, MA2, Arthur Kramer, Ph.D.3, Edward
McAuley, PhD4, Sean P. Mullen, PhD1

1
University of Illinois at Urbana-Champaign, Urbana, IL; 2University of Illinois at Urbana-
Champaign, Champaign, IL; 3Northeastern University, Boston, MA; 4University of Illinois at
Urbana Champaign, Urbana, IL

Exercise behavior maintenance is challenging, particularly for people with many barriers. It
would benefit researchers and practitioners to have an accurate method to identify
individuals exhibiting characteristics associated with a high likelihood of dropout from an
exercise trial. Using brief measures, serving as proxy measures of poor self-regulatory
functioning, Mullen et al (2013) provided initial evidence for such a profile for attrition from
exercise based on 4 variables (i.e., memory complaints, self-efficacy for overcoming barriers
to exercise, left leg balance time, and stairs down time) in a sample of older adults (65+)
involved in an exercise trial. In the current NHLBI funded study (n=133 middle-aged adults, 45-
64 years; 3 values were imputed for 4 participants using EM algorithm imputation), we
replicated the original model using discriminant function analysis (DFA); note the perceived
memory item was taken from a different scale and a longer, 60s balance assessment was
used. The DFA found that 71.3% were correctly classified vs. 72.22% of the original sample,
however the memory complaint item was a non-significant discriminating variable (p=.68).
Self-efficacy for overcoming barriers (66.74 vs. 56.14), left leg balance time (38.46s vs. 30.32),
and stairs down time (5.61s vs. 6.65) all differentiated (p < .05) "completers" from "dropouts,"
respectively. Upon re-examination of theory and research, and given the age difference
between samples, a revised model was tested. Current physical activity level (3-day average
Fitbit step count) replaced memory complaints and the revised model (imputed data showed
similar results, in parentheses) correctly classified 72.1% (71%) of the total sample (Wilk's =
.84, 2 (4) = 22.57, p < .01, eigenvalue = .20, canonical correlation = .41). Classification results
determined that the new model accurately classified 74% (72%) of those who remained for
the entirety of the exercise trial, as well as 66.7% (69.7%) of those who officially requested to
withdraw from the study or discontinued contact. These findings support using a brief battery
for identifying middle-aged and older adults who may need some additional support to
enhance their self-regulatory skills.
S1314 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Jason D. Cohen, B.S., University of Illinois at Urbana-Champaign,

Urbana, IL, 61801; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1315

B240 6:00 PM-7:00 PM

SLEEP DURATION AND ITS EFFECT ON DAILY MVPA

Janine Nguyen, N/A 1, Chrisanta Tomaneng, N/A 2, Margaret Schneider, PhD3

1
University of California, Irvine, Orange, CA; 2University of California, Irvine, Winchester, CA;
3
University of California, Irvine, Irvine, CA

Background: Although physical activity is a key factor to a healthy lifestyle, much of todays
youth are not getting the recommended amount of physical activity. Only 24.8% of
adolescents in the U.S. engage in at least 60 minutes per day of moderate to vigorous physical
activity (MVPA) and it is suspected that sleep duration may influence activity. Findings have
been inconclusive with respect to this relationship.

Methods: Middle school students (N=182, mean age = 11.06 years) participated in this study.
MVPA was objectively monitored for seven days with an ActiGraph. During the same 7-day
period, sleep data were collected using a self-report log that documents wake and sleep time.
Participants also completed a test of cardiorespiratory endurance (VO2peak) and a depression
inventory and were assessed for height and weight to determine Body Mass Index percentile
(BMI). Correlational and regression analyses assessed the relationship of average sleep
duration with average minutes of MVPA per day.

Results: The participants average sleep time was 9.55 hours per night (SD = 0.52 hours) and
average MVPA was 49.05 minutes/day (SD = 22.47 minutes). After removing participants with
incomplete data, correlation analysis (N = 151) showed no association between sleep and
activity (p > 0.05). Results were the same for both males and females and for teens with
relatively low versus higher scores on a depression inventory, and the association remained
non-significant after controlling for VO2peak and BMI.

Conclusions: Average sleep duration had no correlation with average MVPA in this study,
suggesting that young adolescents participation in physical activity is not influenced by their
usual amount of sleep. It is possible, however, that unmeasured aspects of the environment
such as parental control of time to bed may have influenced self-reported sleep duration, thus
obscuring the sleep-activity relationship. Using an objective measure of sleep duration would
strengthen future studies.

CORRESPONDING AUTHOR: Janine Nguyen, N/A , University of California, Irvine, Orange, CA,
92869; [email protected]
S1316 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B241 6:00 PM-7:00 PM

THE RELATIONSHIP BETWEEN SELF-EXPANSION AND FITBIT-MEASURED PHYSICAL ACTIVITY:

DAILY DATA ACROSS 4 WEEKS

Xiaomeng (Mona) Xu, PhD1, Samantha J. Tupy, M.A.2, Ashley L. Miller, B.S. in Psychology3,
Danielle Correll, BS4, Rick Tivis, MPH5, Claudio R. Nigg, PhD, FSBM6

1
Idaho State University, Pocatello, ID; 2Idaho State University, Chubbuck, ID; 3University of
Oregon, Eugene, OR; 4University of Colorado Colorado Springs, Colorado Springs, CO; 5Idaho
State University, Meridian, ID; 6University of Hawaii, Honolulu, HI

Physical activity (PA) is related to the prevention of chronic diseases, however most
Americans do not engage in sufficient PA. Translational research promoting PA behavior and
adherence is recommended. One novel factor associated with health behaviors and
adherence is self-expansion. The self-expansion model states that people are intrinsically
motivated to enhance abilities to achieve goals by increasing perspectives, identities, and
resources. Self-expanding activities are characterized by novelty, excitement, and interest or
challenge (e.g., hobbies). Self-expansion has been associated with better outcomes in
behavioral weight loss, including greater self-reported PA. This study is the first to investigate
the link between self-expansion and objective PA and their relationship on a daily basis (past
research has only examined pre-post). Community participants (n= 50, 72% female, 88%
white, Mage = 39.96, SDage = 12.14) completed baseline assessment, wore a Fitbit OneTM and
completed a daily survey for 4-weeks, and completed follow-up. Participants were not given
PA directions as this was not an intervention. Retention was 100% and compliance was
99.29% for wearing the Fitbit and 96.86% for completing the 28 daily surveys. Baseline
measures included general self-expansion (M = 74.12, SD = 15.5, max possible 98) and PA-
specific self-expansion (M = 62.92, SD = 18.54, max possible 91). Daily surveys included short
versions of the general and PA-specific self-expansion scales. Across the 4-weeks, participants
averaged 8589 daily steps (SD= 5151). Baseline general self-expansion did not significantly
predict steps taken over the 4-weeks, however baseline PA-specific self-expansion did (t(48)=
2.06, p < 0.05). At the daily level, repeated measures analyses showed that steps taken was
significantly correlated with both general (rs .14-.17, ps .01) and PA-specific self-expansion
(rs .40-.49, ps < .001) across the 4-weeks. This data implies that self-expansion (particularly
PA-specific self-expansion) may be a promising avenue to pursue for promoting PA.
Additionally, the association with day-to-day variation of PA suggests that self-expansion may
be well suited for real-time interventions.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1317

CORRESPONDING AUTHOR: Xiaomeng (Mona) Xu, PhD, Idaho State University, Pocatello, ID,
83201; [email protected]
S1318 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B242 6:00 PM-7:00 PM

URBAN YOUTHS PHYSICAL ACTIVITY EXPERIENCE, PHYSICAL ACTIVITY LEVELS, AND USE OF
SCREEN BASED MEDIA DURING LEISURE TIME

Hui Xie, Ph.D1, Linda Caldwell, PhD2, Jason Scott, Ph.D.3

1
California State University Northridge, Northridge, CA; 2Penn State, University Park, PA;
3
University of Tennessee-Knoxville, Knoxville, TN

Urban Youths Physical Activity Experience, Physical Activity Levels, and Use of Screen Based
Media During Leisure Time

Introduction: Physical activity (PA) and sedentary behaviors are two important factors
associated with childhood obesity. However, there is no guarantee that participation in PA will
lead to decreased sedentary behaviors. If youth do not obtain positive experiences from PA,
they may seek desirable experiences from sedentary activities, implicating the need for
studying youths affective responses to PA. Therefore, this study examined the relationships
between youths positive PA experience (LTPA-E), PA level (LTPA-L), and use screen-based
media (LTUS) during leisure time.

Methods: The sample consists of 305 7th and 8th graders (47% female) from 4 schools in an
urban, under-resourced area in the US. 90% to 95% of students in those schools were African
or Latino Americans. LTPA-L was measured by asking participants to recall their LTPA in the
last 7 days via 18 questions that addressed 3 domains: walking, jogging, and other physical
activity. LTPA-E was measured by 5 items (e.g., When you are doing physically active things in
your free time, how often do you enjoy it). Items were aggregated into 3 parcels for LTPA-E
and LTPA-L. Three types of LTUS were included: watching TV/movie, using computer/Internet,
and non-active videogame. Two questions (one for weekday use and the other for weekend
use) were used to measure each type of LTUS and were aggregated. Multiple-group structural
equation models were used to examine the relationships between LTPA-E, LTPA-LE and each
type of LTUS for males and females.

Results: The measurement models had a good fit (non-significant 2; RMSEA = .045 to .051)
and were equivalent across genders. The path coefficients were also equivalent between two
genders except for the relationship of LTPA-E to LTUS-videogame. LTPA-E had a significant
positive effect on LTPA-L. LTPA-L in turn had a significant effect on LTUS-TV/movie, but not on
LTUS-computer/Internet and LTUS-videogame. LTPA-E directly had a significant negative
effect on LTUS-videogame (only for males) and LTUS-computer/Internet.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1319

Conclusion: LTPA-E predicts LTPA-L, and has an influence on LTUS-computer/Internet and

LTUS-videogame that is independent of LTPA-L. The effect of LTPA-E on LTUS-TV/movie is
mediated by LTPA-L. The findings suggest PA experience has an important role in promoting
PA and thwarting use of screen based media among youth.

CORRESPONDING AUTHOR: Hui Xie, Ph.D, California State University Northridge, Northridge,
CA, 91330-8269; [email protected]
S1320 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B243 6:00 PM-7:00 PM

VARIABILITY IN SLEEP AND PHYSICAL ACTIVITY RELATED TO BMI AMONG HEALTHY ADULTS

Stephanie A. Hooker, M.S., M.P.H.1, Laura C. Bouchard, M.S.2, Kelly Glazer. Baron, PhD, MPH,
CBSM3

1
University of Colorado Denver, Denver, CO; 2University of Miami, Coral Gables, FL; 3Rush
University Medical Center, Chicago, IL

Objectives: Variability in health behavior is receiving increased attention as a marker of health

and health outcomes. However, little is known about the relationships between variabilities
among different health behaviors and how these relate to measures of health, such as body
mass index (BMI). This study assessed the relationship between variability in sleep and
physical activity (PA) and tested the relationship between variability and BMI in a sample of
healthy participants.

Method: Healthy participants with at least 6.5 hours habitual sleep duration were recruited
from the greater Chicago area (N=75) as part of a larger study of circadian timing and health.
Participants wore Actiwatch Spectrum wrist bands and SenseWear arm bands for 7
consecutive days to provide daily data on sleep duration (hours per night) and PA (minutes >3
METS). Linear regression analyses were used to assess the relationships among sleep and PA
variabilities as well as BMI. Analyses controlled for age and sex.

Results: Two thirds of this sample were female (66%) with an average age of 26.8 years
(SD=7.6). Participants had an average sleep duration of 6.7 hours (SD=1.5) and engaged in PA
for an average of 147.2 minutes (SD=87.8) per day. Average BMI was 23.5 (SD=4.2). Sleep
variability was not related to PA variability (p=0.71). Greater sleep variability was related to
shorter sleep duration (r=-0.30, p=0.002). In contrast, greater PA variability was related to
greater PA duration (r=0.73, ps>0.18). There was a significant interaction between PA
variability and average PA related to BMI (p=0.021). At low levels of average PA, greater PA
variability was related to lower BMI, whereas lower PA variability was related to greater BMI.
At high levels of average PA, variability was unrelated to BMI.

Conclusion: Among healthy adults, variability was related to total levels of each behavior but
variabilities in sleep and PA were not correlated. Results suggest that PA variability may be
more related to health among those who are less active. Further research is needed to clarify
relationships among variability in health behaviors in relation to cardiometabolic outcomes,
such as BMI.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1321

CORRESPONDING AUTHOR: Stephanie A. Hooker, M.S., M.P.H., University of Colorado Denver,

Denver, CO, 80217; [email protected]
S1322 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B244 6:00 PM-7:00 PM

WALK AND ROLL TO SCHOOL

Carolyn C. Johnson, PhD MS, John Marmion, MPH

Tulane University School of Public Health and Tropical Medicine, New Orleans, LA

Background: Safe Routes to School (SRTS) is a federally-funded program partnering with states
for the purpose of increasing community health by making walking and biking to schools safe
for the nations children. SRTS Louisiana funded a Tulane Prevention Research Center project
in partnership with Bike Easy.

Purpose: To implement and evaluate safety education classes for elementary schoolchildren,
assess attitudes of parents about children walking and biking to school, develop a crossing
guard manual and provide safety equipment to school crossing guards for the purpose of
increasing the number of children biking and walking to school.

Method: Bike Easy provided two one-hour courses Safe Steppers and Bicycling Skills 123to
4th and 5th graders in 10 New Orleans public schools. Classes were interactive with practice
with bikes and helmets. Post-surveys were administered. Brief surveys were distributed to
parents via the children. The crossing guard manual was developed through a literature
search and review of manuals from other states.

Results: A total 335 4th and 5th graders completed the bike quiz and 332 completed the
walking quiz. The bike training was the more successful with 15 of 19 classes obtaining a class
mean of >70% correct responses, while only 5 of 19 classes achieved the same for the walking
quiz. Parental response was low (14%). More than 50% of those responding lived more than 2
miles from school and child transport was via school bus or private vehicle (bus=65%;
car=24%; walk=10%; public transit=1%). Generally, parents were not in favor of children
walking or biking to school with distance to school being the most frequently identified barrier
(53%), and sidewalk quality (20%), traffic (16%), and time (14%) also being cited as barriers.
The crossing guard manual has been completed and is under review by the state.

Conclusions: Although the education classes were interactive and fun, learning from the
walking class was limited and, though better for the biking class, was still not optimal. While
these education classes probably will not contribute to increased walking and biking to
schools because of the attitudes and barriers cited by parents, improving the safety of
children wherever they bike and walk is a worthwhile public health effort.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1323

CORRESPONDING AUTHOR: Carolyn C. Johnson, PhD MS, Tulane University School of Public
Health and Tropical Medicine, New Orleans, LA, 70112; [email protected]
S1324 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B245 6:00 PM-7:00 PM

FAMILY COHESION, BURDEN, AND HEALTH-RELATED QUALITY OF LIFE AMONG PARKINSONS

DISEASE CAREGIVERS IN MEXICO

Stephen K. Trapp, PhD1, Melissa M. Ertl, BA2, Susana Gonzalez Arredondo, Lic3, Yaneth
Rodriguez Agudelo, Dr4, Juan Carlos C. Arango-Lasprilla, PhD5

1
University of Utah, Salt Lake City, UT; 2University at Albany, State University of New York,
East Greenbush, NY; 3Student, Mxico, Distrito Federal, Mexico; 4Instituto Nacional de
Neurologia, mexico, Distrito Federal, Mexico; 5Ikerbasque. Basque Foundation for Science.,
Bizkaia, Pais Vasco, Spain

Family cohesion is a known predictor of health-related quality of life (HRQOL) for family
caregivers (CG) of someone with Parkinsons disease. Unfortunately, CG tend to experience
numerous psychosocial stressors that can lead to CG burden and diminished HRQOL. Although
past research illustrates these relations within samples in the US, little is known about the role
of CG burden on the association between family cohesion and CG HRQOL in Mexico. The
present study examines whether CG burden mediates the relation between family cohesion
and HRQOL among Parkinsons disease CG in Mexico.

Method

The sample consists of family caregivers (N = 102; Mage = 50.34, SD = 14.02; 77% women, 23%
men) of patients with Parkinsons disease from Mexico City, Mexico.

Outcome measures of the study include aggregate factors of mental and physical HRQOL from
the SF-36 Health-Related Quality of Life (Ware, Snow, Kosinski, & Gandek, 2000). The Family
Adaptability and Cohesion Evaluation Scale-Fourth Edition (FACES-IV; Olson, 2010) assessed
family cohesion, and the Zarit Burden Inventory (Zarit, Orr, & Zarit, 1985) measured perceived
burden.

Results

Perceived burden was hypothesized to mediate the relation between family cohesion and
mental and physical HRQOL among caregivers. Results indicated that family cohesion
significantly predicted perceived burden (b = -5.77, SE = 1.47, p = .0002 for the mental HRQOL
model and b = -5.90, SE = 1.80, p = .002 for the physical HRQOL model), and that perceived
burden was a significant predictor of mental (b = -.80, SE = .13, p < .0001) and physical HRQOL
(b = -.73, SE = .13, p < .0001). Family cohesion, which significantly predicted mental HRQOL (b
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1325

= 6.98, SE = 2.23, p = .002), was no longer a significant predictor of mental HRQOL after
controlling for perceived burden (b = 2.38, SE = 2.05, p = .25), consistent with full mediation.
In terms of physical HRQOL, family cohesion remained a significant predictor, b = 6.88, SE =
2.31, p = .004, providing evidence of partial mediation. Approximately 14% of variability in
mental HRQOL (R2 = .14) and 10% in physical HRQOL (R2 = .10) were accounted for by the
predictors. Indirect effects were tested using bootstrap estimation with 1000 samples.
Indirect coefficients were significant for mental (b = 4.60, SE = 1.56, 95%CI = 1.99, 8.04) and
physical HRQOL (b = 4.33, SE = .2.04, 95%CI = 1.00, 8.92).

Conclusion

Identifying contributing factors to improved HRQOL among Parkinsons CG is critical. Family

CG are vulnerable to experiencing health deterioration. Present study findings provide
support that caregiver burden is an intermediary variable for the relations between family
cohesion and HRQOL among Mexican CG. Understanding these factors provides an
opportunity to intervene and improve HRQOL of CG in this area. Important cultural
implications will be discussed.

CORRESPONDING AUTHOR: Stephen K. Trapp, PhD, University of Utah, Salt Lake City, UT,
84132; [email protected]
S1326 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B246 6:00 PM-7:00 PM

INSTILLING THE ART OF HAPPINESS IN COMMUNITY DWELLING OLDER ADULTS

Elizabeth M. Orsega-Smith, PhD, Steve Goodwin, PhD, Katie Greenawalt, BS, Jennie Turner,
BS, Erica Rathie, Student

University of Delaware, Newark, DE

The population of older adults is increasing exponentially as we know that by 2030 older
adults will comprise 20% of the population. As the US population ages, late life depression
becomes more of a public health problem. Depression is associated with increased mortality
and decreased physical, cognitive and social functioning, which can impact the ability to
maintain independence which is important to overall well-being.

Methods:

The Art of Happiness is an 8-week intervention based on the teachings of the Dalai Lama with
the goals of enhancing happiness levels and overall mental well-being in older adults. The
pilot program was conducted at 3 senior centers with 90 minute classes that examined a
different weekly topic including; (1) defining happiness, (2) stress management, (3) reflecting
on happiness, (4) compassion and human connection, (5) forgiveness, (6) transforming
suffering, (7) mindfulness, and (8) humor. Participants were provided with The Art of
Happiness by the Dalai Lama as well as a journal to complete homework assignments. Each
session included lecture and discussion portions with homework that focused on that weeks
topic. Pre and post questionnaires were used to assess participant gratitude, life satisfaction,
depression, stress, mindfulness, arousal states, subjective happiness, and general
demographic and health information.

Results:

The 31 participants who completed the course were mostly married (43.8%), female (87.5%)
and Caucasian (91%) with an age range of 53-93 years of age (mean age= 74 years). Paired t-
tests (pre and post program measures) showed significant improvements in participants
subjective happiness (p

Conclusion:

This suggests that it is possible to influence the older adults frame of mind in terms of
outlook on life and improving quality of life. These types of programs which influence positive
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1327

mental health in the older adult population may be a mechanism by which to improve the
overall well- being in the older adult population

CORRESPONDING AUTHOR: Elizabeth M. Orsega-Smith, PhD, University of Delaware, Newark,

DE, 19716; [email protected]
S1328 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B247 6:00 PM-7:00 PM

PHYSICAL ACTIVITY, SELF-REPORTED PHYSICAL FUNCTIONING AND QUALITY OF LIFE IN

STROKE SURVIVORS

Erin Snapp, MS1, Bradley J. Kendall, B.A., M.S.2, Neha Gothe, MA, MS, PhD3

1
Wayne State University, Dearborn, MI; 2Wayne State University, Grosse Pointe Park, MI;
3
Wayne State University, Detroit, MI

Background

Stroke is the third leading cause of death in the United States and the leading cause of
disability among its survivors. Recent reviews in this population have shown that regular
physical activity (PA) can lead to improved quality of life (QOL), functional capacity and
mobility among stroke survivors. The goal of this study was to examine the relationship
between PA levels in stroke survivors and their satisfaction with life (SWL) and overall physical
functioning (PF).

Methods

Participants consisted of 31 stroke survivors who were at least 6 months post-stroke (mean
age:61.9711.17). PA was measured using an Actigraph accelerometer that was worn by
participants during waking hours for 7 days. Using NHANES cut-offs, PA data was categorized
by counts per minute as sedentary (0-100), light (101-2019), and moderate to vigorous (MV;
2020). The 36 Item Short Form Health Survey (SF-36) was used assess overall PF and the
Satisfaction with Life Scale (SWLS) was used to asses perceived SWL.

Results

Correlation and regression analyses showed that light PA was correlated with SWLS
(r=.404, p =.03, =.40) and SF-36 PF (r=.454, p =.01, =.45). MVPA was not significantly
correlated with SWLS (p=.12) but was significantly correlated with SF-36 PF (r=.49, p =.009,
=.53). Sedentary time was not significantly correlated with either self-report measure.

Conclusion

Light PA consistently predicted self-reported PF and SWL in stroke survivors. MVPA was
predictive of PF but comparable with light PA which may be more effective to adopt and
maintain for this population. Future research should examine light PA based interventions
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1329

such as stretching and strengthening, yoga, tai-chi etc., and their efficacy in improving QOL
among stroke survivors.

CORRESPONDING AUTHOR: Erin Snapp, MS, Wayne State University, Dearborn, MI, 48124;
[email protected]
S1330 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B248 6:00 PM-7:00 PM

THE INFLUENCE OF DIET, SLEEP AND PHYSICAL ACTIVITY ON WELL-BEING - PERCEPTIONS AND
BEHAVIORS

Sandra Jane. Winter, PhD, MHA1, Tia Rich, Ph.D., M.A., M.S.W.2, Erin Avery, M.S., B.S.3, Haley
Hedlin, PhD4, Vidhya Balasubramanian, MS4, Catherine Heaney, PhD, MPH5

1
Stanford University School of Medicine, Palo Alto, CA; 2Stanford School of Medicine,
Stanford, CA; 3Stanford Prevention Research Center, Santa Cruz, CA; 4Quantitative Sciences
Unit, Department of Medicine, Stanford University, Palo Alto, CA; 5Stanford Prevention
Research Center, Stanford, CA

Background: The association between health behaviors and chronic diseases is well
documented. Less well researched is the association between health behaviors and perceived
wellbeing. This study examined the perceived influence of diet, sleep and physical activity on
wellbeing and associations between self-reported behaviors and perceived wellbeing.

Methods: Participants were mainly older, well-educated women (n=222, mean age=59.20.7,
years of education=16.70.2, women=90%) who reported good, very good or excellent health
(88.3%) and wellbeing (91%). For the previous two weeks, participants reported: 1) the
perceived influence of diet, sleep and physical activity on their wellbeing (none, positive or
negative influence) and 2) measures of these behaviors.

Results: The reported influences of these behaviors on wellbeing were: diet : none=48%,
positive=27%, negative=26%; sleep: none=38%, positive=27%, negative=35%; physical
activity: none=64%, positive=16%, negative=20%. Comparing the groups of participants who
reported no influence, positive influence or negative influence of these behaviors on their
wellbeing, there were statistically significant differences in the reported behaviors by group.
Those reporting a negative influence of these behaviors on wellbeing also reported eating
more desserts and processed food daily and sleeping fewer minutes per night compared to
those reporting no or a positive influence of these behaviors on wellbeing: mean servings of
desserts/day (none=1.61.9, positive=11.3; negative=3.14.1; p

Conclusion: More than half of the participants in this study reported that diet and sleep
positively or negatively influenced their wellbeing, and over one third reported that physical
activity positively or negatively influenced their wellbeing. There was no association between
reported physical activity and the influence of physical activity on wellbeing, but eating more
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1331

servings of desserts and processed foods per day and sleeping fewer minutes per night were
negatively associated with perceived wellbeing.

CORRESPONDING AUTHOR: Sandra Jane. Winter, PhD, MHA, Stanford University School of
Medicine, Palo Alto, CA, 94304-1334; [email protected]
S1332 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B249 6:00 PM-7:00 PM

WELLNESS IN WOMEN: LONG TERM FREEDOM FROM A DIAGNOSIS OF INVASIVE BREAST

CANCER

Winnifred B. Cutler, Ph.D.

Athena Institute for Women's Wellness, Chester Springs, PA

Overestimates of health risks can impair decision-making and paradoxically reduce wellness.
Records of 2,305,427 screened-symptomless women were used to estimate their risk of being
diagnosed with invasive breast cancer over their next 25 years, and determine the percentage
of women actually likely to remain free of invasive breast cancer.

Prior estimates of breast cancer incidence started with Diseased Women from cancer
registries and used census population estimates. Our study intentionally did neither, thereby
increasing accuracy in determining the likelihood of a womans living free of breast cancer.

Method: Our systematic review identified 19 peer-reviewed published studies of 2,305,427

women meeting 5 stringent criteria. Incidence of first invasive breast cancer was estimated
using logistic regression based on follow-up duration

Results: Over 25 years of follow-up, an estimated 94.55% of women will remain breast
cancer-free (95% CI: 93.97, 95.13). The mean cumulative incidence rate of first invasive
breast cancer increased by 0.2% for each year of follow-up (95% CI: 0.17, 0.23; p < 0.01; R2 =
0.90). There was no evidence of an age-related increase; but there was evidence of a higher
rate for those who became menopausal through surgery.

Conclusion: The vast majority (99.75%) of asymptomatic peri and postmenopausal women will
not be diagnosed with invasive breast cancer each year, and 95% will live well free of a
diagnosis of invasive breast cancer. For those who avoid the 7 known risk factors for being
diagnosed with breast cancer an even higher percentage will live free of invasive breast
cancer.

Weight gain, inadequate exercise, excess alcohol, and inadequate solar radiation generating
vitamin D deficiencies are the 4 major risks factors well documented to increase the risk of
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1333

breast cancer. Hormonal declines following menopause and several specific hormonal therapy
regimens are also associated with increased risk. Converging evidence suggests that we
should now add mammography itself to our list of hazards.

Ever more sensitive scanning is increasingly likely to detect very early stage cancers, 30-50%
of which we now know would be self-limiting and benign. However, once detected, such
cancers usually prompt treatment and expose women to unnecessary iatrogenic harms.

The apparent success of screening depends heavily on inclusion of cases where treatment was
directed at breast cancer that was most likely to never develop to clinical significance. Like
prostate cancer, as many as 50% of women develop non-life threatening breast cancer before
they die. Overestimates of ones risk of developing deadly invasive cancer fuel this excessive
fear and detrimentally drive womens decision to undergo regular mammography screenings.
A broad convergence of recent research shows that these decisions paradoxically compromise
wellness.

CORRESPONDING AUTHOR: Winnifred B. Cutler, Ph.D. , Athena Institute for Women's

Wellness, Chester Springs, PA, 19425; [email protected]
S1334 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B250 6:00 PM-7:00 PM

ACCULTURATION MODERATES THE EFFECTS OF POSITIVE EMOTIONALITY ON THE CORTISOL

AWAKENING RESPONSE IN MEXICAN AMERICAN ADULTS

Antonio Garcia, M.A., Deborah Mangold, Ph.D.

The University of Texas at San Antonio, San Antonio, TX

Mexican Americans face unique acculturation-related stressors that affect the neurohormonal
stress response. Previous research has shown that Mexican Americans with greater Anglo
orientation evidence an attenuated cortisol awakening response (CAR) and that stress
associated with the acculturation process may render some Mexican Americans more
susceptible to adverse health outcomes. The tendency to experience positive emotions is a
personality trait with particular relevance for examinations of resilience to the adverse effects
of stress in this underserved minority group. Numerous studies have demonstrated a strong
relationship between stress-induced alterations of the hypothalamic-pituitary-adrenal (HPA)
axis and certain negative health outcomes, and suggest that extraversion is an important
factor promoting psychological adaptation and resilience to the adverse effects of stress.
Additional evidence suggests that highly resilient individuals proactively cultivate positive
emotionality, a key facet of extraversion, by strategically eliciting positive emotions through
the use of humor, relaxation techniques, and optimistic thinking. However, to our knowledge,
no studies have examined the relationship between the tendency to experience positive
emotions and the CAR in Mexican Americans.

In the current study, the effects of positive emotionality and acculturation on the CAR were
examined in adults of Mexican descent. Salivary cortisol samples were collected at awakening
and 30, 45, and 60 minutes thereafter, on two consecutive weekdays from 89 healthy
Mexican American adult males (43) and females (46), ages 18 to 38 years. Results of general
linear mixed modeling (GLMM) showed that lower positive emotionality was associated with
an attenuated CAR. Stronger Anglo orientation was also associated with an attenuated CAR.
Follow-up analyses showed that in Mexican Americans with strong Mexican orientation,
positive emotionality was associated with a robust CAR, whereas in Mexican Americans with
moderate or strong Anglo orientation, CAR was attenuated, and positive emotionality showed
no effect. The current findings suggest that in less acculturated Mexican Americans with
strong connections to traditional cultural values and behaviors, positive emotionality
functions as a buffer against stress-induced alterations in the CAR. However, in more
acculturated Mexican Americans with fewer connections to cultural values and traditions, the
tendency to experience positive emotions may not be an effective buffer against stress-
induced alterations in the CAR. It is possible that participation in cultural traditions and
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1335

behaviors facilitates a social milieu responsive to positive emotionality, fostering resilience to

stress-induced alterations in the CAR.

CORRESPONDING AUTHOR: Antonio Garcia, M.A., The University of Texas at San Antonio, San
Antonio, TX, 78230; [email protected]
S1336 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B251 6:00 PM-7:00 PM

HISPANIC ETHNICITY AND PRIMARY CARE UTILIZATION AMONG MILLENNIALS

Hannah Kimmel, BA, MPH '17

Brown University School of Public Health, Providence, RI

Background: Hispanics are the largest ethnic minority in America, and experience racial health
disparities on outcomes for various health measures. Using a primary care provider minimizes
some of these outcome disparities. Persons aged 18-34 years old (Millennials) occupy an
age range more likely to be uninsured than older and younger individuals. Implementation of
the Affordable Care Act (ACA) in 2010 sought to increase preventive primary care access and
utilization. There exists a gap in the literature of Millennial primary care utilization by ethnicity
after the ACA.

Objective: This study examines the characteristics of primary care utilization among
Millennials, exploring differences in this utilization by Millennial ethnicity.

Methods: Data are collected from the 2011-2012 National Health and Nutrition Examination
Survey (NHANES). This study accounts for complex survey design and uses bivariate analysis
and multivariate logistic regression to assess the relationship between ethnicity and routine
health care utilization habits. Primary care utilization is defined as routine care provided at a
doctors office, clinic, or Health Maintenance Organization (HMO).

Results: Among Millennials (N=1,794), Hispanics (N=397) have 1.75 times (95% CI 1.25, 2.46)
the odds of not utilizing primary care compared to Non-Hispanic Whites. This relationship is
not significant after adjusting for gender, citizenship status, health insurance status, and
education level (OR 1.04; 95% CI 0.731, 1.48). Hispanic Millennials are 2.63 times (95% CI 1.64,
4.21) more likely to be uninsured than Non-Hispanic Millennials holding citizenship status,
gender, and education level constant.

Conclusion: The unadjusted results of Millennial routine care utilization habits show Hispanic
Millennials are 1.75 times as likely to not use primary care when compared to Non-Hispanic
White Millennials. When adjusting for insurance utilization, this difference is no longer
significant. This reinforces previous findings of health care utilization and access disparities by
race and ethnicity, even among Millennials. This could explain the difference in primary care
utilization among these two groups.

CORRESPONDING AUTHOR: Hannah Kimmel, BA, MPH '17, Brown University School of Public
Health, Providence, RI, 09206; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1337

B252 6:00 PM-7:00 PM

OUTCOMES OF ELEMENTARY AND MIDDLE SCHOOL REFUSERS IN JAPAN

Mari Saito, M.D.1, Alan Lefor, Ph.D.2, Yutaka Kikuchi, M.D.1

1
Haga Red Cross Hospital, Mouka, Tochigi, Japan; 2Department of surgery, Jichi Medical
University, Shimotsuke, Tochigi, Japan

Background: Japans modern education system is built on the principle that every child has
equal educational opportunity. All schools are the same so every child gets the same
education. However, children who are unwilling to attend school have few options to
continue their education. In 2015, the Education Ministry reported that school refusers
represent at least 1.5% of elementary and middle school aged children, but detailed data is
unavailable. We provided behavioral approaches to school refusers and aim to describe the
outcomes of elementary and middle school aged school refusers who received therapy at our
institution.

Methods: Data were extracted from medical records of children with insurance codes
including school refusal, psychogenic reaction or somatoform disorders. Data were
analyzed if they underwent psychological therapy because of prolonged physical symptoms
related to school activities or because of prolonged school refusal.

Results: A total of 118 children (boys, n=53) underwent therapy at a mean age of 12y. The
most frequent concomitant physical symptoms were headache and digestive symptoms
(nausea, vomiting, abdominal pain) related to school activities. The common trigger for school
refusal was fear of other classmates, such as being talked about behind their back or being
excluded from class activities. Of these, 31 (26%) were diagnosed with borderline mental
retardation or a developmental disorder including autism spectrum disorder, attention deficit
hyperactivity disorder and Asperger syndrome. Half of children improved after a mean of 10
months of cognitive-behavioral therapy and resolving problems with classmates. Twenty-
three (21%) children quit psychological therapy at an average of 7 months and 87% had
continued difficulty attending school. Children who quit therapy tended to be absent from
school for longer periods, be totally absent from school and have a previous history of
stopping therapy.

Conclusion: Most school refusers are triggered by difficulties with other classmates. They had
headaches and digestive symptoms related to school activities, however, one-half had a good
response to psychological therapy within a year. School refusers who did not have a good
S1338 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

response after 7 months of therapy, had decreased compliance with appointments and poor
outcomes.

CORRESPONDING AUTHOR: Mari Saito, M.D., Haga Red Cross Hospital, Mouka, Tochigi, 321-
4306; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1339

B253 6:00 PM-7:00 PM

SOCIOCULTURAL AND ENVIRONMENTAL INFLUENCES ON BRAZILIAN MOTHERS BELIEFS &

PRACTICES RELATED TO CHILD FEEDING & WEIGHT STATUS

Ana Cristina. Lindsay, DDS, MPH, Dr.PH1, Sherrie Wallington, PhD2, Mary L. Greaney, PhD3,
Maria Helena Hasselmann, RN, PhD4, Marcia Maria Tavares Machado, N/A5

1
University of Massachusetts Boston, Boston, MA; 2Georgetown University Medical Center,
Lombardi Comprehensive Cancer Center, Rockville, MD; 3University of Rhode Island,
Kingstom, RI; 4Institute of Nutrition, Rio de Janeiro, Rio de Janeiro, Brazil; 5PROFESSOR OF
UNIVERSITY FEDERAL OF CEARA, FORTALEZA, N/A, Brazil

Background: Length of residence in the United States (US), changes in dietary and physical
activity behaviors, and economic and social barriers contribute to high childhood obesity rates
among children from immigrant families in the US. Brazilians comprise a fast-growing
immigrant population group in the US, yet little research has focused on health issues
affecting Brazilian children in immigrant families. Understanding sociocultural and
environmental influences on parents beliefs and practices related to child feeding and weight
status is essential to altering obesity trends in this group.

Methods: Qualitative study consisting of five focus groups with a convenience sample of 29
Brazilian immigrant mothers.

Results: Analyses revealed that the sociocultural and environment transitions faced by
Brazilian immigrant mothers influence their beliefs and practices related to child feeding and
weight status. Additionally, acculturation emerged as a factor affecting mothers feeding
practices and their childrens eating habits, with mothers preferring Brazilian food
environments and that their children preferring American food environments. Mothers
viewed themselves as being responsible for promoting and maintaining their childrens
healthy eating and feeding behaviors, but changes in their social and cultural environments
due to immigration and the pressures and demands of raising a family in a new country make
this difficult.

Conclusions: Health promotion interventions to improve healthful eating and feeding

practices of Brazilian children in immigrant families must account for social and cultural
changes and daily life demands due to immigration as well as potential variation in the levels
of acculturation between mothers and their children.
S1340 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Ana Cristina. Lindsay, DDS, MPH, Dr.PH, University of

Massachusetts Boston, Boston, MA, 02125; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1341

B254 6:00 PM-7:00 PM

MERITORIOUS AWARD WINNER

THE ROLE OF RABBIS IN DECISION-MAKING ABOUT BRCA TESTING AMONG ORTHODOX

JEWISH WOMEN

Haeseung Yi, MPA1, Meghna Trivedi, MD2, Rivka Starck, PsyD3, Alejandro Vanegas, MS2,
Raven David, MS2, Isaac Schechter, PsyD3, Rita Kukafka, DrPH, MA2, Katherine Crew, MD, MS2

1
Columbia University, Edgewater, NJ; 2Columbia University Medical Center, New York, NY;
3
Institute for Applied Research & Community Collaboration, Monsey, NY

BACKGROUND The prevalence of founder mutations in the BRCA1 and BRCA2 genes is 1 in 40
among Ashkenazi Jews. Despite the potential benefits of BRCA testing to inform cancer
screening and risk-reducing surgeries, there are still concerns about adverse psychological and
social consequences, which may vary by cultural and religious backgrounds. Orthodox Jews
often consult with Rabbinic and communal authorities in medical decision-making, which is
consistent with their religious values. The aim of this study was to examine whether Rabbis
have a role in medical decision-making about BRCA genetic testing among Orthodox Jewish
women.
METHODS We applied a mixed-methods approach to analyze quantitative survey data and
qualitative transcripts. We distributed a survey via community-based and religious email
listservs in Upper Manhattan in New York City. We collected data on demographics, breast
cancer risk factors and religious/cultural factors influencing a decision to undergo BRCA
genetic testing. Survey participants were re-contacted to participate in two focus groups
separated into single (n=8) and married (n=10) women. The sessions were audio-recorded and
transcribed. For the qualitative approach, a content analysis was conducted using Atlas.ti
software. The Scotts Pi value for the inter-rater reliability of two coders was 0.7091.
RESULTS Of 505 Orthodox Jewish women invited to participate in the survey study, 269 (53%)
completed the survey. The median age was 25 years (range, 19-86) and 67% identified as
Modern Orthodox. Sixty-six percent stated that they were moderately to very likely to consult
a Rabbi to make a significant medical decision. However, less than half (46%) thought that
their Rabbi was moderately to very medically informed. When making an important medical
decision, 88% replied that they were moderately to very likely to seek advice from a primary
doctor/family physician and 43% said that they would consult with a Rabbi or Rebetzin
(Rabbis wife). Themes that emerged from qualitative analysis confirmed the results we
obtained from the survey. While some participants said that they would consult a Rabbi, some
preferred discussing health decisions with their doctors. By analyzing the transcripts, we
S1342 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

found that Orthodox Jewish women tend to seek medical advice from Rabbis especially for
the halachic implications (Jewish law or code of ethics).
CONCLUSION A key finding from our study is that Rabbis and health professionals influence
medical decisions about genetic testing in the Orthodox Jewish population. Therefore,
educating Rabbis on the medical aspects of BRCA genetic testing, including cancer risk and risk
reduction strategies, will be crucial. We believe that education and decision support tools for
both Rabbis and health professionals will be important to disseminate information on BRCA
testing in the Orthodox Jewish community.

CORRESPONDING AUTHOR: Haeseung Yi, MPA, Columbia University, Edgewater, NJ, 07020;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1343

B255 6:00 PM-7:00 PM

UNDERGRADUATES DESIGNATED ORGAN DONOR STATUS ON THEIR DRIVERS LICENSE: AGE,

RACE/ETHNICITY, GENDER, AND RELIGION MATTER

Cheryl Albright, PhD, MPH1, Kara Saiki, MPH2, Lynne Wilkens, DrPh2, Ashley M. Biddle, MA2,
David Devine, MS2, Misty Pacheco, DrPH3, Pamela Smith, APRN, RNFA, FNP-BC4

1
Univeristy of Hawaii at Manoa, Honolulu, HI; 2University of Hawaii, Honolulu, HI; 3University
of Hawaii at Hilo, Hilo, HI; 4Hawaii Pacific University, Kaneohe, HI

On average, 22 people die each day waiting for an organ transplant and over half of the
119,967 people on the U.S. transplant waiting lists are ethnic minorities. Specifically, the
proportion of Asian, Native Hawaiian, Pacific Islanders (ANHPI), on waiting lists exceeds their
representation in the general population; thus, the need for ANHPI organ donors is high. The
iDecide Hawaii Project was designed to educate ANHPI undergraduates about the decision to
be a Designated Organ Donor (DOD) on their next (often an adult 21+ years of age) drivers
license. The project used a gamified online eHealth intervention (e.g., Instagram, YouTube) to
both educate and engage undergraduates in issues surrounding DOD status and incentivized
student organizations to be DOD advocates via pledge contests with cash prizes. From 2014-
2016, 2,138 students completed baseline surveys before the intervention was launched
(Mage= 22.8 years 5; 76% female; 15% Hispanic, 3% African American, 4% American
Indian/Alaska Native, 32% Asian, 17% Native Hawaiian/Pacific Islander, 31% White, and 9%
multiracial (largely Asian/White). One-third (35%) were Protestants, 24 % Catholic, and 28%
were not religious or were Agnostics/Atheists. Of those who had a drivers license, 59.6%
were already a DOD; but, DOD rates were significantly lower for students with a < 21year
license (57%) versus those with 21+ year license (63%) (p = 0.006). There were also differences
by age with 58% of 18-24y/o, 59% of 25-29y/o, and 68% of 30+ y/o being a DOD (p = 0.0001).
Significantly more females (62%) were DODs than males (54%; p = 0.003). There were
significant differences by race with 74% of Whites, 56% of Native Hawaiians/Pacific Islanders,
44% of Asians, and 65% of multiracial persons reporting being a DOD on their license
(p=0.0001). There were significant DOD differences within Asians: Filipinos= 38%,
Koreans=62%, Japanese=63%, Asian/White multiracial= 70% (p=0.001). There were significant
differences by religion with 53% of Protestants, 46% Catholics, and 62% not
religious/Agnostics reporting being DODs on license (p=0.0001). Thus, prevalence of DOD was
higher for students who were older, female, not religious, or White; but, within Asians, DOD
prevalence ranged from 38% to 70%. Changes in knowledge, attitudes and
S1344 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

actions/engagement of both DODs and non-DODs exposed to our eHealth, gamified

intervention will be assessed via post-test surveys. Moderator analyses will determine any
differential impact by age, race, gender or religion. Engagement in the gamified eHealth
intervention varied by site and contest; but, hundreds of undergraduates entered the DOD
pledge contests and the Instagram picture contests that followed an organ donation-related
theme. Examples of contest winners will be displayed.

CORRESPONDING AUTHOR: Cheryl Albright, PhD, MPH, Univeristy of Hawaii at Manoa,

Honolulu, HI, 95822; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1345

B256 6:00 PM-7:00 PM

ASSOCIATIONS BETWEEN PHYSICAL ACTIVITY, SEDENTARY BEHAVIOR, AND SLEEP HABITS

AMONG AN ATLANTIC CANADIAN COHORT

Cynthia C. Forbes, PhD1, Vanessa DeClercq, PhD2, Melanie Keats, PhD1

1
Dalhousie University, Halifax, NS, Canada; 2Atlantic PATH, Halifax, NS, Canada

Background: Lack of sleep and poor sleep quality has been linked to many poor physical and
mental health outcomes. Engaging in regular PA has been linked to changes in sleep circadian
rhythms, thermogenic regulation, body weight, physical fitness, anxiety, depression, and pain;
any of which could be a possible mechanism by which PA impacts sleep. The goal of this study
was to determine any associations between physical activity or sedentary behaviour (SB) and
sleep duration or quality among a cohort of Atlantic Canadians.

Methods: 31,173 participants, aged 35-69 years, living in Atlantic Canada (Nova Scotia, New
Brunswick, Newfoundland & Labrador, and Prince Edward Island), completed a series of
assessments. Domain-specific physical activity and sedentary behaviour was assessed using
the International Physical Activity Questionnaire-Long Form (IPAQ). Sleep duration and sleep
quality were measured with two items; first asking participants to report in hours and minutes
On average, how many hours per day do you usually sleep, including naps? and then, How
often do you have trouble going to sleep or staying asleep? on a 5-point scale from never
to all the time. Objective measures were used in Physical Activity Level (PAL) calculations if
available; in cases where they were not, self-report measures were used.

Results: 18,258 participants had both PA measures and sleep data. PAL computation revealed
59.7% of the sample was inactive with a further 14.5% in the low active category. 25.7% of
the sample were considered active or very active. Participants averaged 5.8 hours of SB per
day with most of the sample sedentary between 4-7 hours per day (37.7%). Most (65.5%)
reported getting between 7-9 hours of sleep per night with an actual average of 7.2 hours per
night. Most of the sample reported having sleep troubles some of the time or less frequently
(77.5%). As sleep duration increased, the percentage of those inactive also increased. U-
shaped trends emerged with sleep disturbances either none of the time or all of the time at
higher levels than some of the time among those classified as very active. An inverse linear
relationship was found for sleep quality and SB.

Conclusion: As sleep duration increased, so did the proportion of those classified as inactive.
This is contrary to research that shows a u-shaped relationship between sleep duration and
PA. Results also showed that a higher proportion of those who slept more than 11 hours and
S1346 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

had sleep disturbances all the time were those who were sedentary for more than 7 hours.
There is less research on SBs link to sleep quality but what literature there is indicates a link
similar to PA. More research is needed to confirm a link between SB and sleep duration and
quality.

CORRESPONDING AUTHOR: Cynthia C. Forbes, PhD, Dalhousie University, Halifax, NS, B3H
1V7; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1347

B257 6:00 PM-7:00 PM

ASSOCIATIONS OF CHRONOTYPE WITH ALCOHOL AND MARIJUANA USE IN A NATIONAL

SAMPLE OF U.S. HIGH SCHOOL STUDENTS

Denise Haynie, PhD, MPH1, Daniel Lewin, Ph.D.2, Kaigang Li, PhD3, Leah Lipsky, PhD4, Danping
Liu, PhD1, Joseph Bible, PhD1, Ronald J. Iannotti, PhD5, Bruce Simons-Morton, EdD, MPH4

1
National Institute of Child Health and Human Development, Bethesda, MD; 2Children's
National Health System, Washington, DC; 3Colorado State University, Fort Collins, CO; 4NICHD,
Bethesda, MD; 5The CDM Group, Inc, Bethesda, MD

Changes typifying adolescence have profound effects on adolescent sleep patterns with
implications for health and well-being. Research has demonstrated relations between poor
sleep quality and increased alcohol use. Preferred sleep period, or chronotype, has emerged
as potentially useful in understanding this association. This study examined adolescents
chronotype in association with alcohol and marijuana use.

Data come from the NEXT Generation Study, a national cohort of US 10 thgrade students
(2009-10). Latent class analysis using data from the first 3 years of the study (W1-W3)
identified chronotype classes by preference for 1) earlier sleep period (21%), 2) later sleep
period (16%) and 3) middle sleep period (61%). W3 weekday and weekend sleep durations
were categorized as short (< 7 hours per night), inadequate (7-8 hours), optimal (8-9.5 hours)
and long (> 9.5 hours). Logistic regression models examined the associations of chronotype
class and W3 sleep duration with W3 30-day alcohol and marijuana use. Models adjusted for
design variables, gender, race/ethnicity, and W2 substance use.

Participants (n=2168) were 59% (weighted) female, 62% white, 13% African American, and
19% Hispanic. 30-day reports of W3 alcohol use were: 38% any use, 27% binge drinking, 25%
drunk, and 7% blackout from alcohol; 19% reported any W3 30-day marijuana use. In adjusted
logistic regressions, compared to members of the mid-type class, the early chronotype class
had lower odds of drinking (odds ratio=0.48, 95%CI=0.32-0.71), binging (OR=0.44, CI=0.29-
0.67) or being drunk (OR=.0.37, CI=0.23-0.61); the late chronotype class had higher odds of
binging (OR=1.87, CI=1.17-2.99) and being drunk (OR=1.68, C=1.06-2.67). Odds of marijuana
use were lower for those reporting long sleep duration on unscheduled days
(OR=0.44,CI=0.29-0.67); marijuana use was not associated with chronotype.

Three distinct classes of chronotypes were identified in US adolescents, with late chronotype
associated with alcohol use but not marijuana use, which was positively associated instead
S1348 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

with long sleep duration. Research is needed to differentiate the psychosocial, behavioral and
biological causes of associations between chronotype, sleep duration and substance use.

CORRESPONDING AUTHOR: Denise Haynie, PhD, MPH, National Institute of Child Health and
Human Development, Bethesda, MD, 20817; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1349

B258 6:00 PM-7:00 PM

ASSOCIATIONS OF MULTIPLE SLEEP PROBLEMS WITH AFFECTIVE AND COGNITIVE

VULNERABILITIES

Catherine Rochefort, MS1, Erin Koffel, PhD2, Michael Chmielewski, PhD1

1
Southern Methodist University, Dallas, TX; 2Minneapolis VA Health Care System, Mineapolis,
MN

Recently, there have been calls for additional research examining the combined influence of
cognitive and affective factors on sleep problems. Past research has demonstrated that
increased rumination is associated with poor sleep, but less research has examined other
cognitive vulnerabilities. Furthermore, although studies have examined the influence of
depression/anxiety (as broad classes) on sleep, very little research has examined the
association of specific symptoms of depression/anxiety with sleep problems. This is an
important limitation because depression and anxiety are heterogeneous constructs comprised
of multiple symptoms, each of which might demonstrate differential associations with sleep
problems. In this cross-sectional study, college students (N = 611) completed measures
assessing a variety of sleep problems (sleep quality [WHIIRS], daytime sleepiness [ESS], and
insomnia [IDAS]), cognitive vulnerabilities (trait rumination [RSQ], automatic thoughts [ATQ],
fear of negative evaluation [BFNE], dysfunctional attitudes [DAS]), and ten symptoms of
depression and anxiety (IDAS, with items referencing sleep removed). Results indicated that
cognitive vulnerabilities (including rumination) demonstrated limited incremental validity over
symptoms of depression and anxiety for all sleep problems (R2 = .00 - .02). In contrast, the
depression and anxiety symptoms demonstrated significant (p < .01) incremental validity over
and above all of the cognitive vulnerabilities for each sleep problem (WHIIRS: R2 = .10; ESS
R2 = .09, IDAS R2 = .35). Interestingly, each sleep problem was predicted by different
symptoms of depression and anxiety, indicating unique associations between specific types of
sleep problems and specific symptoms. Given that between 15 56% of the variance in sleep
problems was accounted for by symptoms of depression and anxiety, our results suggest that
they are more important for sleep problems than cognitive vulnerabilities, which appear to
have limited influence across the included sleep problems. Because quality sleep is integral to
physical and mental health in clinical and nonclinical populations, sleep interventions may
need to place more emphasis on improving affect.

CORRESPONDING AUTHOR: Catherine Rochefort, MS, Southern Methodist University, Dallas,

TX, 75206; [email protected]
S1350 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B259 6:00 PM-7:00 PM

INFANT SLEEP TRAINING AND MATERNAL HEALTH BEHAVIORS

Megan Clarke, Ph.D.1, Amy A. Gorin, Ph.D.2, Jennifer Selensky, B.A.3

1
University of Connecticut, West Hartford, CT; 2University of Connecticut; Institution for
Collaboration on Health, Intervention, and Policy, Storrs, CT; 3University of Connecticut,
Storrs, CT

Background: Women are vulnerable to excess weight gain and retention during their
childbearing years. This study explored the relationships between infant sleep training,
maternal physical activity levels, and weight loss during the first postpartum year by
comparing mothers who reported successfully sleep training their infants with those who did
not. Group differences were examined in 1) quantity of maternal sleep, 2) quantity of infant
sleep, 3) maternal physical activity and weight loss, 4) psychosocial variables including
postpartum depression and perceived social support and 5) mothers preference for potential
weight loss and infant sleep training interventions.

Methods: Mothers of 6-12 month old infants (N=150, Mage=32.24, MBMI=27.07) were
recruited with study flyers both online and in the community. Participants completed
questionnaires online, including item 8 from the Paffenbarger Activity Questionnaire, the
Postpartum Depression Screening Scale, and the Brief Infant Sleep Questionnaire.
Participants also answered questions about how they had engaged with various sleep training
methods with their infants. Two groups were established: Sleep Trainers were those who
endorsed successfully sleep training their infants, and Non-Sleep Trainers who did not
endorse the item. Data was analyzed using chi-square and independent t-tests.

Results: Sleep Trainers slept more than Non-sleep Trainers ( p < .05) and woke up significantly
fewer times overnight (p < .01). The children of Sleep Trainers also slept significantly more
overnight (p < .01), and their mothers reported higher levels of perceived social support ( p <
.05). There were no significant group differences in physical activity levels, but differences in
weight retention approached significance (p = .10) with mothers who successfully sleep
trained indicating that they were closer to their prepregancy weight. Mothers in both groups
endorsed the highest level of interest in online interventions. Non-Sleep Trainers were
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1351

significantly more likely to express interest in potential sleep training interventions, with most
expressing interest in an online format.

Discussion: These results suggest that interventions targeting postpartum weight retention
should take the mother-infant dyad into account in study design. Infant sleep modification
presents a possible target for intervention with potential benefits for both mothers and
infants.

CORRESPONDING AUTHOR: Megan Clarke, Ph.D., University of Connecticut, West Hartford,

CT, 06107; [email protected]
S1352 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B260 6:00 PM-7:00 PM

INFLUENCES OF INSOMNIA SUBTYPES ON SUBSEQUENT ANXIETY AND DEPRESSION: A

POPULATION-BASED LONGITUDINAL COHORT STUDY

Ping-Jen Chen, MD1, Shih-Feng Weng, PHD2, Ming-Ping Wu, MD, PHD3, Chung-Han Ho, PHD3,
Charles Lung-Cheng Huang, MD3, Jhi-Joung Wang, MD, PHD3, Ya-Wen Hsu, PHD, MPH4

1
Chi-Mei Medical Center, Tainan, Taiwan, Tainan, Tainan, Taiwan (Republic of China);
2
Kaohsiung Medical University/Department of Healthcare Administration and Medical
Informatics, Kaohsiung, Kaohsiung, Taiwan (Republic of China); 3Chi-Mei Medical Center,
Tainan, Tainan, Taiwan (Republic of China); 4Chia Nan University of Pharmacy & Science,
Tainan, Tainan, Taiwan (Republic of China)

Background: Though prior studies have investigated the associations between insomnia and
mental health, research assessing the causal relationships using population-based study with
long-term follow-up is still lacking.

Objective: To examine the causal influences of insomnia on the onset of depression only,
anxiety only, depression or anxiety, and depression and anxiety during the 4-year follow-up.

Methods: Participants were identified from the National Health Insurance enrollees in Taiwan
during 2002~2009. The study included 19,273 subjects with insomnia (52.116.0 years old)
and 38,546 matched subjects without insomnia (51.216.0 years old).

Results: Compared with non-insomniacs, insomniacs had a higher risk of developing anxiety
only [adjusted hazard ratio (HR) =8.83, 95% CI=7.59-10.27)], depression only (adjusted
HR=8.48, 95% CI=6.92-10.39), either anxiety or depression (adjusted HR=9.90, 95% CI=8.83-
11.1), as well as both anxiety and depression (adjusted HR=17.98, 95% CI=12.65-25.56). When
breaking down into insomnia subgroups, individuals with a relapse of insomnia (adjusted
HR=10.42-26.80) had the highest risk of anxiety only, depression only, either anxiety or
depression, and both anxiety and depression, followed by persistent insomnia (adjusted
HR=9.82-18.98) than remission of insomnia (adjusted HR=4.50-8.27) in comparison to the
non-insomniacs. When comparing the 4-year cumulative incidence between insomniacs to
non-insomniacs for anxiety only, depression only, either anxiety or depression, and both
anxiety and depression, we found that the relapse group (p-value for log-rank test
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1353

Conclusions: Insomnia is a strong predictor for developing depression and anxiety. Patterns of
insomnia should be monitored to help treating and controlling subsequent psychiatric
disorders.

CORRESPONDING AUTHOR: Ping-Jen Chen, MD, Chi-Mei Medical Center, Tainan, Taiwan,
Tainan, Tainan, 710; [email protected]
S1354 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B261 6:00 PM-7:00 PM

SLEEP AND HEALTH BEHAVIORS AMONG AFRICAN AMERICAN CONGREGATION MEMBERS

Natasha Aduloju-Ajijola, PhD, MPH, CHES1, Jannette Berkley-Patton, PhD2, Carole Bowe
Thompson, BS3, Kelsey Christensen, MA4, Alexandria G. Booker, M.A.5, Sheila Lister, B.S.4, Nia
Johnson, B.A.6, Therese Ruhland-Petty, MA7

1
University of Missouri-Kansas City School of Medicine, Kansas City, MO; 2University of
Missouri-Kansas City, Kansas City, MS; 3Community Health Ressearch Group/School of
Medicine/University of Missouri-Kansas City, Kansas City, MO; 4University of Missouri-Kansas
City, Kansas City, MO; 5University of Missouri-Kansas City, Raytown, MO; 6University of
Missouri- Kansas City, Kansas City, MO; 7University of Missouri- Kansas City, Des Moines, IA

Sleep disorders are often comorbid with a broad range of medical and psychiatric conditions
that have a negative impact on health, mood, and quality of life. African Americans (AAs)
report higher rates of extreme sleep behaviors, including hypersomnia and hyposomnia, and
are at an increased risk for related consequences, including cardiovascular disease, diabetes
mellitus, and respiratory issues. Research has demonstrated that there is a reciprocal
relationship between sleep and these health issuesthus, health conditions impact sleep and
sleep subsequently impacts health outcomes. The Black church is an integral part of the AA
community, as AAs have the highest rate of church attendance compared to all other racial
and ethnic groups in the United States. The Black church has the ability to address health
concerns affecting members, as well as their communities through outreach ministries.
However, previous studies have not examined the sleep behaviors of AA church-affiliated
persons. The purpose of the current study was to examine the demographic, social, and
health factors associated with lack of sleep among AA church-going populations. This study
used baseline data from Project FIT (Faith Influencing Transformation), a health promotion
intervention piloted in AA faith-based settings. The study participants (N = 352) were
predominantly female (n = 240, 69.4%), with an average age of 53.6 years (SD = 13.1). Results
showed that 30% of participants had less than 6 hours of sleep a night, 70% received 6 to 10
hours of sleep a night, and 1% received over 10 hours of sleep a night. Findings indicated that
income (r=.159, p=.008), stress levels (r=-.163, p=.007), and gender (r=.140, p=.017) were
significantly correlated with sleep. In addition, diagnoses of heart disease (r=-099 p=.067) and
polycystic ovarian syndrome (r=.160, p=.008); previous screening for mental health (r=-.159,
p=.018) and level of education completed (r=.176, p=.004) were also significant sleep
predictors. Linear regression analyses revealed that income ( =.155, p=.020), polycystic
ovarian syndrome ( =.165, p=.012), stress levels ( =-.120, p=.068), and gender ( =.133,
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1355

p=.039) together were able to predict sleep behavior. Sleep behaviors and their subsequent
impact on health should be further explored among the African American church-going
populations.

CORRESPONDING AUTHOR: Natasha Aduloju-Ajijola, PhD, MPH, CHES, University of Missouri-

Kansas City School of Medicine, Kansas City, MO, 64110; [email protected]
S1356 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B262 6:00 PM-7:00 PM

SLEEP PROBLEMS IN A LARGE SAMPLE OF CANCER PATIENTS BEFORE INITIATING TREATMENT

Eric S. Zhou, PhD1, Karen Clark, MS, MS2, Christopher Recklitis, PhD. MPH3, Richard A.
Obenchain, N/A4, Matthew Loscalzo, LCSW5

1
Dana-Farber Cancer Institute, Boston, MA; 2City of Hope, Carlsbad, CA; 3Dana-Farber Cancer
Institute/Harvard Medical School, Boston, MA; 4City of Hope National Medical Center,
Pasadena, CA; 5City Of Hope-National Medical Center, duarte, CA

Background: Sleep problems are common in adult cancer patients and have significant health
consequences. However, few studies have examined multiple disease sites, or captured sleep
function at the critical time post-diagnosis, but prior to initiating treatment. This study sought
to address these limitations by exploring the prevalence of sleep problems and its correlates
in a large sample of cancer patients.

Methods: 5,702 cancer patients (67.5% female; 76.9% White, 17.2% Asian, 5.5% Black)
completed a routine electronic distress screen, including a question on sleep problems,
following diagnosis and prior to initiating cancer treatment. Patients across 8 different cancer
diagnoses (breast, gastrointestinal, gynecological, head and neck, hematological, lung,
prostate, urinary) were included. Patients endorsing Severe or Very Severe sleep
problems-related distress were classified as having a sleep problem.

Results: 12.5% of patients reported sleep problems, though significant differences in the
proportion of patients with sleep problems were seen across diagnoses (4.3-13.8%; p < .001),
with prostate cancer patients least likely and gastrointestinal cancer patients most likely to
experience sleep problems. Logistic regression analyses indicated demographic factors (lower
household income and female gender; not race, education, or marital status) were associated
with greater prevalence of sleep problems. A logistic model adjusted for demographic
variables revealed anxiety and depression, but not concerns about potential treatment-
related side effects, were associated with sleep problems (p < .001).
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1357

Conclusions/Implications: A sizeable proportion of cancer patients already report major

problems with their sleep, before any treatment has occurred, and this is an issue across racial
groups and educational levels. Key demographic and mood factors are associated with sleep
problems. This can have important implications for identifying and treating patients at
greatest risk for sleep problems during a vulnerable period early in their cancer trajectory.

CORRESPONDING AUTHOR: Eric S. Zhou, PhD, Dana-Farber Cancer Institute, Boston, MA,
02215; [email protected]
S1358 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B263 6:00 PM-7:00 PM

SLEEP, RESILIENCE, AND PSYCHIATRIC DISTRESS IN UNITED STATES MILITARY VETERANS

Jaime M. Hughes, PhD, MPH, MSW1, Christi S. Ulmer, PhD2, Susan N. Hastings, MD, MHS2,
Jennifer M. Gierisch, PhD, MPH3, Mid-Atlantic VA MIRECC Workgroup, .4, Matthew O.
Howard, PhD5

1
Center for Health Services Research in Primary Care, Durham VA Health Care System, Cary,
NC; 2Center for Health Services Research in Primary Care, Durham VA Health Care System,
Durham, NC; 3Duke Univeristy, Durham, NC; 4Durham VA Health Care System, Durham, NC;
5
UNC School of Social Work, Chapel Hill, NC

Sleep problems are a common symptom of several mental health conditions, including
depression, PTSD, and anxiety. However, roughly two-thirds of Veterans without an active
mental health diagnosis also complain of poor sleep. Left untreated, sleep problems may be
associated with psychiatric distress that could contribute to a subsequent mental health
diagnosis. Few studies have examined the clinical correlates of sleep problems in Veterans
without current mental health diagnoses. Identifying risk and protective factors related to
poor sleep and increased psychiatric distress could strengthen pertinent prevention efforts.
Research suggests psychological resilience, defined as positive stress-coping ability, protects
against poor mental health outcomes in Veterans exposed to stress or trauma. However, the
relationship between sleep and resilience has not been previously studied. Utilizing data from
the Study of Post-Deployment Mental Health, a volunteer research registry housed at the
Mid-Atlantic VAs Mental Illness Research, Education, and Clinical Center, this research
examined the demographic, health, sleep and resilience characteristics of post-9/11 military
Veterans with one or more deployments who did not meet past-month criteria for a DSM-
defined Axis I mental health disorder (N=1,117). Hierarchical linear regression and moderation
analyses were used to examine the relationship between poor sleep quality (Pittsburgh Sleep
Quality Index) and psychiatric distress (Global Symptom Index, Symptom Checklist-90) and to
test whether psychological resilience (Connor-Davidson Resilience Scale) moderated this
relationship. Results revealed Veterans with poor sleep quality had worse physical and
psychological health, lower resilience, and endorsed more lifetime traumatic events.
Controlling for health and demographic characteristics, poor sleep quality was associated with
greater psychiatric distress [F (5, 1111) = 84.7, R2=0.159, p < 0.0001]. Both factors of resilience
adaptability and self-efficacy were found to have a significant buffering effect between
poor sleep quality and greater psychiatric distress (ps < 0.05). Additional research is needed
to better understand the relationship between sleep and resilience in Veterans, and how
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1359

interventions targeting each may help to curb psychiatric distress and ultimately reduce risk of
subsequent mental health diagnoses.

CORRESPONDING AUTHOR: Jaime M. Hughes, PhD, MPH, MSW, Center for Health Services
Research in Primary Care, Durham VA Health Care System, Cary, NC, 27519;
[email protected]
S1360 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B264 6:00 PM-7:00 PM

ARE THERE RACIAL DISPARITIES IN THE SOCIAL INTEGRATION/MORTALITY ASSOCIATION?

Steven D. Barger, PhD1, Bert N. Uchino, PhD2

1
Northern Arizona University, Flagstaff, AZ; 2University of Utah, Salt Lake City, UT

Objective: Social integration, as indicated by being married, having recent contacts with other
people, participating in group activities, etc., is strongly and inversely associated with
mortality. However, there is little data evaluating whether this association is consistent across
racial groups. Method: In a representative US sample we examined the presence and form of
the association of social integration with all-cause mortality within non-Hispanic Blacks (n =
4,201) and non-Hispanic Whites (n = 20,217). Vital status was ascertained ten years following
the baseline assessment. Social integration scores ranged from zero to eight. The two lowest
categories were combined and this single social integration variable was entered as a
predictor in Cox regression models. We statistically controlled for age, sex, chronic disease
prevalence, received emotional support, socioeconomic status (education, employment,
home ownership) and smoking. Results: After multivariate adjustment social integration was
inversely related to ten-year survival for both groups. Among Whites each unit increase in
social integration was associated with a 10% reduced risk of dying over ten years, hazard ratio
[HR] = 0.90, 95% confidence interval [CI], 0.87-0.92. A similar association was observed among
Blacks, HR = 0.92, 95% CI, 0.86-0.98. There was no evidence of heterogeneity across the two
hazard estimates (z = -0.64, p = 0.52). Conclusions: These data provide evidence that the
social integration/mortality association is evident for both Whites and Blacks.

CORRESPONDING AUTHOR: Steven D. Barger, PhD, Northern Arizona University, Flagstaff, AZ,
86011; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1361

B265 6:00 PM-7:00 PM

CHRONIC DISEASE PREVENTION AND TREATMENT AMONG MEXICAN MIGRANTS: FINDINGS

FROM A BINATIONAL PROJECT

Xiao Zhang, PhD1, Gudelia Rangel, PhD2, Eduardo Gonzalez-Fagoaga, PhD3, Melbourne Hovell,
PhD4, Sylvia Guendelman, PhD5, Ana Martinez-Donate, PhD6

1
University of Wisconsin-Madison, Verona, WI; 2El Colegio de la Frontera Norte, San Antonio
del mar, Baja California, Mexico; 3Comisin de salud fronteriza mxico- estados unidos, San
Diego, CA; 4San Diego State University, San Diego, CA; 5University of California-Berkeley,
Berkeley, CA; 6Drexel University, Philadelphia, PA

Background: Circular Mexican migrants represent a socioeconomically disadvantaged

population and are more likely to experience limited access to health care in both the U.S. and
Mexico.The situation may worsen if the new administration implements stringent immigration
and health policy reforms that target immigrants. Guided by the Andersen health care
utilization model, this study aims to 1) examine the levels of preventive and treatment
services receipt for chronic diseases, including hypertension, hypercholesterolemia and
diabetes, across the migration phases; and 2) identify predisposing, enabling, and need factors
that were associated with the receipt of preventive and treatment services for each migration
phase, among Mexican migrants that converged in the border city of Tijuana, Mexico.

Methods: We used data from a 2013 cross-sectional survey of Mexican migrants conducted as
part of Project Migrante, a binational collaboration between the U.S. and Mexico (N=2115).
Based on spatial trajectories and migration history, Mexican migrants were categorized into
five migration phases: pre-departure, return, transit, destination, and interception. We
conducted multiple logistic regressions to model receipt of preventive and treatment services
for chronic diseases within each migration phase and to compare receipt levels across
migration phases.

Results: Overall, 23.0% of Mexican migrants represented by our sample self-reported at least
one type of chronic diseases. The age- and gender-standardized prevalence rates of any
preventive services receipt ranged from 23.0% among return migrants to 29.2% among transit
migrants; and the standardized rates of any treatment services receipt among those with any
chronic diseases ranged from 39.5% among migrants in the interception phase to 54.5%
among those in the destination phase. After statistical adjustment for other factors,
destination (AOR=1.49) and interception migrants (AOR=1.84) were more likely to receive
preventive services than their pre-departure counterparts. In contrast, receipt of treatment
S1362 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

services did not vary significantly across migration phases. Different sets of predisposing,
need, and enabling factors were associated with the receipt of preventive and treatment
services for different migration phases.

Conclusion: Mexican migrants had suboptimal levels of receipt of preventive and treatment
services for chronic diseases across all migration phases. Tailored intervention programs
should be provided to Mexican migrants at different migration phases to increase health care
receipt and reduce the burden of chronic diseases.

CORRESPONDING AUTHOR: Xiao Zhang, PhD, University of Wisconsin-Madison, Verona, WI,

53726; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1363

B266 6:00 PM-7:00 PM

COUPLE EMOTION REGULATION SKILLS AND PARTNER SUPPORT OF PHYSICAL ACTIVITY IN

ADULTS WITH OSTEOARTHRITIS

Cynthia M. Khan, PhD1, Christine Rini, PhD2, Robert F. DeVellis, Ph.D.2, Ida J. Griesemer, BA1,
Kent M. Lee, MA2, Mary Altpeter, PhD, MSW, MPA3, Leigh F. Callahan, PhD4

1
University of North Carolina at Chapel Hill, Carrboro, NC; 2University of North Carolina at
Chapel Hill, Chapel Hill, NC; 3UNC at Chapel Hill, Chapel Hill, NC; 4University of North Carolina,
Chapel Hill, Chapel Hill, NC

Physical activity is crucial to management of osteoarthritis (OA). Even small increases in

activity can improve health and well-being. Getting social support for physical activity (e.g.,
encouragement, working together) from a partner is a robust predictor of physical activity,
making it a potentially useful target for couples-focused interventions to promote OA
management. Yet, social support involves complex, skill-based interpersonal interactions, so
partner support for physical activity is most likely to be beneficial when both members of a
couple have adequate support-related skills. We propose that benefits of partner support
depend on the emotion regulation skills of both members of couples (i.e., ability to
understand their own and each others emotions and respond adaptively to discordance). In
169 couples that included a person with OA (PWOA) who was insufficiently active and wanted
to increase activity and that persons partner (Partner), we investigated associations between
couples emotion clarity (i.e., their ability to understand their own emotions) and partner
support received by PWOA after a couples-focused OA and physical activity class. We
hypothesized that the association between PWOA emotion clarity and partner support would
be moderated by partner emotion clarity. After controlling for PWOA gender, findings from
hierarchical linear regression analyses partially supported our hypothesis: PWOA emotion
clarity was positively associated with positive partner support for physical activity ( = .25, p =
.029) when partners had better emotion clarity. However, when partners had poorer emotion
clarity, PWOA emotion clarity was not associated with positive partner support ( = -.10, p =
.349). PWOA emotion clarity was negatively associated with partner unsupportive behaviors
(e.g., criticism, minimizing the importance of physical activity; = -.20, p = .010), but this
association was not moderated by partner emotion clarity ( = .02, p = .842). Findings suggest
that efforts to leverage partner support to promote PWOA physical activity may be optimized
by helping each couple member attain or improve emotion regulation skills such as emotion
clarity.
S1364 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Cynthia M. Khan, PhD, University of North Carolina at Chapel Hill,
Carrboro, NC, 27510; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1365

B267 6:00 PM-7:00 PM

FORGIVENESS & GRATITUDE VS. SHAME DUE TO HETEROSEXISM & DEPRESSION IN OLDER
GAY MEN AND LESBIANS

Tosha Griggs, Student

Center for Psychosocial Health Research, Lake Dallas, TX

Older lesbians and gay men (LG) are at greater risk for depression and shame due to
heterosexism (SDH) because of a lifetime of victimization (Fredriksen-Goldsen et al., 2013).
SDH is associated with negative mental health outcomes, is positively correlated with age and
more prevalent in older LG that live alone (Meyer, 1995; Grossman, D'augelli & O'connell,
2002). However, forgiveness of others may help older LG combat depression and shame
because forgiveness is empowering (Cass, 1984). According to Cass (1984) forgiveness of
others is associated with self-acceptance and lower mental distress (Kim & Enright, 2014).
Forgiveness of others is also associated with increased gratitude and greater life satisfaction
(McCullough, Tsang & Emmons, 2004). Gratitude is associated with positive mental health
(Snyder & McCullough, 2000) and lower depression (Musick & Wilson, 2003). Through a
minority stress (Meyer, 2003) lens, we hypothesize that SDH is positively associated with
depression, forgiveness of others and gratitude are negatively associated with depression and
that SDH, forgiveness of others and gratitude account for a significant proportion of variance
in depression. Participants (Mage = 58.99, SD= 6.48) were 50% gay men, 50% lesbians and self-
identified as 68% European American, 17% African American and 15% other ethnicity. A
hierarchical regression analysis (F [10, 89] = 7.12, p < .001) indicated that, in aadition to
gratitude (=-.34, p < .001) and forgiveness of others (=-.22, p < .05), being single (=.42, p <
.001) was also a significant predictor, accounting for approximately 38% of the variance in
depression (adjusted R=.38, p < .001).
Our findings suggest that depression is positively associated with SDH and negatively
associated with gratitude and forgiveness of others. SDH is negatively associated with
gratitude and forgiveness of others. Being single is negatively associated with forgiveness of
others and positively associated with SDH. Because depression in older adults is associated
with a decrease in physical, social and cognitive function and increased mortality (Blazer,
2003), therapeutic priorities should focus on the aspects of life to be thankful for, while
encouraging older LG to use forgiveness as a coping strategy to counteract SDH and
depression.

CORRESPONDING AUTHOR: Tosha Griggs, Student, Center for Psychosocial Health Research,
Lake Dallas, TX, 75065; [email protected]
S1366 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B268 6:00 PM-7:00 PM

GENDER DIFFERENCES: IMPACT OF VIOLENCE EXPOSURE ON DRUG USE AND HIV RISK
BEHAVIORS AMONG YOUNG ADULT AFRICAN AMERICANS

Forough Saadatmand, PhD1, Roderick Harrison, PhD2, Jennifer Bronson, PhD3, Deanna Crouse,
MHS4

1
Howard University, College of Medicine, Department of Pediatrics and Child Health,
Washington, DC, N/A; 22M Research Services, LLC, Reston, VA; 3Office of Justice Program /
Department of Justice, Laurel, MD; 4Howard University / College of Medicine / Dept. of
Pediatrics and Child Health, Washington, DC, DC

BACKGROUND: This presentation provides baseline preliminary analysis of data collected in a

longitudinal study from 440 African Americans young adults ages 18-25 living in Washington
DC. OBJECTIVE: One component of the study is designed to test the effects of social and
environmental stressors such as exposure to community violence (including homicide) on
alcohol, tobacco and other drug (ATOD) use and sexual risk behaviors that may render African
American young adults more vulnerable to contracting HIV/AIDS. The objective here is to
understand how exposure to violence impacts ATOD use and HIV risk behaviors, and to
determine if the strengths of the associations vary based on violence type and gender.
METHODS: We first conducted a factor analysis of the exposure to community violence scales
to identify domains or types of violence. We then conducted a series of regression analyses to
assess the various community violence types (e.g., witnessing a murder) on the dependent
variables representing ATOD use, and knowledge and perception of HIV risk factors. RESULTS:
Our data showed that as children, 33% of the participants reported seeing a parent get hit by
another parent or a boyfriend/girlfriend, and 42% said someone close to them was murdered.
As adults, 66% have seen other people using or selling drugs in real life, and 36% have
actually been asked to get involved in aspects of selling or distributing illegal drug. Women
and men differed in the types of violence to which they were likely to have been exposed.
Women were significantly more likely to have been exposed to personal violence and attacks,
which explained a mean of 75 percent of the total variance in lifetime ATOD use and age of
first ATOD use for women. Men, in contrast, were more likely to have been direct victims of
violence. Experiencing gun use or seeing violent deaths explained 51.6 percent of the total
variance for men and 69 percent of the total variance in the model for male marijuana use.
CONCLUSION: The results suggest that a large percentage of young African American adults
are exposed to community violence, that community violence has a strong impact on ATOD
use and HIV risks behaviors, and that African American women and men are exposed to
different kind of community violence.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1367

NOTE: "The views here do not necessarily represent the Bureau of Justice Statistics or the
Department of Justice "

Funding: This research is funded by NIH/National Institute on Minority Health and Health
Disparities grant # 4R01MD005851-05.

CORRESPONDING AUTHOR: Forough Saadatmand, PhD, Howard University, College of

Medicine, Department of Pediatrics and Child Health, Washington, DC, N/A, 20001;
[email protected]
S1368 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B269 6:00 PM-7:00 PM

GREEN SPACES AND GROCERY STORES: URBAN ENVIRONMENTAL DETERMINANTS OF WEIGHT

AMONG MEXICAN-AMERICAN YOUTH

Dorothy McLeod, MA, Amy Bohnert, PhD, Catherine DeCarlo Santiago, PhD

Loyola University Chicago, Chicago, IL

Latino children have the highest rates of obesity of any racial/ethnic group in the U.S. (Ogden,
Kit, Carrol, & Flegal, 2014), and are therefore at increased risk for the many negative sequelae
of child obesity. Several studies have demonstrated that neighborhood factors, such as the
distance to or density of grocery stores or other food outlets, as well as the distance to or
density of green spaces and parks, are associated with obesity and related risk factors among
Latino youth (e.g. Chaparro et al., 2014, Hsieh et al., 2014). Studies of these relations among
Latino youth have suggested that these factors may be differentially impactful by gender (e.g.,
Hsieh et al., 2014). This study will examine these neighborhood factors and their impact on
obesity among Mexican-origin immigrant youth, and whether these effects have differential
significance for girls and boys.

Data were collected in the homes of 104 Mexican-origin immigrant families with a child aged
6-11 years (M=8.91, SD=1.34), in which at least one parent was a first-generation immigrant
and whose household income fell at or below 150% of the poverty line. A sample of 89
participants with valid zBMI data was used for initial analyses. Parents provided child/family
demographic information. Anthropometric measurements were used to calculate child zBMI
based on CDC growth charts (Kuczmarski et al., 2002). Address data was used to plot the
participants home locations in a Geographic Information System (GIS). Data derived from the
City of Chicago Public Data Portal were used to plot grocery store and park locations. ArcGIS
was used to conduct geospatial analyses of participants home locations and their relation to
the stated neighborhood factors.

There was substantial variation in child zBMI (M=1.01, SD=1.07). Overall, 44.9% of children
were normal weight, 14.6% were overweight, and 40.4% were obese, demonstrating a heavier
sample than the overall US Hispanic child population (Ogden et al., 2014). A map of
participants locations and weight statuses revealed no geographic clustering of participants
falling into the overweight or obese categories. Contrary to previous research, preliminary
analyses revealed that proximity to the nearest park was negatively associated with zBMI,
such that the greater distance between the participants home and the nearest park, the
lower the participants zBMI (=-.07, p=.01). However, there was no direct association
between proximity to grocery stores and child weight. These preliminary findings suggest that
home distance to parks and green spaces may have unique effects among this population.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1369

Future analyses will confirm this finding by examining density of parks and grocery stores
(rather than distance) and will explore the moderation effects of gender on these relations.

CORRESPONDING AUTHOR: Dorothy McLeod, MA, Loyola University Chicago, Chicago, IL,
60605; [email protected]
S1370 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B270 6:00 PM-7:00 PM

HEALTH-RELATED SELF-EFFICACY: THE ROLE OF SOCIAL SUPPORT IN CANCER, DISABILITY, AND

MULTIMORBIDITY

Kelly D. Blake, ScD1, Devlon Jackson, PhD, MPH2, Anna Gaysynsky, MPH3, Erin Kent, PhD4,
Neetu Chawla, PhD, MPH5, Danielle Blanch-Hartigan, PhD, MPH6, Lila J. Finney-Rutten, PhD,
MPH7

1
Behavioral Research Program, Division of Cancer Control and Population Sciences, National
Cancer Institute, National Institutes of Health, Bethesda, MD; 2University of Maryland, College
Park, College Park, MD; 3ICF International, Bethesda, MD; 4National Cancer Institute,
Rockville, MD; 5Kaiser Permanente Northern California, Oakland, CA; 6Bentley University,
Waltham, MA; 7Mayo Clinic, Rochester, MN

Background: Self-efficacy consistently has been shown to play a role in health behavior, and
social support is associated with health-related self-efficacy. Our study provides population
estimates of disability, multimorbidity, social support, and health-related self-efficacy in
adults ever diagnosed with cancer and adults with no cancer history, and examines the
relationship between social support and health-related self-efficacy in the context of cancer,
disability, and multimorbidity. Methods: Data were from the National Cancer Institutes
cross-sectional Health Information National Trends Survey (HINTS) 4, Cycle 2 (2012-2013).
Multivariable logistic regression models were employed to assess the probability that cancer,
cancer plus the presence of any disability, cancer plus the presence of any comorbid
condition, and social support are independently associated with two measures of health-
related self-efficacy (confidence in ability to take care of own health, confidence in ability to
get advice about cancer) in the overall sample (N=3,266) and the subsample of individuals
ever diagnosed with cancer (N=422). Results: Those with a cancer diagnosis and cancer plus a
comorbid condition were no more or less likely to be confident that they could take care of
their health or get advice or information about cancer compared to those never diagnosed
with cancer or those without cancer and multimorbidity (OR 0.93, CI 0.66-1.32 and OR 0.71, CI
0.49-1.04, respectively). In contrast, those who had been diagnosed with cancer and also
experienced a disability were significantly less likely to report being confident in their ability
to take care of their health (OR 0.30, CI 0.16-0.56). Social support was independently and
significantly associated with health-related self-efficacy: those reporting that they had some
type of social support (emotional, informational, instrumental) were more than twice as
confident on both measures of health-related self-efficacy across all health conditions under
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1371

study. Conclusion: Social support may play an important role in bolstering health-related self-
efficacy among cancer patients, particularly in the context of multimorbidity and disability.

CORRESPONDING AUTHOR: Kelly D. Blake, ScD, Behavioral Research Program, Division of

Cancer Control and Population Sciences, National Cancer Institute, National Institutes of
Health, Bethesda, MD, 20892-9671; [email protected]
S1372 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B271 6:00 PM-7:00 PM

PRESSURE ULCER DEVELOPMENT AMONG ETHNICALLY DIVERSE, COMMUNITY DWELLING

ADULTS WITH SPINAL CORD INJURY

Lucia I. Florindez, MA1, Florence Clark, Phd, OTR/L, FAOTA2, Mike Carlson, PhD2

1
University of Southern California, Burbank, CA; 2University of Southern California, Los
Angeles, CA

Background:

Medically serious pressure ulcers (MSPUs) are the most prevalent secondary complication for
individuals with a spinal cord injury (SCI), compromising physical health and quality of life.

Purpose:

The objective of this study was to identify the life circumstances that lead to MSPU
development while people with SCI participated in a lifestyle-based pressure ulcer prevention
program, and to provide recommendations for future intervention design.

Methods:

This study entailed summarizing baseline co-morbidity data and a qualitative thematic
analysis of 1800 treatment notes from the Pressure Ulcer Prevention Study (PUPS), a
randomized controlled trial evaluating the efficacy and cost-effectiveness of a multifaceted
lifestyle-based intervention in reducing incidence of MSPUs. Participants were ethnically and
socio-economically diverse community-dwelling adults with SCI recruited from a major
rehabilitation facility in Los Angeles County. Of the 83 participants enrolled into the 12-month
PUPS intervention, 25 developed an MSPU during the course of the intervention and are the
focus of this sub-study.

Results:

In total, 25 patients enrolled in the intervention developed 40 unique MSPUs. Analysis of data
from eligible PUPS participants revealed themes that captured the range of circumstances
that led to MSPU development during study participation. The six factors that describe MSPU
development are as follows: 1) circumstances unrelated to the intervention, 2) lack of
rudimentary knowledge pertaining to wound care, 3) equipment and supply issues, 4) co-
morbidities, 5) non-adherence to prescribed bed rest, and 6) passivity.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1373

Conclusions:

Many factors may contribute to the development of MSPUs in ethnically diverse adults with
SCI. These conditions, especially when compounded by limited financial resources, seemed to
have undermined the effectiveness of a multifaceted intervention program. Several
programmatic modifications have potential for increasing the interventions effectiveness for
individuals similar to those in the sample we studied, including a focus on addressing
equipment needs, and tailoring the intervention to the health literacy levels of participants.

Funding and Acknowledgements:

This work was supported by grant R01HD056267 from the National Center for Medical
Rehabilitation Research (NCMRR) within the Eunice Kennedy Shriver National Institute of Child
Health & Human Development, National Institutes of Health to the University of Southern
California. Sincere thanks to members of the PUPS Study Team.

CORRESPONDING AUTHOR: Lucia I. Florindez, MA, University of Southern California, Burbank,

CA, 91505; [email protected]
S1374 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B272 6:00 PM-7:00 PM

RELATIONSHIP BETWEEN CROSS-CULTURAL AND SEXUAL MINORITY SPECIFIC-CULTURAL

COMPETENCE: A STUDY OF HEALTHCARE PROVIDERS

Joanne G. Patterson, MSW, MPH1, Jennifer M. Jabson, PhD, MPH1, Charles Kamen, PhD, MPH2

1
University of Tennessee, Knoxville, TN; 2University of Rochester Medical Center, Rochester,
NY

Background: Cross-cultural competence (CCC) improves health care providers (HCPs) skills for
providing culturally competent healthcare to racial/ethnic minority patients. It is not known
how CCC relates to HCPs cultural competence in treating sexual minority patients (lesbian,
gay, and bisexual patients; SMP). Given the pervasive influence of homophobia and
heterosexism, associations between CCC and SMP specific cultural competence (SMPCC) may
be moderated by HCPs attitudes about SMP.

Methods: HCPs recruited from a Southeast region completed a self-report questionnaire

including measures of CCC, SMP-specific cultural competency (SMPCC), and attitudes toward
SMP. Descriptive statistics were calculated on each measure. Nested multivariable linear
regression was calculated to test for associations between CCC and SMPCC. Attitudes toward
SMP were tested as a moderator of this relationship. Analyses were adjusted for HCP age,
number of years of practice, prior sexual minority training, and gender.

Results: Mean SMPCC scores (x= 2.47, SD = 0.54) indicated low SMPCC whereas mean CCC
scores (x= 2.88, SD = 0.32) indicated average CCC. Mean attitudes toward SMP scores (x=
14.84, SD = 2.16) indicated neutral attitudes toward treating SMP. CCC was associated with
SMPCC; for every unit increase in CCC, SMPCC increased 1.72 units (b = 1.72, t(17) = 4.27, p <
.001). HCP attitudes about SMP moderated the association between CCC and SMPCC (b =
0.23, t(17) = 2.28, p = .04).

Discussion: While CCC and SMPCC were generally associated, HCPs attitudes toward SMP
moderated this association. At high CCC, HCPs evidenced high SMPCC regardless of attitudes
toward SMP. At low CCC, HCPs with negative attitudes towards SMP had significantly lower
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1375

SMPCC than HCPs with positive attitudes toward SMP. It is unclear whether negative
attitudes towards SMP exacerbate low CCC, leading to significantly lower SMPCC, or
whether positive attitudes towards SMP make providers more competent in caring for these
patients, even in the context of low CCC. Future research should examine HCPs experiences
with CCC and SMPCC, elements of CCC associated with SMP cultural competency, and how
HCPs reconcile attitudes toward SMP with CCC and provision of SMPCC.

CORRESPONDING AUTHOR: Joanne G. Patterson, MSW, MPH, University of Tennessee,

Knoxville, TN, 37912; [email protected]
S1376 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B273 6:00 PM-7:00 PM

ROLE OF RACIALIZED POLICE VIOLENCE IN REDUCING BLACKS MENTAL HEALTH:

#BADGEONBLACK CRIMES

Aerielle Allen, M.A.1, Blair T. T. Johnson, Ph.D.2, Mora A. Reinka, M.S.3

1
University of Connecticut, Cromwell, CT; 2University of Connecticut, Storrs, CT; 3University of
Connecticut, Manchester, CT

Research has shown that members of stigmatized groups experience chronically high levels of
stress and poor health due to experiences of prejudice and discrimination. Experiencing
racialized police violence and police (mis)conduct arguably raise stress levels among
marginalized individuals. While these incidents usually directly involve only a few people, they
are not isolated events. Social media and national media attention allows such events to
affect communities across the country, not just those directly involved. As such, racialized
policing events are potentially a major concern to public health, especially for those who are
most similar to the victims. Using data from the Center for Disease Controls Behavioral Risk
Factors Surveillance System, the current study examines how the Michael Brown and Eric
Garner incidents related both to Black and White individuals mental health. These events are
of particular interest because they gave national attention to the Black Lives Matter activist
movement, which has provided a platform to campaign against violence and racism targeting
Black people. Results demonstrated that in the thirty days following the shooting of Michael
Brown in Ferguson, MO, there were significant increases in the mean number of days Missouri
residents aged 18-24 and 35-44 reported stress, depression, or overall poor mental health,
compared to those polled outside the state, p=.009. Within Missouri, Black respondents also
reported an increase in the mean number of days they experienced poor mental health,
compared to Whites, p<.01. Similar trends emerged regarding the death of Eric Garner in
Staten Island, NY. Thirty days following Garners death, New York respondents aged 25-34
reported more stress, depression, or overall poor mental health compared to residents of
other states, p < .01. Although not significant, similar trends were seen for respondents not
living in New York. Mirroring the Michael Brown analyses, Black respondents in New York also
demonstrated an increase in the mean number of reported days where they experienced poor
mental health, compared to White New Yorkers, p < .01. Additional analyses examine trends
for mental health one year following each event and limitations and implications are
discussed.

CORRESPONDING AUTHOR: Aerielle Allen, M.A., University of Connecticut, Cromwell, CT,

06416; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1377

B274 6:00 PM-7:00 PM

UNDERSTANDING THE SOCIAL DETERMINANTS OF HEALTH IN UNDERSERVED COMMUNITIES

Alexa J. Meinhardt, Department of Biological Sciences, University of Delaware, Newark, DE1,

Jaclyn M. Natalone, Department of Biological Sciences, University of Delaware, Newark, DE2,
Maggie Norris Bent, MPA3, Mia A. Papas, PhD4

1
University of Delaware, Lititz, PA; 2University of Delaware, Dix Hills, NY; 3Westside Family
Healthcare, Wilmington, DE; 4Value Institute, Christiana Care Health System, Newark, DE

There is growing recognition of the strong correlation between social determinants and
community health outcomes. Despite efforts to address these issues, Delawareans living in
underserved communities continue to experience poor health outcomes. The Healthy
Neighborhoods Project, a partnership between Westside Family Healthcare and the Delaware
Center for Health Innovation, provides a formal approach for organizations to develop and
implement strategies that improve community health. As part of the community planning
process, a neighborhood-level community needs assessment was conducted in the urban
areas of Wilmington and Claymont, Delaware to identify the strengths and resources available
within specific neighborhoods. Guided by the precede-proceed model, this investigation
helped communities prioritize their health targets and determine barriers that hinder meeting
these targets. A survey was administered to residents in the communities to understand
concerns from a local perspective, and a resource library was developed that identified all
available community health promotion activities. Activities were divided into themes
including maternal/child health, chronic disease prevention/management, and mental
health/addiction. Programs were divided among these themes, with 255 locally available
programs and an additional 56 that could be accessed statewide. The locations of each
program were mapped using a Smartsheet feature that allows for visualization of programs by
theme. Community health indicators for the Wilmington and Claymont neighborhood
revealed poor health outcomes with an increased prevalence of obesity (29.7%), smoking
(18.9%) and infant mortality (6.9 per 1,000 live births). Although there are many programs
targeting the Wilmington and Claymont neighborhood, community health indicators are
poor. It is important to understand how this community utilizes health promotion programs
and determine current unmet needs. Understanding the social determinants of health from a
unique community-based perspective will aid in developing health promotion programs to
successfully meet the goals of this initiative.
S1378 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Alexa J. Meinhardt, Department of Biological Sciences, University

of Delaware, Newark, DE, University of Delaware, Lititz, PA, 17543; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1379

B275 6:00 PM-7:00 PM

DO LONELINESS AND SELF-ESTEEM ALTER THE INFLUENCE OF SELF-AFFIRMATION ON STRESS

RECOVERY?

Lauren Hofschneider, B.A.1, Mikaela Grimm, B.A.2, Hong Di Wang, Student1, Katlyn King,
Student3, Kallie Ochiai, Student1, Paula Ong, Student1, Marie Cross, B.A., M.A.1, Sarah
Pressman, Ph.D.1

1
University of California, Irvine, Irvine, CA; 2University of California, Irvine, San Juan
Capistrano, CA; 3University of California, Irvine, Newport Beach, CA

Recent research has illustrated that self-affirmation can ameliorate damaging physiological
stress responses. What has gone unstudied, however, is the possibility that some individual
differences interact with forms of affirmation to determine whether these benefits are
present or absent. Since self affirmation often focuses on esteem enhancement and affirming
on important social relationships, we examine the relationship between trait loneliness (an
important determinant of social perceptions) and self-esteem in the context of different
forms of affirmations and how these variables relate to physiological stress recovery. One
hundred and ten undergraduates (59% female, Mage=19.4) were randomly assigned to one of
three conditions: a non-social self-affirmation group (n=33) who wrote about an important
personal value, a social self-affirmation group (n=42) who wrote about an important
relationship, or a control group (n=35) who wrote about daily activities. Participants then
completed two stress tasks (a speech and a math task). Blood pressure was recorded
approximately every two minutes during the study. Trait self-esteem and loneliness were
assessed via self-report at baseline. Within the non-social self-affirmation condition, both self-
esteem, b = .494, t(28) = 2.215, p = .035, and loneliness, b = -.627, t(28) = -2.963, p = .006,
significantly predicted recovery diastolic blood pressure after controlling for age, sex, and
ethnicity. Within the social self-affirmation condition, only loneliness significantly predicted
recovery diastolic blood pressure after controlling for age, sex, and ethnicity, b = -.622, t(37) =
-2.44, p = .02. Within the control condition, neither self-esteem nor loneliness predicted
recovery diastolic blood pressure. Neither self-esteem nor loneliness predicted recovery
systolic blood pressure in any of the conditions. The different rates of diastolic blood pressure
recovery indicate that the benefits derived from self-affirmation exercises may be contingent
upon psychosocial variables. For example, individuals who report being more lonely may
benefit from affirming the existence of strong social relationships, and individuals who report
higher levels of self-esteem may benefit from affirming the existence of strong personal
values. Future research incorporating self-affirmation manipulations should focus on
S1380 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

investigating similar individual differences because they may play an important role in how
stress interventions work.

CORRESPONDING AUTHOR: Lauren Hofschneider, B.A., University of California, Irvine, Irvine,

CA, 92612; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1381

B276 6:00 PM-7:00 PM

EXPERIENCES AND IMPACTS OF STRESS RELATED TO PARENTING CHILDREN WITH LEARNING

DISABILITIES: A QUALITATIVE ANALYSIS

Emma Chad-Friedman, BA1, Karen Kuhlthau, PhD2, Jacqueline Proszynski, BA1, Thomas
Muzzonigro, BA1, Elyse R. Park, PhD, MPH1

1
Massachusetts General Hospital, Boston, MA; 2Massachusetts General Hospital, boston, MA

Background: There is growing literature on the increased levels of parental stress associated
with caring for children with learning disabilities (LD). High levels of parental stress are
associated with clinical depression, low levels of self-efficacy in caregiving, maladaptive
parenting practices, decreased social support, and poor perceived health. This qualitative
study explored the perspectives of parents of LD and experts working with these parents to
identify 1) sources and impacts of parental stress, and 2) resources available to parents; this
study also informed development of a targeted mind-body resiliency program for parents of
children with LD.

Methods: From June-July 2016 individual and focus group in-depth interviews were
conducted with 5 clinicians, 2 organizations and 4 focus groups of parents of children with LD
(N=25). Content analyses were conducted using NVivo 11 software; reliability of kappa = 0.93
was achieved.

Results: Results from expert and parent interviews demonstrated high levels of parental
stress associated with caring for children with learning disabilities. Sources of stress were
primarily navigating the educational system to advocate for the childs needs and
communicating with other parents, family members, and the child. Both parents and experts
identified impacts of stress on emotional and physical exhaustion, strained familial
relationships, and social isolation. Experts identified heightened guilt and concerns about
stigma as main stressors, whereas parents additionally revealed being burdened by feeling
misunderstood. Experts identified many informational resources available to parents, but
most parents were not aware of these resources. Moreover, no skills-based psychosocial
programs were revealed.

Conclusions: Parents caring for children with LD experience high levels of stress, particularly
affecting relationships, leading to high levels of social isolation. Parents also utilize few
emotional or informational support systems. A skills-based resiliency intervention could
provide useful skills for parents to manage their stressors and provide support to their
children and family.
S1382 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Emma Chad-Friedman, BA, Massachusetts General Hospital,

Boston, MA, 02114; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1383

B277 6:00 PM-7:00 PM

FINDINGS OF A MIND-BRAIN-BODY INTERACTION STUDY OF HYPERTENSION IN BLACK MEN

Linda Oakley, PhD

University of Wisconsin-Madison, Madison, WI

Findings of a Mind-Brain-Body Interaction Study of hypertension in Black Men. PROBLEM

Hypertension, a well characterized inflammatory illness response to the stress of racialized
social and economic inequities, is a prevalent population health problem in Black men. AIM
We assessed the interaction of mind, brain, and body factors with racial stress and blood
pressure in community Black men. METHODS Over 3 study visits, 2-3 weeks apart, a diverse
sample of right-handed, adult Black men (N = 20), age 24-64, completed surveys and were
assessed for biomarkers of stress reactivity. Participants responded to self-report measures of
daily racial micro-aggressions, civilian Post Traumatic Stress, resilience, positive thinking, daily
life stress, depression, childhood stress, incarceration, days of poor physical-mental health,
and interoception (self-awareness and self-observation of body sensations). At visit 2
Electroencephalography (EEG), using a 19-sensor cap, was used to produce 2 10-minute (eyes-
open/ closed) maps of brain electrical activity (frequency and aptitude). Following the EEG,
participants were assessed for heart rate variability (HRV) using a 3-lead computerized
electrocardiogram (ECG) program and completed measures of positive and negative affect
and daily spirituality. A brief third study visit was used to obtain a venous blood sample to test
non-fasting plasma level of high sensitivity C-reactive protein (hs-CRP), body weight, and
digital blood pressure readings. ANALYSES Structural equation modeling (SEM) is used to
estimate parameters of racial social stress related to mind-brain-body interaction and
hypertension. The brain score used in the model is the EEG estimate of Left-Right prefrontal
cortex asymmetry (F3 minus F4) or greater left (social challenge) relative to right (social
threat) reactivity to perceived social rejection. Body variables in the model are the RMSSD
(Square Root of the Mean Squared Difference between Adjacent Normal to Normal Beat
Intervals) estimate of HRV and plasma hs-CRP as an estimate of autonomic nervous system
interaction linking cognitive perception and blood pressure. RESULTS Our results present a
SEM test of mind-brain-body x environment interaction of stress related hypertension in adult
Black men. DISCUSSION Findings are discussed relative to their contributions to symptom
science and neurofeedback training in stress self-management for Black men.

CORRESPONDING AUTHOR: Linda Oakley, PhD, University of Wisconsin-Madison, Madison,

WI, 53705; [email protected]
S1384 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B278 6:00 PM-7:00 PM

HAIR BIOMARKERS AMONG MIDDLE-SCHOOL CHILDREN IN RELATION TO SCHOOL STRESS AND

ASTHMA

Eyal Oren, PhD, MS1, Erika Barrett, MPH1, Julia Fisher, MS2, Dean Billheimer, PhD1, Mark L.
Laudenslager, PhD3, Lynn B. Gerald, PhD, MSPH1

1
University of Arizona, Tucson, AZ; 2Arizona BIO5 Institute, Tucson, AZ; 3University of
Colorado Denver Anschutz Medical Campus, Aurora, CO

Background: Children and adolescents with chronic exposure to disruptive events have
prolonged physiological and health consequences. Given the extensive time spent at school, it
is possible that school social and physical environments strongly impact students through the
effects of stressors such as school-based conflict, discrimination and violence, as well as
academic and social demands. There has been little research examining the role of chronic
school stress and asthma status on hair biomarkers among low SES children and adolescents.

Methods: In this feasibility study, 52 children were recruited from a large, low socioeconomic
status, majority-Latino middle school. On two study visits, participants completed a self-report
questionnaire including perceived chronic school stressors (CASE), asthma outcomes, and
demographics and provided hair samples to test for cortisol and dehydroepiandrosterone
(DHEA). We examined the effects of predictors on cortisol and DHEA. A separate linear mixed
model was run on outcome variables for each predictor of interest. Each model included the
given predictor and visit as fixed effects and participant as a random intercept.

Results: High overall school stress was reported, with slightly increased stress during the
second visit (P=0.103). On the second visit, which corresponded to final school examinations,
average cortisol (marginal mean 4.0 vs. 6.5, P < 0.001), but not DHEA (marginal mean 14.5 vs.
15.7, P=0.316), was significantly higher. Neither overall stress, nor any of ten domain-specific
stress questions were correlated with higher cortisol or DHEA levels. Self-reported asthma
status was also not correlated. Females showed significantly lower cortisol (4.4 vs. 6.4,
P=0.049) but not DHEA (14.9 vs. 15.3, P=0.858) than males.

Discussion: Middle-schoolers reported high perceived school stress, but questionnaire

responses did not correlate with hair biomarker measures. Chronic stress may induce
hyporesponsiveness of the hypothalamic-pituitary adrenal axis and result in a blunted cortisol
response. This preliminary study demonstrated feasibility of collecting hair biomarkers in a
school setting and lower reported levels among females but found no correlation with self-
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1385

reported measures. Future research can examine these associations in larger samples among
asthmatic children, to better understand mechanisms mediating chronic stress and asthma
relationships.

CORRESPONDING AUTHOR: Eyal Oren, PhD, MS, University of Arizona, Tucson, AZ, 85724;
[email protected]
S1386 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B279 6:00 PM-7:00 PM

SELF-ENHANCEMENT AND EMOTIONAL EXPERIENCE AMONG PATIENTS WITH THALASSEMIA

AND MULTIPLE SCLEROSIS

Shaima Almahmoud, MA1, Karin Coifman, PhD2

1
Kent state University, kent, OH; 2Kent state University, Kent, OH

Selfenhancement (SE) is the tendency toward overly optimistic or unrealistic self-serving

biases. It has been shown to be a useful disposition under circumstances of extreme diversity
with different populations such as patients with chronic illnesses (e.g., HIV; Stein et al., 2011;
cancer; Zsuzsanna et al., 2012). While SEers show resilience and good adjustment in managing
threatening events, SE has potential social costs too (Bonnano et al., 2002). For example,
SEers have difficulty forming or maintaining friendships (Fraley & Bonanno, 2004; Yan &
Bonanno, 2015), which may conflict or undermine the benefits. Although there is evidence of
regulatory benefits to SE following threatening situations, little is known about the beneficial
effects of SE on different emotion response dimensions (especially emotion facial behavior).
This study examines emotional responses in a sample of chronically ill adults (Thalassemic/MS
patients). Thalassemia: an inherited blood disorder causing severe anemia that may lead to
frequent complications due to transfusion-transmitted infections and iron overload (Galanello
& Origa, 2010). Multiple Sclerosis (MS): a kind of autoimmune diseaseusually diagnosed
between 20-50, which involves abnormal response of the bodys immune system against
the central nervous system leading to unpredictable exacerbations (Goretti et al.,
2010).Participants were 27 adults with transfusion-dependent thalassemia, 41 adults with MS,
73% female. Mean age= 35, SD= 8.86 years. They were recruited from medical centers in the
Northeastern USA.They completed questionnaires and were interviewed in private in the
medical exam room or in the lab. Multiple regression analyses, controlling for psychological
distress, age, and disease burden were conducted to examine whether SE could significantly
predict positive or negative affect or facial behavior and how the nature of the relationship
between SE and emotion changed as a function of the type of illness. Results showed that
higher SE was associated with lower levels of psychological distress, r=-0.38, p=.001 and
higher reported positive affect, r=0.47, p < .001 across both samples. There was an interaction
of SE with the patient group when predicting negative and positive facial emotion, =.32,
p=.04, =-.39, p=.004, respectively. Moreover, there was an interaction of SE with the patient
group when predicting close informants social attributes and the availability of the close
informant to the participant, =.27, p=.05, =.43, p=.002, respectively. These results indicate
the importance of illness history in the development and influence of self-enhancing biases
and show that SE has different impacts on the regulatory responses of emotion.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1387

CORRESPONDING AUTHOR: Shaima Almahmoud, MA, Kent state University, kent, OH, 44242;
[email protected]
S1388 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B280 6:00 PM-7:00 PM

THE ACCURACY OF BRIEF STRESS MEASURES IN PREDICTING ILLNESS

James Amirkhan, Ph.D.

California State University Long Beach, Long Beach, CA

Medical and psychological theories indicate stress to be an etiological factor in a variety of

illnesses; therefore, stress measures hold promise for the prediction of imminent pathology.
However, most stress measures are too lengthy to be used for rapid diagnosis, and too
cumbersome to be imposed on distressed individuals. The present study examined two brief
measures, the popular 10-item Perceived Stress Scale (PSS-10) and the recently published 10-
item Stress Overload Scale-Short (SOS-S), in terms of their accuracy in identifying symptomatic
persons. A demographically diverse sample (n = 249) was drawn from two different
community sites, which were purposefully selected to yield a wide range of stress levels.
Survey packets included the PSS-10, the SOS-S, and the Patient Health Questionnaire (PHQ-
15), which assesses somatic symptoms common to physical and psychiatric disorders, in
counterbalanced orders. Participants completed these packets on site, with full anonymity of
response. Partial correlations showed PSS-10 and SOS-S totals to inter-correlate, and to
correlate significantly with PHQ-15 symptoms. However, a test comparing the magnitude of
these correlations revealed the SOS-S to be a significantly stronger predictor of PHQ-15 scores
than the PSS-10. Next, categories were constructed from stress scores to divide participants
into High Risk vs. Low Risk groups for each measure. ANOVA tests showed significant
differences in the number of symptoms reported across risk categories for the PSS-10 and for
the SOS-S. In short, both total scores and categorical scores, on both short stress measures,
worked to differentiate highly symptomatic from other participants. It was concluded that
brief stress scales, which are well suited to urgent situations and/or distressed victims, are
indeed viable diagnostic tools for assessing the risk for stress sequelae. Moreover, having
been tested in a general population at community venues, the PSS-10 and especially the SOS-S
appear to be good choices for use across a broad demographic spectrum and a variety of
contexts.

CORRESPONDING AUTHOR: James Amirkhan, Ph.D., California State University Long Beach,
Long Beach, CA, 90840; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1389

B281 6:00 PM-7:00 PM

THE ROLES OF STRESS EATING AND MINDFULNESS IN THE RELATIONSHIP BETWEEN RECENT
LIFE STRESSORS AND INDICATORS OF OBESITY RISK

Elizabeth Cotter, PhD1, Nichole Kelly, PhD2

1
American University, Washington, DC; 2University of Oregon, Eugene, OR

Research has established a link between stress and increased risk for obesity. This study
examined whether stress eating mediates the relationship between recent life stressors and
indicators of obesity risk, including body mass index (BMI) and waist circumference. We also
explored whether mindfulness moderates the relationship between recent life stressors and
stress eating. Cross-sectional data were examined from 4963 adults in the Midlife in the U.S.
study (MIDUS II). Hierarchical regression analyses indicated that recent life stressors were
associated with higher BMI and waist circumference, even after controlling for age, race,
gender, education level, and annual household income. Using a nonparametric bootstrapping
approach, we found that stress eating partially mediated the association between recent life
stressors and BMI/waist circumference. Both recent life stressors and mindfulness were
independently associated with stress eating; however, their interaction was not significant.
These results provide further evidence of the contributions of psychosocial factors to obesity
risk and suggest that stress reduction and mindful eating techniques might be beneficial
components to obesity-related interventions.

CORRESPONDING AUTHOR: Elizabeth Cotter, PhD, American University, Washington, DC,

20016; [email protected]
S1390 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B282 6:00 PM-7:00 PM

A COMPARISON OF TRAUMATIZED AND NON-TRAUMATIZED INJURED PATIENTS WHO

SCREENED POSITIVE FOR HEAVY DRINKING

Dylan Richards, BS1, Emmanuel Sequeira, BS2, Craig Field, PhD, MPH1

1
Latino Alcohol and Health Disparities Research (LAHDR) Center, Department of Psychology,
The University of Texas at El Paso, El Paso, TX; 2Latino Alcohol and Health Disparities Research
(LAHDR) Center, The University of Texas at El Paso, El Paso, TX

Health care settings are regularly accessed by non-treatment-seeking heavy drinkers. Thus,
Screening, Brief Intervention, and Referral to Treatment (SBIRT) has been implemented as a
public health approach to address the unmet need for alcohol treatment. The purpose of the
current study was to examine potential differences between two groups of heavy drinking
trauma patients: those who experienced their injury as traumatic and those who did not
experience their injury as traumatic.

Exploratory analyses were conducted using baseline data from a multisite randomized
controlled trial of brief intervention to reduce drinking in the trauma care setting. Participants
were 596 injured patients from three urban Level I trauma centers who screened positive for
heavy drinking. Participants completed questionnaires that assessed sociodemographics,
injury history, and alcohol-related variables. Univariate analyses were conducted to assess
potential differences between traumatized and non-traumatized patients. Statistical
differences between the groups included gender (2[1] = 7.55, p < .01), ethnicity (2[3] =
27.95, p < .001), intent of injury (2[1] = 8.19, p < .01), and readiness to change alcohol use
(2[2] = 7.91, p < .05). A multivariate logistic regression (0 = non-traumatized; 1 = traumatized)
assessed potential predictors of injury-related trauma. Correlates that met the criteria of p <
.25 in the univariate analyses were simultaneously entered into a logistic regression model.
Gender (OR = 0.46, = -0.78, p < .01 [reference group was female]) and identification as
African American and Hispanic (OR = 2.70, = .99, p < .001 and OR = 3.14, = 1.14, p < .001,
respectively) were statistical predictors in the model.

The results of the current study suggest that women and ethnic minorities were more likely to
experience their injuries as traumatic. Admission to a trauma care center provides an
opportunistic moment for alcohol intervention, and those who experience an injury as
traumatic may be more receptive to healthy behavior change. Understanding the
characteristics of traumatized and non-traumatized patients provides insight into potential
avenues for tailoring intervention.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1391

CORRESPONDING AUTHOR: Dylan Richards, BS, Latino Alcohol and Health Disparities Research
(LAHDR) Center, Department of Psychology, The University of Texas at El Paso, El Paso, TX,
79912; [email protected]
S1392 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B284 6:00 PM-7:00 PM

CHARACTERISTICS OF COMORBID MARIJUANA USE AND INTERNET GAMING DISORDER

Travis Hyke, B.A. 1, Charlotte Beard, M.S.2, Amie Haas, Ph.D.3

1
Palo Alto University, Sunnyvale, CA; 2Palo Alto University, Milpitas, CA; 3Palo Alto University,
Palo Alto, CA

The current study explored the association between Internet Gaming Disorder (IGD) and
comorbid marijuana use. Internet gaming addictions are highly prevalent in adolescents,
affecting about 3% of persons under age 20 (Muller at al., 2015, Ferguson, Coulson, & Barnett,
2011). Similar etiological pathways have been found between IGD and substance use
addictions, with both disorders linked to negative psychosocial consequences (Van Rooij et al.,
2014) and co-occurrence (Castro-Calvo et al., 2016). In particular, IGD appears to co-occur
with substances that are expand the users sensory experience, like marijuana, and it has been
posited that behavioral addictions like IGD activate cannabinoid receptors in the brain (Lopez-
Moreno et al., 2008) and comorbid use has been linked to IGD severity (Van Rooij et al.,
2014). However, few studies have provided more analysis of marijuana use among individuals
with IGD. The present study sought to explore two primary hypotheses concerning Internet
gaming and marijuana use: i) individuals who use marijuana while playing internet games
would show more IGD symptoms than those who abstained while gaming and ii) individuals
who reported using marijuana to enhance gameplay would have an increased severity of IGD
symptoms related to mood modification. Concurrent cannabis and video game users (n = 68,
63% male) were recruited from Amazon's Mechanical Turk as part of a larger study related to
Internet gaming and substance use. Participants completed a survey that included
questionnaires regarding domains of IGD, substance use, and psychological functioning. As
expected, individuals with simultaneous use had significantly greater IGD scores relative to
conjoint users (t (61) = -2.94, p < .01). Overall frequency of marijuana use and using marijuana
to enhance gaming experience failed to predict IGD (F(1, 61) = 2.35, p = 0.13 and F(1, 61) =
2.59, p = 0.11, respectively). This study provides preliminary evidence that the link between
marijuana use and IGD can be differentially affected by how and when individuals administer
marijuana. The present study has implications for future treatment of comorbid behavioral
and substance use addictions, as when and how individuals are using both can help inform
evidence-based treatment.

CORRESPONDING AUTHOR: Travis Hyke, B.A. , Palo Alto University, Sunnyvale, CA, 94089;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1393

B285 6:00 PM-7:00 PM

EPIDEMIOLOGY OF SUBSTANCE USE AMONG YOUNG AFRICAN AMERICAN WOMEN IN NEW

ORLEANS, LOUISIANA

Aubrey S. Madkour, PhD1, Gretchen Clum, PhD2, Norine Schmidt, MPH2, Carolyn C. Johnson,
PhD MS2, Patricia Kissinger, BSN, MPH, PhD2

1
Tulane University, New Orleans, LA; 2Tulane University School of Public Health and Tropical
Medicine, New Orleans, LA

Background: While studies exist describing differences in substance use across racial/ethnic
and gender groups, patterns of substance use and for correlates of substance use for late
adolescent African American women are not well understood.

Purpose: Few studies have examined the epidemiology of substance use among young
African American women, including potential correlates of substance use patterns within this
group.

Methods: Young African American women, ages 18-19 yrs, were recruited from a community
college, historically black universities, and community sites to participate in a randomized
controlled trial for online teen pregnancy prevention. At baseline participants reported
lifetime and past three-month alcohol, marijuana and other illicit drug use. Respondents also
reported indicators of health and well-being, characteristics of sexual partnerships, and family
background characteristics. Latent class analysis (LCA) and multivariable multinomial logistic
regression determined the association between potential risk factors and latent classes of
substance use.

Results: Of 459 participants, 60% reported drinking alcohol, 24% reported binge drinking, and
32% reported marijuana use in the past three months. LCA results suggested four latent
typologies of substance use: recent abstainers (31%), recent drinkers (30%), marijuana users
(8%), and polysubstance users (31%). In multivariable analyses, factors found to be
significantly associated with increased risk of substance use was higher maternal education,
early sexual initiation, and having a recent male sexual partner.

Conclusions: A substantial proportion of late African American adolescents exhibited one or

more recent substance use behaviors, with nearly a third evidencing use of two or more
substances. Future analyses should examine how substance use behaviors change over time,
as well as the consequences of longitudinal patterns of use for this group.
S1394 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Aubrey S. Madkour, PhD, Tulane University, New Orleans, LA,
70112; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1395

B286 6:00 PM-7:00 PM

EXAMINING CHARACTERISTICS AND CLINICAL TREATMENT OF SMOKERS IN A SUBSTANCE USE

DISORDER TREATMENT CLINIC

Julie C. Gass, PhD, MA1, David H. Morris, Ph.D.2, Jamie Winters, PhD3, Joseph VanderVeen,
PhD4, Stephen Chermack, Ph.D.5

1
VHAWNY Center for Integrated Healthcare, Buffalo, NY; 2Veteran Affair Ann Arbor
Healthcare System, Ypsilanti, MI; 3VA Ann Arbor Healthcare System & University of Michigan,
Ann Arbor, MI; 4Ann Arbor VA Healthcare System, Canton, MI; 5University of Michigan/VA
Ann Arbor, Ann Arbor, MI

Individuals with a substance use disorder (SUD) diagnosis are more than twice as likely to
smoke cigarettes as the general population. Emerging research has suggested that treating a
substance use disorder simultaneously with tobacco use leads to a higher rate of treatment
success for both substances. Despite this, substance use treatment protocols tend not to
focus on tobacco use; in fact, traditional substance use treatments discourage patients from
attempting to quit smoking. One rationale is that patients may not be motivated to quit
smoking. In the current study, data from Veterans enrolled in outpatient treatment for a SUD
were examined to assess for general characteristics of smokers as compared to non-smokers
as well as to examine motivation to quit smoking. Baseline (i.e., pre-treatment) data from 278
Veterans were used. Charts of smokers in the SUD clinic (SUDC) were reviewed to assess how
smoking is handled by SUDC providers, and if smokers to attempt cessation. Of 278, 163 SUDC
patients reported that they currently smoke cigarettes (M = 16.3 cigarettes per day, SD =
11.1). Smokers in the clinic reported greater general impairment than nonsmokers,
F(1,248)=10.2, p=.002, as well as greater specific impairment: Physical Problems,
F(1,258)=14.8, p=.000; Interpersonal Problems, F(1,262)=6.7, p=.010; Intrapersonal Problems,
F(1,260)=7.2, p=.008, and Social Responsibility, F(1,262)=16.5, p=.000. Smokers in the sample
were marginally more anxious than their non-smoking counterparts as measured by the
GAD=7, F(1,254)=3.8, p=.053, though they were not significantly more depressed (p=0.193).
On a 1-10 scale, smokers reported moderate levels of importance (M=5.4, SD=3.1), readiness
(M=5.6, SD=3.2), and confidence (M=5.0, SD=3.0) regarding quitting smoking. Review of
smokers medical records reveal that while SUDC providers assess tobacco use at intake (90%)
and offer treatment (86.5%), a substantially small portion of smokers attempt cessation
(41.1%) while enrolled in SUDC. No patients enrolled in smoking-specific treatment in SUDC
with a behavioral health provider, though 78 patients did obtain nicotine replacement or
another smoking cessation medication (41% were prescribed by a SUDC provider).These
preliminary analyses demonstrate that, contrary to the belief that treatment-seeking
S1396 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

substance users are not motivated to quit smoking, Veterans were at least moderately ready
to quit smoking while they were enrolled in substance use treatment though a majority do not
attempt to quit smoking. Further, there is evidence that those who are smokers have greater
impairment caused by substance use, suggesting that this subgroup is of particular high need.
Specific treatment recommendations will be discussed, including how behavioral health
providers in SUD clinics may be better able to capitalize on patients moderate motivation to
quit at intake.

CORRESPONDING AUTHOR: Julie C. Gass, PhD, MA, VHAWNY Center for Integrated
Healthcare, Buffalo, NY, 14217; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1397

B287 6:00 PM-7:00 PM

INITIATION INTO INJECTION DRUG USE AND CROSS-BORDER MIGRATION AMONG PEOPLE
WHO INJECT DRUGS IN TIJUANA, MEXICO

Claudia Rafful, MSc, PhD candidate1, Maria Luisa Mittal, MD2, Jason S. Melo, BS Public
Health3, Dan Werb, PhD4, Andy Guise, PhD 3

1
UCSD, La Jolla, CA; 2University of California San Diego, La Jolla, CA; 3University of California
San Diego, San Diego, CA; 4University of California, San Diego, La Jolla, CA

Background: People who inject drugs (PWID) play a key role in facilitating the entry of others
into this behavior. In Tijuana, Mexico, recent findings suggest that the initiating of others into
injecting is associated with reporting short-term residence (i.e., 1-5 years) in the United
States. However, the nature of the association between cross-border migration and injection
initiation of others and own remains unexplored. Methods: Preventing Injecting by Modifying
Existing Responses (PRIMER, NIDA DP2-DA040256-01) is a mixed methods multi-site study
assessing social and structural factors associated with PWID initiating others into drug
injection. For the present qualitative study, PWID enrolled since 2011 in a prospective cohort
study in Tijuana (Proyecto El Cuete IV, NIDA R01-DA019829)were purposively sampled for the
qualitative component of PRIMER according to sociodemographic factors and drug use
behaviors indicating their initiation of others into injection drug use. Semi-structured
interviews were performed during June- August 2016, which explored the context and
characteristic of individuals own injection initiation events as well as their initiating others
into drug injection. Thematic analysis was used to develop recurring and significant categories
in the data. Results: Twenty-one participants were interviewed (women=8, men=13) with a
median age of 40 years (Interquartile Range: 36 46). Migration between Mexico and the
United States was core in participant narratives on their own and others injection initiation.
The majority of participants that lived in the United States (66.7%) reported that they had
initiated their injection drug use there. Migration from Mexico to the United States was linked
to participants own initiation into injecting, which seems to be related to the normalization of
injection in social networks as well as underlying social and economic alienation and
marginalization. Additionally, participants reported on a moral code whereby the initiation of
others into injecting was highly stigmatized. However, participants also reported that
increases in economic, social and physical insecurity after returning to live in Mexico made
them more susceptible to initiating others. Conclusion: This is the first study to qualitatively
investigate the initiation of injection drug use in Mexico. Migration stressors appear key to
individuals own initiation and their initiation of others. Further research is needed to analyze
the social normalization and expansion of injection drug use among migrant populations in
S1398 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

the Mexico-U.S. border in order to address the unacceptably high burden of injection-related
harms among this vulnerable population. The impact of migration should be considered
within binational injection prevention and social programs.

CORRESPONDING AUTHOR: Claudia Rafful, MSc, PhD candidate, UCSD, La Jolla, CA, 92093-
0507; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1399

B288 6:00 PM-7:00 PM

MEDICAL MARIJUANA DISPENSARY ROLL-OUT: BASELINE MEASURES OF MARIJUANA USE

AMONG ADULTS AND YOUTH IN HAWAII FROM 2010-2014

Codie M. Garza, B.S.1, Claudio R. Nigg, PhD, FSBM2

1
Office of Public Health Studies/University of Hawaii at Manoa, Honolulu, HI; 2University of
Hawaii, Honolulu, HI

BACKGROUND: Medical marijuana has been legal in Hawaii since 2000. Dispensaries became
legal in 2015, were permitted as of July 2016, and are expected to open in December 2016.
With rates of prescription drug use without a prescription stable in Hawaii, monitoring and
prevention of non-medical marijuana use should be a focus among these changes in access.

PURPOSE: To provide baseline marijuana use among adults and youth in Hawaii prior to the
launch of dispensaries.

METHODS: The National Survey on Drug Use and Health data regarding adult and youth
marijuana use for years 2010-2014 was examined. Rates for past year and past month
marijuana use are presented for youth (1217), young adult (1825), and adult (26+).

RESULTS: Youth marijuana use remained stable from 2010-2014. Past year use rates were:
20102011: 14.95% [95%CI: 12.5417.73], 20112012: 16.6% [95%CI: 14.0419.54], 2012
2013: 15.33% [95%CI: 12.8218.23], 20132014: 13.9% [95%CI: 11.6116.57]. Past month use
rates were: 20102011: 8.29% [95%CI: 6.5410.46], 20112012: 9.69% [95%CI: 7.7412.06],
20122013: 9.55% [95%CI: 7.6911.81], 20132014: 7.65% [95%CI: 6.049.65].

Young adult marijuana use remained stable from 2010-2014. Past year use rates were: 2010
2011: 29.53% [95%CI: 25.8733.47], 20112012: 31.51% [95%CI: 28.0535.19], 20122013:
31.27% [95%CI: 27.6935.09], 20132014: 27.32% [95%CI: 23.7531.2]. Past month use rates
were: 20102011: 17.58% [95%CI: 14.6520.95], 20112012: 18.15% [95%CI: 15.4221.25],
20122013: 18.84% [95%CI: 15.9422.13], 20132014: 15.87% [95%CI: 13.0719.14].

Adult marijuana use remained stable from 20102014. Past year use rates were: 20102011:
8.79% [95%CI: 710.97], 20112012: 10.14% [95%CI: 8.1712.53], 20122013: 10.35% [95%CI:
8.2812.87], 20132014: 10.13% [95%CI: 8.1812.48]. Past month use rates were: 2010
2011: 5.13% [95%CI: 3.866.8], 20112012: 5.69% [95%CI: 4.317.48], 20122013: 6.91%
[95%CI: 5.288.99], 20132014: 6.68% [95%CI: 5.128.66].
S1400 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CONCLUSIONS: The data suggests a fairly low level of use in youth and adults but higher use in
young adults. Relatedly, other data indicates young adults have the lowest perception of great
risk about marijuana use. With recreational marijuana use illegal in Hawaii and change in
access to come, it is vital to monitor marijuana use.

FUNDING: Strategic Prevention Framework Partnerships for Success 2013 Initiative (SPF-PFS
2013) of the Substance Abuse and Mental Health Services Administration (SAMHSA), under
grant number:1U79SP020167

CORRESPONDING AUTHOR: Codie M. Garza, B.S., Office of Public Health Studies/University of

Hawaii at Manoa, Honolulu, HI, 96822; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1401

B289 6:00 PM-7:00 PM

MENS AND WOMENS PERCEPTIONS OF TREATMENT OUTCOMES IN A CLINICAL TRIAL OF

INJECTABLE OPIOIDS FOR SEVERE OPIOID USE DISORDER

Heather Palis, BA, MSc, Kirsten Marchand, BSc, Eugenia Oviedo-Joekes, PhD

University of British Columbia, Vancouver, BC, Canada

Background: Studies of opioid use disorder and clinical trials of treatments for severe opioid
use disorder consistently suggest that treatment outcomes rarely differ by gender. The
present study aims to explore whether men and women have differing perceptions of
treatment effectiveness.

Methods: Baseline characteristics and six-month treatment outcomes were analysed by

gender. Participants were asked to share their reasons for perceived treatment effectiveness.
A thematic analysis was carried out using an inductive approach and themes and relationships
across participant comments were developed using the strategy of constant comparison.

Results: Women in treatment presented with a number of vulnerabilities as compared to

men. These included higher rates of HIV (22.6% vs. 11.4%), having ever engaged in sex work
(83.9 % vs. 22.1%), and significantly worse health. After six months of treatment, the only
outcome that differed by gender was physical health, with men having a significantly better
score. Themes surrounding perceptions of treatment effectiveness were similar for men and
women, the most common of which were improved health and improved quality of life,
however mens and womens descriptions surrounding these themes differed.

Conclusions: The collection of participant perceptions of effectiveness provides an

opportunity for understanding the patient experience that cannot be revealed with the use of
standardized tools and researcher defined health measures. The systematic use of
information collected from participants perceptions of effectiveness has strong clinical
significance, and can work to ensure clinical practice is designed in a way that can effectively
meet individual patient needs.

CORRESPONDING AUTHOR: Heather Palis, BA, MSc, University of British Columbia, Vancouver,
BC, V6J 0C9; [email protected]
S1402 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B290 6:00 PM-7:00 PM

SOCIAL DESIRABILITY BIAS AND SELF-REPORTS OF MENTAL HEALTH, DRUG USE, AND HEALTH
SERVICE UTILIZATION AMONG SUBSTANCE USERS.

Carl A. Latkin, Ph.D, Melissa Davey-Rothwell, Ph.D, Karin Tobin, Ph.D

Johns Hopkins School of Public Health, Baltitmore, MD

Social desirable response bias may lead to inaccurate self-reports and erroneous study
conclusions. We examined the relationship between social desirability bias and self-reports of
substance use, mental health, and health service utilization among a community sample of
591 inner-city substance users who were enrolled in a behavioral change intervention to
address HIV risk behaviors and depressive symptoms. The study was conducted Baltimore,
Maryland from 2009-2013. Modified items from the Marlowe-Crowne Social Desirability Scale
were included in the survey, which was conducted face-to-face and using Audio Computer Self
Administering Interview (ACASI) methods. Less than half (44%) the participants were female,
54% had less than 12 years of education, 61% reported main partners. The median age was 45
and 95% reported that they had been unemployed within the last six months and 24%
homeless. There were highly statistically significant differences in levels of social desirability
response bias by levels of depressive symptoms, drug user stigma, subjective health status,
recent drug use, and size of social networks. There were no associations between health
service utilization measures and social desirability bias. In multiple logistic regression models,
even after including the CES-D, as a measure of depressive symptomology, social desirability
bias was associated with drug use stigma, drug use duration, and network size. We also found
that social desirability bias was not associated with enrollment in prior research studies,
indicating that professional subjects do not appear to provide more socially desirable
responses. These findings suggest that social desirability bias is associated with key health
measures. These associations are not primarily due to depressive symptoms and research
methods are needed to reduce social desirability bias. Such methods may include the wording
and prefacing of questions, clearly defining the role of study participant, and assessing and
addressing motivations for socially desirable responses.

CORRESPONDING AUTHOR: Carl A. Latkin, Ph.D, Johns Hopkins School of Public Health,
Baltitmore, MD, 21205; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1403

B291 6:00 PM-7:00 PM

YOUNG ADULTS RIDING WITH A MARIJUANA-, ALCOHOL-, AND DRUG-IMPAIRED DRIVER

Kaigang Li, PhD1, Elizabeth Ochoa, B.A2, Federico E. Vaca, MD, MPH3, Bruce Simons-Morton,
EdD, MPH4

1
Colorado State University, Fort Collins, CO; 2Colorado State University, Fort collins, CO; 3Yale
University School of Medicine, New Haven, CT; 4NICHD, Bethesda, MD

Purpose: Examine the prevalence and correlates of young adults riding with an impaired
driver (RWI) due to marijuana (MJ), alcohol (ALC), or illicit drug (ID).

Methods: Data were from the NEXT Generation Health Study waves 4 and 5 (W4&5 [2013-14],
1 and 2 years after high school, with N=1979 at W4 and N=2142 at W5). W5 data was
specified for ALC-, MJ-, and ID-related RWI in a vehicle that was driven by a young or older
adult. Multinomial logistic regressions estimated the association of W5 substance-specific RWI
with W5 binge drinking, MJ use, and environmental status variables (i.e., school-, residence-,
and work- status). Overall W4 RWI and demographic variables were included.

Results: The prevalence of W5 RWI was 33% (n=694). Specifically, 17% of participants
reported riding with a young ALC-impaired driver and 4% with an older ALC-impaired driver;
21% reported riding with a young MJ-impaired driver and 2% with an older MJ-impaired
driver; and 5% reported riding with a young ID-impaired driver and 0.7% with an older ID-
impaired driver. The adjusted models included residence and work status separately so two
adjusted odds ratios [OR] were reported. W4 RWI was associated with W5 ALC- (OR=2.6-4.3, p
< .05).

Conclusion: The results show that RWI is prevalent in young adults two years after high
school, with 33% of participants reporting RWI at least once in the past year. MJ-related RWI
prevalence was greater than ALC-related prevalence, and occurred more often with an
impaired young driver than with an impaired older driver. MJ use was associated with all ALC-,
MJ-, and ID-related RWI. The findings support the reduction of early RWI, substance use, and
heavy drinking as strategies to reduce impaired driving in young adults.

CORRESPONDING AUTHOR: Kaigang Li, PhD, Colorado State University, Fort Collins, CO,
80523; [email protected]
S1404 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B292 6:00 PM-7:00 PM

AURORA SMOKING AND HEALTH SURVEY: A COMMUNITY-BASED SURVEY ON STRESS, COPING

AND SMOKING STATUS

Patricia Valverde, PhD MPH1, Kathleen Garrett, MA1, Robert McGranaghan, MPH2, Natalie
Slopen, ScD3, Margaret A. Chesney, PhD4, Arnold Levinson, PhD MJ1

1
Colorado School of Public Health, Aurora, CO; 2Colorado School Of Public Health, Aurora, CO;
3
School of Public Health, University of Maryland College Park, College Park, MD; 4UCSF School
of Medicine, Menlo Park, CA

SIGNIFICANCE: The community-academic partnership project was designed to assess smoking

patterns, coping skills, knowledge, motivation to quit smoking and barriers to cessation
including chronic life stressors in a low socioeconomic status (SES) population. We
hypothesized that stress levels would be similar among smokers and non-smokers and coping
self-efficacy would be associated with smoking status. METHODS: The cross sectional survey
respondents were English and Spanish speaking adult smokers and non-smokers recruited
through recreation centers, a local food bank program, multi-unit housing complexes and a
public library in the targeted neighborhood. The survey instrument included validated
measures and a stress instrument of 6 domains comprised of life stressors, stress related to
work, relationships, neighborhood, unfair treatment and finances. Each stress domain score
was standardized into a Z-score distribution. Coping was measured using the Coping Self-
Efficacy instrument. RESULTS: A total of 405 individuals were surveyed including 295 current
smokers (72.8%) and 110 non-smokers (27.2%). Respondents were 48% male, 42.6% African
American and 31.4% Hispanic and 45.8% were single and only 11% of the population
completed college or higher. Three fourths of the respondents were not working. The
majority had some type of health insurance (79.4%) with most reporting Medicaid as their
insurance (72.4%). Most respondents smoked cigarettes (95.6%) compared to cigars (8.8%). A
third of respondents reported smoking their first cigarette within five minutes of waking up
(35.4%). Regarding past quit attempts, respondents who did use evidenced based treatments
reported NRT (26.6%), and QuitLine (18.3%) use. Those more likely to be smokers were men
with less than college education and non-Latinos. Those experiencing higher levels of stress
due to discrimination were at higher odds (aOR 1.82, 95% CI 1.25-2.67) of being a smoker
than those with lower stress levels after controlling for demographic and other stress
variables. Higher levels of stress across the 6 stress domains were associated with smoking
(aOR 1.37 95% CI 1.07-1.76) even after adjustment. Those with higher levels of coping self-
efficacy were more likely to be non-smokers (OR 0.99 95% CI 0.985-0.996). CONCLUSION:
Different types of stressors and the coping skills to handle them are an under-researched area
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1405

for smoking cessation interventions. Assisting the low SES smoker to quit and adhere amidst
chronic stressors will be essential to further reduce smoking prevalence.

CORRESPONDING AUTHOR: Patricia Valverde, PhD MPH, Colorado School of Public Health,
Aurora, CO, 80045; [email protected]
S1406 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B293 6:00 PM-7:00 PM

CIGARETTE SMOKING AND ELECTRONIC CIGARETTE VAPING PATTERNS AS A FUNCTION OF E-

CIGARETTE FLAVORINGS

Mark Litt, PhD1, Valerie Duffy, PhD2, Cheryl Oncken, MD3

1
University of Connecticut Health Center, Farmington, CT; 2University of Connecticut, Storrs,
CT; 3University of Connecticut School of Medicine, Farmington, CT

Objective: The present study examined the influence of flavoring on the smoking and vaping
behavior of cigarette smokers asked to adopt e-cigarettes for a period of six weeks.

Methods: Subjects were 88 current male and female smokers with no intention to stop
smoking, but who agreed to substitute e-cigarettes for their current cigarettes. Upon intake,
subjects were administered tests of taste and smell for e-cigarettes flavored with tobacco,
menthol, cherry, and chocolate, and were given a refillable e-cigarette of their preferred
flavor or a control flavor. Subjects monitored their daily cigarette smoking for 1 week, and
were then asked to adopt e-cigarettes for a period of 6 weeks. Subjects completed daily logs
of cigarette and e-cigarette use and were followed each week.

Results: Analyses over days indicated that, during the 6-week e-cigarette period, cigarette
smoking rates dropped from an average of about 16 cigarettes/day to about 7/day. E-
cigarette flavor had a significant effect on smoking, such that the largest drop in cigarette
smoking occurred among those assigned menthol e-cigarettes, and the smallest drop in
smoking occurred among those assigned chocolate and cherry flavors. These results were
confirmed by weekly CO readings. E-cigarette vaping rates also differed significantly by flavor
assigned, with the highest vaping rates for tobacco and cherry flavored e-cigarettes, and the
lowest rates for those assigned to chocolate. These results persisted regardless of preferred
flavor, whether the subject was a menthol smoker, or the level of nicotine in the e-cigarette.

Conclusions: The findings suggest that adoption of e-cigarettes in smokers may influence
smoking rates, and that e-cigarette flavorings can moderate this effect. E-cigarette vaping
rates are also influenced by flavorings. These findings may have implications for the utility of
e-cigarettes as a nicotine replacement device, and for the regulation of flavorings in e-
cigarettes for harm reduction.

CORRESPONDING AUTHOR: Mark Litt, PhD, University of Connecticut Health Center,

Farmington, CT, 06030-3910; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1407

B294 6:00 PM-7:00 PM

ELECTRONIC NICOTINE DELIVERY SYSTEM (ENDS) USE DURING SMOKING CESSATION: A

QUALITATIVE STUDY OF 40 OKLAHOMA QUITLINE CALLERS

Katrina Vickerman, PhD1, Laura Beebe, PhD2, Gillian Schauer, PhD, MPH3, Brooke Magnusson,
MA1, Brian King, PhD, MPH4

1
Optum Center for Wellbeing Research, Seattle, WA; 2University of Oklahoma Health Sciences
Center, Oklahoma City, OK; 3Centers for Disease Control and Prevention, Office on Smoking
and Health, Atlanta, GA; 4Office on Smoking and Health, Centers for Disease Control and
Prevention, Atlanta, GA

Objectives: Approximately ten percent (40,000 each year) of U.S. quitline enrollees who
smoke cigarettes report current use of Electronic Nicotine Delivery Systems (ENDS); however,
little is known about callers ENDS use. Our aim was to describe why and how quitline callers
use ENDS, their beliefs about ENDS, and the impact of ENDS use on callers quit processes and
use of FDA-approved cessation medications.

Design: Qualitative interviews conducted 1-month post-registration. Interviews were

recorded, transcribed, double-coded, and analyzed to identify themes.

Setting: Oklahoma Tobacco Helpline

Participants: 40 callers aged 18 who were seeking help to quit smoking, were using ENDS at
registration, and completed 1 program calls.

Results: At 1-month post-registration interview, 80% of callers had smoked cigarettes in the
last 7 days, almost two-thirds were using ENDS, and half were using cessation medications.
Nearly all believed ENDS helped them quit or cut down on smoking; however, participants
were split on whether they would recommend cessation medications, ENDS, or both together
for quitting. Confusion and misinformation about potential harms of ENDS and cessation
medications were reported. Participants reported using ENDS in potentially adaptive ways
(e.g., using ENDS to cut down and NRT to quit, and stepping down nicotine in ENDS to wean
off ENDS after quitting) and maladaptive ways (e.g., frequent automatic ENDS use, using ENDS
in situations they didnt previously smoke, cutting down on smoking using ENDS without a
schedule or plan to quit), which could impact the likelihood of quitting smoking or continuing
ENDS use.
S1408 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions: These qualitative findings suggest quitline callers who use ENDS experience
confusion and misinformation about ENDS and FDA-approved cessation medications. Callers
also use ENDS in ways that may not facilitate quitting smoking. Opportunities exist for
quitlines to educate ENDS users and help them create a coordinated plan most likely to result
in completely quitting combustible tobacco.

CORRESPONDING AUTHOR: Katrina Vickerman, PhD, Optum Center for Wellbeing Research,
Seattle, WA, 98104; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1409

B295 6:00 PM-7:00 PM

IDENTIFYING PREDICTORS OF SMOKING IN THE SMOKEFREE.GOV ADULT TEXT-MESSAGING

SMOKING CESSATION PROGRAM

Kara P. Wiseman, MPH, PhD1, Kisha I. Coa, PhD, MPH2, Augustson Erik, PhD, MPH3

1
National Cancer Institute, Bethesda, MD; 2ICF International, Rockville, MD; 3National Cancer
Institute, Rockville, MD

Background: Text-based smoking cessation programs are effective, however, many users fail
to initiate their quit attempt or quickly relapse. A more thorough understanding of predictors
of early failures can help inform program changes targeted to improving treatment outcomes.
Methods: Items were included on the program registration page of a cessation text program
to assess: time to first cigarette, cigarette cravings, frequency of smoking reminders,
frequency around smokers, external and internal motivation to quit, confidence in the quit
attempt, and long-term intention to be smoke free. Associations between each characteristic
and smoking status were evaluated using logistic regression modeling (n=6226). Two
outcomes were considered: smoking on the quit day (yes/no) and smoking one week into the
quit attempt (yes/no). Smoking on the quit day was included in all models of smoking at one-
week. Results: Three user characteristics were associated with smoking on the quit day. Users
who smoked within five minutes, or between six and 30 minutes of waking up were 1.63 and
1.54 times more likely to smoke on their quit day, respectively compared to users who
smoked after more than 30 minutes (95% CIs: 1.202.21, 1.142.10). Compared to users who
had low confidence in their quit attempt, users with high confidence or some confidence were
50% and 40% less likely to smoke on their quit day, respectively (ORs: 0.50, 0.60; 95%CIs:
0.33-0.75, 0.40-0.89). Users who had low long-term smoke free intention were 1.79 times
more likely to smoke on their quit day compared to those with high long-term intention
(95%CI: 1.11-2.88). Two user characteristics were associated with smoking at one-week. Users
who were often around smokers were 2.04 times more likely to smoke at one-week compared
to users who were infrequently around smokers (95%CI: 1.36-3.09). Users who reported high
external motivation to quit smoking were 1.52 times more likely to smoke at one-week
compared to users with low external motivation to quit (95%CI: 1.04-2.23). Conclusions:
Within a short time window, the predictors of smoking differed, which suggests that the
relevance of specific messages might change over the course of the quit attempt. Quit success
might be increased if messages early in a program focus on preparing users for withdrawal
symptoms, and building confidence in quitting, and by saving messages about handling
cravings and maintaining motivation to stay smoke free until after the quit day.
S1410 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Kara P. Wiseman, MPH, PhD, National Cancer Institute, Bethesda,
MD, 20892-9761; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1411

B296 6:00 PM-7:00 PM

IMPULSIVITY AND ALTERNATIVE TOBACCO USE AMONG YOUNG ADULT NON-DAILY SMOKERS

Kristin Brikmanis, BA, Angela Petersen, PhD, Neal Doran, PhD

University of California, San Diego, La Jolla, CA

Introduction: Impulsivity has been prospectively associated with cigarette smoking behaviors
among young adults, but little is known about how impulsivity affects alternative nicotine and
tobacco product (ANTP) use. One possibility is that impulsive individuals find nicotine
particularly reinforcing, and thus may be vulnerable to rapid progression of tobacco use. The
purpose of the present study was to examine the association between impulsivity and ANTP
use over time.

Method: Participants (n = 392, 52% male) aged 18-24 were recruited for a longitudinal study
of tobacco use. Eligibility criteria included non-daily cigarette smoking for 6 months with no
history of daily smoking. Five components of impulsivity (lack of premeditation, lack of
perseverance, sensation seeking, and negative and positive urgency) were measured at
baseline, and participants reported recent tobacco use at baseline and follow-ups at 3, 6, and
9 months. Assessments were conducted online or via mobile app.

Results: Across the 4 assessments, 37-55% of participants reported recent ANTP use.
Frequency of use did not vary significantly over time. Sensation seeking and lack of
premeditation prospectively predicted frequency of ANTP use, such that participants higher in
those aspects of impulsivity used ANTPs more often. This association did not change over
time.

Conclusion: Findings suggest a high prevalence of ANTP use among young adult non-daily
smokers. Data also indicated that those who lack premeditation and/or are prone to
sensation seeking may be more likely to use ANTPs.

CORRESPONDING AUTHOR: Kristin Brikmanis, BA, University of California, San Diego, La Jolla,
CA, 92037; [email protected]
S1412 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B297 6:00 PM-7:00 PM

JOB SEEKING SMOKERS: SCREENING & ELIGIBILITY IN A TOBACCO TREATMENT TRIAL

Beatriz Anguiano, B.A1, Anna Epperson, PhD2, Nicole E. Anzai, BS3, Adrienne Lazaro, BS4,
Judith J. Prochaska, PhD, MPH3

1
Stanford University, Oakland, CA; 2Stanford Prevention Research Center, Palo Alto, CA;
3
Stanford University, Stanford, CA; 4Stanford Prevention Research Center, Palo, CA

Background/Objective: While the US adult smoking prevalence has declined greatly over the
past 4 decades, disparities exist for those with less education, lower socioeconomic status,
and the unemployed. In California, 1.4 million adults are unemployed and seeking work and
16% are smokers, compared to 12% of employed Californians. Our prior research has found
unemployed smokers have a significantly lower likelihood of reemployment and a lower
hourly wage compared to nonsmoking jobseekers. The current study summarizes screening
and recruitment strategies of the Total IMPACT study, a randomized controlled trial testing a
smoking cessation intervention for unemployed smokers with the objective of increasing
abstinence and aiding re-employability.

Methods: Jobseekers presenting to three Employment Development Departments (EDD) in

the San Francisco Bay Area were approached about the study and screened for eligibility. Pro-
active face-to-face recruitment and reactive methods (e.g., flyers) were used and reasons for
ineligibility and refusals were coded. Eligibility criteria included unemployed < 2 years,
currently job seeking, smoking > 1 cigarette per day, with a carbon monoxide breath reading >
7 ppm, able to read and write in English, and currently residing in the Bay Area.

Results: Of the 1,934 adults approached between October 2015 and August 2016, 37% were
current smokers (n = 720). Those ineligible were nondaily smokers (23%), currently employed
(23%), or non-job-seeking (16%). Among eligible smokers (n = 231), 67% declined. Key reasons
for declining enrollment were the length of time needed to complete the study (21%) and lack
of interest (67%). To date, 144 participants have enrolled into the study. Participants are
predominantly men (73%) and ethnically/racially diverse with 52% non-Hispanic African
American, 30% non-Hispanic White, 8% Hispanic, and 4% Asian/Pacific Islander. Among those
eligible, men were more likely to enroll than women (43% vs. 29%) and African American
individuals were more likely to enroll than other ethnic/racial groups (49% vs. 38%), p-values <
.05.

Conclusions: While integrating a tobacco cessation intervention into EDD settings offers
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1413

opportunity for reach and engagement of under-served and racially/ethnically diverse

smokers, recruitment is challenging and points to the need for flexibility in criteria (i.e.,
broadening to include nondaily smokers) and minimizing study time demand.

CORRESPONDING AUTHOR: Beatriz Anguiano, B.A, Stanford University, Oakland, CA, 94602;
[email protected]
S1414 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

B298 6:00 PM-7:00 PM

METHODS OF CESSATION AND ROLE OF ELECTRONIC CIGARETTES AMONG CURRENT AND

FORMER SMOKERS: A CROSS-SECTIONAL SURVEY OF US ADULTS

Pratibha Nayak, PhD1, Terry Pechacek, PhD1, Joanna-Grace Manzano, MD2, Michael Eriksen,
ScD1

1
Georgia State University, Atlanta, GA; 2The University of Texas MD Anderson Cancer Center,
Houston, TX

Background: Electronic cigarettes (EC) are increasingly used by smokers as cessation or harm
reduction tools without FDA approval. We examined smoking cessation methods of current
and former smokers, EC use with nicotine, perceived influence on craving, and attitude
toward continued use of EC.
Participants: Nationally representative 2015 cross-sectional survey of U.S. adults.
Respondents were current (n=1,280) and former smokers (n=341; within five years of quitting
cigarette use).
Methods: Main exposure was EC use to substitute for or complement cigarettes. Outcome
measures were weighted adjusted odds ratios (AOR) for successful quitting based on reported
smoking status.
Results: The most commonly reported method to quit smoking was gradually reducing
cigarettes among current smokers (64%) and completely giving up cigarettes among former
smokers (80%). Respondents who reported having tried switching completely to EC while
trying to quit were more likely to report having a successful quit attempt compared to those
who did not [AOR=1.53, 95% confidence interval [CI], 1.07-2.19]. Respondents who had
completely switched to EC during their last quit attempt were more likely (AOR=2.81, 95%CI,
1.28-6.15) to report having a successful quit attempt. Compared to respondents who did not
use EC as cigarette substitutes, those who did were less likely to report a successful quit
attempt (AOR=0.33, 95%CI 0.17-0.64). About 100% of current and 95% of former smokers
who used EC daily included nicotine. Almost 35% of current smokers and 80% of former
smokers reported that EC helped reduce cigarette cravings. Daily EC users intend to use EC for
a long time.
Conclusions: Switching completely to EC may help smokers successfully quit; however, using
EC to substitute for some cigarettes was less likely to support quitting. Thus, EC use during
quit attempts may have positive and negative effects on population trends in quitting.

CORRESPONDING AUTHOR: Pratibha Nayak, PhD, Georgia State University, Atlanta, GA,
30303; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1415

B299 6:00 PM-7:00 PM

PSYCHOLOGICAL FACTORS AMONG COPD PATIENTS: RELATIONS WITH SMOKING STATUS AND
COPD SYMPTOM IMPACT

Amanda Mathew, Ph.D., Susan Yount, Ph.D., Allison Rogowski, M.S., Neil Shea, M.D., Ravi
Kalhan, M.D., M.S., Brian Hitsman, Ph.D.

Northwestern University Feinberg School of Medicine, Chicago, IL

Background: Chronic obstructive pulmonary disease (COPD) is a tobacco-related disease

associated with several comorbid conditions, including elevated rates of depression and
anxiety. Psychological factors that commonly underlie nicotine dependence, depression, and
anxiety may represent novel treatment targets, but have not yet been examined among COPD
patients. We assessed three psychological factors anxiety sensitivity (AS; fear of anxiety-
related sensations), distress intolerance (DI; inability to withstand distressing states), and
anhedonia (Anh; diminished sense of pleasure or interest) in relation to smoking status,
COPD symptom impact, and negative response to COPD symptoms.

Methods: We conducted a single-session laboratory assessment with 40 COPD patients (20

current smokers and 20 former smokers). All participants completed self-report measures of
psychological factors (Anxiety Sensitivity Index, Distress Intolerance Index, and Snaith-
Hamilton Pleasure Scale, respectively), COPD symptom impact (COPD Assessment Scale),
response to COPD symptoms (Breathlessness Catastrophizing Scale), and general anxiety and
depression symptoms (PROMIS-Depression and Anxiety short forms).

Results: Participants were older adults (66.7 + 6.8 years; 21 men, 19 women) with moderate
levels of COPD symptom impact (CAT score 16.6 + 6.7). Current vs. former smokers reported
higher levels of AS (19.0 vs. 9.9; p=.01) and marginally higher levels of DI (16.1 vs. 11.5; p=.07).
In univariate regression models, AS, DI, and Anh were each associated with greater COPD
symptom impact (s .31-.64; all pss .36-.67; all ps=.36, p=.03) and breathlessness
catastrophizing (=.31, p=.04) after adjusting for general depression and anxiety symptoms.

Conclusion: Preliminary findings show that AS and DI are higher among current vs. former
smokers and all three psychological factors are positively associated with COPD symptom
impact and negative response to symptoms. AS showed unique associations with COPD
functioning even after adjusting for general depression and anxiety symptoms, perhaps
because it specifically indexes risk for negative reactions to bodily sensations such as
shortness of breath. AS may be an important target for behavioral interventions to improve
symptom burden and quality of life among COPD patients.
S1416 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Amanda Mathew, Ph.D., Northwestern University Feinberg School

of Medicine, Chicago, IL, 60611; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1417

B300 6:00 PM-7:00 PM

RACIAL AND GENDER DIFFERENCES IN SMOKING, SECONDHAND SMOKE AND RELATED

FACTORS FROM THE 2009 CALIFORNIA HEALTH INTERVIEW SURVEY

Katrina Obieta, MPH(c), Hee-Soon Juon, PhD

Thomas Jefferson University, Philadelphia, PA

Background: Lung cancer is the number one cause of cancer-related death in the U.S. and
worldwide. Overall lung cancer incidence and mortality rates have decreased for all races over
the past 20 years, however, racial differences still persist. When looking at smoking rates by
race, Asians and Hispanics have the lowest prevalence of smoking. The data is consistent in
that Asians and Hispanics also have the lowest lung cancer incidence and mortality rates.
When these racial groups are further divided by ethnic subgroups, however, there are large
disparities in trends for lung cancer between subgroups. Particularly, large disparities persist
in Asian American ethnic groups. The purpose of this study is to examine the prevalence of
current smokers and secondhand smoke by race, Asian American ethnicity, and gender. The
study also examines the risk factors for smoking.

Methods: We used data from the 2009 California Health Interview Survey (CHIS) and
performed univariate analysis to examine racial/ethnic and gender differences in prevalence
of current smokers as well as racial differences in prevalence of secondhand smoke exposure.
We also looked at related factors for smoking among residents of California. A multivariate
logistic regression model was conducted with weighted analyses using svy from STATA version
13.

Results: Among 47,614 California adults, the prevalence of current smokers was 13.57% (95%
CI 12.83-14.34%). American Indian / Native American (26.64% 95% CI 18.61-36.58) and African
Americans (16.22%, 95% CI 13.55-19.3%) had the highest rate of current smoking while Asian
Americans had the lowest among all races (10.58%, 95% CI 8.45-13.17%). When Asian
Americans are disaggregated, however, rates of prevalence are different among subgroups.
Among Asian Americans, Vietnamese (30.96%, 95% CI 16.1-51.17), Korean (20.37%, 95% CI
9.13-39.42%), and Filipino (19.35%, 95% CI 12.13-29.42%) men and highest prevalence rates
for current smokers. In Asian American women, Korean (19.74%, 95% CI 10.26-34.6%) have
the highest prevalence of smokers. Asians had the highest prevalence of secondhand smoke
exposure (10.39%, 95% CI 8.12-13.19%), measured by whether or not smoking was allowed in
the home, of all the races. In multivariate analysis, factors associated with current smoking
were acculturation, gender, age, general health perception, marital status, education, and
alcohol consumption.
S1418 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusion: Results dispel the notion that Asian Americans are not high risk groups ideal for
intervention programs. This has a significant implication in increasing awareness of lung
cancer screening since lung cancer screening rates are low in Asian Americans despite high
smoking prevalence in certain ethnic groups. Moreover, the study shows that both men and
women in specific Asian American groups can greatly benefit from culturally integrated
cessation programs.

CORRESPONDING AUTHOR: Katrina Obieta, MPH(c), Thomas Jefferson University,

Philadelphia, PA, 19107; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1419

B301 6:00 PM-7:00 PM

SUPPORTING SMOKERS WHO ARE NOT PREPARED TO QUIT PERMANENTLY: PRACTICE QUIT
AND DAILY CHALLENGES TEXT MESSAGING PROGRAMS PILOT

Kisha I. Coa, PhD, MPH1, Amy Sanders, M.A.1, Jessica Havlak, BA, MPH Candidate2, Tessa
Heydinger, BS 3, Augustson Erik, PhD, MPH4, Meredith Grady, MPH4

1
ICF International, Rockville, MD; 2ICF, Rockville, MD; 3ICF International, Coronado, CA;
4
National Cancer Institute, Rockville, MD

Background: Observations from research and clinical settings indicate that the challenges of
successfully sustaining cessation often leave smokers discouraged, demotivated, and
ambivalent about trying again, leading to long lags between attempts. These ambivalent
smokers are typically not the target of smoking cessation programs which often build the
intervention around a selected quit date. Such smokers may benefit from interventions that
provide motivational enhancement and skills training without the need to commit to long
term cessation. The current project is a preliminary analysis of two text messaging programs
designed to engage and support smokers not yet ready to make a sustained quit attempt.

Programs: Practice Quit (PQ) supports smokers through a time-limited (1, 3, or 5 day) quit
attempt. Daily Challenges (DC) is designed for smokers who are not ready to quit for a full day
and includes texts that suggest one cessation-related activity (e.g., resist a craving for 10
minutes) each day for 7 days. Both programs include texts which provide skills training and
values clarification.

Methods: 760 users enrolled in the programs between April and June 2016. Program data
were available for all users, and descriptive analysis were calculated.

Results: Users were primarily female (74.6%), daily smokers (95.7%), and a mean age of 37.3
years. Most users (71.6%) selected the DC program. Of those who selected the PQ program,
the 5-day option was chosen most frequently. Dropout rates were relatively low (15.9% of
users). At the end of each program attempt, users were given the option to complete another
attempt, and 24.7% of users chose to complete more than one attempt. There was some
indication of passive disengagement evidenced by declining response rates to program
assessment questions over time. 42.7% (day 1) to 8.3% (day 7) of users responded to the daily
text indicating whether they completed the daily challenge (DC), and 38.1% (day 1) to 8.7%
(day 5) responded to the daily text indicating whether they had smoked (PQ).
S1420 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions: The results of this pilot study have established feasibility and user interest in
these programs. User data also identified areas for enhancements to improve the resources
before a larger scale evaluation is launched. If successful, this type of approach may have
significant value for broadening the reach and engagement of cessation interventions.

CORRESPONDING AUTHOR: Kisha I. Coa, PhD, MPH, ICF International, Rockville, MD, 20850;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1421

B302 6:00 PM-7:00 PM

THE CLINICAL UTILITY OF DEPRESSION AND WEIGHT-RELATED SUBTYPES AMONG WEIGHT-

CONCERNED WOMEN SMOKERS

Rachel H. Salk, PhD1, Rachel P. Kolko, PhD2, Lisa J. Germeroth, PhD3, Rebecca Emery, MS3,
Michele D. Levine, PhD4

1
University of Pittsburgh Medical Center, Pittsburgh, PA; 2University of Pittsburgh School of
Medicine, Pittsburgh, PA; 3University of Pittsburgh, Pittsburgh, PA; 4University Of Pittsburgh,
Pittsburgh, PA

Weight concerns and depression symptoms are barriers to quitting smoking among women.
Weight-concerned and depressed women smokers are more likely to drop out of treatment
and have poorer cessation outcomes in treatment. However, extant research has not
examined subtypes of women who may have different symptom profiles that are amenable to
more specialized intervention. Thus, we aimed to examine whether a group-based analysis
could meaningfully capture individual differences on symptom measures of depression, body
esteem, and eating behaviors, and demonstrate clinical utility.

Weight-concerned women (N = 349; 86% white) received smoking cessation counseling and
were randomized in a double-blind, placebo-controlled trial to one of two adjunctive
counseling components and one of two medication conditions. At baseline, women
completed measures of depression, body esteem, and eating behaviors (restraint and
disinhibition). Women also completed measures to assess BMI, nicotine dependence, number
of cigarettes smoked per day, age, and education. Latent profile analysis (LPA) was used to
identify subtypes of weight-concerned women smokers based on dimensions of depression,
body esteem, and eating behaviors. ANOVA was used to characterize baseline differences
among identified subtypes.

Results supported a three-profile model: LP1 (14% of the sample; high depression and
restraint, low body esteem), LP2 (35%; moderate on all measures), and LP3 (51%; low
depression, high body esteem and disinhibition). Entropy for the model was .79 and average
latent class probabilities ranged from 90% to 91%. Groups did not differ on age, education, or
nicotine dependence. All groups differed significantly on BMI: LP1 (M= 30.55); LP2 = (M =
24.75), and LP3 (M = 28.21). Groups also differed on number of cigarettes smoked per day,
with women in LP1 smoking significantly more cigarettes (M = 24.14) than the other two
groups: LP2 (M = 19.67) and LP3 (M = 20.47).
S1422 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Among weight-concerned women smokers, heterogeneity in the symptom presentation of

depression, body esteem, and eating behaviors was captured by three distinct groups. These
groups demonstrated clinical utility, as the high-risk group (14% of the sample) consisted of
obese women who smoked over a pack of cigarettes a day. These women may need more
specialized, intensive intervention. Thus, identifying groups with distinct symptom profiles can
lead to more personalized medicine strategies to improve smoking cessation outcomes
among women.

CORRESPONDING AUTHOR: Rachel H. Salk, PhD, University of Pittsburgh Medical Center,

Pittsburgh, PA, 15206; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1423

B303 6:00 PM-7:00 PM

THE INTERACTION BETWEEN DEPRESSION AND SOCIAL ANXIETY AS A PREDICTOR OF

MOTIVATION AND CONFIDENCE TO QUIT SMOKING

Hannah L. Shoemaker, B.S. , Jennifer Dahne, PhD, Bryan Heckman, PhD, Matthew Carpenter,
PhD

Medical University of South Carolina, Charleston, SC

Background: Smokers with elevated depressive symptoms are at heightened risk for poor
cessation outcomes, including a decreased likelihood of initiating a quit attempt and an
increased likelihood of relapsing. Social anxiety (SA) is highly comorbid with depression and
may convey singular or iterative risks for poor cessation outcomes. Despite this comorbidity,
little is known about the interaction between depression and SA for predicting cessation-
related outcomes. Both motivation to quit (or to remain quit after initiating an attempt) and
confidence to quit (or remain quit) are robust predictors of sustained cessation. Thus, the
present study examined the interaction between depression and SA as predictors of
motivation to quit/remain quit (MTQ) and confidence to quit/remain quit (CTQ).

Methods: Participants included 543 smokers (28% female, 76% white, Age M(SD)=34.4 (8.7)
recruited nationally for an online survey who either recently quit (n=242) or planned to quit in
the next month (n=301). Depression was assessed via the Patient Health Questionnaire-8, SA
was assessed via the Social Phobia Inventory, CTQ and MTQ were assessed via contemplation
ladders (0-10) and nicotine dependence was assessed using the Heaviness of Smoking Index.

Results: Moderation analyses were performed using PROCESS bootstrapping procedures

(10,000 samples). Covariates included nicotine dependence and group (current smoker,
recent quitter). For MTQ, there were significant main effects of depression (b = -0.04, SE =
0.02, p = 0.03) and of SA (b = -0.07, SE = 0.03, p = 0.03), though the interaction between the
two was not significant. For CTQ, there was a significant interaction between depression and
SA (b = -0.02, SE = 0.01, p < 0.001), which was then probed at high and low levels of SA.
Depression significantly predicted CTQ at both high (b = -0.25, SE = 0.03, p < 0.001) and low (b
= -0.12, SE = 0.03, p < 0.001) levels of SA. The negative relationship between depression and
CTQ was stronger for those with high levels of SA than those with low levels of SA.

Conclusions: Results highlight the importance of symptoms of depression and SA for

predicting MTQ and CTQ. Depressed smokers and socially anxious smokers may be at
heightened risk for low MTQ. In contrast, smokers who experience both elevated depressive
symptoms and SA may be at comparatively higher risk for low CTQ than depressed smokers
S1424 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

without SA. Thus, to promote sustained cessation, interventions should consider targeting
both symptoms simultaneously.

CORRESPONDING AUTHOR: Hannah L. Shoemaker, B.S. , Medical University of South Carolina,

Charleston, SC, 29401; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1425

B304 6:00 PM-7:00 PM

UNDERSTANDING ADULT HOOKAH SMOKING THROUGH PATTERN IDENTIFICATION, 2016.

Haneen S. Abudayyeh, MPH1, Allison M. Glasser, MPH1, Amanda L. Johnson, MHS1, Amy M.
Cohn, PhD1, Theodore L. Wagener, PhD2, Jessica M. Rath, PhD, MPH, CHES3, Andrea C.
Villanti, PhD, MPH, CHES1

1
Schroeder Institute at Truth Initiative, Washington, DC; 2University of Oklahoma Health
Sciences Center, Oklahoma City, OK; 3Truth Initiative, Washington, DC

Introduction: The prevalence of hookah use has been increasing in U.S. adults; however, little
is known regarding correlates and patterns of hookah use in this population. The current
study examines correlates and the frequency of different patterns of hookah use in a national
sample of U.S. adults.

Methods: Data were drawn from Wave 10 (November 2016) respondents aged 1844 years
from the Truth Initiative Young Adult Cohort Study (n =4,091). Respondents were asked about
past hookah use; those who reported someday or everyday use were queried about number
of days used in the past 30 days and if they had ever or currently used hookah regularly (i.e.,
at least once a week, for at least a month). Weighted bivariate analyses were used to estimate
correlates and frequency of ever and past 30day hookah use, and differences between
current and non-current ever hookah users on past 30-day alcohol, marijuana, drug use, and
use of other tobacco products.

Results: Of the 21% of the sample who had ever used hookah, 9% used hookah in the past 30
days. Ever hookah users were significantly more likely than never users to report alcohol,
marijuana, and other drug use in the past 30 days (p < 0.005). Additionally, respondents who
had ever used hookah were significantly more likely to report past 30-day use of any other
tobacco product and e-cigarettes than those who had never used hookah (p < 0.01).

Compared to non-current ever hookah users, past 30-day users were also more likely to have
used other tobacco products and e-cigarettes (ps < 0.01). Few ever hookah users (8%)
reported having used hookah regularly, while 36% of past 30-day users reported regular
use (p < 0.0005). Forty percent of ever hookah users described themselves as presently using
hookah regularly; this was also higher in past 30-day users than in ever users (68.0% vs.
6.8%, p < 0.001). The majority (68%) of past 30-day users reported hookah use on fewer than
five of the past 30 days and 42% engaged in a single hookah session on those days.
S1426 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions: Ever and past 30-day hookah use is significantly associated with concurrent
tobacco, e-cigarette, and substance use. While most ever hookah users did not report
frequent (> 5 days/month) hookah use, past 30-day users describe themselves as regular
users more often than ever hookah users do. This suggests a difference in patterns of use and
uptake between potential experimenters and regular users. Identifying ways to curb regular
or weekly use may be one method to counteract the progression from experimentation to
regular use.

CORRESPONDING AUTHOR: Haneen S. Abudayyeh, MPH, Schroeder Institute at Truth

Initiative, Washington, DC, 20001; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1427

B305 6:00 PM-7:00 PM

WHEN LESS IS MORE: SUPPORTIVE POSTPARTUM SMOKING RELAPSE PREVENTION IS MORE

EFFECTIVE AND LESS COSTLY THAN AN ENHANCED VERSION

Rachel P. Kolko, PhD1, Rebecca Emery, MS2, Jennifer Grace, MS3, Andrea Kass, PhD4, Rachel H.
Salk, PhD3, Lisa J. Germeroth, PhD2, Yu Cheng, PhD2, Michele D. Levine, PhD5

1
University of Pittsburgh School of Medicine, Pittsburgh, PA; 2University of Pittsburgh,
Pittsburgh, PA; 3University of Pittsburgh Medical Center, Pittsburgh, PA; 4The University of
Chicago, Chicago, IL; 5University Of Pittsburgh, Pittsburgh, PA

Postpartum smoking is linked to deleterious outcomes for women and their infants, leading to
increased health care costs for individuals and society. Behavioral interventions for smoking
are effective but can be costly to implement; therefore, it is beneficial to understand costs to
achieve efficacy for behavioral interventions, given the implications for adoption of services.
We aimed to assess the incremental cost-effectiveness of two empirically-based interventions
designed to increase rates of sustained postpartum smoking abstinence among women who
quit smoking during pregnancy.

Women (N=300) were randomized to an intervention condition prior to delivery and received
treatment through 24-weeks postpartum. We compared two interventions (matched for time
and attention) that were adapted for the postpartum period and addressed behavioral
smoking relapse prevention. One intervention comprised an enhanced cognitive-behavioral
intervention with additional specialized focus on womens postpartum concerns about mood,
stress, and weight (STARTS).The other intervention involved a supportive behavioral
intervention (SUPPORT). Biochemically confirmed smoking abstinence was assessed at 52-
weeks postpartum. Costs to deliver the interventions from the payer, participant, and societal
(payer + participant) perspectives were calculated for each intervention. The incremental cost
effectiveness ratio (ICER) was derived to compare the interventions on the resources required
(i.e., dollars) to obtain a unit of effectiveness (i.e., % achieving abstinence).

By 52-weeks postpartum, 21% and 27% of STARTS and SUPPORT participants, respectively,
demonstrated abstinence (p=.28, number needed to treat=16.67). From the societal
perspective, the mean cost per participant for STARTS was $971 (payer=$909,
participant=$63) and for SUPPORT was $837 (payer=$760, participant=$77). Compared to
STARTS, SUPPORT was dominant, meaning it was less costly and more efficacious (ICER=-
$22.39 per % achieving abstinence).
S1428 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Results show that SUPPORT yielded similar rates of smoking abstinence in postpartum women
and was less costly than the more intensive STARTS intervention, suggesting SUPPORT may be
well suited for dissemination. These results have clinical and policy implications for
intervening in the postpartum period, as there is great importance in helping women to
increase rates of sustained abstinence from smoking via interventions that reduce
implementation costs while optimizing impact.

CORRESPONDING AUTHOR: Rachel P. Kolko, PhD, University of Pittsburgh School of Medicine,

Pittsburgh, PA, 15213; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1429

Poster Session B 6:00 PM-7:00 PM

COMPARING CHANGES TO HEDONIC HUNGER FOLLOWING A SURGICAL AND NON-SURGICAL

WEIGHT LOSS INTERVENTION

Win Guan, PhD1, Valerie H. Myers, PhD2, Melissa N. Harris, MPH1, Phillip Brantley, PhD1,
Robbie Beyl, PhD1, Robert L. Newton, Jr., PhD1

1
Pennington Biomedical Research Center, Baton Rouge, LA; 2Klein Buendel, Inc., Golden, CO

Hedonic hunger is defined as a subjective state of hunger that is prompted by exposure to

food-related cues but by definition they do not occur in response to prolonged food
deprivation." Previous research has shown a decrease in hedonic hunger following both
bariatric surgery and non-surgical behavioral interventions. However, no research has directly
compared changes to hedonic hunger for surgical and non-surgical interventions. The current
study aimed to compare changes in hedonic hunger at 6 months, 1 year, and 2 years following
bariatric surgery (N=148) and a non-surgical intensive medical intervention (N=180). Data
utilized in this study come from an observational study (Heads Up) of obese participants
undergoing either bariatric surgery (gastric banding, bypass or sleeve) or a non-surgical
intensive medical intervention (IMI) composed of a liquid diet and behavior therapy to
promote calorie restriction and increased physical activity). Participants completed the Power
of Food (PFS) scale at baseline, 6 months, 1 year, and 2 years following intervention. A mixed
effects regression model was utilized to estimate changes to hedonic hunger after adjusting
for age, sex, race, and baseline BMI. Hedonic hunger decreased dramatically 6 months
following both bariatric surgery and IMI. This decrease was significantly stronger for the
surgical group compared to the IMI group but only for the aggregate PFS score and not for the
subscale PFS scores. The change in Hedonic Hunger plateaued between 6 months and 1 year
for both groups. However, hedonic hunger for participants in the IMI group slightly increased
toward baseline levels 2 years following intervention while the decreased level of hedonic
hunger for surgical patients was sustained. The findings in this study demonstrate 1) a greater
initial decrease to hedonic hunger following bariatric surgery compared to a behavioral
intervention and 2) sustained low levels of hedonic hunger for surgical patients compared to
behavioral intervention participants. Further research will be analyzed to assess the
implications of these findings for weight loss maintenance following surgical and behavioral
interventions.

CORRESPONDING AUTHOR: Win Guan, PhD, Pennington Biomedical Research Center, Baton
Rouge, LA, 70808; [email protected]
S1430 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Friday

March 31, 2017

8:00 AM-9:15 AM
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1431

Symposium 38 8:00 AM-9:15 AM

COMMUNICATING UNCERTAIN CANCER INFORMATION IN THE ERA OF PRECISION MEDICINE

Emily B. Peterson, PhD1, Paul K.J. Han, MD, MA, MPH2, Kimberly Kaphingst, ScD3, Wen-Ying
Sylvia Chou, PhD, MPH4, Jennifer Elston. Lafata, PhD5

1
National Cancer Institute, Rockville, MD; 2Maine Medical Center Research Institute, Portland,
ME; 3University of Utah and Huntsman Cancer Institute, Salt Lake City, UT; 4Behavioral
Research Program, Division of Cancer Control and Population Sciences, National Cancer
Institute, Bethesda, MD; 5Linieberger Comprehensive Cancer Center and Eshelman School of
Pharmacy, The University of North Carolina at Chapel Hill, Chapel Hill, NC

Communication about cancer-related risk occurs frequently in clinical interactions across the
cancer control continuum from prevention, diagnosis, treatment, to end-of-life. These
conversations inevitably contain uncertain and sometimes conflicting or confusing
information. For example, consider a primary care provider recommending a mammography
screening test, a genetic counselor reviewing the return of genetic testing results, and an
oncologist discussing prognosis and treatment options: they must all navigate these
discussions with the intent to facilitate informed decision making, despite the fact that the
information they are communicating often contains caveats or may not be actionable.
Evolving national/clinical guidelines and the emphasis on personalized risk factors, rapid
advances in cancer treatment, and the promises behind the Precision Medicine movement
in cancer suggest that communication about these issues will only continue to grow more
complex and uncertain.

Our proposed symposium will highlight key research about this communication challenge with
an overarching goal of developing a broad framework to understand communication about
uncertain information in various cancer-related contexts. The panelists represent cutting-edge
science in clinical communication about uncertain or complex information in three distinct but
related domains, including cancer screening recommendations, genetic testing, and prognosis
and goals of treatment for patients diagnosed with advanced cancer. The symposium will 1)
review theories, frameworks and approaches that have been applied to communication about
uncertain information; 2) identify research priorities in cancer communication in the era of
Precision Medicine; 3) consider implications for health disparities. After individual
presentations on diverse contexts of risk communication, ample time will be allotted for a
discussion to unite the common characteristics of these areas and discuss the overarching
implications of communicating complex and uncertain cancer information.
S1432 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

The learning objectives of this symposium are: 1) Understand the role of communicating
about uncertain information on patient decision making and behavior; 2) Discuss current
priorities and challenges in communicating complex information with leading experts; 3) Gain
a broader understanding of issues related to communication about uncertain information that
transcends a specific context or research area.

Symposium 38A

UNCERTAINTY IN THE COMMUNICATION OF GENETIC AND GENOMIC INFORMATION

Kimberly Kaphingst, ScD

Many types of uncertainty are embedded within genetic risk information, including
uncertainty about penetrance (whether a gene variant will be expressed in an individual),
clinical validity (strength of association between a gene variant and disease or condition) and
clinical utility or implications for patient management. As exome and genome sequencing
become more common in the clinical setting, more variants with uncertain clinical validity and
utility and penetrance are being identified for each patient, raising the issues of what of this
uncertain and complex information to return to patients and how. This presentation will first
address these issues by presenting results from a case study. The case study was a survey of
1080 women diagnosed with breast cancer at age 40 or younger. The survey focused on
preferences for the communication of secondary findings from whole genome sequencing,
that is, findings not directly related to the breast cancer indication but that might still be of
value to patients. The survey findings showed that women ranked their interest in results with
more uncertain clinical implications (e.g., variants of uncertain significance) lower than their
interest in sequencing results with clearer clinical implications (e.g., results with implications
for prevention or medications). When asked about their preferences for timing of the delivery
of results for variants of uncertain significance, about as many women never wanted to
receive this information (32%) as women who wanted the information at or before the time
of diagnosis with breast cancer (35%). Significant predictors of wanting to receive this type of
uncertain result by the time of diagnosis included having less knowledge about the limitations
of genome sequencing (OR = 0.93; 95% CI=0.86, 0.99) and being more health conscious
(OR=1.35; 95% CI= 1.03, 1.77). The results from this case study will be discussed in the context
of a scoping review of communication of genetic and genomic risk information across the
cancer control continuum based on frameworks of uncertainty management and risk
information processing. The presentation will summarize the current state of this science and
discuss future directions for investigating uncertainty in the communication of genetic and
genomic information.

Symposium 38B
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1433

COMMUNICATION ABOUT PROGNOSIS AND GOALS OF CARE IN ADVANCED CANCER CARE

Dr. Wen-Ying Sylvia Chou, PhD, MPH

One of the major challenges in caring for patients diagnosed with advanced or recurrent
cancer concerns making decision about optimal treatment and care. One component of this
challenge is that there are significant misalignments between providers and patients with
regard to their understanding of prognosis (Prigerson et al. 2016, Gramling et al. 2016). This
misalignment occurs in the context of well-publicized advances in cancer science which
suggest the likelihood of life-prolonging treatment and which may lead patients to
overestimate their chances of survival. There are numerous challenges in communicating
prognosis, and the inevitable uncertainty associated with a patients prognosis further
complicated the clinical communication. Understanding and improving patient-provider
communication about uncertainty when discussing prognosis and goals of care should be an
urgent priority for cancer clinical practice.

This presentation describes two qualitative studies centered around the communication of
prognosis in the context of advanced cancer. The first is a qualitative discourse analysis of
transcripts of oncologist-patient conversations, in which dominant linguistic features in
oncologists language are identified. For instance, oncologists are frequently observed to use
euphemisms, optimistic expressions, and ambiguous language in discussing the status of the
disease and goals of treatment. Such expressions contribute to the limited and vague
discussions of prognosis and the sense of uncertainty in patients. The second study, based on
key informant interviews (N=19) members of multidisciplinary teams caring for patients
enrolled in cancer clinical trials, describes the diverse range of perspectives on goals of patient
care. The study findings highlight the need for communicating and managing uncertainty in
caring for patients on cancer clinical trials, particularly in balancing patients clinical goals with
research goals. The presentation concludes with a summary of the current state of science
and relevant theories in prognosis communication in cancer, with a focus on how to consider
and manage uncertainty in caring for patients with advanced cancer.

Symposium 38C

UNSPOKEN RISK AND UNCERTAINTY: THE CASE OF CANCER SCREENING

Jennifer Elston. Lafata, PhD

Recent studies highlight the risk and uncertainties associated with common cancer screening
tests, including screening mammography, prostate specific antigen (PSA), and low dose
computed tomography (LDCT). Despite known risks and uncertainties, the US population
remains enthusiastic about cancer screening in general, and many patients and physicians fail
S1434 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

to understand the risk of overdiagnosis and other uncertainties related to cancer screening.
Given the importance of physician recommendation to cancer screening use, particularly
troubling is the void in risk and uncertainty discussion at the time of most physician
recommendations for screening. As our understanding of the balance of screening risks and
benefits increases, and as cancer screening capabilities are enhanced via genomics and other
medical advances, cancer screening decisions are likely to become more complex and
personalized.

Using data from audio-recorded primary care office visits joined with pre-visit patient survey
data (N=415) and a national survey of primary care physicians (PCPs) (N=278), we highlight
current challenges to supporting informed decision making in primary care. For example,
while 86% of patients reported understanding testing pros and cons as important to their
cancer screening decisions, only 5% of colorectal cancer screening conversations included any
mention of a screening con, and only two patients explicitly asked a question about screening
risks during the consultation. The presentation will describe when and how uncertainty was
communicated when these risk conversations did occur. Furthermore, PCPs are more likely to
report a discussion of cancer screening benefits than they are a discussion of cancer screening
risks (79% vs, 64%, respectively). Likewise, PCPs are more likely to report that shared decision
making is important in cancer screening contexts where there are less uncertainties (i.e.,
relatively more evidence that benefits of screening outweigh harms as determined by the US
Preventive Services Task Force) compared to those where there is a high degree of
uncertainty or when expected balance of harms and benefits is similar. Combined, these
findings suggest that PCPs may view shared decision making as a tool to help convince
patients to adhere to recommended screening, and not as a means to provide balanced
information, including uncertain information, to patients. Ultimately, the lack of uncertainty
discussions surrounding screening tests may lead to patient misunderstanding and
unnecessary testing. Using these and other findings, an ecological framework for the
challenges of discussing risk and uncertainty in the context of cancer screening will be
presented, and the multi-levels of intervention needed for the inclusion of risk and
uncertainty within cancer screening discussions highlighted.

CORRESPONDING AUTHOR: Richard Brown, PhD, Virginia Commonwealth University School of

Medicine; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1435

Symposium 39 8:00 AM-9:15 AM

EXPANDING THE HORIZONS OF BEHAVIORAL MEDICINE WITH QUALITATIVE AND MIXED-

METHODS RESEARCH: VOICES OF DIVERSE CANCER PATIENTS

Tess Thompson, PhD, MPH1, Lailea Noel, PhD2, Tess Thompson, PhD, MPH1, Kristi Graves,
PhD3, Krystal Warmoth, PhD4, Qian Lu, Ph.D.4

1
Washington University in St. Louis, Webster Groves, MO; 2New York University, New York,
NY; 3Georgetown University, Washinton, DC; 4University of Houston, Houston, TX

Breast cancer is the most commonly diagnosed non-skin cancer among African American,
Latina, and Asian American women. This symposium examines how qualitative and mixed-
methods research can allow researchers to better understand the experiences of diverse
breast cancer patients in order to inform interventions, enhance treatment, and reduce
health disparities. The four presenters will describe their approaches to qualitative and mixed-
methods research and incorporate cross-cutting perspectives from disciplines including
medicine, public health, and social work.

The first presenter will share data from a phenomenological study of the lived experience of
African American women from a socially and economically isolated urban environment who
delayed initiation of breast cancer treatment for more than six months after identification of a
breast abnormality. This presenter will discuss how phenomenological qualitative
methodology allows for rich, in-depth data to emerge about this vulnerable group of women.
The second presenter will examine how face-to-face qualitative interviews were used to delve
more deeply into specific themes with a subsample of participants from a randomized,
controlled trial of a video narrative intervention. In this study, married African American
breast cancer survivors described how they perceived cancer to have affected their social
support, their marital relationship, and their spouses physical and mental health. The third
presenter will describe the social needs and challenges of Chinese American immigrant breast
cancer survivors based on an analysis of expressive writing from three writing sessions. The
final presenter will discuss mixed-methods results from a study of Latina breast cancer
survivors in which qualitative approaches were used to help design an intervention that was
tested in a randomized controlled trial, collect evaluation information about the intervention,
and measure the study teams engagement with the intervention.

Based on these four presentations, we will discuss the challenges and rewards of using
qualitative and mixed-methods techniques in diverse populations. We will address rationales
for using qualitative methods, techniques for ensuring methodological rigor, and ways of
S1436 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

communicating qualitative findings. The goal is to deepen researchers and clinicians

understanding of the experiences of diverse breast cancer patients in order to support
patients through diagnosis, treatment, and survivorship.

Symposium 39A

EXPLORING PARTICIPATORY HEALTH DECISION MAKING AND BREAST CANCER TREATMENT

INITIATION FOR MEDICALLY MARGINALIZED WOMEN

Lailea Noel, PhD

African American women have a 42% higher mortality rate from breast cancer than White
women. While there are many factors contributing to this disparity, the timely initiation of
treatment is growing area of interest. An under-explored area of attention for scientific
studies is how to identify and engage in care women who do not follow up with treatment
following a breast abnormality. Often these are among the most medically marginalized
women of our society who have some of the highest rates of mortality. This study explored
the lived experience of eight African American women, living in a socially and economically
isolated environment in a large metropolitan area. In all cases the initiation of treatment was
delayed for more than six months from the identification of a breast abnormality, with a delay
in the initiation of treatment of more than a year in four of the eight examples. In accordance
with phenomenological qualitative methodology, the interviews were open-ended, allowing
for rich, in-depth data to emerge relatively unbiased by the assumptions and expectations of
the researcher. Key findings suggest that engaging patients in health care decision making
necessitates a better understanding of the processes involved in treating breast cancer, the
impact that the side-effects may have on her quality of life, and access to patient support
services to counter these side-effects. The findings from this study highlight that most of the
communication about treatment initiation occurred at the primary care level, prior to
initiating contact with oncology services. Furthermore, the context within which they
experienced the phenomenon is structured by underlying conceptualizations of chronic
traumatic events coupled with community resource deserts. These results indicate that future
interventions focused on engaging women in care decisions and decreasing socioeconomic
barriers to the initiation of the treatment for breast cancer should take place between
primary care providers, oncologists, and their patients and focus on de-mystifying the process
of treatment. Incorporating this information, as a type of barrier-focused community-based
intervention, or improved case management services to assist cancer patients in the primary
care setting, may impact cancer mortality outcomes.

CORRESPONDING AUTHOR: Lailea Noel, PhD, New York University; [email protected]

Symposium 39B
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1437

WERE IN THIS TOGETHER: SPOUSAL SOCIAL SUPPORT IN AFRICAN AMERICAN BREAST

CANCER PATIENTS

Dr. Tess Thompson, PhD, MPH

Social support provides health benefits both in cancer patients and in the general population.
Spouses are often key sources of social support for breast cancer patients, and patients often
provide their partners with support as well. Recent research has also shown small but
significant effects of a cancer diagnosis on the physical and mental health of spouses of cancer
patients. Less is known, however, about how cancer affects relationships in African American
couples. This qualitative study examined how married African American breast cancer
survivors perceived cancer to have affected their social support, their relationship with their
spouse, and their spouses physical and mental health. We interviewed a subset of 15 married
women who participated in a randomized controlled trial of a culturally tailored cancer
communication intervention. We collected demographic data, and participants responded to
semi-structured interview questions. Two coders coded interview transcripts independently
and then reached consensus on codes in order to conduct a thematic analysis. At the time of
interview, participants mean (SD) age was 60.2 (7.4) and mean length of marriage was 23.3
(12.4) years. All reported their spouses were African American. Eight had early-stage cancer,
11 received breast-conserving surgery, and 7 received chemotherapy. Eight participants
believed their relationships stayed the same or became better after their diagnosis. Five
reported marital problems, however, and 3 thought that their cancer had negatively affected
their marriage. Most participants were uncertain whether their cancer diagnosis and
treatment had affected their husbands physical and emotional health, often because
husbands did not discuss their reactions to their wives cancer. Many participants reported
their husbands provided both emotional and tangible support, and several said their husbands
helped them adjust to sexual side effects of treatment and changes in body image. Many
participants also reported providing support to their husbands, and many believed that
clinicians could play an important role in helping couples adapt together to a cancer diagnosis
and demands of treatment. Clinicians should take into account patients social context and
provide referrals for couples with marital difficulties and for patients and partners who would
benefit from psychosocial interventions.

CORRESPONDING AUTHORS: Maxine Davis, MSW, Washington University in St. Louis;

[email protected]; Maria Perez, MA, Washington University in St. Louis;
[email protected]; Donna Jeffe, PhD, Washington University in St. Louis;
[email protected]

Symposium 39C

PATIENT-REPORTED OUTCOMES AMONG LATINA BREAST CANCER SURVIVORS FOLLOWING

THE NUEVA VIDA INTERVENTION
S1438 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Kristi Graves, PhD

Introduction: Few research studies have evaluated evidence-based psychosocial interventions

to improve patient-reported outcomes among Latina breast cancer survivors and caregivers.
Using strong community-academic partnerships and mixed methods approaches, we sought
to examine patient-reported outcomes among a diverse sample of Latina breast cancer
survivors following the Nueva Vida Intervention.

Method: We recruited Latina breast cancer survivors and their caregivers through 4
community based-organizations across the U.S. Following feedback from qualitative
interviews with survivors and caregivers, we refined the intervention. We then conducted an
RCT in which survivor-caregiver dyads were randomized to either an 8-session psychosocial
intervention called Nueva Vida or to usual services (e.g., support groups, patient navigation).
We completed quantitative assessments at baseline, immediately post-intervention and 6-
months post intervention to measure patient-reported outcomes using the Patient Reported
Outcomes Measurement Information System (PROMIS). We collected intervention evaluation
data using brief exit interviews from study participants and intervention process surveys from
interventionists. We measured team engagement using both quantitative and qualitative
approaches.

Results: We randomized 135 Latina survivors and their caregivers to the Nueva Vida
Intervention (n=70) or usual care (n-65). Latinas were diverse in terms of their country of
origin (>80% immigrants) and years in the US. To date, 78 survivors have completed the 6-
month follow-up. Compared to survivors in usual care, intervention participants reported
statistically and clinically significant lower anxiety (t(75)=2.03, p<.05); trends were noted for
less fatigue (p<.10) among intervention participants. Exit interviews with participants
indicated a high level of satisfaction with the intervention; interventionists provided feedback
on strategies to further tailor the intervention to their specific Latino population. Team
members reported consistent and strong engagement through the study period.

Conclusions: Early results from the Nueva Vida Intervention trial indicate the ability to recruit,
retain and improve specific patient reported outcomes among a diverse sample of Latina
breast cancer survivors. Future efforts to evaluate evidence-based interventions with Latino
families facing cancer will benefit from strong community-based partnerships and efforts to
ensure cultural and linguistic relevance of the intervention.

CORRESPONDING AUTHORS: Ysabel Duron, BA, Latinas Contra Cancer;

[email protected]; Ivis Febus-Sampayo, SHARE;
[email protected]; Migdalia Torres, LCSW-R, Gilda's Club New York City;
[email protected]; Claudia Campos, MA, Nueva Vida, Inc.; ccampos@nueva-
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1439

vida.org; Christina Rush, MA, UC Denver; [email protected]; Adriana Serrano, MA,

Georgetown University; [email protected]

Symposium 39D

EXPLORING THE SOCIAL NEEDS AND CHALLENGES OF CHINESE AMERICAN IMMIGRANT

BREAST CANCER SURVIVORS

Krystal Warmoth, PhD

Purpose: Little is known about the psychological wellbeing and social barriers among
immigrant Chinese American breast cancer survivors. The aim of the present study was to
explore the social needs and challenges of Chinese American immigrant breast cancer
survivors.

Methods: This study used the expressive writing approach to explore the experiences among
27 Chinese American breast cancer survivors. The participants were recruited through
community-based organizations in Southern California, most of whom were diagnosed at
Stages I and II (33% and 48% respectively). All the participants had been living in the USA for
at least 5 years (on average 19 years). Participants were asked to write three 20-minute
essays related to their experience with breast cancer (in 3 weeks). Participants writings were
coded with line-by-line analysis, and categories and themes were generated.

Results: Emotion suppression, self-stigma, and perceived stigma about being a breast cancer
survivor were reflected in the writings. Interpersonally, participants indicated their reluctance
to disclose cancer diagnosis to family and friends and concerns about fulfilling multiple roles.
Some of them also mentioned communication problems with their husbands. Related to life in
the USA, participants felt unfamiliar with the health insurance and health care system, and
they encountered language barriers.

Conclusions: Counseling services addressing concerns about stigma and communication

among family members may benefit patients adjustments. Tailor-made information in
Chinese about diagnosis and treatment for breast cancer and health insurance in the USA,
may also help patients go through the course of recovery.

CORRESPONDING AUTHORS: Bernice Cheung, University of Houston; [email protected];

Jin You, PhD, Wuhan University; [email protected]; Nelson Yeung, PhD, The Chinese
University of Hong Kong; [email protected]; Qian Lu, PhD, University of Houston;
[email protected]
S1440 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 40 8:00 AM-9:15 AM

USING E-HEALTH COMMUNICATION AND TECHNOLOGY TO REDUCE HEALTH CARE

DISPARITIES

Charles R. Jonassaint, PhD, MHS1, Linda Fleisher, PhD, MPH2, Betina Yanez, Ph.D.3

1
University of Pittsburgh, Pittsburgh, PA; 2Children's Hospital of Philadelphia, Philadelphia, PA;
3
Northwestern University Feinberg School of Medicine, Chicago, IL

Racial and ethnic minorities are at an increased risk for poor health, yet receive lesser quality
healthcare than their white counterparts. Despite the availability of evidence-based
interventions, there are significant disparities in the reach and impact of these treatments for
underrepresented minority groups.

Over the past decade, mobile technology ownership and Internet use has drastically
increased, particularly among minority groups. The pervasiveness of technology across race,
ethnicity and socioeconomic strata has led to excitement surrounding the potential for e-
Health to bridge the disparities gap and deliver evidence-based care to underserved, hard-to-
reach populations. However, there is limited literature examining the effectiveness of e-
Health interventions among minority groups. Further, the majority of available e-Health
interventions have not been developed with a focus on reaching underrepresented patients
and may ignore the needs and sociocultural barriers faced by racial and ethnic minorities.
Tailoring e-Health intervention for racial and ethnic minorities may improve their
acceptability, effectiveness and future dissemination.

In this symposium, we first present a systematic review examining the feasibility and efficacy
of cancer-related e-Health interventions among racial and ethnic minorities and highlight the
needs and barriers to racial and ethnic minorities utilization of e-Health technologies. Next,
we present data from a large, pragmatic e-Health intervention trial aimed at improving mental
health outcomes among a racially diverse group of primary care patients. Finally, we review
the process of designing a user-centered intervention to enhance clinical trial participation
among African American cancer patients.

Overall, this symposium demonstrates that e-Health interventions have the potential to
improve health among underserved groups and bridge the health care disparities gap.
However, there are several critical issues to consider regarding the design and
implementation of eHealth to ensure these interventions are maximally effective with
underrepresented groups. Future e-Health research should adopt user-centered approaches
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1441

to intervention development that take into account the specific needs of the target
population.

Symposium 40A

ACCEPTABILITY AND IMPACT OF COMPUTERIZED COGNITIVE BEHAVIORAL THERAPY FOR

DEPRESSED AND ANXIOUS AFRICAN AMERICANS AND WHITES

Dr. Charles R. Jonassaint, PhD, MHS

Background: Access to effective treatment fordepression and anxiety in primary care is often
limited, particularly for minority patients who also experience racial disparities in treatment
outcomes. Recent trials demonstrate that computerized cognitive behavioral therapy (CCBT)
programs may help improve access to treatment. However, it is unclear if these programs are
equally acceptable to African Americans (AAs) as Whites, and whether these groups
experience comparable benefits. We examined these questions among patients enrolled in
our NIMH-funded trial of online collaborative care for mood and anxiety disorders.

Methods: Patients aged 18-75 with current depression, panic and/or generalized anxiety
disorder were referred to our trial from 26 Pittsburgh-area primary care practices. We
randomized protocol-eligible and consenting patients who reported elevated levels of mood
and/or anxiety symptoms (PHQ-9 or GAD-7 10) with Internet access to receive their doctors
usual care or one of two other groups (1-3-3 ratio) that included 6-months of care manager-
guided access to the U.S. version of the 8-session Beating the Blues CCBT program. Blinded
assessor determined PROMIS depression and anxiety outcomes by phone at 6-months follow-
up.

Results: Between 8/1/12 and 9/30/14, we randomized 91 AA and 499 White participants to
our two CCBT study arms (mean age: 43 years; SD=14; range 18-75). AA patients, compared
to Whites, had higher baseline PHQ-9 scores (14.4 vs. 13.1; pp=0.08 and 4.7 vs. 5.5; p=0.03,
respectively). However, AAs showed a greater 6-month decrease in depression (-4.76 vs. -
1.73; p=.024) and anxiety symptoms (-5.43 vs. -1.44; p=.083) compared to whites.

Conclusion: Internet-delivered and care manager-guided CCBT is an acceptable and effective

means to provide depressed and/or anxious AA patients with access to mental health care at
scale and potentially reduce racial disparities in treatment outcomes.

CORRESPONDING AUTHORS: Bea Belnap, Dr Biol Hum, University of Pittsburgh;

[email protected]; Jordan Karp, University of Pittsburgh; [email protected]; Abebe
Kaleab, PhD, University of Pittsburgh; [email protected]; Bruce Rollman, MD, University of
Pittsburgh; [email protected]
S1442 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 40B

BUILDING A BETTER E-HEALTH INTERVENTION: USING FORMATIVE RESEARCH TO GUIDE THE

DEVELOPMENT OF AN E-HEALTH COMMUNICATION TOOL

Linda Fleisher, PhD, MPH

Designing salient and relevant e-Health interventions requires target audience insights
throughout the development process and using formative research approaches can provide
rigor and substance to those insights. Although best practice approaches and current efforts
in patient engagement support patient and consumer input, a systematic approach that
includes patients perspectives on the eHealth tool in the concept phase (refining the
problem), prototype development and testing and early pilot testing is often not actualized.
This presentation will include an overview of formative approaches in e-Health intervention
development and case study. mychoice, an e-Health tool to increase informed decision about
clinical trials for African American cancer patients was developed using a systematic, patient
engagement process to ensure that the intervention was relevant, appropriate and addressed
the complex and emotional issues related to treatment trials. The formative research
included 16 in-depth interviews with African American cancer patients who had and had not
participated in clinical trials and a perceptual mapping survey to identify key messages with 41
African American cancer patients. The mychoice prototype was then shown to African
American cancer patients for additional feedback and user testing. This presentation will
provide practical approaches and describe in detail the methods, data collection strategies,
recruitment approaches, the application of the formative research and user testing, and
patient insights across the development and design of the mychoice communication tool.

Symposium 40C

A SYSTEMATIC REVIEW OF E-HEALTH INTERVENTIONS IN ONCOLOGY FOR DIVERSE

INDIVIDUALS

Betina Yanez, Ph.D.

Background: The number of e-health interventions in oncology has been increasing over the
past decade and research has demonstrated that these interventions are feasible and can
yield favorable health outcomes. Although racial/ethnic minority patients are accessing the
Internet and using technology devices at equivalent or higher rates of non-minority patients,
it is not well documented whether e-health interventions are feasible, acceptable, and
efficacious among these groups. Therefore, this systematic review sought to elucidate the use
of e-health, oncology-related interventions in diverse populations. Methods: Using methods
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1443

consistent with PRISMA guidelines, we identified 772 articles published between January 2000
and June 2016. Of these, 14 articles met inclusion criteria for this review. Results: Of these 14
articles, 86% included African Americans, 13% included Hispanics, and 7% included Asian
Americans. Fifty-seven percent of the articles focused on cancer prevention, 29% on cancer
control, and 14% on cancer survivorship. All articles evaluated a web-based computer or
tablet intervention. Sixty-four percent of the 14 articles reported on the efficacy of the e-
health intervention and thirty-six percent of the 14 articles only investigated usability,
feasibility, acceptability, and technology use as outcomes. Four e-health interventions showed
increased knowledge and cancer screening intentions, but did not increase cancer-screening
behaviors relative to a control condition. All 3 of the efficacy studies focusing on cancer
control and survivorship revealed a significant, favorable effect of the e-health intervention
(e.g., decreased depression, quality of life) when compared to a control condition. The extent
to which participants self-selected into, interacted with, and used various components of the
interventions varied by race/ethnicity. Conclusion: Findings suggest that e-health
interventions among minorities are feasible and acceptable. However, findings regarding the
efficacy of e-health interventions among racial/ethnic minorities are equivocal, with
interventions focusing on cancer control and survivorship demonstrating better efficacy than
cancer prevention interventions. Although e-health interventions demonstrate the potential
to reduce health disparities, more research is needed to delineate the mechanisms by which
researchers can adapt e-health interventions to best serve the distinct needs of racial/ethnic
minorities.

CORRESPONDING AUTHORS: Cody Boland, B.A., Northwestern University Feinberg School of

Medicine; [email protected]; Frank Penedo, Ph.D., Northwestern University
Feinberg School of Medicine; [email protected]
S1444 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 41 8:00 AM-9:15 AM

MODELS OF INTEGRATING BEHAVIORAL HEALTH SERVICES IN PRIMARY CARE: GOLD

STANDARD OR SUBSTANDARD?

Sherri Sheinfeld Gorin, PhD, FSBM1, Kathryn E. Kanzler, PsyD, ABPP2, Jrgen Untzer, MD,
MPH, MA, BS3, Rodger Kessler, ABPP4, Deborah Cohen, PhD5, Jeffrey Goodie, Ph.D., ABPP6

1
New York Physicians Against Cancer, http://nypac.center/, New York, NY; 2University of
Texas Health San Antonio, San Antonio, TX; 3University of Washington School of Medicine,
Seattle, WA; 4The University of Vermont, Burlington, VT; 5Oregon Health and Science
University, Portland, OR; 6Uniformed Services University, Gaithersburg, MD

Most behavioral and mental health care occurs in the primary care setting. Psychosocial issues
are involved in approximately 70% of primary care visits, with conditions such as depression
among the most common complaints; these symptoms are often comorbid with other chronic
diseases such as diabetes. While greater collaboration between behavioral health (including
mental health and substance use) and primary care clinicians, as supported in the Affordable
Care Act, could mitigate symptom-related negative outcomes for patients, collaborations are
few in practice. Further, few of the existing integrated care programs are based on research
evidence. The symposium will present major extant models of behavioral health in primary
care, and review research and practice-based evidence for their uptake and effectiveness in
diverse primary care settings. Presenter 1 will review the findings from RCTs, large-scale
implementation trials, and practice evidence for the collaborative care approach. He will
describe adaptations of the collaborative care approach for diverse primary care settings and
populations, as well as in delivering its core components, for example, by tele-mental health.
Presenter 2 will describe how to integrate behavioral health within the current fragmented
health system, with the findings from a mixed methods study of 19 US based primary care
practices and community mental health centers. She examines structures, processes, clinical
workflows, and the interactional behaviors of primary care and behavioral health clinicians
integrating care. Relying on research findings from national implementation trials and national
expert panel consensus, Presenter 3 will focus on the gaps in current integration research. He
will address the comparative effectiveness of different models and the degree of
organizational integration. He will propose a framework for the future of research in
integrated primary care settings.

Symposium 41A
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1445

MOVING FROM COLLABORATIVE TO STEPPED CARE: EVIDENCE AND TRANSLATION

Dr. Jrgen Untzer, MD, MPH, MA, BS

The evidence is mounting from RCTs and large-scale implementation studies for intervention
approaches to caring for individuals with common mental disorders in primary care. The
presenter will examine the evidence for collaborative care, an approach in which primary care
providers are supported by a behavioral health professional and a psychiatric consultant to
treat a population of patients with common mental disorders such as anxiety or depression in
primary care settings. The presenter will also review adaptations of core components of the
model for diverse primary care settings and populations and variations such as delivering core
components of the care using tele-mental health care. Finally, the presenter will discuss the
translation of research evidence to practice experience via implementation of evidence-based
programs in diverse primary care settings. This presentation will explore barriers and
facilitators to implementation from policy, regulatory, and cultural perspectives. Based on the
current literature and clinical data, a stepped approach to integrating care will be proposed. A
stepped approach to integrating care has the potential to reach and benefit a large number of
individuals with common behavioral health challenges (e.g., mental health and substance
abuse) in primary care.

Symposium 41B

WHY ARE WE TALKING ABOUT INTEGRATION POLICY WHEN WE DO NOT HAVE COMPELLING
INTEGRATION METRICS?

Dr. Rodger Kessler, ABPP

Not much has changed in research and practice since the 2008 AHRQ report (Butler and Kane
et al 2008) concluding that we know one effective model of care for treating depression and
anxiety in primary care. While, in fact, there are a number of exemplar practices and
organizations, we still know little about models of care. The current movement to put
integrated practice into policy is premature due to lack of solid data and metrics. We cannot
identify the key elements that contribute to positive patient outcomes, nor do we know the
appropriate intervention dose(s) for specific patient populations and, at what cost. Further,
the prevalent research paradigm does not address between-model comparative effectiveness
nor does it address necessary elements or dose(s) of organizational integration. We are
further hindered by a proliferation of checklists and data collection tools that have no
psychometric validation and thus, no standardized comparative metrics. This presentation will
discuss the findings from national and local trials and national expert panels to identify
research gaps and to suggest a framework that will allow us to move past Butler and Kanes'
S1446 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

observed limitations into the future.

Symposium 41C

FITTING THE INTEGRATION APPROACH TO THE LOCAL PRACTICE CONTEXT

Dr. Deborah Cohen, PhD

Compelling research evidence, provider and patient needs are driving the integration of
health care. What is urgently needed now is empirical data on how to integrate physical,
behavioral, emotional, and substance use care (hereafter referred to as behavioral health)
within the current fragmented health system. We studied 19 US based primary care practices
and community mental health centers that varied with respect to type, size, ownership, and
location. Using mixed methods, including direct observation, interviews, surveys, assessment
of expenditures, and Reach (e.g., RE-AIM), we comprehensively examined structures,
processes, clinical workflows, and the interactional behaviors of primary care and behavioral
health clinicians integrating care. We also evaluated the impact of practice integration efforts
on expenditures to implement innovations, Reach, and patient outcomes. Our findings show
that 5 key organizing constructs (integration REACH, establishing care pathways for a range of
patient illness, approaches to patient transitions, location of workforce, and mental model for
integration) intertwine to produce locally adapted approaches to integrating care. We
observed significant decreases in adjusted mean PHQ-9 among patients. This finding was
corroborated by accounts of patients experience of care, where they reported the benefits of
receiving integrated care such as acquiring skills to cope with adverse situations. Additionally,
we show that average start-up and ongoing effort expenses for integrating care are non-
trivial. These results, together with a robust body of evidence from RCTs, makes a compelling
case for practices to adopt integrated care and for payers to make this model feasible through
finance reform.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1447

Symposium 42 8:00 AM-9:15 AM

LEVERAGING TECHNOLOGY TO SUPPORT VULNERABLE AND DISADVANTAGED PEOPLE WITH

DIABETES: A REVIEW, EXAMPLES, AND RECOMMENDATIONS

Lindsay S. Mayberry, MS, PhD1, Courney R. Lyles, PhD2, Monica E. Peek, MD, MPH, MSc3,
Brian Oldenburg, BSc, MPsychol, PhD4, Chandra Y. Osborn, PhD, MPH5

1
Vanderbilt University Medical Center, Nashville, TN; 2University of California San Francisco,
San Fransisco, CA; 3University of Chicago, Chicago, IL; 4The University of Melbourne,
Melbourne, Victoria, Australia; 5Informed Data Systems Inc., Nashville, TN

Technologies are being used to engage, educate, and support people with self-care. Hard-to-
reach, low-income, less-educated, and/or racial/ethnic minorities who are often more
affected by chronic health conditions, and, once diagnosed, experience more difficulty w/self-
management, creating a cascade of costly health problems. Mobile self-care support can
reach a wide audience, including the underserved. However, rigorous and novel methods are
needed to design and evaluate mobile self-care support among vulnerable populations. A
combination of limited access to certain types of devices (smartphones, smartwatches),
unique barriers to self-care, and low engagement with self-care support are common among
disadvantaged groups. These and other factors should be identified and addressed in the
design, development, implementation and evaluation of mobile self-care support for
vulnerable populations. This symposium will therefore provide:

- A case study of adapting tablet-collected patient experience data among persons with
low-literacy

- An overview of a hybrid implementation-effectiveness approach to evaluating two

mobile health (mHealth) interventions in community health clinics

- A review of the literature on internet and mobile interventions to improve diabetes

control among disadvantaged and vulnerable people with type 2 diabetes

Each speaker will provide recommendations for improving the design and evaluation of
technology-delivered tools for disadvantaged and vulnerable populations with diabetes.

Symposium 42A
S1448 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

BUILDING SUSTAINABILITY ALONGSIDE EVIDENCE FOR MOBILE HEALTH INTERVENTIONS IN

UNDERSERVED PATIENTS

Dr. Lindsay S. Mayberry, MS, PhD

Background: Technology-based interventions are challenging among underserved patients

and must reach end-users quickly to remain relevant and for evidence-based health
technologies to keep pace with marketed ones. We used community engagement and
iterative mixed-methods to understand and overcome implementation barriers throughout
the design and evaluation of two diabetes support mHealth interventions in underserved
populations.

Methods: The REACH study examines the effectiveness of two mHealth for improving self-
care and glycemic control among adults with type 2 diabetes (T2D) receiving care from
Federally Qualified Health Centers (FQHCs) in Nashville, Tennessee. The two interventions
evaluated are REACH, a tailored, theory-based text messaging intervention, and FAMS, a
family-focused intervention which incorporates phone coaching and text messages to a
support person (e.g., friend or family member).We employed community-based strategies to
develop a sustainable and implementable intervention: a) eliciting and incorporating input
from FQHCs and patients before the evaluative trial to inform the content and form of the
interventions; b) sustaining FQHC engagement throughout the trial; c) expanding to new
clinics at each stage; and d) gathering information about barriers and facilitators to
implementation at each stage.

Results: Through iterative mixed-methods (from intervention conception, pilot, refinement,

usability/feasibility testing) we made intervention design decisions (eg, using resources
patients and clinics already have; automating text messaging for minimal time, financial, or
personnel costs) and trial design decisions (eg, an active control; not providing or paying for
participants cellphones/plans, allowing participants to self-tailor the intervention) to enhance
clinic engagement and intervention sustainability. During the trial, we maintain community
partnerships via clinic/community events and quarterly study newsletters tailored for patients
and, separately, clinic staff. By expanding from two clinics (formative research and piloting) to
four clinics (usability/feasibility testing) to seventeen clinics (for the trial), we created
intervention elements likely to be preservable across settings. To sustain engagement and
momentum generated by the initiation of the evaluative trial, we will conduct rolling focus
groups with FQHC providers and administrators and participants completing the trial to
discuss barriers and facilitators to sustainability should REACH and/or FAMS prove effective.

Conclusion: This presentation will discuss design decisions and provide a clear example of a
hybrid effectiveness-implementation approach to mHealth among disadvantaged patients.

CORRESPONDING AUTHORS: Chandra Osborn, PhD, One Drop; [email protected];

Lydnsay Nelson, PhD, Vanderbilt University Medical Center; [email protected];
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1449

Ken Wallston, PhD, Vanderbilt University School of Nursing; [email protected];

Sunil Kripalani, MD, MPH, Vanderbilt University Medical Center;
[email protected]

Symposium 42B

IN-DEPTH CASE STUDY OF ADAPTING PATIENT EXPERIENCE DATA COLLECTION FOR DIVERSE
PATIENTS USING TABLETS IN CLINIC

Courney R. Lyles, PhD

Background: Healthcare policy supports the inclusion of patient experience in quality

measurement and reporting. However, response rates to the gold-standard, paper-based
Consumer Assessment of Healthcare Providers and Systems (CAHPS) surveys are low (<25%),
and differ by racial/ethnic and language subgroups.

Methods: We sought to 1) design and implement tablet-based CAHPS administration in safety

net clinical practices in the San Francisco Department of Public Health, and 2) conduct
formative qualitative work to create shorter, lower literacy patient experience items
(informed by CAHPS) and explore broader concepts of patient experience among vulnerable
patients. We created a visually attractive, simple-to-use tablet interface with multi-lingual
questionnaires. Next, we completed 25 in-depth interviews with patients to validate a new
literacy-appropriate tablet questionnaire and compare it to the paper version.

Results: Over 6 months, 400 patients completed tablet-based CAHPS questionnaires in 4 SF

clinics. The overall response rate and representation of Latino, Black, and younger
respondents was higher for tablet vs. paper-based collection. During the interviews, paper vs.
tablet CAHPS versions performed similarly, but patients strongly preferred the tablet even
those without previous experience with using mobile technology. The final tablet application
improved on the CAHPS items, by 1) reducing the number of items from 31 to 17 while
preserving core domains, 2) reducing the reading level of survey questions from 7th to 5th
grade, and 3) adding open-ended questions.

Conclusions: Federal policy supporting patient experience data collection is moving forward.
This study provides clear next steps to ensure underrepresented and vulnerable patient
perspectives are engaged and represented in this process.

CORRESPONDING AUTHORS: Lina Tieu, MPH, UCSF; [email protected]; Erin Curtis, SF

Department of Public Health; [email protected]; Urmimala Sarkar, MD, MPH, UCSF;
[email protected]

Symposium 42C
S1450 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

MHEALTH & INTERNET INTERVENTIONS FOR DISADVANTAGED PEOPLE WITH TYPE 2

DIABETES: CONSIDERATIONS & RECOMMENDATIONS FOR RESEARCH

Dr. Monica E. Peek, MD, MPH, MSc

Background: Mobile and Internet-based interventions are supporting patients self-care

efforts and, in turn, their glycemic control, but knowledge about the acceptance and benefits
of technology-delivered interventions among disadvantaged patients with T2D is surprisingly
limited.

Methods: We conducted a review of experimental or quasi-experimental studies evaluating

the effects of technology-delivered interventions on the glycemic control (A1c) of
disadvantaged adults with T2D from 2011-2015. We operationalized being disadvantaged as
samples with 50% patients with low socioeconomic status (SES), living in rural settings, or
racial/ethnic minority.

Results: Twelve studies were identified and summarized to characterize the state of
knowledge with recommendations to advance the field in three key areas: the development
and evaluation of technology-delivered interventions for (1) low SES and rural patients, (2)
racial/ethnic minorities, and (3) patients with limited literacy and numeracy. Within-group
improvements in A1c were consistently demonstrated in both mobile and Internet-based
interventions, however, trials have not found between-group differences. Three studies
report reductions in inpatient or emergency department use.

Conclusion: Current evidence suggests that mHealth interventions may be effective at

improving diabetes outcomes within vulnerable populations. Novel study designs are needed
to address questions of reach and cost. Future studies should identify best practices for
developing and delivering effective content to disadvantaged patients, quantifying the reach
of technology interventions to vulnerable patients, and evaluating cost-effectiveness.

CORRESPONDING AUTHORS: Lindsay Mayberry, PhD, Vanderbilt University Medical Center;

[email protected]; Courtney Lyles, PhD, University of California, San
Francisco; [email protected]; Brian Oldenburg, PhD, University of Melbourne;
[email protected]; Chandra Osborn, PhD, MPH, One Drop;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1451

Symposium 43 8:00 AM-9:15 AM

FEASIBILITY STUDIES TO INFORM PHYSICAL ACTIVITY AND WEIGHT MANAGEMENT

INTERVENTIONS IN WOMEN: THREE REAL-WORLD EXAMPLES

Scherezade K. Mama, DrPH1, Danielle Symons Downs, Ph.D.2, Danielle Arigo, Ph.D., Licensed
Psychologist3, Shannon L. Mihalko, Ph.D.4

1
The Pennsylvania State University, State College, PA; 2The Pennsylvania State University,
University Park, PA; 3The University of Scranton, Scranton, PA; 4Wake Forest University,
Winston Salem, NC

Feasibility studies are the foundation for intervention studies and are used to explore whether
an intervention should move forward to efficacy testing. However, there are not set standards
to guide the design and evaluation of feasibility studies in behavioral medicine. Feasibility
studies are needed but are often overlooked as the first step in physical activity and weight
management intervention development, where it may be necessary to test different
recruitment and research strategies, such as intervention dosage, measurement methods and
tools, and user acceptability of intervention approaches. This is especially important when
working with hard to reach or underserved populations whom may require more tailored or
targeted recruitment and intervention approaches. This symposium presents examples of
physical activity and weight management feasibility studies among three distinct samples of
women. The first presentation will explore the feasibility of a wearables-supported physical
activity intervention for women. The second presentation will describe the feasibility of
conducting an intervention to prevent weight gain in breast cancer survivors during active
treatment. The third presentation will explore the feasibility of an adaptive intervention to
manage weight in overweight/obese pregnant women and will describe how feasibility
outcomes were used to inform the development of the efficacy study. The symposium will
conclude with a discussion of the areas of focus of feasibility studies, feasibility study designs
and outcomes, and how feasibility studies are used in a two-phased approach to inform the
development of intervention efficacy and effectiveness studies.

Symposium 43A

INFLUENCE OF A FEASIBILITY STUDY ON THE DESIGN OF AN INDIVIDUALLY-TAILORED,

ADAPTIVE INTERVENTION TO MANAGE WEIGHT IN OVERWEIGHT/OBESE PREGNANT WOMEN

Danielle Symons Downs, Ph.D.

S1452 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Half of all pregnant women in the U.S. begin their pregnancies overweight or obese (OW/OB)
and 60% of these women exceed gestational weight gain (GWG) guidelines. Despite focused
prevention efforts, interventions have generated limited evidence to prevent GWG outside of
the Institute of Medicine (2009) guidelines, suggesting an important need for innovative
weight management interventions in this population. An individually-tailored intervention
that promotes healthy eating, physical activity, weight management, and self-monitoring and
adapts to the unique needs of OW/OB pregnant women over the course of pregnancy may be
necessary to manage GWG. However, little to no past research has used an adaptive
intervention to manage weight in pregnant women. We developed a 2-phase approach to
effectively and efficiently manage GWG whereby phase I (feasibility study) directly informed
the development of the phase 2 randomized control trial (RCT). The goals of this presentation
are to: (a) describe the conceptual design of the adaptive intervention and explain the
feasibility and RCT studies; (b) discuss findings and highlight key lessons learned from
implementing the feasibility study; and (c) illustrate how the feasibility study informed the
RCT. The feasibility study (N=22) OW/OB pregnant women randomized to 1 of 7 dosages over
a brief 6-week period) tested user acceptability of components (education, goal-setting, self-
monitoring, healthy eating/physical activity behaviors), dosages, and intensive data collection
procedures. Key aspects of the feasibility study that informed the RCT design were subject
inclusion criteria/recruitment strategies, user acceptability of dosage components and
intensity, use of mHealth tools for real-time feedback, data collection procedures
(timing/schedule of measures), and decision rules for adapting the intervention. The RCT
(N=30 OW/OB pregnant women) is currently underway and preliminary findings from this
study will also be discussed.

CORRESPONDING AUTHORS: Jennifer Savage, Ph.D., Penn State Univ; [email protected]; Daniel
E. Rivera, Ph.D., Arizona State Univ; [email protected]; Krista S. Leonard, Penn State Univ;
[email protected]; Abigal M. Pauley, MS, Penn State Univ; [email protected]; Emily H Hohman,
Ph.D., Penn State Univ; [email protected]; Penghong Guo, Arizona State Univ;
[email protected]

Symposium 43B

TESTING THE FEASIBILITY OF FITBIT AND ONLINE PARTNERSHIPS TO PROMOTE PHYSICAL

ACTIVITY AMONG WOMEN

Dr. Danielle Arigo, Ph.D., Licensed Psychologist

Regular physical activity (PA) reduces health risks, yet the majority of U.S. women fall far short
of recommended PA levels. Women endorse unique barriers to PA, including lack of
motivation and lack of social support. Wearable PA tracking technology, which shares users
PA data in online social networks, has the potential to address the need for scalable PA
programs for women. However, two key aspects of this potential are not clear: (1) how best
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1453

to use wearables to promote PA among women, and (2) whether women will use and respond
to wearables as intended. This presentation will describe a series of studies designed to assess
the feasibility and acceptability of a wearables-supported, online PA program for women.
Novel features of this program were the use of Fitbit and an associated online social network,
and the assignment of PA partners for support and accountability. Feasibility outcomes
included the ability to meet recruitment targets, enrollment and retention, adherence to
technology guidelines (Fitbit daily wear, number of posts to the online message board),
frequency of partner communication, and perceived utility of/barriers to partner
communication. Acceptability outcomes included ratings of participant satisfaction and
willingness to recommend the program to other women. In Study 1, conducted entirely
online, participants (n= 12, MAge=46, MBMI=32.60 kg/m2) received a webinar to introduce
behavioral PA promotion skills and communicated with partners via private message boards
on the Fitbit social network. This study showed positive feasibility/acceptability outcomes
(e.g., Fitbit daily wear=93%, retention=93%, M=2 partner contacts per week over 4 weeks)
and identified key opportunities for improvement, which informed Study 2. Participants
(n=20, MAge=50, MBMI=30.9 kg/m2) attended a single face-to-face skills session, and
communicated with their partner and with the group via Fitbit social network.
Feasibility/acceptability outcomes were maintained or improved in Study 2 (e.g., daily
wear=97%, retention=100%, M=2 partner exchanges per week over 6 weeks). Discussion of
these studies will focus on their use in the development of an NIH R21 proposal to pilot test
this PA program among women with elevated cardiovascular risk.

Symposium 43C

THE THRIVE STUDY: FEASIBILITY OF AN INTERVENTION TO PREVENT WEIGHT GAIN IN BREAST

CANCER SURVIVORS DURING CHEMOTHERAPY

Shannon L. Mihalko, Ph.D.

The American Cancer Society estimated that more than 230,000 new cases of invasive breast
cancer would be diagnosed among women in 2015. Weight gain, a common side effect
experienced by breast cancer survivors receiving adjuvant chemotherapy, has been shown to
negatively impact cancer outcomes, general health and morbidity, and positive well-being.
Although weight gain can be prevented by healthy lifestyle behaviors, the task of adopting
and maintaining physical activity and an energy-balanced diet is a daunting one for most
adults and especially for women undergoing an intensive cancer treatment protocol.
Consequently, the timing of successful weight management interventions is an ongoing
debate. The primary objective of the Taking Health Realization into Valued Eating and Exercise
(THRIVE) study was to demonstrate the feasibility of conducting an intervention to prevent
weight gain for breast cancer survivors during active chemotherapy. This presentation will
provide an overview of the theoretical framework and highlight findings from a preference
S1454 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

survey that informed the design and timing of the intervention. Feasibility was determined by
recruitment, accrual and adherence rates, perceived barriers, adverse events, and program
satisfaction. Reasons for ineligibility and the need to reevaluate inclusion/exclusion criteria in
this population will be discussed, as well as barriers to recruitment and retention. Thirty
women participated in THRIVE with an average age of 51 years. The adherence rate to both
testing and intervention sessions was greater than 90% with time constraint and distance the
most often cited barriers to participation. There were no adverse events and participants
reported an overall high program satisfaction (3.7 on 4.0 scale). This feasibility study will set
the stage for a multi-center clinical trial to test the efficacy of weight gain prevention during
chemotherapy that may lead to beneficial effects on recurrence of breast cancer, risk of
comorbid conditions, and both stress reduction and improved quality of life in cancer
survivors.

CORRESPONDING AUTHORS: Allison Yax, B.S., Wake Forest University; [email protected];

Sarah Oliver, M.S., Wake Forest School of Medicine; [email protected]; Gary Miller,
Ph.D., Wake Forest University; [email protected]; Susan Melin, M.D., Wake Forest School of
Medicine; [email protected]; Tiffany Avery, M.D., Wake Forest School of Medicine;
[email protected]; Judith Hopkins, M.D., Novant Health; [email protected];
Greg Russell, M.S., Wake Forest School of Medicine; [email protected]; Peter
Brubaker, Ph.D., Wake Forest University; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1455

Symposium 44 8:00 AM-9:15 AM

TARGETING HEALTH BEHAVIOR CHANGE AND CLINICAL OUTCOMES WITH MINDFULNESS-

BASED MOBILE INTERVENTIONS

Jessica Chandler, PhD1, Judson A. Brewer, MD/PhD2, Ashley E. Mason, PhD3

1
Medical University of South Carolina, Charleston, SC; 2UMass Medical School, Worcester,
MA; 3University of California San Francisco, San Francisco, CA

This symposium comprises presentations that detail, from start to finish, three mobile-
delivered interventions targeting health behavior change and improved clinical outcomes
using mindfulness-based intervention techniques. Each presentation reviews processes of trial
conduct, spanning from participant recruitment to results. By reviewing the entire process of
conducting such trials, presenters will maximize attendees learning about how to conduct
research testing the impacts of mobile-delivered interventions on health behavior change and
clinical outcomes. Each presenter will speak about interventions that delivered mindfulness
and meditation practices over smartphones with the goals of improving health behavior and
health conditions, including pre-hypertension, cigarette smoking, or obesity. The first
presentation will report on a smartphone-delivered breathing awareness meditation (BAM)
that focused on reducing blood pressure among adults with pre-hypertension. The second
presentation will report on a smartphone-delivered mindfulness program focused on
engaging individuals who smoke cigarettes to complete ecological momentary assessments
(EMA) about episodes of craving and cigarette smoking while they work toward complete
smoking cessation. The third presentation will report on a smartphone-delivered mindful
eating intervention focused on reducing craving-related eating and weight loss among
overweight and obese individuals. All presentations will review lessons learned in conducting
trials that test smartphone-delivered interventions that employ mindfulness and meditation
practices in the service of improving health behavior and clinical outcomes. Presentations will
also report on the effects of these interventions on metrics of health and health behavior,
including change in blood pressure levels, cigarette craving and smoking, and craving-related
eating and weight loss.

Symposium 44A

SMARTPHONE DELIVERED BREATHING MEDITATION FOR BLOOD PRESSURE CONTROL AMONG

ADULTS WITH PREHYPERTENSION
S1456 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Jessica Chandler, PhD

Background: Pre-hypertension (pre-HTN) affects ~33% of adults and is a primary risk factor for
hypertension (HTN) and cardiovascular disease (CVD). Breathing awareness meditation (BAM)
can reduce stress levels and blood pressure (BP) but the optimal dosage level has not been
determined. Furthermore, dose adherence has often been based upon self-report. In this trial
we tested the acceptability and efficacy signals of Tension Tamer (TT), an iOS and Android
smartphone BAM app targeting reductions in BP. Methods: Tension Tamer (TT) uses
text/audio instructions with proprietary embedded, validated reflective photo
plethysmography software that captures real-time heart rate (HR) from a users fingertip
placed over video camera lens during meditation sessions. Users receive feedback graphs
after each session showing their HR changes. TT time stamps all HR data, which provides an
index of adherence to completing sessions. TT provides tailored reinforcement and
motivational text messages along with educational information about benefits of BAM.
Results: Participants were 62 adults (mean age: 35.7 years; 32 males; 33 White; 29 Black) with
pre-HTN on 3 consecutive sessions (mean SBP 128.8 mmHg; 95% recruitment rate). They were
randomly assigned to 5, 10 or 15 min sessions, twice daily for 6 months. Average adherence
declined from month 1 through months 3 and 6 across dosage groups (ps<.001); 5 min (mth 1:
82% to mth 6: 78%), 10 min (mth1: 80% to mth 6: 70%) and 15 min (mth 1: 79% to mth 6:
53%). We observed clinically meaningful and statistically significant reductions in SBP at
months 1, 3 and 6 for all 3 dosage groups (ps<.001): 5 min: -8.0, -6.4, -6.2; 10 min: -7.8, -7.6, -
9.2 and 15 min: -11.1, -9.4, -9.8 mmHg. Conclusion: The TT app is acceptable and useful in
objectively tracking adherence and shows promise in reducing SBP in people with pre-HTN.
We will present additional quantitative and qualitative analyses of post-trial informant
interviews, as well as TT refinements and dosage determination for our upcoming 12-month
efficacy RCT.

CORRESPONDING AUTHORS: John Sieverdes, PhD, MUSC; [email protected]; Mercedes

Cain, HS, College of Charleston; [email protected]; Spencer Wilder, HS, College of
Charleston; [email protected]; Jenna Gunsolley, BS, MUSC; [email protected]; Luke Sox,
BS, MUSC; [email protected]; Brenda Brunner-Jackson, MPH, MUSC; [email protected];
Patel Sachin, MS, MUSC; [email protected]; Frank Treiber, PhD, MUSC; [email protected]

Symposium 44B

MINDFULNESS TRAINING FOR SMOKING CESSATION DELIVERED VIA SMARTPHONE

Judson A. Brewer, MD/PhD

Background: Digital Health interventions for smoking cessation have become more widely
used, yet challenges exist in documenting their mechanism of action and efficacy. Further,
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1457

randomized controlled trials in mHealth face new opportunities and challenges: Recruitment
from a broader and real world population is more accessible, yet participant screening and
protocol adherence are more difficult to assess. Methods: We evaluated recruitment,
engagement and retention in a randomized trial of a 21-day smartphone app for mindfulness
training for smoking cessation: Craving to Quit. The app uses experience sampling to query
smokers behavior and experience in real time. Queries ask participants questions such as,
how much are you craving a cigarette right now? and participants use a sliding scale to
indicate craving level. This trial compared Craving to Quit to a 21-day control app that
provided goal setting, reminders, and ecological momentary assessment of cigarette craving
and smoking. Participants were randomized to experimental (n=247) or control (n=263).
Results: Enrollment: 5300 people accessed online screening, 5% were excluded for not having
a smartphone, and 509 were randomized to treatment. We randomized twelve participants
per week to treatment over 42 weeks. Demographics: Participants were mostly female (70%
female), and 50% of participants used iPhone, with the other 50% using Android. The mean
age was 42 years. On average, participants reported smoking 17 cigarettes/day and 6 prior
quit attempts. Of participants who started treatment (treatment starters), those randomized
to Craving to Quit tended to complete more treatment days (133 treatment starters) relative
to those assigned to the control app (186 treatment starters), p = .051. Among treatment
starters, completion rates were higher for the Craving to Quit group (47%) relative to the
control group (27%), X2=14.3, p<.001. High engagement in experience sampling was achieved
and did not differ between groups (p=.74). Treatment starters checked-in on average 55.3
48.5 times across the study, on more than 60% of treatment days (mean=148 days). The
average number of check-ins per day was 3.6 1.7 and did not differ across groups (p=.19). At
6-month follow-up, the experimental (78%) and control (74%) groups evidenced similar
retention (p=.34) and retention was greater for treatment starters (85% overall). Conclusion:
Craving To Quit effectively engages individuals who are attempting to quit smoking cigarettes
to check in about their cigarette cravings and smoking behavior. Craving to Quit is a promising
mobile intervention for individuals attempting to quit cigarette smoking.

CORRESPONDING AUTHORS: Kathleen Garrison, PhD, Yale University;

[email protected]; Prasanta Pal, PhD, UMass Medical School;
[email protected]; Stephanie O'Malley, PhD, Yale School of Medicine;
[email protected]

Symposium 44C

A MOBILE-DELIVERED MINDFUL EATING INTERVENTION TARGETING CRAVING-RELATED

EATING AND WEIGHT LOSS

Dr. Ashley E. Mason, PhD

S1458 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Background: Most weight loss interventions are handicapped by sampling biases due to
requiring in-person meetings. Attrition rates are often high due to unrealistic expectations
about abilities to adhere to diet and exercise prescriptions. In this study, we tested a 28-day
self-paced smartphone-delivered intervention that targets craving-related eating using
mindful eating practices. The intervention minimizes user burden by asking users to view daily
lessons that are approximately 5 to 10 minutes at any time of the users choosing, which
facilitates easy incorporation into ones lifestyle. Methods: In total, 104 adult overweight or
obese women (Mean age = 46.2 years, SD = 14.1 years; Mean BMI = 31.5, SD = 4.5) began the
intervention. We recorded anthropometric data pre-intervention and one month after the
participant completed day 28. We assessed participants food cravings before, during, and
after the intervention using ecological momentary assessment (EMA) delivered via text
message (SMS). Results: Recruitment & Enrollment: Of 1,651 participants who completed an
online screener, 170 completed the baseline SMS-based craving assessment, and 141
qualified by responding to at least 7 of 9 texts and endorsing at least 3 food craving
experiences. Ninety-eight (94.2%) participants indicated how they learned about the study. Of
these, 40.8% (n=40) learned about the study through Facebook, 25.5% (n=25) through campus
email announcements, 18.4% (n=18) through a letter in the mail, 14.3% (n=14) through other
means, and 1.0% (n=1) through Craigslist. Retention: Of the 104 participants who began the
intervention, 6 withdrew due serious life events (e.g., death in family, overseas relocation).
Seventy-six participants completed the 28-day intervention (73.1% completion rate), of which
72 attended an in-person follow-up visit one month after completing the intervention.
Intervention effects: In per protocol analysis (intervention completion within 3 months)
participants (N=61) lost 1.8 lbs (p = .019) and reported a statistically significant reduction in
craving-related eating from pre- to post-intervention (Mean diff = -0.19, b=-.46, p<.001). In
intention to treat analysis (intervention completion as of 7 months), participants (N=72) lost
1.0 lbs (p = .160) and reported a statistically significant reduction in craving-related eating
(Mean diff = -0.20, b=-.42, p<.001). Conclusion: Mobile-delivered interventions that target
mindful eating practices, rather than prescriptive diet and exercise recommendations, may
represent a low-burden, low-cost method to reduce craving-related eating among overweight
individuals.

CORRESPONDING AUTHORS: Nina Jhaveri, MBA, University of Southern California;

[email protected]; Michael Cohn, PhD, University of California San Francisco;
[email protected]; Judson Brewer, MD, PhD, University of Massachusetts Medical
School; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1459

Symposium 45 8:00 AM-9:15 AM

USING ECOLOGICAL MOMENTARY ASSESSMENT (EMA) TO ADVANCE HEALTH BEHAVIOR

THEORY

Jaclyn P. Maher, Ph.D.1, Genevieve F. Dunton, PhD, MPH1, Chih Hsiang Yang, MEd2, Meg
Bruening, PhD, MPH, RD3

1
University of Southern California, Los Angeles, CA; 2Penn State University, State College, PA;
3
Arizona State University, Phoenix, AZ

Health behavior theories serve as a guiding framework for understanding and intervening on
weight-related behaviors (e.g., physical activiy, sedentary behavior, dietary intake). However,
traditional health behavior theories such as the Theory of Planned Behavior or Social
Cognitive Theory face growing criticism. Traditional health behavior theories are limited in
that they often (1) consider behavior as static phenomena examined at a single time point, (2)
overlook contextual influences on behavior, and (3) do not incorporate concepts such as
stability or instability of behavior and relevant covariates. Ecological momentary assessment
(EMA) is a real-time data capture strategy that can help to address the limitations of health
behavior theory by intensively sampling behavior and other relevant factors as they occur in
the context of everyday life. Specifically, EMA can be used to determine (1) the extent to
which explanatory factors co-occur in time and space with health behaviors (i.e.
synchronicity), (2) the temporal sequence of antecedents to and consequences of health
behaviors (i.e., sequentiality), and (3) patterns of fluctuation and change in explanatory
factors and behavior (i.e., instability). This symposium will address how EMA is stimulating
advances in health behavior theory with specific emphasis placed on the role of EMA to
address sequentiality. The first presentation will explore time-varying predictors of short-term
physical activity intention formation and short-term physical activity intention-behavior
coupling using data collected through EMA and ambulatory monitoring. The second
presentation will explore the role of mothers momentary stress on mothers subsequent
parenting practices regarding their childs physical activity, sedentary behavior, and eating.
The third presentation will explore associations between momentary positive and negative
affect and physical activity and sedentary behavior in college students. Overall, these
presentations will reveal both the benefits and challenges of using EMA methods to advance
our understanding of the processes underlying health behavior change. The discussant will
address both the specific contributions of these presentations and the broader role of using
EMA in the advancement and refinement of health behavior theory.

Symposium 45A
S1460 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

MOMENTARY ASSESSMENT OF PHYSICAL ACTIVITY INTENTION-BEHAVIOR COUPLING IN

ADULTS

Jaclyn P. Maher, Ph.D.

Background: Research attempting to elucidate physical activity (PA) intention-behavior

relations has focused on differences in long-term behavior forecasting between people.
However, regular PA requires a repeated performance on a daily or within-daily basis.
Intensive longitudinal data (ILD) collected through methods such as Ecological Momentary
Assessment (EMA) and mobile sensors can examine within-person differences in PA intentions
over shortened timescales to understand how they might predict variations in one's
immediate behavior. This study used ILD collected through EMA to (a) describe the extent to
which short-term intention-behavior coupling occurs and (b) explore time-varying predictors
of intention formation and short-term intention-behavior coupling. Methods: Adults (n =116)
participated in three separate four-day waves of EMA. Each day, participants received eight
EMA questionnaires assessing short-term PA intentions (i.e., over next few hours) and
current affect and context (i.e., right now). Participants wore accelerometers at the same
time as completing EMA to assess whether they engaged in 10 minutes of moderate-to-
vigorous physical activity (MVPA) in the 3-hour period after each EMA prompt. Results:
Participants reported having short-term intentions to engage in PA in 41% of EMA prompts.
However, participants engaged in 10 minutes of MVPA following only 16% of the prompts
that short-term PA intentions were reported. Participants were more likely to intend to be
active in the mornings (OR = 2.16, p < 0.01), but less likely to intend to be active in the
evenings (OR = 0.32, p < 0.01) compared to other times during the day. Participants were also
more likely to intend to be active on weekend days compared to weekdays (OR = 1.69, p <
0.01). Odds of intentions followed by PA were greater on occasions when individuals reported
higher levels of positive affect than was typical for them (OR = 2.20, p < 0.01). Conclusion:
Results from this study are the first of its kind to indicate the possibility of time-varying factors
predicting the formation and enactment of short-term PA intentions across the day.
Understanding the role that these processes play in intention formation as well as intention-
behavior coupling is necessary to determine the processes that regulate PA in the context of
everyday life and ultimately develop more effective, tailored interventions for individual
behavior change.

CORRESPONDING AUTHORS: Ryan Rhodes, PhD, University of Victoria; [email protected]; Eldin

Dzubur, MS, University of Southern California; [email protected]; Jimi Huh, PhD, University of
Southern California; [email protected]; Stephen Intille, PhD, Northeastern University;
[email protected]; Genevieve Dunton, PhD, MPH, University of Southern California;
[email protected]

Symposium 45B
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1461

MATERNAL STRESS AND WEIGHT-RELATED PARENTING PRACTICES: AN ECOLOGICAL

MOMENTARY ASSESSMENT STUDY

Dr. Genevieve F. Dunton, PhD, MPH

Background: According to theories of ego depletion and self-regulation, psychological stress

may compromise parenting practices related to childrens dietary intake and physical activity.
However, past research in this area used retrospective measures of stress and overlooked
within-person processes. This presentation will demonstrate how Ecological Momentary
Assessment (EMA) can be used to examine micro-temporal sequences underlying maternal
stress and subsequent weight-related parenting practices. Methods: Mothers (N = 194) (mean
age = 41.0,48.5% Hispanic) of children 8-12 years old participated in 7 days of EMA with up to
7 randomly-prompted EMA surveys per day (every 2 hours) during childrens non-school
time. EMA items assessed perceived stress (right now), exposure to stressors (i.e., work at
home, family demands, spousal tension) (over the past 2 hours), and weight-related
parenting practices (i.e., limiting of high fat/high sugar foods) (over the past 2 hours).
Multilevel models examined prospective associations between maternal stress variables
reported at any given prompt T and subsequent weight-related parenting reported at the next
prompt T+1 (occurring 2 hours later), disaggregating within-subject (WS) and between-subject
(BS) effects of stress and adjusting for relevant covariates (e.g., time of day, child
age). Results: BS effects showed that mothers who experience greater stress from work at
home (b=-0.33), family demands (b=-0.30), spousal tension (b=-0.50), and general perceived
stress (b=-0.08), on average compared to other mothers, reported less limiting of high
fat/high sugar foods (ps <.05). However, WS effects showed that when mothers reported
experiencing greater stress from work at home (b=0.37), family demands (b=0.32), and
spousal tension (b=0.54) than their usual, they subsequently engaged in more limiting of high
fat/high sugar foods in the next 2 hours (ps<.05). Conclusion: Among mothers, overall levels
of stress may deplete emotional and cognitive resources needed to successfully perform
weight-related parenting. However, during times of family and home stress, mothers'
vigilance toward weight-related parenting may be temporarily enhanced. The subpopulation
of mothers who tend to experience the most stress may be in greatest need for interventions
targeting weight-related parenting, and the effectiveness of such interventions could be
maximized by enhancing parenting vigilance during situations with lower relative levels of
stress. This presentation will conclude by discussing the strengths and challenges of using
EMA to test and advance theories of ego depletion and self-regulation.

CORRESPONDING AUTHORS: Wangjing Ke, MS, University of Southern California;

[email protected]; Eldin Dzubur, MS, University of Southern California; [email protected];
Adam Leventhal, PhD, University of Southern California; [email protected]; Jimi Huh,
PhD, University of Southern California; [email protected]; Gayla Margolin, PhD, University of
Southern California; [email protected]; Stephen Intille, PhD, Northeastern University;
[email protected]
S1462 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 45C

AFFECTIVE ANTECEDENTS OF SITTING, STANDING, AND MOVING IN DAILY LIFE

Chih Hsiang Yang, MEd

Background: Cognitive processes feature prominently in theories of physical activity and

sedentary behavior. The role of affective processes, such as stress, anxiety, and pleasure, has
been less well developed. Affective processes can fluctuate rapidly throughout the day and
these fluctuations cannot be captured adequately cross-sectional, prospective, and panel
datasets. Affective experiences during physical activities predict adherence, presumably
because people become conditioned to associated activity with a reward or punishment
based on this affective experience. Affective experiences may also energize or inhibit proximal
activity patterns but this association has not been established. We used an experience
sampling research design to test hypotheses about the role of affective processes in
predicting the odds of future sedentary behavior, standing, and physical activity of different
intensities. Methods: Over 14 days, 158 university students responded to eight randomly-
timed prompts to complete momentary self-reports using the Personal Analytics Companion
(Paco) smartphone application. A total of 14,591 experience samples were collected
(M=92/participant). Results: A series of logistic multilevel models found that, after controlling
for gender, the time of the day, day of the week, and duration between momentary
responses, within-person deviations in prior stress levels significantly predicted increased
odds of reporting sitting (odds ratio= 1.07, p<.05), and decreased odds of reporting standing
(odds ratio= 0.86, p<.01) in the following responses. The odds of engaging in vigorous physical
activity were also negatively predicted by prior stress levels (odds ratio= 0.81, p<.05) and
positively predicted by prior levels of anxiety (odds ratio= 1.19, p<.05) and pleasure (odds
ratio= 1.23, p<.01). At the between-person level, average levels of stress, anxiety and pleasure
did not predict the odds of sitting or engaging in light/moderate/vigorous physical activity.
Interaction effects were further identified between the latency of responses and levels of
momentary pleasure in predicting subsequent light (p<.01) and moderate physical activities
(p<.01). Conclusion: These results supported the hypothesis that within-person affective
dynamics influence proximal activity patterns. Thus, affect regulation interventions may be
useful for reducing sedentary behavior and promoting physical activity. Further work with
both ecological momentary assessment designs and experimental trials of affective regulation
interventions is needed to elaborate the role of affective processes in regulating physical
activity and sedentary behavior.

CORRESPONDING AUTHOR: David Conroy, Ph.D., Penn State University; [email protected]

ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1463

Symposium 46 8:00 AM-9:15 AM

MECHANISMS OF TREATMENT CHANGE: RESULTS FROM ACCEPTANCE AND COMMITMENT

THERAPY FOR CHRONIC PAIN, WEIGHT LOSS, AND BINGE EATING

Kevin E. Vowles, PhD1, Evan M. Forman, PhD2, Stephanie Manasse, MS2

1
University of New Mexico, Albuquerque, NM; 2Drexel University, Philadelphia, PA

Psychosocial treatments for physical illness have amassed considerable evidence of

effectiveness, yet the mechanisms by which they achieve this success remains unclear.
Consequently, there are few data to guide treatment development and implementation
efforts focused on these active processes, or mechanisms, of change. Recent calls from
several research bodies, such as the NIH, have highlighted the need to address this issue. This
symposium will present the results of several studies which specifically evaluate treatment
mechanisms in one model of intervention, that of Acceptance and Commitment Therapy
(ACT). Data from three representative health conditions, chronic pain, obesity, and binge
eating, will be presented. The presentations will share a focus on treatment processes
specifically targeted by ACT, that of increases in acceptance and awareness (e.g., of pain,
current weight, and urges to binge) and engagement in valued activities and will evaluate how
changes in these processes relate to treatment outcomes. Dr Vowles will first discuss the
results of several Latent Growth Modeling analyses using data collected as part of an
interdisciplinary program of ACT for chronic pain. Results indicated two distinct patterns of
change in pain acceptance and engagement in valued activities, one of substantial
improvement and one of no change. Distal outcome analyses, investigating the relation of
trajectory of change in treatment mechanisms (i.e., improved/unchanged) with disability and
pain-related functioning at three-month follow-up indicated that membership in the
improved acceptance and values trajectory was associated with significantly greater
improvements in disability and functioning. Dr. Forman will next describe mediational
analyses conducted on a randomized clinical trial of an acceptance-based behavioral
treatment versus gold standard behavioral treatment for obesity. Bootstrapping analyses
indicated that the advantage of the acceptance-based treatment in the maintenance of
weight loss through 3-year follow-up was mediated by earlier changes in eating-related
psychological acceptance and autonomous motivation. Finally, Ms. Manasse will present the
results from a preliminary investigation of cognitive and affective mechanisms of change in an
ACT for binge eating disorder. Results indicated that improvements in emotion regulation and
decreases in experiential avoidance were associated with improvements in eating pathology.
Exploratory analyses also indicated that improvements in working memory and inhibitory
control were associated with decreases in binge eating. These data provide guidance on active
treatment mechanisms within one distinctive model of treatment and the overlap across
S1464 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

conditions suggests these processes may be generalizable across healthcare conditions,

thereby providing a potential model for future work.

Symposium 46A

ACT FOR CHRONIC PAIN: WITHIN-TREATMENT TRAJECTORIES FOR ACCEPTANCE & VALUES
PREDICT IMPROVED FUNCTIONING AT FOLLOW-UP

Dr. Kevin E. Vowles, PhD

A key issue in chronic pain treatment concerns the changes necessary for improved
functioning. Traditionally, reducing pain has been viewed as a prerequisite for improvement -
alternatively, Acceptance and Commitment Therapy (ACT), a behavior change approach,
theorizes that pain reduction is not necessary for improvement, rather responses to pain must
change, such that behavior become less restricted, more flexible, and more persistent in the
pursuit of adequate activity.These improvements in functioning are hypothesized to result
from improvements in pain acceptance and engagement in valued activity. Support for these
two treatment mechanisms have been provided via correlational findings (e.g., pre-post
changes in acceptance are significantly correlated with magnitude of improvement in
functioning), but modern statistical methods allow for more nuanced and informative
analyses which examine how specific aspects of change relate to treatment outcome. For
example, Latent Growth Modeling (LGM) allows the examination of longitudinal data to
determine if multiple trajectories of change occur (e.g., a trajectory of improvement and one
of no change) and if aspects of these trajectories, including intercept and slope, are
statistically predictive of improvement in key treatment outcomes, such as distress and
disability. This presentation will detail a LGM analyses of trajectories of pain acceptance and
valued action, as well as the relation of these trajectories to outcomes at a three-month
follow-up, over the course of a 4 week program of interdisciplinary ACT for chronic pain. Pain
acceptance and values were assessed weekly during treatment; distress and disability were
assessed at treatment onset and follow-up. Results indicated two latent trajectories for both
acceptance and values, one with no change and one with an increasing quadratic slope
indicating significant improvement during treatment. Further, change trajectory during
treatment was significantly related with improvement at follow-up, such that greater
increases (i.e., positive slope) in acceptance and values during the four weeks of treatment
were statistically predictive of larger improvement at follow-up. This pattern of findings
provides support for the ACT model in chronic pain and suggests that the mechanisms
identified in the theoretical model have a significant statistical relation with degree of
improvement following treatment.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1465

CORRESPONDING AUTHORS: Julie Ashworth, FRCP, IMPACT Pain Service;

[email protected]; Gail Sowden, PT, IMPACT Pain Service; [email protected];
Jayne Levell, DClinPsy, IMPACT Pain Service; [email protected]

Symposium 46B

MECHANISMS OF ACTION EXPLAINING THE SUPERIORITY OF ACCEPTANCE-BASED

BEHAVIORAL OVER STANDARD BEHAVIORAL WEIGHT LOSS TREATMENT

Evan M. Forman, PhD

Acceptance-based behavioral treatment (ABT) has demonstrated superior weight loss to gold
standard behavior treatment (SBT). However, the mechanisms of action that carry the effects
of ABT are not well understood due to the fact that the treatment contains a suite of
strategies (e.g., behavioral, tolerance of difficult internal experiences and loss of pleasure,
mindful decision making and commitment to valued behavior), a paucity of data and
measurement limitations. In a recent randomized controlled trial, 190 overweight or obese
participants were assigned to 25 sessions of ABT or SBT over a 1-year period. At baseline,
mid-treatment (6 months) and post-treatment (1 year), postulated mechanisms of action
(autonomous motivation and eating-related psychological acceptance) were measured, and
weights were taken at these points as well as at 1-year follow-up (2 years) Bootstrapping
analyses indicated that the advantage of ABT at post-treatment and follow-up assessments
was mediated by 6-month change in autonomous motivation and eating-related psychological
acceptance (CIs>95%). In addition, change in eating-related psychological acceptance (but not
autonomous motivation) from baseline to 6 months mediated the advantage of ABT during
the maintenance period between years 1 and 2 (post-treatment to 1-year follow-up). Thus,
some psychological mechanisms have effects long after they have been activated. Future
investigations should work to develop and test methods of enhancing behavioral weight loss
treatment on the strength of accumulating knowledge about the most (and least) potent
mechanisms of action.

CORRESPONDING AUTHORS: Meghan Butryn, PhD, Drexel University; [email protected];

Stephanie Manasse, MS, Drexel University; [email protected]

Symposium 46C

AN EXAMINATION OF POTENTIAL COGNITIVE AND AFFECTIVE MECHANISMS OF CHANGE IN

ACCEPTANCE-BASED TREATMENT FOR BINGE EATING

Stephanie Manasse, MS
S1466 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Standard treatments for binge eating disorder (BED) produce significant improvements in
eating pathology, but room for improvement remains. Strategies from acceptance-based
treatments (ABTs; e.g., Acceptance and Commitment Therapy and Dialectical Behavioral
Therapy) show promise for enhancing the treatment of binge eating disorder (BED)
specifically by providing strategies for tolerating negative affect, and slowing down decision
making processes. However at this time, it is unknown whether ABT works via its
hypothesized mechanisms. Examination of mechanisms is critical for eventually developing
more efficacious and optimized interventions.

As such, we conducted a preliminary pilot open trial (n=19) of ABT for BED and assessed
whether pre-post changes in potential process measures were associated pre-post
improvements in eating pathology. Specifically, we examined whether changes in emotion
regulation and experiential avoidance were associated with changes in binge eating and global
eating pathology after 10 weeks of treatment. Given that the focus of several ABT strategies is
to slow down in-the-moment decision-making, as an exploratory analyses, we examined
several executive functioning variables (i.e., cognitive control processes that govern goal-
directed behavior) as potential treatment processes. Results indicated that changes in
improvements in eating pathology were associated with increases in emotion regulation
(r=.54) and decreases in experiential avoidance (r =-.41). In addition, improvements in working
memory (r=.39) and inhibitory control (r=.54) were associated with decreases in eating
pathology.

These preliminary results from our small pilot trial show promising support for increases
emotion regulation and decreases in experiential avoidance as potential mechanisms of ABT.
In addition, improvements in working memory were associated with treatment gains, perhaps
because slowing down in-the-moment decision making impacts the ability to keep goal-
oriented information in mind. ABT may also work to decrease impulsive tendencies via
improving inhibitory control. Given the small sample size and correlational nature of analyses,
replication with larger samples is necessary.

CORRESPONDING AUTHORS: Evan Forman, PhD, Drexel University; [email protected];

Adrienne Juarascio, Ph.D., Drexel University; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1467

Symposium 47 8:00 AM-9:15 AM

NEXT-GEN MHEALTH: INTEGRATING PHYSIOLOGICAL MEASURES WITH SMART TECHNOLOGY

TO INFLUENCE HEALTH BEHAVIOR CHANGE

Yue Liao, MPH, PHD1, Susan M. Schembre, MS PhD RD2, David W. Wetter, PhD3, Bonnie
Spring, PhD4, Tiffany Cvrkel, PhD5

1
University of Texas MD Anderson Cancer Center, Houston, TX; 2The UT MD Anderson Cancer
Center, Houston, TX; 3University of Utah, Salt Lake City, UT; 4Northwestern University,
Chicago, IL; 5University of California, Los Angeles, Los Angeles, CA

Advancements in wearable sensor technology now make it possible to assess, in real-time,

physiological measures that influence or are influenced by health behaviors; representing an
exciting and potentially transformative development in health behavior research. The utility of
integrating objective measures reflecting a persons physiological state with traditionally used
self-report measures as well as ubiquitous mobile and smartphone technologies has inciting
implications on both the assessment of and intervention on health behaviors. In particular,
this integrative approach enables the study of the acute effects that physiological states, such
as stress, have on health-related behaviors, and reciprocally, how a persons behavior could
affect physiological states that have important health consequences. Importantly, this
approach has the greatest potential to proactively intervene at critical moments in peoples
natural environments in an effort to support health-related decisions. Areas of research
enhanced by the integration of body sensor technology and health behavior change that will
be discussed during this symposium include dietary intake and smoking cessation and relapse
prevention. In the first talk we will discuss the feasibility and value of using continuous
glucose monitors to objectively assess dietary intake and to provide proactive dietary decision
support for weight control and healthy eating. Our second speaker will discuss how body
sensors can be used to facilitate the real-time assessment of stress and smoking behavior by
tracking physiologic and behavioral data. Our final speaker will discuss the translation of these
objective assessment techniques to the development of a real-time adaptive stress reduction
intervention that is triggered by sensed physiological stress with the aim of preventing relapse
to smoking. During the session the speakers will discuss the challenges of this research related
to big data analytics and intervention timing. The session will highlight the need for behavioral
theories that integrate physiological measures with behavior change constructs and clarify the
relationship between physiological state and perceived state. In addition, the discussant -
a bioethicist specializes in emerging biomedical technologies will lead the discussion around
the topic of the ethical challenges around mHealth, eHealth, and Big Data.
S1468 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 47A

USING MOBILE GLUCOSE MONITORING AS BIOLOGICAL FEEDBACK ON DAILY EATING

BEHAVIORS

Dr. Yue Liao, MPH, PHD

Background: A number of theoretical frameworks emphasize the importance of using

feedback as a strategy for health behavior change (e.g., social-cognitive theory, theory of
reasoned action). Real-time physiological feedback, objectively assessed by body sensors, has
the potential to influence behavior change by informing health-related decisions at critical
moments. This study aims to examine the feasibility of collecting continuous glucose
monitoring (CGM) data from free-living individuals as real-time physiological feedback about
short-term energy status (pre-meal glucose levels) and metabolic responses to dietary intake
(post-meal glucose dynamics) to influence diet-related decisions that promote dietary
behavior change for weight control. Methods: Ten healthy participants (M age=28.9, 70%
female, 30% Hispanic, 60% overweight) have completed a 7-day monitoring period. Blood
glucose (BG) levels were recorded every 5 minutes using the Dexcom G4 PLATINUM CGM
system. Participants also completed diet records using MyFitnessPal and indicated perceived
pre-meal hunger levels in the CGM receiver using a 10-point scale. CGM data from monitoring
days 3 to day 7 were analyzed. In addition, diet data will be analyzed and presented for
glycemic quality using Nutrition Data Systems for Research. Results: On average, participants
reported 33.3 (SD=12.8) eating events across the 5 analytical days. Of these eating events,
61.7% (SD=.20) were initiated when a participants pre-meal BG level was greater than 85
mg/dl (a standard fasting BG level) and 48.8% (SD=.19) were initiated when a participants BG
level was greater than their own fasting BG level. On average, participants had 2.9 (SD=3.0)
hyperglycemic events (post-meal BG rise above 140 mg/dl). The average mean amplitude of
glycemic excursions (MAGE), which indicates the BG variability from peaks to nadirs, for all
eating events was 35.7 mg/dl (SD=10.3). Preliminary analyses showed that overweight
participants reported more frequent eating above 85 mg/dl than normal weight participants
(t=-2.22, p=.058). Discussion: Serial CGM data provides valuable physiological feedback
reflecting eating behaviors and dietary glycemic quality in free-living settings. Findings provide
early evidence that pre-meal BG levels could guide decisions about when to eat. Future
studies testing the use of CGM data as physiological feedback in real-time interventions that
promote dietary behavior change and weight control are warranted.

CORRESPONDING AUTHORS: Liang Li, PhD, UT MD Anderson Cancer Center;

[email protected]; Minxing Chen, UT MD Anderson Cancer Center;
[email protected]; Naifa Busaidy, UT MD Anderson Cancer Center;
[email protected]; Karen Basen-Engquist, UT MD Anderson Cancer Center;
[email protected]; Jimi Huh, University of Southern California; [email protected];
Susan Schembre, UT MD Anderson Cancer Center; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1469

Symposium 47B

ACUTE STRESSORS, DISCRIMINATION, AND SMOKING LAPSE LIKELIHOOD AMONG

UNDERSERVED GROUPS

Dr. David W. Wetter, PhD

Several conceptual models posit that greater exposure to stressors is a key mechanism linking
underserved status to lower rates of smoking cessation. Surprisingly, there are only a handful
of studies examining the impact of acute stress on real time, real world cessation outcomes,
and to the best of our knowledge, none that examine the acute effects of minority specific
stressors such as discrimination. Two longitudinal cohort studies examined the influence of
acute stressors on the likelihood of smoking lapse. In both studies, smokers received smoking
cessation treatment (behavioral, patch) and were tracked both pre- and postcessation using
Ecological Momentary Assessments (EMA). The first study included 364 predominantly low
socioeconomic status smokers evenly split among African Americans, Latinos, and Whites.
Participants reported the presence of a new stressor in 3%, an ongoing stressor in 10%, and
both new and ongoing stressors in 31% of EMAs. The presence of any stressor (new OR=2.1,
p<.001; ongoing OR=1.4, p=.003; both OR=1.8, p<.001) increased the likelihood of a lapse in
the next 4 hours, even after controlling for smoking in the previous 4 hours. There was no
synergistic effect of experiencing both new and ongoing stressors. The second study examined
the impact of acute discrimination on lapse likelihood among 159 Spanish speaking Latino
smokers. Participants reported that they were certain they had experienced discrimination in
<1% of EMAs. The presence of discrimination increased the likelihood of a lapse in the next 4
hours (OR=6.1, p<.03), even after controlling for smoking in the previous 4 hours. The results
from these studies suggest that acute stressors place underserved populations at greater risk
for poor cessation outcomes. Ongoing research is building on these data using state of the
science methodologies such as AutoSense and geographic positioning system (GPS).
AutoSense tracks behavioral and physiologic data in real-time and can objectively detect when
an individual smokes or encounters a stressor. GPS permits real-time spatial mapping, which
can be paired with EMA and Autosense, and with relevant environmental exposures (e.g.,
tobacco outlets; area-level poverty).

CORRESPONDING AUTHORS: Aaron Haslam, B.A., Texas Tech University;

[email protected]; Christine Vinci, Ph.D., Moffitt Cancer Center;
[email protected]; Cho Lam, PhD, University of Utah; [email protected]; Liang Li, PhD,
MD Anderson Cancer Center; [email protected]

Symposium 47C
S1470 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

OPTIMIZING SENSE2STOP: A JUST-IN-TIME ADAPTIVE MHEALTH STRESS INTERVENTION TO

PREVENT SMOKING RELAPSE

Prof. Bonnie Spring, PhD

Technical advances in the design of wearable sensors make it feasible to passively monitor
changes in physiological state, creating the potential for real time adaptive intervention. The
presenter will describe the development and implementation of a stress reduction
intervention triggered by sensed physiological stress and with the aim of preventing relapse to
smoking among recently quit smokers. She will describe how the interdisciplinary
intervention team worked through intervention development challenges that included: 1.
construct validation of the physiological stress indices, 2. translation between computer
science and behavioral science metrics in operationalizing the physiological trigger for
intervention, 3. designing stress management interventions that can be triggered based on
momentary physiological stress indices, 4. providing processing capacity to perform big data
analytics needed to infer stress accurately and make intervention decisions in real time, 5.
micro-randomization approach to developing evidence based decision rules about
intervention timing, and 6. managing data quality and participant burden. The Sense2Stop
development process highlights the need for a more dynamic theory of stress, one that
clarifies the relationship between momentary physiological and subjective stress, pinpoints
substrates for stress management intervention, and posits behavior change mechanisms
underlying the optimal timing of intervention triggers.

CORRESPONDING AUTHORS: Gwen Ledford, M.A. , Northwestern University;

[email protected]; Susan Murphy, PhD, University of Michigan;
[email protected]; Inbal Nahum-Shani, PhD, University of Michigan; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1471

Symposium 48 8:00 AM-9:15 AM

MEASURING AND TAILORING WEIGHT MANAGEMENT TREATMENT FOR THE UNIQUE NEEDS
OF VETERANS

Robin Masheb, Ph.D.1, Lindsey Dorflinger, PhD2, Phoutdavone Phimphasone, MA3, Kathryn
Godfrey, MS4, Niloofar Afari, PhD5

1
Yale School of Medicine/VA CT Healthcare System, North Haven, CT; 2Walter Reed National
Military Medical Center, Bethesda, MD; 3University of North Carolina, Charlotte, New Haven,
CT; 4SDSU/UCSD Joint Doctoral Program in Clinical Psychology, La Jolla, CA; 5University of CA,
San Diego & VA San Diego Healthcare System, La Jolla, CA

Studies show that Veterans have been disproportionately affected by overweight/obesity over
the last decade, but few studies have examined eating behaviors and associated factors in this
population. Given the high prevalence of binge eating among overweight/obese Veterans as
well as the growing number of women Veterans, there is a pressing need to enhance
assessment and better understand the unique treatment needs among Veterans seeking
services related to weight and eating behaviors. The objective of this symposium is to
enhance our understanding of eating behaviors, and correlates of eating disorders and
obesity, among this vulnerable patient group. Data was collected from Veterans attending
orientation sessions for the VA national weight management program utilizing measures of
eating, weight, and psychiatric functioning, and the electronic health record. Each presenter
will report on a timely and relevant aspect of eating and weight. The first presenter will
discuss findings of the relatively new and controversial construct of food addiction. Pertinent
to the inclusion of Binge Eating Disorder in DSM-5, the second presenter will introduce a new
single-item screener for binge eating that is feasible for use in medical settings. Given the high
rates of comorbid pain and overweight/obesity in the Veteran population, and the
implications for the treatment of each condition, the third presenter will discuss the unique
characteristics and needs of Veterans with overweight/obesity who meet criteria for clinical
pain. Finally, the last presenter will examine weight reporting biases in this sample. The
implications of these findings for assessment and treatment of weight management and
associated eating behaviors in the Veteran population will be discussed.

Symposium 48A

FOOD ADDICTION AMONG VETERANS WITH OVERWEIGHT/OBESITY

S1472 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Dr. Robin Masheb, Ph.D.

Objective: Obesity prevalence is higher among veterans compared to the general population.
and thus places veterans at greater risk for morbidity, mortality, and compromised quality of
life. The relationship between maladaptive eating and obesity within an addiction framework
is an emerging field of study. The purpose of the present study was to investigate food
addiction in a sample of veterans with overweight/obesity.

Methods: Participants were veterans with overweight/obesity who attended consecutive

orientation session at VA Connecticut Healthcare System for the MOVE! program, VAs weight
management program which has been implemented nationally. Veterans (n = 119, mean age
61.7 years, mean BMI 37.9, 90% male, 75% Caucasian) completed the Yale Food Addiction
Survey; measures of eating behavior, including the Eating Disorder Examination-
Questionnaire, Yale Emotional Overeating Questionnaire, Night Eating Questionnaire, and
Questionnaire on Eating and Weight Patterns Revised; and psychiatric screening
questionnaires for depression (PHQ-2), posttraumatic stress disorder (PC-PTSD) and alcohol
abuse (AUDIT-C).

Results: Twelve participants (11%) screened positive for Food Addiction (FA). Analyses of
covariance (ANCOVAs) controlling for BMI were conducted to compare those who screened
positive versus negative on FA. Veterans with FA had significantly higher scores on screeners
for depression (F=8.18, p

CORRESPONDING AUTHORS: Kathryn Min, VA Connecticut Healthcare System;

[email protected]; Su Cho, PsyD, VA Connecticut Healthcare System; [email protected];
Christopher Ruser, MD, VA Connecticut Healthcare System; [email protected];
Lindsey Dorflinger, PhD, Walter Reed National Military Medical Center;
[email protected]

Symposium 48B

A SINGLE-ITEM SCREENING MEASURE FOR BINGE EATING

Lindsey Dorflinger, PhD

Binge eating disorder (BED) is rarely screened for or detected as part of standard medical
care, largely because there are currently no assessment tools that are sufficiently brief and
specific enough to make routine screening feasible. However, given that BED is associated
with a number of medical and psychiatric issues, including cardiovascular disease, metabolic
syndrome, depression, and anxiety, and because effective and accessible treatments for BED
have been established, the identification of BED in medical settings could facilitate timely
referral to treatment, and potentially subsequent improvement in comorbidities. Therefore,
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1473

the present study sought to examine the validity of a single-item screening measure for binge
eating among veterans, who have shown to have high rates of binge eating as well as
associated medical and psychiatric comorbidities. Participants were 116 veterans referred to
the MOVE! program, a primary care-based weight management program implemented
throughout VA nationally. Participants had a mean age of 61.66 years (SD = 8.73) and mean
BMI of 37.90 (SD = 7.35). The item, the VA Binge Eating Screener (VA-BES), was derived from
the widely-used VA MOVE!23 and MOVE!11 surveys, which have been used as part of clinical
care within VA to assess factors that affect weight and weight-related behaviors and tailor
treatment to each veterans needs. All participants completed the Questionnaire of Eating and
Weight Patterns Revised (QEWP-R) to assess for BED, the VA-BES, and measures of
disordered eating and depressive symptoms. Patient responses to the QEWP-R showed a
prevalence of BED of 7.76%, and that frequency of binge eating ranged from zero to 21
episodes per week. The VA-BES was compared to the QEWP-R to determine the sensitivity,
specificity, positive predictive value, and negative predictive value for each cutpoint. One
cutpoint (2 binge eating episodes per week) maximized these values, demonstrated excellent
agreement with the QEWP-R (2 = 24.79, p < .001), and had significant associations with
measures of disordered eating, emotional overeating, and depressive symptoms. The VA-BES
appears to be a useful and valid measure of binge eating in veterans, and is brief and targeted
enough to be routinely used in primary care and other outpatient medical settings in which
the identification of binge eating and referral to treatment could ultimately have beneficial
effects on the medical issues at hand. The VA-BES demonstrated good concurrent validity,
sensitivity, and specificity in this sample of overweight and obese veterans; future studies
should also assess its utility in other samples.

CORRESPONDING AUTHORS: Christopher Ruser, MD, VA Connecticut Healthcare System;

[email protected]; Robin Masheb, PhD, Yale School of Medicine;
[email protected]

Symposium 48C

VALIDITY OF SELF-REPORTED WEIGHT: MEASURING WEIGHT REPORTING BIAS IN VETERANS

WITH OVERWEIGHT/OBESITY

Phoutdavone Phimphasone, M.A.

Understanding how individuals self-report height and weight is important given that
measurement of objective heights and weights is not feasible or practical for many large-scale
studies. Previous research has investigated reporting biases in US general populations and
clinical samples, with participants generally under-reporting weight. To date, only one study
investigated the accuracy of self-reported height and weight among Veterans.
S1474 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

This study investigated the accuracy of Veterans self-reported weights compared to

measured weights obtained from electronic medical record (EMR). Veterans (N=127) were
recruited from a weight management program at VA Connecticut Healthcare System and
completed a battery of self-report questionnaires at their orientation session. Veterans
reported their current weight, and their weight 3, 6 and 12 months ago. The orientation
session date was considered the index date. Measured weights were obtained from EMR from
time points of interest (orientation, and 3, 6, and 12 months prior to orientation) with
acceptable window of (+/- 30 days) from each time point. The Bland-Altman plot method was
used to quantify the agreement between the self-report and EMR weights by constructing
limits of agreement.

Analyses have been conducted on 88 Veterans for current and 3-month self-reported and
EMR weights. Participants (n= 80 male; n = 8 female) had mean age of 62.03 (SD=8.9), average
BMI of 38.03 (SD = 7.47), and 77% were Caucasian. For current weight, the average difference
was -2.8lbs (0.91kgs, SD = 15.33) with Veterans under-reporting weight than EMR weight;
limits of agreement -27.56 to 32.87. For 3-month weight recall, the average difference was -
0.73 lbs (0.33 kg, SD = 11.76); limits of agreement -22.70 to 23.76. Linear regression analyses
revealed that there were no significant proportional biases in weight recall; standard
deviations were large indicating significant individual variability in recall. Final analyses will
include the entire sample (N=127) and all four weight collection time periods (orientation, and
3-, 6,- and 12-months).

Preliminary analyses demonstrated that Veterans accurately recalled their current weight and
their weight three months compared to EMR; they were also better reporters than the
general population who significantly under-report their current weight by 6-13lbs (3-6 kg).
Understanding biases in reporting of weights strengthens our confidence in studies that
utilizes self-report methods for collecting weight and weight trajectory data.

CORRESPONDING AUTHORS: Dominica Hernandez, PhD, University of Connecticut;

[email protected]; Kathryn Min, VA Connecticut Healthcare System;
[email protected]; Lindsey Dorflinger, PhD, Walter Reed National Military Medical Center;
[email protected]; Christopher Ruser, MD, VA Connecticut Healthcare System;
[email protected]; Robin Masheb, PhD, Yale School of Medicine;
[email protected]

Symposium 48D

MODIFIABLE RISK FACTORS AMONG WEIGHT LOSS SEEKING VETERANS WITH PAIN

Kathryn Godfrey, MS
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1475

Rates of overweight/obesity approach 75% in some Veteran samples. Chronic pain is also
highly prevalent among Veterans with up to 70% of Afghanistan and Iraq Veterans endorsing
pain lasting 3 months or longer. Overweight/obesity and chronic pain frequently co-occur and
demonstrate a reinforcing, bidirectional relationship. Individuals with pain and obesity often
attribute weight gain to difficulty with exercise due to pain. However, less is known about
how pain relates to other potentially modifiable factors, such as eating behaviors and
psychiatric symptoms. This study examined the characteristics of Veterans with and without
pain that were initiating weight loss treatment (MOVE!) at VA Connecticut. Veterans enrolling
in MOVE! (N = 126) completed measures of psychiatric, eating, sleep, and pain symptoms. The
sample was mostly male (90%), White (75%), and non-Hispanic (93%) with average age of 62
years (SD = 8.6) and average body mass index (BMI) of 38 (SD = 7.4). Mean self-reported pain
rating was 4.11 (SD = 2.26, range = 0-10). Moderate to severe pain, defined as a current pain
rating of 4 and above, was endorsed by 57% of the sample. BMI was not significantly different
between Veterans with and without pain; however, Veterans reporting pain reported higher
levels of weight concern (p = 0.017), emotional overeating (p = 0.030), and night eating (p =
0.033). They also had significantly higher scores on measures of depression (p = 0.002), post-
traumatic stress disorder (p = 0.005), and insomnia (p < 0.001).For Veterans seeking
behavioral weight loss treatment, moderate to severe pain is associated with more severe
mental health problems, disordered eating symptoms, and sleep disturbance, all of which may
negatively impact weight loss outcomes. A better description of the clinical characteristics of
Veterans with pain who participate in MOVE! can highlight their unique needs and improve
treatments to address chronic pain in the context of weight loss treatment.

CORRESPONDING AUTHORS: Lindsey Dorflinger, PhD, Walter Reed National Military Medical
Center; [email protected]; Kathryn Min, VA Connecticut Healthcare System;
[email protected]; Christopher Ruser , VA Connecticut Healthcare System;
[email protected]; Robin Masheb, PhD, Yale School of Medicine;
[email protected]
S1476 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 49 8:00 AM-9:15 AM

PEER SUPPORT: GLOBAL DISSEMINATION AND LESSONS LEARNED

Edwin B. Fisher, Phd1, Henriette Curtz Jansen, Master of Sports Science2, Yuenxing Liu, PhD3,
Guadalupe X. Ayala, Ph.D. MPH4, John Elder, PhD, MPH4

1
Gillings School of Global Public Health, University of North Carolina-Chapel Hill, Chapel Hill,
NC; 2Municipality of Copenhagen, Farum, Hovedstaden, Denmark; 3Shanghai Jiaotong
University Affiliated Sixth People's Hospital, Shanghai, Shanghai, China (People's Republic);
4
San Diego State University, San Diego, CA

Abundant evidence supports the value of peer support interventions including involvement of
community health workers, Promotores de Salud, coaches and others. Despite their strengths,
implementation of such programs presents numerous challenges. Who will be the peer
supporters? How will they be trained? Will they work with groups, individually, by phone?
Who will supervise them? How will quality be assured? In what organization will they be
based? Will they be part of clinical teams or community based? If dealing with complex
diseases, what are the limits of their roles and the linkages with professionals? This
symposium presents the lessons learned from peer support programs in Copenhagen,
Shanghai, and along the California-Mexico border. In Copenhagens Health in Your Language,
peers carry out dialogue based meetings and varied activities to address risks for
noncommunicable diseases and HIV among varied immigrant groups. Men in Copenhagen
promotes strengthened social relations and mental health among unemployed, mostly
unmarried men with limited education and who are at risk for type 2 diabetes. The Shanghai
Integration Model (SIM) brings together specialty/hospital and primary care for diabetes.
Drawing on training that has led to over 35 programs in China and successful models for peer
support in Community Health Centers, programs in Shanghai have addressed how to place
peer supporters in Community Health Centers, link them with hospital-based clinical and
training resources, and deploy them to facilitate self management as well as planned care
within the SIM. In southern California, an academic-community partnership led to the
development and implementation of peer support programs to promote healthy eating,
diabetes control, and asthma control, and to prevent and control childhood obesity in the US
Hispanic/Latino community. Program structures have varied from stand-alone programs to
integration into clinical care. Across all of these programs, the continuum of peer support
from community to clinic, mutual help to highly trained peer coaches, individual to group,
general health to disease-specific objectives emerges as both a challenge of peer support
and the base of its vitality.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1477

Symposium 49A

PEER SUPPORT IN THE COPENHAGEN MUNICIPALITY HEALTH DEPARTMENT

Mrs. Henriette Curtz Jansen, Master of Sports Science

Copenhagen Municipality has initiated three innovative peer support approaches as part of a
strategy to curb inequality in health and reach populations at special risk and hardly reached
by traditional health services. The respective target groups are:

1. Ethnic minority groups at risk for noncommunicable diseases (especially immigrants from
Middle Eastern countries)

2. Ethnic minority groups experiencing high rates of HIV (especially from sub-Saharan Africa,
Eastern Europe, Thailand and the Philippines)

3. Men at risk of type 2 diabetes (especially unemployed single men with limited education,
over the age of 45).

For the two first target groups the program Health in Your Language emphasizes peer
education and role modeling. Trained peer health advisors, recruited from relevant ethnic
groups, facilitate dialogue-based meetings on health topics in everyday settings. Peer health
advisors also facilitate activities with professional health workers, bridge building citizens and
municipal staff. In 2015 more than 6000 people attended activities facilitated by peer health
advisors. Based on Health in Your Language, a special team of ethnic minority health advisors
from sub-Saharan Africa, Thailand and the Philippines offer tailored activities such as health
advice, health quizzes, checks and on-site HIV tests.

For men at risk for diabetes, a program starting in March 2016, Men in Copenhagen addresses
health in a broader perspective. Here, peers facilitate informal social activities (walks and
talks, social networking, games, social cooking etc.) to the target group and when needed
bridge build to health related services such as smoking cessation classes, etc. To facilitate
access, there are no requirements of registration to attend activities. In contrast to Health in
Your Language, the focus is not primarily on delivering health advice and knowledge, as it is
assumed to create distance to the target group. Rather social relations and mental health is
strengthened as a pathway to improved health. All initiatives seek to meet participants
where they are, thus working with religious and voluntary institutions, housing areas,
shelters etc. Peers are paid by the hour and continuously guided and trained.

Organizational strengths of working with peers in these programs are:

1) Continuous generation of knowledge about hardly reached groups through their
interactions with peers
S1478 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

2) Peer workers' ability to navigate easily in settings where health professionals, even those
focused on community outreach, cannot.

Organizational challenges include supporting and maintaining peer teams which can be time
consuming and challenging.

CORRESPONDING AUTHORS: Nanna Ahlmark, PhD, National Institute of Public Health,

University of Southern Denmark, Denmark; [email protected]; Camilla Dindler, PhD,
Department of Communication and Psychology, Aalborg University, Denmark;
[email protected]; Bodil Harboe, MA, Health Department, City of Copenhagen;
[email protected]

Symposium 49B

PEER SUPPORT IN THE SHANGHAI INTEGRATION MODEL

Yuenxing Liu, PhD

In China, recent government reforms to establish community health centers (CHCs) as the
principal providers of primary and routine health care have led to a need for innovative
approaches to care of chronic conditions like diabetes. New approaches aim to redefine the
model of chronic disease care in China by reversing the underutilization of CHC-based primary
care, linking CHCs with hospital-based specialty care, and strengthening the capacity of CHCs
through the training of care teams. Shanghai has been a national leader in the modern era of
healthcare reform. The Shanghai Integration Model (SIM) brings together primary and
specialty/hospital care for diabetes through shared treatment algorithms, patient education,
referral plans, and information systems as well as extensive training of CHC clinicians in
diabetes care. It encourages those with diabetes to receive routine care through CHCs near
where they live, relying on specialty/hospital care for complex problems and complex medical
treatment. Peer support can facilitate this transition, promoting excellent care as a
collaboration with CHC primary care to promote daily management and quality of life, backed
up by appropriate specialty/hospital care. Our previous collaborations have documented the
benefits of peer support through CHCs and trained over 500 professionals, resulting in over 35
ongoing peer support programs. Applying this experience to peer support in the SIM,
objectives include: 1) expand the capacity of CHCs to provide ongoing diabetes self-
management support, 2) strengthen linkages between primary and specialty care, 3) improve
the efficiency of healthcare utilization, and 4) improve patient outcomes and quality of life.
Peer support in the SIM encourages regular care through CHCs, promotes active management
of diabetes, extends the education provided to patients by professionals, works with CHC staff
to provide individualized care, and reinforces referrals to specialty/hospital care as well as the
return to CHCs for regular care. How to organize this is challenging. Hospital-based staff and
resources assist CHCs with program development and resources for training peers. CHCs
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1479

recruit and supervise peer supporters and provide their base of operations. Collaboration
among CHCs and hospital staff includes common peer support protocols, materials, and
monitoring procedures to provide standardization while also encouraging tailoring to
individuals and communities served. This approach harmonizes the strengths of both sectors,
CHCs and hospitals, in extending integration of care to integrated support for those with
chronic disease.

CORRESPONDING AUTHORS: Patrick Yao Tang, MPH, Peers for Progress, Gillings School of
Global Public Health, University of North Carolina-Chapel Hill; [email protected];
Maggy Muchieh Coufal, MPH, MA, Asian Center for Health Education; [email protected];
Cai Chun, MPH, Division of Disease Control & Prevention and Maternal & Child Healthcare,
Shanghai Municipal Health Bureau; [email protected].; Weiping Jia, PhD, MD, Shanghai
Sixth People's Hospital, Shanghai Diabetes Institute, and Jiao Tong University, Shanghai, China;
[email protected]

Symposium 49C

PROMOTING HEALTH THROUGH PEER SUPPORT IN THE US HISPANIC/LATINO COMMUNITY:

VARIATIONS ON A THEME

Dr. Guadalupe X. Ayala, Ph.D. MPH

Over the past 10 years, an academic-community partnership between the Institute for
Behavioral and Community Health and Clinicas del Salud del Pueblo, Inc. has developed and
evaluated a number of peer support programs to reach predominantly Mexican-origin
individuals and families to prevent and control a variety of chronic diseases. Among several
considerations to program development have included the extent to which the peer
supporters are integrated into clinical care and whether or not they are paid for their efforts.
The inception of this work began with a stand-alone model of peer support to promote
healthy eating through a family-based program. Using a video based on edutainment
principles, community health workers or promotoras from the clinic conducted home visits
with families to promote the adoption of healthy eating behaviors such as family meals. This
study resulted in improvements in mothers and childrens dietary intake, mothers parenting
behaviors, and in the healthfulness of the home environment. A recent innovation involved
the integration of community health workers into a clinic-based obesity care model to prevent
and control childhood obesity. In this study, pediatric patients and their families were referred
to a Family Wellness Program organized by the clinics Outreach Department to extend
healthcare into the community. Community health workers led the Family Wellness Program
which provided parents and children with the skills to modify their behaviors, relationships
and environment to promote healthy eating, physical activity, water consumption and quality
sleep. This study resulted in significant decreases in child BMI, as well as healthier eating
behaviors. In addition to testing approaches that vary on level of integration, we have also
S1480 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

considered the ongoing challenges of sustaining such models of care without adequate health
care funding. As such, the team has evaluated volunteer models of peer support, including
within the context of diabetes control. These efforts have resulted in improvements in
glucose control as measured by glycated hemoglobin and improvements in fitness as
measured by a step-test. However, sustainability remains a challenge despite the volunteer-
status of the peer supporters. Peer support models can vary on a number of dimensions,
including those discussed here. Researchers and practitioners need to consider the models
that are most suited for the specific problems, groups, and settings programs address.

CORRESPONDING AUTHORS: Leticia Ibarra, MPH, Clnicas de Salud del Pueblo, Inc.;
[email protected]; Luis Olmedo, Comit Cvico del Valle; [email protected]; Nadia
Campbell, MPH, Institute for Behavioral and Community Health; [email protected];
Esther Bejarano, Comit Cvico del Valle; [email protected]; Lucy Horton, MS, MPH,
Institute for Behavioral and Community Health; [email protected]; John P. Elder, PhD,
MPH, San Diego State University and the Institute for Behavioral and Community Health;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1481

Symposium 50 8:00 AM-9:15 AM

THE IMPORTANCE OF THE FAMILY CONTEXT IN PREVENTING AND TREATING CHRONIC HEALTH
CONDITIONS IN DIVERSE CHILDREN

Sarah Domoff, PhD1, Michelle Eakin, PH.D.2, Kristi E. Gamarel, PhD, EdM3, Daniel Mello, N/A4,
Elizabeth Towner, PhD5

1
Central Michigan University, Mt. Pleasant, MI; 2Johns Hopkins University, Baltimore, MD;
3
Brown University School of Public Health, Providence, RI; 4University of California, Merced,
Merced, CA; 5Wayne State University, Detroit, MI

Chronic health conditions, such as asthma, obesity, diabetes, and HIV, disproportionately
affect low-income and racially/ethnically diverse children and their families. Understanding
and intervening on the social and contextual determinants of health may facilitate efforts to
eliminate health disparities. Addressing the cross-cutting themes of the Society of Behavioral
Medicine 38th Annual Meeting (i.e., addressing health disparities and the social context and
health behaviors), this symposium will present four papers that evaluate the role of family
contextual factors in health conditions salient to racially/ethnically diverse children. The first
paper will examine how family routines associate with asthma management in preschool-age
children recruited from an inner city Head Start. The second paper will characterize how
mealtime TV viewing associates with family functioning and mother engagement with child
mealtime behaviors in low-income families with school-age children. The third paper tests
whether HIV stigma predicts mental health symptoms in parent-child dyads in South Africa.
Finally, in a sample of late adolescents from impoverished neighborhoods, the fourth paper
will evaluate the role of family relationship quality on diabetes management. Particular
strengths of this symposium include multidisciplinary presenters (e.g., pulmonary medicine,
child clinical psychology, public health, health psychology, and family medicine) and the
utilization of diverse methodologies (e.g., home observations, dyadic analyses, biological
assays). The discussant will facilitate conversation among presenters and attendees on how
clinicians can intervene on family and other social contextual factors to reduce health
disparities in ethnic/minority children.

Symposium 50A

TV USE DURING MEALTIME IN LOW-INCOME FAMILIES: ASSOCIATIONS WITH FAMILY

FUNCTIONING AND MEALTIME INTERACTIONS
S1482 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Dr. Sarah Domoff, PhD

Background: Mealtimes have been proposed as a family context that could be targeted to
prevent childhood obesity. However, TV viewing during mealtimes (more prevalent in low-
income households) may interfere with the psychosocial benefits of shared family meals, such
as parent-child engagement. The aim of this study was to examine how TV use during
mealtime associates with family-level functioning and maternal engagement among low-
income children and their families. These behaviors are hypothesized to be important
mechanisms through which family meals may be protective against child obesity risk.

Methods: This study is part of an ongoing longitudinal study of low-income mothers and their
children, originally recruited from Head Start preschools. Family mealtimes were video-
recorded by mothers when children were 4-8 years old (M age = 5.9 years, SD = 0.7; 48.2%
female). The majority of children (53.4%) were non-Hispanic White, with 16.7% non-Hispanic
Black, 17.1% Biracial, and 12.0% Hispanic. Of the 301 participants, 251 had videorecordings
(83.4%). The primary variables examined here were collected through observational coding of
family mealtime recordings. After achieving reliability ( > .70), trained research assistants
coded mealtimes for family-level functioning, mother-child engagement and affect, and
mealtime TV exposure (i.e., duration of audible TV during meal) and mealtime TV engagement
(i.e., amount of time child/mother directed eye gaze to TV). Linear regression analyses were
conducted to test whether greater mealtime TV exposure and engagement predicted poorer
family-level functioning and less maternal engagement or positive affect with child.

Results: After controlling for important confounders, analyses indicated that greater TV
exposure during mealtime was associated with poorer family functioning during mealtime (B =
-.13, p < .05) and less maternal positive affect (B = -.23, p < .05). Mothers of children who had
greater TV engagement during mealtime engaged less with their child overall (B = -.14, p <
.05), and engaged less with their childs eating during the meal (B = -.13, p < .05).

Conclusions: Utilizing observational methodology to characterize family meals, this study

elucidated potential mechanisms by which mealtime TV use may confer risk for child obesity
in low-income families.

CORRESPONDING AUTHORS: Julie Lumeng, MD, University of Michigan; [email protected];

Alison Miller, PhD, University of Michigan; [email protected]

Symposium 50B

MEDICATION ROUTINES ARE ASSOCIATED WITH HIGHER MEDICATION ADHERENCE AND

WORSE ASTHMA CONTROL AMONG URBAN PRESCHOOL CHILDREN

Michelle Eakin, PH.D.

ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1483

INTRODUCTION: Management of a child with asthma includes multiple behavioral and

environmental strategies. Families of school age children who integrate these strategies into
their routines report higher adherence but it is not known how these routines may impact
asthma management behaviors and overall asthma control in an urban preschool population.

METHODS: Children with asthma were recruited from Baltimore City Head Start to participate
in a randomized trial of asthma education. Families complete an in-home baseline assessment
that included the Asthma Routines Questionnaire (ARQ), Family Asthma Management System
Scale (FAMSS) and Test of Respiratory Asthma Control in Kids (TRACK).

RESULTS: 328 primary caregivers (94% African American, 63% reported earning less than
$20,000/year) of a child with asthma (61% male, 3.50.7 years) completed the assessment.
Medication routine burden was associated with higher asthma knowledge (r=0.16) symptom
awareness (r=0.17), balanced integration (r=0.18) and medication adherence (r=0.19) on the
FAMSS using Spearman correlations. However, there was a negative association between
medication routines and asthma control (r=-0.15).

CONCLUSIONS: Preliminary results indicate that better medication routines were associated
with higher self-reported adherence and asthma management skills but an inverse related to
asthma control. Families with children with uncontrolled asthma are more likely to have to
administer a daily medication and set up a medication routine, while families of children with
mild asthma only manage it sporadically. It is important for clinicians to integrate targeted
patient education about medication adherence, including the establishment of behavioral
routines to reduce known asthma medication adherence health disparities.

CORRESPONDING AUTHORS: Shang-En Chung, Johns Hopkins University; [email protected];

Cynthia Rand, PhD, Johns Hopkins University; [email protected]; Kristin Riekert, PhD, Johns
Hopkins University; [email protected]

Symposium 50C

THE DYADIC EFFECTS OF HIV STIGMA ON THE MENTAL HEALTH OF CHILDREN AND THEIR
PARENTS IN SOUTH AFRICA

Dr. Kristi E. Gamarel, PhD, EdM

Background: Evidence illustrates that HIV stigma both enacted stigma towards people
living with HIV and stigma by association towards their families is linked with adverse
individual mental health outcomes. However, questions remain unanswered with regards to
family dynamics of HIV stigma, specifically whether parental experience of HIV adversely
influence child mental health outcomes and vice versa. This study examines reciprocal impacts
of HIV stigma on parent and child mental health in South Africa, a country with a large
S1484 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

generalized epidemic.

Methods: Parent-child dyads (n=2477) in KwaZulu-Natal Province, South Africa participated in

a cross-sectional survey. Parents and children completed measures of enacted HIV stigma,
stigma by association, depressive symptoms, and anxious symptoms. Parents reported their
HIV status, child orphan status, age, gender, food insecurity, and education. Analyses using
structural equation modeling were used to examine the influence of HIV stigma on depressive
and anxious symptoms.

Results: Accounting for sociodemographic characteristics, both parents and childrens reports
of greater HIV stigma were independently associated with higher levels of anxious and
depressive symptoms (all p < 0.001). Parents reports of enacted HIV stigma were
independently associated with childrens greater depressive symptoms (p<0.01) and anxious
symptoms (p<0.01). Similarly, childrens reports of HIV stigma by association were
independently related to parents greater depressive and anxious symptoms (both p<0.01).

Conclusions: Findings provide support for a cross-over effect of HIV stigma on the mental
health of families affected by HIV, such that each family members experience of stigma
negatively impacts their own and their family members mental health. Future research and
family-based approaches to stigma-reduction interventions to alleviate mental health
problems will be discussed to guide programs to improve the mental health of parents and
children affected by HIV in South Africa.

CORRESPONDING AUTHORS: Caroline Kuo, DPhil., Brown University;

[email protected]; Mark Boyes, PhD, Curtin University; [email protected];
Lucie Cluver, PhD, Oxford University; [email protected]

Symposium 50D

FAMILY RELATIONS AND EXECUTIVE FUNCTION MEDIATE NEIGHBORHOOD DISORDER RISK IN

ADOLESCENTS WITH TYPE 1 DIABETES

Daniel Mello, N/A

Characteristics of impoverished neighborhoods (e.g., % unemployed or receiving public

benefits) are linked to poorer health (Chen & Paterson, 2006; Dulin-Keita, Casazza, Fernandez,
Goran, & Gower, 2012), but the mediating processes remain unclear. Such characteristics
contribute to family stress and parenting behaviors (Drew et al., 2011) that may undermine
the development of adolescents self-regulatory skills (Schroeder & Kelley, 2009), which
support type 1 diabetes management (Berg et al., 2014). The current study examined whether
the association of Neighborhood Disorder (ND; Dulin-Keita et al., 2012) to diabetes
management during late adolescence was mediated by lower parentchild relationship
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1485

quality and its links to poorer executive functions. Late adolescents (N = 247; M Age = 17.8, SD
= .40; 65% female; 75% non-Latino White) with type 1 diabetes completed measures of
parental acceptance, executive functions (BRIEF; Guy, Isquith, & Gioia, 2004), and adherence;
glycemic control was indexed by HbA1c assay kits. ND was captured by combining five U.S.
Census-derived indicators (e.g., % youth living in poverty, number of vacant units). Mediation
models revealed ND was linked to poorer HbA1c partially through lower parental acceptance
and, in turn, poorer executive function (bias-corrected 95% confidence interval range for
standardized indirect effects [ab = .006]: .001 .021). ND was linked to poorer adherence only
by full mediation through parental acceptance and executive function (bias-corrected 95%
confidence interval range for standardized indirect effects [ab range = -.025 to -.015]: -.059 to
-.003). Overall, family relationship quality may be a modifiable factor that underlies the risk of
ND on diabetes management.

CORRESPONDING AUTHORS: Deborah Wiebe, MPH, Ph.D., University of California, Merced;

[email protected]; Ashley Baker, B.A., University of California, Merced;
[email protected]; Cindy Berg, Ph.D., University of Utah;
[email protected]
S1486 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 51 8:00 AM-9:15 AM

EMERGING ISSUES IN GENOMICS IMPLEMENTATION AND TRANSLATION

Barbara B. Biesecker, PhD, MS1, Jennifer M. Taber, PhD2, Catharine Wang, PhD, MSc3, Colleen
M. McBride, PhD4

1
National Human Genome Research institute, National Institutes of Health, Bethesda, MD;
2
Kent State University, Kent, OH; 3Boston University School of Public Health, Boston, MA;
4
Emory University, Rollins School of Public Health, Atlanta, GA

The landscape of genomics continues to evolve rapidly and has outpaced our understanding
of how to best implement and translate genomic information to improve the health of
populations. Conventional models to convey genomic information have been challenged as
scientific discoveries occur and the demand for genetic services increases. Moreover, the
changing nature of genomic information and the online availability of personal genomics
testing underscore the need for evidence of clinical and personal utility. This symposium
highlights research examining the translation of genomics within this changing landscape.
Presentations on evidence from three endeavors will contribute to informing future
implementation: 1) models to return carrier results, 2) re-contacting individuals to convey
reinterpretation of variants and 3) interpretation of raw DNA. The first speaker will present
outcomes from a randomized controlled trial of 446 participants examining the effect of
receiving carrier results from a genetic counselor or through an online interface. This work
tests the relative value of two models for conveying carrier status from genome sequencing.
The second speaker will present qualitative and quantitative data from a study designed to
assess the outcomes of conveying reinterpretation of variants to 29 individuals, in comparison
to a matched control group (n=28). The third speaker will present results from two
complementary surveys conducted with (a) 321 consumers and (b) 83 genetic counselors, to
examine respective experiences with online raw DNA interpretation services, and describe
outcomes of genetic counseling sessions with consumer clients. Finally, the discussant will
consider the implications of these findings for genomics translation, highlight related
emerging issues in ongoing precision (behavioral) medicine efforts, and offer insights into
future genomics translation research priorities and how the field can build upon the findings
presented in this symposium.

Symposium 51A
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1487

RESULTS FROM A RANDOMIZED CONTROLLED TRIAL COMPARING TWO MODES OF DELIVERY

OF CARRIER RESULTS ASCERTAINED BY GENOME SEQUENCING

Dr. Barbara B. Biesecker, PhD, MS

A growing volume of genomic testing has fueled studies to identify alternative modes of
results delivery that may be as effective as genetic counseling but more efficient. We
conducted a trial of 461 adult participants enrolled in a genome sequencing study to assess
equivalence between return of carrier results from a genetic counselor and a tailored web-
based platform. All participants were randomized to delivery mode. Those in the genetic
counselor arm had significantly higher formal education, but there were no other significant
differences in socio-demographics. Follow up survey data were ascertained at one month and
six months after receiving results. Primary outcomes were knowledge, psychological
responses including uncertainty, and communication of results to children. Participants
receiving results from the genetic counselor scored higher in knowledge of heredity concepts
(p<0.05) at one month. There were no significant differences in knowledge domains between
delivery modes at six monthds. No difference was seen in perceptions of positive responses to
results or uncertainty at one month and at six months. Participants receiving results from the
website had significantly higher distress at one month (p<0.05) but not at six months. The
aggregate distress ratings at both times were <0.3 on the MICRA subscale (0-5). Responses to
open-ended questions about results demonstrated misunderstanding about the implications
of carrier results among 20% of participants who received results from the website,
suggesting that harms may have been incurred that did not occur in the genetic counselor
arm. Seventy-two percent and 78% of participants communicated the carrier results to their
adult children with no significant differences between arms. Additional analyses of
independent variables associated with outcomes, fidelity to the intervention, participant
preferences and feasibility are underway. Early results suggest little difference in primary
outcomes between the study arms. Yet there is qualitative evidence of misunderstanding that
may be sufficiently high to suggest that even highly educated participants may not understand
their carrier results as well after receiving them from a website and that equivalence was not
attained.

CORRESPONDING AUTHORS: Katie Lewis, ScM, NHGRI; [email protected]; Jennifer

Johnston, PhD, NHGRI/NIH; [email protected]; Kristen Fishler, BS, NHGRI/NIH;
[email protected]; Leslie Biesecker, MD, NHGRI/NIH; [email protected]

Symposium 51B
S1488 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

RESPONSES TO REINTERPRETATION OF A GENE VARIANT BY PARTICIPANTS IN A SEQUENCING

STUDY

Dr. Jennifer M. Taber, PhD

As discoveries in genome science advance, people who received personalized genetic risk or
carrier information may subsequently receive corrections. Little is known about the
consequences of these corrections. We conducted an exploratory study examining
participants immediate and short-term responses to such a correction for one gene variant.
Data were collected from participants (n=57, ages 50-70) in an NIH sequencing study
(ClinSeq). Some (n=29) originally learned they were carriers of Duarte variant galactosemia.
However, the variant was later reclassified as benign and participants sequences were
reassessed. Positive testers (n=19) had a newly identified causative variant and were told they
were carriers. Negative testers (n=10) were told they were not carriers. Data were also
collected from 28 controls who had learned that they were carriers for a disease of
comparable severity and had not received a correction. A survey was administered
immediately after participants received corrected results. Controls received the survey after
being reminded of their previous result. All participants completed a 3-month follow-up.
When positive and negative testers were asked to recall the correction information, only 1
participant in each group indicated a misunderstanding or did not recall the information (for a
total of 4 out 29 participants). All others demonstrated complete or partial recall accuracy.
When asked how the information made them think or feel, most positive (63.2% at 3 months)
and half of negative testers (50% at 3 months) indicated either positive (i.e., reassured) or
neutral reactions. Positive testers reported lower intentions to share their results than
negative testers (p=.016) and controls (p=.003; F(2)=5.46, p=.007). Positive (p=.026) and
negative testers (p=.018) had higher intentions than controls to learn results for preventable
disease in the future (F(2)=4.23, p=.020). Intentions to learn results for unpreventable disease
showed a similar pattern (F(2)=3.39, p=.041). There were no differences among controls and
positive and negative testers in perceived ambiguity, accuracy of the results, negative
emotions (feeling upset, anxious, regret), or uncertainty about results implications. Although
it is unknown to what extent these results will generalize to actionable diseases and other
samples as genome sequencing becomes more accessible, results suggest that informing
people of corrections to personally relevant risk information may not have adverse effects.

CORRESPONDING AUTHORS: Katie Lewis, National Human Genome Research Institute;

[email protected]; William Klein, National Cancer Institute; [email protected]; Leslie
Biesecker, National Human Genome Research Institute; [email protected]; Barbara Biesecker,
National Human Genome Research Institute; [email protected]

Symposium 51C
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1489

DO-IT-YOURSELF (DIY) GENOMICS: CONSUMER AND GENETIC COUNSELOR RESPONSES TO

ONLINE RAW DNA INTERPRETATION

Catharine Wang, PhD, MSc

Online services are currently available to assist consumers in interpreting their raw DNA, yet
little is understood about the utility of these services and the downstream implications, if any,
on genetic service providers. To gain insight on these issues, complementary surveys were
administered online to (1) 321 consumers of raw DNA interpretive services and (2) 83 genetic
counselors. Surveys were designed to examine respective experiences with raw DNA
interpretation services, and corresponding issues that may have arisen from counseling these
consumers. The majority of consumers use one of five companies to interpret their raw DNA;
73% used more than one. Company choice was driven by the type of results offered (51%),
price (45%), online reviews (31%) and recommendations by friends/family (15%). Only 5%
sought advice from a medical professional before using a service. Among the 30% who shared
results with a medical provider, 80% shared with their primary care doctor, 14% with a genetic
counselor, and 25% with other specialist. Overall, 81.5% of consumers reported being satisfied
or very satisfied with the information they received. Among the 18.5% who were less
satisfied, qualitative reasons included confusion with the information presented and feelings
that the information was incomplete. Among genetic counselors, 53% reported having ever
been contacted by a consumer following the use of a service to interpret raw DNA data.
Among counselors who had been contacted, 72% saw at least one or more cases. Patients
requested counseling for the following disease specialty areas: cancer (52%), neurology (21%),
metabolism (17%), pediatrics (10%), psychiatric (10%), hematology (7%), dementia (7%), other
(14%). Counselors quantitative ratings of patients emotional state pre- versus post-
counseling included worried (67% vs 12%, p<.001), confused (43% vs 27%, ns), upset (18% vs
31%, ns), ambivalent (11% vs 15%, ns), reassured (0% vs 27%, p=.006) and relieved (0% vs
31%, p=.003). Counselors overall impressions of the session were mostly negative (48%) or
mixed (20%) versus positive (32%). Challenges during counseling included patients unrealistic
expectations about what testing would tell them and overemphasis on the validity of results
reported by these services. Counselor ratings of the clarity and informativeness of the DNA
reports were low. In sum, although consumer respondents reported high satisfaction with
online raw DNA interpretative services, genetic counselors expressed many challenges with
counseling these types of patients. The downstream implications resulting from consumer
efforts to interpret their raw DNA will be discussed.

CORRESPONDING AUTHORS: Maureen Flynn, MS, CGC, MPH, MGH Institute of Health
Professions; [email protected]; Natalie Thomas, MPH, Stanford University;
[email protected]; Tricia Norkunas, MS, MPH, Boston University; [email protected]; Tiernan
Cahill, Boston University; [email protected]; Andrew Parlato, Boston University;
[email protected]; Blake Wertz, Boston University; [email protected]; Qiankun Zhong, Boston
S1490 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

University; [email protected]; James Cummings, PhD, College of Communication, Boston

University; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1491

Symposium 52 8:00 AM-9:15 AM

EXPLORING INTERSECTIONALITY AND WOMENS HEALTH

Ariella R. Tabaac, MS1, Veronica Njie-Carr, PhD2, Ivy Ho, PhD3, Jasmine Abrams, PhD, MS, BS4

1
Virginia Commonwealth University, Richmond, VA; 2University of Maryland School of
Nursing, Baltimore, MD; 3University of Massachusetts Lowell, Lowell, MA; 4University of
Maryland, Baltimore County, Baltimore, MD

Conceptualized by theorist Kimberl Crenshaw, Intersectionality is a framework that posits

that people can belong to multiple social categories (e.g., race, gender, sexual orientation)
that intersect and reflect differing forms of privilege or systemic discrimination (e.g., racism,
sexism, heterosexism). From a public health perspective, this entails examining how people
experience different forms of societal-level discrimination, barriers to care, or health risk
behaviors in relation to the intersection of personal identities. In addition, it challenges how
societal-level discrimination impacts health behaviors and patient outcomes from a
multidimensional approach. Given the multidimensional and multi-level approach of the
Intersectionality theory, it can be adapted to provide conceptual background for behavioral
health research through the use of identity-based determinants of health and the interaction
of health disparities and social injustice. The primary purpose of this symposium is to present
behavioral health research from an Intersectionality perspective to address the following
aims: 1.) To discuss how the Intersectionality can be applied to different populations and
health behaviors in womens health research, 2.) To showcase research projects that have
taken an Intersectional approach, and 3.) Discuss challenges and next steps in applying
Intersectionality to womens health research. Individual presentations will focus on unique
areas of identity and its intersection with gender and will include theoretical and empirical
presentations. Themes will include the intersection of gender with sexual orientation,
socioeconomic status, race and ethnicity, culture, and immigration status, and will incorporate
micro- and macro- levels of determinants of health.

Symposium 52A

USING INTERSECTIONALITY AS A FRAMEWORK TO EXPLORE CERVICAL CANCER SCREENING

DISPARITIES AMONG SEXUAL MINORITY WOMEN

Ariella R. Tabaac, MS
S1492 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Sexual minority women (SMW; i.e., lesbian, bisexual, and queer women) are at a heightened
risk for cervical cancer due to low rates of cervical cancer screening and a greater incidence of
smoking and other cancer risk behaviors. Research has found that primary reasons for
nonscreening include lack of provider, lack of provider referral, or misconceptions about
susceptibility. Provider heterosexism is one barrier to cervical cancer screening, and can
influence perceptions about susceptibility to human papillomavirus (HPV), the primary cause
of cervical cancer development, or lead to health care avoidance due to anticipated
discrimination. Research on screening has yet to account for how identification with multiple
minority groups may amplify cervical cancer risk, screening behavior, or health care access.
The goal of this presentation is to address current research on cervical cancer screenings in
SMW, and to discuss how an Intersectionaly framework can be applied to cervical cancer
research in this population. The intersection of sexual orientation, gender identity, race, age,
and disability categories will be discussed.

Symposium 52B

THE INTERSECTION OF PSYCHOSOCIAL DETERMINANTS OF HEALTH AMONG SOUTHEAST

ASIAN AMERICAN WOMEN

Dr. Ivy Ho, PhD

Southeast Asian American women, specifically those who are of Vietnamese, Cambodian, Lao
and Hmong descent, face significant health disparities. For example, compared to their non-
Hispanic white counterparts, Southeast Asian American women are more likely to die of
cervical cancer and less likely to participate in cervical cancer screening. Breast cancer is also
disproportionately common among Southeast Asian American women, and screening rates
are low compared to White Americans. Although these cultural groups are distinct from one
another, they share various historical and social similarities such as tumultuous history of
political unrest, war, and genocide. The environmental and structural challenges of living for
years in refugee campsunder harsh conditions exposed survivors to emotional and physical
trauma and diseases. Refugees en route to the United States underwent a harrowing voyage,
fraught with danger from pirates who robbed and murdered the refugees and raped the
women. Contrary to the model minority myth that posits Asians Americans are successful,
financially secure and well educated, Southeast Asian American families face multiple
disadvantages. They are of among the lowest household incomes and educational attainment.
Therefore, in this presentation, the intersectionality of historical, cultural, and social
(socioeconomic and polical) is reviewed focusing on how immigration history, social class,
gender and race intersect to influence health status among Southeast Asian American
women. The presentation will include qualitative data, gathered from research on the
Cambodian American community in Lowell, Massachusetts, to illustrate the psychosocial
factors discussed. These data were gathered from interviews with healthcare providers who
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1493

provided their perceptions of their Cambodian American female patients with regard to
participation in preventative care, adherence, and health behaviors.

Symposium 52C

USING INTERSECTIONALITY TO UNDERSTAND GENDER ROLES AND HEALTH IMPLICATIONS FOR

AFRICAN AMERICAN AND VIETNAMESE AMERICAN WOMEN

Jasmine Abrams, PhD, MS, BS

Discourse in Intersectionality provides dialogue relevant to racism and sexism to address and
overcome dominant cultural hegemony. Underlying Intersectionality is the recognition that
identity is experienced as a single, synthesized experience of many constructs. Numerous
studies have applied the perspective of Intersectionality in understanding the divergence in
gender roles found between African American women and White women. However, limited
research has offered insight into how gender roles may differ across ethnic minority women.
Indeed, research identifying similarities or disentangling differences in ethnic minority gender
role beliefs has been largely absent in the literature suggesting need for qualitative
exploration to increase our understanding of the phenomena. The purpose of this study was
to explore and compare gender role beliefs among African American and Vietnamese
American women. Thematic analyses were conducted with data gathered from eight focus
groups with 44 African American women (mean age = 44) and four focus groups 47
Vietnamese American women (mean age = 42). Women were diverse in generational,
religious, and educational backgrounds. Two similar primary themes emerged: (1) womens
roles as chief caretakers and (2) womens responsibilities to fulfill multiple roles. There were
also similar experiences of a need to convey strength and be self-sacrificial. Two distinct
differences that emerged from the focus groups were beliefs about interpersonal interactions
and perceptions of societal expectations. For example, women expressed contrasting views
about perceptions of how a woman should act among others when in public (e.g., mild
mannered versus openly strong and resilient). This study demonstrates that the
conceptualization of gender role beliefs, while at times similar, diverges among culturally
different groups. Health research and programming needs to account for these and other
culturally nuanced differences via cultural tailoring. Implications for research and health
related programming are discussed.

CORRESPONDING AUTHORS: Morgan Maxwell, PhD, Global Community Health Promotion

Network; [email protected]; Faye Belgrave, PhD, Virginia Commonwealth University;
[email protected]
S1494 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 53 8:00 AM-9:15 AM

THE ROLE OF RELIGION AND SPIRITUALITY IN BEHAVIORAL INTERVENTIONS

Karen Yeary, PhD1, Crystal L. Park, PhD2, Aasim Padela, MD MSc3, Ananya T. Banerjee, R.Kin
PhD4, Donna Jeffe, PhD5

1
University of Arkansas for Medical Sciences, Little Rock, AR; 2University of Connecticut, Storrs
Mansfield, CT; 3The University of Chicago, Evanston, IL; 4University of Toronto, Toronto, ON,
Canada; 5Washington University School of Medicine, Saint Louis, MO

Faith-based interventions have been successful in changing multiple health behaviors, such as
fruit and vegetable intake, cancer screening, smoking, and physical activity. However the
majority of these interventions have been conducted with Protestant or other Christian
religious affiliations, and studies engaging other faiths are necessary to more fully understand
faith-based interventions. Further, given the prominence of religiousness and spirituality in
the lives of many individuals, consideration of faith may also be relevant for non-faith-based
interventions.

This symposium will present the results of two qualitative studies and one quantitative study
to inform attenders of key factors to consider in designing health behavior interventions for
Muslim groups and a general patient population. Aasim Padela et al. will describe the key
facilitators at multiple levels of the socioecological model for successful participation in an
exercise intervention for South Asian Muslim women. Ananya Banerjee et al. will report the
important faith beliefs identified in their qualitative work with Muslims that were used to
design a faith-based intervention to increase mammography screening in Muslims. Finally,
Donna Jeffe et al. will present results showing of a randomized controlled trial designed to
increase adherence to recommended treatment in newly diagnosed African American breast
cancer patients, showing how religion/spirituality predicted dietary change over 2-year
follow-up.

At the conclusion of the individual presentations, Karen Yeary will discuss important faith
components to consider in developing interventions for diverse groups, future directions for
faith-based interventions, and consideration of religion and spirituality in non-faith-based
interventions.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1495

Symposium 53A

A RELIGIOUSLY TAILORED GROUP EDUCATION INTERVENTION TO ENHANCE MAMMOGRAPHY

UPTAKE AMONG MUSLIMS- DESIGN ELEMENTS AND OUTCOMES

Dr. Aasim Padela, MD MSc

Background: Faith-based health interventions adopt varied approaches to engage religious

theology and faith leaders in program design. Insights into strategies and design elements that
breed success will allow for greater theorization and model portability across faiths. This
project fills a knowledge gap regarding religiously-tailored interventions among Muslims.

Objective: To describe the design of, and participant-level outcomes related to, a religiously-
tailored peer-led group education program that addressed salient mammography-related
beliefs of American Muslims.

Methods: A multi-disciplinary community advisory board worked with academicians to design

an intervention project focused on cancer screening. Phase 1 involved focus groups and
interviews with an ethnically-diverse group of women aged 40 and older sampled from
Muslim organizations to identify salient behavioral, normative and control beliefs regarding
mammography. Phase 2 involved interviews with the same target population to elicit ideas
about intervention design. CAB members and staff used these data to design the curriculum
and messaging for Phase 3 which involved peer-led education classes aimed to enhance
mammography intention.

Results: The structural elements and messages of the classes addressed barrier beliefs in at
least one of 3 ways (i) Reprioritizing- introducing another religious belief that has greater
resonance with participants such that the barrier belief is marginalized, (ii) Reframing the
belief within a religious worldview such that it is consistent with the health behavior desired,
and (iii) Reforming- using a religious scholar to provide correct interpretations of religious
doctrine. 50 Muslim women from diverse backgrounds completed both the program and
surveys that recorded changes mammography intention and resonance with barrier and
facilitator beliefs.

Discussion: This feasibility trial demonstrated that there is great potential for leveraging
religion in designing health promotion messages, yet there is also risk for co-opting religious
S1496 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

teachings in the service of biomedicine. As behavioral interventions move from being faith-
placed to faith-based, dialogue over the theoretical bases and ethical justifications of
religiously-tailoring health messages is needed.

CORRESPONDING AUTHORS: Milkie VA, MA, Emory University; [email protected]; Hadiyah

Muhammad, MPH, n/a; [email protected]; Shaheen Nageeb, MD, Initiative on
Islam and Medicine, University of Chicago; [email protected]; Sana Malik, MSW, MPH,
DrPH, Initiative on Islam and Medicine, University of Chicago; [email protected]; Monica
Peek, MD MPH, the University of Chicago; [email protected]; Michael
Quinn, PhD, University of Chicago; [email protected]

Symposium 53B

FACTORS INFLUENCING THE IMPLEMENTATION AND SUSTAINABILITY OF MOSQUE-BASED

EXERCISE PROGRAMS FOR SOUTH ASIAN MUSLIM WOMEN

Dr. Ananya T. Banerjee, R.Kin PhD

Research shows that South Asian (SA) Muslim women have lower rates of physical activity
participation and higher rates of mortality and morbidity from diabetes compared to the
general population. A six-month mosque-based exercise intervention for SA Muslim women at
high-risk for diabetes was implemented in Ontario, Canada. Our evidence demonstrated the
exercise intervention was feasible, acceptable and sustainable. The purpose of our follow-up
qualitative study was to examine the multi-level factors influencing successful participation in
the exercise intervention among SA Muslim women. Key facilitators emerged from the direct
content analysis of twelve semi-structured interviews and were classified at four levels based
on the Socio-Ecological Model: personal, community, built environment and social.
Participants identified improved personal health, Imam leadership, accessible and gender-
specific spaces and social support from peers as pertinent factors for the implementation and
sustainability of exercise programs in mosques. The findings have particular relevance for
physical activity promotion and public health because they suggest factors necessary to
design faith-based exercise interventions for ethnic communities at risk for chronic diseases.

CORRESPONDING AUTHORS: Meena Bhardwaj, MPH, City of Toronto;

[email protected]; Paula Harvey, BMBS, PhD, FRACP, Women's College
Hospital; [email protected]; Jennifer Price, RN PhD, Women's College Hospital;
[email protected]

Symposium 53C
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1497

FAITH/SPIRITUALITY IS ASSOCIATED WITH DIET CHANGE IN NEWLY DIAGNOSED AFRICAN

AMERICAN BREAST CANCER PATIENTS

Donna Jeffe, PhD

Few studies have explored associations between ones faith/spirituality and health behavior
change over the course of cancer treatment and recovery. We prospectively enrolled 227
African American breast cancer patients (stages 0-3; mean age 56 [range 33-81] years; 28%
married; 52% >high school graduate) to a randomized controlled trial of a cancer-information
interventions impact (vs. standard of care) on behavioral outcomes. At enrollment (baseline),
a mean 6 days from first surgical post-op visit or start of neoadjuvant therapy, we collected
demographic, psychosocial, and clinical data and administered the 15-item Systems of Belief
Inventory (total scores 0-45) and the 4-item Concerns about Recurrence Scale (mean scores 1-
6); higher scores reflect greater levels of faith/spirituality and concerns about recurrence. At
2-year follow-up, we asked patients if a doctor/health professional advised them to make
changes in their diet since completing breast cancer treatment to improve prognosis and if
they changed their eating habits to improve overall health over the past two years.
Multivariable logistic regression analysis identified independent predictors of diet change
(yes/no) at 2-year follow-up among baseline factors significantly associated with diet change.
A greater proportion of patients who were advised to make a diet change actually changed
their diet compared with the proportion of patients who were not advised to make a diet
change (69/76 [90.8%] vs. 87/114 [76.3%], respectively; p=.011). Higher baseline levels of
faith/spirituality (mean 41.0 vs. 37.3; p = .001) and of concerns about recurrence (mean 3.2
vs. 2.5; p = .026) were reported by patients who changed their diet at 2-year follow-up
compared with those who did not. Greater faith/spirituality (odds ratio [OR] = 1.07, 95%
confidence interval [CI] = 1.01-1.15), being advised to make a diet change (OR = 3.97, 95% CI =
1.50-10.47), and use of endocrine therapy (OR = 0.19, 95% CI = 0.06-0.56) were independently
associated with diet change, but not study arm or concern about recurrence. Studies
examining African American breast cancer patients faith/spirituality in association with
change in other health behaviors are warranted to design targeted behavior change
interventions.

CORRESPONDING AUTHORS: Maria Perez, Washington University in St. Louis;

[email protected]; Kelsey Ruland, Washington University in St. Louis;
[email protected]; Julie Margenthaler, Washington University in St. Louis;
[email protected]; Matthew Kreuter, Washington University in St. Louis;
[email protected]; Graham Colditz, Washington University in St. Louis; [email protected]
S1498 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 54 8:00 AM-9:15 AM

THE PRESIDENT'S CANCER PANEL ON CONNECTED HEALTH: IMPLICATIONS FOR BEHAVIORAL

MEDICINE RESEARCH

David K. Ahern, PhD1, Bradford W. Hesse, PhD1, Paul Jacobsen, PhD1, Ellen Beckjord, PhD,
MPH2

1
National Cancer Institute, Rockville, MD; 2UPMC Health Plan, Pittsburgh, PA

The Presidents Cancer Panel is an external advisory committee authorized by the National
Cancer Act of 1971 with a chairperson and members appointed directly by the President of
the United States. The Panels charter calls for it to review all facets of the National Cancer
Program and then to issue reports directly back to the President on any blockages or delays
hindering the rapid execution of the program writ large. From 2014 to 2015, this external
advisory panel convened a series of workshops on one such blockage: the discontinuities in
care that occur when patients lack personal data for self-management or when
communications between patients and their care teams are jumbled or missing. The purpose
of the series was to evaluate the potential of using connected health technologies to
address these discontinuities and by so doing to improve support for proactive patient
engagement, for improved communications, and for accelerated discovery in research.
Connected health technology, in this context, refers to the broad array of mHealth, eHealth,
patient portals, wearable devices, and wireless sensors that allow for greater support to
patients away from the clinic. Although there have been significant improvements in cancer
prevention, early detection, and treatment over the past several decades, cancer care delivery
will face significant challenges in coming years as the U.S. population ages, the oncology
workforce shrinks, the costs of cancer care continue to grow, and medical advances expand
treatment options. Technology-based approaches have been proposed to address these
challenges in cancer care but research is urgently needed to identify effective and scalable
solutions. As Chair, David Ahern will provide an overview of the workshop series and the
recommendations of the Report. Brad Hesse will highlight and review the recommendations
regarding research priorities. Ellen Beckjord will discuss the relevance of the key findings and
report recommendations from the perspective of SBMs Digital Health Council (DHC). Finally,
as Discussant, Paul Jacobsen will review and discuss the implications of the Report for
behavioral medicine research on cancer prevention and control.

Symposium 54A
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1499

THE PRESIDENTS CANCER PANEL (PCP) REPORT ON CONNECTED HEALTH: RESEARCH

PRIORITIES

Dr. Bradford W. Hesse, PhD

The Presidents Cancer Panel offers recommendations to the highest levels of authority for
work that is critical to accomplish across multiple levels of influence in order to accelerate
progress against the disease. These recommendations frequently provide guidance to policy
makers and to administrative heads of government, but they also provide guidance to the
biomedical community for areas of high priority research and funding a point of keen
interest to the scientific membership of the Society of Behavioral Medicine. The specific text
for recommendations will be released in the last quarter of 2016, well before the 2017 SBM
annual meeting. Although the specific details of the recommendations will not be available
until that time, it is safe to infer from publically available materials that research will be
needed to gain a better understand of how connected health applications can serve to
support patient engagement as well as individuals pursuit of healthy behavior. Given the
topic of connected health, there will also likely be an emphasis on tool development for
consumers, with a scientific focus on collecting evidence for the efficacy and effectiveness of
tools in practice. Because of the implications of what happens when healthcare is not
connected, there will also be a need for communication research aimed at bolstering
continuity in care. Another area of likely interest from the report stems the Panels stated
interest in fostering a learning healthcare environment an objective shared by the Vice
Presidents Cancer Moonshot initiative with data from connected devices informing a new
generation of discovery and translational research. In this talk, Dr. Hesse will describe the
specific recommendations for research articulated by the Presidents Cancer Panel in detail,
highlighting their relevance to behavioral medicine. Parallels will be drawn from the ways in
which connected health research in cancer apply to more general applications across diseases.

Symposium 54B

THE PRESIDENTS CANCER PANEL (PCP) REPORT ON CONNECTED HEALTH: PERSPECTIVES

FROM SBMS DIGITAL HEALTH COUNCIL (DHC)

Dr. Ellen Beckjord, PhD, MPH

The PCP Report on Connected Health calls for creation of a learning healthcare system for
cancer that will improve prevention, detection, treatment, and survivorship. High priority
research in several areas could help achieve the goal of establishing a learning health care
system and increase the benefits of connected health for cancer. Better tools and interfaces
could be developed if more were known about how healthcare providers work together and
the factors that enhance peoples engagement with their health and healthcare. Research also
S1500 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

is needed to ensure that the vast quantities of data being generated can be used in
meaningful ways to support person-centered care. Dr. Hesse will delineate and discuss the
specific research priorities of the Report and their relevance to the field of behavioral
medicine.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1501

Paper Session 26: Enhancing Decision-Making in the Context of Prevention and Treatment
10:45 AM-11:00 AM

DIFFERENCES IN PERCEIVED BARRIERS TO HCV TREATMENT INITIATION IN METHADONE

USERS VS. IVDUS: APPLICATION OF PERCEPTUAL MAPPING

Sarah Bauerle. Bass, Ph.D., MPH1, Amy Jessop, Ph.D., MPH2, Laurie Maurer, MA, Ph.D(c)3,
Mercedes Gutierrez, MD3, Muhamed Gashat, MD, MPH2, Elizabeth Carmody, N/A3

1
Temple University College of Public Health, Philadelphia, PA; 2University of the Sciences,
Philadelphia, PA; 3Temple University, Philadelphia, PA

Background: IV drug users (IVDUs) and those on methadone have a 67-96% HCV seropositivity
rate and without treatment may develop liver cirrhosis or hepatocellular carcinoma.
However, only about 11% initiate treatment. Despite increased access to new treatments,
limited research has focused on the unique perceptions of treatment initiation in these
populations.

Methods: A survey of barriers and benefits to HCV treatment was administered to 100
methadone clients at four centers and 50 active IVDUs who utilize a needle exchange in
Philadelphia; all were HCV+. Perceptual mapping (multidimensional scaling) and vector
modeling methods were used to create 3-dimensional maps to show how barriers and
facilitators to HCV treatment decision making are related, and how message/intervention
strategies can be tailored for a specific audience. These commercial marketing techniques,
used to influence consumers toward product purchasing, can be used to understand the
unique perceptions of a decision to enhance informed medical decision making.

Results: Perceptual maps show clear conceptual differences about barriers and benefits,
beliefs about healthcare, and overall knowledge of HCV between methadone users and IVDUs.
MMT clients believed that treatment was worth it, but also were concerned about the
length of time it might take to get treated and tangible barriers such as transportation and
paperwork. IVDUs self-concept of being a drug user was seen as a significant barrier to
initiating treatment, and concern with insurance and ability to pay was also important. IVDUs
also believed that treatment was not needed because their HCV was not advanced and they
had no symptoms. These were not significant in the MMT clients, suggesting better
awareness of HCV and its progression. Neither group saw significant benefits to treatments,
although the concepts of being cured and being in charge were closest, indicating a
significant message strategy for an intervention. No issues of trust in healthcare providers
were observed, a significant finding.
S1502 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions: These methods are useful in helping understand these at-risk groups unique
perceptions regarding HCV treatment and how messages might have to differ based on drug
use. To increase informed decision making, interventions must include messages that address
negative perceptions of treatment and promote their benefits, rather than focusing on
mistrust of medical providers. But IVDUs would need better information regarding HCV in
general whereas MMT clients need strategies to overcome tangible barriers.

CORRESPONDING AUTHOR: Sarah Bauerle. Bass, Ph.D., MPH, Temple University College of
Public Health, Philadelphia, PA, 19122; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1503

Paper Session 26: Enhancing Decision-Making in the Context of Prevention and Treatment
11:00 AM-11:15 AM

SHOW ME MY HEALTH PLANS: USING A DECISION AID TO IMPROVE DECISIONS IN THE

FEDERAL HEALTH INSURANCE MARKETPLACE

Mary Politi, PhD1, Marie Kuzemchak, BA2, Abigail Barker, PhD3, Kimberly Kaphingst, ScD4,
Jingxia Liu, PhD1, Timothy McBride, PhD5, Matthew Kreuter, PhD, MPH3, Sydney Philpott, BS1

1
Washington University in St Louis School of Medicine, St Louis, MO; 2Vanderbilt University
School of Medicine, Nashville, TN; 3Brown School of Social Work, Washington University, St
Louis, MO; 4University of Utah and Huntsman Cancer Institute, Salt Lake City, UT; 5Brown
School, Washington University, St Louis, MO

Introduction: Since the Affordable Care Act was passed, over 12 million individuals have
enrolled in the health insurance marketplace. Without support, many struggle to make an
informed plan choice that meets their health and financial needs.

Methods: Show Me My Health Plans (SMHP) provides education, preference assessment, and
annual out-of-pocket cost calculator with plan recommendations produced by a tailored, risk-
adjusted algorithm incorporating age, gender, and health status. We evaluated whether
SMHP compared to healthcare.gov improved health insurance decision quality and the match
between plan choice, needs and preferences among 328 Missourians enrolling in the
marketplace.

Results: Participants using SMHP had higher health insurance knowledge (LS-Mean=78 vs. 62;
p < 0.001), decision self-efficacy (LS-Mean=83 vs. 75; p < 0.002), confidence in their choice (LS-
Mean= 3.5 vs. 2.9; p < 0.001) and improved health insurance literacy (OR 2.52, p < 0.001)
compared to participants using healthcare.gov. Those using SMHP were 10.3 times more likely
to select a silver-tier plan (p < .0001).

Discussion: SMHP can improve health insurance decision quality and can improve odds that
consumers select an insurance plan with coverage likely needed to meet their health
needs. This study represents a unique context through which to apply principles of decision
support to improve health care choices.

CORRESPONDING AUTHOR: Mary Politi, PhD, Washington University in St Louis School of

Medicine, St Louis, MO, 63110; [email protected]
S1504 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 26: Enhancing Decision-Making in the Context of Prevention and Treatment
11:15 AM-11:30 AM

FEASIBILITY AND PATIENT PERCEPTIONS OF VIDEO DECLARATIONS REGARDING END-OF-LIFE

DECISIONS BY HOSPITALIZED PATIENTS

Lisa M. Quintiliani, PhD1, Jessica Murphy, MD1, Pablo Buitron de la Vega, MD1, Katherine R.
Waite, BA2, S. Emily Armstrong, BA2, Joseph Davenport, BA3, Lori Henault, MPH2, Angelo
Volandes, MD MPH4, Michael Paasche-Orlow, MD MA MPH1

1
Boston University, Boston, MA; 2Boston Medical Center, Boston, MA; 3Boston Medical
Center, Watertown, MA; 4Massachusetts General Hospital, Boston, MA

Patients are asked to complete advanced directive forms, but may benefit from additional
ways to express themselves. A videotaped declaration could be a supplemental opportunity
and a source of information for family and clinicians. Our aim was to evaluate the feasibility
and patient perceptions of video declarations, the term we use to describe the process of
patients videotaping themselves describing their wishes. We approached hospitalized patients
at an urban safety-net hospital with attending referral. After a baseline survey, we showed a
brief video on an iPad describing 3 approaches to end-of-life care (i.e., life prolonging care,
limited care, and comfort care). We then invited participants to video themselves with the
iPad, making their own declaration about their wishes if they were very sick and asked
participants about the usefulness of this process. Videos were transcribed and coded by a
team, with multiple stakeholder input. There were 16 participants; mean age was 60 (SD=14)
years, the majority (62%) reported being either Hispanic or Black or African American and
having a high school education or lower. One participant declined. Out of 15 video
declarations, the majority were able to express their wishes about CPR (n=12) and intubation
(n=13). Nine video declarations had directives that were judged to be unclear. Participants
also frequently discussed their family structure (n=11), religious/spiritual well-being (n=8),
legacy/dignity issues (n=6), and physical symptoms (n=6). Average video length was 3:10
minutes. The majority (66%) thought this process was quite a bit or extremely helpful, and
73% would be quite a bit or extremely likely to recommend this to others. These findings
show that asking hospitalized patients to make videos describing their end-of life-care goals
was feasible and acceptable. While the majority of participants were able to describe their
wishes around CPR and intubation, a fair amount of uncertainty remained. Further research is
needed to better understand how to support patients in describing their wishes clearly.
Additional work is also needed regarding embedding video declarations into the EMR and
other platforms for maximum impact.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1505

CORRESPONDING AUTHOR: Lisa M. Quintiliani, PhD, Boston University, Boston, MA, 02118;
[email protected]
S1506 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 26: Enhancing Decision-Making in the Context of Prevention and Treatment
11:30 AM-11:45 AM

SHARED DECISION MAKING IN THE UNITED STATES: PREDICTORS AND DISPARITIES

Kiara Spooner, DrPH, MPH1, Jason L. Salemi, PhD, MPH2

1
Baylor College of Medicine, Department of Family and Community Medicine, Houston, TX;
2
Baylor College of Medicine, Department of Family & Community Medicine, Houston, TX

Background: Effective communication between patients and providers is essential to

achieving positive health outcomes and high quality of care. Shared decision making
collaborative processes involving agreed-upon approaches, goals, and expectations by both
the patient and healthcare provideris a key element of patient-centered health care.
Although research supports the value of shared decision making in clinical practice, more
evidence is needed to better understand the factors which may influence patient involvement
in health decisions. The objective of this study is to describe the national prevalence of shared
decision making in the United States (U.S.) and identify differences across patient-level
sociodemographic, health and risk perception characteristics.

Methods: Nationally representative cross-sectional data from the 2012 and 2014 iterations of
the Health Information National Trends Survey were analyzed. Administered biennially by the
National Cancer Institute (NCI) to U.S. adults 18 years and older, HINTS monitors changes in
the evolution of health communication. HINTS 4 Cycles 2 (2012) and 4 (2014) data were
collected via self-administered mailed questionnaires. Descriptive statistics and multivariable
regression analyses were conducted to examine associations between patient characteristics
and shared decision making behavior.

Results: Among survey respondents reporting at least one non-emergency visit to a

healthcare provider (N=5,631), most were Non-Hispanic (NH)-white, female, aged 35-64
years, and college educated. A majority of respondents also indicated being diagnosed with at
least one chronic disease (e.g., diabetes, hypertension, cancer) and having a regular
healthcare provider (73.7%). Slightly more than half (53.5%) of survey respondents reported
always engaging in shared decision making with a healthcare provider. After adjustment for
confounding factors, respondents perceiving themselves as completely confident in their self-
care ability, as well as those with a regular provider (odds ratio [OR], 1.62; 95% confidence
interval [CI], 1.27-2.07), were significantly more likely to indicate always participating in
shared decision making; while there was a lower odds of shared decision making among
Hispanics (OR, 0.73; 95% CI, 0.54-1.00), compared to NH-Whites.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1507

Conclusion: Patients involvement in shared decision making varies across sociodemographic

and health-related factors. These findings may inform efforts aimed at improving patient-
provider communication, increasing the uptake of shared decision making and achieving
higher quality of care.

CORRESPONDING AUTHOR: Kiara Spooner, DrPH, MPH, Baylor College of Medicine,

Department of Family and Community Medicine, Houston, TX, 77098;
[email protected]
S1508 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 27: Understanding the Role and Impact of Physical Activity among Cancer
Survivors 10:45 AM-11:00 AM

CHANGES IN MUSCLE GENE-EXPRESSION, STRENGTH AND CANCER-RELATED-FATIGUE IN

PROSTATE CANCER PATIENTS

Karen Mustian, PHD, MPH, MS1, Michelle Janelsins, PhD, MPH1, Charles Kamen, PhD, MPH1,
Ian Kleckner, PhD1, Matthew Asare, PhD, MBA, CHES2, Calvin L. Cole, PhD2, Luke J. Peppone,
PhD, MPH3, Charles E. Heckler, Ph.D., M.S. 1

1
University of Rochester Medical Center, Rochester, NY; 2University of Rochester, Rochester,
NY; 3University of Rochester, Wilmot Cancer Institute, Rochester, NY

Background: Radiation therapy (RT) and androgen deprivation therapy (ADT) impair muscular,
mitochondrial and immune function resulting in weakness and fatigue in prostate cancer
patients. We investigated the influence of an exercise intervention (EXCAP ) on expression
of muscle genes, muscular strength, and cancer-related-fatigue.

Methods: Prostate cancer patients (N=58; mean age=67), receiving RT (47%) or ADT (53%),
were randomized to 6 wks of EXCAP (7 days/wk) or standard care (RT or ADT with no
exercise) in this phase II clinical trial. RNA was isolated from muscle biopsies for microarray
analyses of 4825 mitochondrial and nuclear genes. Physical strength was assessed using
multiple repetition maximum testing (chest press and leg extension). Cancer-related-fatigue
was assessed via the Brief Fatigue Inventory (BFI). Assessments were pre- and post-
intervention. Robust multi-array average normalization was used. Analyses included analyses
of covariance (ANCOVA), correlations, and partial least squares (PLS) with cross-validation.

Results: MYH8, MYL5, ACTN3, XIRP1, MTTM, and HLA-DQB1 were significantly correlated with
muscle strength and cancer-related-fatigue (all p>2-fold down regulation in MYH8 and XIRP1
in the exercise group, no >2-fold changes in expression in the control group, and a >2-fold
difference between groups on MTTM where MTTM was down-regulated >1.5-fold in controls
with no change in exercisers (all p<0.05). ANCOVAs revealed a trend for group differences in
strength (all p<0.10) with significant group differences in fatigue (p<0.05): exercisers
improved while controls worsened. PLS suggested down-regulation of 3 novel biomarkers,
MYL5, ACTN3, and HLA-DQB1, may predict increases in cancer-related-fatigue among prostate
cancer patients.

Conclusions: Results suggest that exercise influences muscles expression of genes that
regulate muscular, mitochondrial and immune function and are linked to inflammatory
muscle disease in prostate cancer patients; these gene expression changes may mediate
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1509

changes in cancer-related-fatigue in prostate cancer patients due to cancer, RT or ADT, and

exercise. Further research is needed to validate these novel findings, to expand our
knowledge regarding the role of these biomarkers in the development of cancer-related-
fatigue, and to understand the influence of exercise on these biomarkers in prostate cancer
patients receiving RT and ADT. ClinicalTrials.Gov: NCT00815672 Funding: DOD PC061518, NCI
CA102618B, NCI CA189961.

CORRESPONDING AUTHOR: Karen Mustian, PHD, MPH, MS, University of Rochester Medical
Center, Rochester, NY, 14620; [email protected]
S1510 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 27: Understanding the Role and Impact of Physical Activity among Cancer
Survivors 11:00 AM-11:15 AM

REPLACING SEDENTARY TIME WITH PHYSICAL ACTIVITY: EFFECTS ON CANCER-RELATED

COGNITIVE IMPAIRMENT

Diane Ehlers, PhD1, Jason Fanning, PhD2, Elizabeth Awick, MS3, Susan Aguiaga, PhD3, Josh
Cosman, PhD4, Joan Severson, MS5, Arthur Kramer, PhD6, Edward McAuley, PhD7

1
University of Illinois Urbana-Champaign, Urbana, IL; 2Wake Forest University, Winston Salem,
NC; 3University of Illinois at Urbana-Champaign, Urbana, IL; 4Pfizer, Inc., 610 Main St, MA;
5
Digital Artefacts, Iowa City, IA; 6Northeastern University, Boston, MA; 7University of Illinois at
Urbana Champaign, Urbana, IL

Cancer-related cognitive impairment (CRCI) is recognized as a high priority in clinical research

due to its significant, often long-term impact on daily functioning and quality of life. The
influence of lifestyle behaviors, such as sedentary and physical activity behaviors, on CRCI is
largely unexplored. Recent research has emphasized the important health implications of
reducing sedentary time in cancer survivors. The purpose of this study was to examine the
cognitive effects of substituting 30 minutes of sedentary behavior with 30 minutes of light
activity or moderate-to-vigorous physical activity (MVPA) in breast cancer survivors. Using an
iPad-based platform, 307 breast cancer survivors aged 57.299.82 years completed
questionnaires assessing their demographics and breast cancer history, and cognitive tests
assessing their executive function and processing speed (Task-Switch, Trail Making, Swap).
Participants also wore an accelerometer for seven consecutive days to assess their sedentary
and activity behaviors. Isotemporal substitution was used to analyze the data. The
substitution of sedentary time with MVPA was associated with better reaction time on the
Task-Switch stay (=-0.19, p=0.007) and switch (=-0.19, p=0.006) trials, and on the Swap task
(=-0.15, p=0.04). A positive association was also observed between light activity and Task-
Switch stay reaction time (=0.16, p=0.05). The substitution of sedentary time with MVPA was
marginally associated with better total time on Trails A (=-0.13, p=0.06). Positive associations
with total time on Trails A (=0.22, p=0.005) and Trails B (=0.18, p=0.02) were observed
when substituting sedentary time with light activity. All effects were evident independent of
age and months since last chemotherapy treatment. Substituting 30 minutes of sedentary
time with 30 minutes of MVPA may help ameliorate impairments in executive function related
to breast cancer treatment. Interestingly, substituting sedentary time with light activity
resulted in slower times on measures of executive function and processing speed. Additional
research investigating moderators and latent classes of breast cancer survivors may illuminate
further information to explain these relationships between physical activity and CRCI.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1511

CORRESPONDING AUTHOR: Diane Ehlers, PhD, University of Illinois Urbana-Champaign,

Urbana, IL, 61801; [email protected]
S1512 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 27: Understanding the Role and Impact of Physical Activity among Cancer
Survivors 11:15 AM-11:30 AM

EXERCISE-INDUCED STRENGTHENING OF THE INFLAMMATORY CYTOKINE NETWORK IS

ASSOCIATED WITH REDUCED FATIGUE IN 348 CANCER PATIENTS

Ian Kleckner, PhD1, Calvin Cole, PhD1, Charles E. Heckler, Ph.D., M.S. 1, J. Philip Kuebler, MD2,
Amy Brockmeyer, MD3, Rezwan Islam, MD4, Michelle Janelsins, PhD, MPH1, Shelli Kesler,
PhD5, Karen Mustian, PHD, MPH, MS1

1
University of Rochester Medical Center, Rochester, NY; 2Columbus NCORP, Columbus, OH;
3
Virginia Mason, Federal Way, WA; 4Marshfield Clinic, Weston, WI; 5MD Anderson, Houston,
TX

Background. Chemotherapy is hypothesized to cause fatigue partly via dysregulation of the

inflammatory cytokine network; namely, by weakening the normal co-regulatory couplings
between concentrations of pro- and anti-inflammatory cytokines. Although preliminary
research suggests that exercise treats fatigue in cancer patients by favorably influencing
individual markers of inflammation, no studies have examined whether exercise influences
the co-regulatory couplings between concentrations of pro- and anti-inflammatory cytokines.
The purposes of this study were to assess how regulation of the inflammatory cytokine
network is (1) affected by exercise and (2) related to fatigue.

Methods. We performed simple network analyses (i.e., correlations matrices) on data from
348 cancer patients enrolled in a randomized trial of chemotherapy plus 6 weeks of
moderate-intensity walking and resistance exercise (N = 173) vs. chemotherapy alone (N =
175) in mixed-type, early/mid-stage cancer patients (mean age = 56, 93% female). At pre- and
post-intervention, patients (1) provided blood to assess concentrations of IL-1, IL-6, IL-8, IL-
10, IFN, and sTNFR1 via ELISA and (2) reported fatigue using the multidimensional fatigue
symptom inventory.

Results. Compared to chemotherapy alone, exercise plus chemotherapy strengthened the

normal co-regulatory couplings between changes in concentrations of several pro- and anti-
inflammatory cytokines; specifically, between IL-6 and all other cytokines (p < 0.05) and
between IL-10 and all other cytokines (p < 0.05). Among exercise participants, decreased
fatigue was associated with strengthened co-regulatory coupling between changes in
concentrations of IL-6 and IL-10 (p = 0.08).
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1513

Conclusions. Exercise strengthened the regulation of the inflammatory cytokine network via
strengthened co-regulatory couplings between concentrations of pro- and anti-inflammatory
cytokines. The strength of co-regulatory coupling between IL-6 and IL-10 may mediate the
beneficial effects of exercise on fatigue, considering that IL-6 and IL-10 are known to
contribute to the physical conditioning effects of exercise. Our novel methods to analyze
cytokine data may complement traditional analytic approaches in the investigation of
immune-mediated phenomena beyond this study.

CORRESPONDING AUTHOR: Ian Kleckner, PhD, University of Rochester Medical Center,

Rochester, NY, 14642; [email protected]
S1514 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 27: Understanding the Role and Impact of Physical Activity among Cancer
Survivors 11:30 AM-11:45 AM

PHYSICAL ACTIVITY AND SEDENTARY BEHAVIOR IN OLDER AND YOUNGER BREAST CANCER
SURVIVORS

Gillian Lloyd, BA1, Elizabeth Awick, MS2, Whitney A. Welch, Ph.D.3, Edward McAuley, PhD2,
Siobhan Phillips, PhD, MPH1

1
Northwestern University Feinberg School of Medicine, Chicago, IL; 2University of Illinois at
Urbana Champaign, Urbana, IL; 3Northwestern University, Chicago, IL

Physical Activity and Sedentary Behavior in Older and Younger Breast Cancer Survivors

Gillian R. Lloyd1, Elizabeth A. Awick2, Whitney A. Welch1, Edward McAuley2 & Siobhan M.
Phillips1

1
Northwestern University; 2University of Illinois at Urbana-Champaign

Background: Approximately 72% of breast cancer survivors (BCS) are over the age of 60. The
combination of treatment side-effects and comorbid conditions increases older BCS risk for
frailty, disability and premature mortality. Reducing sedentary behavior (SB) and increasing
physical activity (PA) may prevent or delay these adverse outcomes, however little is known
about how these behaviors may differ between older and younger BCS. The purpose of this
study is to examine whether SB and PA differ between older and younger BCS, overall and in
various contexts (i.e. work, leisure).

Methods: Participants (N=375) wore an accelerometer on their non-dominant hip for 7

consecutive days and self-reported their time spent in PA and SB in various contexts. We used
general linear modeling analyses to examine whether differences existed between older [ 60;
n=157; Mage=65.1 (SD=4.4)] and younger [age=50.5 (SD=6.8) BCS on total self-reported and
accelerometer-assessed PA and SB and self-reported context-specific PA and SB when
controlling for relevant covariates.

Results: Total average daily minutes self-reported and accelerometer-assessed SB were highly
correlated (r=1.0; pp=0.003). Older BCS had higher levels of total daily average minutes of
accelerometer-assessed [Mdiff= 29.3 (SE=7.9); p=0.001] and self-reported [Mdiff= 29.4 (SE=7.8);
p=0.001] SB. Younger BCS reported more time SB at work [Mdiff=119.2, (SE=20.6); p=Mdiff=36.7
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1515

(SE=8.2); p=Mdiff= 9.4 (SE=1.9); p0.001] and light [Mdiff=19.6, (SE=6.7); p=Mdiff=1.2, (SE=0.43);
p=0.01].

Conclusions: Older BCS engage in more SB and less PA and report greater SB during leisure
time than their younger counterparts. These findings suggest it may be important to consider
age when designing PA promotion and SB reduction interventions so intervention strategies
can be tailored toward relevant contexts. Future research should further explore the varying
contexts of PA and SB in older BCS to further elucidate what specific activities they participate
in during leisure time in order to target these behaviors through tailored interventions.

CORRESPONDING AUTHOR: Gillian Lloyd, BA, Northwestern University Feinberg School of

Medicine, Chicago, IL, 60611; [email protected]
S1516 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 28: Games for Health 10:45 AM-11:00 AM

ENJOYABLE, HIGHLY ACTIVE VIDEOGAMES DECREASED PHYSICAL ACTIVITY IN CHILDREN

Kristin Schneider, Ph.D.1, Jocelyn S. Carter, Ph.D.2, Cynthia Putnam, PhD2, Jacey Keeney, M.S.3,
Draycen DeCator, M.A.2

1
Rosalind Franklin University of Medicine & Science, North Chicago, IL; 2DePaul University,
Chicago, IL; 3Rosalind Franklin University of Medicine & Science, Marseilles, IL

Several reviews suggest that merely providing an active videogame (AVG) to children will not
increase their physical activity. Whether an AVG increases physical activity may vary
depending on the AVG played since AVGs vary on the extent to which they are enjoyable and
require full-body movement. We aimed to identify AVGs preferred by children that required
full body movement and explore whether providing them to children increased their physical
activity. Children aged 8-14 years (n=95; 61% male; 51% overweight/obese) listed the AVGs
they played the most. The most popular games were dance-related games (e.g., Just Dance)
and adventure-related games (e.g., Kinect Adventures). Trained research assistants who
played >100 different AVGs also rated the dance- and adventure-related games as high in
requiring vigorous movement. A subsample of participants (n=21, 52% male, 43%
overweight/obese) who owned AVGs, logged their AVG use and wore a pedometer for one
week at baseline . They then received their preferred choice from one of the dance- or
adventure-related AVGs; 12 weeks later, to reduce novelty effects, they again wore the
pedometer and logged minutes of AVG use for one week. Mixed models that controlled for
age, BMI category (lean vs overweight & obese) and month the participant wore the
pedometer (time-varying covariate), revealed a significant time x AVG minutes interaction (B =
-2378.87; 95th CI: -4712.23, -45.50) for pedometer steps. At baseline, average minutes of AVG
use was positively associated with steps (r=0.45, p=.10), but at time 2, AVG minutes was
negatively associated with steps (r=-0.59, p=.03). AVG use increased over time (B=2.26; 95th
CI: 0.50, 4.02). Contrary to our hypothesis, providing children with enjoyable AVGs that
require full body movement decreased physical activity. AVG use may have supplanted non-
AVG physical activity. Continued caution for using AVGs in the home to increase child physical
activity is warranted. Integrating all of a childs physical activity into the AVG (e.g., receive
points in the AVG for going for a walk outside) may be more successful at increasing physical
activity in children.

CORRESPONDING AUTHOR: Kristin Schneider, Ph.D., Rosalind Franklin University of Medicine

& Science, North Chicago, IL, 60064; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1517

Paper Session 28: Games for Health 11:00 AM-11:15 AM

IMPACT OF POKEMON GO ON PHYSICAL ACTIVITY: A PRE-POST ANALYSIS

Leslie Oley, MS1, David Stuck, PhD1, Luca Foschini, PhD1, Jessie Juusola, PhD1, Shefali Kumar,
MPH1, John Torous, MD2

1
Evidation Health, San Mateo, CA; 2Beth Israel Deaconess Medical Center /Harvard Medical
School, Brookline, MA

Background: In July 2016, a location-based augmented reality mobile game called Pokmon
GO was released and quickly saw millions of players. Anecdotal accounts suggest the game
can increase physical activity, as players follow prompts to move around outdoors. We sought
to quantitatively evaluate whether the game has improved activity levels and wellness by
analyzing behavior before and after starting play.

Methods: To identify Pokmon GO players and non-players, we utilized a proprietary platform

(Achievemint.com) where members can connect trackers and apps to receive rewards for
healthy behaviors. We distributed a single question survey to members who had an activity
tracker connected asking whether they were playing Pokmon GO. Affirmative responders
were sent a follow-up survey with questions such as when they started playing, general
experience, and perceived impact of the game on activity and mood. We analyzed both survey
responses as well as activity tracker step data for 30-days pre/post each players start date.
Those who responded that they were non-players composed a control group, using step data
pre/post the games launch date.

Results: Of the 1,721 people who responded to the single question survey, 597 (35%)
reported playing Pokmon GO. Of those, 340 (57%) completed the follow-up survey. Prior to
starting Pokmon GO, players were less active than non-players based on activity tracker step
data (8,726 vs 9,430 average daily steps, p=0.002). After starting, players daily average steps
increased significantly by 906 (p < 0.001) in the first 2 weeks; however, in weeks 3 and 4,
average daily steps significantly decreased from that peak by 477 (p=0.015). Users who
reported a body mass index (BMI) higher than 30 increased their daily average step count by
890 more than those with BMI lower than 30 (p=0.057).

Conclusions: Pokmon GO players had significantly increased activity levels in the short-term
after starting play, and the increase in steps was two times greater in obese players. However,
the increase in physical activity waned over time. This analysis highlights the potential for
augmented reality mobile game concepts to be utilized for improving physical activity and
wellness in certain populations.
S1518 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Leslie Oley, MS, Evidation Health, San Mateo, CA, 94401;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1519

Paper Session 28: Games for Health 11:15 AM-11:30 AM

MONSTER APPETITE: EFFECTS OF MESSAGE FRAMING ON NUTRITIONAL CHOICES IN A DIGITAL

GAME ENVIRONMENT

Maria Hwang, Ed.D.

Columbia University Medical Center (CUMC), New York, NY

Americans' health has reached a dangerous epidemic of obesity and diabetes from over
consumption and unhealthy food choices, one that creates disparities among race, ethnicity,
region, and income. In response to this national health threat, a body of research on games
for health that strive to motivate individuals to adopt healthy behaviors are growing. The
most typical approach to many health games is to provide positive reinforcement for healthy
behaviors. However, there exist other approaches that may have potential to further enhance
health awareness and experience, for example, by rewarding players to make the
conventionally unhealthy behaviors in a game environment to solidify their original belief in
healthy behaviors, otherwise known as inoculation theory.

In this study, we explore two novel approaches to encourage healthy nutritional choices in
real life by vicariously experiencing an unconventional, subversively-framed or inoculation-
based calorie game, Monster Appetite (MA), through monster avatars. We developed MA,
and an online snack shop, Snackazon, to test whether subversive or inoculation gameplay led
to healthy snack shopping behavior. There were two treatment conditions: 1) subversive
framing - MA's original goal to consume the highest-calorie snack items to keep their monster
avatar overweight, sluggish, and inactive was maintained and the end-of-day pop-up
messages were negatively-framed, and 2) inoculation-based framing: MA's goal was flipped
(consume the lowest-calorie snack items to keep their monster avatar slim and active) and the
pop-up messages were positively-framed. The two conditions with pre-existing information on
participants food habits (snacking behavior) resulted in a 2 x 1 study design (N=225).

The study showed that even if participants indicated low behavioral intention to look up
caloric information, on Snackazon both subversive and inoculation treatment participants
showed statistically significant nutritional information seeking behavior (ISB) for the online
snack items, and ISB was significantly correlated with what snack items they chose and the
reasons they list for choosing those snacks. This result fit well with the Theory of Planned
Behavior (TPB) as it emphasizes the progression from awareness and attention to a health
issue, followed by intention to change, culminating in actual behavioral change. In addition,
for the inoculation group, participants showed a statistically significant increase in healthier
snack choices post-gameplay compared to pre-gameplay. However, while the subversive
group did not choose healthier snacks, they provided statistically significantly
S1520 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

healthier reasons for making their snack choices. The study shows promising results for
unconventional approaches for nutrition game-based behavioral change studies as well as
support for traditional theories such as TPB.

CORRESPONDING AUTHOR: Maria Hwang, Ed.D., Columbia University Medical Center (CUMC),
New York, NY, 10029; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1521

Paper Session 28: Games for Health 11:30 AM-11:45 AM

DEVELOPMENT OF A MOBILE, AVATAR-BASED APPLICATION TO MONITOR TEENS PERSONAL

BODY SHAPE GOALS

Annmarie Lyles, PhD, RN1, ashish Amresh, PhD2, Michael Todd, PhD1, Rebecca E. Lee, PhD1

1
Arizona State University, Phoenix, AZ; 2Arizona State University, mesa, AZ

Background: During adolescence participation in physical activity decreases while concern

about weight increases as teens get older. Using Internet-based healthy strategies with teens
who need or want to make changes to their bodies may be a practical method to help them
participate in physical activity to attain and maintain healthy bodies.

Objective: The purpose of this pilot test was to develop and assess acceptability and feasibility
of an avatar-based, theoretically derived, mobile application entitled, Having A Positive
Perception of You Application (HAPPY App).

Methods: The HAPPY App was engineered for teens to identify what they thought they looked
like, what they wanted to look like, and what they actually looked like based on body
measurements using three avatars.

Results: The HAPPY App was pilot tested with teens ages 15 to 18 to assess for acceptability
and feasibility. A total of 42 students created and viewed their avatars. A majority of the
students were female (67%), age 16 (38.1%), white (74%), non-Hispanic (86%), and in Grade
10 (47.5%). The students had positive reactions to the avatar application. Almost half thought
it was good to be able to see their actual selves. Most of the students were comfortable
creating and viewing the avatars (95.2%) and would use the application in the future to see
how their bodies change over time (95.2%).

Conclusion: Avatar-based mobile applications, such as the HAPPY App, provide immediate
feedback and allow users to engage with images that are personalized to represent their
perceptions and actual body images. This pilot study adds to the increasing but limited
research of using games to improve health outcomes among teens. There is a need to further
adapt the HAPPY App and implore feedback from a larger number of teens including those
from diverse backgrounds.

CORRESPONDING AUTHOR: Annmarie Lyles, PhD, RN, Arizona State University, Phoenix, AZ,
85004; [email protected]
S1522 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 29: Early Influences: Child Nutrition and Obesity 10:45 AM-11:00 AM

SOCIAL NETWORK INFLUENCES IN A HOME-BASED CHILDHOOD OBESITY PREVENTION

PROGRAM

Kayla de la Haye, Ph.D.1, Sarah-Jeanne Salvy, Ph.D.2

1
University of Southern California, Los Angeles, CA; 2University of Alabama at Birmingham,
Birmingham, AL

Social Network Influences in a Home-Based Childhood Obesity Prevention Program

Social networks have been found to exert considerable influence on obesogenic behaviors and
resulting obesity in observational studies, yet are rarely the focus of obesity prevention or
treatment efforts. Social network characteristics likely to be important to the adoption and
maintenance of a healthy weight and energy balance behaviors include: social isolation,
because social connections provide support and resources needed to engage in healthy
behaviors; and weight and behavior norms, because social connections are a source of
influence via several mechanisms (e.g., normative influence, mimicry).

This study builds upon a childhood obesity prevention program called COPE (Childhood
Obesity Prevention at homE), which is a new, add-on module for existing Home Visitation
Programs (HVPs). HVPs provide in-home services to low-income mothers and their infants
across the U.S.; a population with health and obesity disparities. Specifically, this 6-month
pilot study assessed if mothers social network characteristics were related to the focal
outcomes of the COPE intervention (mother diet, physical activity, weight status; infant diet,
weight trajectory), and tested if social network characteristics moderated the intervention
effects. Fifty mothers and infants (70% Hispanic/Latino) were recruited through our
community HVP partner and randomized to receive (1) the HVP core curriculum only or (2)
the HVP core curriculum + COPE module, for six months. Assessments, conducted at baseline
and post-intervention included: (1) mothers social network characteristics, (2) mother/infant
food intake and physical activity, and (3) mothers postpartum weight retention and childrens
growth velocity.

Several features of mothers social networks, including the proportion of females in their
network, network density, and the proportion of network members who provided social
support, were significantly associated with the intervention outcomes (p < .05), and
moderated change in mother and infant outcomes. These findings suggest that family-based
obesity interventions should consider the social networks in which families are embedded,
and develop appropriate strategies to foster social connections that provide support and
healthy social influence, to bolster healthy behavior and weight outcomes.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1523

CORRESPONDING AUTHOR: Kayla de la Haye, Ph.D., University of Southern California, Los

Angeles, CA, 90089-9239; [email protected]
S1524 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 29: Early Influences: Child Nutrition and Obesity 11:00 AM-11:15 AM

CITATION AWARD WINNER

SELF-RATED HEALTH AMONG CHILDREN: WHAT DOES IT MEASURE?

Noel Kulik, PhD CHES1, Mary Cornwell, MPH, CHES2

1
Wayne State University, Detroit, MI; 2Macomb County Health Department, Fair Haven, MI

Adults self-rated health (SRH) has been a consistent independent predictor of morbidity and
mortality, and has been associated with obesity and level of physical activity (PA) among both
adolescents and adults. Little is known about SRH and children. The purpose of this study is to
examine SRH by race/ethnicity, sex, and income (school free/reduced lunch status), and to
determine its relationship to childrens self-reported quality of life (PedsQL: physical,
emotional, social, school and overall health), fruit and vegetable (F&V) consumption and PA.
Participants were 3rd through 5th grade students (N= 427; 48.5% female) from lower-, middle
or higher-income elementary schools in the Midwestern USA. Children ranged in age from 8
to 13 years (Mean: 9.7 1.0) and were mostly African American (n=156; 37.3%), Caucasian
(n=181; 43.3%) or multiracial/other (n=81; 19.5%). Demographic, PedsQL, F&V, PA and SRH
data were collected via questionnaire during one classroom visit. Children self-reported their
health as excellent (n=114; 26.7%) very good (n=171; 40.0%), good (n=111; 26.0%), fair (n=27;
6.3%) or poor (n=4; 0.9%), with a mean of 3.85 .92. SRH did not differ by school, sex, age,
grade or race/ethnicity. There was a significant, small positive correlation between F&V and
SRH (Fr: r = 0.16, p = .001; Veg: r = .20, p < .001) and PA and SRH (r = 0.11, p = .022) such that
children who ate more F&V and engaged in more PA reported higher SRH. When looking at
PedsQL variables, SRH was significantly positively correlated with physical (r = 0.28, p < .001),
emotional (r = 0.25, p < .001), social (r = 0.19, p < .001), school (r = 0.14, p = .004), and overall
(r = 0.27, p < .001) health. A multiple regression was run to predict SRH from F&V, PA and
overall PedsQL health. The model statistically significantly predicted SRH, F(4,415) = 11.824, p
< .001, adj. R2 = .094. Only overall health and vegetable consumption significantly added to
the prediction, p < .05. Findings from this study suggest that SRH can discriminate between
self-reported quality of life in children as young as 8 years of age, and may be a valid and quick
1-iterm measure of overall health in children, especially related to diet behaviors. Future
research on SRH in children is needed to understand how children qualitatively view health,
and how SRH is related to objectively assessed health by a physician.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1525

CORRESPONDING AUTHOR: Noel Kulik, PhD CHES, Wayne State University, Detroit, MI, 48202;
[email protected]
S1526 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 29: Early Influences: Child Nutrition and Obesity 11:15 AM-11:30 AM

MERITORIOUS AWARD WINNER

MORE THAN JUST FOOD: A META-ANALYSIS OF FAMILY MEALTIME PRACTICES ASSOCIATED

WITH CHILDRENS NUTRITIONAL HEALTH

Mattea Dallacker, PhD Student1, Jutta Mata, Professor2, Ralph Hertwig, Professor1

1
Max Planck Institute for Human Development, Berlin, Berlin, Germany; 2University of
Mannheim, Mannheim, Baden-Wurttemberg, Germany

Family meals can be seen as the cradle of eating behavior: by the age of 10, a child has eaten
about 10,000 meals, most of them in a family setting. A higher frequency of family meals is
known to be associated with better diet quality and lower body mass index (BMI) in children.
But what aspects of family meals are healthy for children? This meta-analysis systematically
summarizes studies on social, environmental, and behavioral attributes of family meals that
have the potential to positively influence childrens nutritional health. Six frequently
investigated family mealtime practices (43 studies, 57 effect sizes, 42,929 participants) were
identified. Positive associations between the following practices and childrens nutritional
health were consistently found in separate meta-analyses: TV off during meals (r = .08), food
quality (r = .11), parental modeling (r = .11), positive atmosphere (r = .12), and longer meal
duration (r = .20). Childrens involvement in meal preparation was associated with better diet
quality (r = .08), but also with higher BMI (r = .06). Mechanisms potentially underlying these
effects are proposed, building on Herman et al.s model of how the presence of others shapes
eating behaviors (Herman, Roth, & Polivy, 2003). The generalizability of the identified
mealtime practices is discussed, as are potential policy implications.

CORRESPONDING AUTHOR: Mattea Dallacker, PhD Student, Max Planck Institute for Human
Development, Berlin, Berlin, 14195; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1527

Paper Session 29: Early Influences: Child Nutrition and Obesity 11:30 AM-11:45 AM

WEEKEND SLEEP DELAY IS ASSOCIATED WITH SEVERITY OF OBESITY AND INCREASED SCREEN
TIME IN ADOLESCENTS WITH OBESITY

Jacqueline Hayes, MA1, Myra Altman, MA1, Katherine N. Balantekin, PhD, RD2, Denise Wilfley,
Ph.D.1, C. Barr Taylor, Ph.D.3, Joanne Williams, Ph.D.4

1
Washington University in St. Louis, St. Louis, MO; 2Washington University School of
Medicine, St Louis, MO; 3Palo Alto University/Stanford Medical School, Palo Alto, CA; 4Deakin
University, Waurn Ponds, Victoria, Australia

Weekend Sleep Delay is Associated with Severity of Obesity and Increased Screen Time in
Adolescents with Obesity

Jacqueline F. Hayes1, Myra Altman1, Katherine N. Balantekin1, Denise E. Wilfley1,

C. Barr Taylor2, & Joanne Williams3

1
Department of Psychiatry, Washington University in St. Louis

2
Center for m2Health, Palo Alto University and Stanford Medical Center

3
School of Health and Social Development, Deakin University

Sleep duration is often linked to childhood obesity; however sleep patterns, specifically
weekend sleep delay (e.g. later bedtimes on weekends than weekdays), have also been
associated with behavioral and metabolic changes that promote obesity, independent of sleep
duration. Given that adolescents commonly have later bedtimes on weekends than weekdays,
this sleep pattern may play a role in adolescent obesity and obesogenic behaviors. This study
assesses weekend sleep delay and its relation to relative weight as well as physical activity and
weekend screen time (i.e., TV, computer, video games) in adolescents with
overweight/obesity participating in an online behavioral weight loss study. At baseline,
adolescents (n=189, mean age = 14.661.60, mean zBMI = 2.050.41) were weighed and
measured, reported typical sleep and wake times on weekdays and weekends, and responded
to questionnaires assessing typical physical activity habits and number of weekend hours
engaged in screen time. Results showed that, on average, adolescents reported going to bed
S1528 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

1.381.07 hours later on weekends than weekdays. In a hierarchical regression model

controlling for age, gender, and sleep duration on weekends and weekdays, greater weekend
sleep delay was associated with greater severity of overweight (=.16, p < 0.05). In similar
follow-up models predicting behavioral outcomes, weekend sleep delay was not associated
with any differences in physical activity; however, it was significantly related to greater
weekend screen time (=.23, p < 0.01). Thus, above and beyond sleep duration, weekend
sleep delay is positively related to severity of overweight/obesity and time spent on the
computer/playing video games in treatment-seeking adolescents with obesity. Reducing sleep
delay on the weekends and promoting a consistent sleep schedule throughout the week may
be a worthwhile treatment target to optimize behavioral and weight outcomes in adolescent
obesity treatment.

CORRESPONDING AUTHOR: Jacqueline Hayes, MA, Washington University in St. Louis, St.
Louis, MO, 63110; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1529

Paper Session 30: The Role of Affect in the Context of Physical Activity 10:45 AM-11:00 AM

CITATION AND MERITORIOUS AWARD WINNER

AEROBIC TRAINING IMPROVES AFFECTIVE AND COGNITIVE RESPONSES TO DAILY STRESSORS

IN FAMILY CAREGIVERS

Adam M. Caplin, B.A & Sci1, Jordan Weiss, BSc2, Samantha J. Schilf, BA3, Kirsten Johansen,
MD3, Ben A. Hives, BKIN1, Elissa S. Epel, PHD3, Eli Puterman, PhD1

1
University of British Columbia, Vancouver, BC, Canada; 2University of Pennsylvania,
Philadelphia, PA; 3University of California, San Francisco, San Francisco, CA

Introduction: Individuals differ in their cognitive and emotional responses to stressful events,
with maladaptive responses predicting the development of physical and mental health
conditions. Observational studies show that physically inactive individuals have worse mood
and are more anxious in response to laboratory or daily stressors than those who are active.
Here, we report a randomized controlled trial of the effects of aerobic training on measures of
psychological reactivity to daily stressors in previously inactive caregivers.
Methods: Sixty-three informal family caregivers of patients with Alzheimers disease or other
dementia were randomized to either a 6-month exercise program (N=30) or a waitlist control
(N=33) if they reported < 150 minutes/week [the Center for Disease Control and Prevention
recommended minimum] of moderate-to-vigorous physical activity and high levels of
psychological stress (0.5 SD above the national mean on Cohens Perceived Stress Scale).
Participants completed ecological momentary assessments 6 times/day for 7 days pre and
post intervention. At each assessment, they reported whether a stressor occurred since their
last assessment, and if so, answered questions on stress appraisals (perceived stress,
perceived control), coping (rumination, positive reappraisal) and negative affective (NA)
responses (mean of anger, anxiety, sadness, and shame). Intent-to-treat mixed model
analyses were completed with random intercepts, fixed slopes, and restricted maximum
likelihood estimation.
Results: Change over time () between exercise and control group caregivers was significantly
different for perceived stress (p = .015) and NA (p = .004), and marginally significant for
perceived control (p = .089). Although caregivers in the control group did not change in
appraisals or NA, caregivers in the exercise group significantly increased in control [0 = 39.95,
SE = 3.01, CI = 33.95, 45.96; = 6.65, SE = 3.22, CI =0.32, 12.98] and decreased in perceived
stress [0 = 75.10, SE = 2.09, CI = 70.92, 79.29; = -5.56, SE = 1.89, CI = -9.27, -1.85] and NA
[0 = 43.44, SE = 2.43, CI = 38.58, 48.31; = -5.93, SE = 1.96, CI = -9.78, -2.08] when a
stressful event occurred.
S1530 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Summary: The benefits of exercise are widespread and alter the ways individuals respond to
stressful life events. We demonstrated novel findings showing that physical activity
strengthens adaptive appraisals and reduces negative affect in response to stressors
experienced by a previously inactive group of caregivers.

CORRESPONDING AUTHOR: Adam M. Caplin, B.A & Sci, University of British Columbia,
Vancouver, BC, V6T 1Z3; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1531

Paper Session 30: The Role of Affect in the Context of Physical Activity 11:00 AM-11:15 AM

OUTCOME EXPECTANCY MEDIATES THE RELATIONSHIP BETWEEN AFFECTIVE AND PHYSICAL

FEELING STATES AND NEXT-DAY PHYSICAL ACTIVITY

Yue Liao, MPH, PHD1, Jaejoon Song, MS2, Michael Robertson, MPH3, Karen Basen-Engquist,
PhD, MPH2

1
University of Texas MD Anderson Cancer Center, Houston, TX; 2The University of Texas MD
Anderson Cancer Center, Houston, TX; 3MD Anderson Cancer Center, Houston, TX

Background: A growing number of studies have shown affective and physical feeling states as
predictors of daily physical activity (PA) in free-living settings, yet little is known about the
mechanism underlying such effects. This study aims to test social cognitive states as
mediators of the relationship between affective and physical feeling response to PA and PA
levels next day using ecological momentary assessment (EMA). Methods: The EMA protocol
was administered via a handheld computer (PDA) for 7 consecutive days at baseline, and 5
consecutive days at 7 different time points across a 6-month period (42 EMA days total)
among 99 endometrial cancer survivors participating in the Steps to Health Study. In the
morning of each EMA day, the PDA prompted participants to record their self-efficacy (SE) for
exercising that day (1 item), as well as expected positive (EPO; 6 items) and negative
outcomes (ENO; 3 items) of exercising that day. Participants were also asked to record all
exercise sessions using the PDA, and to complete the Exercise-induced Feeling Inventory (EFI;
12 items) and Somatic Sensations (SS; 10 items) both pre- and post-exercise. Difference scores
were calculated to indicate feeling responses to exercise. Each feeling response was linked
with the SE/EPO/ENO and the overall physical activity level (PAL; derived from a composite
measure of accelerometer and EMA self-report data) the following day. Multilevel mediation
models (1-1-1) were used to examine the within-person (WP) and between-person (BP)
effects using EFI/SS as the predictor, SE/EPO/ENO as the mediator, and PAL as the outcome.
Results: ENO mediated the effect of affective response to exercise (EFI) on next day PAL at the
BP level (beta=0.176, SE=0.086, p=0.040), suggesting more positive affective response led to
higher PAL through a decrease in ENO (beta=-0.036, SE=0.009, pConclusion: Affective and
physical feeling responses to exercise predict daily PA level partly through their effects on
individuals negative outcome expectancy. Future interventions could consider targeting this
mechanism for PA promotion.

CORRESPONDING AUTHOR: Yue Liao, MPH, PHD, University of Texas MD Anderson Cancer
Center, Houston, TX, 77230-1439; [email protected]
S1532 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 30: The Role of Affect in the Context of Physical Activity 11:15 AM-11:30 AM

MERITORIOUS AWARD WINNER

THE ACUTE BIDIRECTIONAL RELATIONSHIPS OF CHILDRENS AFFECTIVE STATES AND PHYSICAL

ACTIVITY: THE ROLE OF CHRONIC STRESS

Cheng K. Fred Wen, MPH1, Yue Liao, MPH, PHD2, Jimi Huh, PhD1, britni R. belcher, PhD, MPH1,
Eldin Dzubur, MS1, Adam Leventhal, PHD3, Genevieve F. Dunton, PhD, MPH1

1
University of Southern California, Los Angeles, CA; 2University of Texas MD Anderson Cancer
Center, Houston, TX; 3USC, Pasadena, CA

Introduction: There is a growing literature suggesting that affective states are acutely and bi-
directionally related to physical and sedentary behavior. Yet, these relationships and potential
moderators of these relationships are rarely examined in children.

Methods: A sample of 180 children (mean age=9.6 years old, 51.7% girl, 53.9% Hispanic) and
their mothers completed an ecological momentary assessment (EMA) study that examined
parenting factors and obesity risk. Across 1 week, children wore accelerometers on waist to
measure moderate-to-vigorous physical activity (MVPA) and sedentary behavior (SB), and
received 3-7 EMA prompts/day by smartphones asking about their current positive and
negative affective states. Perceived chronic stress, age, sex, and ethnicity were measured at
baseline for children and mothers. Multilevel regression models tested the within-(WS) and
between- person (BS) associations of: (1) time spent in MVPA or SB within the 60 mins before
each EMA prompt predict affective states at that prompt and (2) affective states at each EMA
prompt predict time spent in MVPA or SB in the 60 mins after that prompt. Interaction terms
were created by multiplying the both the WS- and BS- versions of the main predictor of each
model with childrens and mothers chronic stress scores. Models controlled for child sex, age,
race, body mass index (BMI) percentile, and affective states and activity levels at the prior
EMA prompt.

Results: When engaged in more SB than usual in the 60 mins before an EMA prompt, children
reported lower positive affect at that prompt (WS; b=-0.01, SE=0.01, p < 0.1). In contrast,
engaging in more MVPA than usual before a prompt was associated with higher positive affect
at that prompt (WS; b=0.02, SE=0.01, p < 0.1). Mothers baseline chronic stress moderated
this association (WS; b =-0.02, SE=0.01, p < 0.05), such that the positive affective response to
MVPA was weaker for children of mothers with higher levels of chronic stress.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1533

Conclusions: Promoting MVPA and decreasing SB may help acutely improve childrens
affective states. These acute effects may be impacted by mothers stress, underscoring the
potential importance of modulating mothers stress and the need to further understand the
mechanism underlying childrens affective response to MVPA.

CORRESPONDING AUTHOR: Cheng K. Fred Wen, MPH, University of Southern California, Los
Angeles, CA, 90013; [email protected]
S1534 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 30: The Role of Affect in the Context of Physical Activity 11:30 AM-11:45 AM

AEROBIC EXERCISE TRAINING IMPROVES AFFECT IN NORMAL HEALTHY YOUNG ADULTS

Richard Sloan, PhD1, Paula McKinley, PhD2, Peter Shapiro, MD1, Eli Puterman, PhD3, Matthew
Bartels, MD2, Keith Diaz, PhD1, Kathleen McIntyre, LCSW1, Martina Pavlicova, PhD1, Jennifer
Scodes, MS4, Pamela Flood, MD5

1
Columbia University Medical Center, New York, NY; 2Montefiore Medical Center, Bronx, NY;
3
University of British Columbia, Vancouver, BC, Canada; 4New York State Psychiatric Institute,
New York, NY; 5Stanford University, Palo Alto, CA

Objective: Aerobic conditioning is associated with beneficial effects on affect but most data
come from observational studies. Here we report the effect of a randomized controlled trial
of aerobic training program on measures of affect.

Method: Normal healthy sedentary young (age 20-45 y) adults were randomized to a 12-
week aerobic training program or a waitlist control condition and measures of depressive
affect (Beck Depression Inventory, BDI), hostility (Cook Medley scale, CKM), and anxiety
(Spielberger Trait index) were collected at study entry (T1) and at 12 weeks (T2). Longitudinal
linear mixed effect models with generalized estimating equations were fit separately for each
outcome. Final models were adjusted for age, sex, BMI, body fat percentage, and the
alternative Mediterranean diet index.

Results: 119 participants were randomized to the exercise (n=60, 32 women) or wait list
(n=59, 31 women) conditions. Mean age was 31.3 ( 5.9) years. Mean study entry VO2max
was 30.4 6.9 and 31.0 6.4 ml/K/min in the exercise group and control groups respectively.
After training, VO2max increased to 34.7 8.4 in the exercise group but did not change (30.1
6.2) in the control group. Within the exercise group, BDI scores significantly decreased by
29% from T1 to T2 (ES = 0.71, p = .0470); within the waitlist group, the BDI scores increased by
9%, though not significantly (ES = 1.09, p = .4871), while adjusting for covariates. Within the
exercise group, CKM scores significantly decreased by 1.50 points from T1 to T2 (ES = -1.50, p
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1535

= .0091); within the waitlist group, CKM scores increased by 0.25 points, though not
significantly (ES = 0.25, p = .5838), in the adjusted model. Exercise training had no effect on
trait anxiety.

Conclusions: In this randomized controlled trial of a 12-week aerobic training program

administered to healthy, young, sedentary adults, aerobic capacity increased significantly as
predicted in the training but not the control group. Training also led to significant reductions
in depressive affect and hostility but not anxiety. These findings are noteworthy especially in
light of the fact that participants had normal levels of affect at study entry.

CORRESPONDING AUTHOR: Richard Sloan, PhD, Columbia University Medical Center, New
York, NY, 10032; [email protected]
S1536 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 31: Engagement with Genetic Testing and Information 10:45 AM-11:00 AM

LONG-TERM OUTCOMES OF TELEPHONE VS. IN-PERSON GENETIC SERVICES DELIVERY

Mary K. Interrante, MA1, Beth N. Peshkin, MS, CGC1, Heiddis Valdimarsdottir, PhD2, Marc D.
Schwartz, PhD3

1
Georgetown University, Washington, DC; 2Reykjavik University, Reykjavik, N/A, Iceland;
3
Georgetown University/Lombardi Cancer Center, Washington, DC

Genetic counseling and testing for BRCA1/2 mutations has become integral to the clinical care
of women at risk of hereditary breast and ovarian cancer (HBOC). As the demand for cancer
genetic counseling and testing has risen, alternatives to traditional face-to-face genetic
counseling and testing are needed to ensure widespread access. We conducted a multi-site,
randomized noninferiority trial comparing telephone delivery to standard in-person delivery
of genetic counseling and testing for HBOC. Over the short-term, we found that telephone
delivery was less expensive and yielded comparable (i.e., noninferior) genetic counseling
(knowledge, decision conflict, satisfaction) and psychosocial (distress, quality of life)
outcomes. In this report, we examined 12-month outcomes from this trial, to determine if
telephone delivery remains noninferior to in-person delivery over the long-term.

Eligible women were age 21-85 who did not have newly diagnosed or metastatic cancer, were
at high-risk for HBOC and were self- or physician-referred for genetic counseling at one of our
study sites. Overall, 669 women were randomized to receive usual care (UC) or telephone
counseling (TC). Of those, 514 completed the 12-month follow-up assessment. There were no
differences in attrition across the two study groups (2= 0.88, p = 0.767).

At 12-months post-randomization, TC was statistically noninferior to UC on all outcomes.

Cancer distress (d = -0.50, upper bound of 97.5% CI, 1.99) did not cross the 4-point
noninferiority limit, genetic testing distress (d = -0.54, upper bound of 97.5% CI, 1.08) did not
cross the 3-point noninferiority limit, physical function (d = 0.77, lower bound of 97.5% CI, -
0.74) and mental function (d= -0.17, lower bound of 97.5% CI, -1.73) did not cross their 2.5
point noninferiority limits and satisfaction with genetic testing decision (d = 0.15, lower bound
of 97.5% CI, -0.31) did not cross its 1-point noninferiority limit. In follow-up analyses adjusting
for genetic test result and potential confounders, the results were unchanged.

These results, confirming the long-term noninferiority of TC compared to UC, provide further
evidence that remotely delivering genetic services is a safe and effective approach to
increasing access and reducing the costs of genetic testing.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1537

CORRESPONDING AUTHOR: Mary K. Interrante, MA, Georgetown University, Washington, DC,

20007; [email protected]
S1538 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 31: Engagement with Genetic Testing and Information 11:00 AM-11:15 AM

ENGAGEMENT WITH GENETIC INFORMATION AND UPTAKE OF GENETIC TESTING: THE ROLE
OF TRUST AND PERSONAL CANCER HISTORY

Megan Roberts, PhD1, Jennifer M. Taber, PhD2, William Klein, PhD3

1
NCI, Rockville, MD; 2Kent State University, Kent, OH; 3NCI/NIH, Rockville, MD

Background: Targeted screening and treatment using genetic technologies is increasingly

common. For individuals with a family and/or personal history of cancer, genetic testing might
identify those at especially high risk. This is particularly important when there are
corresponding guidelines for the management of high risk individuals (e.g., those with
BRCA1/2 mutation, Lynch Syndrome). However, use of genetic testing remains low among
those eligible. This may be partially due to lack of communication about genetic testing and
low trust in genetic testing information sources. No studies have explicitly examined the
association between receiving genetic information from trusted sources and cancer-related
genetic test uptake within a national sample of US adults. Our objectives were to better
understand (1) the extent to which individuals trust the sources from which they receive
information about health-related genetic tests, (2) whether level of trust for sources of
genetic information might be related to health-related genetic testing uptake, and (3)
whether factors such as cancer history or numeracy play a role in the latter association.

Methods: We used cross-sectional data from the Health Information National Trends Survey
(HINTS 4, Cycle 3). Our study sample included individuals who responded yes to the
following survey item: Genetic tests that analyze your DNA, diet, and lifestyle for potential
health risks are currently being marketed by companies directly to consumers. Have you
heard or read about these genetic tests? (n=1117). All analyses accounted for the complex
survey design to achieve population-based estimates.

Results: Although respondents trusted information from health professionals the most, they
were less likely to report hearing about genetic testing from them compared to the television
(p < 0.01). Regardless of source, higher levels of trust in the information source from which
participants heard about genetic tests were associated with increased odds of genetic testing
uptake. This was particularly true among individuals with a personal history of cancer.
Numeracy did not moderate the association between trust and uptake of testing.

Conclusion: Overall, these results suggest that trust is important among individuals with a
personal cancer history. Because providers were among the most trusted sources of health
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1539

information, they may be important sources for genetic testing information, particularly for
individuals with a personal cancer history.

CORRESPONDING AUTHOR: Megan Roberts, PhD, NCI, Rockville, MD, 20814;

[email protected]
S1540 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 31: Engagement with Genetic Testing and Information 11:15 AM-11:30 AM

COMMUNICATING COMPLEXITY: ANALYSIS OF THE COMMUNICATION OF CANCER GENETIC

TEST RESULTS THREE MONTHS POST DISCLOSURE

Charite Ricker, MS, LCGC1, Rachel Koff, MS, CGC2, Chenxu Qu, N/A3, Julie O. Culver, MS, LCGC,
CCRP4, Duveen Sturgeon, MSN, ACNP, AGN5, Katrina Lowstuter, MS6, Christine A. Hong, B.S.,
M.S.7, Brian Allen, M.S., C.G.C8, Courtney Rowe-Teeter, MS, LCGC9, Alexandra Lebensohn, MS,
CGC10, Kerry Kingham, MS, CGC, LGC11, Nicolette Chun, MS12, Peter Levonian, MS, LCGC13,
Meredith Mills, BA14, Anne-Renee Hartman, Uri Ladabaum, MD, MS15, Kevin McDonnell, MD1,
James M. Ford, MD16, Stephen Gruber, MD PhD MPH1, Allison W. Kurian, M.D., M.Sc.17,
Gregory Idos, MD4

1
University of Southern California, Los Angeles, CA; 2Stanford Health Care, Stanford, CA; 3USC
Norris Comprehensive Cancer Center, LA, CA; 4USC Norris Comprehensive Cancer Center, Los
Angeles, CA; 5Norris Comprehensive Cancer Center, Los Angeles, CA; 6University of Southern
California, South Pasadena, CA; 7University of Southern California/Norris Comprehensive
Cancer Center, Anaheim, CA; 8Myriad Genetics, Inc., Salt Lake City, UT; 9Stanford Healthcare,
San Jose, CA; 10Stanford Healthcare, Stanford, CA; 11Stanford Healthcare, Palo Alto, CA;
12
Stanford Health Care Systems, Stanford, CA; 13Stanford, Stanford, CA; 14Stanford University,
Stanford, CA; 15Stanford University School of Medicine, Stanford, CA; 16Stanford University
School of Mecidine, Stanford, CA; 17Stanford University, Atherton, CA

Research on the communication of genetic test results in families have focused on

predominately non-Hispanic White mutation positive families. Little is known about this
process for racially and ethnically diverse individuals who undergo genetic testing. As part of a
prospective clinical trial of patient experience with a 25-gene test, participants were sent a
questionnaire three months after results disclosure, including questions about sharing test
results. This is an interim analysis of the first 562 56.2% (n=562/1000) respondents.
Responses were analyzed by results groups (Negative Neg, n=297; Variant of Uncertain
Significance VUS, n=207; Positive Pos, n=59).

Self-reported race/ethnicity is 44.8% non-Hispanic White, 37.9% Hispanic, 10.1% Asian, and
2.8% Black. The majority (97.2%, n=546) had shared their test results. The Pos group was
over six times more likely to encourage family members to undergo genetic testing than the
Neg (OR=6.6, p < 0.0001, 95% CI [3.2, 14.7]) and VUS groups (OR=6.1, p < 0.0001, 95% CI [2.9,
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1541

13.8]). When compared to the Neg and VUS groups the Pos group was 12.4 and 13.8 times
more likely, respectively, to report family members having undergone genetic
testing (OR=12.4, p p < 0.00001, 95%CI [5.7,35.3]). All Pos respondents (n=59) shared their
test results. When asked if they had encouraged family to undergo genetic testing, Asian
respondents less frequently reported having done so than non-Hispanic Whites (p < 0.05).
Additionally, Asian participants in the Pos group responded in the negative more frequently
when asked if their family members underwent genetic testing (p < 0.05).

Three months post genetic testing, communication of results was very high among all result
groups. However, individuals in the Pos group were more likely to encourage testing in their
family members and their family members were more likely to undergo genetic testing,
suggesting that participants correctly understood the implication of their results and were
able to convey that information to their relatives. Differences by race/ethnicity merit further
exploration. Longitudinal data will continue to be collected on these and the subsequently
accrued participants, allowing for expanded analyses.

CORRESPONDING AUTHOR: Charite Ricker, MS, LCGC, University of Southern California, Los
Angeles, CA, 90033; [email protected]
S1542 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 31: Engagement with Genetic Testing and Information 11:30 AM-11:45 AM

BRCA-POSITIVE PREVIVORS INFORMATION NEEDS REGARDING HEREDITARY BREAST AND

OVARIAN CANCER RISK

Marleah Dean, PhD1, Courtney L. Scherr, PhD2, Meredith L. Clements, BA, MA1

1
University of South Florida, Tampa, FL; 2Northwestern University, Chicago, IL

Introduction: BRCA1 and BRCA2 carriers are at significantly increased risk to develop breast
and ovarian cancer. Because of this genetic risk, BRCA-positive, unaffected patientsreferred
to as previvorsexperience uncertainty regarding decision-making. Information is a key
mechanism for managing such uncertainty. While previous research has explored cancer
patients and survivors information needs, the information needs of previvors are likely to be
distinct. Framed by the Messaging Model for Health Communication Campaigns (MMHCC)
which incorporates components of social marketing to provide a rigorous approach for the
development of communication campaignswe assessed previvors information needs and
behaviors in order to inform the development of an educational intervention to manage
uncertainty and promote uptake of appropriate prevention activities.

Methods: 25 qualitative, in-depth interviews were conducted with BRCA-positive, female,

English-speaking previvors. Interviews were audio-recorded, transcribed verbatim by a
professional transcription service, and checked for accuracy by study team members. The
constant comparison method of grounded theory was utilized to code the interview data.
Inter-coder reliability was established using Krippendorfs alpha. Coders were reliable at
=.92.

Results: Qualitative analysis of previvors experiences revealed information needs were

related to the stage in their health journey. Specifically, a four-stage model was identified: 1)
pre-testing information needs, 2) post-testing information needs, 3) pre-management
information needs, and 4) post-management information needs. Two recurring dimensions of
desired knowledge also emerged within the stagesmedical knowledge and personal/social.

Conclusion: Results suggest previvors information needs are distinct from cancer patients
and survivors needs and change overtime as they navigate their health journey. Future
research should confirm these results with a more diverse population, and if confirmed,
develop educational materials to assist patients in managing uncertainty and reducing
decisional conflict.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1543

CORRESPONDING AUTHOR: Marleah Dean, PhD, University of South Florida, Tampa, FL,
33620; [email protected]
S1544 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 32: Helping Mom Know Best: Maternal Health Research 10:45 AM-11:00 AM

IMPROVING MOTHERS DIETARY INTAKE AN ANALYSIS OF INDIVIDUAL AND FAMILY

MEDIATORS IN A FAMILY-BASED INTERVENTION

Lucy Horton, MPH, MS1, Donald J. Slymen, PhD2, Elva M. Arredondo, PhD2, Humberto Parada,
Jr., PhD, MPH3, Guadalupe X. Ayala, PhD, MPH2

1
San Diego State University Research Foundation, San Diego, CA; 2San Diego State University,
San Diego, CA; 3University of North Carolina at Chapel Hill, Chapel Hill, NC

Purpose. The US Hispanic/Latino population is estimated to almost double between 2014 and
2050, representing 25% of the US population. Alongside this rapid growth, disparities in health
outcomes between racial/ethnic groups persist, and continue to lead to more illnesses and
premature death for communities such as US Hispanics/Latinos. Implementing effective,
culturally appropriate interventions to improve the dietary intake of Hispanics/Latinos is,
therefore, a priority to reduce health disparities in this rapidly growing population.

Methods. This study describes dietary changes among Hispanic/Latino mothers, and
mediators of those changes over a 10-month period from an RCT conducted in a rural region
along the US-Mexico border in California, US. A convenience sample of 361 families (children
between 7-13 years old and their mothers, and in a 25% subsample, fathers) was recruited
from community locations and using flyers and mass media. Intervention families (n=180)
completed an 11-session community health worker-delivered home-based intervention
designed to improve dietary outcomes. Remaining families (n=181) were part of a delayed-
treatment control condition.

Results. The mean age of mothers was 39 years (SD=7.86). Most were born in Mexico (82%),
and were either overweight (36%) or obese (50%). Daily servings of fruits was higher among
mothers in the intervention condition (mean=1.86, SE=0.14) compared to those in the control
condition (mean=1.47, SE=0.14) at 10-months post-baseline (p0.05). Mothers in the
intervention condition reported consuming a lower percent energy from fat (mean=30.0%,
SE=0.26% versus 31.0%, SE=0.27%; p0.01) and a higher diet quality (mean=2.93, SE=0.07)
compared with mothers in the control condition (mean=2.67, SE=0.07; p0.01). These
changes were mediated by improvements in behavioral strategies to increase fiber and lower
fat intake and family support for vegetable purchasing, as well as decreased unhealthy eating
behaviors and perceived family barriers to healthy eating.

Discussion. Both individual (behavioral strategies to increase fiber and lower fat intake,
reducing unhealthy behaviors) and family (family support for vegetable purchasing, perceived
family barriers to healthy eating) mediated changes in mothers dietary intake. Future
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1545

interventions should consider approaches that target mediators of change, in addition to

focusing on the dietary behavior itself.

CORRESPONDING AUTHOR: Lucy Horton, MPH, MS, San Diego State University Research
Foundation, San Diego, CA, 92123; [email protected]
S1546 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 32: Helping Mom Know Best: Maternal Health Research 11:00 AM-11:15 AM

BEHAVIORAL & PSYCHOSOCIAL SCREENING & DECISION AIDS FOR MOTHERS OF YOUNG
CHILDREN: DESIGN & LESSONS LEARNED

Lorraine Walker, MPH, EdD1, Heather Becker, Ph.D.2, Bo Xie, Ph.D.3, Bobbie S. Sterling, RN,
MSN, PhD4, Sherry Hendrickson, PhD, APRN, CNS-BC2

1
University of Texas at Austin, Austin, TX; 2The University of Texas at Austin, Austin, TX;
3
School of Nursing, Austin, TX; 4University of Texas at Austin, Georgetown,, TX

Behavioral and psychosocial health (B&PH) is important to health after pregnancy, but is often
overlooked during healthcare visits. Screening tools are a strategy to increase attention to
B&PH (including diet, physical activity, tobacco and alcohol use, mood, and body image).
Moreover, inclusion of decision aids eliciting womens preferences may promote a more
patient-centered screening process. Aim 1 of this study was to identify an acceptable format
and method of receiving results after B&PH screening. Aim 2 was to identify key
values/preferences for inclusion in decision aids related to B&PH change-decisions. Individual
interviews were done with 26 women of diverse income and race/ethnicity (7 White, 16
Latina, 2 Black, 1 Asian). Women were recruited by mailed flyers or flyers posted in a clinic
serving children of low-income families. Womens mean age was 28 (range 18 to 40), with 1-5
children and the youngest 3 years. During interviews, women first completed a B&PH
screening questionnaire, and then responded to open- and closed-ended questions about
design of materials to convey screening results. The interview also included their ratings of 11
values/preferences (from behavioral theories and maternal health research) for use in a
decision aid. Analysis involved computing percentages and the chi-square, where appropriate.
Findings included: women preferred a feedback form entitled Moms Health (54%) over
using terms such as dashboard or report card. They preferred screening results shown in a
graphic format (56%) with 3-level, positively worded labels (42%) rather than other wording.
Hispanic versus non-Hispanic ethnicity was unrelated to format preferences. Phone/computer
was preferred for receiving results over receiving results from a care provider. The 4 top
ranked values/preferences for inclusion in a decision aid were: most important to being a
mother, most needing improvement, most motivated to change, and most family support for
a change. 24 of 26 women rated the screening and decision aid as very useful in helping
mothers to make changes to improve health. Our findings provide direction for
communicating screening results and design of decision aids for B&PH changes to improve
health of mothers of young children.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1547

CORRESPONDING AUTHOR: Lorraine Walker, MPH, EdD, University of Texas at Austin, Austin,
TX, 78701; [email protected]
S1548 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 32: Helping Mom Know Best: Maternal Health Research 11:15 AM-11:30 AM

A MULTIFACETED COMMUNITY PROGRAM TO SUPPORT BREASTFEEDING IN KANSAS

Lisette Jacobson, PhD, MPA, MA1, Brenda Bandy, BS, IBCLC2, Martha Hagen, MS, RD, LD,
IBCLC3

1
University of Kansas School of Medicine-Wichita, Wichita, KS; 2Kansas Breastfeeding
Coalition, Inc., Manhattan, KS; 3Kansas Department of Health and Environment, Topeka, KS

Background: The Communities Supporting Breastfeeding (CSB) program is a multifaceted

community approach to improve breastfeeding support in Kansas. This session covers all
aspects of the program including goals, program evaluation, and shared communities
experiences.

Methods: The CSB designation as defined by the Kansas Breastfeeding Coalition (KBC)
requires meeting specific criteria within six sectors of the community: local breastfeeding
leadership, peer support, maternity care practices, societal support, employer support, and
childcare provider support. The KBC provides an assessment, hospital staff education,
coalition building assistance, continuity of breastfeeding care, and education to home
visitors. The CSB program is now in its third year. During year one and two, a survey to assess
progress toward the designation was administered to all mothers who benefited from
breastfeeding services.

Results: Since July 2014, eleven communities representing diverse populations have achieved
the CSB designation. During program years one and two, five and six communities
respectively achieved the CSB designation within the allotted 10-month time period. About
20 mothers in each of the respective communities participated in the survey (N=211). The
majority of women believed local breastfeeding leadership, peer support, and hospital
support provided adequate breastfeeding support in their community. Women were less
familiar with programs that promote breastfeeding among employers and childcare
providers. Further, year one communities appear to benefit from improved exclusive
breastfeeding rates that may be associated with the CSB designation.

Conclusion: The CSB program informs other Kansas communities in their effort to build a
culture of breastfeeding support and is now in its third year with five new communities. The
program is currently listed as an emerging practice on the Innovation Station website of the
Association of Maternal & Child Health Programs.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1549

CORRESPONDING AUTHOR: Lisette Jacobson, PhD, MPA, MA, University of Kansas School of
Medicine-Wichita, Wichita, KS, 67214; [email protected]
S1550 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 32: Helping Mom Know Best: Maternal Health Research 11:30 AM-11:45 AM

CHANGES IN FRUITS AND VEGETABLES AVAILABILITY AND CONSUMPTION AMONG LOW-

INCOME PREGNANT WOMEN: HEALTHY EATING ACTIVE LIVING

Ru-Jye Chuang, DrPH1, Shreela Sharma, PhD RD LD2, Gregory Bounds, MPH2

1
UTHealth School of Public Health, Houston, TX; 2University of Texas School of Public Health,
Houston, TX

Availability of fruits and vegetables (FV) increases consumption among adults, and is known as
a protective factor to obesity. Healthy Eating Active Living (HEAL) is a multi-component,
community-based, Community Health Worker (CHW)-led six-week program designed to
promote healthy eating, physical activity, and intention to breastfeed among Medicaid-
eligible pregnant women in Houston, TX. Pregnant women are recruited through clinics and
participate in weekly 90-minute facilitated and experiential, group-education sessions. The
purpose of this pilot study is to evaluate the feasibility of HEAL in increasing home nutrition
environment and daily FV consumption, and if the impact vary by participants perceived food
security (PFS).

This was a one-group pre-post evaluation design. Availability to FV (23 items), daily
consumption (24 items), and PFS (6 items) were measured using self-administered surveys.
Ninety-two Medicaid-eligible pregnant women completed both surveys at baseline and post-
intervention from October 2014 - September 2015. More than half (52.2%) of the women
reported low food security at baseline. McNemars test was conducted to evaluate the change
of home availability of FV between the high/low-PFS groups, and Wilcoxon signed-rank test
for FV consumption.

Results showed that participation in HEAL significantly increased home availability of 11 FV (p

< 0.05) and increased womens daily consumption of FV (p < 0.05). When stratified by
perceived food security, low-PFS women had increased availability of 9 FV and daily
consumption of FV pre-to-post-intervention (p < 0.05). For their high-PFS counterparts, HEAL
helped increase availability of 5 FV items and daily consumption of vegetables (p < 0.05) but
not fruit.

This pilot study demonstrates the initial feasibility of using multi-component, community-
based, CHW-led intervention approach to increase home availability and daily consumption of
FV among low-income pregnant women, especially in women with low perceived food
security.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1551

CORRESPONDING AUTHOR: Ru-Jye Chuang, DrPH, UTHealth School of Public Health, Houston,
TX, 77030; [email protected]
S1552 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Panel Discussion 14 12:15 PM-1:15 PM

WEB AND MOBILE TECHNOLOGIES: ACTIVITY TRACKERS, FOOD INTAKE PROGRAMS AND
BEHAVIORAL ASSESSMENTS FOR PERSONALIZED HEALTH

Sachiko T. St Jeor, Ph.D., RDN, FADA,FAHA1, Craig A. Johnston, PhD2, John P. Foreyt, Ph.D.3,
Steven N. Blair, PED4, Barbara E. Millen, DrPH., RD, FADA5

1
University of Nevada School of Medicine, Reno, NV; 2University of Houston, Houston, TX;
3
Baylor College of Medicine, Houston, TX; 4University of South Carolina, Columbia, SC;
5
Boston Nutrition Foundation, Inc, Westwood, MA

Consumers are motivated to use physical activity trackers and web-based food intake and
health programs, but their effectiveness and potential in improving and maintaining overall
health needs further investigation. The use of such innovative and interactive mobile
technologies provide powerful research opportunites that can help us better target
a personalized and integrative health approach. Research can also be facilitated using the
consumer as the informant providing important data and insight into individualized
approaches, changing behaviors and understanding variations in outcomes. This panel aims
to present personalized approaches and discuss their strengths,limitations and applications to
research.

CORRESPONDING AUTHOR: Sachiko T. St Jeor, Ph.D., RDN, FADA,FAHA, University of Nevada

School of Medicine, Reno, NV, 89519; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1553

Panel Discussion 15 12:15 PM-1:15 PM

STRATEGIES FOR ENGAGING, RECRUITING AND RETAINING DIVERSE WOMEN IN MEDICAL

RESEARCH

Aisha T. Langford, PhD, MPH1, Jennifer Wenzel, PhD, RN, CCM, FAAN2, Ruth J. Geller, MHS3,
Marjorie R. Jenkins, M.D., M.Ed., F.A.C.P.3

1
New York University School of Medicine, New York, NY; 2Johns Hopkins School of Nursing,
Baltimore, MD; 3U.S. Food and Drug Administration, Silver Spring, MD

The importance of increasing the representation of women and minorities in medical research
has gained recognition. Barriers to the participation of these groups in research have been
documented and discussed. Progress will be achieved by moving from understanding barriers
to implementing and evaluating solutions. In January 2016, the US FDA Office of Womens
Health launched the Diverse Women in Clinical Trials Initiative. The chair of the proposed
panel led a group of national expert collaborators in the development of professional
education webinars hosted through the Association of Clinical Research Professionals. The
chair is a national leader in sex and gender-specific womens health research and educational
programming innovations. The panelists are academic faculty with extensive research and
leadership experience in relevant fields including decision science, health communication,
health disparities, and the participation of underrepresented groups in clinical trials. Panelists
will describe strategies implemented in their work to increase the participation of diverse
women in medical research, discussing the evidence base, results, and feasibility of each
approach. Aspects of diversity to be discussed include race/ethnicity, age, co-morbidities,
rural location, and socioeconomic status. A Q&A session will provide opportunity for
attendees to interact with panelists. After attending this session, attendees will be able to
describe and apply appropriate strategies to increase the feasibility of research participation
for diverse women; describe effective applications of digital health to the engagement,
recruitment, and retention of diverse women in research; and develop culturally competent
communication about research participation tailored to diverse women. The panel will equip
attendees with cutting-edge strategies to engage diverse women in behavioral and medical
research, in order to achieve effective, acceptable, and culturally competent interventions.
This panel represents a collaboration between the US FDA Office of Women's Health and
innovative academic leaders within the clinical trials space.

CORRESPONDING AUTHOR: Aisha T. Langford, PhD, MPH, New York University School of
Medicine, New York, NY, 10016; [email protected]
S1554 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Panel Discussion 16 12:15 PM-1:15 PM

THE INTEGRATED PRACTICE UNIT (IPU) AS TRANSLATION ACCELERATOR: REAL-WORLD

IMPLEMENTATION IN DIABETES, PAIN, AND PRIMARY CARE

James E. Aikens, PhD1, William D. Tynan, Jr., PhD, ABPP2, Ellen Poleshuck, PhD3, Andrea
Cherrington, MD MPH4, Amy Huebschmann, MD, MS5, Lara Dhingra, PhD6, Kathryn E. Kanzler,
PsyD, ABPP7

1
University of Michigan, Ann Arbor, MI; 2American Psychological Association, Washington, DC;
3
University of Rochester Medical Center, Rochester, NY; 4University of Alabama at
Birmingham, Division of Preventive Medicine, birmingham, AL; 5University if Colorado School
of Medicine, Aurora, CO; 6MJHS Institute for Innovation in Palliative Care, New York, NY;
7
University of Texas Health San Antonio, San Antonio, TX

With the advent of population-based health, successful translation and implementation of

behavioral medicine interventions will require a clear economic understanding of healthcare
delivery. Although most chronic conditions require a coordinated treatment team, healthcare
is often fragmented, redundant, and inefficient. The fee-for-service payment model is a
significant obstacle to improving healthcare delivery because it only reimburses for the
quantity of service, not its quality, outcome, or efficiency.

This panel will provide in-depth examination of the Integrated Practice Unit (IPU; Porter and
Lee, 2013) model for improving illness management, outcomes, quality, and costs for virtually
any chronic condition. In this model, treatment units are organized around specific patient
problems rather than clinical specialties.

Although many existing multidisciplinary clinical teams are structured this way, few are true
IPUs. A fully-developed IPU also focuses on systematically improving care quality, coordinating
patients transitions across inpatient and outpatient settings, and monitoring key clinical
outcomes and treatment costs. Importantly, its clinical objectives are to address quality and
outcomes rather than RVU generation, and its business objectives are to target improved
outcomes, efficiency, and growth. Effective IPUs then expand to new geographic locations and
use their data to establish their bundled care rates for defined populations.

Following an overview of the general IPU model, panel members will present real-world
exemplars of how IPUs are being implemented to address diabetes, chronic pain, and primary
care. The panel will also discuss how researchers can use IPUs existing data to increase the
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1555

utility of clinical trials, advance the most effective treatment models, approach population
care from an economically-informed perspective, and advocate for improved healthcare
policies in a language understood by administrators and insurers.

CORRESPONDING AUTHOR: James E. Aikens, PhD, University of Michigan, Ann Arbor, MI,
48103; [email protected]
S1556 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Panel Discussion 17 12:15 PM-1:15 PM

BEHAVIORAL MEDICINE SCIENTISTS IN INDUSTRY: OPPORTUNITIES, CHALLENGES AND

LESSONS LEARNED

Cynthia M. Castro Sweet, PhD1, Heather Cole-Lewis, PhD, MPH2, Leanne Mauriello, Ph.D.3,
Danielle Blanch-Hartigan, PhD, MPH4

1
Omada Health, Inc., San Francisco, CA; 2Johnson and Johnson Health and Wellness Solutions,
New York, NY; 3Spectrum Health System, Grand Rapids, MI; 4Bentley University, Stoneham,
MA

The explosion of mHealth and eHealth technologies has opened new career pathways for
behavioral medicine scientists, and created innovative opportunities to apply expertise in
behavior theory and scientific methods to diverse, health-related organizations outside of
academic settings. Many health-related organizations (from small start-ups to large multi-
national organizations) benefit from behavioral medicine specialists to inform the creation,
application and evaluation of new health behavior-based tools and services for varied
audiences. This panel of SBM members will present their experiences in bringing behavioral
medicine expertise to life in careers outside of academia, and will highlight the unique
opportunities and obstacles across different organizations and settings. The panelists will
discuss the roles and responsibilities of their positions, the impact of their work on their
organizations, and challenges experienced as academically-trained scientists in non-academic
settings. The first speaker will discuss her role as the lead outcomes researcher for a digital
behavioral medicine company focused on chronic disease prevention. The second speaker will
discuss her career integrating evidence-based science on human behavior change with
technology across various sectorsconsumer healthcare, medical devices, and
pharmaceutical--in a large healthcare corporation. The third speaker will discuss her role in
bringing innovative lifestyle management programs to patients within a large, not-for-profit
health care system. The final speaker will discuss her perspective as an early-career
researcher consulting with a start-up company. Through sharing their experiences, the
panelists aim to expose students and early-career scientists to the varied career paths after
behavioral medicine training, and further collaborative discussion between behavioral
medicine scientists across industry and academia.

This presentation is supported by the Industry Collaboration subcommittee of the SBM Digital
Health Council.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1557

CORRESPONDING AUTHOR: Cynthia M. Castro Sweet, PhD, Omada Health, Inc., San Francisco,
CA, 94111; [email protected]
S1558 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Panel Discussion 18 12:15 PM-1:15 PM

INTEGRATING PATIENT REPORTED DATA INTO BEHAVIORAL PRAGMATIC TRIALS: CHALLENGES

AND SOLUTIONS

Wendy J. Weber, ND, PhD, MPH1, Susan M. Czajkowski, Ph.D.2, Lynn DeBar, PhD, MPH3, Doug
Zatzick, MD4

1
National Center for Complementary and Integrative Medicine/ National Institutes of Health,
Bethesda, MD; 2National Cancer Institute, Rockville, MD; 3Kaiser Permanente, Center for
Health Research, Portland Oregon, Portland, OR; 4University of Washington, Harborview,
Level 1 Trauma Center, Seattle, WA

In the last few years there has been a growing interest by the public and the scientific
community in pragmatic clinical trials that test research hypotheses, which will directly inform
the health care system. This interest has been driven by a number of factors including the
high cost of traditional efficacy studies and the exclusion of many individuals from efficacy
trials resulting in results that do not generalize to many patients. The proposed discussion
session will include presentations that will provide an overview of pragmatic trials for
behavioral interventions; how patient reported outcomes can be incorporated into pragmatic
trials; examples from two large scale pragmatic trials that are incorporating patient reported
outcomes; and provide resources for planning and conducting pragmatic trials. Dr. Weber will
highlight an overview of pragmatic trials for behavioral interventions. Dr. Czajkowski will
present on how pragmatic trials can capture patient reported outcomes (PROs), and give
examples of some common PRO measures used in these trials. Dr. Debar will present the
challenges of including patient reported outcomes of pain measures in a large pragmatic trial
evaluating integration of multidisciplinary services within the primary care environment
versus usual care. Dr. Zatzick will present about the challenges of collecting patient reported
information to determine participant eligibility for a large pragmatic trial to evaluate a
coordinated care delivery program for patients treated for acute trauma to prevent PTSD and
other related conditions. Dr. Weber will provide a summary of resources developed by the
NIH Health Care System Research Collaboratory for investigators who are planning and
conducting pragmatic trials (www.nihcollaboratory.org). The session will close with a panel
discussion with all presenters to answer questions from the audience.

This session will provide attendees with a strong overview of how to overcome challenges to
incorporating patient-reported outcomes into pragmatic trials. Attendees will also be
informed of resources and tools available to help them better understand the complexity of
planning and conducting pragmatic trials.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1559

CORRESPONDING AUTHOR: Wendy J. Weber, ND, PhD, MPH, National Center for
Complementary and Integrative Medicine/ National Institutes of Health, Bethesda, MD,
20892; [email protected]
S1560 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Panel Discussion 19 12:15 PM-1:15 PM

USING CONSUMER PRODUCTS IN BEHAVIORAL RESEARCH: LESSONS LEARNED FROM THE

FRONTLINES

Ernesto Ramirez, PhD, MS1, Sarah Kunkle, MPH2, Job Godino, Ph.D.3, Sheri J. Hartman, PhD4,
Chad D. Rethorst, Ph.D.5, Lisa Gualtieri, PhD, ScM6, Valentin Prieto-Centurion, MD7

1
Fitabase, San Diego, CA; 2Fitbit, San Francisco, CA; 3UC San Diego, La Jolla, CA; 4Department
of Family Medicine and Public Health, UC San Diego, La Jolla, CA, USA, La Jolla, CA; 5UT
Southwestern Medical Center, Dallas, TX; 6Tufts University School of Medicine, Boston, MA;
7
University of Illinois at Chicago, Chicago, IL

Over the last five years, Fitbit and other consumer health devices and applications have
rapidly become key tools for measurement and intervention studies. Consumer products
provide novel and user-friendly methods for behavioral data collection including, but not
limited to, physical activity, heart rate, sleep, dietary behaviors, and weight. Additionally,
most, if not all, consumer products are designed to collect data over longer periods of time
than what is commonly gathered by typical research grade devices (e.g., accelerometers)
and subjective measurement methods (e.g., questionnaires). In this panel, we will discuss the
opportunities available to behavioral researchers when consumer devices are included as part
of a measurement and/or intervention study. This panel represents key stakeholders in the
current consumer device and research ecosystem. Each panelist will present their work and
participate in a moderated discussion. The panelists have been involved with supporting
and/or conducting research using consumer products, including: 1) a representative from
Fitbit, the leading consumer physical activity and sleep tracking device brand; 2) an industry-
academic partner with Fitabase, specializing in data management of behavioral data produced
by consumer tools; 3) researchers who have deployed various consumer products as part of
validation, measurement, and intervention studies across diverse participant populations.
These researchers have designed and conducted innovating technology-enabled studies with
children and adolescents, cancer survivors, patients with COPD, mental health patients, and
historically underserved populations. Audience members will learn about the current best
practices for using consumer products, including current validation outcomes, and how to
best design studies for longitudinal data collection, interactivity, and maximal participant
engagement.

CORRESPONDING AUTHOR: Ernesto Ramirez, PhD, MS, Fitabase, San Diego, CA, 92116;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1561

Panel Discussion 20 12:15 PM-1:15 PM

ACCELERATING INNOVATION WITH THE SPEEDING RESEARCH INTERVENTIONS (SPRINT)

PROGRAM

Cynthia A. Vinson, PhD, MPA1, Heather Greenlee, ND, PhD, MPH2, Lisa S. Miller, PhD3, Nora L.
Nock, PhD, PE4, Edmund Pendleton, MS5, Amanda Marin-Chollom, MA6

1
Division of Cancer Control and Population Sciences, National Cancer Institute, Bethesda, MD;
2
Mailman School of Public Health, Columbia University, New York, NY; 3University of
California, Davis, Davis, CA; 4Case Western Reserve University, Cleveland, OH; 5University of
Maryland, College Park, MD; 6Columbia University Medical Center Mailman School of Public
Health & The Graduate Center, City University of New York, New York, NY

During the summer of 2016, the National Cancer Institute (NCI) kicked off the SPeeding
Research INTerventions (SPRINT) Program which is designed to foster, grow and nurture an
innovation ecosystem for behavioral interventionists. The program is run by instructors with
extensive startup and teaching experience and provides real world, hands-on training on how
to successfully incorporate innovations in cancer control into successful products. The
ultimate goal is to create scalable research-tested behavioral interventions that are ready to
be put into real world practice and which will reach a large audience of users.

Ten teams consisting of a Principal Investigator, Mentor and Entrepreneurial Lead participated
in the 8-week pilot of SPRINT. The teams participated in both a 3-day opening and a 1-day
closing in-person training session at NCI and weekly webinars. Teams also interviewed at
least 30 potential customers, competitors and stakeholders. Teams integrated the
information acquired during the interview with quantitative data to refine their respective
intervention/product and develop a viable business model for their NCI-funded behavioral
interventions.

The first team will share how the SPRINT training influenced plans for dissemination and
implementation (D&I) of a Spanish language nutrition and physical activity program for breast
cancer survivors. At the beginning of the SPRINT training, the team had the goal of scaling
their intervention to reach a wider audience of Latina breast cancer survivors. After
completing 50 interviews they pivoted to creating a website and/or phone app focused only
on nutrition, which can be used by a larger and more diverse group of cancer survivors. Going
forward, in order to facilitate delivering the product to the consumer at a faster pace, the
team intends to think of scalability when beginning to design and test new interventions.
S1562 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

The second team will discuss how SPRINT helped the team in pursuing D&I of their
intervention which involves an 'assisted' exercise technology. They will discuss the challenges
faced in scaling-up their 'assisted' cycling prototype and modifications made based on lessons
learned from SPRINT training and 'customer' interviews.

The third team will present on SPRINTs impact on D&I of a research-tested food label training
program. The team started the program with a vague idea of selling an app to consumers to
improve their food label reading skills. They ended SPRINT with a strategy for adoption,
implementation and sustainability of a web-based nutrition literacy tool to help employers
promote healthy food choices and reduce costs associated with diet-related chronic
conditions.

An overview of the course design and content as well as evaluation results will be presented
along with a discussion of future opportunities for SPRINT training by NCI representatives.

CORRESPONDING AUTHOR: Cynthia A. Vinson, PhD, MPA, Division of Cancer Control and
Population Sciences, National Cancer Institute, Bethesda, MD, 20892; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1563

Panel Discussion 21 12:15 PM-1:15 PM

GETTING STARTED WITH QUALITATIVE METHODS IN CANCER RESEARCH: TIPS FOR DESIGN,
DATA COLLECTION AND ANALYSIS

Shirley M. Bluethmann, PhD, MPH, Michelle A. Mollica, PhD, MPH, RN, OCN, Katharine
Rendle, PhD, MSW, MPH, Erin Kent, PhD

National Cancer Institute, Rockville, MD

There is increasing emphasis on use of qualitative and mixed-methods approaches in the field
of cancer prevention and control research. Qualitative methods can be used to systematically
capture contextual data in ways that are difficult to do with quantitative approaches alone.
These methods can be used to provide rich insights on a particular topic, or they can serve as
the foundation for additional quantitative research and intervention development. Planning
and managing qualitative research, however, can be challenging for those new to the field.
This panel will present a broad overview of how to conceptualize, conduct, and analyze
qualitative research projects, with a focus on interview methods in cancer prevention and
control research. This session is geared toward early-stage investigators and those that are
new to qualitative research.

CORRESPONDING AUTHOR: Shirley M. Bluethmann, PhD, MPH, National Cancer Institute,

Rockville, MD, 20850; [email protected]
S1564 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Panel Discussion 22 12:15 PM-1:15 PM

APPLYING FOR GRADUATE FUNDING FROM THE NATIONAL SCIENCE FOUNDATION AND THE
NATIONAL INSTITUTE OF HEALTH

Margarita Sala, BS1, Amy Heard, BA2, Stephanie Manasse, MS3, Austin S. Baldwin, PhD4

1
Southern Methodist University, Dalas, TX; 2Loyola University Chicago, Chicago, IL; 3Drexel
University, Philadelphia, PA; 4Southern Methodist University, Dallas, TX

For graduate students and trainees, applying for grants from the National Science Foundation
(NSF) or the National Institute of Health (NIH) is an important experience in their professional
development. Receiving one of these grants guarantees funding during graduate school,
opens opportunities for professional development, allows for the freedom to conduct
independent research, and positions the trainee well to be competitive for future funding.
However, these grants are highly competitive and guidance is limited, and thus the application
process may seem overwhelming and daunting. In this panel, three graduate students who
have been awarded pre-doctoral funding from the NIH or NSF will discuss their perspectives
on preparing applications for the NSF Graduate Research Fellowship Program (GRFP) and the
NIH Pre-doctoral Fellowship (NRSA F31), emphasizing their thoughts regarding what
constitutes a successful application. In addition, an experienced reviewer will discuss his
experience reviewing NIH Pre-doctoral Fellowship grants, emphasizing the review criteria set
by NIH and his insights into what constitutes a good application. Attendees will have the
opportunity to ask questions to the panelists.

CORRESPONDING AUTHOR: Margarita Sala, BS, Southern Methodist University, Dalas, TX,
75204; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1565

Panel Discussion 23 12:15 PM-1:15 PM

RESULTS AND LESSONS FROM COMPASS: NATIONAL IMPLEMENTATION OF EVIDENCE-BASED

COLLABORATIVE CARE MODEL

Claire Neely, MD, MD1, Karen J. Coleman, MS, PhD2

1
Institute for Clinical Systems Improvement, Bloomington, MN; 2Kaiser Permanente Southern
California, Pasadena, CA

Introduction: Collaborative care for the treatment of mental and physical health conditions in
primary care settings has been shown to be effective in controlled trials. We present
outcomes from a national dissemination and implementation project for depression and
comorbid diabetes and/or cardiovascular disease, funded by the Center for Medicare &
Medicaid Innovation. Barriers and facilitators that influence successful implementation and
sustainability of this model across diverse populations, organizations and geographies will be
discussed.

Background: The comorbidity of depression with diabetes and cardiovascular disease

increases health care costs. Recently, randomized trials have shown that collaborative care
models for depression can be expanded to comorbid conditions such as diabetes. Until now,
wide-scale dissemination and implementation of this collaborative approach had not been
tested.

Methods: 3,609 patients in 18 primary care systems and 172 clinics were enrolled in the Care
for Mental, Physical and Substance Use Syndromes (COMPASS) program as part of a project to
implement a proven chronic care model for patients with depression and diabetes and/or
cardiovascular disease. COMPASS involved a minimum of monthly contact with care
managers and weekly case review by a team of consulting physicians and psychiatrists. Eligible
adult patients met these criteria: Patient Health Questionnaire (PHQ9) score 10 or higher and
another poorly controlled medical condition; diabetes mellitus (HgbA1c > 8%) and/or
cardiovascular disease (blood pressure > 140/90 mmHg).

Results: Of 3,609 patients with uncontrolled disease at enrollment, 40% achieved depression
remission or response, 23% glucose control, and 58% blood pressure control during a mean
follow-up period of 11 months. Depression improvement was directly related to the
frequency of care management contact. Patients and clinicians were satisfied with COMPASS
care. There were large variations in outcomes across medical groups. Challenges in
implementation included changing scope and practice of physicians, psychiatrists, and existing
care managers and care management systems.
S1566 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions: Results from COMPASS provide support for institutionalizing proven

collaborative care models in high volume primary care settings. Widespread implementation
and improved patient outcomes were feasible through a collaborative effort among diverse
health care systems; however, both implementation and outcomes varied greatly.

CORRESPONDING AUTHOR: Claire Neely, MD, MD, Institute for Clinical Systems Improvement,
Bloomington, MN, 55425; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1567

Panel Discussion 24 12:15 PM-1:15 PM

MIND-BODY RESEARCH: A PANEL DISCUSSION OF SCIENTIFIC AND FUNDING TRENDS,

PRIORITIES AND STRATEGIES

Crystal L. Park, PhD1, Lanay M. Mudd, PhD2, Frederick (Rick) Hecht, MD3, Beth C. Bock, PhD4

1
University of Connecticut, Storrs Mansfield, CT; 2National Center for Complementary and
Integrative Health, Bethesda, MD; 3University of California, San Francisco, San Francisco, CA;
4
Brown Medical School, Providence, RI

This panel session will provide a stimulating discussion of the current landscape and future
prospects of mind-body science and is designed to promote extensive interaction between
panelists and audience. The session will begin with a brief overview of current NCCIH funding
priorities and opportunities by Lanay Mudd, program director at NCCIH. Two grantees with
substantial NCCIH support, Rick Hecht and Beth Bock, will provide their perspectives on
cutting-edge developments in mind-body research and their own strategies for maintaining
adequate resources to pursue research. Panelists will discuss the importance of thinking
systematically about sustained funding for ongoing lines of research and consider what makes
funding proposals stand outand what can sink them. The moderator, Crystal Park, will take
questions from the audience and will also be ready with prepared questions to foster dialog.

CORRESPONDING AUTHOR: Crystal L. Park, PhD, University of Connecticut, Storrs Mansfield,

CT, 06269; [email protected]
S1568 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Panel Discussion 25 12:15 PM-1:15 PM

COMPLEX BEHAVIORAL MEDICINE PATIENTS: ADVANCING THE NEED FOR BEHAVIORAL

MEDICINE SERVICES VIA CASE STUDIES

Kelly Gilrain, PhD1, Cori E. McMahon, PsyD2, David A. Moore, PsyD3, Gonzalez A. Efrain,
Psy.D., MSClinPharm, ABPP4, Kaplan M. Lynne, Ph.D.5, Petrongolo Jennifer, MSW, MA6

1
Cooper University Hospital, Haddonfield, NJ; 2Cooper University Hospital/ MD Anderson
Cancer Center at Cooper, Camden, NJ; 3La Salle University, Philadelphia, PA; 4University of
Miami School of Medicine, Miami, FL; 5Children's Hospital of Philadelphia, Philadelphia, PA;
6
La Salle University, Lancaster, PA

Behavioral Medicine IP Consultation Teams are becoming more prevalent across hospital
systems to address the psychological needs of medical patients in a manner that differs
greatly from Social Work and Psychiatry services. The reasons for consulting an IP Behavioral
Medicine Team are varied from adjustment concerns of mood and anxiety related to medical
issues, trauma symptoms after injuries, new medical diagnoses that are causing emotional
distress and varying types of losses related to medical concerns, as well as advisement on
cognitive issues for our patients struggling with dementia versus delirium versus depression.
While many of these cases are straightforward, Behavioral Medicine teams are more regularly
being asked to provide insight into the more complex patient, particularly when there is no
medical rationale to explain a patients pain, nausea, inability to use a limb, refusal to eat or
engage with care or understand altered mental status. Many consultations now cross over to
more complex challenges in assessment, management, and recommendations. These cases
are typically associated with factitious, somatization and conversion disorders, which may also
be complicated further by personality issues. Additionally, the capacity evaluations requested
are oftentimes multifaceted due to cognitive issues, patient versus familial desires, as well as
ethical concerns.

This Panel Discussion will showcase a variety of complex patient consultations received by
Behavioral Medicine Teams across the country as well as those patients who would benefit
from having Behavioral Medicine input. These new and complex consults are expanding the
horizons of typical Behavioral Medicine requests from adjustment, anxiety or depression to a
more intricate consult which must take into account biopsychosocial factors, the
interconnectedness of physical and emotional wellbeing and how psychological functioning
impacts medical presentations and subsequent medical care. The panel will provide insight on
how to assess, conceptualize, and work with patients and their medical teams to ensure
understanding of the medical patients psychological needs within the context of a complex
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1569

and often vague symptom presentation. Clinicians from Cooper University Hospital (Camden,
NJ), University of Miami, Department of Psychiatry and Behavioral Sciences (Miami, FL);
WellSpan Behavioral Health (York, PA), and CHOP (Philadelphia, PA) will discuss these complex
cases.

CORRESPONDING AUTHOR: Kelly Gilrain, PhD, Cooper University Hospital, Haddonfield, NJ,
08033; [email protected]
S1570 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Panel Discussion 26 12:15 PM-1:15 PM

GROUP VISITS AND GROUP PSYCHOTHERAPY IN INTEGRATED PRIMARY CARE: PROMOTING

CHILDRENS MENTAL AND PHYSICAL HEALTH.

Amber J. Landers, PhD1, Shireen F. Cama, MD1, Amelia Swanson, PhD2, Bianca S. Shagrin, MD3

1
Cambridge Health Alliance/Harvard Medical School, Cambridge, MA; 2University of
Massachusetts Medical School/UMass Memorial Medical Center, Worcester, MA; 3Cambridge
Health Alliance, Cambridge, MA

Group interventions are a useful intervention to promote change via accessible exercise,
social and education opportunities for families at their primary care clinic. This panel will
discuss two group modalities in primary care (a psychotherapy group and a group visit) that
were developed for promoting health behaviors among children and families These
interventions will be considered from a multidisplinary team-based model and cultural-
ecological theory that considers interplay of the social environment on children and families.
Panel members will discuss the following groups from these frameworks, including lessons
learned and implications for training.

Parent group Led by psychologist, psychiatry resident, and psychology fellow in primary care
clinic. Offered supportive therapy, mindfulness and education, as well as a training
opportunity. Will discuss challenges and opportunities at the intersection of adult and child
work in clinical practice and training, logistics and future goals.

Wellness group led by pediatrician and conducted with multidisciplinary team (psychology,
social work, nursing, nutrition). Offered health intervention and exercise activity (zumba!) as
well as training opportunity. Will discuss logistics, outcomes of team-based approach, and
future goals.

CORRESPONDING AUTHOR: Amber J. Landers, PhD, Cambridge Health Alliance/Harvard

Medical School, Cambridge, MA, 02138; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1571

Symposium 55 1:30 PM-2:45 PM

BEHAVIORAL INTERVENTIONS FOR OBESITY: A DEBATE ON THE STATE OF THE EVIDENCE

Dawn K. Wilson, Ph.D., FSBM1, Peter Kaufmann, PhD2, Robert M. Kaplan, PhD3, Karina
Davidson, PhD, MASc4

1
University of South Carolina, Columbia, SC; 2NHLBI, Bethesda, MD; 3Stanford University, Palo
Alto, CA; 4Columbia University Medical Center, New York, NY

This debate will focus on several themes common to the proposed presentations that include:
1) The importance of measuring patient centered health outcomes in contrast to surrogate
markers, 2) Interpretation of null RCTs in light of study power, patient population, and study
designs, and 3) Evidence requirements for public policy. A brief overview of these themes and
the overall importance of the session from a national perspective will be provided by the
chair.

Symposium 55A

A NULL RESULT IS NOT A TRIAL FAILURE: AN IN-DEPTH EXAMINATION OF THE IMPACT OF

TRIAL DESIGN ON CLINICAL OUTCOMES.

Dr. Peter Kaufmann, PhD

Many large NIH trials produce null results. Dr. Kaufmann was the co-author of an important
paper in Clinical Trials that described changes in the primary outcome of the Look AHEAD trial
necessitated by low event rates (http://www.ncbi.nlm.nih.gov/pubmed/22334468). The
presentation will describe how clinical trial design influences outcome, with implications for
clinical application and future clinical trials.

Symposium 55B

EFFECT SIZES AND PRIMARY OUTCOMES IN LARGE NIH-BUDGET, CARDIOVASCULAR-RELATED

BEHAVIORAL RANDOMIZED CONTROLLED TRIALS
S1572 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Robert M. Kaplan, PhD

Dr. Kaplan will summarize the results of all large-budget behavioral trials funded by NHLBI and
by NIDDK. In the context of current results, the presentation will suggest new directions for
behavioral clinical trials. Dr. Kaplan published a review article on this topic in Feb. 2016,
Annals of Behavioral Medicine, http://www.ncbi.nlm.nih.gov/pubmed/26507906.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1573

Symposium 56 1:30 PM-2:45 PM

DIGITAL HEALTH TOOLS TO SUPPORT POPULATION HEALTH IN AN INTEGRATED DELIVERY AND

FINANCING SYSTEM

Ellen Beckjord, PhD, MPH, Darren Olson, MBA, Ryan E. Cummings, MS, Amber Blackwood,
MPH, CHES

UPMC Health Plan, Pittsburgh, PA

Evidence-based digital health interventions have the potential to extend the reach and
enhance the impact of behavioral medicine. As part of an integrated delivery and financing
system responsible for three million covered lives across commercial and governmental
products, the UPMC Health Plan has been working to leverage digital tools to optimally
manage the health and wellbeing of its members. This work includes the design,
implementation, and evaluation of digital solutions focused on a variety of target behaviors in
the service of multiple health-related goals. In this symposium, we will feature three digital
tools developed and used by the UPMC Health Plan for population health management. The
first presentation will focus on a member-facing mobile application; its features; and its
importance as a gateway for members into a longitudinal and positive relationship with
their health insurer (Olson). Next, we will discuss the translation of an internally-developed,
evidence-based stress management program into a mobile solution, with an emphasis on the
practices used to develop and evaluate a minimum viable product (Cummings). The third
presentation will detail analytical approaches to optimizing member use and completion of a
web-based, cognitive behavioral intervention for mild to moderate depression and anxiety
(Blackwood). Our Discussant (Beckjord) will review what we have learned about best practices
as well as challenges across these three solutions, including the importance of iterative
development; the capture and use of consumer-generated data; defining member
engagement in actionable ways; and using member input to provide digital solutions that
meet and are consistent with consumers needs, preferences, and values.

Symposium 56A

USING A MEMBER APP TO ALIGN CONSUMER EXPECTATIONS IN THE MARKETPLACE WITH

THEIR HEALTH PLANS

Darren Olson, MBA

S1574 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Consumer expectations are at an all-time high when it comes to top-of-the-line service and
user experience, and differentiating oneself from the competition is a challenge for many.
Health insurance is not excused. As Moores law accelerates the affordability and availability
of technology, consumers expect their health plans to keep pace. In 2012, the UPMC Health
Plan developed and released a mobile application for members and over the past year has
conducted six studies with 143 participants to iteratively improve the solution and align it
with member expectations. User testing focused on consumers attitudes about and behaviors
with the app as well as roadblocks or barriers to utilization. We used the findings to prioritize
the roadmap of feature development in our sprint cycle. We will describe multiple features of
the app; their utility for serving health plan members; and iterative improvements made to
features including enabled fingerprint ID login; a streamlined registration process; use of push
notifications; and the integration of other digital health tools into the mobile app
environment. Results show that moving to fingerprint ID login reduced login errors by 30%;
that redesigning key elements of the information architecture in response to user feedback
decreased time spent on the login screen by 50%; increased monthly app downloads by 5%;
increased app utilization by 25%; and increased monthly unique return visitors by 10%; and
that streamlining the registration process decreased the time required to create a new
account by two to three minutes. We will discuss these features and associated results in the
context of supporting preventive health behavior; health behavior change; and optimizing
consumer experiences as part of an integrated delivery and financing system serving three
million members across commercial and governmental insurance products.

CORRESPONDING AUTHORS: Darren Olson, M.B.A, UPMC Health Plan; [email protected];

Ryan Cummings, M.S., UPMC Health Plan; [email protected]; Alexis Miller, UPMC
Health Plan; [email protected]; Jeremy Burton, MSIS, UPMC Health Plan;
[email protected]; Himen Patel, BA, UPMC Health Plan; [email protected]; Ellen
Beckjord, PhD, UPMC Health Plan; [email protected]; Kim Jacobs, UPMC Health Plan;
[email protected]

Symposium 56B

THE BREATHE STRESS MANAGEMENT APP: LONGITUDINAL EVALUATIVE RESEARCH TO

VALIDATE A LIFESTYLE DIGITIZATION CONCEPT

Ryan E. Cummings, MS

Stress is a common experience among health care consumers, and often the most frequently
endorsed lifestyle risk factor reported on health risk assessments. The UPMC Health Plan has
had an evidence-based stress management intervention available to members for over a
decade; however, digitizing this intervention and offering it as a mobile solution is necessary
to achieve adequate reach and to optimize impact. In 2015, in collaboration with a digital
entertainment and gaming company, we designed and user-tested a prototype of a mobile
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1575

stress management solution called Breathe. This minimum viable product (MVP) used
ecological momentary assessment (EMA) to measure levels of stress throughout the day and
provided guided instruction on deep breathing for stress reduction. Using qualitative
methods, we conducted a 14-day longitudinal study to assess the effectiveness of the app to
reduce and manage stress over time. We guided the study with 30 individuals (15 iPhone
users and 15 Android users). We recruited participants through a remote usability platform,
UserTesting. We conducted an initial evaluative exercise wherein users access the application
and are asked to rank their overall health and stress levels and their expectations of the app.
We then guided them through the setup process. asking them to speak aloud about their
experiences. Data collection included recording participants on-screen device interactions,
voice, and inter-study response to survey questions. At one week, we conducted a "check-in"
with our users to validate their on-going participation. The test concluded with asking
participants to explain their experience through both written and verbal prompts. Results
showed that most users expected the app to be contextually aware, with notifications
functioning as suggestions for techniques. Regarding app features, users viewed tagging
reports of stress with triggers as cumbersome and confusing. However, they found the check-
in interaction (EMA) to be very easy and intuitive. Participants found milestone badges very
motivating and perceived the color palette as calming, and some users felt less stress just
from viewing the app. Finally, participants stated that the setup process was lengthy, and
warned it might deter people from completing the setup process. These results allowed us to
provide recommendations to improve the overall experience and align continued
development with prioritization of feature sets as we continue to translate evidence-based
behavior change interventions into mobile solutions for health plan members.

CORRESPONDING AUTHORS: Kelsey Humphries, MHCI, UPMC Health Plan;

[email protected]; Darren Olson, MBA, UPMC Health Plan; [email protected]; Alexis
Miller, UPMC Health Plan; [email protected]; Timothy Cline, Ph.D., MCC, UPMC Health
Plan; [email protected]; Ellen Beckjord, Ph.D., UPMC Health Plan; [email protected];
Manoj Anand, MET, Schell Games; [email protected]; Sabrina Culyba, MET, Schell
Games; [email protected]; Jesse Schell, Schell Games; [email protected]; Kim
Jacobs, UPMC Health Plan; [email protected]

Symposium 56C

BEATING THE BLUES: USING CONSUMER USAGE DATA TO DRIVE ENGAGEMENT AND
INCENTIVE STRATEGIES

Amber Blackwood, MPH, CHES

Many patients with anxiety or depression seek mental health treatment in a primary care
setting where medication is often prescribed. The UPMC Health Plan is enhancing their
members experiences by offering digital tools as an alternative or supplement to traditional
S1576 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

treatment. Beating the Blues is a web-based, cognitive behavioral intervention for mild to
moderate depression and anxiety that is made up of 31 interactive modules. While the
program design is evidence-based and theory-driven, completion of all 31 modules remains
low (n=1334; completion=8.8%). Consumer usage data from digital health tools can be
utilized to drive incentive strategies that encourage participation and increase program
completion. We analyzed usage data to determine whether the time elapsed between
completing modules was associated with continued participation in the program. We
identified two high risk points where attrition was greater than 18%, and the number of days
between modules was significantly (p<0.05) different between those who completed the next
consecutive module and those who did not. 19% of individuals who completed module 2 did
not go on to complete module 3, and 18% of individuals who completed module 11 did not go
on to complete module 12. Among those who completed module 3 (n=935), the average time
between completing modules 1 and 2 was significantly (p<0.05) shorter (0.93 days; SD=4.53)
than among those who did not complete module 3 (n=212; 2.59 days (SD=9.55)). Among
those who completed module 12 (n=414), the average time between completing modules 10
and 11 was significantly (p<0.05) longer (1.38 days; SD=3.61) than among those who did not
complete module 12 (n=13; 0.19 days (SD=0.87)). These data suggest that an opportunity
exists for engagement to be optimized through strategically incentivizing the pace at which
individuals progress through web-based mental health treatment. When managing digital
health intervention tools that involve a high level of patient effort and are of high value to the
payer/provider, analytical approaches should be used to guide the design of strategic
incentive structures to optimize engagement and achieve positive health outcomes.

CORRESPONDING AUTHORS: Ellen Beckjord, PhD, MPH, UPMC Health Plan;

[email protected]; Stephen Doyle, MS, MBA, UPMC Health Plan; [email protected]; Kim
Jacobs, UPMC Health Plan; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1577

Symposium 57 1:30 PM-2:45 PM

ADVANCING BEHAVIORAL MEDICINE IN OBESITY RESEARCH ACROSS THE LIFESPAN

Danielle E. Jake-Schoffman, PhD1, E. Amy Janke, PhD2, Claudio R. Nigg, PhD, FSBM3, Andrea T.
Kozak, PhD4, Bernard F. Fuemmeler, PhD, MPH5, Jennifer Huberty, PhD 6

1
University of Massachusetts Medical School, Worcest, MA; 2University of the Sciences,
Philadelphia, PA; 3University of Hawaii, Honolulu, HI; 4Oakland University, Rochester, MI;
5
Massey Cancer Center, Virginia Commonwealth University, Richmond, VA; 6Arizona State
University School of Nutrition and Health Promotion, Phoenix, AZ

The future of behavioral medicine lies in our ability to transform our science into action. Our
field sits at the intersection between communication science, consumer education, industry,
medicine, public health, and technology. The Society of Behavioral Medicine (SBM) has the
unique ability to bring together experts from diverse fields to facilitate a collaborative
environment through which synergy can be created across behavioral medicine disciplines.
Over the past year, members of SBMs Special Interest Groups (SIGs) have prepared
manuscripts highlighting the future directions of behavioral medicine research in their
respective fields of interest, to contribute to a landmark special issue for the Journal of
Behavioral Medicine, a call for action to promote research advancing the science to inform
policy and practice. Among the common themes that emerged across the special issue
manuscripts was the critical issue of advancing obesity prevention and treatment across the
lifespan. This symposium will highlight the future of behavioral medicine research in obesity
prevention and treatment as outlined by 4 SBM SIGs: Obesity and Eating Disorders, Child and
Family Health, Womens Health, and Multiple Health Behavior Change. The Obesity and Eating
Disorders SIG presentation will focus on advances in technology-based solutions for weight
management, particularly for younger and older adults and the accompanying research gaps
in this emerging field. The Child and Family SIG presentation will focus on the prevention of
childhood obesity through lifecourse and systems perspectives, highlighting critical milestones
for intervention and the multilevel nature of effective preventive medicine approaches. The
Womens Health SIG presentation will focus on the need for effective interventions to help
women with weight and stress management during the critical inter-conception period.
Finally, the Multiple Health Behavior Change SIG presentation will focus on areas of future
research to maximize health behavior change for the prevention and treatment of multiple
chronic diseases concurrently, including obesity, cardiovascular disease, and type 2 diabetes.
Together, the symposium will present a range of perspectives on the future of obesity
research that will stimulate an engaging dialogue among the presenters and the audience to
further advance the conversation about the advancement of obesity research.
S1578 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 57A

FUTURE DIRECTIONS OF MULTIPLE BEHAVIOR CHANGE RESEARCH

Dr. Claudio R. Nigg, PhD, FSBM

Non-communicable diseases (i.e., chronic diseases including cardiovascular disease, cancer,

chronic respiratory disease, diabetes and obesity) result in 36 million deaths each year.
Individuals habitual participation in a single health-risk behaviors can seriously contribute to
morbidity and mortality (e.g., tobacco use, daily fast food intake, etc.); however, more
concerning is the typical co-occurrence or clustering of multiple health-risk behaviors. This
burden can be minimized through successful cessation of health-risk behaviors and adoption
of healthy behaviors; namely healthy lifestyle adoption or multiple health behavior change
(MHBC). MHBC is an emerging field and, in order to advance MHBC research, future research
recommendations are provided. A valid measure of MHBC (i.e., lifestyle) is warranted to
provide the needed basis for MHBC investigations and evaluations. MHBC is thought to occur
through shared co-variation of underlying motivating mechanisms, but how these
relationships influence behavior remains unclear. A better understanding of the relationship
between behaviors and the related motivating mechanisms (and potential cross-relationship
of influences) is needed. Future research should also aim to improve lifestyles through
understanding how to change multiple health behaviors. Finally, MHBC research should target
the development of sustainable interventions which result in lasting effects (e.g., capacity,
systems, policy and environmental changes), with dissemination considered during
development. Focusing MHBC research in these areas will increase our understanding and
maximize the impact on the health of populations.

CORRESPONDING AUTHORS: Claudio Nigg, PhD, University of Hawaii; [email protected]; Karly

Geller, PhD, Miami University; [email protected]; Sonia Lippke, PhD, Jacobs University
Bremen gGmbH; [email protected]

Symposium 57B

CURRENT RCT EVIDENCE AND FUTURE DIRECTIONS FOR TECHNOLOGY-BASED WEIGHT

MANAGEMENT AMONG ADULTS

Andrea T. Kozak, PhD

Obesity is a prevalent health care issue associated with disability, premature mortality, and
high costs. Traditional, face-to-face behavioral weight management interventions lead to
clinically-significant weight losses in overweight and obese adults; however, individuals might
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1579

not be able to participate in these programs due to limited access as well as the cost and time
constraints associated with participation. Technological advances such as widespread access
to the Internet, increased use of smartphones, and newer behavioral self-monitoring tools
have resulted in the development of a variety of eHealth weight management programs. This
presentation will provide a review of the current literature, potential solutions to
methodological limitations (e.g., high attrition, minimal participant racial/ethnic diversity,
heterogeneity of technology delivery modes), and future directions for the field, including
important dissemination and policy implications. Younger and older generations comfort
with technological intervention modalities as well as the expectations of digital natives versus
novices will be highlighted.

CORRESPONDING AUTHORS: Joanna Buscemi, PhD, DePaul Univeristy; [email protected];

Misty Hawkins, PhD, Oklahoma State University; [email protected]; Monica Wang,
ScD, Boston University; [email protected]; Jessica Breland, PhD, VA Palo Alto Health Care
System and Stanford University School of Medicine; [email protected]; Kathryn Ross,
PhD, MPH, University of Florida; [email protected]; Anupama Kommu, PsyD, Mind Body
Connection, Inc.; [email protected]

Symposium 57C

CHILDHOOD OBESITY IN THE ERA OF PREVENTION: NEW OPPORTUNITIES AND CHALLENGES

Bernard F. Fuemmeler, PhD, MPH

Since passage of the Patient Protection and Affordable Care Act (ACA) in 2010, there is
revitalized interest in prevention and population health management. This increased focus on
prevention requires new ways of thinking about child and family health in research, practice,
and policy contexts. Given childhood obesity increases risk for chronic health conditions into
adulthood, it will become necessary for prevention models to address childhood obesity
research from a broader perspective that embrace both a life course perspective and systems
or multilevel thinking.

There are a number of touch-points that offer opportunities for implementing prevention
strategies throughout the life course and across social systems that could be leveraged to
affect changes in population obesity prevalence. For instance, the first 1,000 days from
conception to 24 months after birth represents a critical window to address risk factors
associated with childhood obesity, including maternal weight, tobacco use, and lack of
breastfeeding. In addition, as a child matures from early childhood to adolescence there is a
potential to shape weight-related trajectories by intervening on social-contextual factors, such
as the family, school and community contexts.
S1580 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Child and family health researchers will need to advance models and methods that
incorporate a life course and systems perspective. Effective prevention efforts will require a
better understanding of how early life factors shape future health trajectories and how
broader family, school and community systems can be strategically leveraged to influence
these trajectories in such a way as to maximize child health and minimize later adult chronic
health conditions. This presentation will provide a brief overview of research on risk factors
for childhood obesity in the first 1,000 days through early childhood with a focus on findings
addressing a life course and systems perspective.

CORRESPONDING AUTHORS: Pam Behrman, PhD, College of Mt St Vincent;

[email protected]; Maija Taylor, Bowling Green State University;
[email protected]; Rebeccah Sokol, University of North Carolina; [email protected];
Emily Rothman, PhD, Boston University; [email protected]; Lisette Jacobson, PhD, University
of Kansas; [email protected]; Danielle Wischenka, Yeshiva University;
[email protected]; Kenneth Tercyak, PhD, Georgetown University
Medical Center; [email protected]

Symposium 57D

EXPLORING THE NEED FOR INTERVENTIONS TO MANAGE WEIGHT AND STRESS DURING INTER-
CONCEPTION

Dr. Jennifer Huberty, PhD

More than half of women in the U.S. enter pregnancy overweight or obese. More than 2/3 of
these women exceed Institute of Medicine recommendations for healthy weight gain during
pregnancy leading to adverse maternal, infant, and child health outcomes. Managing weight is
a critical component of health during pregnancy. Strategies that encourage weight
management, with emphasis on those that include ways to manage stress during the inter-
conception period (i.e., time immediately following childbirth to subsequent pregnancy) are
needed to promote optimal maternal and infant health outcomes.

The purpose of this presentation is to 1) summarize the current state of knowledge, 2)

critically evaluate the research focused on weight and stress management during the inter-
conception period, and 3) provide future recommendations for research in this area.

CORRESPONDING AUTHORS: Jenn Leiferman, PhD , Colorado School of Public Health;

Leiferman, Jenn ; Abbey R Kruper, Psy.D. , Department of Obstetrics & Gynecology Medical
College of Wisconsin; Kruper, Abbey ; Lisette Jacobson, PhD, MPA, KU Med School of Wichita;
Lisette Jacobson ; Danielle Wishenka, MA, Yeshiva University Verkauf; Danielle Wischenka ;
Jeni Matthews, MS, ASU; Jeni Matthews (Student) ; Molly Waring, PhD, University of
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1581

Massachusetts Medical School; Waring, Molly ; Betty Braxter, PhD, University of Pittsburg;
betty braxter ; Sara Kornfield, PhD, Washington University School of Medicine; Kornfield, Sara
S1582 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 58 1:30 PM-2:45 PM

TRANSLATING LIFESTYLE INTERVENTIONS FOR REDUCING CARDIOMETABOLIC RISK IN

VULNERABLE AND DISADVANTAGED POPULATIONS.

Delwyn Catley, Ph.D.1, Thandi Rose. Puoane, Dr PH.2, Brian Oldenburg, BSc, MPsychol, PhD3,
Christie Befort, PhD4, Ken Resnicow, Ph.D.5

1
Children's Mercy Hospital & University of Missouri-Kansas City, Kansas City, MO; 2University
of the Western Cape, Bellville, Western Cape, South Africa; 3The University of Melbourne,
Melbourne, Victoria, Australia; 4University of Kansas Medical Center, Kansas City, KS;
5
University of Michigan, Ann Arbor, MI

While rigorous RCTs often demonstrate the efficacy of behavioral medicine interventions,
their translation into practice and public health impact remains limited. There are increasing
efforts to bridge this gap as well as an interest in understanding methods for successfully
translating effective interventions into real-world settings. One important example is a
lifestyle intervention approach to modify behavioral risk factors for cardiovascular disease and
Type 2 diabetes. Although large trials in the US and Finland have provided strong evidence for
the efficacy of lifestyle interventions, the challenge is to translate these interventions in such
a way as to have high public health impact across a wide variety of real-world contexts. In this
symposium we will describe trials being undertaken in three countries to translate effective
lifestyle interventions to real-world practice. One presenter will describe the ongoing process
of adapting a proven intervention for delivery by Community Health Workers in a poor urban
township in South Africa in preparation for a large randomized effectiveness trial. Another
presenter will describe the process they used for adapting and scaling up a peer-led, group
based intervention in rural India for a large effectiveness trial. The third presenter will
describe an on ongoing comparative effectiveness trial of alternative methods of delivering a
lifestyle intervention in rural primary care practices in the U.S. Each presenter will describe
the rationale for adapting the intervention for the targeted setting, the conceptual or
theoretical approach to adaptation, the methods and processes of adaptation, the approach
to measurement of the adaptation process, and the results. The discussant will integrate
these presentations by highlighting commonalities in the approach to adaptation as well as
key differences and lessons learned.

Symposium 58A
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1583

ADAPTING THE U.S. NATIONAL DIABETES PREVENTION PROGRAM FOR THE DEVELOPING
WORLD

Dr. Thandi Rose. Puoane, Dr PH.

Increasing rates of cardiovascular disease and diabetes in developing countries are presenting
significant challenges for the health care systems of countries with scarce resources. The
National Diabetes Prevention Program (DPP) is a proven lifestyle intervention for reducing
cardiometabolic risk that could be a powerful prevention tool, however the lack of health
professionals to deliver the program, the high literacy and numeracy levels required of
participants, and cultural differences in lifestyle (e.g., food preferences) need to be addressed.
Informed by the RE-AIM framework and using a Community Engaged approach we are
translating the DPP for delivery to overweight/obese community members in two
impoverished urban townships in South Africa. The adapted program will be evaluated in a
large cluster-randomized trial. Three major adaptations are: 1) to deliver the program using
video and Community Health Workers (CHWs), 2) to capitalize on the widespread cell phone
use to add interactive text-messaging support, and 3) to enhance CHWs communication skills
through simplified Motivational Interviewing training. These adaptations together with lower
literacy requirements, and cultural and language adaptations are expected to enhance
program adoption, reach, and effectiveness. Community engagement is consistent with the
implementation principle and is achieved through the recruitment of study team members
drawn from the community and the formation of Community Advisory Boards that guide the
project throughout. Support and engagement from the community is enhanced by years of
preliminary work in the community. The research team drafts session scripts and expert
consultants and Community Advisory Board members provide feedback to ensure the
intervention will be adoptable and maintained. Implementation evaluation measures will
include quantitative and qualitative measures of implementation feasibility and fidelity,
perceived strengths and weaknesses of the program from the perspective of participants,
CHWs and the NGO. Results of pilot sessions conducted by CHWs indicated 100% completion
of fidelity requirements with rating of 4 out of 5 on overall quality). Session attendance and
completion of homework was acceptable and 100% gave the highest satisfaction rating
possible. Upon completion of the adaptation process the cluster randomize efficacy trial will
begin and, if successful, dissemination will begin.

CORRESPONDING AUTHORS: Thandi Puoane, DrPH, University of the Western Cape;

[email protected]; Kathleen Goggin, Ph.D., Children's Mercy Hospital; [email protected];
Lungiswa Tsolekile, M.P.H., University of the Western Cape; [email protected]; Resnicow
Ken, PhD, University of Michigan; [email protected]; Rafia Rasu, PhD, University of Kansas;
[email protected]; Delwyn Catley, Ph.D., Children's Mercy Hospital; [email protected];
Kandace Fleming, PhD, University of Kansas; [email protected]

Symposium 58B
S1584 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

EVIDENCE-BASED DIABETES PREVENTION IN INDIA: THE KERALA DIABETES PREVENTION

PROGRAM

Professor Brian Oldenburg, BSc, MPsychol, PhD

A number of real world diabetes prevention programs have been well implemented and
evaluated in high income countries; however, efforts to adapt to low and middle-income
countries have been lacking. A strong, community-based model is required in India because
the majority of people still live in rural areas and accessibility to health services is quite
limited. The Kerala Diabetes Prevention Program (K-DPP) has been adapted to rural Kerala,
India from Finlands GOAL Lifestyle Implementation Trial. This has occurred in four steps over
the last 5 years: 1) Needs assessment and participatory design; 2) Program translation and
modelling; 3) Piloting of program delivery model; 4) Program refinement, active
implementation and controlled evaluation. We used Intervention Mapping to identify
program objectives, key determinants of target behaviors, learning and change objectives,
behavior change techniques and practical strategies for implementation. The peer-led
intervention program was also guided by learnings from US Global Peers for Progress
program. Needs assessment identified environmental and personal determinants of
unhealthy lifestyle changes. Barriers included limited state-wide policies and programs for
diabetes prevention, poor accessibility to and availability of health services and low self-
efficacy regarding lifestyle change. Program piloting also identified the need to improve
knowledge and awareness of diabetes prevention, increase male participation and simplify
intervention delivery. The resulting program, K-DPP, includes four components: 1) a peer-led
program conducted in community-based groups; 2) a training and support program for lay
peer leaders; 3) resource materials; 4) specific strategies for community engagement. A
cluster RCT has been undertaken in 60 rural communities in Kerala and 24-month findings
show a reduction in key cardiometabolic risk factors in intervention communities as well as a
36% reduction in diabetes incidence in the impaired glucose tolerance subgroup, compared to
control communities. Scalability is now being evaluated in a further 15,000 communities in
India.

CORRESPONDING AUTHORS: Emma Thomas, The University of Melbourne;

[email protected]; Thirunavukkarasu Sathish, The University of Melbourne;
[email protected]; Pilvikki Absetz, University of Tampere;
[email protected]; Kavumpurathu Thankappan, Sree Chitra Tirunal Institute for
Medical Sciences and Technology; [email protected]

Symposium 58C
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1585

IMPLEMENTING BEHAVIORAL OBESITY TREATMENT IN RURAL PRIMARY CARE: THE RE-POWER

COMPARATIVE EFFECTIVENESS TRIAL

Christie Befort, PhD

Rural residents have higher obesity rates and less access to evidence-based lifestyle
interventions. Primary care has the potential to fill a major gap in the provision of weight
management services for rural communities. The objective of this cluster-randomized
pragmatic trial is to evaluate the comparative effectiveness of three obesity treatment models
in rural primary care. The three models are operationalized as Fee-for-Service (FFS; one-on-
one office visits delivered by local clinic staff), Patient-Centered Medical Home (PCMH; group
visits delivered by local clinic staff), and Disease Management (DM: group phone visits
delivered by centralized interventionists). 36 practices from the Midwestern U.S. were
randomized to deliver one of the three interventions to 40 patients (n=1440) age 20 to 75
with a BMI 30-45 kg/m2. The trial design was guided by patient and provider input since its
inception. The Patient Advisory Board includes 10 men and women living in rural communities
across 4 states. Provider stakeholders are rural PCPs with a vested interest in improving the
treatment of obesity in their practices. To enhance engagement, a day-long kick off meeting
was held where approximately 35 patient and provider stakeholders and investigators shared
personal and practice-level experiences, discussed intervention design across the three arms,
and brainstormed potential barriers and solutions to both patient and practice-level
engagement. Resulting adaptations have included reducing the level of training for group
counselors in the PCMH arm to represent what a practice might typically cover for a chronic
disease care coordinator, and providing an intervention toolkit (previously unplanned but
deemed necessary) for sites randomized to the FFS approach. To reach the targeted 36 sites,
a total of 75 practices were approached. Participating practices includes 10 Rural Health
Clinics (9 hospital-owned, 1 physician-owned), 11 Federally Qualified Health Centers (8
healthcare system-owned, 3 non-profit), 2 VA clinics, and an additional 6 hospital-owned
practices and 7 physician-owned practices. Only 14 practices have co-located behavioral
health specialists. An ongoing mixed methods process evaluation is examining characteristics
influencing practice-level implementation and adoption. In sum, practice-based trials in
geographically dispersed regions require working with heterogeneous practices with limited
access to behavioral health and where local adaptations are necessary.

CORRESPONDING AUTHORS: Jeffrey VanWormer, PhD, Marshfield Clinic Research Foundation;

[email protected]; Cyrus DeSounza, MD, University of Nebraska
Medical Center; [email protected]; Kim Kimminau, PhD, University of Kansas Medical
Center; [email protected]; Allen Greiner, MD, University of Kansas Medical Center;
[email protected]; Edward Ellerbeck, MD, University of Kansas Medical Center;
[email protected]
S1586 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 59 1:30 PM-2:45 PM

USING EHEALTH AND MHEALTH METHODS TO PROMOTE HEALTHY BEHAVIORS AMONG

WOMEN

Lisa Cadmus-Bertram, Ph.D.1, Emily L. Mailey, Ph.D.2, Danielle Symons Downs, Ph.D.3, Danielle
Arigo, Ph.D., Licensed Psychologist4, Katarzyna Wac, PhD5

1
University of Wisconsin - Madison, Madison, WI; 2Kansas State University, Manhattan, KS;
3
The Pennsylvania State University, University Park, PA; 4The University of Scranton, Scranton,
PA; 5Stanford University, Stanford, CA

Electronic and mobile health tools have great potential to improve the effectiveness of health
promotion programs for women, as they can increase the reach of behavioral interventions
and respond to users needs in real time. Electronic health (eHealth) tools such as web-based
intervention programs allow women to access audiovisual content at their convenience,
which can address problems such as difficulty scheduling appointments (e.g., around childcare
needs). Similarly, mobile health (mHealth) tools such as wearable activity sensors and
smartphone applications are carried with the user, and can passively collect health data to
address womens individual health behaviors (e.g., reminders to increase physical activity or
log food intake). Yet currently, these tools are not optimally engaging nor well matched to
womens changing health needs across the lifespan; as a result, commercially available tools
alone rarely foster sustained behavior change. The goal of this symposium is to describe the
design and preliminary outcomes of three technology-supported interventions tailored to
womens health behavior change needs. Each speaker will address the theoretical basis for
integrating and tailoring eHealth and mHealth tools for women; these tools were used to
augment traditional behavioral intervention to actively engage the specific populations of
interest. First, Dr. Mailey will describe a web-based health promotion program for military
spouses, which incorporated tailored podcasts and group dynamics features. Second, Dr.
Downs will discuss an individually-tailored intervention to prevent excessive gestational
weight gain among overweigtht/obese pregnant women, which incorporated Jawbone activity
sensors and Aria wireless scales. Third, Dr. Arigo will present on the feasibility and preliminary
effectiveness of a physical activity promotion program for obese post-menopausal women,
which incorporated the self-monitoring and online social networking features of Fitbit. The
discussant, Dr. Katarzyna Wac, will synthesize lessons learned from these studies and discuss
necessary next steps for improving the use of eHealth/mHealth tools for health promotion
among women.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1587

Symposium 59A

WEB-BASED HEALTH PROMOTION FOR MILITARY SPOUSES: INCREASING ENGAGEMENT AND

EFFECTIVENESS

Dr. Emily L. Mailey, Ph.D.

Military spouses report numerous barriers to health behaviors and could benefit from
interventions that empower them to prioritize their own health and wellness. Web-based
interventions would allow content to be highly accessible for this population, but are often
plagued by low engagement and retention. The purpose of this study was to compare the
effects of a comprehensive, theory-based web-delivered intervention (INT) and a standard
educational web-delivered intervention (CONT) on physical activity, stress, depression, and
anxiety among military spouses. CONT (n=46) provided links to health-related resources on
the militarys Operation Live Well website and a weekly discussion board. INT (n=73) was
based on Self-Determination Theory and utilized original podcasts to deliver weekly content
related to stress management, physical activity, and nutrition. In addition, all INT participants
were placed on teams with 4-5 other spouses and engaged in weekly team challenges to
facilitate social support. Both interventions lasted for 10 weeks. INT participants logged onto
the website more often [median=12] than CONT participants [median=3.5]. Mixed design 2
(group) X 2 (time) repeated measures ANOVAs revealed significant main effects for stress
[F(1,76)=35.0, pF(1,79)=30.2, pF(1,78)=39.0, pF(1,81)=6.61, p=.01] and moderate activity
[F(1,81)=5.77, p=.02], but not vigorous activity [F(1,81)=2.27, p=.14], on the International
Physical Activity Questionnaire. There were no significant between-group differences. These
results suggest web-delivered interventions can improve physical activity and mental health
among military spouses; however, additional work is needed to optimize effectiveness and
engagement. In addition to discussing the primary results, this presentation will describe the
formative research that informed the development of the theory-based intervention, and will
highlight qualitative and quantitative follow-up data from participants that offer valuable
insight about how to make web-based interventions maximally engaging and beneficial.

CORRESPONDING AUTHORS: Brandon Irwin, PhD, Kansas State University; [email protected];

David Warnky, Kansas State University; [email protected]; Chelsey Glatz, BS, Kansas State
University; [email protected]

Symposium 59B

MHEALTH SELF-MONITORING OF PHYSICAL ACTIVITY AND WEIGHT WITHIN A PRENATAL

INTERVENTION TO MANAGE GESTATIONAL WEIGHT GAIN

Danielle Symons Downs, Ph.D.

S1588 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Physical inactivity is prevalent in pregnant women and it can increase the risk of excessive
gestational weight gain (GWG), which is associated with pregnancy and delivery
complications. Innovative strategies to promote perinatal physical activity and reduce
excessive GWG are strongly needed. Novel mHealth tools such as wrist-worn activity
monitors, Wi-Fi weight scales, and smartphone applications can assist with self-monitoring to
promote increased physical activity and weight management. However, little to no past
research has utilized these tools to promote prenatal activity and manage GWG. We have
developed an individually-tailored behavioral intervention that adapts intervention dosages to
the unique needs of overweight and obese pregnant women and uses mHealth methods to
promote physical activity and manage GWG. The goals of this presentation are to: (a) describe
the intervention and mHealth methods for using the Jawbone activity monitor and Aria Wi-Fi
weight scale for self-monitoring of physical activity and weight and (b) discuss preliminary
findings and highlight key lessons learned from the initial trial on the use of the mHealth
devices within the prenatal intervention. Overweight/obese pregnant women (N=22) were
randomized to one of seven intervention dosages over a brief 6-week intervention period.
Daily activity steps/minutes and weight were analyzed with SPSS and MATLAB; visualizations
were used to provide real-time feedback to participants. The use of mHealth data to develop
participant-specific energy balance models and reporting energy intake underreporting will be
described. Average participant compliance rates were 95% (Jawbone) and 84% (Aria). Key
lessons learned to be discussed will include designing the mHealth device self-monitoring
procedures to provide real-time feedback to participants, overcoming technology issues, and
user acceptability of the devices.

CORRESPONDING AUTHORS: Jennifer Savage, Ph.D., Penn State Univ; [email protected]; Daniel
E. Rivera, Ph.D., Arizona State Univ; [email protected]; Krista S. Leonard, Penn State Univ;
[email protected]; Abigal M. Pauley, MS, Penn State Univ; [email protected]; Emily H Hohman,
Ph.D., Penn State Univ; [email protected]; Penghong Guo, Arizona State Univ;
[email protected]

Symposium 59C

FEASIBILITY OF A FITBIT-SUPPORTED BEHAVIORAL INTERVENTION TO INCREASE PHYSICAL

ACTIVITY AMONG OBESE POST-MENOPAUSAL WOMEN

Dr. Danielle Arigo, Ph.D., Licensed Psychologist

Exercise is strongly recommended for obese post-menopausal women in order to mitigate

cardiovascular risk, but less than 15% currently meet exercise guidelines. These women cite
lack of both time and social support for exercise as primary barriers, indicating two critical
targets of intervention. To address these targets, this study tested the feasibility of a
technology-supported, social intervention to promote exercise among obese post-
menopausal women. Participants (n=11, MAGE=59, MBMI=32.0 kg/m2) increased either
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1589

traditional endurance exercise (EE, e.g., brisk walking; n=5) or short-duration interval exercise
(SDIE, guided by DVDs; n=6). EE participants progressed to 250 exercise minutes per week,
whereas SDIE participants progressed to 75 minutes per week, over 16 weeks. All participants
attended one group treatment session per month. They tracked their exercise using the Fitbit
system, and communicated with their treatment group between sessions using the Fitbit
social network. This network included a group message board and a ranking of participants
exercise minutes (highest to lowest). Using a short recruitment window, enrollment
approached the intended 12 participants (11/12, 92%), and retention over 16 weeks was
10/11 (91%); 8/11 participants (73%) attended all four group sessions. Among completers,
daily device wear was 92% across groups. The average participant also responded to 11/16
weekly Fitbit social network threads (70%) and viewed the exercise rankings twice per week.
Regarding exercise adherence, 3/5 EE participants met prescribed exercise minutes at both
mid- and end-of-treatment, whereas 5/5 SDIE participants (who completed) met prescribed
exercise at both time points. Findings demonstrate that introducing the novel components of
this intervention (i.e., Fitbit self-monitoring and social networking, SDIE) to behavioral
treatment is feasible and acceptable to obese post-menopausal women, and that adherence
to SDIE might be greater than to EE. Consequently, this intervention may effectively address
both lack of time and lack of social support for exercise as perceived barriers to engagement
in the target population, and warrants investigation on a larger scale.

CORRESPONDING AUTHORS: Joan Grossman, Ph.D., R.D., The University of Scranton;

[email protected]; Jessica Bachman, Ph.D., R.D., The University of Scranton;
[email protected]
S1590 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 60 1:30 PM-2:45 PM

USING DIGITAL SOCIAL TOOLS FOR SMOKING CESSATION: INTERVENTION ENGAGEMENT AND
EFFICACY

Judith J. Prochaska, PhD1, Lorien Abroms, ScD, MA2, Danielle Ramo, PhD3, Cornelia Pechmann,
MS, MBA, PhD4

1
Stanford University, Stanford, CA; 2George Washington University, Washington, DC;
3
University of California, San Francisco, San Francisco, CA; 4University of California Irvine,
Irvine, CA

Tobacco use is the leading preventable cause of death in the US. Novel interventions are
needed for prevention and intervention to reduce smoking among high risk groups. With wide
reach and integration into the lives of users, digital tools (e.g., text messaging, social media)
hold promise to help smokers quit. Yet scientific evidence for the utility of some digital tools is
limited and inconclusive. This symposium will present the results of three randomized trials
testing the efficacy of digital interventions for smoking cessation. Presentations will focus on
strategies used to adapt interventions to a digital format, characterize engagement, and
report efficacy in helping smokers quit. The first presentation is a text-messaging intervention
for pregnant smokers (Quit4baby). Results provide support for the interventions efficacy after
3 months of enrollment. The second presentation is a Facebook-based intervention for young
adults tailored to readiness to quit smoking (The Tobacco Status Project) reporting 6-month
outcomes. Results demonstrate efficacy with respect to biochemically-verified smoking
abstinence at 3 months, reduction in smoking at 6 months, and the utility of using Facebook
to engage the young adult population and tailoring to readiness to quit. The final talk is a trial
testing the efficacy of a Twitter intervention for relapse prevention. Results demonstrate
efficacy at 6 months among adult smokers, and highlight moderators of engagement and quit
smoking status. The discussant will highlight lessons learned in carrying out trials with digital
tools, unique strategies for measuring engagement in a digital environment, and the impact
digital social tools can have on delivering engaging evidence-based smoking cessation
interventions. The symposium will include dedicated time for a question-and-answer session
with the panel of presenters.

Symposium 60A

A RANDOMIZED TRIAL OF A TEXT-MESSAGING PROGRAM TO PROMOTE SMOKING CESSATION

IN PREGNANT SMOKERS
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1591

Dr. Lorien Abroms, ScD, MA

While automated text messaging programs have been found to be successful for adult
smoking cessation, their efficacy for pregnant smokers is unknown. The Quit4baby study
explores whether an interactive and intensive text messaging program can promote smoking
cessation in pregnant smokers. Women were recruited from Text4baby, an existing text
messaging program aimed at pregnant women. From July 2015 to February 1016, over 35,000
Text4baby subscribers were sent a recruitment text message. Interested and eligible women
were enrolled (n= 497) and randomized to receive Text4Baby + Quit4Baby (intervention) or
Text4Baby (control). Participants were surveyed at 1-month, 3-month and 6-months post-
enrollment, and saliva samples were collected at 3-months for biochemical verification of
smoking status. Using an Intention-to-Treat analysis, where missing are assumed to be
smoking, 28.80% of the intervention group and 15.79% of control group reported not smoking
in the past 7 days at 1-month (p<.01) and 35.20% of the intervention group and 22.67% of the
control group reported not smoking in the past 7 days at 3-month (p<.01). Biochemical
verification of smoking status at 3-months indicated no significant differences among groups,
though quit rates favored the intervention group. Results will also be presented for the 6-
month follow-up. Current results provide mixed support for the efficacy of the Quit4baby
program. Future study is warranted.

CORRESPONDING AUTHORS: Lorien Abroms, ScD, GWU; [email protected]; Leah Leavitt, MPH,
GWU; [email protected]; Pamela Johnson, PHD, Voxiva; [email protected];
Sean Cleary, PhD, GWU; [email protected]; Thomas Brandon, PhD, Moffitt Cancer Center;
[email protected]; Jennifer Schindler-Ruwisch, MPH, GWU;
[email protected]; Jessica Bushar, MPH, Zero to Three; [email protected]

Symposium 60B

THE TOBACCO STATUS PROJECT: OUTCOMES FROM A RANDOMIZED TRIAL OF A FACEBOOK

SMOKING CESSATION INTERVENTION FOR YOUNG ADULTS

Dr. Danielle Ramo, PhD

Introduction: Social media represents a promising strategy to deliver evidence-based smoking

cessation treatment to young adults. Our group has developed the Tobacco Status Project
(TSP) smoking cessation intervention for young adults, tailored to readiness to quit smoking.
Previously we demonstrated feasibility and promising 12-month quit rates. Here we present
outcomes from a randomized trial comparing the TSP to an online control condition. Methods:
Young adult cigarette smokers (N=500; age 18-25) were recruited online and randomized to
either the 3-month TSP intervention or a referral to the National Cancer Institute
Smokefree.gov website, including Internet, Smartphone, and social media features. TSP
S1592 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

included assignment to a private Facebook group tailored to Transtheoretical Model readiness

to quit smoking (precontemplation, contemplation, preparation), daily Facebook contacts,
weekly live counseling sessions, and for those ready to quit, six Cognitive Behavioral Therapy
counseling sessions. Results: The sample was 73% White, 55% female, 87% daily smokers; 48%
smoked 10 or fewer cigarettes per day, and averaged 2.8 years smoking; 30% were in
precontemplation; 49% contemplation, and 21% preparation for quitting smoking.
Posttreatment abstinence was modeled via logistic GEE. At treatment end (3 months), the
follow-up rate was 70% (67% treatment, 74% control; 2=.124, p=.14). In a modified intent-to-
treat analysis, biochemically-verified 7-day point prevalence abstinence was significantly
higher for TSP than control (8.3% vs 3.3%; odds ratio [OR]=2.52; 95% confidence interval
[CI]=1.56, 4.04; p.0001). A greater proportion of TSP participants reduced the number of
cigarettes they smoked in the past week by at least half from baseline to 3 months (52.7% vs.
38.8%; OR=1.82; 95% CI=1.33, 2.49; p=.0002). There were no differences in likelihood of
having made a quit attempt during treatment or readiness to quit smoking in the next month
across groups. 6-month follow-up rate was 68% (64% treatment, 73% control; 2=.124, P=.14).
Only reduction in smoking was still significantly different at 6 months (58.2% vs. 51.7%;
OR=1.37; 95% CI=1.04, 1.83; p=.026). TSP participants who commented at least once (75%)
averaged 42 comments, baseline stage of change and monetary incentive predicted comment
volume, and comment volume marginally predicted abstinence at 3 months. Conclusion: A
novel Facebook intervention was engaging and associated with smoking abstinence at 3
months and smoking reduction at 6 months. Social media intervention could be disseminated
widely, and be helpful even for those not ready to quit smoking.

CORRESPONDING AUTHORS: Johannes Thrul, PhD, University of California, San Francisco;

[email protected]; Kevin Delucchi, PhD, University of California, San Francisco;
[email protected]; Sharon Hall, PhD, University of California, San Francisco;
[email protected]; Pamela Ling, MD, MPH, University of California, San Francisco;
[email protected]; Alina Belohlavek, BA, University of California, San Francisco;
[email protected]; Shirley Zhao, BA, University of California, San Francisco;
[email protected]; Judith Prochaska, PhD, MPH, Stanford University; [email protected]

Symposium 60C

PREDICTORS AND OUTCOMES OF DYADIC TIES IN ONLINE SMOKING CESSATION GROUPS

Cornelia Pechmann, MS, MBA, PhD

Tweet2Quit is a smoking-cessation intervention that involves creating 20-person groups of

smokers on Twitter who agree to quit together and support each others cessation for 90
days, aided by daily automated discussion-topics tweets and individualized texts on
engagement. An initial randomized controlled trial found 40% sustained abstinence for
Tweet2Quit vs. 20% for control. Also strong dyadic ties related to engagement and
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1593

engagement to abstinence (ps < .01). This study examined what facilitated dyadic ties and we
studied 8 online groups. A dyad was identified when a group member sent a tweet to a
specific individual, and 630 dyads were identified. Dyadic tie strength was measured as the
number of tweets exchanged between two members who tweeted. Engagement of the dyad
with the group was measured as the total count of tweets sent by dyad members to
themselves or others. Dyadic goal attainment assessed whether one or both dyad members
attained abstinence. Based on past research, people perceive that self-disclosing
dissimilarities may result in rejection, yet there are potential benefits to self-disclosure in
terms of rapport building. We tested a mediational model that predicted that dyadic
dissimilarity would relate negatively to dyadic self-disclosure, but dyadic self-disclosure would
relate positively to dyadic tie strength, engagement of the dyad with the group, and dyadic
goal attainment (abstinence). The results supported the mediation model for gender = -.02,
employment status = -.02, and marital status = -.01 (ps < .05), but not age = -.0003 which
was rarely self-disclosed. When one member of a dyad self-disclosed a demographic
dissimilarity, this discouraged the other dyad member from self-disclosing that same
demographic, but self-disclosing dyads benefitted greatly in terms of tie strength,
engagement and abstinence. Our findings indicate that making online groups demographically
homogeneous may encourage self-disclosure but this may run counter to diversity goals.
Alternatively one can create dyads (buddy pairs) that are demographically similar and
encourage self-disclosure. In our larger randomized controlled trial of Tweet2Quit, we will test
both women-only and coed groups, assign people to demographically similar dyads, and
encourage self-disclosure.

CORRESPONDING AUTHORS: Kelly Yoon, UG, University of California, Irvine;

[email protected]; Judith Prochaska, PhD, MPH, Stanford Prevention Research Center;
[email protected]
S1594 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 61 1:30 PM-2:45 PM

EXPANDING THE HORIZONS OF BEHAVIOR CHANGE RESEARCH WITH AN EXPERIMENTAL

MEDICINE APPROACH: THE NIH SOBC RESEARCH NETWORK

Jennifer A. Sumner, PhD1, Christine M. Hunter, PhD2, Donald Edmondson, PhD, MPH3, Jun Ma,
MD, PhD4, Susan M. Czajkowski, Ph.D.5

1
Columbia University Medical Center, Center for Behavioral Cardiovascular Health, New York,
NY; 2Division of Diabetes, Endocrinology, and Metabolic Diseases, Bethesda, MD; 3Columbia
University Medical Center, New York, NY; 4University of Illinois at Chicago, Chicago, IL;
5
National Cancer Institute, Rockville, MD

Unhealthy behaviorsincluding smoking, overeating, medication nonadherence, and

sedentary lifestylescontribute to negative health outcomes and disease. Unfortunately, it is
extremely difficult to initiate and maintain lasting healthy behavior changes. A major
challenge facing the field in addressing these often intransigent behaviors is insularity of
scientific domains that together might inform more effective prevention and intervention
efforts. In response to these challenges, the NIH Common Funds Science of Behavior Change
(SOBC) program was formed to bring together researchers across disciplines to develop a
more unified science of behavior change. The program represents a fundamental shift in the
approach to behavior change research and aims to improve our understanding of mechanisms
of behavior change by supporting cross-cutting research that integrates basic and
translational science. In 2015, the SOBC Research Network was formed and brought
together several interdisciplinary research teams to employ a mechanisms-focused,
experimental medicine approach to behavior change research. These researchers are
developing assays to measure putative intervention targets related to common mechanisms
of behavior change, including self-regulation, stress reactivity and resilience, and
interpersonal and social processes. The teams will then engage these targets through
experimentation and/or intervention and examine subsequent behavior change. In this
symposium, we will introduce the SOBC Research Network and provide examples of a
mechanisms-focused, experimental medicine approach to behavior change research. First, an
NIH representative from the SOBC program will describe the rationale and goals of the SOBC
Research Network. Next, researchers supported by the SOBC program will present their
approach to examining mechanisms of change common to multiple health behaviors and
demonstrate how the experimental medicine paradigm and collaborative SOBC approach
have been fundamental to their work. The symposium will conclude by having an NIH
representative review the implications of the SOBC research approach for future basic and
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1595

translational research and discuss how basic science findings can be applied to develop and
test innovative prevention and treatment strategies.

Symposium 61A

THE NIH SCIENCE OF BEHAVIOR CHANGE (SOBC) COMMON FUND PROGRAM: RATIONALE AND
GOALS

Dr. Christine M. Hunter, PhD

The Science of Behavior Change (SOBC) Common Fund Program at the National Institutes of
Health is supporting a mechanisms-focused, experimental medicine approach to behavior
change research. An experimental medicine approach involves identifying putative
intervention targets, developing assays (measures) to permit verification of target
engagement, engaging the target through experimentation or intervention, and testing the
degree to which target engagement produces the desired behavior change. Putative
intervention targets are synonymous with "mechanisms of action" and with processes that
are hypothesized to be malleable and to play a causal role in producing behavior change. By
integrating work across basic science disciplines and between basic and clinical scientists, this
effort will lead to an improved understanding of the underlying principles of behavior change
as they apply to the initiation and maintenance of change, and to the personalization of
intervention strategies. The current stage of the program supports the SOBC Research
Network, which focuses on three classes of intervention targets that are highly relevant to
understanding the mechanisms of behavior change: self-regulation, stress reactivity and stress
resilience, and interpersonal and social processes. This introductory presentation will explain
the rationale and steps involved in the experimental medicine approach as implemented in
the SOBC Program and discuss how it is being used to examine mechanisms of change
common to multiple health behaviors. The overarching goal of the SOBC Program is to
transform the approach to behavioral mechanisms research, develop tools that can support
the implementation of this approach in future research, and enhance the development of
future behavioral interventions.

Symposium 61B

SYSTEMATIC REVIEW OF MECHANISMS OF ACTION IN NIH-FUNDED TRIALS TO IMPROVE

MEDICATION ADHERENCE

Dr. Donald Edmondson, PhD, MPH

S1596 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Background: Many patients do not take their medication as prescribed and although many
interventions have been tested to increase adherence, few effective approaches exist. The
NIH Science of Behavior Change (SOBC) program approach offers an experimental medicine
method to optimize behavioral interventions by targeting underlying mechanisms of behavior
change. We conducted a systematic review and meta-analysis to determine the proportion of
NIH-funded adherence trials that measured a mechanism of action, and determined the
strength of association between those mechanisms and nonadherence.

Methods: Multiple electronic databases were searched using all relevant subject headings
and free-text terms to represent medication adherence and behavior change. We applied a
search filter to limit results to studies funded by NIH only. Reference lists and cited reference
searching were also conducted. Unpublished studies were sought through trial registries and
grey literature sources, and by contacting authors of included studies. Eligibility criteria were
the following: population, all adult human participants taking medication for any condition;
intervention, any that aims to change the behavior of participants adherence to medication;
study design, randomized and quasi-randomized studies funded by NIH. Screening, data
extraction, and quality assessment were conducted by two independent reviewers, with
disagreements resolved through discussion. In our overall analysis, we will examine whether
interventions are more effective when mechanisms were measured, and we will also analyze
the associations of different mechanisms/mechanism classes with adherence. This systematic
review and meta-analysis will be complete by February 2017.

Conclusion: We will discuss the proportion of NIH-funded intervention trials that tested
mechanisms of behavior change hypothesized as influential in medication adherence. We will
assess the relative strengths of association between mechanisms and adherence outcomes.
This research will illustrate the current state of the science of behavior change, and highlight
opportunities for behavior change research going forward.

CORRESPONDING AUTHORS: Kevin Sundquist, Center for Behavioral Cardiovascular Health,

Columbia University Medical Center; [email protected]; Laura Meli, Center for
Behavioral Cardiovascular Health, Columbia University Medical Center;
[email protected]; Louise Falzon, Center for Behavioral Cardiovascular Health,
Columbia University Medical Center; [email protected]

Symposium 61C

MECHANISTIC SELF-REGULATION TARGETS IN INTEGRATED BEHAVIOR THERAPY FOR OBESE

AND DEPRESSED ADULTS: RAINBOW-ENGAGE STUDY

Dr. Jun Ma, MD, PhD

ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1597

Obesity and depression are top contributors to the global burden of disease and disability;
when comorbid, the burden is exacerbated. Understanding the processes and mechanisms
contributing to risk will facilitate the development of targeted and personalized approaches to
behavior change. Self-regulation is a core construct in empirically supported behavioral
interventions for these disorders and a candidate target mechanism for behavior change.
Human neuroscience has identified large-scale brain circuits for key self-regulation targets:
emotion regulation, cognitive control, and self-reflection. Using a target-driven, experimental
medicine approach, the ENGAGE study leverages an ongoing randomized controlled trial
(RAINBOW) of integrated behavior therapy for comorbid obesity and depression in primary
care to measure self-regulation targets among trial participants. The study aims are to identify
and refine the measures (assays), validate target engagement and malleability, and test
target-driven optimization of behavior therapy for comorbid obesity and depression. The
chosen assays of emotion, cognitive and self-reflective regulation involve multiple domains
(neural, physiological, behavioral, and psychological) and settings (brain imaging lab, virtual
reality environments, and smartphone-based passive sensing). The assays are included at
baseline, 2, 6, 12 and 24 months. We will examine the relationships between assays within
and between domains and settings, and how assays of self-regulation targets predict health
behaviors and outcomes through 24 months. ENGAGE will provide a toolkit of validated assays
of self-regulation targets important for health behavior change. Findings on how self-
regulation mechanisms predict behavior and outcomes will advance the development of
targeted and personalized intervention strategies for treating obesity and depression, with
implications for other common lifestyle-related comorbidities.

CORRESPONDING AUTHORS: Lisa Goldman Rosas, PhD, MPH, Palo Alto Medical Foundations
Research Institute; [email protected]; Mark Snowden, MD, MPH, University of Washington
School of Medicine; [email protected]; Brian Wandell, PhD, Stanford University;
[email protected]; Jeremy Bailenson, PhD, Stanford University; [email protected];
Walter Greenleaf, PhD, Stanford University; [email protected]; Paul Dagum, MD, PhD,
Mindstrong Inc; [email protected]; Philip Lavori, PhD, Stanford University;
[email protected]; Trisha Suppes, MD, PhD, Stanford University; [email protected];
Megan Lewis, PhD, RTI International; [email protected]; Joshua Smyth, PhD, Pennsylvania State
University; [email protected]; Leanne Williams, PhD, Stanford University;
[email protected]
S1598 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 62 1:30 PM-2:45 PM

THE IMPACT OF ADVERSE CHILDHOOD EXPERIENCES [ACES] ON CHILD AND YOUNG ADULT
HEALTH FUNCTIONING

Pamela Behrman, PhD1, Kimberly A. Randell, MD, MSc2, Sheela Raja, PhD3, Nisha Beharie,
DrPH4, Emily F. Rothman, ScD5

1
College of Mt St Vincent, New Britain, CT; 2Children's Mercy Kansas City, Kansas city, MO;
3
University of Illinois at Chicago, Chicago, IL; 4NYU Rory Meyers College of Nursing, New York,
NY; 5Boston University School of Public Health, Boston, MA

In 1998, a team of researchers led by Drs. Vincent Felitti, of Kaiser Permanente, and Robert
Anda, of the Centers for Disease Control, found that certain potentially traumatic exposures in
childhood were significantly associated with serious adult health outcomes. These traumatic
exposures, called Adverse Childhood Experiences, or ACEs, include adversities such as
childhood physical abuse, childhood sexual abuse, parental interpersonal violence, parental
substance abuse, parental separation or divorce, economic hardship, and parental
incarceration. In addition to adult psychosocial effects (e.g. depression and homelessness),
ACEs were strongly related to adult chronic diseases such as cancer, heart disease, and COPD.
High reported ACEs were also associated with 20-year reductions in life expectancy. The
authors noted dose-response patterns, with higher reported ACEs more strongly related to
serious adult health outcomes. Since the publication of the original ACE studies, there have
been more than 60 investigations published by the original authors and several hundred peer-
reviewed studies using ACEs as a primary variable.

The researchers in this symposium add to this auspicious body of literature. Using data
collected at two Midwest academic hospitals, Dr. Kimberly Randell presents her study
outcomes on the relations between parents reported ACEs and their school-aged (ages 6-12)
children's adversities. Using data collected at a university-administered dental clinic, Dr.
Sheela Raja illustrates the influences of ACEs and previous negative dental care experiences
on childrens (ages 4 and over) behavior during routine dental care appointments. Finally,
using longitudinal data from a nationally-representative sample of adolescents, Dr. Nisha
Beharie showcases her findings on the impacts of childhood trauma and homelessness on
adult drug abuse. Our discussant, Dr. Emily Rothman, concludes this symposium by analyzing
these study outcomes in the context of existing literature. She also recommends future
directions for child-focused ACEs research and prevention efforts.

Symposium 62A
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1599

ASSOCIATIONS BETWEEN PARENT ADVERSE CHILDHOOD EXPERIENCE SCORE AND CHILDREN'S

ADVERSITY IN PEDIATRIC HEALTHCARE SETTINGS

Dr. Kimberly A. Randell, MD, MSc

Purpose: Limited data suggest parent adverse childhood experiences (ACEs) are associated
with increased risk of child adversity. We aimed to determine associations between parent
ACE score and child adversity in pediatric healthcare settings.

Methods: This was a secondary analysis from a study examining parent ACEs and child weight.
Parents of children age 6-12 years attending general pediatric or pediatric weight
management clinics at two Midwest academic hospitals completed a survey assessing parent
and child conventional and expanded ACEs. ACE scores were calculated by summing individual
ACEs.

Results: 339 parents participated. 58 (17%) parents had a conventional ACE score of > 4, 26
(8%) an expanded ACE score > 4, and 89 (26%) a total ACE score > 4. Univariate analysis
showed parent conventional, expanded and total ACE score > 4 was associated with increased
risk for child exposure to household criminal activity, household mental illness, household
substance abuse, family violence, neighborhood violence, homelessness, racism, bullying,
poverty, and child conventional and total ACE score > 4. Logistic regression found parent total
ACE score > 4 is associated with increased risk for child conventional and total ACE score > 4.

Conclusions: Findings support the call for pediatric healthcare delivery within a two-
generation model. Parental ACE score > 4 was common and associated with increased child
risk for multiple adverse experiences as well as a child ACE score > 4, suggesting parent ACE
history may identify children at increased risk for adverse experiences and the attendant
negative health outcomes. Consideration should be given for parental ACE assessment
and provision of targeted resources for families in which parent ACE score is > 4.

Symposium 62B

THE ROLE OF ADVERSE CHILDHOOD EVENTS IN PEDIATRIC DENTAL CARE: IMPLICATIONS FOR
TRAUMA INFORMED DENTAL CARE

Sheela Raja, PhD

Purpose: Adverse childhood experiences such as exposure to violence, serious illness, and
sudden death can impact childrens health and their ability to cope with stressful situations.
S1600 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

We examined the relationship between a history of childhood trauma and a childs behavior
during non-invasive, routine dental treatment.

Methods: Parents of healthy children ages four years and over attending their initial dental
appointment at a university pediatric dental clinic were asked to complete the Traumatic
Events Screening Inventory- Parent Report Revised. Following the dental appointment, a
pediatric dental resident completed the Frankl Scale, a clinician-based rating of a childs
behavior in the dental chair.

Results: One hundred-seventy parent-child pairs participated. Fifty-three percent of parents

indicated their child had experienced at least one traumatic event. Forty-four percent of
parents indicated their child had a prior negative experience at the dentist. Analysis showed
no significant relationship between traumatic events history and poor dental behavior, but a
significant relationship between a previous negative dental experience and poor dental
behavior.

Conclusion: Children with a history of traumatic events did not exhibit uncooperative behavior
more often than those who did not experience it. However, future studies should examine the
relationship between trauma history and behavior during dentally invasive procedures (e.g.,
fillings, extractions). Additionally, our results suggest that in addition to serious medical
procedures, surveys of childhood trauma should also include dentally stressful experiences,
including the potential consequences of immobilization and protective restraints in dental
treatment.

Symposium 62C

THE SYNDEMIC EFFECT OF CHILD ABUSE AND HOMELESSNESS ON SUBSTANCE USE

Dr. Nisha Beharie, DrPH

Previous research has indicated elevated substance use as well as experiences of childhood
trauma (e.g. physical and sexual abuse) among homeless populations relative to their housed
counterparts. However, no research to date has investigated the syndemic effect of
homelessness and history of abuse on substance use. As such, this study investigates the
independent and combined effects of homeless and child abuse on drug use.

Logistic regression was employed using data from Wave I (adolescence), III (emerging
adulthood), and IV (adulthood) of the National Longitudinal Study of Adolescent to Adult
Health (Add Health) (N = 12,288). The independent variable was a four-level categorical
variable: no experience of homelessness or child abuse, experience of only child abuse,
experience of only homelessness, and experience of both child abuse and homelessness. All
outcome variables of substance use were dichotomous (yes/no) and included binge drinking,
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1601

marijuana, cocaine, methamphetamine, prescription pain medication, and injection drug use.

Homelessness was found to have stronger strengths of association than experiences of child
abuse alone when predicting substance use. However, the strengths of association were
greatest among those who had experienced both homelessness and child abuse. For example,
those who had experienced only childhood abuse or only homelessness were at an
approximately two-fold increased odds of ever using cocaine by emerging adulthood
(OR=1.76, 95% CI:1.41, 2.20; and 2.23, 95% CI:1.45, 3.44 respectively). However, those who
had experienced both childhood abuse and homelessness were at an almost six-fold increased
odds of using methamphetamines by emerging adulthood (OR=6.56, 95% CI:3.85, 11.18).

Findings from this study highlight the crucial need to address past trauma as well as housing
instability in order to reduce substance use among those who have experienced
homelessness.

CORRESPONDING AUTHORS: Joy D. Scheidell, MPH, NYU School of Medicine;

[email protected]; Maria Khan, PhD, NYU School of Medicine; [email protected]
S1602 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 63 1:30 PM-2:45 PM

INNOVATION IN INDOOR TANNING MEDIA, MESSAGING, AND INTERVENTION SCIENCE

Darren Mays, PhD, MPH1, Molly E. Waring, PhD2, Derek Reed, PhD3, W. Douglas Evans, PhD4,
Sherry Pagoto, PhD2

1
Georgetown University Medical Center, Lombardi Comprehensive Cancer Center,
Washington, DC; 2University of Massachusetts Medical School, Worcester, MA; 3University of
Kansas, Lawrence, KS; 4George Washington University Milken Institute School of Public
Health, Washington, DC

Skin cancer is the most common malignancy in the US, and indoor tanning in adolescence and
young adulthood significantly increases skin cancer risks, accounting for an estimated 10% of
all cases. Research indicates patterns of indoor tanning behavior are complex, ranging from
pre-event and seasonal tanning to potential tanning addiction. Like other cancer risk
behaviors, indoor tanning is influenced by a suite of risk factors, including exposure to pro-
tanning messaging in the media. Innovative strategies to investigate the indoor tanning media
environment, how young people may respond to pro-tanning messaging, and how messaging
can in turn be designed and delivered to prevent and reduce indoor tanning is needed to
address the growing burden of skin cancer. This symposium will feature three presentations
reporting on highly innovative approaches to indoor tanning media, messaging, and
intervention science. The first study describes a content analysis of indoor tanning tweets,
demonstrating the vast majority are generated by indoor tanners characterizing their motives
and consequences of tanning, and indicating presence of few public health messages in this
media space. The second study uses a novel cue-reactivity paradigm within a simulated indoor
tanning salon to characterize indoor tanning cravings and motivations among young women.
The study findings point to policy approaches to limit indoor tanning marketing as have been
implemented in other behavioral cancer prevention domains (e.g., tobacco control). The final
study pilot tested the feasibility, acceptability, and preliminary effects of a theory-driven,
mobile phone text messaging intervention among young adult women who indoor tan. With
further testing, this novel intervention approach has the potential for scalability and high
public health impact. Our discussant will provide a critical synthesis of these innovative
studies and lead a discussion of the implications for future research on indoor tanning
behavior and approaches to prevention.

Symposium 63A
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1603

WHO IS TWEETING ABOUT INDOOR TANNING AND WHAT ARE THEY SAYING?

Dr. Molly E. Waring, PhD

Conversations about healthy and unhealthy behaviors have been observed on Twitter and
may provide insight into attitudes and motivations for these behaviors. We content analyzed
tweets that included the terms tanning bed or tanning salon to describe the content of
conversations about indoor tanning on Twitter and to determine what proportion of tweets
were posted by tanners. We extracted 4,691 tweets using the terms tanning bed or
tanning salon over a 7-day period (17-23 March 2016). We randomly selected 1,000 tweets
stratified by search term and timing of tweet (weekday or weekend). After replacing ads
(n=200), we conducted a content analysis of tweet content. The research team developed a
codebook and two coders coded all tweets. Twenty percent of tweets were double-coded
(=0.74, 81% agreement) and consensus on discrepantly-coded tweets was reached through
discussion. Among tweets expressing a desire to tan, we analyzed the proportion of tweets in
which the tanner mentioned mood or appearance as reasons they tan (=0.89, 98%
agreement for mood, =0.92, 99% agreement for appearance). Tweets (n=1,000) were posted
by 978 unique accounts. The majority of tweets (71%) were by tanners (n=699 individuals).
Themes included: tanner expressing desire or eagerness to tan (41%), tanner complaining
about a negative experience at the tanning salon (12%), tanner reporting a burn (11%), tweets
mocking tanners or the act of tanning (11%), tanner reporting sleeping in a tanning bed (6%),
tanning salon employee chatter (4%), anti-tanning message (4%), and tweets arguing against
tanning legislation (1%). Eleven percent of tweets were about other topics or unrelated to
tanning. Among the 405 tweets in which a tanner expressed the desire to tan, 10% mentioned
mood and 9% mentioned appearance as the reason they tan. Findings revealed a rate of 670
tweets per day about indoor tanning using two search terms. The majority of tweets were
from tanners, with some reporting risky habits including burning or sleeping in the tanning
bed. Very few tweets were public health messages. Twitter may provide an opportunity to
study and intervene upon indoor tanners.

CORRESPONDING AUTHORS: Molly Waring, PhD, UMass Medical School;

[email protected]; Katie Baker, DrPH, East Tennessee State University;
[email protected]; Christine May, PhD, UMass Medical School;
[email protected]; Anthony Peluso, BS, East Tennessee State University;
[email protected]; Sherry Pagoto, PhD, UMass Medical School;
[email protected]

Symposium 63B

CUE-REACTIVITY IN INDOOR TANNING DISORDER: STATE-BASED INFLUENCES ON CRAVINGS

AND MOTIVATION TO TAN
S1604 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Dr. Derek Reed, PhD

Excessive indoor tanning is a behavioral excess mimicking behavioral addiction characteristics

associated with other addiction-related disorders such as drug use and gambling. Dense
literature in the addiction sciences suggests that cue-reactivity is a behavioral marker of
conditioned cravings and motivation for substances of abuse. Unlike other commodities of
abuse, there are very few regulations on the prevalence of tanning-related advertisements
and cues in public settings. Our study examined the extent to which indoor tanners exhibited
cue-reactivity in a study translated from those using cigarette and alcohol cue-reactivity
preparations. Findings suggests that tanning-related cues and paraphernalia significantly
increased markers of craving and motivation to tan in college-aged indoor tanners. These
tanning cue-reactivity findings provide novel insights into the role of tanning-related
messages and visual cues in tanning disorders; such findings also contribute to a growing
empirical body of support for tanning as a behavioral addiction. Notably, these cue-reactivity
findings suggest that tanning disorder may be a function of both trait and state influences.
Finally, the impact of cues on cravings and motivation for tanning also provide novel
implications for regulating the extent of tanning-related cues in advertisements, messages,
and media targeting youth.

CORRESPONDING AUTHORS: Amel Becirevic, MA, University of Kansas; [email protected]; Michael

Amlung, PhD, McMaster University; [email protected]

Symposium 63C

INNOVATION IN INDOOR TANNING MEDIA, MESSAGING, AND INTERVENTION SCIENCE

Dr. W. Douglas Evans, PhD

Background: Although skin cancer is largely preventable, it affects nearly 1 of 5 U.S. adults.
There is a need for research on how to optimally design persuasive, public health indoor
tanning (IT) prevention messages.

Objectives: To examine whether framed messages on indoor tanning behavioral intentions

delivered through text messaging (SMS) will produce: 1) positive responses to the messages,
including message receptivity and emotional response; 2) indoor tanning efficacy beliefs,
including response efficacy and self-efficacy; and 3) indoor tanning risk beliefs.

Methods: We conducted a pilot study of IT prevention messages delivered via mobile phone
text messaging in a sample of 21 young adult women who indoor tan. Participants completed
baseline measures, were randomized to receive gain-, loss-, or balanced-framed text
messages and completed post-exposure outcome measures on IT cognitions and behaviors.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1605

Participants received daily mobile phone indoor tanning prevention text messages for 1 week,
and completed the same post-exposure measures as baseline.

Results: Over the 1-week period there were trends or significant changes after receipt of the
text messages including decreased IT behavioral intentions (p = 0.105) and increased
perceived susceptibility (p p p = .027). OLS stepwise linear regression models showed an
effect of text message exposure on self-efficacy to quit indoor tanning (t = -2.475, p < .023).
OLS linear regression including all measured scales showed a marginal effect of texts on self-
efficacy (t = 1.905, p < .077). Participants endorsed highly favorable views towards the text
messaging protocol.

Conclusions: This study supports this use of mobile text messaging as an IT prevention
strategy. Given the nature of skin cancer risk perceptions, the addition of multimedia
messaging service (MMS) is another area of potential innovation for IT prevention messages.

CORRESPONDING AUTHOR: Darren Mays, Ph.D., Georgetown University;

[email protected]
S1606 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 64 1:30 PM-2:45 PM

THE TRANSFORMATION FROM 'I EXERCISE' TO 'I AM AN EXERCISER': HOW SELF-IDENTITY

INFLUENCES EXERCISE BEHAVIOR & MAINTENANCE

Angela D. Bryan, PhD1, Ryan E. Rhodes, PhD2, Arielle S. Gillman, MA1, Courtney J. Stevens,
MA3, Bess H. Marcus, Ph.D.4

1
University of Colorado Boulder, Boulder, CO; 2University of Victoria, Victoria, BC, Canada;
3
Alpert Medical School of Brown University, Providence, RI; 4University of California, San
Diego, La Jolla, CA

The transformation from I exercise to I am an exerciser: How self-identity influences

regular exercise behavior and maintenance

An estimated 80 percent of US adults do not meet government recommendations for physical

activity, and interventions to increase exercise, primarily based on social cognitive models,
continue to have only modest success, especially when it comes to behavior maintenance
over the long-term. To find insight into potential solutions to this problem, our symposium
turns to a long history of social psychological research on identity and behavior, and applies
these findings to exercise interventions to understand and enhance long-term maintenance of
physical activity behavior. We present novel mixed-methods research addressing this
question. Rhodes presents meta-analytic findings demonstrating a reliable relationship
between exercise identity and regular exercise behavior, and identifies correlates of identity
formation that link to theoretical frameworks that can guide intervention content. Gillman
examines the link between past behavior and identity formation in the exercise context, and
finds that regular exercise as a result of study participation can lead to spontaneous exercise
identity formation in previously sedentary women, leading to maintenance of exercise
behavior several months after study conclusion. Stevens examines psychological mechanisms
through which an acceptance-based clinical intervention to increase exercise behavior may
enhance exercise identity, and suggests that treatment-related effects on values congruent
behavior and remembered affect during and after exercise may support increases in exercise
identity, leading to greater maintenance at follow-up. Taken together, these investigations
highlight important connections between self-identity and the exercise experience that may
influence the maintenance of exercise behavior over time. Discussion will focus on factors
that may help guide incorporation of exerciser into the self-identity of insufficiently-
exercising individuals, and thus can be used in interventions to boost long-term behavioral
sustainability.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1607

LO1: To highlight a growing body of research demonstrating that exercise identity and regular
exercise behavior are reliably linked.

LO1: To translate meta-analytical findings on identity and exercise into novel interventions to
increase exercise behavior and maintenance

Symposium 64A

FORMING AN EXERCISE IDENTITY: A REVIEW OF CORRELATES AND POTENTIAL THEORETICAL

FRAMEWORKS

Professor Ryan E. Rhodes, PhD

Forming an Exercise Identity: A Review of Correlates and Potential Theoretical Frameworks

Exercise identity represents a reflexive, self-regulating behavioral maintenance system that

has shown a reliable, medium-sized correlation with regular physical activity (PA). Despite this
evidence, the process of how an exercise identity forms has received limited research
attention in the PA domain. Furthermore, the integration of exercise identity into
contemporary theories of PA has been under-developed. The purpose of this symposium
presentation is to provide a thematic review and analyses of the correlates of exercise
identity, as well as the contemporary theories that address identity formation in the PA
domain in order to facilitate future intervention research. Eligible studies were from English,
peer-reviewed published articles that examined identity (and/or schema) in the context of PA.
Searches were completed in June 2015 in five databases. Sixty-two independent data-sets,
primarily of modest quality, were identified. Results showed that identity was associated with
commitment, perceived capability, affective judgments, identified/integrated regulation and
social comparison but not associated with perceived benefits of PA or introjected/extrinsic
regulation. Theoretical models for understanding identity formation in the PA domain
included adapted self-efficacy and self-determination theories, the PA self-definition model,
multi-process action control, and variants of social identity, group cohesion and self-
perception theories. Overall, contemporary research on the correlates of exercise identity
support specific variables that span commitment, capability, affect, and social comparison and
these link to several applicable theoretical frameworks in order to explore identity formation.
Interventions employing experimental manipulations of these proposed antecedents are now
needed for a better understanding of exercise identity formation.

CORRESPONDING AUTHORS: Alison Quinlan, University of Victoria; [email protected]; Navin

kaushal, University of Montreal; [email protected]

Symposium 64B
S1608 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

EXERCISE IDENTITY AND BEHAVIOR MAINTENANCE AFTER A SUPERVISED EXERCISE

INTERVENTION

Arielle S. Gillman, MA

Given recent meta-analytic work demonstrating that exercise identity is reliably associated
with long-term maintenance of physical activity behavior, the next logical question to explore
is how to encourage inactive individuals to incorporate an exercise identity into their self-
concept. A long history of social psychological research has demonstrated that individuals are
likely to explain their past behavior as being the result of stable, internal characteristics. As
such, we examined whether simply committing to regular participation in exercise as part of a
short-term study could lead women to associate with an exerciser social category label,
leading to maintenance of exercise behavior after the study. Our participants were 240
previously sedentary women participating in a larger study that examined changes in
biological markers of breast cancer risk after 16 weeks of supervised exercise at differing
levels of intensity and duration. Participants were not given explicit guidance regarding
exercise maintenance after the intervention. Before and after the exercise program,
participants completed the Exercise Identity Scale (Anderson & Cychosz, 1994), a measure
examining the degree to which one identifies with exercise as an integral part of his or her
self-concept. 6 months post-intervention, participants self-reported their exercise behavior, as
well as psychosocial predictors including attitudes, norms, self-efficacy, and intentions to
exercise. We found that across our sample, exercise identity increased from baseline to post-
intervention (t=8.18, pr=.32), norms (r=.29), intentions, (r=.18), and minutes of exercise
(r=.18). These findings suggest useful implications for behavior change. It appears that
individuals can relatively spontaneously incorporate healthy behavior engagement into their
self-concept, even without explicit encouragement to do so. Continuous participation in
exercise behavior might thus create lasting effects on identity that can translate to behavior
maintenance over the long-term.

CORRESPONDING AUTHORS: Courtney Stevens, University of Colorado Boulder;

[email protected]; Angela Bryan, University of Colorado Boulder;
[email protected]

Symposium 64C

ASSOCIATIONS AMONG IDENTIFIED REGULATION, VALUES CONCORDANCE, AND EXERCISE

BEHAVIOR OVER TIME: AN ACCEPTANCE-BASED APPROACH

Courtney J. Stevens, MA
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1609

Regular exercise participation is associated with reduced risk for all-cuase mortality and has
also been shown to confer both prophylactic and therapeutic benefits for mental and
emotional health. Unfortunately, half of American adults are insufficiently active and another
quarter engage in no exercise at all. Prior work has shown that discomfort/displeasure
experienced during exercise is associated with less esercise engagement, but how to best
address these factors is not well understood. Rather than attempt to change or control the
experience of exercise, acceptance-based interventions seek to facilitate values identification
and promote behavior engagement consistent with those values (even in the presence of
experiential discomfort). Identified regulation is a construct posited by self-determination
theory to underlie motivation for behaviors that individuals personally identify with or value.
The aims of the present investigation were to (1) evaluate the influence of an acceptance-
based exercise intervention on exercise identified regulation and associated exercise behavior
change over time; and (2) explore the associations between identified regulation, values
concordance, and exercise behavior adoption and maintenance. The study was divided into
two phases: Phase 1 (adoption, baseline visit post intervention), and Phase 2 (maintenance,
3 and 6 months follow up). Data were collected at these 4 time points from a sample of N =
119 insufficiently active women randomly assigned to 1 of 3 conditions: (1) an acceptance-
based health coaching intervention, (2) an information-based health coaching intervention, or
(3) a no-health coaching control intervention. Results showed that identified regulation
increased through Phase 1 among the acceptance-based intervention participants, but
decreased over time among education and control participants, t(107) = 2.13, p = .03. Further,
across conditions, identified regulation at post-intervention was associated with more
exercise minutes completed at 3- and 6-months follow up, r = .19 - .29, respectively. Daily
journal data for participants in the acceptance-based condition indicated that acting in
accordance with values was positively associated with more minutes of exercise completed
during both the exercise adoption, r = .66, and maintenance phases, r = .27 - .45. These results
point to meaningful associations between identified regulation, values concordance, and
exercise behavior adoption and maintenance.

CORRESPONDING AUTHORS: Arielle Gillman, University of Colorado Boulder;

[email protected]; Angela Bryan, University of Colorado Boulder;
[email protected]
S1610 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 65 1:30 PM-2:45 PM

MILITARY AND VETERAN HEALTH BEHAVIOR RESEARCH AND PRACTICE

Jeffrey Haibach, PhD, MPH1, Daniel Cassidy, PhD2, Margaret Dundon, PhD3, Jennifer S.
Funderburk, PhD4, Gerald W. Talcott, PhD, ABPP5

1
U.S. Department of Veterans Affairs, Washington, DC; 2U.S. Air Force, San Antonio, TX; 3VHA,
Buffalo, NY; 4VA Center for Integrated Healthcare, Syracuse, NY; 5University of Tenessee,
Lackland AFB, TX

There are2.1 million current military Servicemembers and 21 million living Veterans in the
United States. Although they were healthier upon entering military service compared to the
general U.S. population, in the longer term Veterans tend to be of equivalent or worse health
than civilians. One primary explanation for the Veterans health disparity is poorer health
behaviors during or after military service, especially areas of physical activity, nutrition,
tobacco, and alcohol. In response, the Department of Defense and Department of Veterans
Affairs continue to develop, evaluate, and improve health promotion programs and
healthcare services for military and Veteran health behavior in an integrated approach. In this
symposium, the chair will first present the context and overview of Military and Veteran
health behavior research and practice, followed by presenters on specific interventions and
programs in key areas. Presenter 1 will discuss tobacco research and prevention through
DoDs initial military entrance training. Presenter 2 will discuss a quality improvement
implementation of VAs new primary care group health promotion program, Gateway to
Healthy Living. Presenter 3 will discuss the latest status on VAs primary care behavioral
health program including DoD collaboration. Symposium presenters and participants will then
discuss health behavior research and practice in both general and specific areas in forward
movement for the health of Servicemembers, Veterans, and their Families.

Symposium 65A

INTERVENING DURING THE TRANSITION INTO MILITARY SERVICE ON TOBACCO AND HEAVY
ALCOHOL USE

Daniel Cassidy, PhD

ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1611

Every year, nearly 40,000 men and women enlist in the United States Air Force (USAF). Prior
to arrival at Basic Military Training, Lackland Air Force Base, TX, these individuals will have met
or exceeded the militarys rigorous medical screening standards. While each of these Airmen
enters the USAF with fundamentally good health, many also arrive with or subsequently
develop health risk behaviors that have long-term costs of increased healthcare expenditure,
reduced operational readiness, and lower quality of life. The Department of Defense spends
an estimated $2.5 billion each year managing the ramifications of tobacco use and alcohol
misuse. In the critical transition from civilian to military life, the Military Health System strives
to apply the best available behavioral science for the short and long term health and
wellbeing of servicemembers. Early efforts to reduce tobacco use leaned heavily upon an 8
week or greater tobacco ban during initial entry training, and the initial ban remains in place.
Despite the majority of trainees espousing complete confidence in their ability to remain
tobacco-free after initial training, only 17-28% of those who used tobacco prior to enlistment
remain abstinent for at least 1 year, and new users are added. The resulting total percentage
of tobacco users remains higher at 1-year follow-up and, in general, higher in the military than
the civilian population. The story is much the same with regard to heavy alcohol use, which is
also prohibited during initial entry training: The ban on consumption is of enduring benefit for
some while others initiate or resume heavy use after the restrictions are lifted, suggesting a
persisting challenge and opportunity to intervene during this period. Brief individual and
group-based interventions delivered during and after initial entry training have demonstrated
utility in the reduction of both health risk behaviors. For example, the Alcohol Misuse
Prevention Program (AMPP) was associated with a 36% reduction in alcohol-related incidents
in one evaluation (e.g., driving under the influence; underage drinking) and can be delivered
to a large audience in less than 45 minutes. The AMPP data outcomes, along with similar
projects targeting tobacco use, cast light on characteristics common to successful population-
level prevention efforts. These interventions are generally delivered in a Socratic style, and
preliminary evidence suggests that effects are stronger when the verbal behavior of the
interventionist or program is aligned with the tenets of motivational interviewing. Programs
under investigation also leverage behavioral economic principles (e.g., choice architecture;
temporal discounting) to preserve freedom of choice while at the same time promote
healthier behavior that persists across the military and beyond.

CORRESPONDING AUTHOR: Melissa Little, PhD, University of Tennessee; [email protected]

Symposium 65B

VAS GATEWAY TO HEALTHY LIVING

Dr. Margaret Dundon, PhD

The Veteran population has high rates of unhealthy eating, physical inactivity and tobacco
use, similar to the general population. These 3 health behaviors are linked to many diseases
S1612 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

and half of global mortality. The challenges of supporting health behavior change include
helping people feel motivated, confident and willing to engage in evidence-based
interventions. The Gateway to Healthy Living Program was designed by the VHA National
Center for Health Promotion and Disease Prevention (NCP) to serve as an entry point for
Veterans who might benefit from self-care and healthy living programming. Gateway is
designed to evoke and increase motivation and confidence for health behavior change, and
link Veterans with effective interventions, programs (e.g., weight management, increasing
physical activity, tobacco cessation, limiting alcohol, and managing stress) and self-
management tools. Gateway offers participants help with goal setting, problem-solving and
change strategies as well as access to existing VHA clinical programs. Following the 90 minute
group session, there are two subsequent clinician/participant contacts to ensure access to
their program of choice, problem-solve, and affirm success. Gateway was piloted in 2015 in six
sites across the VA. All of the 83 participants reported being very satisfied or "somewhat
satisfied with their Gateway session. The majority of patients reported that Gateway helped
them set a goal to improve health. Sixty three percent of the 78 who completed the second
follow-up call reported linking to a program or resource.Staff were also surveyed (n = 50),
with 86% reporting that Gateway was very helpful in connecting Veterans to programs and
resources. Fifty-six percent indicated that Gateway was helpful for supporting health behavior
change, 74% reported Gateway helped improve efficiency of referrals to behavior change
resources, and 68% reported that it saved them time. Based on the pilot success, Gateway
training and program support are now being offered to all facilities in the VHA.

Symposium 65C

CURRENT STATUS AND RESEARCH FOR VA PRIMARY CARE AND BEHAVIORAL HEALTH
INTEGRATION

Dr. Jennifer S. Funderburk, PhD

One of the largest evidence-based transformations that the VA has implemented to help
address the high rates of mental and behavioral health symptoms within Veterans is the
integration of behavioral health services into primary care. To accomplish this, the VA has
integrated behavioral health providers into each primary care setting to provide focused
assessment, intervention, and consultation, and offers evidence-based care coordination via
care management models. This has resulted in over 1600 staff being hired to provide these
services in 89% of the 365 VA primary care sites across the nation. These integrated service
delivery models afford opportunities for trained mental health providers to provide both
secondary and tertiary prevention interventions targeting a variety of mental health (e.g.,
anxiety, depression) and behavioral health (e.g., diabetes management, insomnia) concerns.
Due to the increasing implementation of integrated delivery models outside of the VA, this
presentation will provide an overview of the current status of integration and evaluation
efforts within the VA to stimulate ideas for research and practice across both VA and non-VA
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1613

settings. This overview will help identify opportunities for behavioral medicine researchers to
expand their current work into the primary care setting. Although research has found
integration to improve patient access to services, many additional gaps remain. Therefore,
this presentation will also summarize gaps in our knowledge regarding integrated primary
care, including the effects of brief interventions on clinical outcomes, the best methods to
address multimorbidities, the most efficient ways to coordinate complementary services, keys
to improving communication and collaboration among the team, the need for brief measures
to assess patient functioning, and barriers and facilitators to implementation of integrated
primary care service delivery models. Discussion will focus on opportunities for future
research at a system- to patient-level on integration.

CORRESPONDING AUTHOR: Robyn Shepardson, PhD, VA Center for Integrated Healthcare;

[email protected]
S1614 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 66 1:30 PM-2:45 PM

MECHANISMS WITHIN ROMANTIC RELATIONSHIPS THAT IMPACT HEALTH BEHAVIORS AND

OUTCOMES: IDENTIFYING INTERVENTION TARGETS

Krista Ranby, PhD1, Jennalee Wooldridge, MA1, Michelle vanDellen, PhD2, Megan Lewis, PhD3,
Sydneyjane Varner, M.A.1

1
University of Colorado Denver, Denver, CO; 2University of Georgia, Athens, GA; 3RTI
International, Research Triangle Park, NC

An individuals social environment impacts their health behaviors, yet, the processes by which
this influence occurs is not well understood (LaGuardia & Patrick, 2014). Correlational
research indicates a positive association between social support and health behaviors yet
interventions focused on enhancing social support to change behaviors have fallen short. For
this reason, additional targets relevant to interpersonal relationships that may be useful for
health behavior interventions will be the focus of our symposium. With data from 280
engaged couples in Denver, CO, we found couples who had been living together for over 2
years were more similar on physical activity and BMI as compared to couples living together
for less than 2 years. This study highlights how individuals may change some health behaviors,
particularly daily habits, to match ones partner. Second, we will present an experimental
study in which 100 dual-smoker couples received either a couple-focused or individual-
focused message promoting smoking cessation. Message type and the relationship context
(i.e. conflict) interacted to impact motivations to quit smoking. We will discuss how the
effectiveness of dyadic interventions may be improved by considering qualities of the couples
relationship. Third, we examined relationship functioning and illness uncertainty at four time
points during the year following a prostate cancer diagnosis. Couples (N=165) were recruited
from Duke University Medical Center. Unsupportive behaviors increased and support
provision decreased following diagnosis; baseline illness uncertainty negatively related to
later measures of relationship functioning. Ways to intervene with couples to improve their
adjustment will be discussed. Finally, in a national sample of couples participating in a physical
activity (PA) intervention for Type II diabetes patients, we examined whether collaborative
implementation intentions (e.g., joint planning) improved accelerometer assessed PA and
whether this effect is mediated by partner investment in the PA goal and effectiveness of
social support and social control attempts. This symposium will encourage researchers to
consider the complex nature of the interpersonal environment for health behavior
engagement and change. Further, we will discuss the importance of considering both
partners perceptions, unique aspects of life transitions, and appropriate methodological
approaches for addressing these research questions.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1615

Symposium 66A

HOW COHABITATION AFFECTS HEALTH BEHAVIOR CONCORDANCE AMONG ENGAGED

COUPLES

Dr. Krista Ranby, PhD

Long time married couples experience similar health behaviors and health outcomes (Meyler,
Stimpson, & Peek, 2007). Potential causes of this concordance includes partnering with similar
others, shared resources and environments, and direct influences between partners. The
current study adds to our understanding about how couples come to be similar by examining
health behaviors within young adult, engaged couples. Couples who were over 18 and
considered themselves engaged were recruited from the Denver, Colorado metro area. Both
partners within 280 couples answered questions about their health behaviors, health goals,
and relationship. Participants were an average of 27.2 years old (SD=4.0) and had been in a
relationship with their current partner for an average of 46.5 months (SD=26.4); 90% of
couples were heterosexual relationships. When considering cohabitation, 18% did not live
together, 20% had lived together for less than a year, 22% had lived together for 13-24
months, and 41% had lived together for more than 2 years. Correlations of self-reported
health behaviors and outcomes from both partners were examined. Stronger correlations
emerged on BMI and days of exercise for couples who had lived together longer. This same
pattern was not found for smoking and alcohol use which were correlated among all couples
or sleep quality which was not correlated within couples. Longer relationship length did not
relate to stronger correlations to the same extent as did longer cohabitation time suggesting
daily habits may become more similar as people live together. Although couples were more
similar after living together, couples who had been living together longer were not
significantly more or less healthy than other couples. More work is needed to identify
predictors of which partner is likely to change within the relationship. It seems that in some
couples, one partner is improving their behaviors and in others, one partner is developing
poorer habits. Given that behavioral changes are naturally occurring during this life transition,
prevention interventions that consider relevant goals, motivations, and relationship dynamics
may be better able to shift behavioral habits in a positive direction so that habits established
within this interpersonal context might be maintained over time.

CORRESPONDING AUTHORS: Jennalee Wooldridge, MA, University of Colorado Denver;

[email protected]; Matthew Klein, University of Colorado Denver;
[email protected]

Symposium 66B
S1616 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

COLLABORATIVE IMPLEMENTATION INTENTIONS AMONG COUPLES FACING TYPE 2 DIABETES:

MECHANISMS OF BEHAVIOR CHANGE

Jennalee Wooldridge, MA

Many people with type 2 diabetes (T2D) do not engage in recommended self-management
behaviors (SMB), particularly physical activity (PA). Relationship partners are highly
interdependent and any significant lifestyle change made by one partner likely impacts the
other. Considering the interpersonal context of behavior change and necessity of maintaining
health behavior over time, couple-level processes are one promising area to incorporate into
behavior change interventions. The current study tests a PA intervention with couples in
which at least one partner had been diagnosed with T2D. The experimental manipulation
involved having couples make collaborative implementation intentions (CIIs), or specific plans
for how the diagnosed partner would overcome barriers to PA and follow specific plans to
engage in PA. We theorized this manipulation would be successful by changing several
important mediators. First, we believed CIIs would increase partner investment, or the degree
to which ones partner has a shared responsibility and takes action to support the goal of the
patient engaging in PA. Second, we believed CIIs would increase patients self-efficacy if their
partner is supporting the goal and possibly modeling the behavior themselves. Third, we
believed CIIs would increase health-related social support and positive control behaviors
provided by partners. Increases in all of these putative mediators are hypothesized to be
positively related to PA. PA (Actigraph accelerometers) and self-management behaviors were
assessed at baseline and 6 weeks post manipulation. Mediators were assessed from both
partners at baseline, 3 weeks, and 6 weeks. The relationship of the experimental manipulation
to the proposed mediators as well as the relationship of the mediators to PA and other SMB
will be discussed. This talk will focus on theoretical mechanisms, within relationship partners,
for behavior initiation, particularly within management of a chronic illness.

CORRESPONDING AUTHORS: Krista Ranby , PhD, University of Colorado Denver;

[email protected]; Susane Ogilby , BA, University of Colorado Denver;
[email protected]

Symposium 66C

THE INTERPLAY BETWEEN MESSAGE FRAME AND CONFLICT DURING SUPPORT ATTEMPTS IN
DUAL-SMOKER COUPLES

Dr. Michelle vanDellen, PhD

Nearly 2/3 of smokers are partnered with other smokers. These dual-smoker couples are less
likely than other smokers to try to quit. Very little is known about how relationship
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1617

dynamicsincluding social supportrelate to smoking cessation in dual-smoker couples. In

the present talk, we will discuss the interplay between dyadic interventions and the
relationship context among 100 dual-smoker couples. Specifically, we examined the impact of
a 2 (gain/loss) by 2 (couple/individual) message framing manipulation on motivation to quit
smoking. After completing baseline measures (including conflict during social support such as
nagging, criticism, and distrust), partners independently read scenarios discussing either the
benefits of quitting (gain-frame) or costs of not quitting (loss-frame) as it affected the
individual or the couple. After reviewing the scenarios and reporting their desire to quit
smoking, the couple members came together to discuss views on smoking. After the
discussion they again reported desire to quit. Post-discussion desire to quit varied by whether
the outcomes of framing focused on the individual or the couple and by baseline level of
conflict p < .05. Individuals who reported less conflict were more motivated to quit when the
messages were couple- rather than individual-focused. In contrast, individuals who reported
more conflict were more motivated to quit when the messages were individual- rather than
couple-focused. Thus, when individuals felt the relationship context was characterized by low
conflict, focusing on couple outcomes increased motivation to quit. These results suggest that
relationship dynamics may moderate the effectiveness of certain kinds of messages. Tailoring
messages to couples based on their relationship functioning (e.g., conflict) could increase
effectiveness. Future research examining how the relationship context changes motivations to
quit smoking may improve effectiveness of interventions for dual-smoker couples.

CORRESPONDING AUTHORS: Megan Lewis, PhD, Research Triangle International;

[email protected]; Benjamin Toll, PhD, Medical University of South Carolina; [email protected];
Isaac Lipkus, Phd, Duke University; [email protected]

Symposium 66D

ILLNESS UNCERTAINTY AND RELATIONSHIP FUNCTIONING AMONG COUPLES FACING

PROSTATE CANCER: A DYADIC, LONGITUDINAL EXAMINATION

Sydneyjane Varner, M.A.

Couples, both patients and partners, high in relationship functioning (RF) experience
improved biopsychosocial outcomes in the illness recovery process for prostate cancer (PC).
Illness uncertainty (IU), a major stressor for couples facing a PC diagnosis, may negatively
affect several domains of RF, thus impacting important health outcomes. Little is known,
however, about how the stress of IU impacts specific domains of RF between partners over
time during the year following diagnosis. Couples (N=165) in which one partner was
diagnosed with early stage PC were recruited from a urology clinic. Couples were assessed for
a year following diagnosis and prostatectomy. Participants completed measures of illness
uncertainty (IU), received partner support (PS), unsupportive behaviors (USB), and sexual
functioning (SF) at diagnosis, 1-month, 6-months, and 12-months. In the year following
S1618 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

diagnosis, USB increased significantly in both partners, while PS decreased significantly in both
partners. USB and PS are significantly correlated across time points within couples indicating
that there is variability in couples adjustment during the illness recovery process. Further,
significant correlations between baseline IU and relationship measures at various points
throughout the year suggest the stress of illness may be having a deleterious effect on
couples relationship functioning. These findings reveal trends in distinct measures of
relationship functioning during the year following PC diagnosis that have important
implications for future intervention development.. Rather than focusing on increasing general
relationship quality, targeting specific interaction patterns within couples, such as the ones in
this study, may improve relationship functioning in couples coping with illness recovery.

CORRESPONDING AUTHORS: Sydneyjane Varner, M.A., University of Colorado Denver;

[email protected]; Krista W. Ranby, Ph.D, University of Colorado Denver;
[email protected]; Cary Robertson, M.D., Duke University School of Medicine;
cary.robertson@dm. duke.edu; Issac Lipkus, Ph.D, Duke University School of Nursing;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1619

Symposium 67 1:30 PM-2:45 PM

ADVANCES IN UNDERSTANDING HEALTH EFFECTS OF STIGMA AND DISCRIMINATION IN

FAMILY CONTEXTS

Lisa Rosenthal, PhD1, David M. Frost, PhD2, Nicole M. Overstreet, PhD3, Hector F. Myers, PhD4

1
Pace University, New York, NY; 2University of Surrey, Guildford, England, United Kingdom;
3
Clark University, Worcester, MA; 4Vanderbilt University, Nashville, TN

Much research demonstrates that stigma and discrimination have adverse consequences for
mental and physical health.The majority of this work has focused on the individual level,
finding links between ones experiences of stigma or discrimination and ones own mental and
physical health outcomes. However, recent research and theory suggest stigma and
discrimination is not only experienced at the individual level, but rather should be understood
within the context of social networks, such as within family contexts. This symposium includes
three presentations of empirical studies that make advances in our understanding of health
effects of stigma and discrimination in family contexts across a range of populations, types of
stigma, and mental and physical health outcomes. The first presenter will report results
finding that among same-sex couples, both individual- and couple-level expectations of
rejection due to their socially stigmatized relationship were associated with greater
depressive symptoms and more problematic drinking behavior. The second presenter will
report results finding that among women exposed to intimate partner violence, stigmatizing
reactions to disclosure of that violence were associated with greater depressive symptoms
through the mechanism of greater avoidance coping. The third presenter will report results
finding that everyday discrimination reported by pregnant women of color prospectively
predicted poorer social-emotional development outcomes in their infants at one year old. The
discussant will address the presentations contributions to understanding inter-generational
and couple/family-level effects of stigma and discrimination on health, and implications for
future research and intervention. Overall, this symposium helps move the field forward in
developing a deeper understanding of dynamic health effects of stigma and discrimination
within family contexts, which can improve intervention efforts aimed at reducing health
disparities faced by stigmatized groups in the U.S.

Symposium 67A

DISCRIMINATION IN PREGNANCY PREDICTS POORER INFANT DEVELOPMENT AT ONE YEAR

S1620 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Dr. Lisa Rosenthal, PhD

Black women and some subgroups of Latina women in the U.S. experience disproportionately
high rates of adverse birth outcomes, such as preterm birth and low birth weight. Black and
Latino infants also have poorer behavioral, social, cognitive, and health outcomes than White
infants in the first year of life, and these disparities grow larger over time. Increasing evidence
suggests womens experiences of discrimination predict increased risk of adverse birth
outcomes, supporting that there are inter-generational consequences of discrimination.
However, almost no research has explored inter-generational effects of discrimination beyond
birth outcomes. In a sample of 334 predominantly Black and Latina teens and young women
receiving prenatal care in New York City, we longitudinally tested the hypothesis that
womens experiences of discrimination reported during pregnancy would prospectively
predict poorer infant development outcomes at one year old.

Participants reported their experiences with discrimination during second and third
trimesters, and an aggregate score for pregnancy was calculated. Participants also completed
items from the Infant Toddler Social Emotional Development Scale (ITSEA) assessing
inhibition/separation problems, attention skills, negative emotions, and positive emotions at
one year postpartum. Regression analyses controlling for a range of sociodemographic risk
factors found that womens experiences with discrimination reported during pregnancy
prospectively predicted greater inhibition/separation problems (B=0.87, SE=0.30, p=.026) and
greater negative emotions (B=0.99, SE=0.35, p=.030), but not attention skills or positive
emotion in their infants at one year old.

Findings support that there are inter-generational consequences of womens experiences of

discrimination for their infants development and well-being. Inter-generational effects of
discrimination may help to explain persistent disparities in a range of developmental, physical
health, and mental health outcomes across the life.

CORRESPONDING AUTHORS: Valerie Earnshaw, PhD, University of Delaware;

[email protected]; Joan Moore, BA, Pace University; [email protected]; Darrah
Ferguson, BA, Pace University; [email protected]; Ten Lewis, PhD, Emory
University; [email protected]; Allecia Reid, PhD, Colby College; [email protected];
Jessica Lewis, MFT, Yale University; [email protected]; Emily Stasko, MA, Drexel
University; [email protected]; Jonathan Tobin, PhD, Clinical Directors Network;
[email protected]; Jeannette Ickovics, PhD, Yale University;
[email protected]

Symposium 67B

THE EFFECT OF COUPLE-LEVEL MINORITY STRESS ON THE HEALTH OF SAME-SEX COUPLES

ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1621

Dr. David M. Frost, PhD

Minority stress theory (Meyer, 2003) posits that sexual minority individuals are at risk for poor
health due to the stress they experience stemming from social stigma. Being in a same-sex
couple has been theorized to result in exposure to unique forms of stigma, which are thought
to represent a risk to sexual minorities health, above and beyond the risk associated with the
stigma they experience as individuals (LeBlanc, Frost, & Wight, 2015). This presentation
reports the results of the first empirical test of this theory of Couple-Level Minority Stress.
A purposive sample of 216 partners in 108 sample-sex couples living in the US completed a
web-based questionnaire containing measures of individual minority stress in the form of
expectations of rejection, a newly developed measure of couple-level minority stress,
relationship satisfaction, depression (i.e., CESD) and alcohol use (i.e., AUDIT). The purposive
nature of the sampling design ensured that participants were nearly evenly distributed by
couple gender, relationship duration, and region of the country. Due to the dyadic nature of
the data, the Actor-Partner Interdependence Model (APIM) was used to analyze the data.
Analyses demonstrated that heightened expectations of rejection at both the individual- and
couple-level were associated with poorer health in the form of increased depressive
symptoms and more problematic drinking behavior. Couple-level expectations of rejection
explained poor health above and beyond expectations of rejection at the individual level
(actor effects) for both depression and drinking behavior, as well as at the dyadic level
(partner effects) for drinking behavior. Additionally, evidence for the proliferation of couple-
level minority stress was found in the form of a significant indirect effect of couple-level
expectations of rejection on depression that operated via decreased levels of relationship
satisfaction. Implications for research on sexual minority health disparities and clinical
interventions aimed at improving the health of sexual minority individuals and same-sex
couples will be discussed.

CORRESPONDING AUTHOR: Allen LeBlanc, PhD, San Francisco State University;

[email protected]

Symposium 67C

STIGMATIZING REACTIONS TO INTIMATE PARTNER VIOLENCE DISCLOSURE AS PREDICTORS OF

AVOIDANCE COPING AND DEPRESSIVE SYMPTOMS

Dr. Nicole M. Overstreet, PhD

Women exposed to intimate partner violence (IPV) are at higher risk for elevated depressive
symptoms and major depressive disorder than women not exposed to IPV. Recent research
suggests that negative social reactions to IPV disclosure may be associated with womens
depressive symptoms. Despite increased attention to the relation between negative social
S1622 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

reactions to intimate partner violence (IPV) disclosure and poorer mental health outcomes for
victims, research has yet to examine whether certain types of negative social reactions are
associated with poorer mental health outcomes more so than others. Further, research is
scarce on potential mediators of this relationship. To fill these gaps, the current study
examines whether stigmatizing reactions to IPV disclosure, such as victim-blaming responses
and minimizing experiences of IPV, are a specific type of negative social reaction that exerts
greater influence on womens depressive symptoms than general negative reactions, such as
being angry at the perpetrators of IPV. We also examine avoidance coping as a key mediator
of this relationship. A cross-sectional correlational study was conducted to examine these
relationships. Participants were 212 women from an urban northeast community who
indicated being physically victimized by their male partner in the past six months. Findings
from a multiple regression analysis showed that stigmatizing reactions, not general negative
reactions, predicted womens depressive symptoms. In addition, a multiple mediation analysis
revealed that avoidance coping strategies, but not approach coping strategies, significantly
accounted for the relationship between stigmatizing social reactions and womens depressive
symptoms. Implications of the study findings for improving support from informal and formal
sources and IPV-exposed womens psychological well-being will be discussed.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1623

Symposium 68 1:30 PM-2:45 PM

PROMOTING PSYCHOLOGICAL WELL-BEING IN CANCER PATIENTS AND SURVIVORS: META-

ANALYSES OF RANDOMIZED CLINICAL TRIALS

John M. Salsman, PhD1, Judith T. Moskowitz, PhD, MPH2, Crystal L. Park, PhD3, Thomas V.
Merluzzi, PhD4

1
Wake Forest School of Medicine, Winston Salem, NC; 2Northwestern University Feinberg
School of Medicine, Chicago, IL; 3University of Connecticut, Storrs Mansfield, CT; 4University
of Notre Dame, Notre Dame, IN

Positive affect, meaning and purpose, and self-efficacy represent dominant concepts in
human flourishing and are core components of psychological well-being. In cancer patients
and survivors, these concepts are significantly associated with better health outcomes above
and beyond the influence of distress and dysfunction. Several systematic reviews and meta-
analyses highlight the importance of psychosocial interventions to reduce negative effects of
depression, pain, and fatigue in patients with cancer, but we lack similar meta-analyses of
interventions that promote and sustain well-being. As more people survive cancer,
understanding and facilitating positive adjustment and growth is an important complement to
understanding and mitigating impairment, disability and psychosocial morbidity. In this
symposium, we will provide a summary and meta-analysis of psychosocial intervention effects
on positive psychological outcomes and identify those interventions that have been optimally
beneficial for people with cancer. Our aims are threefold: (1) obtain current, comprehensive,
and robust estimates of the effectiveness of psychosocial interventions on positive affect,
meaning and purpose, self-efficacy among cancer patients and survivors, (2) determine
whether the effects of psychosocial interventions on positive affect, meaning and purpose,
and self-efficacy differ by the intervention focus and modality, and (3) examine patient
sociodemographic (age, gender) and clinical characteristics (cancer type, stage, and phase) as
moderators of the strength of the relationship between psychosocial interventions and
positive affect, meaning and purpose, and self-efficacy. Our first presentation is on the effects
of interventions on positive emotions, optimism, and vigor. Our second presentation
examines the effects of interventions on meaning, life purpose, and posttraumatic growth.
Our final presentation analyses the effects of interventions on general self-efficacy as well as
domain or task-specific self-efficacy. We will close by underscoring important similarities and
differences among the three presentations and delineating the utility of psychosocial
interventions to promote positive psychological outcomes among cancer patients and
survivors. In so doing, we will clarify the role of psychosocial interventions in promoting better
health and well-being among patients with cancer.
S1624 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 68A

DONT WORRY, BE HAPPY? META-ANALYSIS OF RANDOMIZED INTERVENTIONS ON POSITIVE

AFFECT IN CANCER PATIENTS AND SURVIVORS

Judith T. Moskowitz, PhD, MPH

Introduction: Positive affects such as happiness, excitement, and contentment have unique
beneficial effects in the context of health-related stress including improved health behaviors,
better quality of life, and increased longevity. In the past decade, empirical evidence has
emerged suggesting that positive affect may be an important target for behavioral
interventions given its unique beneficial relationship with health outcomes. However, no
systematic reviews exist to quantify the effects of behavioral interventions and better
understand what works, for whom, and under what conditions within the context of cancer.
This meta-analysis investigated the extent to which psychosocial interventions in people with
cancer have effects on positive affect outcomes. Methods: Drawing from seven electronic
databases, 1138 abstracts were identified, evaluated, and coded by reviewer pairs, yielding
thirty-one randomized behavioral clinical trials. Effect sizes were calculated for positive affect
outcomes such as positive emotion, optimism, and vigor. Synthesis was conducted using
correlated random effects models with robust variance estimation to account for dependence
among multiple effect sizes within each study. Potential moderators included sample age,
gender, cancer type, cancer phase, intervention focus, and delivery format. Results: The
weighted average effect across thirty-one studies of positive affect outcomes was g = 0.134,
95% CI [-0.043, 0.312]. There was substantial heterogeneity of effects across studies, with an
estimated between-study variance of 0.193 (I2: 83%). None of the moderators were
statistically significant (all P > 0.574). Results did not exhibit outcome-reporting biases (P =
0.412 from a robust Egger regression test). Discussion: Although several behavioral
interventions included positive affect as an outcome, few targeted positive affect as the
primary outcome and results were typically small in magnitude. Surprisingly, intervention
efficacy did not vary as a function of clinical, demographic, or intervention characteristics.
Tailored interventions to promote positive affect with careful attention to measurement of
positive affect may potentially yield greater effects, but more studies are needed to draw
firmer conclusions.

CORRESPONDING AUTHORS: James E. Pustejovsky, PhD, University of Texas at Austin;

[email protected]; Rosalba Hernandez, PhD, University of Illinois at Urbana-
Champaign; [email protected]; Stephen M. Schueller, PhD, Northwestern University
Feinberg School of Medicine; [email protected]; Mark Berendsen, MLIS,
Northwestern University Feinberg School of Medicine; [email protected];
John M. Salsman, PhD, Wake Forest School of Medicine; [email protected]

Symposium 68B
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1625

DO PSYCHOSOCIAL INTERVENTIONS WITH CANCER PATIENTS AND SURVIVORS INCREASE

MEANING AND PURPOSE? A META-ANALYTIC APPROACH

Dr. Crystal L. Park, PhD

A sense of meaning and purpose are important for individuals dealing with cancer. Meaning
can buffer stress effects by allowing individuals to view their cancer experiences as one part of
a larger, richer whole or to reframe it in more positive ways (post-traumatic growth). Yet,
cancer can threaten individuals sense of meaning by undermining their views of the world as
benevolent and controllable, their identities as healthy and intact, and their futures as
optimistic and long. Identifying whether interventions can promote a sense of meaning,
purpose and growth is important for improving the well-being of cancer patients and
survivors. This meta-analytic study examined the efficacy of interventions that promote
meaning and purpose and potential moderators (age, gender, cancer type, cancer stage,
cancer phase, intervention modality, and delivery format) of this effect. 2080 abstracts from 7
electronic databases were evaluated by reviewer pairs. Inclusion criteria included randomized
controlled trials of a psychosocial intervention with cancer patients and/or survivors that
included an outcome measure of meaning, purpose, or growth. A total of 38 randomized
trials met these criteria. Meta-analyses used correlated random effects models with robust
variance estimation to account for dependence among multiple effect size estimates within
each study. The overall average effect across studies that included over 3000 people was g =
0.313, 95% CI [0.165, 0.461], P < .001. Effects were heterogeneous with an estimated
between-study variance of 0.138 (I2 71%). No moderating effects were identified. There was a
strong indication of publication bias (P = .0196 from a robust Egger regression test). Using
Egger regression to adjust for potential small-sample bias reduced the overall average effect
to g = -0.059, 95% CI [-0.323, 0.205], P = .642, I2 68%. These results suggest that, although
many interventions for improving meaning and purpose have reported salutary effects, the
overall pattern of findings may be a function of outcome reporting biases rather than
evidence that the interventions are effective in bolstering meaning and purpose in the context
of cancer.

CORRESPONDING AUTHORS: James Pustejovsky, PhD, University of Texas;

[email protected]; Kelly Trevino, PhD, Weill Cornell Medical Center;
[email protected]; Allen Sherman, PhD, University of Arkansas for Medical Sciences;
[email protected],; Craig Esposito, PhD, University of Connecticut;
[email protected]; Mark Berendsen, PhD, Northwestern University Feinberg School of
Medicine; [email protected]; John Salsman, PhD, Wake Forest School of
Medicine; [email protected]

Symposium 68C
S1626 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

A META-ANALYTIC REVIEW OF INTERVENTIONS TO ENHANCE SELF-EFFICACY BEHAVIORS IN

CANCER PATIENTS AND SURVIVORS

Dr. Thomas V. Merluzzi, PhD

Background: Peoples beliefs and expectations about their ability to exercise control or
mastery over their behaviors represent an underlying and central mechanism of human
agency and personal efficacy. Self-efficacy (SE) theory posits that greater confidence in the
ability to execute courses of action enhances the probability that desired goals will be
attained. Thus, cancer patients with higher SE for processes like coping, symptom
management, and adherence to medical regimens are more likely to persist in strategies to
achieve desired psychosocial (e.g., better adjustment and quality of life) and medical
outcomes (e.g., less and/or less intense symptoms and side effects) compared with those with
lower SE. This meta-analytic study determined 1) whether interventions to increase SE are
effective and 2) if moderators like intervention focus are important. Method: Synthesis was
conducted using correlated random effects models with robust variance estimation to
account for dependence among multiple effect size estimates within each study. Outcome
reporting bias was examined using funnel plots and a robust variant of Egger's regression.
Moderators examined included age, gender, cancer type, intervention focus, and delivery
format. Results: Across 45 randomized control trials examining the effects of psycho-social
interventions, the weighted average effect on SE outcomes was g=0.338, 95% CI[0.233,
0.443],P<.001. There was substantial heterogeneity of effects across studies, with an
estimated between-study variance of 0.092 (I-squared:72%). Intervention focus explained
significant variation in the effect size estimates (F(2,11.2)=7.73,P=0.008). The average effect
for individual-based interventions was g=0.218 (95% CI[0.121, 0.315], k=28 studies), for dyad-
based interventions g=0.458 (95% CI[0.089, 0.829], k=9 studies), and for group-based
interventions g=0.702 (95% CI [0.409, 0.995], k=7 studies). Effects for group-based
interventions were statistically distinguishable from effects for individual-based interventions
(P=.015) after multiple-comparisons correction). Results were not sensitive to outcome-
reporting or small-sample biases (P=.547, robust Egger regression test). Discussion: Results
suggest that SE does improve in the context of an intervention. Moreover, group
interventions enhance self-efficacy more effectively than individually administered
interventions. Consistent with SE theory, the group format may enhance SE through social
modeling, persuasion in a group context, and support. Thus, in general, this preliminary
evidence supports a preference for group-based rather than individual-based interventions to
enhance SE.

CORRESPONDING AUTHORS: James Pustejovsky, PhD, University of Texas; [email protected];

Errol Philip, PhD, University of Notre Dame; [email protected]; Mark Berendsen, MLIS,
Northwestern University; [email protected]; Carolyn Heitzmann Ruhf, PhD,
University of Southern California; [email protected]; John Salsman, PhD, Wake Forest
University; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1627

Symposium 69 1:30 PM-2:45 PM

MEASURING PERCEIVED RISK: THE PROFOUND IMPLICATIONS OF SEEMINGLY TRIVIAL

CHOICES ABOUT ITEM WORDING AND FORMATTING

Annette Kaufman, PhD, MPH1, Erika A. Waters, PhD, MPH2, Daniel L. Hall, Ph.D.3, Marc T.
Kiviniemi, BA, PhD4, Rebecca A. Ferrer, PhD5

1
Tobacco Control Research Branch, Behavioral Research Program, Division of Cancer Control
and Population Sciences, National Cancer Institute, National Institutes of Health, Rockville,
MD; 2Washington University in St. Louis, St. Louis, MO; 3Harvard Medical
School/MGH/BIDMC, Wellesley, MA; 4University at Buffalo, Buffalo, NY; 5National Cancer
Institute, Rockville, MD

Most health behavior theories assert that higher perceptions of personal risk are associated
with higher engagement in healthy behaviors and lower engagement in unhealthy behaviors.
Yet, many empirical studies have reported conflicting results in which the link between
perceived risk and behavior was positive, negative, or absent. This has led some researchers
to conclude that perceived risk is not an important predictor of behavior change and that
health communication and behavior interventions should focus on addressing other key
concepts (e.g., perceived benefits) or levels of influence (e.g., policy change).

However, conclusions such as these have not adequately addressed whether conflicting
findings might be due to methodological considerations such as the specificity or generality of
item wording and the characteristics of response options. If the inconsistent association
between risk perception and behavior is attributed to limitations of risk perception as an
important construct in health behavior change, rather than to methodological characteristics
that limit the predictive validity of risk perception measures, health communication and
behavior interventions that do not address risk perceptions may be unnecessarily self-
limiting. Furthermore, such methodological variation may cloud our understanding of the
meaning and magnitude of population perceived risk, obscure the true relationship
between perceived risk and behavior, and consequently inhibit the advancement of health
behavior theory and the development of effective health behavior change interventions.

This symposium will provide insight into the effects of common decisions about the wording
and formatting of perceived risk items. It accomplishes this by (1) providing a systematic
S1628 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

review of how risk perceptions are measured in the literature using tobacco use as a case
study; (2) demonstrating that the correlations between perceived risk and other health
beliefs, behavioral intentions, and behavior differ dramatically based on whether the risk
perception items are or are not contingent on plans to engage in risk-mitigating behavior; (3)
describing how perceptions of risk among current and former smokers who undergo lung
cancer screening vary in magnitude depending on the specificity of the risk perception item
and whether the items assess absolute or comparative risk perceptions; and (4) providing
evidence against the common practices of (a) not including a dont know response option
for risk perception items, and (b) treating dont know responses as missing data. The
discussant, a program director at the National Cancer Institute, will identify future research
needs and will describe the theoretical, methodological, and practical implications of the
research.

Symposium 69A

RISK PERCEPTION MEASUREMENT IN TOBACCO CONTROL RESEARCH: A SYSTEMATIC REVIEW

Dr. Annette Kaufman, PhD, MPH

Despite the importance of perceived risk and the extraordinary breadth and volume of
tobacco control research assessing risk perceptions, no studies have been conducted to
review how this construct is measured across tobacco studies. The present review describes
the characteristics of risk perception measures used in tobacco control research and evaluates
whether studies incorporate best practices for risk perception measures put forward by risk
perception researchers. Three databases (PubMed, PsycINFO, and Web of Science) were
searched in March 2015, for published English language peer-reviewed articles measuring
tobacco risk perceptions (N=2,557). Three coders independently coded abstracts for initial
inclusion. A total of 441 articles met initial inclusion criteria and 100 were randomly selected
for full-text review. Three coders independently coded critical aspects of each article, multi-
item measure, and item: 54 articles, 33 measures, and 239 items were coded. We found large
gaps in the use of best practices for measuring risk perceptions. Only 14 articles had a multi-
item measure focused exclusively on risk perception and 11 articles had only one risk
perception item. Many items asked about general health outcomes (43%), did not specify the
person for whom risk was being judged (44%; e.g., self, average person), or did not specify the
conditions for which the risk was being judged (27%; e.g., the product used, intensity of use).
Few items assessed constructs known to provide additional information about how people
deal with risk, including: perceived severity (11%), affective responses to risk (9%), perceived
controllability (8%), or the extent to which risks are seen as unknown to experts (3%). Findings
suggest that tobacco research inconsistently incorporated risk perception measurement
suggested best practices, despite their established importance. Further promulgating best
practice guidelines for assessing risk perception could be useful to tobacco research. A set of
risk perception consensus measures may benefit the field.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1629

CORRESPONDING AUTHORS: Alexander Persoskie, PhD, Center for Tobacco Products, U.S.
Food and Drug Administration; [email protected]; Julie Bromberg, MPH, BLH
Technologies; [email protected]; Jenny Twesten, MPH, BLH Technologies;
[email protected]

Symposium 69B

CONDITIONING PERCEIVED RISK ITEMS ON FUTURE BEHAVIOR: IMPLICATIONS FOR

ADVANCING THEORY AND DEVELOPING INTERVENTIONS

Erika A. Waters, PhD, MPH

Background: Risk perception survey items that are conditioned on behavior (e.g., How likely is
that you will get colon cancer IF YOU DO NOT GET REGULAR EXERCISE) may be interpreted
differently than unconditioned items.

Objective: Examine whether conditioning risk perception items on exercise behavior alters the
association between perceived risk and health beliefs, intentions, and behavior.

Methods: Secondary analysis of data from a risk communication study (N = 812). Participants
were members of GfK Knowledge Networks, an Internet survey panel. Each participant
answered an unconditioned and conditioned version of 4 types of colon cancer risk
perception items: cognitive absolute perceived likelihood, affective absolute perceived
likelihood, cognitive comparative perceived likelihood, and affective comparative perceived
likelihood.

Results: Conditioned items elicited fewer dont know responses than unconditioned items
for the cognitive items (McNemar test ps < .005) but not the affective items (McNemar ps >
.48). Mean risk perceptions were higher for all conditioned than non-conditioned items
(paired t-test ps < .001). All conditioned items were positively correlated with intentions to
exercise (ps < .0001), but none of the unconditioned items were (ps > .10). Conditioned items
were positively correlated with current exercise behavior (ps < .05), but unconditioned items
were either unrelated to (ps > .18) or inversely correlated (ps < .02) with current exercise
behavior. Conditioned items had weaker correlations with worry and stronger correlations
with anticipated regret than unconditioned items (Steiger z-test ps < .001). Perceived severity
was independent from conditioned items (ps > .22) but positively correlated with
unconditioned items (ps < .005). Both conditioned cognitive items and the affective
S1630 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

comparative likelihood item accounted for variance in intentions not accounted for by
response efficacy (ps < .05), which suggests that conditioned risk perception items are not
simply another way to assess response efficacy.

Conclusion: The decision about whether or not to condition perceived risk items on behavior
has critical implications for participant comprehension of survey items, potential missing data,
operationalization of health behavior theory constructs, and drawing correct inferences about
the relationship between perceived risk, other health behavior constructs, and behavior. Not
attending to such considerations may impede intervention development and evaluation.

CORRESPONDING AUTHOR: Courtney Wheeler, MPH, Louisiana State University;

[email protected]

Symposium 69C

RISK PERCEPTIONS AMONG LUNG SCREENING PATIENTS: ONE SIZE DOES NOT FIT ALL

Daniel L. Hall, Ph.D.

Background: Multiple U.S. health agencies have recommended ongoing lung screening for
current and former heavy smokers, which Medicare has recently began covering. Risk
perceptions are known to influence patient uptake of screening guidelines. To date, little is
known about risk perceptions about lung cancer (LC) or smoking-related diseases (SRD), and
whether these perceptions vary by smoking status and/or reference point (i.e., absolute vs.
comparative risk).

Method: This observational, cross-sectional study examined current smokers (n=88) and
former smokers (n=81) who were undergoing lung screening at a large academic medical
hospital. Ten self-report items assessed risk perceptions regarding LC and SRD, framed in
either absolute or comparative terms. Items were scored (1=Low to 5=High), summed, and
averaged to facilitate comparisons (i.e., LC vs. SRD; absolute vs. comparative). Independent
and paired-samples t-tests were conducted to assess (dis)concordance in responses for the
total sample as well as by smoking status.

Results: Overall, lung screening patients perceived moderate levels of health risk (M=3.57,
SD=.68), although there was variability by smoking status and reference point. Specifically,
patients endorsed higher health risk when questions were presented in terms of absolute
health risk versus comparative health risk (p < .001). This pattern was true for both current
and former smokers (ps < .001). When asked about risk for LC and SRD separately, current
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1631

smokers perceived lower risk for LC than for SRDs (p < .001), while former smokers assessed
their risk for both disease groups equally (p = .29).

Discussion: While lung screening patients endorsed moderately elevated health risk overall,
ratings varied by reference point, smoking status, and disease in question. The significant
variability by each of these factors among current and former smokers underscores the
heterogeneity of risk perceptions in the context of lung screening, as well as the complexity
involved in communicating lung screening recommendations and results.

CORRESPONDING AUTHORS: Inga Lennes, MD, MPH, MBA, MGH Cancer Center;
[email protected]; Alaina Carr, BA, University of Colorado at Denver;
[email protected]; Justin Eusebio, BA, MGH Cancer Center;
[email protected]; Elyse Park, PhD, MPH, Harvard Medical School;
[email protected]

Symposium 69D

DO PEOPLE KNOW WHAT THEY SAY THEY DONT? PRESENCE AND POSITIONING OF DONT
KNOW RESPONSE OPTIONS SHIFT PERCEIVED RISK ESTIMATES

Dr. Marc T. Kiviniemi, BA, PhD

BACKGROUND: People sometimes report not knowing their risk for health problems. Some
argue that dont know responses (DKR) reflect satisficing and are essentially error variance.
One should therefore discourage DKR by not providing a response option and/or treat DKRs as
missing data. Others hold that DKRs are meaningful, reflecting differences in health
knowledge and other factors, and should be assessed by including a DKR option. We
conducted two online survey studies of US adults to examine evidence for these competing
assertions.

METHODS AND RESULTS: In Study 1, absolute and comparative risk for both colon cancer and
diabetes was assessed. A DKR option was presented as either the first or last response option.
Participants (N=280) were more likely to use the DKR option when it was presented first (e.g.,
for colon cancer absolute risk, DKRs were 10.4% percentage points higher when presented
first). For colon cancer, for which more people say they dont know their risk than for
diabetes, the means for both types of risk were both significantly higher when DKR was last;
both ts>2.34, both ps<.05. This pattern was also found for diabetes, although differences
were not statistically significant. In Study 2, participants (N=960) reported absolute and
comparative perceived risk for four health problems. Participants received items that either
did or did not have a DKR option (placed last). In all risk domains, the DKR option was used
when presented, with rates ranging from 3.5% (absolute risk, car crash) to 12.7%
(comparative risk, colon cancer). For absolute risk, risk estimates were systematically higher in
S1632 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

the DKR provided condition for all four risk domains, Fs<3.78, all ps<.05. For comparative risk,
means did not differ by condition.

CONCLUSIONS: These findings cast serious doubt on the idea that one should simply reduce
DKRs by not providing an option or treat DKRs as missing data. The differing patterns for
diabetes versus cancer (Study 1) and for absolute versus comparative risk (Study 2) also call
into doubt the idea that DKRs simply reflect satisficing, as this would predict the same
response patterns across domains. The presentation will further explore implications for
understanding of lay estimation of health risks and guidance on how to address DKRs in
questionnaire development and data analysis.

CORRESPONDING AUTHORS: Marc Kiviniemi, University at Buffalo; [email protected];

Heather Orom, University at Buffalo; [email protected]; Caitlin Biddle, University at Buffalo;
[email protected]; Jennifer Hay, Memorial Sloan Kettering Cancer Center;
[email protected]; Erika Waters, Washington University in Saint Louis School of Medicine;
[email protected]; Yuelin Li, Memorial Sloan Kettering Cancer Center;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1633

Paper Session 33: Cancer in Underserved Populations 3:00 PM-3:18 PM

IMPACT OF INDIVIDUAL AND AREA LEVEL RACE/ETHNICITY ON ILLNESS INTRUSIVENESS

AMONG CANCER SURVIVORS

Corinne Leach, MS PhD MPH1, Rhyan Vereen, MPH2, Arthi Rao, PhD2, Katie Ross, BSPH3, Liora
Sahar, PhD, GISP2

1
American Cancer Society's Behavior Research Center, Atlanta, GA; 2American Cancer Society,
Atlanta, GA; 3Rollins School of Public Health, Emory University, Decatur, GA

Illness intrusiveness describes how much a specified disease, such as cancer, disrupts various
aspects of someones life, including physical wellbeing, work, finances, relationships, and
recreational activities. The socioecological model posits that individual health status is
simultaneously produced by individual biology and the surrounding physical, social, and
cultural environment. Despite evidence of the need to consider these hierarchical influences
on cancer outcomes, little is known about the impact of area level factors on cancer-related
outcomes, including illness intrusiveness.

Data from 993 breast, colorectal and prostate cancer survivors within the first year of
completing active treatment were analyzed. Individual level data for the Illness Intrusiveness
Ratings Scale were linked to census tract level data to characterize neighborhood conditions.
As components of area indices, area-level race and education were investigated along with
corresponding individual factors to study their interactive effect on illness intrusiveness.

232 (23.4%) participants reported relatively high illness intrusiveness (score >28). Education
was not a significant predictor of illness intrusion. A model including interaction between area
and individual-level race among other individual confounders exhibited significantly improved
model fit (p < 0.05). The model showed that racial minorities living in areas with a higher
percentage of racial minorities had higher odds of having greater illness intrusiveness when
compared to Whites living in areas with a low percentage of racial minorities (aOR: 1.65, CI:
1.01, 2.68).

The results of this study suggest a differential effect of area level race composition, often
characterized by fewer resources, by individual level race. While this study assesses area-level
factors independently, these measures are also components of combined measures such as
area deprivation indices. Further research will evaluate our research question using an Area
Deprivation Index that is currently being developed. Findings may support the collective
resources model which posits that the effect of area deprivation is greater in certain
S1634 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

sociodemographic groups who may find it difficult to seek potentially beneficial resources
outside of their living area.

CORRESPONDING AUTHOR: Corinne Leach, MS PhD MPH, American Cancer Society's Behavior
Research Center, Atlanta, GA, 30303; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1635

Paper Session 33: Cancer in Underserved Populations 3:19 PM-3:36 PM

REACHING A HARD TO REACH PATIENT POPULATION: USING A MOBILE TECHNOLOGY-BASED

TOOL TO PROVIDE INFORMATION AND SUPPORT

Catherine Benedict, PhD1, Dave Fuehrer, N/A2, Brad Love, PhD3, David Victorson, PhD4

1
Hofstra Northwell School of Medicine, Manhassat, NY; 2Stupid Cancer, New York, NY;
3
University of Texas at Austin, Austin, TX; 4Robert H. Lurie Comprehensive Cancer Center,
Northwestern University, Chicago, IL

Background: Adolescent and young adult (AYA; 15-39 years old) cancer survivors report an
unmet need for emotional and informational support. Instapeer is a mobile app that provides
AYAs access to instant, anonymous peer support and cancer information. Users complete a
profile and are matched with peers based on self-selected criteria or can participate in
unstructured group chats. This study evaluated AYAs use of the app and preferences for peer-
to-peer connection and app-based support.

Methods: In the first 12 months of public launch, 2,343 AYAs representing 44 countries
downloaded the app and completed a profile (demographic/medical information, treatment
side effects, and psychosocial concerns). A product and user analysis was conducted to
evaluate app functioning and user satisfaction including quantitative app metrics and in-depth
qualitative user interviews (N=6).

Results: Users were primarily female (74%), single (41%), childless (65%), and post-treatment
(54%). Most common diagnoses were lymphoma (21%) and breast (18%). 23 unique
treatment side effects and 24 psychosocial concerns were endorsed; 46% of users reported 6
side effects and 65% reported 6 psychosocial concerns. Users averaged 1,121 interactions
per week for a total of 3,595 hours of time spent in the app over 12 months. Group chat
rooms represented the majority of total message exchanges; however, users participating in
one-to-one messaging appeared to make stronger connections with more frequent message
exchanges.Qualitative user feedback centered around 3 primary themes: desire to connect
with others of the same age and diagnosis; interest in connecting with peers based on non-
medical information (e.g., interests/hobbies, location); and hesitation to connect (i.e., send a
request).

Conclusions: Findings provide initial evidence for AYAs interest and enthusiasm for using an
mobile phone app to connect with peers and access support. Features that allow AYAs to
connect with peers based on non-medical similarities and for one-to-one communication may
be critical. Strategies to facilitate initial connections may also be needed to encourage
S1636 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

connections. App-based support tools may offer a promising approach to reach this
underserved patient population. Strategies to promote peer connections and meaningful
exchange of support should be explored based on AYAs' preferences and needs. Future
research should evaluate whether access to support via this technology leads to
improvements in psychosocial outcomes such as perceived support and social isolation.

CORRESPONDING AUTHOR: Catherine Benedict, PhD, Hofstra Northwell School of Medicine,

Manhassat, NY, 11030; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1637

Paper Session 33: Cancer in Underserved Populations 3:37 PM-3:54 PM

MATERIAL FINANCIAL HARDSHIP AND LIMITING CARE DUE TO COST IN A MULTI-RACIAL

COHORT OF CANCER SURVIVORS

Theresa A. Hastert, Ph.D., M.P.P.1, Matthew Banegas, Ph.D., M.P.H., M.S.2, Lauren M. Hamel,
PhD3, Ann G. Schwartz, Ph.D., M.P.H.4, Tara Baird, M.A. 1, Jennifer L. Beebe-Dimmer, Ph.D.,
M.P.H.1

1
Wayne State University School of Medicine/Karmanos Cancer Institute, Detroit, MI; 2Kaiser
Permanente, Portland, OR; 3Wayne State University/Karmanos Cancer Institute, Detroit, MI;
4
Karmanos Cancer Institute/Wayne State University, Detroit, MI

Background: Nearly one-third of cancer survivors in the United States experience financial
hardship due to cancer. We examined whether experiencing material financial hardship was
associated with limiting care due to cost among cancer survivors in Metropolitan Detroit.

Methods: Data from the Detroit Research on Cancer Survivorship (ROCS) cohort include
responses from 682 (275 white, 407 black) adults (ages 27-79) diagnosed with a first primary
breast, colorectal, lung, or prostate cancer since January 1, 2013 and diagnosed or treated at
the Karmanos Cancer Center. Material financial hardship included experiencing a decrease in
income, borrowing from friends or family, remaining in debt, or accessing existing assets
(refinancing/selling a home, selling investments, withdrawing money from savings or
retirement) due to cancer. Limiting care included skipping doses of medication to save money
or refusing treatment or needing to see a doctor but not going due to cost. Logistic regression
tested the association between material financial hardship and limiting care, adjusting for age,
sex, race, marital status, income, education, health insurance, employment, cancer site, and
treatments received. Effect modification by race was assessed.

Results: Overall, 49% of participants experienced material financial hardship, and 19%
reported limiting care due to cost. In adjusted analyses, survivors who experienced material
financial hardship were more than six times as likely to limit care (OR: 6.6, 95% CI: 3.7, 11.5).
This association was strongest for participants who borrowed money from friends or family
(OR: 8.8, 95% CI: 6.6, 16.8) and remained in debt (OR: 5.6, 95% CI: 3.4, 9.1); however,
accessing existing assets (OR: 2.9, 95% CI: 1.6, 5.2) and experiencing a decrease in income
(OR: 2.8, 95% CI: 1.7, 4.5) were also strongly associated with limiting care due to cost.
Marginal effect modification was detected by race, suggesting that the association between
any material financial hardship and limiting care may be stronger among black (OR: 10.0, 95%
CI: 4.8, 21.0) than white survivors (OR: 3.5, 95% CI: 1.3, 9.6; Pinteraction=0.10).
S1638 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusion: Cancer survivors who experienced material financial hardship were much more
likely to report limiting care due to cost than those who did not. Left unaddressed, the impact
of material financial hardship may exacerbate health disparities and lead to worse outcomes
across U.S. cancer survivor populations.

CORRESPONDING AUTHOR: Theresa A. Hastert, Ph.D., M.P.P., Wayne State University School
of Medicine/Karmanos Cancer Institute, Detroit, MI, 48201; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1639

Paper Session 33: Cancer in Underserved Populations 3:55 PM-4:12 PM

CHANGE IN SOCIAL SUPPORT IN AFRICAN AMERICAN BREAST CANCER PATIENTS AFFECTS

DEPRESSED MOOD AND GENERAL HEALTH

Tess Thompson, PhD, MPH1, Maria Prez, MA2, Matthew Kreuter, PhD3, Julie Margenthaler,
MD2, Graham Colditz, MD, DrPH2, Donna Jeffe, PhD4

1
Washington University in St. Louis, Webster Groves, MO; 2Washington University School of
Medicine, St. Louis, MO; 3Washington University in St. Louis, St. Louis, MO; 4Washington
University School of Medicine, Saint Louis, MO

Higher perceived social support is associated with better psychological and quality of life
outcomes in breast cancer patients. Less is known about how changes in social support affect
these outcomes in patients. We interviewed 227 African American women with non-
metastatic breast cancer (72% early stage; 58% annual income < $25,000; 27%
married/partnered; mean age 56, range 33-81 years) in a randomized controlled trial of a
culturally tailored intervention using breast cancer survivor stories. Demographic and
psychosocial interview data were collected at baseline, 1 month later, and then 6 months, 1
year, and 2 years after definitive surgical treatment. Mean Center for Epidemiologic Studies-
Depression Scale (CES-D) score was 11.9 (SD = 11.4) at baseline and 11.1 (SD = 12.1) at 2
years; mean RAND 36-item general health subscale score was 59.1 (SD = 22.4) at baseline and
54.7 (SD = 22.9) at 2 years. An unconditional growth curve model estimated the intercept
(estimated starting point) and slope (change over time) of perceived social support (19-item
Medical Outcomes Study Social Support Survey) from baseline through 1-year follow-up after
definitive surgical treatment. Two separate growth curve models then used this intercept and
slope of social support over the first year after surgery to predict depressed mood and
perceived general health 2 years after surgery. Models that controlled for the baseline
measure of the outcome of interest showed good fit (CFI/TLI > .99, RMSEA .01, SRMR = .04).
Patients who had lower estimated starting points of social support or experienced declines in
social support in the first year after definitive surgical treatment reported higher depressed
mood and lower perceived general health at 2-year follow-up. These results held after
controlling for demographic and cancer-related variables. Thus, declines in social support in
the first year after a breast cancer diagnosis had negative consequences a year later for
African American women in terms of mental and general health. Screening for availability of
social support at diagnosis and monitoring support levels during and after treatment could
help patients with low or declining social support receive support services and psychological
intervention as needed.
S1640 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Tess Thompson, PhD, MPH, Washington University in St. Louis,
Webster Groves, MO, 63119; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1641

Paper Session 33: Cancer in Underserved Populations 4:13 PM-4:30 PM

PSYCHOLOGICAL DETERMINANTS OF EHEALTH ACTIVITY AMONG AFRICAN AMERICAN AND

WHITE CANCER SURVIVORS: A NEW APPLICATION OF THEORY

Nicole Senft, PhD1, Judith Abrams, PhD2, Jennifer L. Beebe-Dimmer, Ph.D., M.P.H.3, Charity A.
Barnes, B.S.4, Anne Katz, PhD5, Ke Zhang, PhD4, Tara Eaton, PhD6, Deborah H. Charbonneau,
PhD4, Elisabeth Heath, MD7, Hayley S. Thompson, Ph.D.8

1
Karmanos Cancer Institute/Wayne State University, Ann Arbor, MI; 2Wayne State
university/Karmanos Cancer Institute, Detroit, MI; 3Wayne State University School of
Medicine/Karmanos Cancer Institute, Detroit, MI; 4Wayne State University, Detroit, MI;
5
Wayne State University/Karmanos Cancer Institute, Detroit, MI; 6Levine Cancer Institute,
Charlotte, NC; 7Karmanos Cancer Institute, Detroit, MI; 8Karmanos Cancer Institute - Wayne
State University, Detroit, MI

eHealth is a promising resource for cancer survivors coping with their physical, psychosocial,
and economic needs, and it may play an especially important role in reducing racial disparities
in cancer survivorship. The current research is a preliminary examination of associations
between beliefs about eHealth and reported eHealth activity. This is part of an ongoing study
using a mixed-method design to assess eHealth activity among African American and white
cancer survivors in the metropolitan Detroit area. The study assesses 17 eHealth activities
spanning 5 categoriesinformation seeking, connecting with informal support, connecting
with healthcare providers, managing health, and medical transactions. We drew from the
Unified Theory of Acceptance and Use of Technology, which suggests the adoption of new
technology can be predicted by beliefs about its ease of use, usefulness, and social
acceptability. We adapted the model to relate to use of technology for health-related reasons
and extended it to include belief in the security of online health services. In a sample of
African American (n=53) and white (n=34) survivors, we tested whether attitudes about ease
(3-items, =.80), usefulness (3-items, =.94), social acceptability (4-items, =.77), or security
(3-items, =.85) of eHealth were associated with eHealth activity, controlling for social-
structural determinants (i.e. age, education, access to internet). The sample was comprised of
primarily older adults (mean age=59, SD=10.37) and survivors of breast cancer (n =67). There
was significantly less eHealth use among African American (M=4.37, SD=3.13) than white
survivors (M=7.97, SD=3.19; t(83)=5.14, p < .001). The belief that eHealth was useful was
positively associated with eHealth activity among both African American (r=.36, p=.02) and
white survivors (r=.36, p=.05). Belief that eHealth was socially acceptable was associated with
increased activity for African American (r=.28, p=.05), but not white survivors (r=.16, p=.41).
S1642 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Beliefs in the ease and security of eHealth were not associated with use for members of either
group (rs < .10, ps>.10). These results suggest interventions aimed at adjusting attitudes about
eHealth may be effective in encouraging the adoption and use of this promising resource.
Furthermore, the particular attitudes associated with eHealth activity differ across racial
groups, suggesting targeted interventions may be most effective in encouraging eHealth use
among cancer survivors.

CORRESPONDING AUTHOR: Nicole Senft, PhD, Karmanos Cancer Institute/Wayne State

University, Ann Arbor, MI, 48103; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1643

Paper Session 34: Novel Tools to Support Digital Health 3:00 PM-3:18 PM

DEVELOPING AUDIENCE PERSONAS THROUGH WEB ANALYTICS TO INFORM HEALTH

PROMOTION AND OUTREACH

Brian Keefe, MA1, Paula Novacki, B.A - Journalism & Mass Communications2, Kisha I. Coa, PhD,
MPH1, Augustson Erik, PhD, MPH3

1
ICF International, Rockville, MD; 2ICF, Washington, DC; 3National Cancer Institute, Rockville,
MD

Introduction: One of the greatest health communication challenges is understanding target

audiences and developing content that meets their needs. In an environment where acquiring
detailed marketing data is often cost-prohibitive and secondary data is often too broad,
gaining detailed insights about the target audience can be difficult. The goal of this study was
to systematically develop audience personas using website usage, demographic, and
engagement data from Google Analytics to describe users of Smokefree.gov. Smokefree.gov is
the National Cancer Institutes digitally based smoking cessation program with a presence on
a number of different platforms (e.g., web, social, text, app). Persona development is a
foundational activity for digital health programs. When appropriately established, personas
represent key audience segments which can be used to inform promotional decisions,
optimization or content for specific browsers or devices, tailoring content, and overall
outreach activities.

Methods: The current project utilized custom development work within Google Analytics to
identify a number of user characteristics, including age, gender, geographic location, content
needs, device preference, and browser preference, in order to build specific Smokefree
personas. Data was collected from January 1, 2016-August 1, 2016, encompassing over 1.1
million visitors to Smokefree.gov.

Results: We identified 4 unique types of Smokefree.gov personas: 1) General website users, 2)

Users entering the site through social media, 3) Users who downloaded the Smokefree mobile
applications, and 4) Users entering the site through the SmokefreeTXT program. These
persona groups differed in a number of ways, mainly their engagement levels with the
website and the devices used to access the content. For example, Group 1 General website
users were younger (age 18-24 (29%)), found the site through an organic search using terms
such as how to quit smoking, and visited an average of 1.23 pages, with the nicotine
withdrawal page being the most popular. In contrast, Group 4 visitors from SmokefreeTXT
were slightly older (25-34), entered the site using an iPhone, landed on the QuitGuide mobile
app page, and visited an average of 2.14 pages.
S1644 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusion: Personas such as these will enable us to create smoking cessation content that
meets the specific needs of various users, is relevant to their lived experiences, and provides
the means to better target outreach activities moving forward. Specifically, Smokefree can
highlight specific resources and content on different channels based on findings from the
personas, re-target promotional efforts to users who more align with the personas, and create
new content based on the characteristics of various user types.

CORRESPONDING AUTHOR: Brian Keefe, MA, ICF International, Rockville, MD, 20850;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1645

Paper Session 34: Novel Tools to Support Digital Health 3:19 PM-3:36 PM

DEVELOPMENT OF THE MOBILE PHONE AFFINITY SCALE: ASSESSING PERSONS RELATIONSHIP

TO THEIR MOBILE PHONE

Beth C. Bock, PhD1, Ryan Lantini, MA2, Herpreet Thind, PhD, MPH, MBBS3, Kristen Walaska,
BS1, Rochelle K. Rosen, PhD4, Joseph L. Fava, PhD5, Lori A. J. Scott-Sheldon, PhD6

1
Brown Medical School, Providence, RI; 2The Miriam Hospital, Providence, RI; 3University of
Massachusetts Lowell, Lowell, MA; 4Brown University School of Public Health, Providence, RI;
5
Centers for Behavioral and Preventive Medicine, The Miriam Hospital, Providence, RI, 02903,
Providence, RI; 6The Miriam Hospital and Brown University, Providence, RI

Background: Existing instruments that assess an individuals relationship with mobile phones
tend to focus on negative constructs such as addiction or dependence and appear to
assume that high mobile phone and technology use reflects pathology. Since mobile phones
can be beneficial for health behavior change, disease management, work productivity and
social connections, there is a need for an instrument that provides a more balanced
assessment of the various aspects of individuals relationship to their mobile phones.

Objective: The purpose of this research is to develop, revise, and validate the Mobile Phone
Affinity Scale (MPAS), a multi-scale instrument designed to assess key factors associated with
mobile phone use.

Methods: Participants (N = 1058; mean age = 32.5, SD=10.3; 50% female) were recruited in a
national sample using Amazon Mechanical Turk (MTurk) to complete a survey that assessed
participants attitudes and use patterns regarding their mobile phone, as well as assessments
of anxiety (State-Trait Anxiety Inventory), depressive symptoms (Centers for Epidemiological
Studies Depression Scale), impulsivity (Barratt Impulsiveness Scale ) and resilience (Brief
Resilience Scale).

Results: Confirmatory factor analysis was used to develop a six-factor model. The final model
fit well (RMSEA = .059, CFI = .941, TLI = .931, SRMR = .042), and consisted of 24 items (4 items
each) measuring six factors: Connectedness, Productivity, Empowerment, Anxious
attachment, Addiction, and Continuous use. The subscales demonstrated strong internal
consistency (coefficient alpha range = 0.76 to 0.88, M = 0.83), and high item factor loadings
(range = 0.57 to 0.87, M = 0.75). Validity analyses further demonstrated support for the
individual subscales. For example, Attachment and Addiction subscales were significantly and
S1646 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

positively correlated with symptoms of depression and anxiety, and correlated negatively with
resilience, while Productivity was positively correlated with resilience.

Conclusions: The Mobile Phone Affinity Scale (MPAS) is a reliable, valid assessment of both
positive and negative characteristics associated with the individuals relationship to their
mobile phone. Mobile phone affinity may have an important impact on the efficacy and
effectiveness of mobile health interventions. However, research is needed to assess its
predictive ability in health behavior change interventions delivered through mobile phones.

CORRESPONDING AUTHOR: Beth C. Bock, PhD, Brown Medical School, Providence, RI, 02906;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1647

Paper Session 34: Novel Tools to Support Digital Health 3:37 PM-3:54 PM

E-EMBRACERS, WARY WEARERS, AND DATA DODGERS: PROFILES OF DIABETES DEVICE USERS

Molly Tanenbaum, PhD1, Esti Iturralde, Ph.D.2, Regan C. Barley, BS2, Sarah Hanes, BA1, Hillary
Kroll, N/A3, Korey K. Hood, Ph.D.2

1
Stanford University, Palo Alto, CA; 2Stanford University School of Medicine, Palo Alto, CA;
3
Stanford University, East Brunswick, NJ

Background: Diabetes devices (insulin pumps, continuous glucose monitors-CGM) are

associated with improved health outcomes and quality of life in adults with type 1 diabetes
(T1D). Rates of device uptake, particularly CGM, are low. This study aimed to understand
profiles of adults with T1D in relation to device use and attitudes to ultimately inform tailored
interventions to address barriers and increase device uptake among each group.

Methods: Participants were 1503 T1D Exchange participants (MSD age 35.314.8 yrs; dx
duration 20.412.5 yrs) who completed surveys on barriers to uptake, diabetes distress, and
technology attitudes. Clinical data available from the T1D Exchange included device use and
health outcomes (glycemic control). K-means cluster analyses were used to group the sample
by barriers to device use and technology attitudes. ANOVAs and chi-square tests assessed
differences among groups by demographic and psychosocial variables (e.g. hypoglycemia
worry, diabetes distress).

Results: Cluster analyses yielded four distinct profiles. The e-Embracers (n=712) endorse few
barriers to device use and have the highest rates of device use, lowest levels of distress, and
lowest A1c (MSD=7.31.1). Data Dodgers (n=170) have the lowest CGM use. The barrier that
most differentiated them was not wanting more diabetes data. Off-the-Grid-ers (n=444) have
the most negative technology attitudes and endorse average levels of barriers to device use.
Wary Wearers (n=177) have the lowest overall device use, are the youngest, and have
shortest T1D duration. They do not want attention or daily hassle of devices, are most
distressed and worried about hypoglycemia, and have highest average A1c (MSD=81.6).
Cost, a commonly endorsed barrier to using devices, did not distinguish among groups.

Conclusion: Off-the-Grid-ers may benefit from addressing attitudes toward diabetes

technology; Data Dodgers may benefit from increasing comfort with handling diabetes-
related information. Wary Wearers may benefit from comprehensive psychosocial
intervention targeting diabetes distress and hypoglycemia worries along with issues related to
S1648 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

wearing devices and having other people notice them. These profiles will help clinicians to
identify adults with T1D needing each intervention to ultimately increase device uptake.

CORRESPONDING AUTHOR: Molly Tanenbaum, PhD, Stanford University, Palo Alto, CA, 94304;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1649

Paper Session 34: Novel Tools to Support Digital Health 3:55 PM-4:12 PM

HABIT APP: FEASIBILITY OF A WEIGHT LOSS PROBLEM SOLVING APP

Sherry Pagoto, PhD1, Jessica Oleski, MA1, Molly E. Waring, PhD1, Emmanuel Agu, PhD2,
Bengisu Tulu, PhD3

1
University of Massachusetts Medical School, Worcester, MA; 2Worcester Polytechnic Inst,
Worcester, MA; 3Worcester Polytechnic Institute, Worcester, MA

A behavioral strategy employed in virtually every visit of weight loss counseling but not yet
included in weight loss mobile apps is problem solving. Problem solving is a process used to
identify barriers to a goal and generate solutions to be iteratively attempted until barriers are
overcome. We developed an app that employs Nezus problem solving algorithm to produce
expert-derived solutions to diet and exercise problems. Each problem solving session on the
app produces a list of solutions or habits. The user selects a habit and then schedules
reminders to execute the habit during the week. We performed a pilot study to evaluate the
feasibility of the Habit app.

Adults who were overweight or obese were enrolled in an 8-week intervention that included
the Habit app and support via a private Facebook group. Participants were instructed to use
Habit app weekly to identify 1-2 habits to work on. In the Facebook group, a counselor
emphasized the importance of trying new habits to overcome weight loss challenges, asked
them to share weekly habit goals and accomplishments, and encouraged dietary self-
monitoring. Problem solving skills were measured with the Social Problem Solving Inventory-R
at baseline and follow-up.

Participants (N=27) were on average 37.2 (sd=11.6) years old, baseline BMI of 31.1 (sd=4.6),
67% female, and 85% non-Hispanic white. Three participants dropped out (88% retention). Of
the 41 diet and exercise challenges identified by participants, only 3 were not addressed in
the app. Participants reported using the Habit app on an average of 7.1 weeks (sd=1.7). Higher
baseline positive problem solving scores were associated with greater use of the app (r= .414,
p=0.04), but no changes were observed in positive problem solving over 8 weeks (p=0.27)
Participants lost on average 0.20% (sd=0.03%) per week (p=0.004), comparable to 0.28% in
the Diabetes Prevention Program Lifestyle Intervention.

The Habit app addressed the majority of participants weight loss challenges, was used on
most weeks, but did not improve problem solving skills. In an ongoing pilot, Habit app is being
delivered with brief problem solving training to better impact problem solving skills. Data
from the ongoing pilot will also be presented.
S1650 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Sherry Pagoto, PhD, University of Massachusetts Medical School,

Worcester, MA, 01655; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1651

Paper Session 34: Novel Tools to Support Digital Health 4:13 PM-4:30 PM

EMBRACING ITERATION GUIDED BY MIXED METHOD EVALUATION DURING MHEALTH

CLINICAL TRIALS

Angela Fidler Pfammatter, PhD1, Gwendolyn Ledford, BA1, Susan Hood, PhD 2, Sean Arca,
B.S.3, Shirlene Wang, BA4, Bonnie Spring, PhD2

1
Northwestern University Feinberg School of Medicine, Chicago, IL; 2Northwestern University,
Chicago, IL; 3Northwestern Medicine, Chicago, IL; 4Northwestern University Feinberg School
of Medicine, Wilmette, IL

mHealth strategies are increasingly used in clinical intervention research, yet much remains
unknown regarding optimization of engagement in interventions delivered by mobile phones.
Although formative work is often involved in intervention design, many interventions and
smartphone applications are not modified based on feedback after launch of a traditional RCT.
As previous research has supported, mHealth app use declines 40-50% over time. To
reasonably expect long term engagement, researchers must identify ways to initiate and
sustain engagement of the app user. Concepts from design research may be used to evaluate
engagement, receive feedback, and make modifications that positively affect engagement and
thus, outcomes. The NUYou Study enrolled 150 freshmen students, in its first year, for a 2-
year cardiovascular health promotion study. The purpose of this work was to systematically
evaluate recruitment and engagement strategies to iterate key trial components, like the
NUYou smartphone application, push notification recommendation system, and social media
groups. Potential participants not enrolled in the trial (n=365) were surveyed about the
effectiveness of advertisement and initial engagement initiatives. Enrolled participant (n=150)
app use was continuously monitored via app log data 6 months after recruitment opened, and
users were interviewed during the trial. While 98% of potential participants recognized and
recalled aspects of the study, only 7% of the total target population enrolled in the trial.
Significant barriers included lack of time and being too busy. App monitoring indicated that
although 14% of the sample used the app consistently, great variability existed in the amount
and length of continued use of the application, pointing to issues with sustained engagement.
Surveys found that although 55% indicated agreement that the app was relevant to their
health goals, 21% indicated agreement that the app helped achieve goal attainment.
Qualitative interviews revealed that students generally liked reminding features, that the app
was easy to use, and that further personalization could help them reach health goals. Results
were applied to inform further iteration of features to optimize engagement in the
intervention and the smartphone application. Specifically, small daily incentives, additional
tracking of stress, happiness, and cognitive performance, and dynamic data visualization were
S1652 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

added to the trial for continuing and new participants in year 2. Opportunities to maximize
learning and optimization during mHealth clinical trials are essential to behavior science and
as such, potential methods and impact are discussed.

CORRESPONDING AUTHOR: Angela Fidler Pfammatter, PhD, Northwestern University Feinberg

School of Medicine, Chicago, IL, 60611; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1653

Paper Session 35: Stress, Pain and Fatigue in Cancer Survivors 3:00 PM-3:18 PM

MINDFULNESS-BASED STRESS REDUCTION FOR FATIGUED CANCER SURVIVORS: EFFECTS ON

MINDFULNESS, ACCEPTANCE, AND SELF-COMPASSION

Kelly Chinh, B.A.1, Catherine E. Mosher, Ph.D.1, Linda F. Brown, PhD2, Kathleen Beck-Coon,
MD3, Kurt Kroenke, M.D.4, Shelley A. Johns, PsyD, ABPP4

1
Indiana University-Purdue University Indianapolis, Indianapolis, IN; 2Indiana University
School of Medicine, Bloomington, IN; 3Indiana University School of Nursing, Indianapolis, IN;
4
Indiana University School of Medicine, Indianapolis, IN

Mindfulness-based stress reduction (MBSR) has resulted in reduced physical and psychological
symptoms in cancer populations, but little is known about theory-based psychological
processes through which mindfulness-based interventions may decrease symptoms. In the
present trial of MBSR for fatigued cancer survivors, mindfulness, acceptance, and self-
compassion were psychological processes theorized to play a role in symptom outcomes. This
trial randomized 35 cancer survivors with clinical levels of cancer-related fatigue (CRF) to
either a 7-week MBSR intervention for CRF or a waitlist control condition. We hypothesized
that the MBSR group would show increases in one aspect of mindfulness (i.e., non-reactivity),
acceptance, and self-compassion relative to controls.

The majority of participants were White (80%), female (94%), and college educated (71%)
with an average age of 57 years (SD = 9.3). Participants were primarily diagnosed with early-
stage breast cancer (80%), and all had completed cancer treatment at least 9 months prior to
enrollment. All 35 participants completed assessments of mindfulness, acceptance, and self-
compassion at baseline, immediately post-intervention, and 1-month follow-up.

Three linear mixed models were conducted to examine main effects of time, study group, and
the time x group interaction on mindfulness, acceptance, and self-compassion. A significant
effect of study group in favor of MBSR was found for non-reactivity (F(1,33)=9.46, p=.004) and
acceptance (F(1,33)=5.75, p=.022). A significant time x group interaction (F(2,66)=5.80,
p=.005) was found for self-compassion, such that the MBSR group showed a steady increase
in self-compassion, whereas the control groups scores remained relatively stable.

Findings suggest that MBSR for fatigued cancer survivors leads to increased levels of
mindfulness, acceptance, and self-compassion. Further research is needed to examine
whether these psychological processes mediate the effects of MBSR on symptom outcomes
during cancer survivorship.
S1654 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Kelly Chinh, B.A., Indiana University-Purdue University

Indianapolis, Indianapolis, IN, 46202; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1655

Paper Session 35: Stress, Pain and Fatigue in Cancer Survivors 3:19 PM-3:36 PM

CITATION AND MERITORIOUS AWARD WINNER

CANCER-SPECIFIC STRESS AND ABSOLUTE LYMPHOCYTE COUNT TRAJECTORIES IN PATIENTS

WITH CHRONIC LYMPHOCYTIC LEUKEMIA

David M. Weiss, M.A.1, Barbara L. Andersen, PhD1, Amy J. Johnson, PhD1, John C. Byrd, MD2,
Kami J. Maddocks, MD1

1
The Ohio State University, Columbus, OH; 2Ohio State University, Columbus, OH

Chronic stress has been commonly observed in cancer patients and is associated with immune
system down regulation. The effect of stress on immunity in hematologic cancers such as
chronic lymphocytic leukemia (CLL) has not been studied despite the role of immune system
dysfunction in CLLs pathogenesis. In a phase II clinical trial, 154 patients with
relapsed/refractory CLL received ibrutinib and provided blood samples and completed a self-
report measure of psychological stress specific to cancer over an 18-month treatment period
(nine assessments). Targeted treatments like ibrutinib have been effective in reducing disease
progression in CLL despite the occurrence of lymphocytosis, which is an increase in absolute
lymphocyte counts (ALC) and has previously been regarded as a marker of active disease.
Controlling for demographic, health status, number of prior treatments, and CLL genetic risk
(del17p) factors, random changepoint models were estimated to evaluate the impact of stress
on ALC trajectories. Stress was associated with pretreatment ALCs (0 = 0.13; 95% CI = 0.02,
0.25) but did not impact the timing of lymphocytosis (x1= 0.03, CI = -0.15, 0.22), or the ALC
trajectory before (x1 = -0.11, CI = -0.23, 0.01) and after (x1=0.10, -0.01, 0.22) lymphocytosis.
Stress affects pre-treatment ALC such that a 10% increase in stress results in a 1.3% increase
in ALC. However, stress has little impact on ALC trajectories after beginning drug therapy.
Additional analysis showed that lymphocytosis occurs later in the treatment trajectory for
individuals classified as drug non-responders compared to responders at 18-months (x2 = -
0.89; CI= -1.48, -0.30). However, there are a variety of ways a patient can exhibit
lymphocytosis (e.g. lymphocytosis before ibrutinib administration or treatment-induced
lymphocytosis) and it is unclear whether the mechanism of lymphocytosis or specific ALC
trajectories can predict treatment response. Overall, results suggest that while stress
increases pre-treatment ALC, patients exhibit similar ALC trajectories after treatment
initiation. Future trials may assess whether clinical, demographic, or other psychological risk
factors can predict ALC trajectories or treatment responses. The current study provides a
S1656 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

novel way to provide knowledge of how psychological stress impacts disease trajectories in
patients undergoing a new cancer therapy.

CORRESPONDING AUTHOR: David M. Weiss, M.A., The Ohio State University, Columbus, OH,
43206; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1657

Paper Session 35: Stress, Pain and Fatigue in Cancer Survivors 3:37 PM-3:54 PM

OMEGA-3 VERSUS OMEGA-6 SUPPLEMENTATION FOR CANCER-RELATED FATIGUE:

EXAMINING INFLAMMATORY AND ANTI-OXIDANT PATHWAYS

Luke J. Peppone, PhD, MPH1, Karen Mustian, PHD, MPH, MS2, Charles Kamen, PhD, MPH2,
Matthew Asare, PhD, MBA, CHES3, Calvin L. Cole, PhD3, Sarah Kerns, PhD MPH2, Michelle
Janelsins, PhD, MPH2

1
University of Rochester, Wilmot Cancer Institute, Rochester, NY; 2University of Rochester
Medical Center, Rochester, NY; 3University of Rochester, Rochester, NY

Background: Cancer-related fatigue (CRF), one of the most troublesome side effects of cancer
and its treatment, is associated with increased levels of inflammation. We conducted a
nationwide, multi-site, phase II RCT through the URCC NCORP Research Base to examine the
effect of omega-3 supplementation (-3) from fish oil versus omega-6 supplementation (-6)
from soybean oil for improving CRF. -3 was chosen for its anti-inflammatory properties while
-6 was chosen as a control condition.

Methods: Breast cancer survivors between 4-36 months post-adjuvant therapy with a CRF
level of 4 or greater (on a 0-10 scale) were stratified by baseline CRF (4-6: moderate and 7-10:
high) and randomized into 3 arms: 1) High-dose -3 (-3; 6 g/day), 2) Low-dose -3/Low-
dose -6 (-3/-6; -3: 3 g/day and -6: 3 g/day) and 3) High-dose -6 (-6; 6 g/day) for 6
weeks. CRF was assessed pre- and post-intervention via the Symptom Inventory (SI). Serum
protein and RNA levels of inflammatory (PGE2, sTNFRI, sTNFRII, and CRP) and anti-oxidant
(SOD2) markers were also measured at pre- and post-intervention. ANCOVAs assessed
associations between changes in CRF and biomarker levels.

Results: 108 female breast cancer survivors were accrued (93% Caucasian, mean age = 59.8).
Mean baseline CRF levels did not differ between groups and serum levels of fatty acids
confirmed high compliance in all study arms with minimal contamination. After controlling for
baseline CRF, -6 supplementation significantly improved CRF in a dose-response manner (SI
Change score: -3 = -2.5 vs. -3/-6 = -2.1 vs. -6 = -0.9; Cohens =0.72; p=0.02). -3
supplementation failed to significantly reduce markers of inflammation, potentially due to low
baseline levels of inflammation. RNA levels of SOD2 significantly increased in the -6 group
(Cohens =0.48; p=0.02), while there was a trending increase in SOD2 protein levels for the
-6 group compared to the -3 group (Cohens =0.22; p=0.17). RNA levels of PGE2 declined
significantly (Cohens =0.78; p0.01) for the -6 group, but there was no change in PGE2
protein levels (p=0.21).
S1658 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions: -6 supplementation from soybean oil significantly reduced CRF compared to -

3 supplementation in breast cancer survivors. These findings are contrary to our hypothesis
and provide new information on the mechanisms underlying CRF. -6 supplementation
resulted in a subtle anti-inflammatory effect, with a more pronounced anti-oxidant effect,
providing a potential mechanism of action. Future studies are needed to confirm these
findings. Funding: NCI UG1CA189961 & R03CA175599. Nordic Naturals, Inc. supplied all study
agents.

CORRESPONDING AUTHOR: Luke J. Peppone, PhD, MPH, University of Rochester, Wilmot

Cancer Institute, Rochester, NY, 14625; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1659

Paper Session 35: Stress, Pain and Fatigue in Cancer Survivors 3:55 PM-4:12 PM

THE ROLE OF ILLNESS PERCEPTIONS IN CANCER-RELATED FATIGUE OUTCOMES IN OVARIAN

CANCER PATIENTS

Crystal Hare, BA, BSc1, Lindsey Torbit, MA1, Cassandra Crangle, MA1, Sarah Ferguson, MD2,
Tae Hart, PhD1

1
Ryerson University, Toronto, ON, Canada; 2Princess Margaret Cancer Centre, Toronto, ON,
Canada

Background: Cancer-related fatigue (CRF) is one of the most common and debilitating
symptoms reported by cancer patients (Jensen et al., 2011). Factors implicated in the severity
of CRF include cancer treatment, age, and pain (e.g., Anderson & Hacker, 2008). Research
suggests that ovarian cancer (OC) patients may be at greater risk for the development of CRF,
compared to other cancer populations (Payne, 2002), yet little research has examined the
factors contributing to CRF in OC. One factor that may contribute to fatigue variability is the
impact of patient illness perceptions (IPs; Leventhal et al., 1980) related to the diagnosis of
cancer and treatment. Using Leventhals Common Sense Model (CSM), this study examined
the role of IPs (i.e., illness identity, timelinechronicity and cyclicality, consequences,
personal and treatment controllability, and coherence) on fatigue severity in patients with OC.

Method: Women with OC (N=265) were recruited from the Princess Margaret Hospital in
Toronto. Participants completed the following self-report questionnaires: Illness Perception
Questionnaire-Revised (Moss-Morris et al., 2002); Brief Pain Inventory (Cleeland & Ryan,
1994); and Functional Assessment of Chronic Illness Therapy Fatigue Scale (Cella et al., 2005).
Controlling for three factors commonly associated with CRF: age, cancer treatment, and pain
severity, a two-step hierarchical regression examined the impact of IPs on fatigue severity.

Results: Step one of the hierarchical regression was significant, F(4,184) = 21.12, p < .001; only
pain severity ( = -.43, p < .001) was significantly associated with more fatigue severity. The
overall model in step two was significant, F(10,178) = 12.65, p < .001. Out of the six IPs
included in the model, greater endorsement that cancer is cyclical in nature ( = -.13, p < .05),
and endorsement of significant consequences associated with OC ( = -.22, p < .001) were
associated with greater fatigue severity, predicting an additional 10.1% of the variance in
fatigue, F(6,178) = 5.12, p < .001. The total model accounted for 38.2% of the variance in
fatigue severity.

Discussion: This is the first examination of the full array of Leventhals illness perceptions in
association with CRF. Believing that ones cancer has serious negative consequences on ones
S1660 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

life and that it is likely to recur appears to be associated with intensified fatigue. These data
provide partial support for the CSM in understanding CRF.

CORRESPONDING AUTHOR: Crystal Hare, BA, BSc, Ryerson University, Toronto, ON, M5B 2K3;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1661

Paper Session 35: Stress, Pain and Fatigue in Cancer Survivors 4:13 PM-4:30 PM

A RANDOMIZED PILOT TRIAL OF AN MHEALTH PAIN COPING SKILLS TRAINING INTERVENTION

FOR HEMATOPOIETIC STEM CELL TRANSPLANT PATIENTS

Caroline Dorfman, Ph.D.1, Sarah Kelleher, Ph.D.1, Rebecca Shelby, Ph.D.1, Hannah Fisher,
B.S.2, Francis Keefe, Ph.D.1, Krista Rowe, RN, MSN, AOCNS3, Tamara Somers, Ph.D.1

1
Duke University Medical Center, Durham, NC; 2University of Miami, Miami, FL; 3Duke
University Health System, Durham, NC

Pain is common and persistent among patients following Hematopoietic Stem Cell Transplant
(HSCT) and associated with decreased quality of life. Current behavioral cancer pain
interventions have limited reach among HSCT patients who, after weeks of intense medical
care, are discharged home often far from the hospital. To address the unique needs of these
patients, we developed a home-based mobile pain coping skills training (mPCST) protocol. A
small RCT was conducted to test the feasibility and acceptability of the protocol and examine
intervention-related changes in pain, pain self-efficacy, pain disability, fatigue, and physical
disability. Participants were randomized to mPCST (n=18) or usual care (n=18). mPCST bridged
hospitalization (1 session) and home (5 video-conferencing sessions) using CBT skills (e.g.,
relaxation, cognitive-restructuring, activity pacing/planning); a website with personalized
messages was used. Participants completed assessments at pre- and post-treatment. mPCST
participants completed M=4.9/6 sessions. Skill use increased over time, and participants
found sessions to be helpful, easy to understand, highly acceptable, and rated the
intervention to be good or excellent overall. mPCST participants evidenced significant
improvements in pain self-efficacy and the two-minute walk test; the usual care arm did not.
Both groups experienced significant improvements in pain disability and fatigue; however, the
magnitude of the effect sizes were greater for the mPCST arm. Participants in both arms
evidenced significant improvements on a self-report measure of physical disability. There
were no significant changes in pain. The results demonstrate that an mPCST protocol tailored
to meet the needs of HSCT patients is feasible and acceptable. Many HSCT patients have pain
that varies over time; patients who improve their pain coping skills may be better able to
manage pain and decrease pain-related disability and interference. Psychosocial interventions
that increase pain self-efficacy and decrease pain disability may lead to improved outcomes
for HSCT patients. Longitudinal studies are necessary to examine the long-term implications of
the observed improvements in pain self-efficacy on patient outcomes.

CORRESPONDING AUTHOR: Caroline Dorfman, Ph.D., Duke University Medical Center,

Durham, NC, 27705; [email protected]
S1662 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 36: Translation of Interventions in Disease Prevention and Management 3:00
PM-3:18 PM

TRANSLATING RESEARCH INTO PRACTICE FOR HEALTH: A COMMUNITY-BASED, LIFESTYLE

INTERVENTION FOR ADULTS WITH TYPE 2 DIABETES

Emily A. Finch, MA, David Liss, PhD, Ronald T. Ackermann, MD, MPH

Northwestern University, Chicago, IL

Background: At least 1 out of 3 people will develop type 2 diabetes in their lifetime and risk of
death is 50% higher for adults with diabetes than for adults without diabetes. The LookAHEAD
study of over 5,000 overweight patients with type 2 diabetes showed that participants in the
intervention program lost more weight and had improvements in urinary incontinence, sleep
apnea, depression, quality of life, physical functioning, and mobility. However, the Look
AHEAD intervention is not currently available on a population scale, because the approach is
prohibitively costly and difficult to translate into real-world settings.

Methods: Working with primary-care providers between January 2013 and July 2014, we
enrolled 331 adults with BMI 24 kg/m2 and prior diagnosis of type 2 diabetes from
two ethnically diverse Chicago area neighborhoods into the study. Participants were
randomized to: (1) brief lifestyle counseling with nearby community resource information; or
(2) brief lifestyle counseling plus encouragement to participate in a group-based adaptation of
the LookAHEAD program offered in both English and Spanish languages, free-of-charge by the
YMCA. The primary outcome was mean difference in body weight at 12 months. Secondary
outcomes included changes in HbA1c, cholesterol, and blood pressures. Interim intent-to-
treat (ITT) analyses accounted for clustering of longitudinal patient-level data and imputed
missing follow-up data.

Results: At baseline, participants had a mean (SD) age of 57.4 11.4 years, BMI 35.6 7.6
kg/m2, HbA1c 7.22 1.25%. Approximately half (50.2%) were women, 30.5% were African
American, 34.4% were non-Hispanic White, and 27.5% were Hispanic. Approximately one
third (38.9%) reported annual household incomes of < $25,000. Overall, 75 participants
(45.7%) participated in at least 1 intervention session. In ITT analysis, the average effect of
randomization to GLI encouragement was -1.58 pounds per six months (95% CI, -2.83 to -0.34;
P=0.01). Over the 12-month follow-up period, this effect equates to an additional 3.16 pounds
of weight loss. Mean weight losses of non-Hispanic white, African American, and Hispanic
participants were not statistically significantly different.

Conclusions: Delivery of a group adaptation of the LookAHEAD program by YMCAs suggests a

ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1663

scalable lower cost mechanism to help a diverse, predominantly low-income adults with type
2 diabetes lose a modest amount of weight. Health care reform models and reimbursement
policies should include access to community-based intensive healthy lifestyle interventions to
translate interventions into practice.

CORRESPONDING AUTHOR: Emily A. Finch, MA, Northwestern University, Chicago, IL, 60611;
[email protected]
S1664 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 36: Translation of Interventions in Disease Prevention and Management 3:19
PM-3:36 PM

CITATION AWARD WINNER

CAN A PEDOMETER-BASED WALKING PROGRAM LOWER HEALTH CARE COSTS AMONG

ADULTS WITH TYPE 2 DIABETES?

Mona AuYoung, PhD, MS, MPH1, Melissa Plegue, M.A.2, Elizabeth Jackson, MD MPH2, Ananda
Sen, PhD2, Kara Zivin, PhD3, Caroline Richardson, MD2

1
VA CCMR, Ann Arbor, MI; 2University of Michigan, Ann Arbor, MI; 3VA Ann Arbor Healthcare
System, Ann Arbor, MI

Physical inactivity is associated with higher health care utilization and costs among adults with
diabetes. Pedometer-based walking interventions have been shown to be effective in
increasing physical activity for diabetics, but the effect on health care utilization and costs is
unclear. This study examines how an incentivized pedometer-based walking intervention
impacts both physical activity and health care costs for adults with diabetes.

In 2009, Blue Cross Blue Care Network enrollees with a body mass index (BMI) of 30 kg/m2 or
greater were given a choice to join an internet-mediated pedometer-based walking program
(Walkingspree). Participation was financially incentivized because individuals could potentially
save an estimated 20% of out-of-pocket expenses by uploading their step count data at least
once every 30 days to the program website and averaging at least 5000 daily steps every three
months. The pedometer assessed both total daily steps and aerobic steps, measured as
continuous steps for at least 10 minutes. A multiple linear regression analysis determined the
impact of average daily steps and aerobic steps on the change in health care costs after one
year, controlling for diabetes status, age, gender, body mass index, length of program
participation and other comorbidities.

Of the 7,594 WalkingSpree participants, 15 percent had diabetes. On average, every 100 daily
steps were associated with a $9.07 decrease in health care costs (p=0.049). Among
participants with a mean of 5000 daily steps, predictive margins show the average person
without diabetes experienced cost savings of -$441.93 (p=0.004). In comparison, those with
diabetes averaged increased costs ($872.67, p < 0.001), particularly for those with diabetes
with complications ($2491.88, p < 0.001). However, these costs decreased for every 100 daily
steps.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1665

Greater daily steps are associated with decreased total health care costs for all three groups,
but there was a nonlinear differential effect by diabetes status. Further investigation is
needed to better understand the relationship between step counts and health care costs for
diabetics, especially those with more serious complications. Although people with diabetes
may incur greater health care costs, daily steps may help to slow these increased costs over
time.

CORRESPONDING AUTHOR: Mona AuYoung, PhD, MS, MPH, VA CCMR, Ann Arbor, MI, 48113-
0170; [email protected]
S1666 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 36: Translation of Interventions in Disease Prevention and Management 3:37
PM-3:54 PM

EFFECTIVENESS OF AN OBESITY TREATMENT APP WITH PROVIDER COUNSELING FOR

MEDICALLY VULNERABLE PATIENTS IN PRIMARY CARE PRACTICE:

Gary G. Bennett, PhD1, Dori M. Steinberg, PhD, RD1, Sandy Askew, MPH1, Erica Levine, MPH2

1
Duke University, Durham, NC; 2Duke Global Health Institute, Durham, NC

INTRODUCTION: Obesity treatment is less successful for those in medically vulnerable

populations. These disparities extend to treatments delivered in the primary care setting,
where there is little evidence that we can produce clinically meaningful weight loss among
populations who have the greatest obesity risk.

OBJECTIVE: We examined whether obesity treatment app, combined with counseling from a
primary care provider and a dietitian, could outperform weight losses produced by usual care
among medically vulnerable primary care patients.

METHODS: Track is a 12-month randomized controlled trial of a digital health weight loss
intervention in a community health center system. Participants were 351 adults (aged 21-65)
with obesity and a diagnosis of hypertension, diabetes, and/or hyperlipidemia. We
randomized participants to usual primary care or the 12-month Track intervention which
comprised: algorithm-generated tailored behavior change goals, weekly self-monitoring via
mobile technologies, daily self-weighing using a network-connected scale, skills training
materials, and 18 counseling phone calls with a dietitian. Primary care providers delivered
counseling to intervention participants using electronic-health record-based
recommendations generated by the Track app. The primary outcome was weight change over
12 months.

RESULTS: Participations (n=337) were largely female (69%) and non-Hispanic Black (52%), with
a mean age of 50.6 years and a mean body mass index of 35.9 kg/m. Most (90%) had less
than a college degree, 32% were unemployed, and half (52%) were living under or near the
poverty line. The Track intervention produced larger weight losses than usual care at 6- (net
effect: -4.4 kg, 95% CI [-5.5, -3.3], P < .0001) and 12-months post-randomization (net effect: -
3.8 kg, 95% CI [-5.1, -2.5], p < .0001). Intervention participants (n=170) completed a median
93.2% (IQR: 54% to 100%) of weekly self-monitoring via mobile device and a median 89%
(IQR: 50% to 100%) of counseling calls. Intervention participants were more likely to lose 5%
or more of their baseline weight at 6- (43% vs. 6%, p < .0001) and 12-months (40% vs. 17%, p
< .0001) post-randomization.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1667

DISCUSSION: We can effectively treat obesity among medically vulnerable patients in

community health centers using a scalable, low cost, digital treatment that both informs and
optimizes provider counseling.

CORRESPONDING AUTHOR: Gary G. Bennett, PhD, Duke University, Durham, NC, 27708;
[email protected]
S1668 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 36: Translation of Interventions in Disease Prevention and Management 3:55
PM-4:12 PM

PROVIDER COUNSELING IS ASSOCIATED WITH INCREASED WEIGHT LOSS IN A PRIMARY CARE-

BASED DIGITAL HEALTH WEIGHT LOSS INTERVENTION

Megan McVay, PhD1, Dori M. Steinberg, PhD, RD2, Sandy Askew, MPH2, Gary G. Bennett, PhD2

1
Duke University School of Medicine, Durham, NC; 2Duke University, Durham, NC

Background: Digital health weight loss interventions offer substantial promise for addressing
obesity in primary care. Primary care providers may play an important auxiliary role in these
treatments by counseling patients to stay engaged in treatment. Methods: This is a secondary
analysis of Track: a randomized clinical trial testing a 12-month digital health behavioral
weight loss intervention delivered at community health centers compared to usual care. At 6
and 12 months, participants reported whether their providers counseled them about weight
generally (general weight counseling) or specifically about intervention participation
(intervention counseling) at their most recent clinic visit, as well as their perceptions of
providers empathy. Weight was also collected at these times points. We used adjusted linear
regression models to compare differences in weight outcomes based on provider counseling
among intervention participants (no counseling vs. general weight counseling only vs.
intervention counseling) and separately among usual care participants (no counseling vs.
general weight counseling). Results: Study participants were 69% female, 52% Black, and had
a mean age of 51 years and BMI of 36 kg/m2. Intervention participants (6 month, n=119; 12
month, n=115) and usual care participants (6 and 12 months, n=124) reported similar
likelihood of receiving any counseling from providers at 6 months (75.6% vs. 69.8%, p=.31)
and 12 months (72.2% vs. 66.9%, p=.38). Provider counseling reported at 6 months was not
associated with weight change at 6 or 12 months among intervention participants, p>.05.
However, provider counseling reported at 12 months was associated with weight change
among intervention participants; those receiving intervention counseling lost more weight
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1669

than those receiving general weight counseling only (adjusted Mdiff = 4.1 kg, 95% CI [0.2, 7.9],
p=.04) and no counseling (adjusted Mdiff = 4.4 kg, 95% CI [0.04, 8.9], p=.047). Among
intervention participants only, perceptions of provider empathy reported at 12 months was
associated with 12 month weight loss, b=-0.13 kg, t(110)=2.0, p=.049. Among usual care
participants, weight loss counseling was not associated with weight loss at any time point.
Conclusions: Our findings suggest that whereas general weight counseling by primary care
providers is unlikely to influence weight, counseling patients to engage in a specific, evidence-
based digital health treatment can contribute to weight loss. Providers conveying empathy
towards patients may also impact weight.

CORRESPONDING AUTHOR: Megan McVay, PhD, Duke University School of Medicine, Durham,
NC, 27701; [email protected]
S1670 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 36: Translation of Interventions in Disease Prevention and Management 4:13
PM-4:30 PM

IMPLEMENTING A PRIMARY CARE ELECTRONIC HEALTH RECORD-BASED MODEL FOR BEST

PRACTICES IN PAIN MANAGEMENT FOR LOW-INCOME PATIENTS

Lara Dhingra, PhD1, Robert Schiller, MD2, Jack Chen, MBS1, Sarah Nosal, MD3, Raymond Teets,
MD2, Nandini Shroff, MPH4, Saskia Shuman, MHS2, Regina Ginzburg, PharmD2, Russell
Portenoy, MD1

1
MJHS Institute for Innovation in Palliative Care, New York, NY; 2Institute for Family Health,
New York, NY; 3Institute for Family Health, Bronx, NY; 4The Institute for Family Health, New
York, NY

Health Information Technology (HIT) can address gaps and variation in provider knowledge
and skills, but studies have noted challenges to the adoption of new technology into daily
patient care. While HIT solutions to improve pain management in primary care may be
valuable, obstacles to adoption must be overcome to realize any benefits. As part of a
randomized trial, we implemented a new electronic health record (EHR)based decision
support tool (the Pain Management Support System for Primary Care [PMSS-PC]) in 6
practices of a large Federally Qualified Health Center, which serve a low-income population in
New York City and upstate New York. We evaluated uptake of the PMSS-PC by primary care
providers (PCPs) in the practices as part of an examination of its implementation. The PMSS-
PC was programmed into Epics platform and used a best practice alert format similar to
extant decision support for health screening and management of chronic diseases, like
diabetes. It included informational screens, checklists, documentation requirements, and
validated practice support tools. Specific elements included screening tools for depression,
anxiety, and substance use; a pain assessment tool; opioid risk assessment, and treatment
guidelines. Treatment guidelines were developed for primary care and included first-line non-
pharmacological approaches, referral for psychological, integrative, and other care, as well as
pharmacotherapy. EHR data extracted between 3/2014-4/2016 showed that 46 PCPs used
the pain assessment tool a total of 1,595 times for 237 patients. Further, 63 PCPs accessed
the treatment guidelines a total of 1,061 times for 433 patients. While few referrals and
treatments were provided using the PMSS-PC (a total of 109 times for 97 patients), overall
referrals increased in the practices between the first 3 months, and last 3 months of
implementation by 110.3% (1,561 to 3,283; physical therapy, 33.5%, 37.8%; psychological and
behavioral care, 29.5%, 29.2%; orthopedic care, 27.0%, 21.9%; pain management, 6.7%, 8.0%;
acupuncture and chiropractic care, 3.3%, 3.1%, respectively). Similarly, the number of
patients who received prescriptions increased by 156.8% (1,376 to 3,533) during this same
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1671

time: 7.9%, 6.9% opioids; 92.1%, 93.1% non-opioids. Only 39.8% of PCPs reported high-
moderate satisfaction with the PMSS-PC. These results suggest that provider use of the
PMSS-PC in this trial was limited, but changes in pain practice were more substantial. We will
report final analyses on uptake and potential modifiers of PCP use of the program over time,
including patient (e.g., depression), provider (e.g., satisfaction), and practice-related (e.g.,
location) characteristics. Future results may show the importance of using HIT to promote
best practices for pain management and its positive effect on patient-reported outcomes.

CORRESPONDING AUTHOR: Lara Dhingra, PhD, MJHS Institute for Innovation in Palliative Care,
New York, NY, 10006; [email protected]
S1672 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 37: Older Adult's Physical Activity: Determinants and Interventions 3:00 PM-
3:18 PM

DEVELOPING A SUSTAINABLE PEER-LED PHYSICAL ACTIVITY PROGRAM FOR SENIOR

Michelle Takemoto, PhD Candidate1, Khalisa Bolling, MPH 1, Brittany Lewars, MPH2, Katie
Crist, MPH3, Eileen Johnson, BA4, Porchia Rich, MD, MPH, MSW1, Kelsie M. Full, MPH, PhDc5,
Jacqueline Kerr, PhD, MSc6

1
University of California, San Diego, San Diego, CA; 2University of San Diego, California, El
Cajon, CA; 3UC San Diego, La Jolla, CA; 4University of California, San Diego, La Jolla, CA;
5
University of California - San Diego, La Jolla, CA; 6Department of Family Medicine and Public
Health, UC San Diego, La Jolla, CA, USA, La Jolla, CA

Less than 3% of older adults meet physical activity (PA) guidelines. Our previous project,
Multilevel Intervention for Physical Activity in Retirement Communities (MIPARC), increased
participants step counts and prevented the inherent decline associated with aging.
Additionally, participants successfully completed advocacy projects to improve walkability in
and around the retirement communities (e.g., safer crosswalks, cleaner pedestrian bridges).

However, once the study ended, transitioning leadership from UCSD staff to individuals within
the community was challenging. Based on these results, the Peer Empowerment Program for
PA (PEP4PA) is designed to promote & assess delivery and sustainment of a peer-led multi-
level PA program in senior centers. PEP4PA transitions intervention delivery from UCSD
research staff to peer health coaches (PHCs) and center staff from within senior centers. PHCs
and a center staff member complete a 16-hour training course and certification to lead the
program and build a sustainment plan.

We employ mixed-methods to evaluate program implementation. Intervention fidelity is

assessed through an online tablet that PHCs use to track intervention delivery including
attendance, participant step counts, and progress towards goals. To promote success, the
intervention dose was simplified to ensure PHCs have capacity to deliver the components.
PHCs meet with participants in person for counseling, use color coding to identify goal setting,
and are encouraged to work with participants who are most in need. As a pragmatic trial,
PEP4PA has rolling recruitment which helps with motivation during the two year project by
continuing to have new participants enroll. Results from the implementation measures are
used to improve the project under the Plan, Do, Study, Act framework.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1673

Training PHCs within the senior centers to lead and deliver the program has empowered
individuals to seek out additional advocacy opportunities. Two PHCs have joined local
committees and all sites secured funding from local businesses to cover some program
expenses. Building capacity within the community and discussing long-term sustainability in
the beginning ensures the program will continue after the research study ends.

CORRESPONDING AUTHOR: Michelle Takemoto, PhD Candidate, University of California, San

Diego, San Diego, CA, 92111; [email protected]
S1674 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 37: Older Adult's Physical Activity: Determinants and Interventions 3:19 PM-
3:36 PM

CITATION AWARD WINNER

FORMATIVE RESEARCH OUTCOMES FOR THE SMART TELEVISION EXERCISE PROGRAM FOR
INDEPENDENT LIVING FACILITIES STUDY (STEP FOR LIFE)

Valerie H. Myers, PhD1, Courtney Fultineer, MPH2, Sophia M. Strickfaden, BA1, Antonio
Galvan, MS3

1
Klein Buendel, Inc., Golden, CO; 2Colorado Foundation for Public Health and the
Environment, Denver, CO; 3Life Care Service, Plainfield, IL

Despite its significance in maintaining functional independence, physical activity (PA) is lowest
among Older Adults (OA). Streaming technology offers innovative potential in delivering
personalized wellness programs to OAs. STEP for LIFE is an NIH-funded study evaluating the
feasibility of a PA program for senior adults residing in Independent Living Facilities (ILF) using
interactive web-enabled television that can be individually tailored. Study goals are to 1)
examine both individual- and organizational-level factors to inform translation of PA for OAs
into a scalable delivery medium, and 2) inform future translation of other evidence-based
health promotion programs with OAs and other senior living settings. To guide prototype
development, a national sample of ILF residents (n=136; mean age 82.4 +/- 6.9 years; 91%
White; 67.9% female) and ILF administrators (n=51; mean age 41 +/- 13.1 years; 80% White;
72% female) were surveyed. Administrators were asked to assess the type, quality, quantity,
and attendance of PA programs available for their residents and the organizational decision-
making process and level of interest in technology approaches for delivery of PA programs.
Results revealed that 88% strongly agreed that providing organized PA programs was
important, that a variety of approaches were used to promote organizational programs (word
of mouth, newsletters, family members, etc.), and on average, half of the facilities had regular
resident participation. Less than 16% of administrators disagreed that a SmartTV exercise
program would be helpful or of interest. Over 70% of administrators indicated that the ability
to tailor the exercise would be highly appealing. ILF residents were surveyed on their exercise
and social activities, and use and comfort with different technologies. Results revealed that
90% rated their health as good or very good, 51.3% participated in a PA program at their
facility 2-3 times a week, and 37.3% reported 5-8 hours a day of sitting. Over 83% reported
comfort using the Internet and 59.7% reported comfort with using a tablet device. Nearly 63%
reported that with some help or training they would be agreeable to trying a SmartTV exercise
program. To further inform prototype development, focus groups and semi-structured
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1675

interviews were also conducted. Real-world usability testing is currently underway. Survey
and formative research results as well as the quantitative outcomes from the completed study
will be presented.

CORRESPONDING AUTHOR: Valerie H. Myers, PhD, Klein Buendel, Inc., Golden, CO, 80401;
[email protected]
S1676 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 37: Older Adult's Physical Activity: Determinants and Interventions 3:37 PM-
3:54 PM

TESTING PSYCHOSOCIAL MEDIATORS OF INTERVENTION STRATEGIES TO PROMOTE PHYSICAL

ACTIVITY IN OLDER ADULTS

Siobhan K. McMahon, PhD, Beth Lewis, PhD, J Michael Oakes, PhD, Weihua Guan, PhD, Jean
Wyman, PhD, Alexander J. Rothman, PhD

University of Minnesota, Minneapolis, MN

Purpose: Evidence is inconclusive about which behavioral change strategies effectively

promote physical activity (PA) among older adults. Thus, we conducted an experiment testing
the relative effects of two theoretically driven sets of behavioral change strategies
interpersonal and intrapersonaldelivered in combination with an evidence-based PA
protocol, on average daily step count. The purpose of this study is to explore whether
psychosocial constructs targeted by these behavioral change strategies can help explain why
specific sets of strategies were and were not effective.

Methods: Data for this study are from a 2x2 factorial experiment in which older adults
(n=102) were randomized to one of 4 conditions: (1) Otago (an evidence-based PA protocol);
(2) Otago+interpersonal behavioral change strategies; (3) Otago+intrapersonal behavioral
change strategies; or (4) Otago+interpersonal and intrapersonal behavioral change strategies.
All participants received a PA monitor (Fitbit OneTM) and intervention content during 8 weekly
small group (4-6 people) meetings, 90 minutes each. Data was collected at 3 time points:
baseline, post-intervention, and 6 months follow-up. PA was measured as average daily step
count via Fitibit monitors. Using the PROCESS macro, ordinary least squares regression models
were assessed in which sets of interpersonal and intrapersonal behavioral change strategies
were considered as independent variables; changes in the psychosocial constructs social
support, readiness, self-efficacy, self-regulation and goal attainment, post-intervention--
controlling for baseline--were considered mediators; and PA, post-intervention and 6 months
follow-up--controlling for baseline--were considered outcomes.

Results: Participants who received interpersonal behavioral change strategies, compared to

those who did not, had significantly greater increases in their average daily step count post-
intervention (p < .001) and six months post-intervention (p < .001). They also reported greater
increases in readiness, self-regulation, and goal attainment at both time points (p's < .05), but
only self-regulation mediated the intervention-PA reltionship. Receipt of intrapersonal
behavioral change strategies did not lead to an increase in PA, but was associated with higher
ratings of social support from family by participants at both time points (p's < .01).
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1677

Conclusion: By exploring the psychosocial constructs targeted by specific sets of interpersonal

and intrapersonal behavioral change strategies, this study contributes to current knowledge
related to the development of PA interventions for older adults and suggests it may be
important to target self-regulation through interpersonal behavioral strategies rather than
intrapersonal strategies.

CORRESPONDING AUTHOR: Siobhan K. McMahon, PhD, University of Minnesota, Minneapolis,

MN, 55410; [email protected]
S1678 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 37: Older Adult's Physical Activity: Determinants and Interventions 3:55 PM-
4:12 PM

CITATION AWARD WINNER

REDUCTIONS IN LONELINESS THROUGH EXERCISE TRAINING IN OLDER ADULTS:

PSYCHOSOCIAL AND NEURAL DETERMINANTS

Diane Ehlers, PhD1, Ana M. Daugherty, Ph.D.2, Agnieszka Burzynska, PhD3, Jason Fanning,
PhD4, Elizabeth Awick, MS5, Laura Chaddock-Heyman, PhD5, Arthur Kramer, PhD6, Edward
McAuley, PhD7

1
University of Illinois Urbana-Champaign, Urbana, IL; 2Beckman Institute for Advanced Science
and Technology, University of Illinois at Urbana-Champaign, Urbana, IL; 3Colorado State
University, Fort Collins, CO; 4Wake Forest University, Winston Salem, NC; 5University of Illinois
at Urbana-Champaign, Urbana, IL; 6Northeastern University, Boston, MA; 7University of Illinois
at Urbana Champaign, Urbana, IL

Despite the prevalence of and negative health consequences associated with perceived
loneliness in older adults, few studies have examined interactions among behavioral,
psychosocial, and neural mechanisms. Research suggests that physical activity and
improvements in perceived social support and stress are related to reductions in loneliness.
Yet, the influence of brain structure on these changes is unknown. The present study
examined whether change in regional brain volume (i.e., amygdala volume) mediated the
effects of changes in social support and stress on change in perceived loneliness after an
exercise intervention. We also examined the extent to which baseline amygdala volume
moderated the relationship between changes in social support, stress, and loneliness.
Participants were 247 older adults (65.44.6 years-old) enrolled in a 6-month randomized
controlled trial comprised of three exercise conditions: Dance (n=69),
Strength/Stretching/Stability (n=70), and Walk (n=108). All groups met for one hour, three
times weekly. Participants completed questionnaires assessing perceived social support,
stress, and loneliness at baseline and post-intervention. Amygdala volume before and after
the intervention was measured with automatic segmentation of each participants T1-
weighted structural MRI. Data were analyzed in a latent modeling framework, with regression
coefficients presented as unstandardized estimates. Perceived social support increased
(p=0.003), while stress (p < 0.001) and loneliness (p=0.001) decreased over the intervention.
Increased social support predicted improved loneliness directly (-0.63, p < 0.01) and indirectly
through decreased stress (-0.10, p = 0.02). Amygdala volume did not change over the
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1679

intervention (p=0.40) and was unrelated to change in psychosocial variables (all p>0.44).
However, individuals with larger baseline amygdalae experienced greater reductions in stress,
and in turn greater decreases in loneliness (0.35, p = 0.02). No group differences in these
pathways were observed. The social support environment of exercise sessions, as opposed to
exercise mode, may represent an important feature of exercise programs aimed at improving
older adults perceived loneliness. As amygdala volume has been linked to anxiety, depression
and impaired cognitive control processes, moderation findings suggest further investigation in
this area is warranted.

CORRESPONDING AUTHOR: Diane Ehlers, PhD, University of Illinois Urbana-Champaign,

Urbana, IL, 61801; [email protected]
S1680 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 37: Older Adult's Physical Activity: Determinants and Interventions 4:13 PM-
4:30 PM

GROUP BASED PHYSICAL ACTIVITY FOR OLDER ADULTS (GOAL) RANDOMIZED CONTROLLED
TRIAL: 3- AND 6-MONTH ADHERENCE OUTCOMES

Mark R. Beauchamp, PhD1, William L. Dunlop, PhD2, Geralyn R. Ruissen, MSc3, Ryan E.
Rhodes, PhD4, Paul A. Estabrooks, PhD5, Samantha M. Harden, PhD6, Svenja A. Wolf, PhD7,
Yan Liu, PhD8, Toni Schmader, PhD1, Eli Puterman, PhD1, Andrew W. Sheel, PhD1

1
University of British Columbia, Vancouver, BC, Canada; 2University of California, Riverside,
Riverside, CA; 3The University of British Columbia, Vancouver, BC, Canada; 4University of
Victoria, Victoria, BC, Canada; 5University of Nebraska Medical Center, Omaha, NE; 6Virginia
Tech, Blacksburg, VA; 7University of Amsterdam, Amsterdam, Noord-Holland, Netherlands;
8
Educational and Counselling Psychology, and Special Education, Vancouver, BC, Canada

Title: GrOup based physical Activity for oLder adults (GOAL) randomized controlled trial: 3-
and 6-month adherence outcomes.

Abstract: The GrOup-based physical Activity for oLder adults (GOAL) trial was a three-arm RCT
(ClinicalTrials.gov # NCT02023632) that was designed to test the efficacy of two different
group-based exercise programs for older adults (informed by the tenets of self-categorization
theory; SCT) in relation to a standard group-based exercise program. 630 older adults (M age
= 71.44, SD = 5.36; 71.1% female) were randomized to similar-age same gender (SASG),
similar-aged mixed gender (SAMG), or standard mixed-aged mixed gender (MAMG) exercise
group conditions. In addition to considering group composition, the SCT-informed conditions
included older-adult instructors, and operationalized key principles and strategies from SCT
designed to foster a sense of social connectedness among exercise group participants. In this
paper we report the results of the trials primary outcome, namely adherence behavior
(assessed through objective measures of class attendance) over 3 and 6 months. Based on
intent-to-treat (ITT) analytic principles, analyses of covariance (controlling for pre-intervention
physical activity behavior) revealed that older adults randomized to the SAMG (3 months d =
.48, p = .001; 6 months d = .44, p= .004) and SASG (3 months d = .34, p= .008; d = .35, p = .01)
conditions adhered to a greater extent than those in the MAMG comparison condition. At
both time points these effects were invariant by gender, and the differences between the
SAMG and SASG conditions were non-significant. In conclusion, the results of this RCT provide
support for the efficacy of group-based physical activity programs informed by the tenets of
SCT. Furthermore, the results suggest that community group-based exercise programs should
attempt to engage in age-targeting but not necessarily gender-targeting among older adults.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1681

CORRESPONDING AUTHOR: Mark R. Beauchamp, PhD, University of British Columbia,

Vancouver, BC, V6T1Z1; [email protected]
S1682 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 38: Understanding Risk Perception 3:00 PM-3:18 PM

CITATION AWARD WINNER

EXAMINING DIFFERENT REACTIONS TO NARRATIVES FROM COLON CANCER SURVIVORS VS.

HEALTHY SCREENERS

Amy McQueen, PhD1, Julianne A. Sefko, M.P.H.2, Charlene A. Caburnay, PhD, MPH3, Matthew
Kreuter, PhD3

1
Washington University, School of Medicine, St. Louis, MO; 2Washington University School of
Medicine, Saint Louis, MO; 3Washington University in St. Louis, St. Louis, MO

The use of patient narratives or testimonials in interventions and on the Internet is becoming
ubiquitous, and has outpaced empirical research regarding how and for whom narratives are
effective. Most research has compared narratives with didactic messages rather than
compare across narratives. To inform future interventions that incorporate narratives, we
tested the effects of different role models (survivors vs. screeners) for increasing
colorectal cancer screening (CRCS) intentions and behaviors. We randomized adults age 50-75
not adherent to CRCS guidelines with no history of cancer to one of three groups: 1) CRCS
information only, 2) CRCS info plus a narrative from a CRC survivor, or 3) CRCS info plus a
narrative from a healthy CRC screener. Role model photos were tailored to the participants
sex, age group and race/ethnicity. Participants (n=487) completed surveys online before and
after the intervention and at 1 & 6 month follow up. We used structural equation modeling
(AMOS 16) to test our conceptual model of the direct and indirect effects of narratives on
CRCS intention and behavior via post-exposure mediators including engagement, emotions,
message acceptance, and psychosocial determinants of CRCS, controlling for significant
covariates. Results showed good model fit. Narratives were more engaging (=.27,p < .001)
and emotionally evocative (=.10,p < .01), but less believable (= -.09,p < .01) than
information only. Engagement was related to greater perceived benefits and social norms for
CRCS (p < .001); greater affect was related to perceived risk, worry and intention (p <
.01). Survivor (vs. screener) stories aroused more negative emotions (=.33,p < .001) which
was related to greater perceived risk, worry, barriers, and info avoidance (p < .001). Survivors
were perceived as less similar to participants (= -.13,p < .05), which was positively related to
message acceptance and social norms (p < .001). Survivor stories produced less self-relevant
engagement (= -.22,p < .001), which was related to greater perceived risk, benefits, norms,
self-efficacy, and intention; less counterarguing and perceived barriers. Only intention post-
intervention predicted CRCS by 6 months follow up. Cancer survivor narratives evoke more
negative emotion and reduced self-related engagement and identification which may
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1683

indirectly undermine message effects on CRCS. Understanding the mechanisms of how

different role models influence health behaviors is important for selecting or creating
effective narratives.

CORRESPONDING AUTHOR: Amy McQueen, PhD, Washington University, School of Medicine,

St. Louis, MO, 63110; [email protected]
S1684 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 38: Understanding Risk Perception 3:19 PM-3:36 PM

REACTIONS TO COLORECTAL CANCER RISK ESTIMATES AMONG A NATIONALLY

REPRESENTATIVE SAMPLE OF ADULTS WHO HAVE NEVER BEEN SCREENED

Constance M. Johnson, PhD1, Sathya Amarasekara, MS2, Wei Pan, PhD1, John A. Updegraff,
PhD3, Isaac Lipkus, PhD4

1
Duke University, Durham, NC; 2Duke University Health System, Durham, NC; 3Kent State
University, Kent, OH; 4Duke University School of nursing, Durham, NC

Screening for colorectal cancer (CRC), the 3rd leading cause of cancer death among men and
women in the United States, saves lives. Yet, among adults ages 50 to 75, about 30% have
never been screened. Internet-based strategies that deliver tailored CRC risk estimates can
promote screening, especially among those at higher risk. To be effective, recipients of
feedback should be able to extract their risk level delivered online, find the risk estimate
accurate and useful, and it should influence their perceived CRC risk. We examined these
reactions to online tailored CRC risk estimates among adults who have never been screened
for CRC. Potential participants were recruited from GfKs Knowledge Networks online panel, a
panel representative of the U.S. population. Eligible participants were between the ages of 50
to 75 who self-reported never having been screened. Using the Harvard Disease Risk Index
algorithm, participants received tailored 10-year risk of getting CRC compared to others their
age and sex ranging from very much below average to very much above average seven
levels of risk total. Participants were asked to recall their estimate and rate on a 7-point Likert
scale its accuracy and usefulness, as well as their perceived comparative 10-year risk from
very much below average to very much above average. Among the 619 participants who
qualified, mean age was 58 (SD=6.4), 46% were male, and 75% were Non-Hispanic, White;
39% of the sample had a high school education or less. The mean level of the tailored
comparative risk estimate was average. Overall, 29% of the sample could not reflect back
their estimated risk. Greater inaccuracy was higher among the less educated. Perceived
comparative risk was slightly below average (M=3.2; SD=1.3) and correlated with the risk
estimate (r=.40, p < .0001). The risk estimate was viewed as somewhat accurate (M=4.0, SD=
1.3) and somewhat useful (M=4.7, SD=4.7). These reactions varied by risk estimate. With
increasing risk, estimates were viewed as less accurate (r=-.14, p < .05) and less useful (r=-.28,
p < .001). Findings suggest that many individuals, especially the less educated, have trouble
extracting their online risk estimate and view it as less acceptable (e.g., accurate, useful) with
higher degree of risk, pointing to defensive processing. Strategies to curb these potential
barriers to online risk communication strategies to promote CRC screening targeting adults
who have never been screened appear warranted.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1685

CORRESPONDING AUTHOR: Constance M. Johnson, PhD, Duke University, Durham, NC, 27710;
[email protected]
S1686 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 38: Understanding Risk Perception 3:37 PM-3:54 PM

DIFFERENT TYPES OF RISK: RISK UNCERTAINTY VARIES BY RISK FRAMING.

Caitlin Biddle, MA, MS1, Jennifer Hay, PhD2, Erika A. Waters, PhD, MPH3, Marc T. Kiviniemi,
BA, PhD1, Heather Orom, PhD1

1
University at Buffalo, Buffalo, NY; 2Memorial Sloan Kettering Cancer Center, New York, NY;
3
Washington University in St. Louis, St. Louis, MO

Purpose: There is debate over the conceptualization and measurement of risk perception. A
variety of measurement approaches have been employed, including assessing perceived risk
as an absolute or comparative judgment and attempting to capture the corrections people
make in their risk judgments based on their past or expected behavior (conditional risk). In
order to assure that our measurement strategies are valid, it is important to understand
whether people are able to make meaningful judgments in response to researchers risk
perception questions. The purpose of this study was to compare participants level of
uncertainty about their risk judgments based on measurement approach.

Methods: Participants were 280 MTURK workers (median education = some college; 80%
White; mean age = 34(SD = 9)) who completed an on-line survey. We assessed conditional
and non-conditional absolute and comparative heart disease risk using a Likert-type scale with
a dont know option. Those who chose a valid response (dont know responders were
excluded) were asked to rate their certainty about their judgment. Paired t-tests were used
to compare mean certainty of participants risk as a function of 1) being conditioned on
healthy behavior change and 2) framed as absolute vs. comparative risk.

Results: More people reported being uncertain or very uncertain about their absolute risk
and non-conditioned risk: 66.36% of participants were uncertain about absolute risk, 44.35%
about comparative risk, 31.45% about conditional absolute risk, and 29.30% about their
conditional comparative risk. For both absolute and comparative heart disease risk,
participants were more uncertain about their non-conditional risk (mean difference= -.5, t = -
8.8, p < .001; mean difference= -.3, t = -6.1, p < .001) than conditional risk. For both
conditional and non-conditional risk, participants were more uncertain about their absolute
risk than comparative risk (mean difference= -.09, t = -2.0, p=.02; mean difference= -.27, t = -
6.4, p < .001).

Discussion: Even after eliminating individuals who say that they dont know their risk,
participants were uncertain about their judgments. The way participants are asked to appraise
their risk contributes to this uncertainty; participants are more uncertain about non-
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1687

conditional than conditional judgments and about absolute than comparative judgments.
While our previous research demonstrates that lack of knowledge contributes to uncertainty
about risk, this association does not explain why levels of uncertainty vary by question format.
Non-conditional and absolute risk items are more ambiguous and may be more difficult to
answer for participants because there a lack of information or no frame of reference, which
may lead to more uncertainty.

CORRESPONDING AUTHOR: Caitlin Biddle, MA, MS, University at Buffalo, Buffalo, NY, 14220;
[email protected]
S1688 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 38: Understanding Risk Perception 3:55 PM-4:12 PM

BARRIERS AND PROMOTERS OF COMMUNICATION ABOUT FAMILY CANCER HISTORY AMONG

AFRICAN-AMERICANS IN BALTIMORE

Melinda Krakow, Phd MPH1, Gypsyamber D'Souza, PhD2, Rebkha Atnafou, MPH2, Anne
Rositch, PhD, MSPH2

1
National Cancer Institute, Washington, DC; 2Johns Hopkins Bloomberg School of Public
Health, Baltimore, MD

Objective. Although cancer is the second leading cause of deaths for Americans, African-
Americans remain more likely to die from cancer than any other racial or ethnic group in the
U.S. Family cancer history (FCH) plays a key role in shaping individual prevention and early
detection behaviors to decrease cancer risk. However, many individuals may lack knowledge
of FCH, and be unaware of increased risk for cancers based on family patterns. For some
African-American communities, a lack of communication about FCH is pronounced and
barriers to this communication have not been well studied. To optimize the use and impact of
FCH, it is crucial to better understand how patients gather and share FCH with family
members, and discuss information with healthcare providers.

Methods. We conducted four community focus groups (40 participants) and seven key
informant interviews (9 participants) to investigate knowledge and experiences of FCH and
barriers/promoters of FCH in the East Baltimore African American community.

Results. Thematic analysis of transcripts identified 11 distinct themes across 4 topic areas:
Definitions of FCH, past FCH communication, and barriers/promoters of FCH. FCH was most
often defined by narratives of cancer experienced by individual family members, rather than
records or patterns of disease over time. The majority of participants had gathered or shared
little or no FCH information with family members. Three psychosocial domains (fear/denial,
pride/dignity, and fatalistic attitudes about cancer) were commonly reported barriers to
sharing family cancer history. Additionally, distrust/skepticism about medical care was a
barrier to sharing FCH with healthcare providers. Diagnosis/death of a loved one and need for
caregiving/social support promoted FCH communication and encouraged cancer prevention
behaviors such as screening.

Conclusion. Although most participants had experienced cancer in their families, reported
communication about FCH was low and psychosocial barriers were common. Understanding
these communication domains in minority populations is crucial to addressing disparities in
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1689

cancer prevention, particularly where effective screening and care recommendations exist for
those with positive family histories. Community-based interventions are needed to improve
understanding of FCH, address common barriers, and provide communication tools to
promote family cancer history communication among African American families in Baltimore.

CORRESPONDING AUTHOR: Melinda Krakow, Phd MPH, National Cancer Institute,

Washington, DC, 20003; [email protected]
S1690 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 38: Understanding Risk Perception 4:13 PM-4:30 PM

PERCEIVED THREAT AND PRESSURE RELATED TO SPORT ACHIEVEMENT EXPLAIN WHY SOME
SPORT PARENTS DON'T TALK ABOUT CONCUSSIONS

Emily Kroshus, ScD MPH

University of Washington, Seattle, WA

Early identification and removal from play of athletes who have sustained a concussion is an
important aspect of reducing concussion-related harm. This process is often dependent on
athlete self-report of injury. Prior research finds that some athletes experience pressure from
their parents to continue playing with concussion symptoms (Kroshus et al., 2015). However,
it is not clear whether this perceived pressure is the result of direct communication or
whether it is inferred from factors such as parent investment in their sporting career. As one
step towards addressing this gap in the literature, the present study examines parent-
described direct communication to encourage concussion reporting, testing three hypotheses.
First, that greater perceived susceptibility to and seriousness of concussion would be
associated with greater parent-child communication about the importance of seeking care for
a suspected concussion. Second, that parents who put more pressure on their child related to
sport achievement would be less likely to encourage care seeking. Third, that there would be
a significant interaction between parent attitudes about the childs sport participation and
perceived seriousness and susceptibility in predicting concussion care seeking communication,
with the relationship between perceived severity, susceptibility and communication being
weaker among parents who put more sport achievement pressure on their child. Participants
were 246 parents of competitive youth club soccer players (ages 8-18 years) in a league
located in a large urban center in the northwest region of the United States who completed
an anonymous online survey. Across the full sample, just over two thirds of parents (68.38%)
indicated that they communicated with their child about the importance of reporting
suspected concussions. Consistent with the Health Belief Model (Rosenstock, 1974) and as
presently hypothesized, parents who believed that their child was more likely to sustain a
concussion were more likely to encourage their child to report symptoms of a suspected
concussion. However, parents who believed that the consequences of this concussion would
be more harmful were significantly less likely than other parents to encourage concussion
reporting. Further, there was an interaction between parent pressure and concussion severity
in that when parents reported greater pressure for sport achievement on their child, the
association between perceiving concussions to be more harmful and encouraging concussion
symptom reporting was significantly less strong. Results will be discussed in the context of
fear appeals, and the implications that these findings have for the development of health
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1691

education targeted at parents of youth sport participants to encourage concussion safety-

supportive communication.

CORRESPONDING AUTHOR: Emily Kroshus, ScD MPH, University of Washington, Seattle, WA,
98105; [email protected]
S1692 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 39: African American Women's Health 3:00 PM-3:18 PM

A QUALITATIVE EXAMINATION OF THE INFLUENCE OF CULTURE ON ENGAGEMENT IN

PHYSICAL ACTIVITY AMONG BLACK WOMEN

Allyson Diggins, M.A. , Nicole Whitehead, PhD

University of Florida, Gainesville, FL

Despite the noted benefits of physical activity, engagement among minorities and women,
particularly Black women, remains low. In response to the lower rates of physical activity
among Black women, there has been an increase in intervention studies promoting physical
activity. However, intervention studies for both inactive and underactive Black women have
produced modest success in short-term engagement and almost no success in long-term
maintenance. Understanding cultural context may be instrumental to developing effective
and sustainable solutions to address health disparities associated with inactivity. Therefore,
the purpose of the current investigation was to gain a better understanding of how culture
influences Black womens experiences with physical activity. Specifically, the associations
between cultural values, beliefs and attitudes and participation in physical activity were
examined. Data were collected using individual semi-structured interviews with participants
(N=12). The interview guide was designed to address common factors related to physical
activity adoption while also discovering additional factors. Women ranged in age from 19-52
(M = 29.38). Mean income for the sample was $33,100 (SD = 25.34). Participant BMIs ranged
from 18.60 to 47.30, with 42% being classified as overweight or obese. Audiotapes of the
interviews were transcribed verbatim and thematic analysis was used to generate key themes.
Emergent themes were then summarized, interpreted and categorized using the PEN-3
Model. Women identified the need for psychoeducation related to health versus aesthetic
benefits of physical activity, significance of communal support, the role of spirituality,
acceptance of larger body sizes as a barrier to a physically active lifestyle, lack of exposure to
physical activity as a familial/cultural norm, and the tendency to prioritize others over
themselves. Analysis of interviews conducted with this sample of women suggests an
interplay among the PEN-3 domains and physical activity behavior, with implications for the
design of culturally sensitive interventions. Results support the importance of incorporating
culture into the development and dissemination of an intervention. Thus, physical activity
interventions should aim to incorporate these perspectives without focusing exclusively on
cultural barriers. Instead, culture should be used to understand and then empower Black
women to increase their levels of physical activity while still remaining true to their culture.

CORRESPONDING AUTHOR: Allyson Diggins, M.A. , University of Florida, Gainesville, FL, 32608;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1693

Paper Session 39: African American Women's Health 3:19 PM-3:36 PM

PATIENT OPINIONS ON THE DESIGN OF A RACIALLY TAILORED EXERCISE INTERVENTION FOR

BLACK INDIVIDUALS WITH MULTIPLE SCLEROSIS

Dominique Kinnett-Hopkins, B.S. 1, Robert W. Motl, PhD2

1
University of Illinois at Urbana-Champaign, Urbana, IL; 2University of Alabama at
Birmingham, Birmingham, AL

Black individuals with multiple sclerosis (MS) experience a more aggressive disease course and
worse comorbidity and prognosis with disease modifying drug therapies than white
individuals with MS. This requires further consideration of approaches for managing MS in this
population, and one approach may be exercise training. There is substantial evidence of the
benefits of exercise training for managing the disease, yet the samples have been mostly
white. We further note that black individuals with MS engage in less physical activity and
exercise than white counterparts. This necessitates the need for developing a racially targeted
exercise training program for black individuals with MS. Such an intervention should include
patients during the designing phase, as the range of input is invaluable when creating the
best-suited exercise program for this population. To that end, we conducted a survey among
black individuals with MS about the development of a targeted-exercise training program.
Participants answered a series of items administered through Qualtircs (Prove, UT) regarding
demographics and opinions of an ideal exercise program, the importance of racial
considerations, and current physical activity levels. 40 black individuals with MS (age of 52.24
8.95, BMI of 30.39 9.28 kg/m, & 92.5% female) completed the survey. The mean physical
activity score was 44.54 22.19. Regarding the exercise program, participants wanted to
exercise 5.0 1.6 days per week at a moderate intensity (50% of participants). 62.5% of
participants desired an exercise program that was longer than 6 months, with 57.5% of
participants wanting the aid of a behavioral coach for supporting exercise behavior change.
The desired program consisted of mainly resistance/weight training exercises (52.5%) and
easy walking (35%). The expectations from exercising were primarily to increase muscle
strength (65%), to improve overall body functioning (60%), increase ability to perform daily
activities (57.5%) and to help manage stress (57.5%). The anticipated barriers to exercise were
largely fatigue (40%) and not having enough time (32.5%). There was variation in the
importance of racial considerations in the design of the program. The data from this study
should be considered when creating an exercise training intervention for black individuals
with MS.
S1694 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Dominique Kinnett-Hopkins, B.S. , University of Illinois at Urbana-

Champaign, Urbana, IL, 61801; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1695

Paper Session 39: African American Women's Health 3:37 PM-3:54 PM

SIX MONTH PHYSICAL ACTIVITY AND PSYCHOSOCIAL OUTCOMES FROM THE HIPP STUDY FOR
AFRICAN AMERICAN WOMEN IN DEEP SOUTH

Dori Pekmezi, PhD1, Matthew C. Ainsworth, MPH1, Rodney P. Joseph, PhD2, Karen Meneses,
PhD, RN, FAAN1, Bess H. Marcus, Ph.D.3, Wendy Demark-Wahnefried, PhD RD4, Renee
Desmond, xx5, Victoria A. Williams, MPH, MED, CHES 6, Elizabeth A. Kvale, MD, MSPH1

1
University of Alabama at Birmingham, Birmingham, AL; 2Arizona State University, Phoenix,
AZ; 3University of California, San Diego, La Jolla, CA; 4UAB, Birmingham, AL; 5xx, xx, AL;
6
University of Alabama at Birmingham, Atlanta, GA

Purpose: To establish the feasibility and acceptability of a Home-based, Individually-tailored

Physical activity Print (HIPP) intervention for African American women in the Deep South.

Methods: A pilot randomized controlled trial was conducted to test the HIPP intervention
(N=43) vs. wellness contact control arm (N=41). Physical activity (7-Day Physical Activity Recall
interviews, accelerometers) and related psychosocial variables were assessed at baseline and
6 months.

Results: The sample was recruited in Birmingham, AL and included 84 overweight or obese
African American women between the ages of 50-69. Retention was high at 6 months (90%).
Group differences in physical activity and related psychosocial variables at 6 months were
examined with a generalized linear model controlling for baseline values. On average, HIPP
intervention participants reported larger increases (M=+73.9 minutes/week, SD=90.9) in
moderate intensity physical activity from baseline to 6 months than the control group (+41.5,
SD=64.4), p> .05. The intervention group reported significantly greater improvements in
physical activity goal-setting and enjoyment from baseline to 6 months compared to the
control group. There were no other significant group differences [weight, BMI, % body fat
(BIA), 6MWT, physical activity planning, behavioral processes of change, stage of change];
however, trends in the data for cognitive processes of change, self-efficacy, outcome
expectations, and family support for physical activity indicated small improvements for the
HIPP arm (P> .05) and declines for the control arm. Significant increases in waist
circumference and decreases in decisional balance and friend support for physical activity
from baseline to six months were observed in the control arm and not the intervention arm.
Of the 36 intervention participants surveyed at follow-up, most reported being satisfied with
the HIPP program and finding it helpful (N=33, 91.67%). Almost all (N=35, 97.22%) would
recommend it to a friend.
S1696 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Discussion: The HIPP intervention has great potential as a low cost, high reach method for
reducing physical activity-related health disparities. The approach was well received by our
sample. The lack of improvement in some domains may indicate that additional resources and
supports will be required to help this target population reach national guidelines for MVPA (>
150 minutes/week of moderate intensity physical activity).

CORRESPONDING AUTHOR: Dori Pekmezi, PhD, University of Alabama at Birmingham,

Birmingham, AL, 35294; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1697

Paper Session 39: African American Women's Health 3:55 PM-4:12 PM

OPTIMIZING SOCIAL COGNITIVE THEORY IN THE DESGIN OF PHYSICAL ACTIVITY PROGRAMS

FOR AFRICAN AMERICAN WOMEN: A QUALITATIVE STUDY

Rodney P. Joseph, PhD1, Barbara E. Ainsworth, PhD, MPH1, Steven P. Hooker, Ph.D.1, Colleen
Keller, PhD2

1
Arizona State University, Phoenix, AZ; 2Arizona State University, phx, AZ

Introduction. The Social Cognitive Theory (SCT) is one of the most widely used theories in
physical activity (PA) promotion research. Through focus group methods, this study examined
the cultural relevance of five SCT constructs in the design and acceptability of a PA
intervention for sedentary and obese African American (AA) women.

Methods. Twenty-five AA women (M age = 38.5 years, M BMI = 39.4 kg.m2) were enrolled in a
series of 3 focus groups held over a 6 week period. Focus group guides were designed to gain
information on how five SCT constructs can be culturally tailored in the design of a PA
program for AA women: self-efficacy, self-regulation, behavioral capability, outcome
expectations, and social support. Focus groups were audio recorded and transcribed
verbatim. Content analysis was used to capture and extend knowledge and understanding of
cultural and contextual resources for participation in a theory-based PA intervention for AA
women.

Results. Participants provided in-depth data on how the five SCT constructs examined were
relevant and can be leveraged in the design of a PA program for AA women. Pertaining to the
construct of behavioral capability, AA women were generally unaware of the amount,
intensity, and types of PA needed for health benefits. Participants discussed a variety of
outcome expectations associated with regular PA engagement, including increased energy,
improved functional status, weight loss, and positive role modelling behaviors. Results for the
constructs of self-efficacy and self-regulation indicated that women perceived their own
intrapersonal processes as their primary barrier to PA and that PA programs should: a)
leverage an AA womens previous positive experiences with PA (i.e., during childhood and
early adulthood), and b) illustrate how PA is relevant to their familial, societal, and caretaking
roles. For the construct of social support, participants endorsed the need of a strong social
support component with a variety of sources acceptable to include in a PA program (i.e.,
family members, other study participants, friends).

Conclusions. Findings highlight the utility of the SCT as a theoretical basis for PA programs for
S1698 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

AA women. PA programs should clearly outline the dose and intensity of PA necessary to
achieve health benefits. Programs should also leverage AA womens previous positive
experiences with PA and illustrate how PA is relevant to their familial, societal, and caretaking
roles.

CORRESPONDING AUTHOR: Rodney P. Joseph, PhD, Arizona State University, Phoenix, AZ,
85004; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1699

Paper Session 39: African American Women's Health 4:13 PM-4:30 PM

EXAMINING PEER SUPPORT AND SURVIVORSHIP OUTCOMES FOR AFRICAN AMERICAN

WOMEN WITH BREAST CANCER

Shannon E. Nicks, PhD, MPH1, Ricardo Wray, PhD, MS2, Sherrill Jackson, MHSA, RN, CPNP3,
Olivia Peavler, MPH4, Stephanie McClure, PhD, MPH2, Theresa Schwartz, MD2

1
Ohio University, Athens, OH; 2Saint Louis University, St. Louis, MO; 3The Breakfast Club, Inc.,
Florrisant, MO; 4Mayo Clinic, Rochester, MN

More than three million female breast cancer (BrCa) survivors live in the United States, and
the number continues to grow. Health status and quality of life amongst survivors is variable,
and African American (AA) BrCa survivors suffer disproportionately from BrCa morbidity and
mortality. Emerging evidence suggests peer support is an effective strategy to promote
positive survivorship outcomes for AA women with BrCa. The aim of this study was to explore
the role of structured peer support in the BrCa experiences of AA survivors. The research
team worked collaboratively with The Breakfast Club, Inc. (BCI), a community-based BrCa
support and advocacy organization with twenty years of experience providing peer support to
survivors, to compare similarities and differences in the survivorship experiences of AA
women with BrCa according to receipt of structured peer support. Data were collected,
analyzed and triangulated from three sources to generate findings. These sources included:
participant observations of BCI members attending monthly support group meetings, key
informant interviews with BCI leadership and semi-structured interviews with two groups of
AA BrCa survivors, such that one group of survivors received structured peer support (in
addition to social support from other sources) and a comparable group of survivors received
standard care (only social support from other sources). Triangulated results indicate that peer
relationships function to assure social connectedness and provide consistent, quality social
support in addition to or in the absence of support from other sources. This consistent peer
support helps survivors cope with the continued stress of BrCa and may result in better
psychosocial health and quality of life. These findings expand our current understanding of
peer support and may enable public health and clinical practitioners to promote improved
quality of life during survivorship by better recognizing and intervening with those for whom
additional support services are needed.

CORRESPONDING AUTHOR: Shannon E. Nicks, PhD, MPH, Ohio University, Athens, OH, 45701;
[email protected]
S1700 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 40: Skin Cancer Prevention: Policies, Intentions and Measurement 3:00 PM-3:18
PM

SUNSCREEN APPLICATION KNOWLEDGE AND REPORTING IN SURVEY FORMAT

Anne Julian, M.A., Sheryl Thorburn, PhD, MPH

Oregon State University, Corvallis, OR

Background: Sunscreen use is the most commonly reported method of UV protection in the
U.S. Despite increasing public awareness of the risks of UV exposure, skin cancer incidence
has increased over the past decades. The effectiveness of sunscreen against the harms of UV
depends on proper application, and consumers frequently do not apply adequate sunscreen
to achieve the advertised SPF. The purpose of this study was to examine reported sunscreen
application, knowledge of proper sunscreen use, and quantity of sunscreen used.

Method: Undergraduate students (N = 335) from a university in Oregon completed an online

survey about their UV exposure beliefs and behaviors. Sunscreen use was measured with the
original and modified versions of NHIS sunscreen use frequency item. Sunscreen quantity
used was measured with two items; response options offered 1) a choice of written
descriptions of sunscreen quantity (i.e., 1/2 tsp.) and 2) an array of photographic depictions of
quantities of sunscreen in a hand, with a coin included for scale. Univariate and bivariate
analyses were performed first on the full sample and then stratified by gender.

Results: Whereas 34% of respondents reported using sunscreen consistently (i.e., almost
always or always) on warm sunny days, only 9% reported consistent sunscreen use on
most days. The majority of participants (89%) either did not know (48%) or reported
incorrectly (41%) the amount of sunscreen needed to achieve the advertised SPF. Sunscreen
use among participants who responded that they did not know the amount of sunscreen
needed did not significantly differ from use among participants who believed they knew the
appropriate amount (p = .22). Reported quantity of sunscreen applied was inconsistent
between the two formats of sunscreen quantity measurement (Weighted kappa = .20).

Discussion: Inconsistent and inadequate application of sunscreen is a challenge for skin cancer
prevention. Perceived lack of knowledge, however, was not associated with use of smaller or
larger amounts of sunscreen used. Inconsistent responses between written and photographic
items measuring sunscreen use highlight the need for more accurate measures of sunscreen
application. Future research is needed to clarify how sunscreen users decide when and how
much sunscreen to apply and to validate items measuring sunscreen use.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1701

CORRESPONDING AUTHOR: Anne Julian, M.A., Oregon State University, Corvallis, OR, 97339;
[email protected]
S1702 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 40: Skin Cancer Prevention: Policies, Intentions and Measurement 3:19 PM-3:36
PM

PROJECT SOL: TESTING A SUN-SAFETY PROGRAM FOR HISPANIC OUTDOOR DAY LABORERS

Elliot Coups, PhD1, Carolina Lozada, MPH2, Michele Ochsner, Ph.D.3, Louis Kimmel, BA4,
Germania Hernandez, N/A4, Reynalda Cruz, N/A4, Ashley Day, Ph.D. 2, Daniel Giovenco, MPH5

1
Rutgers, The State University of New Jersey, New Brunswick, NJ; 2Rutgers Cancer Institute of
New Jersey, New Brunswick, NJ; 3Rutgers University, New Brunswick, NJ; 4New Labor, New
Brunswick, NJ; 5Rutgers School of Public Health, Piscataway, NJ

Introduction: Individuals who work outside often encounter high temperatures, humidity, and
excess exposure to the suns ultraviolet (UV) rays. These factors increase the risk for heat
illness (heat exhaustion/stroke), sunburns, skin damage, and skin cancer. Hispanic individuals
in the United States are over-represented in many outdoor occupations, including landscaping
and construction. Few interventions have been tested to promote sun-safe behaviors among
Hispanic outdoor workers. In Project SOL (Sun-safety for Outdoor Laborers), we evaluated the
impact of a sun-safety program for Hispanic outdoor day laborers (who typically seek work in
informal, open-air urban settings such as street corners).

Methods: 119 Hispanic men (39% Mexican, 28% Ecuadorian, 13% Honduran; M age 39 years)
were recruited from outdoor locations where day laborers commonly congregate in New
Brunswick, NJ. We trained peer trainers, who delivered a Spanish language, culturally
targeted, 2-hour sun-safety program to groups of 722 individuals. Guided by a participant
workbook (developed using preliminary focus group and survey research), trainers led
participants through small group activities focused on the risks and precautions associated
with high heat, humidity, and UV exposure. Participants completed a survey immediately prior
to and after the program that included questions on sunburns and heat illness symptoms last
summer, sun-safety knowledge (8 true-false items), frequency of engaging in sun-safe
behaviors (8 items, 5-point never to always scale: taking regular breaks, drinking water,
working in the shade, and wearing a long-sleeved shirt, long pants, a wide-brimmed hat,
sunglasses, and sunscreen) while working outdoors last summer and intended frequency next
summer, and an evaluation of the programs content and format (9 items, 5-point strongly
disagree to strongly agree scale).

Results: Most participants reported having a sunburn (66%) and/or heat illness symptoms
(66%) last summer.Sun-safety knowledge increased significantly from before to after the
program (Ms = 4.4, 6.5 items correct, t = 13.3, p < .001). The frequency of engaging in sun
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1703

safe-behaviors as a whole last summer was low (M = 3.0) and the intended frequency was
significantly higher after the program (M = 4.2, t = 14.5, p < .001). Statistically significant
increases were also seen separately for each of the 8 behaviors (ts 3.4, ps .001).
Participants evaluated the program very favorably (M = 4.7 out of 5).

Conclusions: We demonstrated the feasibility of delivering a peer-led sun-safety program to

Hispanic outdoor day laborers. The program was well-received by participants and there were
increases in knowledge and sun-safe behavioral intentions after the program. Future research
is warranted to test the impact of the program on behavioral outcomes and to explore
potential dissemination approaches.

CORRESPONDING AUTHOR: Elliot Coups, PhD, Rutgers, The State University of New Jersey,
New Brunswick, NJ, 08901; [email protected]
S1704 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 40: Skin Cancer Prevention: Policies, Intentions and Measurement 3:37 PM-3:54
PM

THE IMPACT OF INDOOR TANNING LEGISLATION ON INDOOR TANNING PREVALENCE IN

ALABAMA

Aaron Blashill, Ph.D. 1, Sherry Pagoto, PhD2

1
San Diego State University, San Diego, CA; 2University of Massachusetts Medical School,
Worcester, MA

A growing list of U.S. states have enacted restrictions on minors use of indoor tanning;
however, few studies have examined the impact of these restrictions.

In 2014, Alabama passed HB 254, which enacted a series of restrictions on indoor tanning
among minors. Analyzing changes in the prevalence of indoor tanning between 2013 and
2015 may provide some evidence for the impact of HB 254. Moreover, data from Florida were
included as a control, as the only restriction in place in Florida during 2013-2015 is a ban in
indoor tanning for minors under 14.

Data were extracted from the 2015 Alabama and Florida YRBS. These surveys are conducted
every two years and assess health risk behaviors among 9-12 grade public and private school
students. The Alabama and Florida surveys are of representative samples of the entire
population of high school students in their respective states. Indoor tanning was assessed
with the item: During the past 12 months, how many times did you use an indoor tanning
device, such as a sunlamp, sunbed, or tanning booth? (Do not count getting a spray-on tan).
As in previous studies, this item was dichotomized as 0 times vs. 1 or more times. The
total employed sample was N = 14,389 (population N = 1,864,241). Mean age of participants
was 16.10 (SD = 1.20). Given differences in the prevalence of indoor tanning by age and
sex, all analyses were stratified by sex and age. Four levels of age were created: 14, 15, 16-17,
and 18 year olds. This approach allows a test of the various indoor tanning regulations
enacted within Alabama. Logistic regression models were used, with the independent
variables of year (2013 vs. 2015), state (Alabama vs. Florida), and the year by state interaction
term. The dichotomous indoor tanning variable was set as the criterion variable. Complex
Samples in SPSS 24 was used to account for weighting, clustering, and stratification, allowing
for population based estimates.

Female High School Students

ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1705

Within each age group, only significant main effects of state emerged, with Alabama girls
reporting higher indoor tanning compared to Florida girls. No significant main effects of time
or state by time interactions were significant for any age group.

Male High School Students

Within each age group, only significant main effects of state emerged, with Alabama boys
reporting higher indoor tanning compared to Florida boys. No significant main effects of time
or state by time interactions were significant for any age group.

The current study found no significant changes in indoor tanning among teens since the
enactment of Alabamas tanning restrictions. More oversight and monitoring is needed to
assure that indoor tanning facilities are compliant with emerging laws. As the FDA considers a
national ban on minors use of indoor tanning, enforcement interventions will be necessary to
realize its impact.

CORRESPONDING AUTHOR: Aaron Blashill, Ph.D. , San Diego State University, San Diego, CA,
92120; [email protected]
S1706 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 40: Skin Cancer Prevention: Policies, Intentions and Measurement 3:55 PM-4:12
PM

RESULTS OF A CROSS-NATIONAL RANDOMIZED TRIAL ON THE EFFECT OF BUILT SHADE ON USE

OF RECREATION AREAS IN PUBLIC PARKS

David Buller, PhD1, Suzanne J. Dobbinson, BSc, MSc, PhD2, Dallas R. English, PhD2, Melanie
Wakefield, Professor3, Mary K. Buller, M.A.1

1
Klein Buendel, Inc., Golden, CO; 2Cancer Council Victoria, Melbourne, Victoria, Australia;
3
Cancer Council Victoria, Melbourne, Victoria, N/A, Australia

Shade use is recommended to prevent skin cancer. A randomized controlled trial tested the
hypothesis that individuals were more likely to use recreation areas in public parks in which
shade was introduced (i.e., ultraviolet protection factor [UPF]-rated shade cloth structures)
than unshaded control areas. Shade was built in public parks in Denver, Colorado (US) and
Melbourne, Victoria (Australia) to test also if built shade would increase use of the recreation
areas more in Australia with its 25-year-long skin cancer prevention campaign than in the
United States. Public parks (n=144) containing two unshaded passive recreation areas (PRA)
suitable for building shade (study and comparison PRAs) were selected in three annual waves.
Following pretest, parks were randomized in a 1:3 ratio to built shade or unshaded control.
Primary outcome was observation of whether the study PRA was in use during four 30-minute
observations at pretest and posttest on weekends between 11 am and 3 pm in summer
months. Adjusting for clustering of observations within parks, analyses showed that shaded
PRAs (probability of PRA in use: pre=0.09, post=0.31) were more likely to be in use at posttest
than unshaded control PRAs (pre=0.13, post=0.15), odds ratio (OR)=3.87, 95% CI=1.70, 8.82
(adjusted for wave, pretest, weather [temperature, cloud cover, and wind], type of recreation,
comparison PRA use, user age and race [proportion white]). City moderated the impact of
shade on PRA use, OR=2.90, 95% CI=1.07, 7.84. Shade increased the probability of PRAs in
Denver being in use (control: pre=0.17, post=0.18; shaded: pre=0.14, post=0.47) but the
increase was not statistically significant in Melbourne (control: pre=0.10, post=0.13; shaded:
pre=0.05, post=0.17). Purpose-built shade may improve sun protection (i.e., shade use) in
public spaces. Shade can be a good public investment because a) many attractive, protective
designs are available and b) it requires less pre-planning than other sun protection practices
(e.g., clothing and sunscreen). Differences in existing shade, amenities in the parks and near
PRAs, and public sun safety campaigns may explain the unexpectedly higher use of shaded
PRAs in Denver than Melbourne.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1707

CORRESPONDING AUTHOR: David Buller, PhD, Klein Buendel, Inc., Golden, CO, 80401;
[email protected]
S1708 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 40: Skin Cancer Prevention: Policies, Intentions and Measurement 4:13 PM-4:30
PM

EFFECT OF AN INTERVENTION ON OBSERVED SUN PROTECTION BY VACATIONERS IN A

RANDOMIZED CONTROLLED TRIAL AT NORTH AMERICAN RESORTS

David Buller, PhD1, Peter A. Andersen, BA, MS, PhD2, Barbara Walkosz, PhD3, michael Scott,
Phd4, Larry Beck, Ph.D.5, Gary R. Cutter, MS PhD6

1
Klein Buendel, Inc., Golden, CO; 2San Diego State Unversity, Jamul, CA; 3Klein Bendel, Inc.,
Wheat Ridge, CO; 4CSU, Chico, Santa Fe, NM; 5San Diego State University, La Mesa, CA; 6UAB
School of Public Health, Vestavia Hills, AL

High-risk sun exposure often occurs during vacations. An intervention based on

Transportation and Diffusion of Innovation Theories comprised of posters, brochures,
audiovisual, and online (website and social media) messages promoting comprehensive sun
protection (clothing, hats, sunscreen and reliable cues to solar UV) targeted vacationers. It
was evaluated with a sample of 41 North American warm-weather destination resorts in 17
states enrolled in a pair-matched, group-randomized pretest-posttest controlled quasi-
experimental design in 2012-14. Resorts were randomly assigned to receive the intervention
or remain no-treatment controls. Sun protection practices were observed in cross-sectional
samples of vacationers who appeared 18 or older at pretest (n=4,347) and posttest (n=3,986)
by trained research staff during two-day visits to resorts. Vacationers sun protection (i.e.,
composite z-score measure of body coverage and use of shade) did not differ by experimental
condition (p>0.05) or extent of intervention implementation (p>0.05). However, recreation
venue moderated intervention impact. The intervention improved comprehensive sun
protection at waterside (pools and beaches) recreation venues (intervention pre=-22.74,
post=-15.77; control pre=-27.24, post=-23.24) but not at non-waterside (golf, courts, and
dining) venues (intervention pre=20.43, post=20.53; control pre=22.94, post=18.03, p pre=-
25.45, post=-14.05; low implementation pre=-24.70, post=-21.40) compared to non-waterside
(high implementation pre=14.51, post=19.98; low implementation pre=24.03, post=18.98, p <
0.01) recreation venues. The intervention was effective with vacationers in recreation venues
with the highest-risk sun exposure, waterside venues. However, it was not broadly effective
throughout the resorts possibly due to the sun-seeking desires of vacationers, information
overload at resorts, and constraints on clothing styles and sun protection practices by
recreation activities.

CORRESPONDING AUTHOR: David Buller, PhD, Klein Buendel, Inc., Golden, CO, 80401;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1709

Paper Session 41: Understanding Lapses and Impulse Management in Diet and Obesity 3:00
PM-3:18 PM

CITATION AWARD WINNER

PHYSICAL, EMOTIONAL AND SITUATIONAL PREDICTORS OF DIETARY LAPSES IN WEIGHT LOSS

TREATMENT

Cara Dochat, B.A.1, Stephanie Goldstein, M.S.1, Evan M. Forman, PhD1, Meghan L. Butryn,
PhD1, Stephanie Manasse, MS1, Ross D. Crosby, PhD2, Graham Thomas, PhD3

1
Drexel University, Philadelphia, PA; 2Neuropsychiatric Research Institute, Fargo, ND; 3Brown
University & The Miriam Hospital, Providence, RI

Frequency of lapsing from a prescribed weight loss diet predicts weight loss treatment failure,
yet inconsistent findings about factors preceding lapsing have prevented effective treatment
targeting. Lapses are categorically heterogeneous and as such may have unique internal and
external antecedents. The current study uses ecological momentary assessment (EMA) to
examine predictors of three lapse typeseating more than intended, eating a forbidden food,
and eating when not intendedin adults (n=189; 82.0% female; 70.9% Caucasian; Mage =
51.81 9.76 years; MBMI = 36.93 5.83 kg/m2) enrolled in a 12-month randomized controlled
trial comparing the effectiveness of standard behavioral weight loss treatment to an
acceptance-based condition. At baseline, participants completed 14 days of EMA in which
they indicated whether and what type of lapse had occurred since the last prompt, and
responded to questions assessing current physical (hunger, tiredness), environmental
(presence of delicious foods), and affective (loneliness, sadness, stress, boredom, feelings of
deprivation) states. Because lapse frequency did not differ by study condition, conditions
were combined for the following analyses. Separate generalized estimating equations were
used to examine whether each state prospectively predicted lapse occurrence at the next
survey. The majority of lapses were categorized as eating a forbidden food (44%), followed by
eating when one had not intended (28%) and eating a larger portion than planned (27%).
Results indicate that likelihood of eating a forbidden food was greater for those who reported
more frequent aversive physical states (p < .05), and for all participants when delicious foods
were present (p < .01). Likelihood of eating larger portions was greater for those reporting
more frequent negative affective states (p < .01). Eating when not intended was more likely to
occur following momentary aversive affective and physical states for all participants (p <
.01). Results support the theory that distinct lapse types exist, and that lapse types can be
predicted by both momentary conditions and individual tendencies toward certain physical
and affective states. Identifying individual differences in tendencies toward sadness,
S1710 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

tiredness, and hunger, might help participants anticipate conditions for lapsing and develop
specific strategies for their most common lapse types. Future research should examine how
lapse types uniquely affect weight loss outcomes.

CORRESPONDING AUTHOR: Cara Dochat, B.A., Drexel University, Philadelphia, PA, 19130;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1711

Paper Session 41: Understanding Lapses and Impulse Management in Diet and Obesity 3:19
PM-3:36 PM

CITATION AWARD WINNER

A RANDOMIZED TRIAL OF AN ACCEPTANCE BASED BEHAVIORAL INTERVENTION FOR WEIGHT

LOSS IN PEOPLE WITH HIGH INTERNAL DISINHIBITION

Jason Lillis, Ph.D.

Brown Medical School, Rocklin, CA

Background: Most individuals completinga behavioral weight loss program achieve

weight loss of approximately 7-10% over 6-9 months and then gradually regain weight
thereafter. Innovative approaches are needed to improve long-term weight control otucomes.
Standard interventions typically include cognitive and emotional control strategies (e.g.
thought replacement, distraction, relaxation) to support weight loss efforts. However
excessive attempts to change or control thoughts, feelings, or bodily sensations, referred to as
experiential avoidance, could paradoxically make it more difficult for individuals with obesity
to cope with food cravings and instead lead to greater caloric consumption. Alternative
treatment components may be necessary to help individuals cope with situations that can
interfere with successful weight loss maintenance. Acceptance-based approaches (e.g.
Acceptance and Commitment Therapy; ACT) may provide a viable alternative thorugh the use
of mindfunlness and acceptance-based strategies that focus on improving felxibile responding
to unwanted thoughts, feelings, and cravings. The aim of this study was to determine whether
an Acceptance-Based Behavioral Intervention (ABBI) produces better weight losses than
standard behavioral treatment (SBT) among individuals reporting high internal disinhibition.

Methods: Participants were 162 adults with overweight or obesity (mean BMI 37.6) randomly
assigned to ABBI or SBT. Intervention was one year with a fading contact schedule and
counterbalanced interventionists. Both interventions provided the same calorie intake target,
exercise goal, and self-monitoring skills training. SBT incorporated current best practice
interventions for addressing problematic thoughts and emotions. ABBI utilized acceptance-
based techniques based on Acceptance and Commitment Therapy. ABBI and SBT were
compared on weight change and disinhibition change over 24 months.

Results: Mixed models analysis showed mean weight loss at 24 months was -4.1% (SE=0.88)
for ABBI and -2.4% (SE=0.87) for SBT (p=.204). Secondary analyses showed that the ABBI
group regained less weight from the end of treatment to the final follow-up (4.6 vs. 7.1kg;
S1712 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

p=.005), and that a significantly higher proportion of ABBI participants achieved a 5% weight
loss (38% vs. 25%; p=.038) at 24 months.

Conclusions: Taken together, results suggest that ABBI had a positive impact on weight
outcomes by reducing the magnitude of weight regain after treatment ended, suggesting that
ABBI could be helpful for improving the maintenance of weight loss for individuals who report
high internal disinhibition. The results of this trial also point to gaps in our current knowledge
of the utility of using mindfulness and acceptance strategies for improving weight control (e.g.
when is the best time to introduce these strategies/ who is a good fit?).

CORRESPONDING AUTHOR: Jason Lillis, Ph.D., Brown Medical School, Rocklin, CA, 95765;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1713

Paper Session 41: Understanding Lapses and Impulse Management in Diet and Obesity 3:37
PM-3:54 PM

CITATION AWARD WINNER

TIME DELAYS PROMOTE HEALTHIER VENDING MACHINE SNACK CHOICES

Brad Appelhans, PhD1, Simone French, PhD2, Lisa Powell, PhD3, Imke Janssen, Ph.D.1,
Elizabeth Avery-Mamer, MS4

1
Rush University Medical Center, Chicago, IL; 2Div. Epidemiology & Community Health,
Minneapolis, MN; 3University of Illinois at Chicago, Chicago, IL; 4Rush University Medical
Center, Morton Grove, IL

In environments replete with palatable foods, a preference for immediate gratification can
frequently outstrip inhibitory control during food choice. Research on delay discounting
suggests that delaying access to tempting high-calorie foods could potentially shift choice to
healthier (though less desired) options. We tested this principle in the context of a novel
vending machine intervention called DISC (Delays to Influence Snack Choice), which selectively
delays the delivery of less healthy snacks. When less healthy snacks are purchased, DISC
retains the purchased snack on a platform until the specified time delay expires. This
ecological study used a repeated-measures design to test the effects of a 25-second time
delay on purchasing rates for healthy snacks in three vending locations at an urban medical
center. The proportion of total sales from healthy snacks was compared for the following
conditions: 1) no intervention, 2) 25-s time delay on less healthy snacks, 3) 25-cent discount
on healthy snacks, 4) a 25-cent tax on less healthy snacks, 5) 25-s time delay plus 25-cent
discount, 6) 25-s time delay plus 25-cent tax. Each condition ran for roughly one month in
each location. Healthy snacks met 5 of 7 nutritional criteria and composed 50% of items sold
in each machine. The proportion of total sales from healthy snacks in each condition was
compared through generalized estimating equations with 31,827 sales. Relative to no
intervention, 25-s time delays yielded a 1.6% to 5.4% absolute increase (p < .001) in the
proportion of total purchases from healthy snacks, depending on location. Healthy snack
purchasing also increased under both 25-cent discounts and taxes (ps < .001). Combining
time delays with the discount doubled its impact (p < .0001). None of the interventions
involving time delays led to a significant reduction in overall vending sales volume. Brief time
delays influence food choice at the point of decision, do not harm vending revenues, and may
augment differential pricing interventions. Future research should test DISC in a larger
multisite trial, and explore the potential impact of time delays on food choice in other
S1714 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

contexts such as fast food restaurants and online grocers. Funded by R21HL121861.
NCT02359916.

CORRESPONDING AUTHOR: Brad Appelhans, PhD, Rush University Medical Center, Chicago, IL,
60612; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1715

Paper Session 41: Understanding Lapses and Impulse Management in Diet and Obesity 3:55
PM-4:12 PM

CITATION AWARD WINNER

ACCEPTANCE-BASED BEHAVIORAL TREATMENT OUTPERFORMS STANDARD BEHAVIORAL

TREATMENT AT POST-TREATMENT AND LONG-TERM FOLLOW-UP

Evan M. Forman, PhD1, Meghan L. Butryn, PhD2, Stephanie M. Manasse, MS2, Ross D. Crosby,
PhD3, Graham Thomas, PhD4

1
Drexel University, philadelphia, PA; 2Drexel University, Philadelphia, PA; 3Neuropsychiatric
Research Institute, Fargo, ND; 4Brown University & The Miriam Hospital, Providence, RI

Acceptance-based behavioral treatments (ABTs; i.e., approaches that fuse behavioral

strategies with tolerance of difficult internal experiences and loss of pleasure, mindful
decision making and commitment to valued behavior) have been recognized as potential
enhancements to the efficacy of gold standard behavior treatment (SBT) for weight loss.
However, only one full randomized controlled trial testing this hypothesis has been published
to date, and no long-term follow-up has been conducted. The NIH-funded Mind Your Health
Project randomly assigned overweight and obese participants (n = 190) to 25 sessions of
either SBT or ABT. At the 1-year (post-treatment) and 2-year (1-year follow-up) assessments,
and using maximum likelihood imputation, ABT produced significantly greater percent weight
loss (1-year: 13.3% 0.83, 2-year: 7.3% 0.88 ) than did SBT (1-year: 9.8% 0.87, 2-year: 4.6%
0.92; linear effect of treatment condition p = .04). Weight regain from 1 to 2 years was
equivalent by condition. Results support the long-term efficacy of ABT. Mediation effects
support the postulated mechanisms of ABT, and moderation effects suggest that those with
greater impulsivity would especially benefit from assignment to ABT. Future directions
include examining outcomes at an even longer follow-up period and examining whether post-
treatment booster sessions strengthen and/or extend the advantage of ABT.

CORRESPONDING AUTHOR: Evan M. Forman, PhD, Drexel University, philadelphia, PA, 19104;
[email protected]
S1716 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 41: Understanding Lapses and Impulse Management in Diet and Obesity 4:13
PM-4:30 PM

CITATION AWARD WINNER

USING MACHINE LEARNING AND JUST IN TIME ADAPTIVE INTERVENTION TO PREDICT AND
PREVENT DIETARY LAPSES: THE DIETALERT APP

Evan M. Forman, PhD1, Stephanie P Goldstein, M.S.2, Adrienne S. Juarascio, Ph.D.1, Meghan L.
Butryn, PhD1, Pramod Abichandani, PhD.1

1
Drexel University, Philadelphia, PA; 2Drexel University, Philadelphi, NJ

Lack of success in weight loss programs can be largely attributed to dietary lapses, i.e., specific
instances of inadherence to dietary recommendations. Lapses have been linked to a select
group of internal and external cues, suggesting that lapse prediction and prevention are
possible, given sufficient data. As such, we have set out to create a smartphone-based system
called DietAlert that utilizes machine learning to predict lapse behavior with increasing
accuracy, and to deliver tailored, momentary interventions designed to prevent lapses from
occurring during a period of elevated risk. This presentation describes four phases of
development and evaluation which involved overweight participants using DietAlert while
following a Weight Watchers weight loss plan. First, we will describe the development of an
experience sampling component of the app, and will report user compliance (94.6%) and user-
rated satisfaction (6.42/7.00). Second, we will describe the creation and evaluation of a
machine learning predictive model of lapses (e.g., the maximizing of sensitivity without
unacceptably low specificity). As such, we will report the final models strongest predictors of
lapse (e.g., socializing, affect, tiredness), sensitivity (70%) and specificity (72%). Third, we will
report on the development of 159 tailored micro-interventions, and implementation of these
interventions in DietAlert. Fourth, we report on an effectiveness trial in which participants
(current n=20; projected n=40) used a full version DietAlert that included experience
sampling, risk alerts, and momentary and tailored micro-interventions. Preliminary results
suggest that DietAlert is reducing lapse frequency (M = -3.00, SD=5.55) and enhancing weight
loss (M = 3.7%, SD=1.6%) versus weight watchers alone (M ~ 2% in an independent sample),
though the uncontrolled nature of the study makes this conclusion temporary.

CORRESPONDING AUTHOR: Evan M. Forman, PhD, Drexel University, Philadelphia, PA, 19104;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1717

Paper Session 42: Advancing Care for Type 1 Diabetes 3:00 PM-3:18 PM

YOUTH WITH TYPE 1 DIABETES: PSYCHOSOCIAL PREDICTORS OF RISK FACTORS AND DIABETES-
RELATED COMPLICATIONS 11 YEARS LATER

Vicki Helgeson, PhD1, Howard Seltman, Ph.D.1, Trevor Orchard, M.D.2, Dorothy Becker, M.D.3,
Ingrid Libman, M.D.3

1
Carnegie Mellon University, Pittsburgh, PA; 2University of Pittsburgh, Pittsburgh, PA;
3
Children's Hospital of Pittsburgh, Pittsburgh, PA

The goal of this study was to examine the extent to which psychosocial variables measured
during adolescence predicted risk factors and early indicators of diabetes-related
complications at age 22. We followed 132 youth with type 1 diabetes, average age 12 at study
entry, for 11 years with 81% retention. At age 22, we obtained measures of glucose control
(HbA1c), cardiovascular risk factors (BMI, cholesterol, blood pressure) and indicators of
diabetes-related complications (nephropathy [albuminuria], autonomic neuropathy [heart
rate from inhalation and exhalation], peripheral neuropathy [from vibratory readings]). Lower
parental social status at study entry and minority race were related to several risk factors and
complications. Current education achieved predicted outcomes: higher education by age 22
was associated with better glycemic control, lower cholesterol, and lower autonomic and
peripheral neuropathy. Thus, we statistically controlled for parent social status, sex, race, and
current educational achievement in all analyses presented below. Overall quality of
relationship with mother and father predicted multiple risk factors, but there were more
consistent links to father relationship. Higher father relationship quality was associated with
lower BMI (B = -1.18, p < .05), lower levels of non-HDL cholesterol (B = -16.11, p < .01), and
marginally higher autonomic neuropathy (B = .06, p < .10). Friend support at age 12 did not
predict risk factors or complications, but friend conflict did. Friend conflict predicted higher
HbA1c (.42, p < .10), higher autonomic neuropathy (B = -.09, p < .05) and higher peripheral
neuropathy (B = .40, p < .05). Greater depressive symptoms and lower feelings of self-worth at
age 12 predicted increased albuminuria (B = 59.78, p < .05; B = -27.0, p < .05) and peripheral
neuropathy (B = 1.32, p < .05; B = -.59, p < .10). These results suggest that youth with type 1
diabetes who are having difficulties in relationships and experiencing psychological distress
are more likely to experience poor health outcomes many years later. There are a number of
potential pathways by which psychosocial variables at age 12 impact physical health 11 years
later.

CORRESPONDING AUTHOR: Vicki Helgeson, PhD, Carnegie Mellon University, Pittsburgh, PA,
15213; [email protected]
S1718 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 42: Advancing Care for Type 1 Diabetes 3:19 PM-3:36 PM

INCONSISTENT SLEEP IS LINKED WITH POOR DIABETES MANAGEMENT AND GLYCEMIC

CONTROL IN ADOLESCENTS

Niral J. Patel, MPH, Kimberly L. Savin, BA, Lauren A. Williams, BS, Beth A. Malow, MD, MS,
Sarah S. Jaser, PhD

Vanderbilt University Medical Center, Nashville, TN

Background: Prior research has focused on improving adherence and glycemic control in
adolescents with type 1 diabetes (T1D), but the effect of sleep on these outcomes has
received little attention. Adolescents, in general, have difficulty meeting sleep
recommendations (8-10 hours/night) due to changes in circadian rhythms paired with early
school start times, and increased time spent on schoolwork, extracurricular activities, and
peer relationships. Insufficient sleep may have serious consequences for adolescents with T1D
due to its link to poor diabetes management. This study aims to examine the links between
inconsistent and insufficient sleep, measured by objective actigraphy data, and measures of
adherence and glycemic control in adolescents with T1D.

Methods: Participants (n = 64, age 13-17 years, 52% female, mean A1C = 8.91.5%) wore
actigraphs for 7 nights. Actigraphy data were analyzed using Philips Actiware software. Sleep
inconsistency was calculated using the standard deviation of total sleep time across 7 nights,
representing the variation in sleep from night to night. Average daily blood glucose
monitoring (BGM), average blood glucose levels (BG), and percent of blood glucose readings
within target range were obtained through 30 day glucometer downloads. A1C was obtained
from medical charts and adherence was measured using the Self-Care Inventory (SCI).

Results: Bivariate analysis revealed a significant association between inconsistent sleep and
daily BGM (r=-0.27, p < 0.05), such that more inconsistent sleep was associated with less
frequent BGM. Further, inconsistent sleep was correlated with A1C (r=0.34, p < 0.01) and BG
(r=0.30, p < 0.05), such that participants with greater sleep inconsistency had higher levels of
A1C and BG. Additionally, sleep consistency was associated with percent of blood glucose
levels within range (r =-0.26, p < 0.05), indicating that greater sleep inconsistency was linked
with fewer blood glucose levels within a specified target range. Inconsistent sleep was not
significantly associated with other measures of adherence such as the parent-report or self-
report SCI. Average total sleep time was not significantly associated with adherence or
glycemic control.

Discussion: In our sample, inconsistent sleep patterns were associated with poorer glycemic
control and less frequent BGM in adolescents with T1D. These findings provide novel insights
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1719

into the importance of sleep in youth with T1D, suggesting that inconsistent sleep may be
more detrimental than total sleep time, and that clinicians should consider sleep patterns as a
factor that might affect adherence and glycemic control. Future research should aim to
determine the directionality of the relationship between sleep patterns and diabetes-related
outcomes through longitudinal studies, and to understand the role of sleep in adolescents'
ability to successfully manage T1D.

CORRESPONDING AUTHOR: Niral J. Patel, MPH, Vanderbilt University Medical Center,

Nashville, TN, 37232; [email protected]
S1720 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 42: Advancing Care for Type 1 Diabetes 3:37 PM-3:54 PM

PERFORMANCE-BASED AND QUESTIONNAIRE MEASURES OF EXECUTIVE FUNCTION LINKED TO

DIABETES OUTCOMES

Emily R. Hamburger, MEd, Niral J. Patel, MPH, Sachini N. Kahanda, BA

Vanderbilt University Medical Center, Nashville, TN

Background: Type 1 diabetes (T1D) is one of the most prevalent medical conditions affecting
teens. A complex treatment regimen is needed to maintain glycemic control, which requires
components of executive function (EF), such as the ability to shift attention, regulate
emotions, and initiate behavior. Although past studies have focused on EF deficits in teens
with diabetes, our study is unique and expands upon current literature, incorporating both
performance-based and parent-reported measures of EF in relation to glycemic control,
treatment adherence, and fear of hypoglycemia.

Methods: Teens with T1D (n=65, age 13-17, 52.3% female, 87.1% White, mean A1c=8.85%)
and their caregivers completed the Self Care Inventory (SCI), a measure of diabetes
management behaviors where higher scores indicate greater adherence. Parents reported on
their teens EF using the Behavior Rating Inventory of Executive Function (BRIEF), with higher
scores indicating greater dysfunction. Teens completed the Fear of Hypoglycemia Survey
(FHS), which assesses behaviors and worries related to hypoglycemia. After teens checked
their blood sugar, the DelisKaplan Executive Function System (D-KEFS) Trail Making subtest, a
performance-based measure of EF, and the Working Memory Index (WMI) of the Wechsler
Intelligence Scale for Children were administered. Clinical data (i.e., A1c) were collected from
medical charts.

Results: Bivariate analysis showed significant relationships between the Worry subscale on
the FHS and the D-KEFS Trail Making (r=-.282, p < .05), and the Behavior Regulation Index
(BRI) on the BRIEF (r=-.273, p < .05). These relationships were not present for the Behavior
subscale of the FHS. The Emotional Control scale on the BRIEF was found to be significantly
associated with A1c (r=.287, p < .05). Several BRIEF subscales, as well as the BRI (r=-.325, p <
.05) and Global Executive Composite (r=-.308, p < .05) were significantly correlated with the
parent report SCI. These relationships were not observed with the self-report SCI. WMI was
not related to any outcomes.

Discussion: As teens become more responsible for their diabetes care and management, the
role of EF in treatment adherence becomes more important. This study reveals that lower
levels of EF are related to worse treatment adherence, and worries about hypoglycemia are
linked to worse performance on an EF task. EF deficits may make it more difficult for teens to
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1721

adhere to their treatment plans and maintain glycemic control. Our study is unique in its
multimethod assessment of EF, using performance-based and questionnaire measures, which
provide valuable and complementary information regarding separate facets of diabetes
management. Worrys relation to EF is a novel finding, and warrants further investigation.
Future research will need to identify which EF components are most important to specific
health behaviors.

CORRESPONDING AUTHOR: Emily R. Hamburger, MEd, Vanderbilt University Medical Center,

Nashville, TN, 37232; [email protected]
S1722 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 42: Advancing Care for Type 1 Diabetes 3:55 PM-4:12 PM

MULTIMETHOD SLEEP ANALYSIS OF ADOLESCENTS WITH TYPE 1 DIABETES

Kimberly L. Savin, BA, Sachini N. Kahanda, BA, Gray A. Lochbihler, BA, Beth A. Malow, MD, MS,
Sarah S. Jaser, PhD

Vanderbilt University Medical Center, Nashville, TN

While sleeps effect on psychosocial and cognitive functioning is well-established, few

researchers have described sleep habits in adolescents with type 1 diabetes (T1D) and few
clinicians regularly assess sleep habits. This study aimed to provide an in-depth description of
adolescents sleep on school nights and weekend nights using a multimethod (actigraphy,
questionnaire, daily diary self-report) approach to measuring sleep habits.

Participants included adolescents with a diagnosis of T1D for at least 1 year (N=65, mean
age=15.0, mean HbA1c=8.9%, 52.3% female). Participants completed the Pittsburgh Sleep
Quality Index (PSQI), wore an actigraph and kept daily diaries recording sleep, activity, mood,
and blood glucose monitoring (BGM) habits for at least 7 days. Actigraphy was analyzed using
Philips Actiware software. Data were analyzed for each participant for one school night and
one weekend night.

On school nights, actigraphy revealed average bed time was 10:55 PM (range: 8:19 PM-2:51
AM) and average wake time was 6:34 AM. Actigraphy showed only 5.8% of participants slept
for 8 hours on school nights, while 38.8% self-reported 8 hours of sleep. School night diary
data revealed 52.3% of the sample slept with 1 electronic device in the room, 60% reported
nighttime BGM, and 76.9% reported morning BGM. On weekend nights, actigraphy revealed
average bedtime was 11:53 PM (range: 9:21 PM-5:06 AM) and average wake time was 8:46
AM. Actigraphy showed 33.3% of participants slept for 8 hours on weekend nights, while
55.8% self-reported 8 hours of sleep. Weekend night diary data revealed that 60% of the
sample slept with 1 electronic device on in the room, 72.3% reported nighttime BGM, and
83.1% reported morning BGM. Using PSQI data, 41.7% self-reported clinically significant poor
overall sleep quality.

Paired t-tests revealed that participants slept more on weekend nights than on school nights,
using both actigraphy (7:31 on weekends, 6:32 on school nights; p < .001) and daily diaries
(8:45 on weekend nights, 7:45 on school nights; p < .005). Bivariate analysis revealed that
greater actigraphy sleep time was correlated with less caffeine intake on the weekend (r=-.27,
p < .05). Additionally, using PSQI report, better overall sleep quality was correlated with less
stress on school nights (r=.34, p < .05) and better mood on the weekend (r=.26, p < .05).
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1723

These findings suggest that few adolescents with T1D meet sleep recommendations on school
or weekend nights regardless of sleep measure used, though they tend to sleep longer and
report BGM more regularly on weekends than school nights. Clinicians could consider
assessing adolescents sleep habits, particularly on school nights, and encouraging earlier
bedtimes and less caffeine intake. Future studies utilizing objective BGM data would elucidate
the relationship between sleep and diabetes management habits.

CORRESPONDING AUTHOR: Kimberly L. Savin, BA, Vanderbilt University Medical Center,

Nashville, TN, 37232; [email protected]
S1724 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 42: Advancing Care for Type 1 Diabetes 4:13 PM-4:30 PM

PATIENT-PROVIDER COMMUNICATION IN PEDIATRIC TYPE 1 DIABETES CLINICAL CARE

April Idalski Carcone, PhD, Deborah Ellis, PhD

Wayne State University, Detroit, MI

Although widely studied in adult health care, few studies have examined patient-provider
communication(PPC) in pediatric medical care and almost none describe PPC in pediatric
chronic illness. A few pediatric PPC studies in primary care and one recent study in chronic
illness describe a pattern of low child participation(age=10.31.1 years, 53% female, 62%
Caucasian, diagnosed 5.12.2 years, 61% insulin pump, MHbA1c=9.2%1.4, 74% 2-parent
homes, Mdnincome=$50-59K). Three coders used an adapted version of the Turn Allocated
System to characterize who initiated a given communication and who the intended recipient
was(e.g., physician speaks to patient). We coded 10,341 speaking turns; 20% of each
recording was co-coded to assess inter-rater reliability(=.680). Patients speaking turns
comprised only 12% of the interactions whereas caregivers represented 39% and physician
48%, a pattern consistent with previous research. Sequential analysis suggested a significant
pattern of conversational turn-taking. Physicians responded directly to patients physician-
directed turns 63% of the time overall with a pattern of increasing response as patients age
increased: physicians responded directly to 45% of 8-9yr olds communications, 63% of 10yr
olds, 64% of 11yr olds, and 75% of 12yr olds. A similar pattern emerged for patient-directed
physician turns and parent interruptions of these turns. Patients responded to physicians 53%
of the time overall with an increasing pattern as patient age increased: 8-9 year olds
responded 42% of the time, 10yr olds responded 44% of the time, 11yr olds responded 58% of
the time, and 12yr olds responded 72% of the time. Parents interrupted these turns 17% of
the time overall, 26% of the time with 8-9yr olds, 21 with 10yr olds, 14% with 11yr olds, and
8% with 12yr olds. These data provide preliminary information on patterns of PPC in pediatric
specialty care. Further research is needed to understand the content and purpose of parent
interruptions, how PPC varies over time, and how to optimize PPC patterns that promote
patient health outcomes in children with chronic health conditions.

CORRESPONDING AUTHOR: April Idalski Carcone, PhD, Wayne State University, Detroit, MI,
48202; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1725

Paper Session 43: Innovation in Tobacco-Based Interventions 3:00 PM-3:18 PM

CAN SMARTER MHEALTH PROGRAMS IMPROVE ENGAGEMENT, ADHERENCE, AND

SMOKING CESSATION OUTCOMES?

Jennifer McClure, PhD1, Melissa Anderson, MS1, Kathy Bradley, MD, MPH1, Larry An, MD2,
Sheryl Catz, PhD3

1
Group Health Research Institute, Seattle, WA; 2University of Michigan, Ann Arbor, MI;
3
University of California - Davis, Sacramento, CA

Most mHealth smoking cessation apps are not particularly smart. Their content does not
follow best practice clinical guidelines and they do not take advantage of smartphone
capabilities to adaptively tailor content or allow two-way communication between users and
clinicians. We developed a comprehensive, interactive mHealth smoking cessation and
medication adherence program (MyMAP; My Mobile Advice Program) that combines standard
cognitive-behavioral self-help with real-time adaptively tailored cessation and adherence
support/advice, and ability to exchange secure messages with a cessation counselor. MyMAP
was pilot tested (n=66; 56% female) against an identical mHealth program minus the
interactive and adaptive support/advice and secure messaging features when both were
paired with varenicline. Those randomized to MyMAP completed self-initiated and prompted
brief check-in surveys about smoking status, cessation medication use, motivation/confidence
for quitting, and common adverse events (AEs; withdrawal symptoms/medication side-
effects). Participants then received tailored support/advice based on this information. Secure
messages were initiated by participants or by counselors, in response to reported AEs, as
needed. Outcome measures of engagement, adherence, and cessation were all positive at 5
months. MyMAP was used more often than the control program (10.6 vs. 2.7 mean log-ins, P
=.0001), driven by use of the experimental features (tailored support/advice and secure
messaging). Experimental participants received help managing 258 withdrawal
symptoms/medication side-effects and exchanged 130 secure messages (22 message strings
initiated by smokers; 44 message strings initiated by clinicians in response to concerns raised
during check-in surveys). Most thought MyMAP could help others quit smoking (92%),
consistently take their medication (97%), and would recommend it (87%). More MyMAP
participants tended to be optimally adherent (36% vs. 24%) and quit smoking (36% vs. 24%;
ORs for each: 1.79; CI: 0.61-5.19). In sum, MyMAPs proactive, adaptive and interactive
treatment approach showed promising effects on engagement, adherence and cessation.
Results have implications for designing future mHealth cessation programs.
S1726 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Jennifer McClure, PhD, Group Health Research Institute, Seattle,
WA, 98101; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1727

Paper Session 43: Innovation in Tobacco-Based Interventions 3:19 PM-3:36 PM

SMOKERS WHO ARE NOT MOTIVATED TO QUIT ARE MORE LIKELY TO QUIT SMOKING WITH
MORE INTENSIVE VS. LESS INTENSIVE INTERVENTIONS.

Romano Endrighi, PhD1, Shira Dunsiger, PhD2, Belinda Borrelli, PhD3

1
Boston University Henry M. Goldman School of Dental Medicine, Boston, MA; 2The Miriam
Hospital and Brown University, Portsmouth, RI; 3Boston University, Henry M. Goldman School
of Dental Medicine, Boston, MA

Parents who smoke and are not motivated to quit are an important group for intervention,
particularly if they have a child with asthma. Previous research indicates that smokers who are
not motivated to quit will be less responsive to intensive smoking interventions, although
motivation by treatment interactions have not been directly tested. The current study is a
secondary analysis examining whether motivation to quit among parents with asthmatic
children moderates the effect of a smoking cessation induction intervention. We hypothesized
that smokers who were not motivated to quit would be more likely to quit smoking and
reduce secondhand smoke exposure (SHSe) in a more intensive intervention vs a less
intensive one.

Parents had a child with an asthma episode requiring urgent care. They were told that they
would receive two home-based asthma education visits and discuss their smoking, but they
did not have to want to quit smoking to enroll. Home visits included asthma education,
motivational interviewing (MI) for smoking cessation and feedback on child SHSe. After the
home visits, participants were randomized (n=339, 79.6% female) to receive either enhanced-
PAM (Precaution Adoption Model; 6 counseling calls using MI for cessation induction and
repeated SHSe feedback) or PAM (6 contact control calls with checks on asthma status) over 4
months. Motivation to quit within 30-days (yes/no) was assessed at baseline. Outcomes were
bioverified 7-and 30-day point prevalence abstinence (PPA) at 6 months, and SHSe was
measured by passive dosimetry (very low/undetectable vs moderate/high).

At baseline, 38.9% were not motivated to quit. There was a significant treatment group by
motivation interaction: Not-motivated smokers in enhanced-PAM were 3-4 times more likely
to be abstinent vs. not-motivated smokers in PAM (7-day PPA OR=3.71, 95%CI=1.06, 12.99;
30-day PPA OR=4.15, 95%CI=1.20, 14.35). Smoking outcomes did not significantly differ
between treatment arms among motivated smokers. Not-motivated smokers in enhanced-
PAM were more than 4 times as likely to have very low/undetectable SHS at follow-up vs not-
motivated smokers in PAM (OR=4.46, 95%CI=1.31, 15.15). SHSe did not significantly differ
S1728 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

between treatment arms among motivated smokers.

Thus, parents who were not-motivated to quit achieved better outcomes with the more
intensive treatment. It cannot be assumed that smokers who are unmotivated to quit will not
be responsive to intensive interventions.

CORRESPONDING AUTHOR: Romano Endrighi, PhD, Boston University Henry M. Goldman

School of Dental Medicine, Boston, MA, 02118; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1729

Paper Session 43: Innovation in Tobacco-Based Interventions 3:37 PM-3:54 PM

IMPLEMENTATION OF MHEALTH TECHNOLOGY FOR HIGH-RISK INPATIENT UNITS:

TECHNOLOGY INPATIENT PROGRAM FOR SMOKERS (TIPS)

Amanda Blok, PhD, MSN, PHCNS-BC1, Rajani Shankar Sadasivam, PhD2, Thomas Houston, MD,
MPH3

1
University of Massachusetts Medical School, Cambridge, MA; 2Univ of Massachusetts
Medical School, Worcester, MA; 3University of Massachusetts Medical School, Worcester, MA

Background. Tobacco use increases the likelihood of hospitalization for cardiac and
pulmonary patients. Hospital stays for these conditions are teachable moments, representing
windows of opportunity for patients to address their risk behavior. The Technology
Inpatient Program for Smokers (TIPS) is a short messaging system (SMS) texting program with
motivational messages for hospitalized smokers. The TIPS implementation program included
1) engaging hospital stakeholders and 2) motivating TIPS usage through a promotional poster
campaign. The goal of this project was to evaluate the implementation.

Methods. The implementation was evaluated quasi-experimentally for 15 weeks on 2 hospital

wards at each of 2 sites. Process measures of the implementation program included weekly
site observation and stakeholder meetings on the wards participating. Reach was calculated
using admission rates of current smokers from electronic medical records and centralized logs
of text messages. Administrative documents were used for a cost analysis.

Results. Posters were hung in patient rooms and hallways on wards by hospital staff. On
weekly rounds, a mean of 10.2% (SD: 8.7%) posters went missing from patient rooms and had
to be replaced. Implementation challenges included coordinating the poster campaign and
establishing leadership buy-in, but were not detrimental to overall Program rollout. Per
hospital floor, 29 to 51 smokers (Mean: 34.7, SD: 8.8) were admitted per week. After 15
weeks, 30 smokers enrolled in the system. Thus, estimating total smokers admitted in 15
weeks, 5.7% (30/520) enrolled in the system. The initial sunk cost of the TIPS intervention,
including posters and stakeholder meetings, was $2238.89. There was a small marginal cost
per smoker of $6.32 ($189.49/30).
S1730 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Discussion. Smokers participated in TIPS with only limited interventional effort and small
marginal costs, although proportionately few smokers signed up for TIPS. Further research will
be conducted to examine the effects of more intensive implementation efforts, including
incorporation of a TIPS Program orientation into nurse workflow.

CORRESPONDING AUTHOR: Amanda Blok, PhD, MSN, PHCNS-BC, University of Massachusetts

Medical School, Cambridge, MA, 02141; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1731

Paper Session 43: Innovation in Tobacco-Based Interventions 3:55 PM-4:12 PM

PROJECT CONNECT: HELPING TOBACCO USERS STAY QUIT AFTER DISCHARGE WITH BEDSIDE
VISITS FROM TRAINED VOLUNTEERS

Denise S. Taylor, MS, RD1, Elisabeth G. Bradley, MS, APRN, AGPCNP-BC, ACNS-BC2, Dominique
Comer, PharmD, MS3, Adebayo Gbadebo, MBA3, Zugui Zhang, PhD3, Claudine Jurkovitz, MD,
MPH3, Freda Patterson, PhD4, Edward Goldenberg, MD1

1
Christiana Care Health System, Wilmington, DE; 2Christiana Care Health System, Lincoln
University, PA; 3Christiana Care Health System, Newark, DE; 4University of Delaware, Newark,
DE

In-patient hospital care represents a window of opportunity to initiate smoking cessation.

However, a lack of cessation support upon discharge limits the extent to which initial
abstinence becomes sustained cessation. To address this gap in care, the current study
evaluated the efficacy of a systems-based smoking cessation program Project Connect-
designed to provide a continuity of smoking cessation care from the bedside back to the
community. Between February and June, 2016, hospital-trained volunteers (former smokers,
former nurses, health behavior science students) administered a brief, smoking cessation
protocol based on the ASK-ADVISE-CONNECT model to active smokers admitted to any
cardiology unit of a single large hospital. If the patient accepted, a fax-referral form was
submitted to the Delaware state quitline to initiate quitline cessation services within 72 hours
of discharge. A total of 434 patients across 462 admissions were included in the study. The
sample was 40% female, 75% white, and on average 57.5 years of age. Across admissions, 30%
resulted in a fax referral to the quitline. Of the patients with complete 90-day follow up data,
50% (122/244) indicated they were ready to quit in the next 6 months. Of those, 49 (40%)
accepted services and enrolled in treatment. To date, 77 patients have reported not using
tobacco within the past 30 days at 90 days post discharge. Project Connect is a novel approach
to help link patients to evidence-based treatment for their tobacco use. Costs are kept low by
using trained volunteers and connecting patients to a state-funded Quitline that offers free
choice of seven FDA-approved cessation medications and coaching over the telephone or in
person. Project Connect represents a value-based approach by treating tobacco use as a
chronic disease that is addressed in the hospital and managed through referral to free
resources in the community.

CORRESPONDING AUTHOR: Denise S. Taylor, MS, RD, Christiana Care Health System,
Wilmington, DE, 19810; [email protected]
S1732 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 43: Innovation in Tobacco-Based Interventions 4:13 PM-4:30 PM

CITATION AWARD WINNER

INCENTIVES FOR HELPING MEDICAID RECIPIENTS QUIT SMOKING: A RANDOMIZED

CONTROLLED TRIAL

Shu-Hong Zhu, PhD1, Christopher M. Anderson, BA1, Sharon E. Cummins, PhD1, Anthony C.
Gamst, PhD1, Neal D. Kohatsu, MD2

1
University of California San Diego, La Jolla, CA; 2California Health and Human Services
Agency, Sacramento, CA

Medicaid recipients smoke at nearly double the rate of the general population and have more
difficulty quitting. Treatment for smoking-related disorders costs Medicaid over $40 billion/
year. CMS launched an initiative to examine incentives for the prevention of chronic disease.
This study was part of that initiative and examined the effect of providing incentives to help
Medicaid smokers quit smoking; the study was embedded in the California Tobacco Quitline.

Quitline callers with Medicaid who consented to the study (N=3,816) were randomly assigned
to: Usual Care (UC), Nicotine Patches (NP), or NP plus Incentives (NP+I) using a ratio of
1.0:1.4:1.4. All groups were provided standard quitline counseling. UC subjects could receive
free nicotine patches by taking a doctors prescription to their pharmacy. NP subjects were
sent free nicotine patches directly to their home by the quitline to minimize barriers. NP+I
subjects were mailed patches directly and were provided with small monetary incentives to
participate in counseling (up to $60 for a complete course of 5 counseling sessions).

Subjects were evaluated on quit attempts made, self-reported prolonged abstinence (30-day
abstinence at 2 months and 180-day abstinence at 7 months), and use of nicotine patches and
other quitting aids.

Incentives increased the median number of counseling sessions delivered by 1. Mailing

patches directly increased use of quitting aids from 68.4% to over 90%. Quit attempt rates
were 88.3%, 91.8%, and 94.5% for the UC, NP, and NP+I groups, respectively. At 7 months,
prolonged abstinence rates were 14.1%, 15.8%, and 19.3%, for the three groups,
respectively. For both quit attempt rates and prolonged abstinence rates, the difference
between UC and NP+I was significant. Results for NP were not significantly different from
either condition.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1733

Barrier free access to patches and small monetary incentives for counseling increase quit
attempts and success for Medicaid recipients. The policy implications of these findings will be
discussed.

Supported by the Centers for Medicare and Medicaid Services (Grant #1B1-CMS330883)

CORRESPONDING AUTHOR: Shu-Hong Zhu, PhD, University of California San Diego, La Jolla,
CA, 92093; [email protected]
S1734 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 44: Correlates of Cardiovascular Disease 3:00 PM-3:18 PM

DEPRESSIVE SYMPTOM PROFILES, CARDIOMETABOLIC RISK AND INFLAMMATION: RESULTS

FROM THE MIDUS STUDY

Diana A. Chirinos, PhD1, Kyle W. Murdock, PhD2, Angie S. LeRoy, MA1, Ryan L. Brown, N/A3,
Christopher P. Fagundes, PhD1

1
Rice University, Houston, TX; 2Rice University, Houson, TX; 3Bryant University, Houston, TX

Depression is an important predictor of cardiometabolic risk; however; the presentation of

depressive symptoms can be highly heterogeneous. In this study, we used latent profile
analysis (LPA) to (1) provide a comprehensive characterization of depressive symptoms
profiles in a healthy sample of adults, and (2) examine their association with cardiometabolic
outcomes, including metabolic syndrome (MetS) and obesity. Participants were 1180 adults
enrolled in the MIDUS-II biomarker study. We used LPA to derive depressive symptom profiles
using subscales of the Mood and Anxiety Symptom Questionnaire, the Center for
Epidemiologic Studies Depression Scale subscales and the Pittsburgh Sleep Quality Index.
Blood was drawn to assess C-reactive protein as a marker of inflammation. Four depressive
symptom profiles were identified. The Low Symptoms subgroup (56.66% of the sample) had
low scores across all subscales. The Somatic Complaints subgroup (25.93%) was
characterized by lower scores across indicators, with subscales measuring somatic symptoms
being the highest within group. The Negative Affect subgroup (11.78%) had high scores
across subscales with those measuring negative affect/loss of interest being the highest.
Finally, the Acute Symptoms subgroup (2.63%) was characterized by high scores on all
subscales. After controlling for sociodemographic characteristics, membership to both the
Somatic Complaints and the Negative Affect subgroups was associated with greater odds
of presenting MetS (ORSom=1.384, p=0.037; ORNegA=2.088, p=0.001) and obesity
(ORSom=1.392, p=0.032; ORNegA=1.934, p=0.001). Inflammation partially attenuated the
association between the Somatic Complaints subgroup and cardiometabolic outcomes,
while associations for the Negative Affect subgroup remained independent. This is the first
study to identify empirically-derived depressive symptom profiles among healthy mid-life
adults in the US and link them to cardiometabolic outcomes. Future work should examine
potential differential pathways to increased risk across depressive symptom profiles and
examine whether tailored interventions have an impact on cardiometabolic risk.

CORRESPONDING AUTHOR: Diana A. Chirinos, PhD, Rice University, Houston, TX, 77098;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1735

Paper Session 44: Correlates of Cardiovascular Disease 3:19 PM-3:36 PM

MERITORIOUS AWARD WINNER

PREOPERATIVE DEPRESSION AND STRESS PREDICT POSTOPERATIVE CARDIOVASCULAR EVENTS

IN PATIENTS UNDERGOING VASCULAR SURGERY

Mallory Rutigliano, N/A1, Tanya Spruill, PhD1, Yu Guo, MA2, Adriana Perez, N/A3, Maya Rubin,
BA4, Caron Rockman, MD, FACS5, Jeffrey S. Berger, MD, MS1

1
NYU School of Medicine, New York, NY; 2New York University school of medicine, New York,
NY; 3N/A, New York, NY; 4New York Medical College, Valhalla, NY; 5New York University
Langone Medical Center, New York, NY

Preoperative depression and stress predict postoperative cardiovascular events in patients

undergoing vascular surgery

Background: Depression and psychosocial stress are associated with cardiovascular morbidity
and mortality, including negative outcomes following cardiac surgery. Hypothesized
mechanisms include poor lifestyle, high blood pressure and increased platelet activity These
mechanisms are also involved in peripheral vascular disease, but little is known about the
effects of psychosocial factors on vascular surgery outcomes. We examined the association
between depressive symptoms and psychosocial stress and cardiovascular events 30 days
after vascular surgery.

Methods: The study sample included 181 adult patients with vascular disease undergoing
non- emergent vascular surgery at NYU Medical Center and Bellevue Hospital. Validated single
item measures of depressive symptoms, perceived stress at home or work, and financial
stress were administered preoperatively and were dichotomized (high versus low) for this
analysis. The two stress items were combined to provide a measure of elevated stress of any
kind. The primary outcome was the 30-day event rate, a composite of stroke/transient
ischemic attack, myocardial infarction, death and troponin elevation.

Results: Mean age was 70 9 and the sample included 29.3% females and 17.7% racial/ethnic
minorities. The 30-day composite event rate was 21.1%. Prior to surgery, 54.2% of
participants reported high depressive symptoms and/or high stress. These participants had
almost twice as many postoperative events as those reporting low depression and low stress
(24.2% vs. 13.2%, p=.07). After adjusting for age, sex, race/ethnicity, body mass index,
smoking status, type of surgery and prior coronary artery disease or myocardial infarction,
S1736 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

high stress/depression was associated with higher odds of postoperative events (OR 3.44, 95%
CI, 1.28 to 9.24). The odds were similar in participants reporting both elevated stress and
depression 3.61 (0.9, 14.55) and those reporting elevated stress or depression 3.38 (1.2, 9.54),
compared with those reporting low stress and low depression.

Conclusions: These results support the hypothesis that high pre-operative stress or depressive
symptoms may increase risk for negative outcomes after vascular surgery.

CORRESPONDING AUTHOR: Mallory Rutigliano, N/A, NYU School of Medicine, New York, NY,
10016; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1737

Paper Session 44: Correlates of Cardiovascular Disease 3:37 PM-3:54 PM

IMPACT OF ANTIDEPRESSANTS ON RISK FOR MAJOR ADVERSE CARDIOVASCULAR EVENTS IN

PATIENTS WITH AND WITHOUT CARDIOVASCULAR DISEASE

Kim L. Lavoie, PhD1, Nicola J. Paine, PhD2, Roxanne Pelletier, PhD3, Andre Arsenault, MD4,
Tavis Campbell, PhD5, Louise Pilote, MD, MPH, PhD6, Simon L. Bacon, PhD2, Jean G. Diodati,
MD7

1
University du Quebec a Montreal, Montreal, PQ, Canada; 2Concordia University, Montreal,
PQ, Canada; 3Research Institute McGill University Health Centre, Montreal, PQ, Canada;
4
Montreal Behavioural Medicine Centre, Montreal, PQ, Canada; 5University of Calgary,
Calgary, AB, Canada; 6McGill University, Montreal, PQ, Canada; 7Hopital Sacre Coeur de
Montreal, Montreal, PQ, Canada

Background: The American Heart Association (AHA) has endorsed depression as a cardiac risk
factor and recommends screening as part of routine practice. This has been met with
controversy due to a paucity of data linking depression treatment to better cardiovascular
outcomes. This study prospectively assessed the association between antidepressant
treatment (defined as being prescribed antidepressants) and major adverse cardiovascular
events (MACE) in men and women undergoing exercise stress tests.

Methods: 2385 participants were recruited on the day of their exercise stress test between
1998 and 2002, and were followed for 8.8yrs. Participants underwent a sociodemographic,
medical, and psychiatric interview (PRIME-MD) and completed the Beck Depression Inventory
(BDI). Antidepressant use was self-reported and verified via chart review. Outcomes included
MACE from administrative databases.

Results: A total of 39% (n=921) of the sample had depression and 8% (n=190) were taking
antidepressants at baseline.Adjusted cox regression analyses revealed that antidepressant use
was associated with a 30% reduced risk of MACE (HR=0.697; 95%CI =0.5040.964; p=.029).
There was no significant interaction between antidepressant use and sex ( = -0.07, 95%CI; ()
= -0.73 0.60; p=.85).

Conclusions and Relevance: Findings indicate a cardio-protective effect of antidepressants on

risk for MACE over an 8.8-year follow-up among patients presenting for exercise stress tests
S1738 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

independent of traditional risk factors, CVD history and baseline depression, which did not
differ by sex.

CORRESPONDING AUTHOR: Kim L. Lavoie, PhD, University du Quebec a Montreal, Montreal,

PQ, H2X 3P2; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1739

Paper Session 44: Correlates of Cardiovascular Disease 3:55 PM-4:12 PM

CLUSTERING OF SOMATIC SYMPTOMS OF DEPRESSION, HRV, AND INFLAMMATION IN

PATIENTS WITH ASYMPTOMATIC HEART FAILURE

Laura Redwine, Ph.D1, Kathy Wilson, MS1, Meredith A. Pung, Ph.D1, Brook Henry, Ph.D.1, Suzi
Hong, Ph.D.2, Paul Mills, Ph.D1

1
University of California, San Diego, La Jolla, CA; 2University of California San Diego, La Jolla,
CA

Stage B, asymptomatic heart failure (HF) presents a therapeutic window for attenuating
disease progression and development of HF symptoms, since mortality rates increase
dramatically when there is progression to symptomatic HF. By identifying biomarkers that
cluster with somatic symptoms of depression, interventions may be developed to target this
high risk group. Methods. Participants were 44 patients with Stage B HF (mean age = 66.2 +/-
7.6 and left ventricular ejection fraction (LVEF%) = 62.8 +/- 8.48) assessed for somatic and
cognitive symptoms of depression, Beck Depression Scale (BDI), heart rate variability (HRV)
(EquaVital, EQ01 heart rate monitor) and plasma inflammatory biomarker, c-reactive protein
(CRP). Results. Regression analyses, adjusting for EF revealed that elevated somatic symptoms
of depression predicted increased HRV measures, RMSSD (delta R2 = .104, p = .037) and HF
(delta R2 = .145, p = .02), and inflammation biomarker CRP (delta R2 = .072, p = .042). A
hierarchical cluster analyses was performed whereby a Dendrogram using Ward Linkage
showed that indeed somatic but not cognitive symptoms of depression clustered with HRV
and CRP. Conclusions. In patients with Stage B heart failure, somatic symptoms of depression
were associated with increased HRV and a marker of inflammation. Somatic symptoms of
depression, potentially denoting sickness behavior associated with increased inflammation,
may result in increased parasympathetic activity to conserve energy.

CORRESPONDING AUTHOR: Laura Redwine, Ph.D, University of California, San Diego, La Jolla,
CA, 92093; [email protected]
S1740 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 44: Correlates of Cardiovascular Disease 4:13 PM-4:30 PM

CHILDHOOD TRAUMA, PERCEIVED STRESS, AND HAIR CORTISOL IN ADULTS WITH AND
WITHOUT CARDIOVASCULAR DISEASE

Stphanie Boss1, Bianca D'Antono, PhD2

1
Montreal Heart Institute/Universit de Montral, Montral, PQ, Canada; 2Montreal Heart
Institute/Universit de Montral, Ste-Julie, PQ, Canada

Background: Traumas experienced in childhood have been associated with greater

psychological or physical morbidity, including a greater risk of developing cardiovascular
disease (CVD). It has been hypothesized that this may reflect, at least in part, disturbances in
how stress is later perceived and regulated.

Objective: To examine the associations of childhood trauma with perceived stress over the
past two years and 3-month hair cortisol levels in adults with and without CVD. Whether
associations differ as a function of sex, age or CVD status is also examined.

Methods: 745 men and 453 women (Age=65 7 years) with and without (N=581) CVD
completed the Childhood Trauma (CTQ) and Perceived Stress Questionnaires. Cortisol
concentrations from 3-cm segments of hair were assessed in 642 participants. Analyses
involved Pearson correlations and hierarchical regressions that controlled for pertinent
covariates.

Results: Moderate to severe childhood trauma was experienced by 374 participants, and was
reported to a similar extent irrespective of sex, age, or CVD status. Women reported more
stress but had lower hair cortisol levels compared to men (all p < 0.001). Those with CVD
exhibited greater cortisol (p < 0.001) but similar stress levels compared to participants
without CVD. Childhood trauma was associated with higher perceived stress levels over the
past two years (r=0.344, p=0.01; =0.309, p < 0.001), but not with hair cortisol concentration
in both univariate and multivariate analyses. Age moderated the relation between childhood
trauma and perceived stress (=-0.061, p=0.023).

Conclusion: Younger individuals who experienced trauma in their youth reported more
chronic stress in late adulthood. However, childhood trauma was not associated with greater
accumulation of cortisol in hair over the past 3 months. Whether this reflects more or less
intact HPA regulation in those having lived trauma in childhood or reflects the characteristics
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1741

of our sample (e.g. 81.1% over the age of 60, 94.5% with at least one illness) requires further
investigation.

Keywords: cardiovascular disease, childhood trauma, hair cortisol, perceived stress

CORRESPONDING AUTHOR: Stphanie Boss, Montreal Heart Institute/Universit de

Montral, Montral, PQ, H3T 1J3; [email protected]
S1742 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 45: Racial Concordance and Provider Communications 3:00 PM-3:18 PM

CITATION AWARD WINNER

RACIAL ATTITUDES AND NONVERBAL CONVERGENCE IN RACIALLY DISCORDANT ONCOLOGY

INTERACTIONS

Lauren M. Hamel, PhD1, Mark A. Manning, PhD2, Louis A. Penner, PhD1, Robert Chapman,
MD3, Terrance L. Albrecht, PhD1, Eggly Susan, PhD1

1
Wayne State University/Karmanos Cancer Institute, Detroit, MI; 2Karmanos Cancer
Institute/Wayne State University School of Medicine, Detroit, MI; 3Josephine Ford Cancer
Center, Detroit, MI

Background: Patient and physician racial attitudes (e.g., physician implicit racial bias; patient
suspicion of medical care racial/ethnic minorities receive) affect perceptions of
communication in racially discordant (Black patients, non-Black physicians) medical
interactions. Racial attitudes are likely expressed and reinforced through nonverbal behaviors.
The Communication Accommodation Theory predicts that nonverbal convergence/divergence
reflects relational convergence/divergence over time in interpersonal interactions. However,
prior research has not investigated the association between nonverbal
convergence/divergence and racial attitudes. We investigated the relationship between
patient and physician racial attitudes and the convergence/divergence of nonverbal behaviors
in racially discordant oncology interactions.

Method: Data were 99 racially discordant video-recorded oncology interactions. Five one-
minute slices from each interaction were extracted, randomized, and coded by trained coders
(N=495 slices). Coded behaviors were selected based on the research literature and
community members input. The coding system assessed frequency/duration of facial
behaviors (e.g., smiling, eye gaze) and body behaviors (e.g., lean, orientation to the other, and
openness). Longitudinal dyad discrepancy score modeling (i.e., scores nested within patient-
physician dyads) was used to determine discrepancy in nonverbal behaviors between patients
and physicians. A three-level longitudinal multilevel model (i.e., time nested within patient,
patient nested within physician) was used to determine to what extent racial attitudes
influence nonverbal convergence/divergence across each oncology interaction for each
behavior.

Findings: Results showed that racial attitudes of patients and physicians influenced the degree
of convergence of facial, but not body, behaviors across the oncology interactions. Physicians
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1743

lower in implicit racial bias converged with their patients more on frequency of smiling than
physicians higher in implicit bias (p < .05). Patients higher on medical suspicion converged
more with their physicians on duration of eye gaze than patients lower on medical suspicion
(p = .07). No significant relationships were found between racial attitudes and body behaviors.

Discussion: Findings suggest racial attitudes may manifest in facial behaviors, lending partial
support to the Communication Accommodation Theory. Next steps including applying the
final phase of the coding system, identification of paraverbal behaviors, to further determine
to what extent racial attitudes impact nonverbal convergence/divergence throughout
oncology interactions, and to link nonverbal behaviors to interaction outcomes.

CORRESPONDING AUTHOR: Lauren M. Hamel, PhD, Wayne State University/Karmanos Cancer

Institute, Detroit, MI, 48201; [email protected]
S1744 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 45: Racial Concordance and Provider Communications 3:19 PM-3:36 PM

RACE, AGE, AND GENDER DIFFERENCES IN CLINICAL TRIAL AWARENESS, PARTICIPATION, AND
SOCIAL SUPPORT IN A SAMPLE OF U.S. ADULTS

Aisha T. Langford, PhD, MPH, Azizi Seixas, PhD, MS

New York University School of Medicine, New York, NY

Background: Historically, racial/ethnic minorities, older adults, and women have been
underrepresented in clinical trials (CTs) and health research studies. Limited knowledge and
awareness, medical mistrust, logistical barriers, and lack of invitation are some of the
commonly cited reasons for these disparities. However, it is not clear how, if at all, these
factors may differ by demographics.

Objective: To assess race, age, and gender differences in awareness, participation, and social
support for CT participation in an online sample of adults across the United States.

Methods: Black (n=1513) and white (n=361) adults were recruited through Survey Sampling
International and completed an online survey about diabetes, health behaviors, CT awareness
and participation, and attitudes about research. Age range of participants was 30 to 85. With
regard to gender, 910 were male and 949 were female. Chi-square tests were conducted for
the following items: 1) Have you ever participated in a research study; 2) Have you ever heard
of a CT; 3) Have you ever been asked by a healthcare provider to participate in a CT; 4) Have
you ever tried to participate in a CT but told that you were not eligible; 5) If you could guess,
how many of your friends and family would participate in a CT (none, some, most, all); and 6)
If you decided to participate in a CT, how many of your friends and family would support that
decision (none, some, most, all)?

Results: There were significant differences between whites and blacks for having heard of a CT
(86% vs. 81%), 2(df=1,N=1865)=4.5, p=.03. There were also racial differences for how many
friends and family would support one's decision to join a CT, 2(df=3,N=1863)=24.75, p < .001,
with whites reporting a higher percentage of most compared with blacks. Among blacks
exclusively, there were significant age differences. Compared with younger black adults, older
black adults were more likely to answer yes to the following items: having participated in a
research study, 2(df=1,N=1481)=4.81, p=.02; heard of a CT, 2(df=1,N=1482)=36.59, p < .001;
ever asked by a provider to participate in a CT, 2(df=1,N=1480)=26.85, p < .001; and tried to
participate in a CT but told you were not eligible, 2(df=1,N=1474)=3.93, p=.04. Additionally,
perceptions of how many friends and family would participate in CTs (i.e., "most" or "all") was
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1745

higher among younger black adults compared with older black adults, 2(df=3,N=1483)=23.12,
p < .001. There were no differences between black men and women for any of the items
evaluated.

Conclusion: More research on the role of social support on CT decision making is needed.
Efforts to raise CT awareness and improve participation of diverse groups may benefit from
targeted approaches that incorporate demographic, health literacy, and psychosocial factors.

CORRESPONDING AUTHOR: Aisha T. Langford, PhD, MPH, New York University School of
Medicine, New York, NY, 10016; [email protected]
S1746 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 45: Racial Concordance and Provider Communications 3:37 PM-3:54 PM

THE INFLUENCE OF HEALTH PROFESSIONAL CONTROL AND QUALITY OF CARE ON WELL-BEING

AMONG BREAST CANCER SURVIVORS

Alexis D. Abernethy, PhD1, Leah S. Fortson, MA2, Nathan Flores, BA3, Kimlin Ashing, Ph.D4

1
Travis Research Institute/ Fuller Theological Seminary, Pasadena, CA; 2Fuller Graduate School
of Psychology, Pasadena, CA; 3Fuller Theological Seminary - Graduate School of Psychology,
Pasadena, CA; 4City of Hope National Medical Center, Duarte, CA

Recent work on locus of health control has clarified the role of both passive and active
dimensions of external locus of control in African American cancer survivors. Locus of control
and perceptions of professional care were examined in African American and Latina breast
cancer (BC) survivors. The hypotheses were as follows: compared to individuals low on quality
of professional care, individuals high on quality of professional care will endorse greater well-
being; and individuals who endorsed higher professional control and higher internal control
(for both control scores, lower scores = higher control) will endorse higher quality of
professional care. This study used baseline data from a parent study examining the effects of
a psychosocial intervention to reduce distress. Participants were recruited from state and
local hospital cancer registries, including the City of Hope cancer registry. The sample included
320 women, 28% African American and 73% Latina, age 26 89 years, and 1-5 years post BC
diagnosis. High quality of professional care was associated with greater well-being ( = .67, p <
.001). Professional control was correlated with professional care. BC survivors who viewed
their doctors as being less in control of their health endorsed greater quality of professional
care (r = .15, p < .01). Internal control was inversely correlated with professional care as
survivors who had high internal control reported greater quality of professional care (r = -.25,
p < .001). There was no significant interaction between professional care and professional
control on well-being. Professional care varied by ethnicity, F(1, 311) = 18.99, p < .001, as
African American survivors endorsed higher quality of professional care than Latina survivors.
English speakers endorsed higher quality of professional care than Spanish speakers, F(1, 311)
= 24.25, p < .01. These results may be partially explained by language differences and Latina
BC survivors lower scores on professional care. These findings support the relationship
between internal control and professional care, but the findings on professional control and
care are unexpected and may provide important insights for care delivery to Latina BC
survivors.

CORRESPONDING AUTHOR: Alexis D. Abernethy, PhD, Travis Research Institute/ Fuller

Theological Seminary, Pasadena, CA, 91101; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1747

Paper Session 45: Racial Concordance and Provider Communications 3:55 PM-4:12 PM

THE IMPACT OF BLACK PATIENTS' RACE-RELATED ATTITUDES ON THE QUALITY OF RACIALLY

DISCORDANT ONCOLOGY INTERACTIONS

Louis A. Penner, PhD1, Felicity Harper, PhD2, Terrance L. Albrecht, PhD1, Eggly Susan, PhD1

1
Wayne State University/Karmanos Cancer Institute, Detroit, MI; 2Karmanos Cancer
Institute/Wayne State University, Detroit, MI

In a seminal report on Unequal Treatment, the IOM panel identified both provider and
patient attitudes as contributing to racial disparities in the quality of healthcare.
Subsequently, considerable attention has been given to the impact of providers racial
attitudes in racially discordant medical interactions. The racial attitudes of Black patients have
received much less research attention. The present study explores how different facets of
Black cancer patients race-related attitudes and beliefs affect patients behavior in racially
discordant oncology interactions and also patients and oncologists post-visit perceptions..
Black patients (n=114) completed baseline measures of perceived past discrimination, trust of
physicians, healthcare suspicion, and racial identity. Approximately one week later, patients
had initial visits with an oncologist (n=18). Ninety-nine of these interactions were recorded
and assessed for patients verbal activity (e.g., how much they talked). Finally, both patients
and oncologists completed post-interaction questionnaires regarding perceptions of each
other and of the interaction. Despite their conceptual similarity, patient attitudes and beliefs
were weakly intercorrelated and showed different effects. Multi-level regression models were
conducted, to account for patients being nested in physicians. Past discrimination was
strongly positively associated with patient verbal activity, but was not associated with patient
perceptions. Suspicion was negatively associated with patients perceptions of oncologist
patient-centeredness, trust in the oncologist, confidence in recommended treatments, and
ability to complete treatments. Suspicion was also negatively associated with physicians
perceptions of patient education. A mediational analysis showed that perceived educational
level led oncologists to believe higher suspicion patients were less likely to follow treatment
regimens and tolerate their treatments. Thus, Black patients race-related attitudes and
beliefs, which are grounded in history and in current experiences as Black people living in the
US, play an important role in racially discordant medical interactions. These attitudes and
beliefs can contribute to racial disparities in patient and physician treatment decisions and
adherence. Differences among patients in these attitudes and beliefs must be considered in
attempts to reduce racial disparities in healthcare.

CORRESPONDING AUTHOR: Louis A. Penner, PhD, Wayne State University/Karmanos Cancer

Institute, Detroit, MI, 48201; [email protected]
S1748 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 45: Racial Concordance and Provider Communications 4:13 PM-4:30 PM

INFLUENCE OF RACE OF PRINCIPAL INVESTIGATOR ON STUDY PARTICIPATION OF AFRICAN

AMERICANS

Martinus Evans, B.S., M.S.1, Christine May, PhD2, Molly E. Waring, PhD1, Danielle E. Jake-
Schoffman, PhD3, Amanda Okpoebo, BS4, Sherry Pagoto, PhD1

1
University of Massachusetts Medical School, Worcester, MA; 2University of Massachusetts
Medical School, Fiskdale, MA; 3University of Massachusetts Medical School, Worcest, MA;
4
UMass Medical School- Preventive Behavioral Medicine, Leominster, MA

African Americans continue to be underrepresented in clinical research studies. This study

aimed to investigate if the race of the principal investigator (PI) would influence African
Americans motivation to participate in research studies, how to convey the race of the PI,
and factors that would help build trust when the PI is not African American. We conducted in-
depth semi-structured interviews (n=4) and two focus groups (n=6) with adults who self-
identified as African American or multiracial African American. Interviews and focus groups
were recorded and transcripts were coded using an inductive method. We conducted a
content analysis on all questions of using an inductive method using two coders. Participants
(n=10) were on average 35 (SD=6.0) years old and 80% were college educated. Results
revealed that for the majority of the participants (n=9), their motivation to participate in a
study would be enhanced if they knew that the PI was African American (kappa =1.0). When
asked what would the best way to convey the PI was African American, a face-to-face
approach was the most common recommendation reported by 4 participants. Other
recommendations included putting a picture of the PI on the recruitment flyers (n=3), and
stating by black people for black people on recruitment materials (n=1) (kappa=.7). Several
themes emerged about how to build trust when the PI is not African American (kappa= .8):
having a presence in the community (n=6), showing transparency in the research process
(n=4), acknowledgement of the African American struggle (n=2), and hiring diverse staff (n=1).
Increasing the diversity of research investigators and research workforce as well as changes in
recruit strategies may increase African American participation in research studies. Further
research is needed to explore which strategies are to increase participation of African
Americans in research studies.

CORRESPONDING AUTHOR: Martinus Evans, B.S., M.S., University of Massachusetts Medical

School, Worcester, MA, 01655; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1749

Paper Session 46: Health Profiles of Veterans 3:00 PM-3:18 PM

CITATION AWARD WINNER

USING MACHINE LEARNING TO IDENTIFY SOCIAL DETERMINANTS THAT PUT VETERANS AT

HIGHEST RISK FOR POOR DIABETES OUTCOMES

Leslie Hausmann, PhD1, Joshua THORPE, PhD, MPH1, Maria K. Mor, PhD1, John Cashy, PhD2,
Amresh Hanchate, PhD3

1
VA Pittsburgh Healthcare System, Pittsburgh, PA; 2VA Pittsburgh Health System, Pittsburgh,
PA; 3Boston University School of Medicine, Boston, MA

Social determinants of health contribute to disparities in diabetes. Whether or how multiple

social determinants combine to put patients at highest risk for poor diabetes management is
unknown. We used classification and regression trees (CART) methods in a national cohort of
188,743 Veterans with type 2 diabetes in 2008 to identify combinations of key social
determinants that predict two indicators of poor diabetes management: diabetes-related
preventable hospitalizations and mortality. Using Veterans Affairs and US Census data, we
computed 45 individual and zip-code level indicators of social determinants of health,
including individuals sociodemographic (e.g., race, marital status, homelessness) and clinical
(e.g., mental illness, HIV) characteristics and their neighborhoods socioeconomic status,
safety, racial/economic diversity, internet connectivity, and grocery stores. In 2009, 0.7% of
the sample were hospitalized and 3.6% died. CART identified 4 individual-level variables that
combined into 5 risk factor profiles for hospitalizations (range: 0.3% to 3.1% hospitalized).
Area under the receiver-operator characteristic (ROC) curve for these 4 predictors was 0.64.
Highest hospitalization rates were among married Veterans with a history of homelessness
(3.1%) and unmarried Veterans (1.2%). Among those remaining, the highest rate was among
black Veterans under age 65 or over 74 (1.0%). For mortality, CART identified 4 individual and
2 neighborhood variables that combined into 8 risk profiles ranging from 1.8% to 6.6% in the
youngest and oldest age groups, respectively (ROC=0.66). In the middle age groups, those
who were unmarried, had hepatitis or used tobacco, and lived in areas with high poverty and
unemployment had increased risk of death. From the 45 possible risk factors, CART analysis
identified a small subset of key social determinants that predicted poor diabetes outcomes in
Veterans. Additionally, these risk factor combinations would be difficult to detect using
traditional statistical methods and can enable interventions to be targeted towards
particularly at-risk populations.
S1750 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Leslie Hausmann, PhD, VA Pittsburgh Healthcare System,

Pittsburgh, PA, 15240; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1751

Paper Session 46: Health Profiles of Veterans 3:19 PM-3:36 PM

MERITORIOUS AWARD WINNER

CHARACTERISTICS OF ARMY RESERVE FAMILIES WHO USE FAMILY READINESS PROGRAMS

Erin M. Anderson Goodell, SM1, D. Lynn Homish, BS2, Gregory G Homish, PhD2

1
Johns Hopkins Bloomberg School of Public Health, Baltimore, MD; 2University at Buffalo,
Buffalo, NY

Background: Previous research suggests that families of Reserve/National Guard (R/NG)

personnel are less integrated into military life compared to active duty families. Special family
readiness programs (FRPs) exist to address this issue, but limited research has been
conducted to examine who uses them. This study uses the Andersen Behavioral Model (ABM)
to examine characteristics of Army R/NG families who access FRPs.

Methods: Data are cross-sectional from Operation: SAFETY (Soldiers And Families Excelling
Through the Years), an ongoing longitudinal study of US Army R/NG Soldiers and their
partners (N=411 couples). The ABM was used to frame predisposing, enabling, and need-
based factors that may influence use of FRPs. Logistic regression models assessed the
likelihood of the respondent or an immediate family member accessing an FRP based on
selected variables.

Findings: Predisposing factors that may influence a family to access FRPs include having more
children (OR=1.14 [1.01-1.30], p < 0.01). A military-specific predisposing factor is a family
experiencing at least 1 deployment (OR=3.18 [2.08-4.85], p < 0.001). Furthermore, having two
spouses who have deployed, as in dual-military couples, also increases the likelihood of
accessing an FRP (OR=4.95 [1.91-12.84], p < 0.01). For enabling factors, families with members
who have longer time in military service are more likely to access FRPs (OR=1.12 [1.07-1.16], p
< 0.001), while families with the wife as the military spouse results in lower likelihood of
accessing FRPs, compared to husband-only and dual-military families (OR=0.26 [0.12-0.57], p
< 0.01 and OR=0.33 [0.12-0.93], p < 0.05, respectively). For need factors, mental health status
(depression, hazardous drinking, and PTSD), of either military or civilian spouse, is not related
to the outcome. However, physical intimate partner violence perpetrated by the husband is
marginally significantly related to accessing an FRP (OR=1.68 [1.00-2.82], p=0.05).
S1752 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions: Findings shed light on the types of military families that may or may not access
FRPs, which might guide future efforts to link underserved families to resources. Future
research should focus on understanding need-based factors and the impact that FRPs might
have on such needs.

CORRESPONDING AUTHOR: Erin M. Anderson Goodell, SM, Johns Hopkins Bloomberg School
of Public Health, Baltimore, MD, 21205; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1753

Paper Session 46: Health Profiles of Veterans 3:37 PM-3:54 PM

LONGITUDINAL EXAMINATION OF RESILIENCE AMONG 3 SAMPLES OF MILITARY PERSONNEL

Gabriel De La Rosa, NA1, Jennifer Webb-Murphy, PhD2, Scott Johnston, CAPT, PhD2, Jeffrey
Millegan, MD, MPH2

1
Naval Center for Combat and Operational Stress Control, San Diego, CA; 2Naval Center for
Combat & Operational Stress Control, San Diego, CA

After exposure to potentially traumatic events (PTEs) some individuals experience great
hardship while others return to normal functioning. This ability to bounce back after
exposure to PTEs characterizes psychological resilience. Given the importance of resiliency in
contributing to individual wellbeing, several researchers have noted a need for considering
the temporal nature of this construct. Research into the stability of resiliency over time is
limited and can provide mixed conclusions. To help add to this growing literature base, the
current study investigates the stability of resiliency among three samples of military
personnel. Sample 1 includes 873 personnel on a combat deployment. Sample 2 includes 177
personnel participating in a training initiative. Sample 3 includes 611 military personnel
presenting for mental health treatment. Resilience was measured using the Response
toStressful Experiences Scale (RSES), a 22-item scale emphasizing coping processes.
Longitudinal growth models using a multilevel random coefficient modeling framework were
employed to examine trajectories of resilience over time (Raudenbush & Bryk, 2002). Results
suggest that from 49% to 64% of variance in resilience scores were explained by between
person differences. Results suggest that lowest initial RSES scores were found among the
sample of treatment seeking military personnel while the highest initial RSES scores were
found among the personnel participating in a training initiative. The treatment seeking sample
displayed the strongest increase in resilience scores over time while the deployed Marine
sample displayed the lowest increase in resilience scores over time. Results from the
treatment seeking sample may suggest that clinicians can provide personnel with the
resources and confidence needed to address future stressors. Results of the current study
suggest that over time, resilience scores are more likely to increase (or remain stable) than to
decrease. This is evidenced by the significant positive relationship between time and resiliency
scores and the positive relationship between age and overall resiliency scores. Even
throughout a stressful combat deployment, resiliency scores tended to increase over time.

CORRESPONDING AUTHOR: Gabriel De La Rosa, NA, Naval Center for Combat and Operational
Stress Control, San Diego, CA, 92103; [email protected]
S1754 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 46: Health Profiles of Veterans 3:55 PM-4:12 PM

12-MONTH WEIGHT LOSS OUTCOMES AMONG VETERANS WITH PREDIABETES IN AN ONLINE

DIABETES PREVENTION PROGRAM

Tannaz Moin, MD, MBA, MSHS1, Mona AuYoung, PhD, MS, MPH2, Laura Damschroder, MS,
MPH2, Caroline Richardson, MD3

1
VA Greater Los Angeles Healthcare System, David Geffen UCLA School of Medicine, Los
Angeles, CA; 2VA CCMR, Ann Arbor, MI; 3University of Michigan, Ann Arbor, MI

There is a significant evidence base supporting the effectiveness of the Diabetes Prevention
Program (DPP), a lifestyle management program to reduce incidence of diabetes, which has
been translated to multiple settings. Online adaptations of the DPP may enhance reach by
allowing individuals to participate in DPP at the time and location of their choosing. However,
most programs have been limited to face-to-face or phone-based modes of delivery. Thus,
the goal of this study was to assess the effectiveness of an online version of the DPP among
Veterans with prediabetes.

Eligible Veterans with prediabetes were invited to enroll in an online DPP, which was added to
an ongoing trial of two in-person programs as an additional parallel, non-randomized arm. The
analytic sample included consented patients who had 12-month weight outcomes available
(via wireless scales or clinical administrative data). Mixed effects models were used to assess
12-month weight change, controlling for demographic factors.

Based on preliminary analyses, on average, online DPP participants (n=268) lost 6.25 lbs (p <
0.001) from baseline to 12 months. This was comparable to the average weight loss of 7.5 lbs
(p < 0.001) among participants of the in-person DPP that was part of the ongoing trial within
the VA. These preliminary findings suggest that online delivery of DPP is feasible and that
results may be comparable to a group-based in-person program for Veterans with
prediabetes.

CORRESPONDING AUTHOR: Tannaz Moin, MD, MBA, MSHS, VA Greater Los Angeles
Healthcare System, David Geffen UCLA School of Medicine, Los Angeles, CA, 90073;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1755

Paper Session 46: Health Profiles of Veterans 4:13 PM-4:30 PM

LIFESTYLE FACTORS AND RISK OF OBESITY AMONG US VETERANS

Deborah B. Bookwalter, ScD1, Isabel G. Jacobson, MPH1, Ben Porter, PhD1, So Yeon Kong,
PhD1, Alyson J. Littman, PhD2, Rudolph P. Rull, PhD1, Edward J. Boyko, MD, MPH3

1
Naval Health Research Center, San Diego, CA; 2VA Puget Sound Health Care System and the
University of Washington, Seattle, WA; 3VA Puget Sound Health Care System, Seattle, WA

Research suggests US veterans have a higher prevalence of obesity (body mass index [BMI]
30 kg/m2) than those with no prior military service and weight gain is particularly high
around the time of military discharge. Despite an established link between individual lifestyle
factors and weight gain, few studies have assessed the combined influence of multiple factors
on obesity risk. This study examined the cumulative effect of several lifestyle factors on risk of
new-onset obesity after separation from the military using data from the Millennium Cohort
Study, an ongoing prospective study designed to investigate health effects associated with US
military service. Participants complete surveys triennially that collect information on
demographics, service-related and lifestyle factors, and physical and mental health.
Participants self-report of lifestyle factors of interest (physical activity, sedentary time, sleep
duration, fast food intake, alcohol use) were dichotomized into healthy and unhealthy
categories (e.g., fast food intake < 1 vs. 1 times per week). Multivariable Cox models
estimated hazard ratios (HR) and 95% confidence intervals (CI) for incident obesity in relation
to the number of healthy lifestyle factors (0 to 5). Among 16 432 participants with a BMI of
18.5 to 29.9 kg/m2 at baseline, 16% became obese over a mean follow-up of 4.8 years.
Obesity risk decreased with an increasing number of healthy lifestyle factors, with a 51%
lower risk for those with all 5 vs. 0 healthy lifestyle factors (HR 0.49; 95% CI 0.34 to 0.69);
evaluated ordinally, risk was 14% lower for each additional healthy lifestyle factor (HR 0.86;
95% CI 0.83 to 0.90; P, trend < 0.001). Estimates were similar by sex but less precise among
women. The population attributable risk for having 1 or more of the 5 unhealthy lifestyle
factors compared with none was 30% (95% CI 1% to 48%), an estimate of the proportion of
individuals with new-onset obesity that would have been prevented if the entire study
population had all 5 healthy factors. The present findings suggest the adoption of an
increasing number of healthy lifestyle factors is associated with an additive reduction in risk of
obesity among veterans.

CORRESPONDING AUTHOR: Deborah B. Bookwalter, ScD, Naval Health Research Center, San
Diego, CA, 92104; [email protected]
S1756 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 47: Behavioral Medicine in Sleep and Insomnia Research 3:00 PM-3:18 PM

MERITORIOUS AWARD WINNER

STRESS REACTIVITY MEDIATES THE WITHIN-DAY LINK BETWEEN SLEEP QUALITY AND
DEPRESSION IN FIBROMYALGIA

Dhwani J. Kothari, M.A., Crystal T. Rivers, J.D., M.A., Mary C. Davis, Ph.D.

Arizona State University, Tempe, AZ

Poor sleep quality is a common correlate of chronic pain. Not only does poor sleep exacerbate
ongoing pain, but it can also impair psychological functioning through heightened symptoms
of depression. The within-day mechanisms linking a night of poor sleep to increased
depression the next day have not been examined. Therefore, the current study examined the
within-day relations among morning reports of sleep quality on the previous night, late
afternoon reports of stress reactivity, and nighttime reports of depressive symptoms. Stress
reactivity was expected to mediate the link between sleep quality and symptoms of
depression. Electronic daily diary data were collected for 21 days from 220 individuals with
fibromyalgia, a chronic pain condition. Multilevel Structural Equation Modeling was
performed to estimate the mediation model at the within-person level. Results demonstrated
that stress reactivity significantly mediated the relation between sleep quality and depressive
symptoms. Specifically, higher than usual morning reports of poor sleep quality on the
previous night predicted greater than usual late afternoon stress reactivity, which
subsequently predicted higher than usual levels of depressive symptoms reported at
nighttime. Further, a secondary within-day analysis showed that nighttime reports of
depressive symptoms did not predict sleep quality reported the next morning. Overall, the
findings suggest that a night of poor sleep may heighten same-day depression in chronic pain
individuals by interfering with their ability to manage stressors that day. There was no
evidence to suggest carryover effects of depression on sleep, suggestive of a vicious cycle,
on subsequent days.

CORRESPONDING AUTHOR: Dhwani J. Kothari, M.A., Arizona State University, Tempe, AZ,
85282; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1757

Paper Session 47: Behavioral Medicine in Sleep and Insomnia Research 3:19 PM-3:36 PM

BRIEF BEHAVIORAL THERAPY FOR CANCER-RELATED INSOMNIA: EFFECT ON QUALITY OF LIFE

AND HEART RATE VARIABILITY DURING CHEMOTHERAPY

Oxana Palesh, PhD1, Michelle Janelsins, PhD, MPH2, Caroline Scheiber, PhD1, Mary Melissa.
Packer, MA3, Shelli Kesler, PhD4, Richard Gevirtz, PhD5, Joseph J. Guido, MS2, Charles E.
Heckler, Ph.D., M.S. 2, Karen Mustian, PHD, MPH, MS2

1
Stanford, Stanford, CA; 2University of Rochester Medical Center, Rochester, NY; 3Stanford
University, Burlingame, CA; 4MD Anderson, Houston, TX; 5Alliant International University, San
Diego, CA

Quality of Life (QOL) is negatively impacted in breast cancer (BC) patients receiving
chemotherapy, and diminished QOL exacerbates psychological and physiological symptoms.
Reduced vagal tone, as indicated via lower heart rate variability (HRV), is hypothesized to
influence the pathophysiology of impaired QOL. To date, there are no effective brief
behavioral interventions that are deliverable to cancer patients in the clinic while they are
receiving chemotherapy. As QOL and HRV are heavily influenced by sleep quality, we aimed to
evaluate the efficacy of a novel intervention, Brief Behavioral Therapy for cancer-related
insomnia (BBT-CI), for improving QOL and HRV in BC patients undergoing chemotherapy.

In our phase II, 2-arm, randomized controlled trial, 71 BC patients (mean age=52.5, SE=1.16;
96% Caucasian) who reported moderate insomnia were randomized to: 1) BBT-CI (2 face to
face sessions + 4 phone calls) or 2) a healthy eating behavioral control (HEAL). BBT-CI and
HEAL were delivered over 6 weeks by trained nurses or clinical research assistants in the
chemotherapy infusion clinic at 5 NCI-funded Community Oncology Research Program
(NCORP) clinical practices across the U.S. QOL was measured via the Functional Assessment of
Cancer Therapy General (FACT-G) and FACT-B, a specific subscale for patients with BC, as
well as HRV via an ambulatory heart-rate device (Firstbeat Bodyguard 2) at baseline and post-
intervention. HRV was measured using time (SDNN, rMSSD) and frequency indices.

The study showed excellent feasibility and acceptability of conducting BBT-CI in the infusion
clinic with a recruitment rate of over 75% of eligible participants and an intervention
adherence of 74%, with 75% of intervention components successfully delivered by trained
community research staff. Analysis of covariance showed that patients who received BBT-CI
exhibited improvements in their QOL compared to those receiving HEAL at 6 weeks (FACT-G,
p=0.009; FACT-B, p=0.016). Furthermore, there was a significant treatment effect in HRV
indices of both time and frequency in which patients receiving BBT-CI remained stable in
S1758 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

contrast to patients receiving HEAL whose HRV declined (SDNN, p=0.047; rMSSD, p=0.040; HF,
p=0.018).

BBT-CI rendered greater improvements than HEAL on QOL and HRV, demonstrating efficacy in
the community oncology setting. BBT-CI is a novel intervention that can yield additional
psychophysiological benefits to patients during their infusion sessions. Bringing efficacious
behavioral interventions to the oncology clinic can change the paradigm of clinical practice,
ameliorating patients burden and symptoms.

CORRESPONDING AUTHOR: Oxana Palesh, PhD, Stanford, Stanford, CA, 94305;

[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1759

Paper Session 47: Behavioral Medicine in Sleep and Insomnia Research 3:37 PM-3:54 PM

SLEEP QUALITY INFLUENCES NEXT-DAY AFFECT AND EMOTION REGULATION: A DAILY DIARY
STUDY OF HIV-POSITIVE GAY AND BISEXUAL MEN

Brett M. Millar, MA1, Jeffrey T. Parsons, PhD2, H. Jonathon Rendina, PhD, MPH3

1
Graduate Center, City University of New York, New York, NY; 2Hunter College and The Center
for HIV Educational Studies and Training, New York, NY; 3Hunter College of the City University
of New York (CUNY), New York, NY

Background: Previous studies have uncovered the impact of poor sleep quality on subsequent
affect and emotion regulation in the general population. Daily diary studies on gay and
bisexual men have also observed event-level associations between both anxious affect and
sexual activation and increased sexual risk-taking. The current study sought to explore links
between sleep quality and daily affect and emotion regulatory capacity amongst HIV-positive
gay and bisexual men using event-level data from an online daily diary study. Method: We
analyzed 798 daily reports from 53 HIV-positive gay and bisexual men enrolled in a 21-day
online diary study. Multilevel logistic modeling was used to test whether previous nights
sleep quality predicted affect (depressive, anxious, anger, fatigue, and sexual, separately) and
emotion regulation. Models were fit with an AR(1) covariance structure and random
intercept. Daily reports of previous nights sleep quality were disaggregated into individual-
level averages and situational fluctuations around ones average, and were used to predict
current-day affect and emotion regulation. Results: The sample was diverse with regard to
race/ethnicity (75% men of color), and had a mean age of 38 years. Daily sleep quality was
significantly associated with each affective outcome, such that an individual having lower-
than-average sleep quality on a given day also reported higher levels of depressive (B = -0.08,
p < 0.001), anxious (B = -0.04, p = 0.02), and angry (B = -0.05, p < 0.01) affect, as well as
greater fatigue (B = -0.08, p < 0.001) and emotion dysregulation (B = -0.04, p< 0.01). Lower-
than-average sleep quality on a given day was also marginally associated with increased
sexual activation (B = -0.05, p = 0.06). Conclusions: Individuals experiencing daily disruptions
in sleep quality reported increased negative emotionality, which is consistent with previous
research on various populations. The current study highlights poor sleep quality as an
important lifestyle variable in these men's emotional and sexual lives--increasing negative
affect and potentially sexual activation, while also decreasing the individual's capacity to
regulate these affective states. Implications for the development of sleep-related
interventions to improve mood and daily coping, and to potentially reduce sexual risk-taking,
will be explored.
S1760 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Brett M. Millar, MA, Graduate Center, City University of New
York, New York, NY, 10023; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1761

Paper Session 47: Behavioral Medicine in Sleep and Insomnia Research 3:55 PM-4:12 PM

SLEEP DISRUPTION AMONG CANCER PATIENTS FOLLOWING AUTOLOGOUS HEMATOPOIETIC

CELL TRANSPLANTATION (HCT)

Ashley M. Nelson, MA1, Heather Jim, PhD2, Brent Small, PhD3, Brian D. Gonzalez, Ph.D.4, Taiga
Nishihori, MD5, Julie Cessna, M.P.H., M.A.6, Kelly Hyland, B.A.6, Meredith Rumble, PhD7, Paul
B. Jacobsen, PhD8

1
University of South Florida & Moffitt Cancer Center, Tampa, FL; 2Moffitt Cancer Center,
Tampa, FL; 3University of South Florida, Tampa, FL; 4Rutgers Cancer Institute of New Jersey,
New Brunswick, NJ; 5H. Lee Moffitt Cancer Center and Research Institute, Tampa, FL;
6
University of South Florida/Moffitt Cancer Center, Tampa, FL; 7University of Wisconsin,
Madison, WI; 8National Cancer Institute, Bethesda, MD

Background: Sleep disruption is a commonly reported quality of life concern among cancer
survivors following HCT. Despite the high prevalence of sleep disruption in HCT recipients,
relatively little research has investigated sleep disruption or explored relationships with
psychological factors such as cognitive or behavioral factors. In addition, no studies have used
actigraphy to characterize sleep disruption in this population.

Method: Autologous HCT recipients who were between 6 to 18 months post-transplant

completed self-report measures of cancer-related distress, fear of cancer recurrence,
dysfunctional sleep cognitions, and maladaptive sleep behaviors upon enrollment. Patients
then wore an actigraph and completed a sleep log for the next 7 days and completed a self-
report measure of sleep disruption on day 7 of the study.

Results: Patients were 84 autologous HCT recipients (age M = 60, 45% female, 69% multiple
myeloma). Forty-one percent met criteria for sub-clinical or clinical insomnia based on patient
self-report (ISI 8). Examination of actigraphy data indicated that, on average, sleep was
disturbed (e.g., wake after sleep onset M = 66 minutes; total sleep time M = 6.5 hours).
Cancer-related distress, fear of recurrence, dysfunctional sleep cognitions, and maladaptive
sleep behaviors were related to self-reported sleep disruption (ps < .05), but were not related
to objective sleep indices. The cognitive and behavioral factors accounted for 27% of the
variance in subjectively reported sleep disruption compared to 7% of variance accounted for
by cancer-related distress and fear of recurrence.

Conclusion: Results suggest that many HCT recipients experience sleep disruption during the
post-transplant survivorship period. Cancer-related distress, fear of recurrence, dysfunctional
S1762 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

sleep cognitions, and maladaptive sleep behaviors are related to self-reported sleep
disruption and can be considered targets for intervention.

CORRESPONDING AUTHOR: Ashley M. Nelson, MA, University of South Florida & Moffitt
Cancer Center, Tampa, FL, 33612; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1763

Paper Session 47: Behavioral Medicine in Sleep and Insomnia Research 4:13 PM-4:30 PM

MIND-BODY-SPIRITUAL PROGRAM REDUCES INSOMNIA IN FEMALE VETERANS WITH PTSD: A

RANDOMIZED PILOT

Jill E. Bormann, PhD, RN, FAAN1, Danielle J. Beck, MPH, CCRC2, Teresa Gacnik, MSN, RN3

1
VA San Diego Healthcare System Center of Excellence for Stress and Mental Health
(CESAMH), San Diego, CA; 2VA San Diego Healthcare System, San Diego, CA; 3VA San Diego
Healthcare System, Carlsbad, CA

Objectives: Prior research on outcomes of the Mantram Repetition Program (MRP) in 80

women has shown significant posttreatment reductions in perceived stress and improvements
in quality of life. The majority of women, however, did not have history of trauma. This study
explored the outcomes of MRP versus Present Centered Therapy (PCT) in female Veterans
diagnosed with PTSD.
Methods: A randomized trial compared the MRP with PCT in 173 Veterans. Of these, 26 were
women randomly assigned to either 8-weeks (1 hour/week) of individual MRP (n=16) or PCT
(n=10). Insomnia Severity Index (ISI), Clinician-Administered-PTSD Scale-version-4 (CAPS), and
PTSD Checklist-version 4 (PCL) were administered at pre-, post-, and 2-month follow-up. Data
were analyzed using mixed effect models.
Results: Participants average age was 39.5+11.6 ranging from 25 to 65 years old. Fifty-four
percent (n=14) were White, 23% (n=6) were married/partnered, and 27% (n=7) were
employed full-time. There was a significant group by time interaction effect for ISI,
F(2,32)=3.70, p=.036, indicating the MRP group had a significantly greater decrease in ISI at 2-
month follow-up (M=18.94+6.3 to 12.5+9.5) versus PCT group (M=18.0+6.4 to 16.4+7.0).
There was strong trend for a group by time interaction effect on the CAPS, F(2,34)=3.08,
p=.059, showing MRP group with a greater decrease at 2-month follow-up (M=81.50+18.9 to
50.30+20.5) compared with PCT (M=76.90+20.4 to 60.33+27.5). There were significant time
effects for CAPS in both groups, F(2,34)=12.43, p=.001. There were significant time effects for
PCL F(2,45)=16.12, p=.0001 with MRP group reporting a slightly greater decrease in PCL at
follow-up (M=60.07+11.7 to 41.10+12.8) compared with PCT (M=60.50+15.5 to 49.50+16.3).
Conclusions: The MRP, a non-trauma-focused, complementary program for PTSD, may
significantly reduce insomnia in female Veterans with PTSD. Despite the small sample size,
these findings are promising and support future research to determine the full effect of the
MRP for PTSD symptoms.
S1764 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Jill E. Bormann, PhD, RN, FAAN, VA San Diego Healthcare System
Center of Excellence for Stress and Mental Health (CESAMH), San Diego, CA, 92161;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1765

Paper Session 48: Caregivers: Providers of and Targets for Behavioral Medicine 3:00 PM-3:18
PM

CHANGES IN NURSE-CAREGIVER COMMUNICATION IN CANCER HOME HOSPICE VISITS OVER

TIME

Maija Reblin, PhD1, Seth Latimer, MStat2, Margaret Clayton, PhD2, Lee Ellington, PhD2

1
Moffitt Cancer Center, Tampa, FL; 2University of Utah, Salt Lake City, UT

Background

Cancer patients often decline quickly at end of life and many choose home hospice care. In
home hospice, family caregivers (CGs) have primary responsibility for patients day-to-day
symptom, psychosocial, and emotional care with support from a nurse-led interdisciplinary
team. Little systematic research has been conducted on nurse-CG communication in home
hospice and thus it is unknown how nurse-CG communication changes with patient decline.
Our goal was to identify the trajectory of home hospice nurse-CG visit communication.

Methods

As part of a large prospective, multi-site observational study, 537 nurse home visits to 101
spouse cancer CGs were audio-recorded from hospice enrollment to patient death.
Communication in physical, psychosocial, and emotional domains were identified for both
nurses and CGs. Linear mixed modelling was conducted on communication domain using visit
as a random effect and CG and nurse as fixed effects.

Results

CG M age=65.6 years (SD=10.2) and 60% were female. Nurses were 90% female and M years
practicing in hospice=4.5 (SD=4.9). Visits averaged 40 minutes and length did not significantly
change over time. Physical care talk was the most common for nurses on average (M=55%,
SD=23%) and CG (M=46%, SD=28%). Psychosocial talk averaged 11% of nurse talk (SD=10%)
and 21% for CGs (SD=19%); emotional talk averaged 12% for nurses (SD=8%) and 8% for CGs
(SD=11%). Over time nurses had significantly less psychosocial (t=-2.56, p.05).
S1766 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Discussion

Findings suggest that nurses and CGs co-construct communication in home hospice over time,
with small but significant changes as patient death approaches. CG physical care talk increases
as their psychosocial and emotional talk is maintained, suggesting CGs are reporting more
physical changes with patient decline. Despite this, nurse physical care talk remains stable.
The reduction in nurse psychosocial and emotional talk frequency may suggest lessening
priority of these domains or providing more space for CGs. Our findings demonstrate a first
step in future research to demonstrate how CG needs are met over time and the impact this
has on patient and CG outcomes.

CORRESPONDING AUTHOR: Maija Reblin, PhD, Moffitt Cancer Center, Tampa, FL, 33612;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1767

Paper Session 48: Caregivers: Providers of and Targets for Behavioral Medicine 3:19 PM-3:36
PM

FEASIBILITY AND ACCEPTABILITY OF A COMMUNICATION INTERVENTION FOR COUPLES

COPING WITH HEMATOPOIETIC CELL TRANSPLANTATION

Shelby L. Langer, PhD1, Joan M. Romano, PhD2, Stephanie Lee, MD3, Tracey Godlewski, PA-C3,
Melissa DuPen, BS2, Laura Porter, PhD4

1
Arizona State University, Phoenix, AZ; 2University of Washington, Seattle, WA; 3Fred
Hutchinson Cancer Research Center, Seattle, WA; 4Duke University, Durham, NC

Hematopoietic cell transplantation (HCT) is an aggressive form of cancer treatment with

medical and psychosocial sequelae. Caregiving partners are also impacted, reporting elevated
levels of distress. Patient-partner communications, moreover, are characterized by avoidance
of treatment-related concerns. This study assessed acceptability and feasibility of a 5-session
couple-based intervention to optimize communication between HCT patients and partners. Of
89 couples approached and deemed eligible, 40 (45%) consented. Demographic
characteristics were: M (SD) age = 55.9 (12.3), 38% female, 5% Hispanic and 92% White
[patients]; M (SD) age = 54.2 (11.9), 62% female, 8% Hispanic and 92% White [partners].
Eligible and consenting patients did not differ from eligible and refusing patients on these
characteristics (p > .05). All enrolled patients and partners completed baseline assessments
prior to HCT. Thirty reached the intervention (TX) phase 1 month post-HCT; 4 patients died
prior to this phase and one was deemed no longer eligible for transplant. All 30 commenced
TX and 25 completed TX and follow-up questionnaires; two remain in treatment (90%
retention). Process data collected after each session revealed that ratings of self-disclosure
and partner responsiveness were high, with means ranging from 4.45 (0.66) to 4.80 (0.35) on
a 1-5 scale. These findings suggest that the intervention facilitated communication as
intended. Number of sessions completed was positively associated with partner ratings of the
extent to which they felt close to and understood/ validated/ accepted by the patient during
the sessions (r values ranging from 0.40 to 0.54, p values < .05), and unassociated with
baseline demographic or psychosocial characteristics (p > .05). Program evaluation ratings
support acceptability: M (SD) = 4.50 (0.59) and 4.17 (0.72) on a 1-5 scale for satisfaction with
services reported by patients and partners, respectively; and 4.17 (0.70) and 4.22 (0.80) for
usefulness of the program. Findings indicate that HCT patients and partners can be recruited
and retained for this intervention, and found it acceptable and useful. Perceived partner
responsiveness may be important for continued retention.
S1768 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Shelby L. Langer, PhD, Arizona State University, Phoenix, AZ,
85004; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1769

Paper Session 48: Caregivers: Providers of and Targets for Behavioral Medicine 3:37 PM-3:54
PM

COPING WITH HOME HOSPICE CANCER CAREGIVING: EMOTIONAL SCAFFOLDING IN NURSE-

CAREGIVER COMMUNICATION

Maija Reblin, PhD1, Janella N. Hudson, PhD1, Margaret Clayton, PhD2, Lee Ellington, PhD2

1
Moffitt Cancer Center, Tampa, FL; 2University of Utah, Salt Lake City, UT

Background

Cancer caregiving can be stressful for families, especially in home hospice as end of life
approaches and more demands are placed on family caregivers (CGs). CGs of home hospice
patients are often dependent on their hospice nurse for guidance in providing patient care
and in helping CGs emotionally cope with the stress of care and an impending loss. Much is
known about how nurses support CGs in providing patient symptom management, but little
research has addressed nurse strategies for scaffolding CG emotional coping during distress.
This study examines the communication strategies utilized by hospice nurses to support
emotional expression and coping and the responses from informal spouse caregivers.

Methods

Data were taken from a larger study of audio recorded nurse visits to families in cancer home
hospice. Using transcripts of recordings, trained coders identified verbal behaviors
representing various types of nurse emotional scaffolding and CG emotional expression. Using
an iterative process of constant comparison, coders inductively categorized the instances of
nurse and CG communication behavior into overarching themes.

Results

19 visits were selected (representing 7 hospice agencies). Average nurse age was 42.5
(SD=8.9), 90% white, 100% female. CGs were spouses of patients (average years in
relationship=35.9, SD=20.0), average age=68.1 (SD=10.6), 100% white, 84% female). Nurses
scaffolding was more common than CG emotional expression. Nurses scaffolding included
encouragement of emotional expression and modeling emotional coping strategies. CGs
expressed positive emotion (e.g. gratitude, humor) and frustration. Common thematic
S1770 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

strategies included positive reframing (e.g. seeking to see the positive side), building
emotional rapport (e.g. using humor to connect), and venting (e.g. giving permission,
providing a safe space, or airing concerns).

Discussion

Emotion-focused coping can reduce stress and improve psychological and physical health.
Nurses provide emotional scaffolding that encourages some CGs to express emotions during
distress, but more research is needed to determine longer-term impacts on CG coping or CG
health outcomes and patient care. These findings may inform educational materials for nurses
or future communication interventions.

CORRESPONDING AUTHOR: Maija Reblin, PhD, Moffitt Cancer Center, Tampa, FL, 33612;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1771

Paper Session 48: Caregivers: Providers of and Targets for Behavioral Medicine 3:55 PM-4:12
PM

SOCIAL SUPPORT FOR LATINO CAREGIVERS

Isabela Marques, BS, Guilherme Balbim, MS, Melissa Magallanes, MS, Judith Rocha, LCSW,
David Marquez, PhD

University of Illinois at Chicago, Chicago, IL

Alzheimers disease and related dementia (ADRD) affect 11% of people age 65 and older in the
US. Latinos are a growing segment of the population and it is expected that by 2050, 128
million Latinos will account for 29% of the population. Loneliness and low levels of support
from family, friends and social groups are important predictors of disease outcomes in older
adults. Latinos report greater social isolation than non-Latino whites, exposing them to
greater risk of chronic diseases. The purpose of this study was to identify the different groups
that provide social support for Latino caregivers, and to understand how the four kinds of
social support (emotional, appraisal, informational, and instrumental support) are offered for
Latino caregivers. Interviews were conducted with Latino caregivers (N=16, 12 females; 4
males), 50 years and older, caring for a relative with ADRD a minimum of 4 hours daily for the
past 6 months. Eight interviews were conducted in Spanish and eight in English. Interviews
about the caregiver role, family and social support, and health behaviors were recorded,
transcribed, translated and coded according to the directed content analysis. Caregivers
mentioned that social support was provided by family, friends, therapist, homemakers,
governmental aid, and organizations. The four types of social support were identified:
emotional, appraisal, informational, and instrumental support. Caregivers reported receiving
emotional support from family and friends; appraisal support from family and therapist;
informational support was provided by family and organizations; and instrumental support
from family, friends, homemakers, organizations, and governmental aid. Family members
were cited as vital piece in caregiving, providing all four types of support. Instrumental
support was mentioned as a needed type of support, mainly offered by family members,
homemakers, and governmental aid. Interventions must consider the Latino family structure
in offering assistance for caregivers, and also the crucial instrumental help desired to reduce
caregivers responsibilities.

CORRESPONDING AUTHOR: Isabela Marques, BS, University of Illinois at Chicago, Chicago, IL,
60612; [email protected]
S1772 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Paper Session 48: Caregivers: Providers of and Targets for Behavioral Medicine 4:13 PM-4:30
PM

ENGAGEMENT IN PLEASANT LEISURE ACTIVITIES AND BLOOD PRESSURE: A 5-YEAR

LONGITUDINAL STUDY IN ALZHEIMERS CAREGIVERS

Taylor Bos, MA1, Brent Mausbach, PhD2, Igor Grant, MD2

1
SDSU/UCSD Joint Doctoral Program in Clinical Psychology, San Diego, CA; 2University of
California San Diego, La Jolla, CA

Objective: Elevated blood pressure is a significant public health concern, particularly given its
association with cardiovascular disease risk, including stroke. Caring for a loved one with
Alzheimers disease has been associated with physical health morbidity, including higher
blood pressure. Engagement in adaptive coping strategies may help prevent blood pressure
elevation in this population. This 5-year longitudinal study examined whether greater
participation in pleasant leisure activities was associated with reduced blood pressure in
caregivers.

Methods: Participants were 126 in-home spousal Alzheimer caregivers (mean age = 74.2 7.9
years) that completed five yearly assessments. Linear mixed effects models analysiswas used
to examine the longitudinal relationship between pleasant leisure activities and caregivers
blood pressure, after controlling for demographic and health characteristics.

Results: Greater engagement in pleasant leisure activities was associated with reduced mean
arterial blood pressure (MAP; p = 0.037). Follow-up analyses indicated engagement in
activities was significantly associated with reduced diastolic (p = 0.031) but not systolic blood
pressure (p = 0.097). In addition, MAP was significantly reduced when caregiving duties ended
because of placement of care recipients in nursing homes (p = 0.005) or death of the care-
recipient (p = 0.019).

Conclusions: Greater engagement in pleasant leisure activities was associated with lowered
caregivers blood pressure over time. Participation in pleasant leisure activities may have both
cardiovascular and psychological health benefits for Alzheimers caregivers.

CORRESPONDING AUTHOR: Taylor Bos, MA, SDSU/UCSD Joint Doctoral Program in Clinical
Psychology, San Diego, CA, 92103; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1773

Paper Session 49: Sexual Behaviors Across Populations 3:00 PM-3:18 PM

MERITORIOUS AWARD WINNER

THE IMPACT OF AGE ON THE ASSOCIATIONS BETWEEN DRUG USE DURING SEX AND HIV RISK
BEHAVIORS AMONG BLACK MEN WHO HAVE SEX WITH MEN

Jessica L. Maksut, M.A.1, Elizabeth J. Siembida, Ph.D.2, Lisa Eaton, PhD3, Kate E. Dibble, M.A.1

1
University of Connecticut, Storrs, CT; 2University of Connecticut, Bethesda, MD; 3InCHIP,
Storrs, CT

Background: The associations between drug use during sex, age, and HIV risk remain under-
studied among samples of exclusively Black MSM (BMSM). Younger BMSM are at highest risk
for HIV infection, and so understanding the influence of age on HIV risk behavior patterns
among BMSM is critical.

Method: 812 BMSM (M age = 33.28 years, SD = 11.15 years) in Atlanta, GA completed a
quantitative assessment via computer-assisted personal interviewing (CAPI). Negative
binomial regression was employed to model the association of number of male partners with
whom condomless anal intercourse (CAI) was had, number of female partners with whom
condomless vaginal intercourse (CVI) was had, depressive symptoms, problem drinking, HIV
treatment optimism, and age with frequency of drug use during sex in the past 3 months.

Results: Of the BMSM who reported drug use during sex (n = 399, 49.14%), the average
number of sex acts under the influence of drugs was 6.28 (SD = 20.47). BMSMs drug use
during sex was positively associated with the number of female partners with whom UVI was
had (AOR = 1.61; 95% CI = 1.39-1.88, p = 0.001), depressive symptoms (AOR = 1.06; 95% CI =
1.01-1.10, p = 0.006), and problem drinking (AOR = 2.00; 95% CI = 1.55-1.58, p = 0.001). In
contrast, HIV treatment optimism (AOR = 0.68; 95% CI = 0.54-0.87, p = 0.002) and younger age
(AOR = 0.95; 95% CI = 0.92-0.98, p = 0.001) were negatively associated with drug use during
sex. The interactions between CAI age (AOR = 1.02; 95% CI = 1.01-1.04, p = 0.003), CVI age
(AOR = 0.98; 95% CI = 0.97-0.99, p = 0.001), and HIV treatment optimism age (AOR = 1.02;
95% CI = 1.01-1.04, p = 0.002) were significant. As age decreased, the association between
CVI with drug use during sex became stronger. Conversely, as age increased, the associations
between CVI and HIV treatment optimism with drug use during sex were strengthened.

Discussion: Age is key to understanding the relationships between drug use during sex, CAI,
CVI, and HIV treatment optimism. Future HIV prevention efforts may consider tailoring their
content to address specific age-related risks for their BMSM clients.
S1774 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Jessica L. Maksut, M.A., University of Connecticut, Storrs, CT,

06269-1058; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1775

Paper Session 49: Sexual Behaviors Across Populations 3:19 PM-3:36 PM

ASSOCIATION BETWEEN ATTACHMENT AND RISK FACTORS FOR ENGAGEMENT IN RISKY

SEXUAL BEHAVIORS AMONG AFRICAN AMERICAN WOMEN

Nicole Gause, M.A.1, Jennifer L. Brown, N/A2, Sarah Whitton, PhD1, Ralph J. DiClemente, PhD3,
Kristen E. Jastrowski Mano, PhD1

1
University of Cincinnati, Cincinnati, OH; 2University of Cincinnati College of Medicine,
Cincinnati, OH; 3Rollins School of Public Health, Atlanta, GA

Background: The study assessed whether attachment-related constructs, working model-

self (WMS) and working model-other (WMO), are associated with risky sexual behaviors, and
whether WMS and WMO are associated with risk factors for sexual behaviors commonly
targeted by HIV prevention interventions for African American women (e.g., fear of condom
negotiation). Furthermore, the study evaluated whether risk factors partially mediate the
hypothesized association between WMS and WMO and risky sexual behaviors among African
American women.

Methods:560 African American females (age 18-24) reported attachment style, condom use
for vaginal and anal sex, and alcohol use prior to sexual encounters during the past 3 months.
A subset of women in relationships reported condom use at most recent sexual encounter
with boyfriend/main sexual partner.

Results: Neither WMS nor WMO were associated with the following outcomes: (a) proportion
condom use for vaginal sex; (b) any condom use for anal sex; (c) condom use at most recent
sexual encounter with boyfriend/main sexual partner; and (d) consuming alcohol prior to any
sexual encounters. As such, no partial mediation hypotheses were supported. However, WMS
was associated with all risk factors for risky sexual behaviors (mediators): partner
communication self-efficacy (for example, in relation to this construct, = 0.26, p > .001); fear
of condom negotiation; peer norms for risky sexual behaviors; partner trust; and sex-related
alcohol expectancies. Participants with lower WMS had (a) lower partner communication self-
efficacy, (b) more fears related to condom negotiation, (c) greater endorsement of peer-
norms for risky sexual behaviors, (d) lower levels of partner trust, and (e) greater sex-related
alcohol expectancies.
S1776 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Discussion: Future research should examine whether women with lower WMS benefit from
more intensive HIV prevention interventions to effectively change these risk factors and,
whether they reduce engagement in sexual risk behaviors.

CORRESPONDING AUTHOR: Nicole Gause, M.A., University of Cincinnati, Cincinnati, OH,

45208; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1777

Paper Session 49: Sexual Behaviors Across Populations 3:37 PM-3:54 PM

GENDER COMPARISONS OF CONDOMLESS SEX INTENTION AMONG COLLEGE VARSITY

ATHLETES IN TAIWAN USING THE THEORY OF PLANNED BEHAVIOR

Jiun-Hau Huang, S.M., Sc.D., Chia-Hsien Hsu, B.S.

National Taiwan University College of Public Health, Taipei, N/A, Taiwan (Republic of China)

Background: Prior research has found that the prevalence of unprotected sexual behavior in
male college athletes was higher than that in female athletes. However, given the different
socio-cultural context in Taiwan, little is known about whether and how male and female
athletes differ in their condomless sex intention (CSI) and associated factors. Hence, this study
aimed to systematically examine the theory-based factors influencing CSI among college
athletes using the Theory of Planned Behavior (TPB).

Methods: A total of 1,652 anonymous survey responses were collected from college varsity
athletes in Taiwan. After excluding non-heterosexual athletes (n=270) and incomplete
responses, a final sample of 1,348 was included in the multivariate logistic regression analyses
to examine the TPB-based factors related to higher CSI (rated on a 7-point scale and
dichotomized by the median). Items derived from the 3 constructs of TPB (Attitude Toward
the Behavior, Subjective Norm, and Perceived Behavioral Control) were factor-analyzed to
form 5 factor-based subscales, whose scores were trichotomized using tertiles as cutoffs.

Results: In this study, male college athletes reported significantly greater prevalence of
higher CSI (53.6%) than their female counterparts (31.7%). In regard to the 5 TPB subscales,
males tended to rate significantly more positively than females concerning attitudinal
evaluations of positive CSI outcomes, subjective norm, and perceived ease of CSI under
constraining conditions. Among male athletes, those perceiving relatively neutral (AOR=1.66)
and relatively positive (AOR=2.59) attitudinal evaluations of positive CSI outcomes were more
likely to have higher CSI. By contrast, females had elevated CSI only when they had relatively
positive evaluations (AOR=3.75). Further, males with relatively neutral (AOR=1.68) and
relatively supportive (AOR=3.22) norm had higher CSI, whereas females reported higher CSI
only when perceiving relatively supportive norm (AOR=3.04). Notably, males perceiving
relatively intermediate level of control over CSI under facilitating conditions (AOR=0.62) had
lower CSI. Lastly, males with relatively high level of control over CSI under constraining
conditions (AOR=1.66) had higher CSI.

Conclusions: In conclusion, male college athletes in this study had higher CSI than their female
counterparts. Using TPB as a framework, this study found that males reported elevated CSI if
they had relatively neutral or higher levels of attitudinal evaluations of positive CSI outcomes
S1778 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

and supportive norm, while females only reported higher CSI if they had relatively positive
evaluations and supportive norm, suggesting a threshold for CSI in female athletes. These
findings along with other differential patterns and associations warrant gender-specific sexual
health education tailored for male and female college athletes in the future.

CORRESPONDING AUTHOR: Jiun-Hau Huang, S.M., Sc.D., National Taiwan University College of
Public Health, Taipei, N/A, 10055; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1779

Paper Session 49: Sexual Behaviors Across Populations 3:55 PM-4:12 PM

EXAMINING THE SYMPTOM NETWORK OF DEPRESSION AND PTSD IN RELATION TO SEXUAL

RISK BEHAVIOR IN MSM WITH CHILDHOOD SEXUAL ABUSE

Karmel W. Choi, MA, PhD Candidate1, Abigail W. Batchelder, Ph.D., M.P.H.2, Michael
Boroughs, PhD3, Peter P. Ehlinger, B.A.4, Conall O'Cleirigh, PhD1

1
Massachusetts General Hospital/Harvard Medical School, Boston, MA; 2Massachusetts
General Hospital/ Harvard Medical School, Boston, MA; 3University of Windsor, Windsor, ON,
Canada; 4Fenway Community Health, Cambridge, MA

Background: Histories of childhood sexual abuse (CSA) are disproportionately common among
men who have sex with men (MSM) and predict high rates of depression and PTSD, which are
known to contribute to sexual risk behavior. To date, research has examined influences of
depression and PTSD on sexual risk at the level of disorders, rather than individual symptoms.
Moreover, while depression and PTSD frequently co-occur, their symptoms are expected to
be meaningfully distinct in presentation and behavioral impact. This study sought to use a
novel network approach to (1) visualize how symptoms of depression and PTSD are inter-
related in MSM with CSA histories, and (2) explore which symptoms of depression and PTSD
are most related to sexual risk behavior.

Methods: In a cross-sectional sample of 296 urban MSM with CSA histories, depressive
symptoms were assessed using the Centre for Epidemiological Studies Depression Scale, and
PTSD symptoms were assessed using the frequency scale of the Davidson Trauma Scale.
Sexual risk behavior was indexed by self-reported frequency of unprotected sex in the past
three months. Network analyses were performed in R using mixed graphical modeling to
account for different variable types.

Results: The resulting network structure showed complex associations (1) between symptoms
across depression and PTSD, and (2) in relation to sexual risk behavior, even when accounting
for the influence of all other variables in the network. While depression and PTSD symptoms
mainly clustered within their respective disorders, they were connected at symptom nodes
related to sleep disruption, concentration difficulties, and limited life outlook. Moreover,
specific symptoms of depression (e.g., trouble keeping mind on things, feeling disliked by
others) and PTSD (e.g., painful reminders of the event, reliving event) were linked to sexual
risk behavior while other symptoms were not.

Conclusions: Network analyses allow associations to be simultaneously mapped according to

S1780 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

the strength and relevance of these relationships. Consistent with existing literature,
depression and PTSD showed some degree of connectivity (comorbidity) across expected
symptoms domains. Of note, this analysis has identified symptoms of depression and PTSD
that may particularly influence sexual risk in MSM with CSA histories, and illustrates a
promising method to refine clinical targets affecting health risk behavior.

CORRESPONDING AUTHOR: Karmel W. Choi, MA, PhD Candidate, Massachusetts General

Hospital/Harvard Medical School, Boston, MA, 02114; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1781

Paper Session 49: Sexual Behaviors Across Populations 4:13 PM-4:30 PM

SEX ED IS NOT A THING!: EARLY HEALTH SEEKING BEHAVIORS AMONG MEN WHO HAVE SEX
WITH MEN IN RURAL AREAS

Danielle N. Lambert, MPH, CHES1, Natalia Truszczynski, MPH2, Anne Marie Schipani-
McLaughlin, MPH2, Carolyn Lauckner, PhD2, Nathan Hansen, PhD2

1
University of Georgia, Atlanta, GA; 2University of Georgia, Athens, GA

Background: Men who have sex with men (MSM) are disproportionately affected by HIV and
STIs. Although at increased risk for negative health outcomes, relatively few sexual health
education curriculums are inclusive to the health needs of men identifying as gay, bisexual, or
queer. This is especially true in rural areas where stigmatization of sexual minorities is even
more prevalent. Thus, in the absence of appropriate programming, little is know about where
and from whom MSM in rural areas seek out and learn sexual education.

Methods: Semi-structured qualitative interviews were conducted with 20 respondents

ranging from 18-60 years old from July 2015 to June 2016. Respondents were eligible for
participation if they identified as a man who dates or has sex with other men, resided in a
non-metropolitan area of the United States, and were at least 18 years old. Interviews were
transcribed verbatim and analyzed by a diverse coding team in NVivo 11 using grounded
theory approaches.

Results: The majority of the respondents sought sexual health information from three
sources: the Internet, their social network, and LGBT resource centers. Geographic isolation
and low population density forced many young MSM to turn to the Internet for information
related to HIV/STIs, condom use, and PrEP. While some recognized the CDC and WHO
websites as trusted forms of information, many did not know where to start searching for
health information on the Web and therefore ended up reviewing non-validated sites, such as
blogs. Web-based social networking platforms, such as dating websites or support groups,
were a way for respondents to obtain health advice from other MSM in rural areas where
inclusive social venues were uncommon. For many respondents, LGBT resource centers
located on college campuses were a key source of health information. They provided
educational materials to improve safe sex behaviors, guidance in accessing healthcare, free
testing, and confidential services in a trusted environment.

Conclusions: The lack of tailored sexual health education for MSM speaks to an immediate
need for more reliable programming and inclusive resource centers. Understanding the
barriers to receiving sex education, and the ways in which MSM are currently obtaining sexual
S1782 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

health information, can help practitioners relay accurate sexual health information to
channels and organizations commonly accessed by MSM.

CORRESPONDING AUTHOR: Danielle N. Lambert, MPH, CHES, University of Georgia, Atlanta,

GA, 30322; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1783

Friday

March 31, 2017

6:00 PM-7:00 PM
S1784 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

6:00 PM-7:00 PM

THE IMPACT OF HOPE ON QUALITY OF LIFE IN LUNG CANCER

Laurie E. Steffen, PhD1, Kevin E. Vowles, PhD2

1
Wake Forest University School of Medicine, Winston-Salem, NC; 2University of New Mexico,
Albuquerque, NM

Goal-setting, problem-solving, and values-based action are common components of quality of

life interventions for cancer (Graves, 2003). These components are sometimes referred to as
hope, which has been associated with quality of life across multiple types of cancer. The
utility of hope has not, however, been tested longitudinally in lung cancer where quality of life
is a key treatment outcome given low survival rates and high symptom and treatment burden.
The goal of this study was to study relations between hope and quality of life (e.g., social
functioning, palliative wellbeing) in lung cancer patients undergoing treatment. Patients with
higher hope were expected to report better quality of life and daily levels of hope were
hypothesized to predict daily quality of life. Fifty lung cancer patients (58% female, 78% non-
small cell, 66% metastatic disease, average age = 68.66, SD = 8.78) completed a baseline
questionnaire and 21 daily assessments (M = 20.3 days, SD = 1.3; 1,042 days of data).
Multilevel modeling was used to analyze same-day and next-day relationships. Pre-specified
models controlled for disease (e.g., time since diagnosis, treatment type, daily disease
symptoms) and psychological (e.g., depression, daily affect) factors. Intra-class correlations
ranged from 0.64 to 0.75. Patients with higher levels of hope reported higher social and role
functioning (estimate = 3.37, SE = 0.90, 95% CI = 1.60, 5.14) and higher palliative wellbeing
(estimate = 0.88, SE = 0.24, 95% CI = 0.40, 1.35). On days where patients reported higher
hope, they reported higher social and role functioning (estimate = 2.36, SE = 0.70, 95% CI =
1.00, 3.73), better physical functioning (estimate = 2.27, SE = 0.81, 95% CI = 0.68, 3.87), and
higher palliative wellbeing (estimate = 0.91, SE = 0.13, 95% CI = 0.66, 1.16). Daily hope
appeared to attenuate the negative impact of treatment days on functioning (estimate = 3.33,
SE = 1.25, 95% CI = 0.88, 5.77). Daily hope did not, however, predict quality of life in any next-
day models. Daily hope was not predicted by cancer symptoms, which suggests that hope may
support quality of life regardless of symptom burden.

CORRESPONDING AUTHOR: Laurie E. Steffen, PhD, Wake Forest University School of Medicine,
Winston-Salem, NC, 27103; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1785

C001 6:00 PM-7:00 PM

A HEALTH MESSAGE INTERVENTION FOR INCREASING PHYSICAL ACTIVITY IN EMERGING

ADULTS AT-RISK FOR DIABETES

Stephanie R. McDonough, M.A., Lillian Slaughter, BA, Barbara A. Stetson, Ph.D.

University of Louisville, Louisville, KY

Background: Millions of adults in the U.S. have type 2 diabetes (T2D), a disease largely
preventable with health behavior change. Rates continue to rise, even among younger
populations. Emerging adulthood (18-25) is an important point in development for
establishing healthful behavior; however most do not meet CDC guidelines for physical
activity (PA). Framed health message interventions may provide an effective and easily
implemented avenue for increasing PA motivation and behavior in emerging adults.

Aim: This pilot study tests the effectiveness of a framed health message intervention focusing
on increasing PA in emerging adults at-risk for T2D.

Method: Ss were 51 sedentary emerging adults with a family history of T2D (68.6% White,
84.3% Female). At baseline and 2-week follow-up, Ss completed validated self-report
measures assessing exercise behavior (SDSCA) and motivation (exercise intention). Ss were
randomly assigned to receive a gain (n=22) or loss (n=29) framed health message to promote
regular PA in an effort to reduce T2D risk. Exercise intention and behavior scores at baseline
and follow-up were compared using paired sample t-tests. Change scores from baseline to
follow-up were calculated and compared by type of framed message using independent
sample t-tests.

Results: Intention to increase exercise was higher at follow-up (M=5.43, SD=1.31), when
compared to baseline (M=4.95, SD=1.52), t(50)=-2.009, p=.050. Frequency of exercise
behavior also increased from baseline (M=2.82, SD=1.20) to follow-up (M=3.78, SD=1.64),
t(50)=-3.663, p=.001. However, no significant differences emerged by frame of message for
exercise intention [Gain: M=.011, SD=1.91; Loss: M=-.853, SD=1.46, t(49)=1.833, p=.073], or
exercise behavior [Gain: M=-.75, SD=1.46; Loss: M=-1.12, SD=2.14, t(49)=.733, p=.467].

Conclusion: Findings suggest that health messages may be an effective strategy for
increasing exercise motivation and behavior in sedentary emerging adults at-risk for T2D,
regardless of message valance. Consistent with previous literature, it was expected that gain-
framed messages would promote greater exercise intention when compared to loss.
However, no differences by message valence emerged. It is possible that simply bringing T2D
risk to awareness, coupled with concrete behavioral suggestions for reducing disease onset,
S1786 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

may be sufficient for promoting health behavior change. More research examining framed
health messages in emerging adults is needed, including exploration of potential mediators
and moderators. Development of efficacious interventions promoting health behavior change
in emerging adults is crucial for reducing disease risk.

CORRESPONDING AUTHOR: Stephanie R. McDonough, M.A., University of Louisville, Louisville,

KY, 40217; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1787

C002 6:00 PM-7:00 PM

A TOBACCO-FREE CAMPUS POLICY REDUCES CIGARETTE SMOKING AND USE OF ELECTRONIC

NICOTINE DELIVERY SYSTEM DEVICES

Sumner Sydeman, Ph.D., Donna Zastrow, Student, Kevin Madden, Student

Northern Arizona University, Flagstaff, AZ

Cigarette smoking is the number one preventable cause of death and disease in the United
States; 90% of all adult smokers start smoking in their teens (Bach, 2016; USDHHS, 2014). As
part of tobacco prevention efforts, college campuses in the United States have started to
implement smoke-free policies in an effort to reduce cigarette smoking and use of Electronic
Nicotine Delivery System (ENDS) devices (e.g., electronic cigarette, electronic hookah). The
purpose of the current research project was to examine the effect of a tobacco and smoke-
free campus policy on cigarette smoking and ENDS use in college freshman living on campus
at a state university in the Southwest United States. Self-report data from 1,419 freshmen
(median age 18, 84% female) were collected the year before the policy took place (Fall 2015
semester) and immediately after the policy was enacted (Fall 2016 semester). The prevalence
rate of cigarette smoking (past 30-day use) dropped from 6.4% in Fall 2015 to 4.0% in Fall
2016, a statistically significant reduction (p = .048). Past 30-day ENDS use decreased from
18.1% in Fall 2015 to 11.7% in Fall 2016, which was also statistically significant (p = .001).
These results provide empirical support that a campus-wide tobacco and smoke-free policy
resulted in less cigarette smoking and ENDS use by freshman living on campus. These findings
are important because smoking has negative effects on health, and these data suggest that
enacting smoke-free policies on college campuses can reduce smoking and ENDS use.

CORRESPONDING AUTHOR: Sumner Sydeman, Ph.D., Northern Arizona University, Flagstaff,

AZ, 86011; [email protected]
S1788 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C003 6:00 PM-7:00 PM

AMBIVALENCE MODERATES THE EFFECT OF MOTIVATIONAL INTERVIEWING ON THE

PROMOTION OF INTERNAL MOTIVATION TO DONATE BLOOD

Kristen R. Fox, M.S.1, Christopher France, PhD1, Janis France, PhD1, Bruce Carlson, PhD1, Lina
Himawan, M.A.1, Victoria Frye, MPH, DrPH2, Louisa Duffy, B.A.3, Debra Kessler, MSRN3, Mark
Rebosa, MBA3, Beth Shaz, MD3

1
Ohio University, Athens, OH; 2City University of New York School of Medicine, New York, NY;
3
New York Blood Center, New York, NY

BACKGROUND: Having a strong internal motivation to donate blood is positively associated

with donation intention and behavior. At a group level, participation in a brief motivational
interview designed to bolster individual autonomy related to the decision to donate has been
shown to increase internal motivation, donation intention, and actual donation behavior.
However, at an individual level, donation-related ambivalence may moderate the effect of a
motivational interview as individuals who hold ambivalent attitudes may be more amenable
to change. Accordingly, the present study sought to determine if individual differences in
ambivalence moderate the effect of a motivational interview on internal motivation to donate
blood.

METHODS: A sample of recent blood donors (N = 299) was randomly assigned to receive an
intervention that either did or did not contain a brief, telephone-based motivational interview
aimed at enhancing individual autonomy related to the decision to donate blood. Measures of
ambivalence and autonomy to give blood were administered at least one week before and
after the interview.

RESULTS: Consistent with a moderation effect, regression analyses revealed a significant

interaction between baseline ambivalence and interview group on post-intervention
autonomy, while controlling for baseline autonomous motivation (p < .05). Specifically,
compared to no-interview controls, participation in a motivational interview was associated
with significantly higher levels of autonomous motivation among donors with high initial
ambivalence (1 SD above the mean) or average levels of initial ambivalence, but no significant
differences were observed among donors with low initial ambivalence (1 SD below the
mean).

CONCLUSIONS: Ambivalent donors appear to derive more benefit from a brief motivational
interview designed to enhance autonomous decision making regarding future blood donation.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1789

Additional research is needed to determine whether the promotion of internal motivation

among ambivalent donors is related to significant changes in donation intention and
behavior.

CORRESPONDING AUTHOR: Kristen R. Fox, M.S., Ohio University, Athens, OH, 45701;
[email protected]
S1790 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C004 6:00 PM-7:00 PM

BREAST CANCER SURVIVORS ATTRIBUTIONS OF LACK OF EXERCISE AS CAUSES FOR CANCER

DIAGNOSIS AND RECURRENCE RISK

Rachel Hirschey, RN, PhDc1, Isaac Lipkus, PhD2

1
Duke University, Durham, NC; 2Duke University School of nursing, Durham, NC

Regular exercise (i.e., 150 vigorous/moderate intensity weekly minutes) is associated with
and thus recommended to decrease breast cancer diagnosis and recurrence risk. However,
most breast cancer survivors (BCS), do not meet this recommendation. Understanding the
extent to which BCS attribute exercise influencing their diagnosis and especially recurrence
can inform how to motivate exercise. Thus we explored the extent to which BCS attribute lack
of exercise to their initial cancer diagnosis and possibility of cancer recurrence, and how the
latter attributions are related to perceived recurrence risk and levels of exercise.

BCS 2 months - 10 years post treatment were recruited through oncologist referral by mail or
in person, at regular follow-up visits. BCS completed online questionnaires. Exercise was
measured using the Godin scale. Attributions for initial diagnosis and recurrence were
assessed by What do you think caused you to be diagnosed with breast cancer? and What
if anything may cause your breast cancer to come back?. Risk recurrence perceptions were
assessed by What do you think is your chance that your breast cancer will come back in your
lifetime if you do not exercise? (1=no chance 7=certain to happen) and Compared to other
breast cancer survivors your age and race, what do you think is your chance that your breast
cancer will come back in your lifetime if you do not exercise? (1=much below average
7=much above average).

The sample (n=111) mean age is 57 (SD=10), mean exercise is 147 min/wk (SD=210), is 87%
Caucasian, and most (81%) have some college education. The two most common attributions
for initial diagnosis were stress and genetics (both cited by 24% of sample); lack of exercise
was mentioned by only 8% (n=9) of the sample. With respect to recurrence risk, lack of
exercise was the second most common attribution reported by 22% of the sample, behind
only diet, mentioned by 26%. Of import, BCS who mentioned lack of exercise as increasing
recurrence risk reported exercising more weekly minutes than survivors who did not mention
lack of exercise (169 (SD=229) vs. 68 (SD=86), p
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1791

BCS attribute lack of exercise as a more frequent cause for possible recurrence than initial
diagnosis. This attribution, along with diet, suggests that BCS perceive causes more under
their control for recurrence than for an initial event. This may serve the purpose of
empowering BCS to believe they have control over lowering their recurrence risk. The
heightened perceptions of recurrence risk among those who mentioned versus did not
mention lack of exercise may serve as a motivational cue. Further studies are needed to
determine how best to incorporate attributions of recurrence risk to motivate exercise.

CORRESPONDING AUTHOR: Rachel Hirschey, RN, PhDc, Duke University, Durham, NC, 27705;
[email protected]
S1792 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C005 6:00 PM-7:00 PM

CHANGES IN BMI, SELF-REPORTED FAMILY NUTRITION, PHYSICAL ACTIVITY, AND SEDENTARY

BEHAVIORS IN THE E.P.I.C. KIDS STUDY

Bonita Sur, M.S.1, Kathylynn Saboda, M.S.1, Scott Going, PhD1, Denise Roe, DrPH1, Craig S.
Stump, MD, PhD2, Randa Kutob, MD, MPH3, Melanie D. Hingle, PhD, MPH, RD1

1
University of Arizona, Tucson, AZ; 2University of Arizona College of Medicine, Southern
Arizona VA Health Care System, Tucson, AZ; 3University of Arizona College of Medicine,
Tucson, AZ

The prevalence of overweight and obesity in youth continues to be high (Ogden et al., 2016).
Obesity in youth is strongly associated with obesity into adulthood, as well as increased risk
for chronic diseases, including type 2 diabetes (T2D) (Field, Cook, & Gillman, 2005; Herman,
Craig, Gauvin, & Katzmarzyk, 2009). Parents shape their childrens home nutrition and
physical activity environments, making them central agents for prevention and treatment of
weight-related problems. However, few effective community-based programs aimed at
reducing obesity prevalence and diabetes risk within the family environment exist. In EPIC
Kids, we tested the effect of a 12-week family-focused YMCA-based lifestyle intervention in
forty-eight 9-12-year-old obese children at risk of T2D and their parents, on behavioral,
anthropometric, and environmental T2D risk factors. We hypothesized that a supportive
home nutrition, physical activity, and media environment would be associated with a lower
child BMI Z score. Child height and weight was measured by trained technicians. Age and sex
specific BMI percentiles were determined using CDC growth charts. Self-reported family food,
physical activity, and media environment was assessed at baseline and follow-up using the
Family Nutrition and Physical Activity (FNPA) questionnaire, a 20-item survey completed by
parents focusing on nutrition, physical activity, and sedentary behavior. Descriptive statistics
and t-tests were calculated. Difference scores were calculated as week 12 minus baseline
scores. BMI Z scores are age and gender standardized using CDC standards. Regression
methods were used to explore the relationship between BMI Z scores and FNPA scale and sub
scales. Thirty-six families completed post-intervention (12-week) measurements.
Improvement (reduction) in BMI Z score (p = 0.02) and in family nutrition (p = 0.0001) and
physical activity (p = 0.002) behaviors were observed from baseline to 12 weeks. No
significant changes in sedentary behaviors were observed (p = 0.063). Similar to previous
research using the FNPA instrument (Ihmels, Welk, Eisenmann, Nusser, & Myers, 2009),
baseline FNPA scores trended toward predicting change in BMI Z scores although did not
achieve significance (p = 0.241). Limitations include potential measurement error from self-
report and small sample size. These preliminary findings demonstrate the potential for a
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1793

YMCA-based lifestyle intervention to improve child BMI Z score, and parent-reported health
behaviors, although change in BMI Z score did not predict changes in health behaviors.

CORRESPONDING AUTHOR: Bonita Sur, M.S., University of Arizona, Tucson, AZ, 85721;
[email protected]
S1794 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C006 6:00 PM-7:00 PM

EXAMINING FAMILY AND GROUP-LEVEL DIFFERENCES ASSOCIATED WITH SELF-MONITORING

IN THE FAMILIES IMPROVING TOGETHER (FIT) TRIAL

Tyler C. McDaniel, M.S., Allison Sweeney, PhD, Dawn K. Wilson, PhD, Juanita Hawkins, B.S.

University of South Carolina, Columbia, SC

The Families Improving Together (FIT) for Weight Loss Project is a randomized controlled trial
testing the efficacy of a motivational and family-based program for improving weight loss
among underserved African American adolescents and their parents. Project FIT integrates
face-to-face and online sessions to build skills around sustaining healthy lifestyle
improvements, including self-monitoring and goal-setting. Parents and adolescents are
encouraged to monitor and track their caloric intake and energy expenditure. Intervention
facilitators review and provide feedback on participants self-monitoring logs. Although
numerous studies have shown that self-monitoring is a critical behavior-change skill, relatively
little is known about the conditions under which people are most likely to self-monitor. The
present research extends past research by examining how: 1) aspects of the family structure
(including marital status and education-level of the parent) and 2) a facilitators ability to
implement behavioral skills training in a given cohort (as assessed by independent process
evaluators) impact self-monitoring among parents and adolescents. A self-monitoring score
was created by summing the total number of weeks (range: 0- 6) that participants reported
engaged in self-monitoring. Among parents (N = 72, Mage = 43.77, 93% female, MBMI = 37.85,
Mincome < $40,000), those who were married tracked significantly more weeks (M = 2.90
weeks) compared to those who were not married (M = 2.00; p < .05). Teens (N = 72, Mage =
13.12, 69.4% female, MBMI% = 97.02) with parents that had at least a college degree tracked
significantly more (M = 2.60 weeks) compared to those with parents that had two years of
college or less (M = 1.80; p < .05). Furthermore, among cohorts with relatively high behavioral
skills implementation, adolescents engaged in greater behavioral tracking (M = 2.54) than
cohorts with relatively low behavioral skills implementation (M = 1.76; p = .03). There were no
differences in self-monitoring by age, BMI, or sex. The present study provides preliminary
evidence that aspects of the family structure as well as group-level factors contribute to self-
monitoring. Given the importance of self-monitoring for promoting health behavior change,
the present results suggest that family structure and parent capacity should be considered
when designing future weight loss interventions for underserved African Americans.

CORRESPONDING AUTHOR: Tyler C. McDaniel, M.S., University of South Carolina, Columbia,

SC, 29201; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1795

C007 6:00 PM-7:00 PM

EXAMINING HEALTH LITERACY, SELF-EFFICACY, AND HEALTH BEHAVIORS IN COLLEGE

STUDENTS WITH CHRONIC CONDITIONS

Jeremy Barsell, B.S., Samantha A. Miadich, M.A. , Michael Trujillo, M.A., Robin S. Everhart, PhD

Virginia Commonwealth University, Richmond, VA

To date, research on how students adapt to college life in the context of managing their
illnesses and their experiences with a chronic condition are limited. In this study, health
literacy and self-efficacy were examined in their associations with health behaviors (e.g.,
general behaviors, wellness maintenance, substance use). Health literacy, an understudied
construct, has high relevance for college students considering many may have recently
reached legal age and are now responsible for seeking and understanding how to use health
services. Participants included 153 students with a chronic condition (18-37 years; 67%
female; 40% White; 23% Black/African-American; 16% Asian; 26% asthma; 6% allergic rhinitis;
4% heart condition) who completed a series of online questionnaires. Health behaviors were
categorized into general behaviors (e.g., I get enough sleep), wellness maintenance (e.g., I
see a doctor for regular checkups), and substance use behaviors (e.g., I do not drink
alcohol). General health behaviors were associated with each subscale of health literacy
(functional, r=.253, p=.003; communicative, r=.459, p < .000; critical, r=.387, p < .000) and self-
efficacy (management, r=.440, p < .000; general, r=.389, p < .000, outcomes r=.404, p < .000)
.Wellness maintenance behaviors were associated with each subscale of health literacy
(functional, r=.206, p=.014; communicative, r=.388, p < .000; critical, r=.485, p < .000), as well
as self-efficacy (management, r=.424, p < .000; general, r=.340, p < .000, outcomes r=.314, p <
.000). Fewer substance use behaviors, however, were only associated with communicative
health literacy (r=.193, p=.023), critical health literacy (r=.277, p=.001), and self-efficacy to
achieve outcomes (r=.233, p < .000). Findings suggest that higher levels of both health
literacy and self-efficacy may have an impact on general health and wellness maintenance
behaviors in college students with chronic conditions. On the other hand, fewer substance
use behaviors were associated with only one dimension of self-efficacy and two dimensions of
health literacy. It may be that in college, students engage in substance use behaviors to
appear more normative to peers, regardless of their own self-efficacy. More research is
needed on college students with chronic conditions in order to elucidate the mechanisms by
which health literacy and self-efficacy can reduce substance use, and also improve positive
health behaviors.

CORRESPONDING AUTHOR: Jeremy Barsell, B.S., Virginia Commonwealth University,

Richmond, VA, 23223; [email protected]
S1796 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C008 6:00 PM-7:00 PM

INTERVENTION MEDIATORS IN AN RCT TO INCREASE COLONOSCOPY UPTAKE AMONG THOSE

AT INCREASED RISK FOR FAMILIAL COLORECTAL CANCER

Barbara H. Brumbach, Ph.D.1, Wendy C. Birmingham, PhD2, Watcharaporn Boonyasiriwat,

N/A3, Scott Walters, PhD4, Anita Y. Kinney, Ph.D.; RN1

1
University of New Mexico Comprehensive Cancer Center, Albuquerque, NM; 2Brigham Young
University, Provo, UT; 3Chulalongkorn University, Bangkok, N/A, Thailand; 4University of North
Texas Health Science Center, Ft. Worth, TX

Background: Understanding pathways by which interventions achieve behavioral change is

important for optimizing intervention strategies. The Extended Parallel Process Model and
implementation-intention strategies were theoretical guides to the development and
implementation of the intervention. Purpose: We examined mediators of behavior change in
a tailored-risk communication intervention that increased guideline-based colorectal cancer
(CRC) screening among individuals at increased familial risk. Methods: Using a population-
based recruitment strategy, participants at increased familial risk for CRC (N=481) were
randomized to one of two intervention arms: 1) remote, tailored-risk communication
(TeleCARE) or 2) mailed educational brochure. Results: A mediation model showed that the
intervention lead to increased colonoscopy uptake through several constructs (threat,
efficacy, emotions, knowledge, intentions). Structural equation modeling showed that
TeleCARE participants were more likely to get a colonoscopy. The effect was partially
mediated through threat perceptions (=0.12, pConclusion: Evaluating mediating variables
between an intervention (TeleCARE) and a primary outcome (colonoscopy) contributes to
understanding underlying mechanisms that lead to health behavior change, thus leading to
better informed future interventions.

CORRESPONDING AUTHOR: Barbara H. Brumbach, Ph.D., University of New Mexico

Comprehensive Cancer Center, Albuquerque, NM, 87131-0001; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1797

C009 6:00 PM-7:00 PM

MATERNAL BMI CHANGE LINKED TO CHILD ACTIVITY CHANGE IN FAMILY-BASED BEHAVIORAL

INTERVENTIONS FOR PEDIATRIC WEIGHT MANAGEMENT

Ashleigh A. Pona, MA1, Jordan A. Carlson, PhD2, Robin P. Shook, PhD2, Meredith L. Dreyer
Gillette, PhD3, Ann M. Davis, PhD, MPH, ABPP4

1
University of Missouri - Kansas City, Kansas City, MO; 2Children's Mercy Hospital, Kansas City,
MO; 3Children's Mercy Kansas City, Kansas City, MO; 4University of Kansas Medical Center,
Kansas City, MO

Objective: Family-based behavioral interventions (FBBIs) for pediatric weight management

encourage weight loss through modification of the familys nutritional intake and physical
activity (PA) habits using behavioral techniques. Studies on FBBIs typically focus on child
weight change as the primary outcome. Little is known about factors associated with change
in PA in FBBIs, but evidence suggests that parent modeling and family environment influence
childrens overall diet and PA habits and are critical components in FBBIs. We investigate
whether change in maternal body mass index (BMI) is associated with change in childs
moderate to vigorous PA (MVPA), prolonged sedentary behavior (PSB), and BMIz during the
course of FBBIs for pediatric weight management.

Methods: Youth (n=126) ages 5-13 (mean age=9.031.7) years with a baseline BMI 85th
percentile (mean BMI%=94.664.6) and their families were enrolled in one of three similar
FBBIs for pediatric weight management and followed over a period of 6-12 months,
depending on length of the FBBI. Child and maternal height and weight were measured by
trained clinicians using standardized procedures. Activity data were collected via
accelerometers. Residualized change scores were computed for each variable of interest
(dependent variables: childrens BMIz, minutes/day of MVPA, proportion of day consisting of
PSB; independent variables: maternal BMI) and each covariate (total wear time, weekdays
worn, weekend days worn) by regressing follow-up (T2) score on baseline (T1) score. Other
covariates included child age, gender, race, study, length of FBBI, and treatment session
attendance.

Results: From T1 to T2, maternal BMI decreased from 30.476.3 to 30.256.0 (p=.26), child
BMIz decreased from 1.80.5 to 1.76.5 (p=.04), child MVPA decreased from 36.6723.4 to
32.0919.1 minutes/day (p=.02), and proportion of day consisting of child PSB increased from
12.068.0% to 14.7811.3% (p=.01). Regression analyses indicated a decrease in maternal
BMI attenuated both the decrease in child MVPA, B=2.77, p=.048, and the increase in
S1798 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

proportion of day consisting of child PSB, B=-.02, p=.02, from T1 to T2. Change in maternal
BMI was not associated with change in child BMIz (B=.03, p=.29) from T1 to T2.

Discussion: Improvement in maternal BMI showed important positive associations with child
MVPA and PSB over the course of FBBIs for pediatric weight management. Our findings
support the theoretical construct that enabling mothers to improve their weight status via
health behavior changes as part of a pediatric weight management intervention is important
in supporting healthy behavior change in their children. Future research should consider
directly measuring maternal PA and/or target maternal health habits more directly.

CORRESPONDING AUTHOR: Ashleigh A. Pona, MA, University of Missouri - Kansas City, Kansas
City, MO, 64111; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1799

C011 6:00 PM-7:00 PM

MOTIVATIONAL PROCESSES AND HEALTH BEHAVIOR OUTCOMES AMONG HIGH-RISK LATINAS

PARTICIPATING IN A LIFESTYLE INTERVENTION

Kelly Biegler, PhD1, Mienah Sharif, PhD1, Laura Arias, BA1, Camila Aguilar, Undergraduate
Student1, Jennifer Escareno, Undergraduate Student1, Emily Dow, MD1, David Kilgore, MD1,
Margaret Schneider, PhD2, Karen S. Rook, PhD1, Dara H. Sorkin, PhD3

1
UC Irvine, Irvine, CA; 2University of California, Irvine, Irvine, CA; 3University of California
Irvine, Irvine, CA

Purpose: Mexican-American women have one of the highest lifetime risks for developing
diabetes, which may stem, in part, from high levels of obesity and physical inactivity.
Research to date has focused on behavioral, socioeconomic, or environmental risk factors,
while few studies have examined the motivational systems underlying poor health behaviors.

Behavioral inhibition and activation are two discrete, opposing motivational systems that are
posited to underlie behavior and mood, and have been widely used as a framework for
examining health behaviors associated with obesity and chronic disease. The behavioral
inhibition system (BIS) promotes avoidance of punishment or negative emotions, and has
been associated with higher levels of anxiety. The behavioral activation system (BAS)
promotes goal directed behavior and is sensitive to reward, and has specifically been
implicated in increased intake of high-sugar, high-fat foods and obesity.

We completed a randomized trial testing the effectiveness of a dyadic weight loss

intervention in Mexican-American mothers with type 2 diabetes and their obese, adult
daughters. Despite high rates of perceived stress in this population, no study has considered
the complex interplay between motivation, stress, social relationships and health
characteristics of two individuals pursuing a shared health behavior change in the context of a
weight loss intervention in this under-resourced, high-risk population.

Objective: The study examined baseline associations between mothers and daughters
ratings of BIS/BAS and: (1) measures of perceived stress and positive and negative social
interactions; and (2) self/other health behavior characteristics (i.e., nutrition intake, BMI).

Method: Data was derived from 89 mother-daughter Mexican-American dyads (N=178).

Results: Higher BIS scores were associated with higher levels of perceived stress (r=0.30, p <
0.001). However, higher BAS scores were associated with lower levels of perceived stress (r=-
S1800 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

0.23, p=0.02) and negative social interactions (r=-0.26, p=0.007). In addition, higher BAS
scores were associated with greater saturated fat intake (r=0.21, p=0.01). Additional analyses
examined the differential associations between mother/daughter ratings of self and other
health behaviors

Discussion: The findings from this study highlight the importance of considering motivational
states in characterizing risk factors of chronic disease and improving health outcomes in this
high-risk population.

CORRESPONDING AUTHOR: Kelly Biegler, PhD, UC Irvine, Irvine, CA, 92697; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1801

C012 6:00 PM-7:00 PM

NUTRITION-BASED HEALTHY LIFESTYLE PILOT PROGRAM FOR FEMALE HOSPITAL EMPLOYEES

Iva Greenshtein-Littman, MPH1, Donna Zwas, MD, MPH2

1
Hadassah Hospital, Jerusalem, Yerushalayim, Israel; 2Cardiology Department, Hadassah
University Medical Center, Jerusalem, Yerushalayim, Israel

Background: Behavioral risk factors for chronic disease (i.e. smoking, unhealthy diet,
sedentary lifestyle) are prevalent in hospital employees (HE). Female HE are at a particular risk
given shift work and stressful work environment. Gender-tailored interventions may improve
outcomes. Identifying effective methods for increasing HL among HE can inform public health
interventions in hospital-based and employee interventions worldwide.

Objectives: This study explored a gender-tailored nutrition-based healthy lifestyle (HL)

program for female HE in a Jerusalem hospital, assessing outcomes of health behaviors and
BMI. It consisted of 8 weekly sessions on the Mediterranean diet and HL monitoring, including
physical activity (PA), with emphasis on providing tools likely to improve outcomes in women
(social support, social and emotional eating, etc.). Participants completed surveys (asking
about nutrition, PA and other HL parameters) and had BMI and blood pressure measured
before and after the program. Program objectives included increased adherence to the
Mediterranean diet, increased PA, reduced smoking, and reduced BMI.

Results: 48 women (ages 36-67) completed surveys. 46% were nurses and 25% from hospital
administration. They reported increased consumption of vegetables (29.8%, p2 (p

Conclusions: This gender-tailored HL pilot program was effective in helping female HE

improve adherence to the Mediterranean diet, reduce smoking, and achieve healthier BMIs.
Tailoring HL programs to women may increase HL in female HE.

Main Messages:

Gender-tailored nutrition-based healthy lifestyle programs can be effective at

empowering female hospital employees to adopt healthy behaviors.
S1802 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Iva Greenshtein-Littman, MPH, Hadassah Hospital, Jerusalem,

Yerushalayim, 9371210; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1803

C013 6:00 PM-7:00 PM

PATTERNS OF TOTAL AND SATURATED FAT CONSUMPTION AND THEIR ASSOCIATIONS WITH
WEIGHT CHANGES OVER A 12-MONTH WEIGHT LOSS STUDY

Yaguang Zheng, PhD, RN1, Susan Sereika, PhD2, Qianheng Ma, MS2, Cynthia Danford, PhD2,
Christopher Imes, PhD2, Yang Yu, MSN2, Ran Sun, MSN2, Rachel Goode, MSW2, Juliet
Mancino, MS2, Lora Burke, PhD, MPH2

1
Boston College, Chestnut Hill, MA; 2University of Pittsburgh, Pittsburgh, PA

Background: Established guidelines recommend restriction of total fat intake to 20-35%

and saturated fat to < 10% of total calories for both general public and those in weight loss
studies. To date, no study has examined temporal patterns of fat intake change. Our aim was
to identify distinct trajectories of total and saturated fat intake and their associations with
weight change.

Methods: This was an analysis of data from a 12-month behavioral weight loss study where
participants were asked to reduce total fat intake to 25%. Dietary intake was recorded using
the Lose It! app on a smartphone and averaged weekly. Weight was measured via a digital
scale every 6 months. Group-based trajectory and mixed modeling were used for analyses.

Results: The sample (N=148) was 90.5% female, 81.1% White, with a mean age (SD) of
51.310.1 years and a mean BMI of 34.14.6 kg/m2. Three trajectory groups were found for %
calories from total fat: recommended & decreasing (n=46, 22.4-26.3%); recommended &
increasing (n=68, 28.9-33.0%); high & increasing (n=34, 32.7-42.5%). For % calories from
saturated fat, 3 trajectory groups were found: recommended & consistent (n=35, 5.3-7.3%),
recommended & increasing (n=89, 8.6-10.0%); high & increasing (n=24, 11.1-16.5%).
Significant differences in % weight change were found among total fat trajectory groups
(p=.04), with mean % weight losses of -10.67.4 for the recommended & decreasing group, -
9.18.8 for recommended & increasing group and -5.8-6.7 for the high & increasing group.
We found no differences in % weight loss among saturated fat trajectory groups (p=.54).

Conclusions: Nearly one third of sample consistently followed recommendations for fat intake
and had greater weight loss. Those consuming fat in excess of recommended levels had less
weight loss and could benefits from additional strategies to assist in the development of skills
and habits to lower fat intake to support weight loss and improve cardiovascular health.
S1804 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Yaguang Zheng, PhD, RN, Boston College, Chestnut Hill, MA,
02467; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1805

C014 6:00 PM-7:00 PM

PILOT TEST OF A SUN SAFETY PROGRAM FOR THE TATTOO COMMUNITY

Robert Dellavalle, MD, PhD, MSPH1, Barbara Walkosz, PhD2

1
University of Colorado Denver, Denver, CO; 2Klein Bendel, Inc., Wheat Ridge, CO

Sun Safety Ink! (SSI!) is a skin cancer prevention program developed for the tattoo
community. Over 5.4 million cases of non-melanoma skin cancers occur annually and
melanoma rates continue to increase. Invasive melanoma of the skin is the third most
common cancer among adolescents and young adults (ages 15-39). Concurrently, the
popularity of tattoos continues to grow in the U.S. and among members of the military. One in
four American adults (23%) have a tattoo, and younger adults, [38% of the Millennial
Generation and 32% of Generation X] are more likely to have tattoos than older adults.
Tattooed individuals need to take extra precautions not only to protect themselves from UVR
but also to reduce damage to and preserve their tattoos. Sun safety is sometimes promoted
with tattoo aftercare instructions.

The SSI! program included a training for tattoo artists with information on skin cancer and
tattoo rates, the relationship of UV and the skin, and skin cancer risk factors and prevention
strategies along with a tip card to be distributed with the tattoo studios aftercare
instructions.

55 tattoo studios were invited to participate. 18 (32%) enrolled, 5 (9%) refused, and 32
(55%) did not respond. [4 dropped out after enrollment and randomization]. 14 tattoo studios
(6 intervention [I], 8 control[C]) participated. 77 clients completed posttest surveys after
receiving a tattoo. Most tattoos were on the arms (50%) and upper torso (47%). 80% reported
a sunburn in the past 12 months; sun protection practices were low (1= never to 3=always):
sunscreen, M=2.16; wide-brimmed hat, M= 1.53; long sleeve shirts and pants, M=1.69;
sunglasses, M=2.57; seek shade, M=1.90. 15.5% engaged in indoor tanning. Strong tanning
preferences were reported.

85% (87% Intervention[I] v. 80% Control [C]) received sun protection information in
the studio: verbal message from tattoo artist: 93% (95%I v. 86%C]; written aftercare
instruction: 92% (90%I v. 100 %C), studio website: 54% (50%I v. 61%C): social media, poster or
sign: 51% (55%I v 41%C), tip card brochure, 73% (76%I v. 64%C).
S1806 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

59% (56%I v. 68% C) of respondents said messages were about protecting tattooed skin
and 40% (44%I v. 31% C), about protecting all of your skin. The results of pilot study support
the need and feasibility to move forward to further test SSI!

CORRESPONDING AUTHOR: Robert Dellavalle, MD, PhD, MSPH, University of Colorado Denver,
Denver, CO, 80220; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1807

C015 6:00 PM-7:00 PM

POSITIVE MOOD IS ASSOCIATED WITH DAILY FRUIT AND VEGETABLE CONSUMPTION DURING
AN INCENTIVE INTERVENTION

Casey K. Gardiner, MA, Angela D. Bryan, PhD

University of Colorado Boulder, Boulder, CO

Monetary incentive interventions to promote health behavior change have gained recent
popularity in both academic research and public health practice, and they have been deemed
effective in numerous health domains. However, much remains unknown regarding the
effects of these interventions on psychosocial factors associated with overall health and well-
being, including affect and perceived stress.

The present research explores these relationships in the context of a brief randomized
controlled trial designed to promote fruit and vegetable (FV) consumption using monetary
incentives. Participants (n = 60) were randomized to receive either daily incentives, delayed
lump sum incentives, or no incentives based upon their FV consumption. Over the course of
21 days, they completed daily reports of their FV consumption, affect, and stress.

Multilevel modeling was employed to examine within- and between-participant relationships

between intervention condition, health behavior engagement (i.e., FV consumption), affect,
and stress. Results demonstrated a significant relationship between affect and FV
consumption both within and across days, over and above the effect of intervention condition
( = 0.90, p = .02). Specifically, on days when participants reported more positive affect, they
consumed more FV ( = 0.07, p = .04). Additionally, participants who experienced more
positive affect across days also consumed more FV ( = 0.52, p = .04). Daily stress, however,
was not significantly associated with FV consumption overall, though a stress X condition
interaction reached a trending level ( = 0.08, p = .10), such that an effect of reduced FV
consumption on days when participants reported more stress was less strong for participants
receiving incentives than those in the control condition.

In sum, this research presents a first examination of the within- and across-day effects of
monetary incentives, affect, and stress on FV consumption. Notably, participants mood was
positively associated with their FV consumption both within and across days, indicating that
healthier behavior is linked to more positive affect, and these relationships were not
adversely affected by the presence of incentives.
S1808 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Casey K. Gardiner, MA, University of Colorado Boulder, Boulder,

CO, 80309; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1809

C016 6:00 PM-7:00 PM

POTENTIAL DRIVERS OF DIETARY CHANGE IN AFRICAN AMERICAN CONGREGANTS

Elizabeth Lynch, PHD1, Laurin Mack, PHD1, Yamin Wang, MS1, Elizabeth Avery-Mamer, MS2

1
Rush University Medical Center, Chicago, IL; 2Rush University Medical Center, Morton Grove,
IL

Introduction: Cardiovascular morbidity and mortality are higher in African Americans than
whites, partly due to differences in diet quality. African Americans tend to have poorer diet
quality than whites, and in particular, eat fewer vegetables. Increased vegetable consumption
is associated with decreased cardiovascular and all-cause mortality. The ALIVE intervention is
a church-based intervention, designed by a partnership among researchers, pastors and
church leaders, to increase vegetable consumption among African Americans. A pilot study
showed that ALIVE was successful at increasing average vegetable consumption by one
serving over 9 months. The purpose of the current analysis is to investigate individual and
social factors related to increased vegetable consumption.

Method: The ALIVE study was a 9-month single-arm trial that was conducted in 5 African
American churches. Participants were exposed to both church-wide and individual-level
intervention components and completed assessments at baseline and 9 months, including 24-
hour dietary recalls and measures of self-efficacy, nutrition knowledge, social norms, and
social support. We measured associations between these psychosocial variables and whether
participants increased vegetable intake by 1 servings per day.

Results: Across the five participating churches, participants (n=206) were mostly female with a
range of socioeconomic status (25% below poverty). 182 participants (88%) completed at
least one dietary recall at 9-month follow-up. Vegetable consumption increased by 1 serving
in 46.6% of participants. Self-efficacy, nutrition knowledge, social norms and social support
increased during the intervention (all ps

Discussion: These findings suggest that clinically significant changes in vegetable consumption
may be related to both individual-level and social factors. Interventions that focus solely on
individual-level factors may be failing to leverage the importance of social norms in driving
dietary behavior.

CORRESPONDING AUTHOR: Elizabeth Lynch, PHD, Rush University Medical Center, Chicago, IL,
60625; [email protected]
S1810 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C017 6:00 PM-7:00 PM

PRECONCEPTION HEALTH (PCH) BEHAVIORS: EXAMINING THE GAP BETWEEN KNOWLEDGE

AND INTENTIONS

Kirsti Toivonen, HBA1, Kirsten Oinonen, PhD1, Katelyn Duchene, MA1, Rosemary Scofich, RN2

1
Lakehead University, Thunder Bay, ON, Canada; 2Thunder Bay District Health Unit, Thunder
Bay, ON, Canada

Background: Positive PCH behaviors (i.e., health changes made before pregnancy) are
important to maximize offspring health. While studies suggest the general public has
knowledge of the importance of PCH behaviors, even many of those with planned pregnancies
do not engage in them. Little research has examined the intention to engage in specific PCH
behaviors. Intentions may represent an intermediate step between PCH knowledge and
behavior change and may help explain the knowledge-behavior gap.

Methods: As part of a larger study on PCH knowledge and behaviors, 170 women from
Thunder Bay, Canada between the ages of 18 and 45 were surveyed regarding their
knowledge of the importance of, and the intention to engage in, 12 different PCH behaviors
prior to a potential future pregnancy (i.e., avoiding alcohol, tobacco, second-hand smoke,
toxic chemicals, personal care products with parabens/phthalates, toxic-cleaners, and fish
high in mercury; reducing stress and caffeine; improving sleep and physical activity; and taking
folic acid). Women were asked whether they intend to engage in the behaviors: in the months
before pregnancy, when they begin to attempt to conceive, when they know they are
pregnant, later in pregnancy, or not at all.

Results: Across the 12 behaviors studied, a mean of 86% of women reported knowledge of
the importance of engaging in them (ranging from 63 to 94%, who knew they should avoid
personal care products with parabens/phthalates and get adequate sleep, respectively). Of
the women with knowledge of the importance of each behavior before pregnancy, a mean of
72% of women reported the intention to engage in each behavior in the months before a
potential pregnancy (ranging from 59 to 82%). Of the women with knowledge, a mean of 28%
did not intend to engage in PCH behaviors in the months before conception (18% endorsed
waiting until trying to conceive, and 10% endorsed waiting until they know they are pregnant,
later in pregnancy, or not engaging in the behavior at all).

Discussion: Although most women report knowing the importance of engaging in positive PCH
behaviors before a potential pregnancy, a mean of 28% of women with knowledge reported
no intention to engage in specific behaviors in the months before attempting conception.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1811

Results indicate that having PCH knowledge does not necessarily lead to intentions. This may
in part explain the knowledge-behavior gap observed in PCH. Future research should
examine reasons for this gap.

CORRESPONDING AUTHOR: Kirsti Toivonen, HBA, Lakehead University, Thunder Bay, ON, P7B
5E1; [email protected]
S1812 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C018 6:00 PM-7:00 PM

PROMOTING PATIENT PARTICIPATION IN HEALTHCARE INTERACTIONS THROUGH

COMMUNICATION SKILLS TRAINING: A SYSTEMATIC REVIEW

Thomas A. D'Agostino, PhD, Thomas M. Atkinson, PhD, Lauren E. Latella, MS Ed, Madeline
Rogers, BA, Dana Morrissey, MPH, LMSW, Antonio DeRosa, MLIS, Patricia A. Parker, PhD

Memorial Sloan Kettering Cancer Center, New York, NY

Background: Communication training is a potentially effective tool for promoting patient

participation in care and helping patients to get the most out of their visits. Previous reviews
of patient communication training literature are either outdated or topical reviews. The
purpose of the current systematic review was to present an overview of the status of patient
communication training research. Our goal was to examine the content and structure of
existing training programs, the design of intervention studies, and evidence concerning the
impact of training on patient behavior, communication process, and other important
outcomes.

Methods: Systematic searches were conducted in six databases: MEDLINE (via PubMed),
Embase, The Cochrane Library, Web of Science, PsycINFO (via OVID), and ERIC (Education
Resource Information Center). References were screened for inclusion through title, abstract,
and full text review phases by a team of five researchers. Extracted data included intervention
study design, sample characteristics, content and structure of training programs, outcomes
assessed, and findings reported.

Results: A total of 32 unique intervention studies met inclusion criteria. Most targeted
primary care (50%) or cancer patients (28%) and used a randomized controlled design (59%).
Interventions used a variety of training formats (i.e., materials only, materials plus individual
coaching, and group-based) and modes of delivering educational material (e.g., written, face-
to-face, and web-based). Reported findings suggest that communication training is an
effective approach to increase patients total level of active participation in healthcare
interactions and that some communication behaviors may be more amenable to training (e.g.,
expressing concerns). Trained patients do not have longer visits and tend to receive more
information from their providers. Most studies have found no link between communication
training and improved health, psychosocial wellbeing, or treatment-related outcomes.

Conclusions: We cannot conclude that a superior patient communication training program

exists or that one method for delivering educational materials is best. If anything, the findings
of our review indicate that there are many different ways of increasing patient participation
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1813

through communication skills training. More work is needed to determine the most
efficacious training programs with the strongest potential for dissemination.

CORRESPONDING AUTHOR: Thomas A. D'Agostino, PhD, Memorial Sloan Kettering Cancer

Center, New York, NY, 10022; [email protected]
S1814 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C019 6:00 PM-7:00 PM

PSYCHOSOCIAL MEDIATORS OF PHYSICAL ACTIVITY ADOPTION IN A RCT OF AN INTERNET-

BASED INTERVENTION FOR LATINAS

Shira Dunsiger, PhD1, Dori Pekmezi, PhD2, Beth C. Bock, PhD3, Britta Larsen, PhD4, Sheri J.
Hartman, PhD5, Sarah Linke, PhD, MPH6, Kim M. Gans, PhD,MPH7, Andrea S. Mendoza-
Vasconez, MPH8, Bess H. Marcus, Ph.D.9, Becky Marquez, PhD, MPH10

1
The Miriam Hospital and Brown University, Portsmouth, RI; 2University of Alabama at
Birmingham, Birmingham, AL; 3Brown Medical School, Providence, RI; 4University of
California, San Diego, San Diego, CA; 5Department of Family Medicine and Public Health, UC
San Diego, La Jolla, CA, USA, La Jolla, CA; 6UC San Diego, La Jolla, CA; 7University of
Connecticut, Storrs, CT; 8University of California San Diego, San Diego, CA; 9University of
California, San Diego, La Jolla, CA; 10University of California San Diego, La Jolla, CA

Theory- and Internet-based interventions have shown great promise in improving physical
activity (PA) among Latinas, a group disproportionately affected by PA-related conditions
(obesity, diabetes). A better understanding of the theoretical constructs driving PA behavior
change in this at-risk group is needed. Pasos Hacia la Salud tested a culturally adapted, theory-
based, individually tailored, Internet-based Spanish language PA intervention vs. an Internet-
based Wellness Contact Control condition among 205 under-active Latinas. For both objective
(accelerometer) and subjective (7-Day PA Recall Interview) measures of PA (minutes/week),
the Intervention group had significantly greater increases in PA than controls. The current
study explores potential psychosocial mediators (self-efficacy, behavioral and cognitive
processes, social support and enjoyment) of the intervention effect on both self-reported and
objectively measured PA. A multiple mediation approach was used based on the product of
coefficients method with bootstrapped standard errors (one model for each of the primary PA
outcomes). There were significant a path coefficients (effects of intervention on mediators)
for self-efficacy, (a=.43, SE=.15,p=.004); behavioral (a=.64, SE=.14, pb paths (effect of
mediator on outcome) such that higher self-efficacy (b=24.54, SE=10.98, p=.03), social support
(friends, b=2.36, SE=1.23, p=.04) and to some degree enjoyment (b=.74, SE=.42, p=.08) were
associated with higher mean min/week of self-reported PA. The only significant b path for
objectively measured PA was for self-efficacy (b=22.36, SE=11.25, p=.04). For both PA
outcomes, there was an indirect effect of the intervention on PA through increases in self-
efficacy (ab=10.49, CI:2.46-24.54 for self-report; ab=11.11, CI:1.37-32.72 for objectively
measured). For self-reported outcomes, there was a significant indirect effect of enjoyment
(ab=7.30, CI: 0.92-21.78). Results indicate that self-efficacy and enjoyment are key
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1815

psychosocial mediators of PA change in Latinas and have important implications for future
intervention targets.

CORRESPONDING AUTHOR: Shira Dunsiger, PhD, The Miriam Hospital and Brown University,
Portsmouth, RI, 02871; [email protected]
S1816 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C020 6:00 PM-7:00 PM

QUALITY OF SLEEP AFFECTS SELF-EFFICACY FOR ADHERING TO A HEALTHY LIFESTYLE PLAN

Lora Burke, PhD, MPH1, Eileen Chasens, phd2, Christopher Imes, PhD1, Christopher Kline,
phd2, Lynn Baniak, phd2, Dara Mendez, PhD1, Stephen Rathbun, phd3

1
University of Pittsburgh, Pittsburgh, PA; 2university of pittsburgh, pittsburgh, PA; 3University
of Georgia, pittsburgh, PA

Background: Prior research suggests that sleep affects adherence to a healthy lifestyle. We
conducted a 12-month behavioral weight loss intervention study and used ecological
momentary assessments (EMA) to collect data about behaviors and emotions. This study is a
secondary analysis of the relationship between self-efficacy (SE) for adherence to a healthy
lifestyle plan and quality and quantity of self-reported sleep.

Methods: Using a smartphone, participants completed a beginning of the day (BOD) survey
reporting quality and quantity of sleep and random assessments of their level of SE for
following a healthy lifestyle. All 12 months of data were used for each participant. Linear
mixed effects models were fit including subject-dependent random effects of intercept, each
sleep variable, and study interval.

Results: The sample (N = 150) was primarily female (90.7%), White 80.7%), middle-aged
(51.1(10.2) years of age), and obese (BMI = 34.0 (4.6) kg/m2). Participants completed 99.6%
of the 44,917 BOD assessments on sleep. They reported a mean of 6.9 (1.3) hrs of sleep and
1.6 (3.5) awakenings. On a 0-10 scale, they rated how well slept as 6.5 (2.1) and trouble
falling asleep as 3.1(2.6), and feeling tired 42% of the days. SE was assessed in 133,508
random assessments and was rated as 7.0 (2.1) on a 0-10 scale. Neither trouble falling asleep
nor hours slept were significantly related to SE; however, SE increased by 0.02 points for each
unit increase in how well slept, increased by 0.15 points when subjects reported that they
were not feeling tired, and decreased by 0.02 points for each awakening the previous night.

Conclusions: While participants averaged nearly 7 hours of sleep per night, they reported
feeling tired over 40% of the mornings. Sleep duration did not affect SE; however, how well a
person slept and not feeling tired increased SE ratings, and awakenings reduced SE ratings.
The mean SE score of 7 suggests that confidence in adhering to the lifestyle met the minimum
level for engaging in adherence to lifestyle behaviors, based on the theoretical basis of the SE
construct. However this level of SE does not ensure they will sustain this behavior in the face
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1817

of challenges. Future work needs to focus on strategies to improve sleep quality among
individuals aiming for weight loss and weight loss maintenance.

CORRESPONDING AUTHOR: Lora Burke, PhD, MPH, University of Pittsburgh, Pittsburgh, PA,
15261; [email protected]
S1818 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C021 6:00 PM-7:00 PM

REACHING THE HARD TO REACH

Tricia K. Gatlin, PhD1, Michael Johnson, Ph.D., RN2

1
University of Nevada, School of Nursing, Las Vegas, NV; 2University of Nevada, Las Vegas,
Henderson, NV

Introduction/Background: Hard-to-reach is a term that has been primarily used by

researchers to describe groups of people who have been historically difficult to find or
contact. The group could be hard to reach for a myriad of reasons, such as geographical
location or social and economic situation, concealment of identity due to fear, social pressure,
or stigma or social invisibility. The term, hard-to-reach, implies that the only barrier to
recruiting the population is reaching or accessing them. Researchers know that is not the case
though. After reaching the hard-to-reach, researchers often encounter challenges in recruiting
and retaining participants. It is important for public interest to include members of hard-to-
reach in research because not including certain sub-populations in research diminishes the
ability to identify groups that potentially have the highest burden of illness and to develop an
understanding of why group differences exist. Thus, the purposes of this presentation are to:
1) describe the challenges in recruiting hard-to-reach population in two separate research
studies, 2) discuss the strategies that were used to overcome those challenges, and 3) provide
recommendations for researchers.

Methods: The authors used two of their own research studies involving hard-to-reach
populations as case studies for this paper. The research studies used in these case studies
involved two very different hard to-reach groups, (1) low-income ethnic minorities who were
un- or under-insured, and (2) lesbian and bisexual women and transgender men. Each author
was able to access their respective population but encountered numerous challenges in
recruiting participants and in collecting all of the data necessary for their study. Several
creative recruiting strategies were utilized to gain adequate sample size for these 2
populations.

Results: Through open discussion between the two authors regarding the challenges to
reaching their hard-to-reach populations, two overarching themes were identified as barriers
to reaching the population of interest: (1) gaining interest and (2) building trust. These
themes add to the literature regarding the multi-prong approach that is needed to recruit
members of hard-to-reach populations. We found that despite having buy-in from
stakeholders and a multi-prong recruiting approach, gaining interest of potential participants
included barriers such as language in recruitment flyers, competing demands for time, and
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1819

transportation to the data collection site. Building trust with interested study participates was
also a large issue noted between both studies; especially concerning sensitive questions or
cultural barriers regardless of the reliability and validity of the tools used in the study.

Discussion: It is widely believed that underserved populations are hard to reach however, it
is because they are unwilling or it is it because we dont know how to reach them? Reaching
the population of interest means that you take into consideration the opportunity to
participate. Gaining interest and building trust with your population of interest is key.
Recruiting in rich environment with known high numbers of the potential participants can
yield low sample size if one does not take the time to consider how to adequately gaining
interest and build trust with potential or actual research subjects.

CORRESPONDING AUTHOR: Tricia K. Gatlin, PhD, University of Nevada, School of Nursing, Las
Vegas, NV, 89154-3018; [email protected]
S1820 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C022 6:00 PM-7:00 PM

RECIPROCAL RELATIONSHIP BETWEEN SEDENTARY BEHAVIOR AND MOOD IN YOUNG ADULTS

Madison M. DeMello, Ph.D. 1, Shira Dunsiger, PhD2, Bernardine M. Pinto, Ph.D.3, Gregory
Hand, PhD MPH4, Steven N. Blair, PED3

1
College of Nursing, University of South Carolina, Columbia, SC; 2The Miriam Hospital and
Brown University, Portsmouth, RI; 3University of South Carolina, Columbia, SC; 4West Virginia
University, Morgantown, WV

With the increasing development of technology, a variety of leisure activities are spent in
sedentary or seated behavior. Numerous associations between sedentary behavior and
mental health outcomes such as mood have been reported, however none have looked at the
reciprocal relationship between mood and sedentary time. The purpose of this study is to
examine the relationship of sedentary behavior with mood and health status in young adults.

Methods: A sample of 430 adults (49.3% male) between the ages of 21-35 provided valid
objective activity data in addition to an assessment of their wellbeing in a year-long
observational study. For the purpose of this study, sedentary behavior is defined as less than
1.5 METS and was measured using a SenseWear mini-armband. The armband was worn for a
period of 10 days (compliance of 7 days), with at least 21 hours of verifiable wear time per
day. In addition, participants mood status and quality of life were assessed by the Profile of
Mood State (POMS) and SF-36 questionnaire respectively. Specifically, this study used the
Total Mood Disturbance (TMD) score from the POMS questionnaire to assess mood.
Assessments of sedentary behavior were conducted at baseline and then quarterly for one
year duration, while assessments of mood and health status were conducted at baseline and
one year.

Results: A cross-lagged, autoregressive clustered model was used to examine simultaneous

changes over time in both mood and sedentary behavior allowing for both clustering and
adjustment of covariates over time. Data suggests that TMD score decreased significantly
over one year, suggesting improvement in mood (p=0.01). Furthermore, there were positive
associations between sedentary time and TMD; this association increased over time
(p=0.001). Specifically, higher sedentary time was associated with more distress and this
association intensified over time. Mean sedentary time remained stable over the course of
the study (p=0.29). However, higher TMD scores (more distress) were associated with greater
mean sedentary time (p=.03), and this association remained stable over the study period
(p=0.51). Taken together, results suggest significant reciprocal associations (cross-lagged
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1821

effects) between mood and sedentary time, with the stronger predictive associations of mood
on sedentary time. Similar patterns of change over time in health status and its reciprocal
relationship with sedentary time will also be presented.

Conclusion: These results indicate a reciprocal relationship between mood and sedentary
time, that is a decrease in sedentary time should improve mood status, likewise, an improved
mood may decrease time spent sedentary. However, the stronger association is mood status
predicting time spent sedentary in young adults. By identifying predictors of seated behavior,
researchers can better understand ways to reduce time spent sedentary.

CORRESPONDING AUTHOR: Madison M. DeMello, Ph.D. , College of Nursing, University of

South Carolina, Columbia, SC, 29208; [email protected]
S1822 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C023 6:00 PM-7:00 PM

REVERENCEA PREOPERATIVE POSITIVE EMOTION PREDICT POST-OPERATIVE LENGTH OF

HOSPITALIZATION

Amy L. Ai, PhD1, Paul Wink, PhD2

1
Florida State University, Tallahassee, FL; 2Wellesley College, Wellesley, MA

There has been growing diversification of the American religious landscape (Pew Forum,
2008), which call for research on its behavioral health implications. Aa a cardinal virtue deeply
rooted in Western and Eastern civilization, reverence is a sacred positive emotion. Reverence
is defined as feeling or attitude of deep respect, love, and awe, as for something sacred,
which can bestow meaning and purpose in life within different worldviews. In this
interdisciplinary study, we investigated the predictive value of traditional religiousness and of
experiencing reverence in religious and secular (e.g., naturalistic, moralistic) contexts in
postoperative hospital length of stay among patients undergoing open-heart surgery
400+. Method: Information on demographics, faith factors, sense of reverence, mental health,
and medical comorbidities was collected from 400+ patients (age 62+12) around two weeks
before surgery via personal interview. Measures included depression, social support,
optimism, religiousness, spirituality, and reverence. Standardized medical indices, including
hospital length of stay, were retrieved from the Society of Thoracic Surgeons' national
database. Results: Hierarchical multiple regression associated secular reverence and shorter
hospitalization, after controlling for demographics, medical indices, depression, and
psychosocial protectors. Other hospital length of stay predictors included female gender, age,
medical comorbidities, low left ventricular ejection fraction, perfusion time, and coronary
bypass graft surgery. Conclusion: Secular reverence exerts a protective role in an objective
measure of cardiac health. More investigation is warrant on naturalistic view of spirituality in
patients with diverse beliefs. If the finding is replicated, health providers may integrate a
naturalistic perspective of spirituality in cardiac behavioral care.

CORRESPONDING AUTHOR: Amy L. Ai, PhD, Florida State University, Tallahassee, FL, 32306;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1823

C024 6:00 PM-7:00 PM

RISK PERCEPTION UNCERTAINTY MATTERS: GREATER UNCERTAINTY PREDICTS A LOWER

LIKELIHOOD OF BEHAVIOR CHANGE.

Caitlin Biddle, MA, MS1, Jennifer L. Hay, Ph.D.2, Marc T. Kiviniemi, BA, PhD1, Erika A. Waters,
PhD, MPH3, Heather Orom, PhD1

1
University at Buffalo, Buffalo, NY; 2Memorial Sloan Kettering Cancer Center, New York, NY;
3
Washington University in St. Louis, St. Louis, MO

Purpose: The association between perceived risk and behavior is often surprisingly weak given
the intuitive nature of the relationship. It is possible that important dimensions of risk
perception have been neglected. We propose that uncertainty about risk perception, or
peoples low confidence in the accuracy of their risk perception appraisals, may also influence
behavior change. This may be the case if uncertainty is indicative of lack knowledge about risk
factors that would guide behavior change or can be used defensively to justify non-action. The
purpose of this study was to test whether risk perception uncertainty predicts health behavior
change.

Methods: Participants were 255 MTURK workers (median education = some college; 83%
White; mean age = 34(SD = 9)) who completed an on-line cross-sectional survey. We used four
logistic regression models to test whether level of uncertainty about perceived risk for heart
disease predicted behavior change, controlling for level of perceived risk. In the first model we
regressed change in diet on perceived conditional absolute risk and associated uncertainty
about risk. In the second, we regressed change in diet on perceived conditional comparative
risk and uncertainty. Corollary models were tested for change in exercise. We controlled for
age, gender, race/ethnicity, education, income, insurance status, and healthcare provider
access.

Results: For conditional comparative and conditional absolute risk uncertainty, 29%-31% of
participants reported being very uncertain or uncertain. Three out of 4 tests supported the
hypothesis that greater uncertainty about perceived risk is associated with lower likelihood of
health behavior change. Participants who were more uncertain about their perceived
conditional absolute, and comparative risk were less likely to have changed their diet (OR =
1.7, p = .04; OR = 1.6, p = .02). Participants who were more uncertain about their perceived
conditional comparative risk were less likely to have changed their physical activity (OR = 1.5,
p = .048). However there was no association between conditional absolute risk uncertainty
and change in physical activity. Only one of the four perceived risk measures was associated
S1824 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

with health behavior change: a higher conditional absolute risk perception of heart disease
was associated with a greater likelihood of diet change (OR = 1.6, p = .03).

Discussion: Risk perception uncertainty predicts health behavior change even when
controlling for level of risk perception. In fact, it was a more reliable predictor of behavior
change than level of perceived risk itself. Dissemination of risk estimation tools that help
people understand their personal risk profiles may help people assess their risk with greater
certainty, and therefore may improve our ability to promote health behavior change.

CORRESPONDING AUTHOR: Caitlin Biddle, MA, MS, University at Buffalo, Buffalo, NY, 14220;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1825

C025 6:00 PM-7:00 PM

SELECTING MOBILE APPS FOR DISTRACTION AND RELAXATION TO HELP SMOKERS DEAL WITH
CRAVINGS

Ariana Kamberi, MBA1, Amanda Blok, PhD, MSN, PHCNS-BC2, Daniel Amante, MPH1, Rajani
Shankar Sadasivam, PhD3, Angela DiBenedetto, MA1, Thomas Houston, MD, MPH1

1
University of Massachusetts Medical School, Worcester, MA; 2University of Massachusetts
Medical School, Cambridge, MA; 3Univ of Massachusetts Medical School, Worcester, MA

Background.

Smart phone mobile applications (or Apps) available at the point-of-need may help smokers in
their quit attempts. Cravings are a major problem for smokers trying to quit. Relaxation and
distraction activities can be used to assist smokers in managing cravings. We identified
existing relaxation Apps and distraction game Apps. Using formative evaluation techniques,
we evaluated smokers preferences for these Apps.

Methods.

Our study was conducted in two phases. Phase1: We conducted a search to identify a set of
relaxation and distraction Apps (iOs and Android) for usability testing. To select a subset of
Apps for patient review, the investigators conducted a heuristic criteria evaluation. Criteria
used included: evidence based Apps that are compatible to both Apple and Android phones,
free to the consumer, frequently downloaded, and low complexity (duration of 2-7 minutes).

Phase 2: We recruited smokers to participate in a think-aloud usability testing. Testing

sessions were coded for thematic content.

Results.

Phase 1: Our search yielded 15 Apps (9 relaxations and 6 distractions). Based on the heuristic
evaluation criteria, a subset of 7 Apps was selected.
S1826 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Phase 2: Participants (N=8) preferred various relaxation Apps based on their visuals and
sounds. Example Comments were: visual (I like scenery, like beach, lakes), and sound (I
like sound of rain, ocean, beach and water running. I love sounds). Among distraction Apps,
a major theme was that users found some distraction games too challenging and frustrating
(This game gives me frustration). Further themes included that engagement was key to
game preference (Game its very engaging, it keeps my mind alert). Also some smokers
commented that game was of appropriate duration to manage a craving (This game is long
enough, it keeps me distracted and away from smoking).

Discussion/Conclusions.

Participants revealed a wide range of personal preferences for both current relaxation and
distraction. Further evaluation of these Apps for current smokers will continue in the larger
randomized trial.

Supported by NIH R01 (CA190866-01A1)

CORRESPONDING AUTHOR: Ariana Kamberi, MBA, University of Massachusetts Medical

School, Worcester, MA, 01605; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1827

C026 6:00 PM-7:00 PM

STRATEGIC SELF-PRESENTATION AND BMI: THE IMPACT OF SOCIAL AND PHYSICAL COPING
STRATEGIES IN THE ACTIVE BY CHOICE TODAY TRIAL

Allison Sweeney, PhD, Dawn K. Wilson, PhD, Nicole Zarrett, PhD

University of South Carolina, Columbia, SC

This study examines the role of social cognitive and motivational theories on increasing
motivation for physical activity (PA) in underserved adolescents. One motivational strategy
known as Strategic Self-Presentation (SSP), involves making a public commitment to an action-
plan and coping strategies. SSP is based on a theoretical framework that integrates elements
from role-playing and commitment, cognitive dissonance theory, and self-perception theory,
all of which predict that a persons public display has a strong impact on ones self-
perceptions and subsequent actions. Past studies have found that SSP facilitates greater
coping skills among chronic disease patients, and positive changes in diet and PA among
minority youth. Despite preliminary support for the efficacy of SSP, few studies have
implemented SSP as a behavioral strategy in a randomized controlled trial. The present study
extends past research by examining the effects of an SSP intervention module of a larger
school-based PA trial on: 1) changes in body mass index (BMI) from pre-to-post intervention;
and 2) the relations of different SSP-derived youth coping strategies on post-intervention BMI.
SSP was an intervention component in the Active by Choice Today (ACT) trial, a 17-week
randomized school-based trial evaluating the efficacy of motivational intervention on
increasing PA among sixth-graders. The SSP task involved completing a video-taped interview
about ones personal coping strategies for overcoming PA barriers and for enlisting peers in
PA. To further induce public commitment, students were allowed to view and revise the
video. SSP interviews (N = 301, 95.7% female, 36.2% African American) were coded by two
raters using a standardized coding scheme ( = 0.75), such that strategies used to enlist peers
in PA were separated into six categories: psychological (greater confidence), emotional (feel
good), social (be with friends), physical (increased ability), health (reduce heart-risk),
academic, and other. Sex, race, baseline PA, and baseline BMI were included as covariates. A
linear regression indicated that generating a greater number of social and physical strategies
was associated with a lower post-intervention BMI, B = -.220, SE = .10, t = 2.21, p = .028. The
results of this study indicate that social and physical coping strategies are especially impactful
for helping underserved adolescents to overcome PA barriers, and should be considered in
future interventions.

CORRESPONDING AUTHOR: Allison Sweeney, PhD, University of South Carolina, Columbia, SC,
29201; [email protected]
S1828 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C027 6:00 PM-7:00 PM

THE EFFECTIVENESS OF HEALTH COACHING: A SYSTEMATIC REVIEW

Jaime M. Hughes, PhD, MPH, MSW1, David Edelman, MD2, Jennifer McDuffie, PhD2, Hayden
Bosworth, PhD2, Eugene Oddone, MD2, Shannon S. Taylor, Ph.D.3, Cindy Swinkels, PhD4,
Andrzej Kosinski, PhD5, Jennifer M. Gierisch, PhD, MPH6

1
Center for Health Services Research in Primary Care, Durham VA Health Care System, Cary,
NC; 2Center for Health Services Research in Primary Care, Durham VA Health Care System,
Durham, NC; 3Durham VA Medical Center, Durham, NC; 4Durham VA Health Care System,
Durham, NC; 5Department of Biostatistics and Bioinformatics, Duke University, Durham, NC;
6
Duke Univeristy, Durham, NC

Health coaching is an emerging patient-centered, collaborative intervention where coaches

work with patients to identify goals and action plans to maximize personal well-being and
overall health. This systematic review evaluated health coaching effectiveness among adults
with chronic medical conditions on clinical, behavioral, and self-efficacy outcomes. We
searched multiple databases for health coaching RCTs, screened studies, and abstracted data
based on pre-determined criteria. When meta-analysis was feasible, we computed summary
effect estimates via standardized mean differences or mean differences in random-effects
models with Knapp-Hartung confidence interval correction. We grouped analyses based on
active (eg, counseling) or inactive (eg, usual care) comparator. When meta-analysis was
infeasible, we synthesized qualitatively. Compared to inactive comparators, health coaching
had a statistically significant effect on A1c (MD -0.30; 95% CI -0.50,-0.10); physical activity
(SMD 0.29; 0.15, 0.43); BMI reduction (MD -0.52; -0.91, -0.14); fat reduction (SMD -0.21; -
0.31, -0.10); and self-efficacy (SMD 0.41; 0.21, 0.62). Only change in physical activity had
sufficient studies to compare effects with active comparators. When compared to active
controls, physical activity change was not significant (SMD 0.17; -0.32, 0.67). Estimates
exhibited moderate to high statistical heterogeneity (I2>50%). In qualitative syntheses, results
were mixed or inconclusive for health coaching on functional status, smoking, and medication
adherence. We explored potential sources of heterogeneity including population
characteristics, intervention dose, intervention delivery mode, coach type, and concordance
with key elements of health coaching. None of these factors were robust predictors of
heterogeneity suggesting that high heterogeneity in pooled estimates may be driven by a
combination of intervention characteristics. These results suggest that health coaching has
the potential to produce significant changes in clinical, behavioral, and psychosocial
outcomes. Changes compared to usual care were similar to published results for other
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1829

counseling or self-management interventions. Variability in what is considered health

coaching may contribute to the overall inconsistency and heterogeneity of effects. Additional
research is warranted on the impact of health coaching and the central elements that
distinguish it from other behavioral counseling and self-management approaches.

CORRESPONDING AUTHOR: Jaime M. Hughes, PhD, MPH, MSW, Center for Health Services
Research in Primary Care, Durham VA Health Care System, Cary, NC, 27519;
[email protected]
S1830 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C028 6:00 PM-7:00 PM

ASSOCIATIONS BETWEEN MEETING COMBINED EXERCISE-ONCOLOGY GUIDELINES AND

QUALITY OF LIFE IN KIDNEY CANCER SURVIVORS

Dominick Strom, Bachelor of Science in Kinesiology1, Kerry S. Courneya, PhD2, Linda Trinh,
PhD3

1
University of Illinois at Urbana-Champaign, Urbana, IL; 2University of Alberta, Edmonton, AB,
Canada; 3University of Illinois at Urbana Champaign, Urbana, IL

Background: Kidney cancer survivors (KCS) are advised to engage in both aerobic and strength
exercise. It is unclear, however, if meeting the combined exercise guidelines confers any
additional quality of life (QoL) benefits compared to meeting only one of the exercise
guidelines.

Purpose: To estimate the prevalence of adherence to the combined exercise guidelines and
determine any associations with QoL.

Methods: All 1985 KCS diagnosed between 1996 and 2010 were identified through a Canadian
Provincial Registry and mailed a survey including the Godin Leisure Time Exercise
Questionnaire and several Functional Assessment of Cancer Therapy QoL scales. Self-reported
demographic and medical variables were collected.

Results: Completed surveys were received from 703 (43%) of the 1,654 KCS that received the
survey. Only 71 KCS (10.1%) reported meeting the combined exercise guidelines, 112 (15.9%)
were meeting the aerobic guideline only, 62 (8.8%) were meeting strength guideline only, and
458 (65.1%) were meeting neither guideline. Analyses of covariance indicated a dose
response association between the exercise guidelines and most QoL outcomes (p < 0.001). For
the primary QoL outcome of the trial outcome index-fatigue (TOI-F), meeting the combined
exercise guidelines was superior to meeting: a) neither guideline (mean difference = 11.8
points; 95% CI: 7.116.3, p < 0.001), b) the strength only guideline (mean difference = 5.4
points; 95% CI: 0.710.1, p = 0.025), and c) the aerobic only guideline (mean difference = 6.1
points 95% CI: 2.49.8, p = 0.001). The aerobic and strength exercise only guidelines were
superior to meeting neither guideline (p = 0.025 for aerobic; p = 0.001 for strength) but were
not different from each other (p = 0.80).

Conclusions: Meeting the combined exercise guidelines is associated with significantly and
meaningfully better QoL in KCS. However, only a small percentage of KCS are meeting the
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1831

combined guidelines. Efforts are needed to encourage KCS to adopt the combined exercise
guidelines to optimize QoL.

CORRESPONDING AUTHOR: Dominick Strom, Bachelor of Science in Kinesiology, University of

Illinois at Urbana-Champaign, Urbana, IL, 61801; [email protected]
S1832 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C029 6:00 PM-7:00 PM

BENEFIT FINDING AND HEALTH BEHAVIORS ONE YEAR AFTER HEMATOPOIETIC STEM CELL
TRANSPLANT

Lisa M. Gudenkauf, Ph.D., Shawna Ehlers, Ph.D., Christi A. Patten, Ph.D., Tabetha A. Brockman,
M.A., James Cerhan, M.D., Ph.D., Dennis Gastineau, M.D.

Mayo Clinic, Rochester, MN

Background/Objectives. The ability of cancer survivors to find benefits in the midst of crisis
has been increasingly recognized. Less is known about benefit finding (BF) among cancer
survivors receiving hematopoietic stem cell transplant (HSCT), especially in association with
health behaviors of prognostic significance and at long-term follow-up. Exercise in particular
has been associated with better mood, fatigue, and speed of rehabilitation post-treatment.
Better understanding of the relationship between BF and health behaviors may suggest
interventions to improve health behaviors as prognostic factors and long-term quality of life
among cancer survivors. Method. 667 adults completed a survey of benefit finding and health
behaviors one year after HSCT. Benefit finding was assessed using the Benefit Finding Scale
(Antoni et al., 2001; Tomich & Helgeson, 2004) 17-item mean score. Multivariate analyses
were used to determine associations between BF and health behaviors. Age ranged from 18
to 78, with an average age of 55. 21. This sample consisted of 60. 4% male, 80. 1% married,
and 69. 7% Caucasian survivors. Most received autologous HSCT (79. 9%). Results. BF was
higher among women than men (F(1, 665)= 8. 716, p = 0. 003) and among those in remission
at one year follow-up (F(1,663) = 5. 359, p = 0. 021). BF was not associated with age at time of
treatment, marital status, or type of transplant. In examination of health behaviors, BF was
associated with eating more servings of fruit per day (r = 0. 081, p = 0. 038) but was not
associated with servings of vegetables or with current exercise at one-year follow-up. BF was
also associated with greater perceived importance of diet (r = 0. 124, p = 0. 001), but not
importance of exercise. BF was associated with intention to change/improve health behaviors,
such that those who intended to improve both diet (F(1,645) = 38. 755, p < 0. 001) and
exercise (F(1,636) = 4. 087, p = 0. 044) reported higher BF compared to those who did not
intend to improve these health behaviors. Finally, BF was associated with greater self-efficacy
for behavior change, including patients confidence in their ability to improve nutrition (r = 0.
237, p < 0. 001) and exercise more (r = 0. 079, p = 0. 046). Associations remained significant
when controlling for gender and remission status, except that BF was no longer associated
with eating more fruits per day. Conclusions. While BF was not generally associated with
current exercise and diet behavior one year after HSCT, BF was associated with intention to
improve health behaviors as well as greater self-efficacy for health behavior change. Future
work should examine whether interventions could capitalize on patients ability to find
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1833

benefits in the cancer experience as well as their intention and self-efficacy for health
behavior change to promote improved health behaviors among HSCT populations.

CORRESPONDING AUTHOR: Lisa M. Gudenkauf, Ph.D., Mayo Clinic, Rochester, MN, 55905;
[email protected]
S1834 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C030 6:00 PM-7:00 PM

CLINIC VISIT WAIT TIME: ADVANCED CANCER PATIENT AND CAREGIVER DISCUSSIONS

Jiayun Xu, PhD, RN1, Lauren Kane, BS2, Echo L. Warner, MPH3, Maija Reblin, PhD4, Margaret F.
Clayton, PhD2, Kathleen Mooney, PhD, RN, FAAN2, Anna Beck, MD5, Lee Ellington, PhD2

1
University of Utah, Rockville, MD; 2University of Utah, Salt Lake City, UT; 3Huntsman Cancer
Institute, Salt Lake City, UT; 4Moffitt Cancer Center, Tampa, FL; 5University of Utah, Huntsman
Cancer Institute, Salt Lake City, UT

Background
In an effort to become more patient and family centered, cancer outpatient care is moving
towards increased consolidation of multiple provider consultations (e.g. oncologist,
pharmacist) into a single visit. For patients with advanced cancer and their family caregivers, it
is especially important that their time is used wisely, focusing on what matters most for them
today and for the remainder of the patients life. Our objective was to describe the length of
wait times and the content of family communication during outpatient oncology
appointments. This information can direct future structuring of appointments in cancer
centers to better meet the needs of families.
Methods
We recorded and analyzed 21 outpatient clinic visits where at least one family caregiver was
present with an advanced cancer patient. Wait times were identified, timed, and patient and
family dialogue was analyzed using an iterative, inductive coding process.
Results
Most patients were female (71.4%) with a mean age of 64.6 (SD=11.7) years; family caregivers
were mostly female (66.7%) averaging 56.3 (SD=16.9) years old and 50% were spouses. Visits
averaged 61.3 (SD=30.5) minutes, with an average of 29.3 (SD=11.2) minutes spent with a
healthcare provider and 25.2 (SD=22.7) total minutes spent waiting. On average 13.1 minutes
of wait time was uninterrupted (range=50 seconds to 81 minutes). All visits except for two
had at least 9.8 minutes of uninterrupted time. During wait times, families spent 53.5% of
time discussing their health care. For example: medication management, the current
appointment or treatment, caregiving responsibilities, and completing clinical or research
paperwork. 46.5% of wait time was spent discussing everyday life such as meals, shopping,
and grandchildren.
Conclusion
We found advanced cancer patients and their family caregivers spend approximately half of
their appointment time waiting for providers. Approximately half of their waiting time is spent
discussing health-care related topics and half on everyday life. Cancer centers may consider
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1835

restructuring appointment timing to further consolidate time with providers, or provide

guidance for discussions during wait times to provide more support to families. Examples
include interventions to facilitate family decision making, sharing values and goals, and care
management problem solving, which have been shown to be an under-addressed need for
advanced cancer patients and families.

CORRESPONDING AUTHOR: Jiayun Xu, PhD, RN, University of Utah, Rockville, MD, 20850;
[email protected]
S1836 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C031 6:00 PM-7:00 PM

DESIGNING PHYSICAL ACTIVITY INTERVENTIONS BASED ON FAMILY CAREGIVERS

PERSPECTIVES: RESULTS FROM THE RECHARGE TRIAL.

Colleen A. Cuthbert, PhD1, Kathryn King-Shier, PhD1, Nicole Culos-Reed, PhD1, Dean Ruether,
MD2, Dianne Tapp, PhD1

1
University of Calgary, Calgary, AB, Canada; 2Alberta Health Services, Calgary, AB, Canada

Background: Research activity on physical activity (PA) interventions to support cancer family
caregiver health and wellbeing has increased over the last 5 years. Recent studies on PA in
cancer caregivers have demonstrated significant effects of exercise on physical health,
psychological health, PA levels, and physical fitness. Very few studies have used qualitative
inquiry to gather the caregivers perspectives about this type of intervention.

Method: This qualitative inquiry was part of a larger mixed methods investigation of a
randomized controlled trial exercise intervention for cancer family caregivers (the RECHARGE
trial). Caregivers participated in group based resistance training classes as well as independent
aerobic sessions over 24 weeks. Our goal was to examine cancer family caregivers
perspectives about their participation in the exercise program. Semi-structured face-to-face
interviews were completed and analyzed using the method of Interpretive Description.

Results: 20 participants were interviewed. The following themes characterized the caregivers
experience of participating in the exercise program. The context of caregiving is an important
component of caregivers ability to regularly exercise, therefore, timing of when caregivers
can participate in PA is highly variable. Caregivers were motivated to participate for many
reasons, including wanting to feel better mentally and physically, and wanting to engage in a
healthier lifestyle with their entire family. Caregivers preferred exercising with other
caregivers in a group based format. The caregivers preferred instructor led (versus
independent) exercise as it meant one less thing to worry about or plan. The resistance
training format was important as it allowed caregivers to focus on one thing and was a
distraction from their worries. Having fun and doing something positive for their own health
was an important benefit of regular exercise.

Conclusions: Our findings highlight important aspects of the design of an exercise program
based on caregivers perspectives. Future research on PA interventions in caregivers should
include qualitative inquiry to better understand how to design interventions based on
caregiver preferences, what outcomes to measure, how to reach more caregivers, and how to
sustain PA programs in the changing context of the caregiving situation.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1837

CORRESPONDING AUTHOR: Colleen A. Cuthbert, PhD, University of Calgary, Calgary, AB, T3L
3C7; [email protected]
S1838 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C032 6:00 PM-7:00 PM

EFFECTS OF A TXT INTERVENTIONS ON PSYCHOSOCIAL OUTCOMES FOR BREAST CANCER

PATIENTS RECEIVING CHEMOTHERAPY: A RCT STUDY

Kuang-Yi Wen, PhD1, Suzanne M. Miller, PhD2, Rita Smith, BS3, Jennifer WINN, MD4, Lori
Goldstein, MD3

1
Fox Chase Cancer Center, Cheltenham, PA; 2Fox Chase Cancer Center/Temple University
Health System, Philadelphia, PA; 3Fox Chase Cancer Center, Philadelphia, PA; 4Fox Chase
Cancer Center, Lansdale, PA

As a core component of cancer care, the use of chemotherapy in the ambulatory care setting
is likely to increase considerably with projected increases in the incidence of breast cancer.
Chemotherapy may improve overall survival; it is also often associated with substantial
treatment-related toxic effects that negatively affect HRQOL. Patients are responsible for
making complex decisions about what to do or when to contact the providers in the event of
new, escalating or unexpected side-effects. To fill this void, we are in the process of evaluating
an innovative and automated text messaging program, designed to facilitate breast cancer
patients self-management in the home setting during adjuvant chemotherapy. The theory-
based mChemoCoping text messaging-based intervention was developed through an iterative
formative evaluation process with targeted population. Using a RCT design, 34 women were
assigned to a control group with a patient education booklet and 31 were assigned to the
mChemoCoping intervention group in the ongoing study. During the 4-month intervention,
participants in the mChemoCoping condition receive one chemotherapy coping tip text
message in the morning and one topic inquiry tip (to generate user-initiated text) in the
afternoon. All participants complete a baseline survey and 4 monthly follow-up survey. For
the 1-month follow-up complete data, Intervention feasibility and acceptability scores are
above average and satisfaction feedback from intervention participants is highly positive
through qualitative interviews. A total of 1649 text messages are sent by intervention
participants and the most prevalent topics requested by the participants are emotional
coping, fatigue, and body changes. The intervention had positive effects on self-efficacy (F =
9.9, p< 0.05) but not symptom distress (F=0.9, p=0.35) at 1-month follow-up. This innovative
mHealth intervention has potential for improving womens coping skills to manage
chemotherapy in the ambulatory care setting.

CORRESPONDING AUTHOR: Kuang-Yi Wen, PhD, Fox Chase Cancer Center, Cheltenham, PA,
19012; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1839

C033 6:00 PM-7:00 PM

EXPERIENCES OF FINANCIAL SACRIFICE AMONG CANCER SURVIVORS UNDER AGE 65

Matthew Banegas, Ph.D., M.P.H., M.S.1, Alexandra Varga, MPH2, Jennifer Schneider, MPH2,
John Dickerson, PhD2, Erin Kent, PhD3, Janet DeMoor, PhD, MPH4, Katherine Virgo, PhD5, Gery
Guy, PhD6, Zhiyuan Zheng, PhD7, Stephanie Nutt, MA, MPA8, Lloyce Pace, MPH9, Lisa
Waiwaiole, MS2, Donatus Ekwueme, PhD10

1
Kaiser Permanente, Portland, OR; 2Center for Health Research, Kaiser Permanente
Northwest, Portland, OR; 3National Cancer Institute, Rockville, MD; 4National Cancer Institute,
National Institutes of Health, Rockville, MD; 5Rollins School of Public Health, Emory University,
Atlanta, GA; 6Centers for Disease Control and Prevention (CDC), Atlanta, FL; 7University of
Maryland School of Pharmacy, Baltimore, MD; 8University of Texas Austin, Dell Medical
School, Austin, TX; 9Global Health Council, Alexandria, VA; 10Centers for Disease Control and
Prevention (CDC), Atlanta, GA

Background: Many cancer survivors experience financial hardship, particularly individuals

under age 65. This study examined qualitative responses of cancer survivors under age 65 to a
survey item requesting description of financial sacrifices associated with cancer.

Methods: We examined data from adults aged 18-64 years with a history of cancer who
responded to a 2012 LIVESTRONG Survey on experiences with cancer. The online survey
contained a section dedicated to the effects of cancer on finances, including borrowing
money, filing for bankruptcy, and other financial sacrifices. Among 4484 survey participants
aged 18-64 years, approximately 37% (n=1656) responded to the survey item Have you or
your family had to make any other kinds of financial sacrifices because of your cancer, its
treatment, or the lasting effects of that treatment? and completed the open-ended question
describing these financial sacrifices. For the qualitative assessment, a codebook was
developed based on initial review of the open-ended responses. Two trained staff classified
each individual response, using codes that were applied as content dictated. We assessed the
frequency of financial sacrifice and examined differences between two reported levels of
medical debt ($10,000 vs. 2 tests. The ns indicate the frequencies of coded financial
sacrifices.
S1840 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Results: Among the 1656 respondents, a majority were women (71%), non-Hispanic white
(87%), between the ages of 40-54 (50%), and married (68%). Approximately 33% of
respondents reported medical debt $10,000. The most commonly described financial
sacrifices included: cutbacks/significant budgeting (n=796); health care insurance/costs
(n=406); career impacts (n=316); asset depletion/reduction (n=273); inability to pay bills
(n=263); family/relationship burden (n=220); housing challenges (n=216); and increased
debt/borrowing (n=184). The type of financial sacrifices differed by level of medical debt. A
significantly greater proportion of respondents with debt $10,000 experienced inability to
pay for bills and services, housing challenges, family/relationship burden, and increased
debt/borrowing compared to those with debt

Conclusion: Findings provide insight into financial sacrifices made by cancer survivors under
age 65 and underscore the sociocultural, individual, interpersonal impact on those who
experience financial hardship.

CORRESPONDING AUTHOR: Matthew Banegas, Ph.D., M.P.H., M.S., Kaiser Permanente,

Portland, OR, 97227; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1841

C034 6:00 PM-7:00 PM

EXPLAINING THE EFFECTS OF MINDFULNESS BASED STRESS REDUCTION ON FATIGUE IN

CANCER SURVIVORS

R. Brian Giesler, PhD1, Shelby Eaton, BA1, Natalie Nusinow, BA1, Shelley A. Johns, PsyD, ABPP2

1
Butler University, Indianapoils, IN; 2Indiana University School of Medicine, Indianapolis, IN

Cancer-related fatigue (CRF) is a significant problem for cancer survivors. In a recent,

randomized, controlled pilot study, Mindfulness-Based Stress Reduction (MBSR) was
compared to an
education/support control group to assess MBSRs capacity to provide relief from fatigue.
Breast (n=60) and colorectal (n=11) cancer survivors (stage 0-III) with persistent, moderately
severe CRF were randomized to MBSR or ES. Outcomes, which included vitality, were assessed
at baseline (T1), post-intervention (T2) and again 6 months later at follow-up (T3). Complete
results are reported elsewhere (Johns, et al., 2016), but by T2, MBSR participants reported
significantly greater vitality (p=0.003) compared to ES participants. To examine the possible
mechanisms behind this difference, several potential mediating variables were identified on
theoretical grounds, including self-regulatory ability, depressive symptoms, pain, attentional
function, and the variables assessed by the Five Facet Mindfulness Questionnaire (FFMQ).
Each potential mediator variable was tested using boot-strap procedures: indirect effects and
the 95% confidence interval were computed for each of 5,000 bootstrapped samples. These
procedures identified one significant mediator: attentional function. For attentional function,
assessed by the Attentional Lapse Subscale of the Attentional Function Index (AFI), the
indirect effect was 2.2602 and the 95% confidence interval ranged from .1147 to 6.9203.
These findings suggest that relative to ES, MBSR exerted its effects, in part, by improving the
attentional function of participants.

CORRESPONDING AUTHOR: R. Brian Giesler, PhD, Butler University, Indianapoils, IN, 46208;
[email protected]
S1842 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C035 6:00 PM-7:00 PM

FACTORS ASSOCIATED WITH TELEVISION VIEWING TIME AMONG BREAST, PROSTATE, AND
COLORECTAL CANCER SURVIVORS

Bennett R. McDonald, MPH1, Corinne Leach, MS PhD MPH2, Elizabeth Fallon, Ph.D.1

1
American Cancer Society, Atlanta, GA; 2American Cancer Society's Behavior Research Center,
Atlanta, GA

Introduction: Sedentary behavior is a distinct risk factor for poor physical and mental health
outcomes independent of moderate to vigorous physical activity (MVPA). Television (TV)
viewing time has been linked to increased risk of cancer mortality. Because cancer survivors
are at higher risk for chronic disease morbidity and mortality and are more likely to be
sedentary compared to individuals without a cancer history, there is need to examine
correlates of TV viewing time in cancer survivors.

Methods: Data from 852 breast, prostate, and colorectal cancer survivors within their first
year post-treatment were analyzed. Hierarchical ordinal regression was used to assess
demographic (age, gender/cancer type, race, education, marital status), medical (months
since treatment completion, treatments received, cancer stage, number of comorbidities,
physical symptom prevalence), behavioral (smoking status, BMI, MVPA), and psychosocial
predictors (depressive symptoms, social support, fear of recurrence) of elevated TV viewing
time per day over the past month (< 2, 2-4, or >4 hours/day).

Results: 26.2% of participants reported viewing less than 2 hours, 41.2% reported 2-4 hours,
and 32.6% reported more than 4 hours of TV viewing time per day. The odds of reporting
elevated TV viewing time were significantly higher among non-whites compared to whites
(OR=1.48; 95% CI=1.01, 2.18) and those with no college education compared to at least some
college (OR=0.69; 95% CI=0.52, 0.92). Obese survivors compared to those of normal weight
(OR=1.67; 95% CI=1.20, 2.33), current smokers compared to never smokers (OR=2.05; 95% CI
= 1.27, 3.31), those not meeting MVPA guidelines compared to survivors who did (OR=1.66;
95% CI = 1.26, 2.18), and those with more comorbid conditions (OR=2.51; 95% CI = 1.61, 3.90)
also reported higher odds of elevated TV viewing time.

Discussion: Subgroups of cancer survivors who have recently completed treatment may be at
higher risk of deleterious physical and mental health outcomes due to elevated TV viewing
time. Future research should validate these findings with objective measures. Modifiable risk
factors, including smoking status, obesity, and MVPA should be considered when developing
interventions aimed at reducing TV viewing among cancer survivors. Further, interventions
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1843

should consider targeting survivors with higher odds of TV viewing, including racial/ethnic
minorities, non-college educated individuals, and/or those with multiple comorbid conditions.

CORRESPONDING AUTHOR: Bennett R. McDonald, MPH, American Cancer Society, Atlanta,

GA, 30312; [email protected]
S1844 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C036 6:00 PM-7:00 PM

GOAL DISRUPTION IN EARLY-STAGE BREAST CANCER SURVIVORS: AN EXAMINATION OF

PREDICTORS AND CORRELATES

Marcie D. Haydon, M.A.1, Annette L. Stanton, Ph.D.2, Patricia A. Ganz, M.D.1, Julienne Bower,
PhD3

1
University of California, Los Angeles, Los Angeles, CA; 2UCLA, Los Angeles, CA; 3UCLA
Departments of Psychology and Psychiatry/Biobehavioral Sciences; Cousins Center for
Psychoneuroimmunology, Los Angeles, CA

Background: Disruption in meaningful life goals is common in cancer patients and has been
shown to predict long-term decrements in well-being. However, to date, few studies have
examined the factors that contribute to goal disruption, particularly in cancer survivors. To
address this gap, we examined the relationship between demographic variables, cancer- and
treatment-related factors, behavioral symptoms (depressive symptoms, fatigue, pain, and
cognitive complaints), and cancer-related goal disruption in early-stage breast cancer
survivors 1 year after the completion of primary treatment.
Methods: A cohort of early-stage breast cancer survivors (n = 190; Mage = 51.75) completed
assessments following primary treatment with surgery, radiation, and/or chemotherapy, and
again 6-months and 1-year later. We focused on the 1-year post-treatment assessment (n =
175) when participants were asked if they had experienced a cancer-related goal disruption.
Behavioral symptoms included depression (Beck Depression Inventory-II), fatigue
(Multidimensional Fatigue Symptom Inventory), pain (RAND SF-36 Pain Subscale), and
cognitive complaints (Patients Assessment of Own Functioning Inventory). These symptoms
were computed as a composite measure and also analyzed individually.
Results: Approximately 27% of women reported a goal disruption due to their cancer
diagnosis. Cross-sectional analyses indicated that both receipt of chemotherapy and
behavioral symptoms were significantly associated with goal disruption; no other
demographic or disease-related factors were associated with goal disruption. In multivariate
analyses, women were more likely to report a goal disruption if they had received
chemotherapy (OR = 3.32, p = .003) and reported greater overall behavioral symptoms (OR =
1.26, p = .004); 2(2) = 26.03, p < .001. When analyzed separately, each behavioral symptom
significantly predicted the likelihood of experiencing a goal disruption (ps .01), holding
chemotherapy constant.
Conclusions: Many women reported a cancer-related goal disruption, even a year after
treatment completion. Disrupted goals were reported as highly meaningful and spanned
various life domains. Chemotherapy and behavioral symptoms were unique predictors,
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1845

suggesting that the impacts of chemotherapy extend beyond its influence on persistent
behavioral symptoms. This is one of the first studies, to date, to examine goal disruption in
early breast cancer survivorship.

CORRESPONDING AUTHOR: Marcie D. Haydon, M.A., University of California, Los Angeles, Los
Angeles, CA, 90095; [email protected]
S1846 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C037 6:00 PM-7:00 PM

HEALTH BEHAVIOR ENGAGEMENT, OBESITY, AND SYMPTOM MANAGEMENT IN LATINA

BREAST CANCER SURVIVORS: A QUALITATIVE STUDY

Gloria Juarez, RN, PhD1, Dara H. Sorkin, PhD2, Mienah Sharif, PhD3, Dana B. Mukamel, PhD4,
Yunan Chen, PHd2, Alfred Kobsa, PhD4, Alice Police, MD4, Sadeeka Al-Majid, PhD5, Kelly
Biegler, PhD4

1
UC Irvine, SanPedro, CA; 2University of California Irvine, Irvine, CA; 3University of California
Irvine, irvine, CA; 4UC Irvine, Irvine, CA; 5California State University, Fullerton, Fullerton, CA

Background: Breast cancer survivors are at risk for weight gain, obesity, and diabetes. Relative
to other ethnic groups, Latinas are at even greater risk of diabetes and obesity, and are more
likely to experience other breast cancer-related symptoms, including fatigue and negative
mood. Symptom burden is further compounded in this population by lower socio-economic
status and limited access to resources (e.g., car transportation, specialty medical care).
Importantly, engaging in health behavior change to achieve weight loss can significantly
reduce the number symptoms and their severity among breast cancer survivors. Thus,
understanding the complex interplay of obesity, health-behavior engagement, and symptom
experience among a group of high-risk, under-resourced Latina breast cancer survivors is a
critical next step for addressing this disparity.

Purpose: To obtain an in-depth understanding of Latina breast cancer survivors perception of

factors that contribute to their own health, and the potential connection between their
engagement in particular health behaviors and whether these behaviors alleviated symptoms.

Methods: We conducted 17 focus groups in both English and Spanish with 70 Latina women
who had undergone treatment for breast cancer to further understand the role of Hispanic
culture in the breast cancer survivor experience, cultural-specific motivators and barriers, as
well as symptom burden, to engaging in a healthy lifestyle during survivorship.

Results: A common theme that emerged from the focus groups was that a healthy lifestyle
helped alleviate some of their symptoms. Some quotes to provide direct examples include:
For me, the more that I exercise the more active I feel. I also want to be healthier and to feel
better. When I do not exercise I feel more fatigued, more tired. The body is not the same after
you have been through a lot. In response to asking what made her start walking another
patient said: Well, because I was feeling like I was not feeling good with myself. I was feeling
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1847

bloated and the doctor said that it would probably help.So I did and I just felt better, you
know. I know, so I started walking a lot and like 30-40 minutes a day.

Conclusions: Findings suggest under-resourced, high-risk Latina breast cancer survivors

recognize a potentially important link between healthy behavior and symptom management.
This awareness has to potential be leveraged to create meaningful lifestyle modification
programs.

CORRESPONDING AUTHOR: Gloria Juarez, RN, PhD, UC Irvine, SanPedro, CA, 90732;
[email protected]
S1848 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C038 6:00 PM-7:00 PM

HEART DISEASE VERSUS CANCER: UNDERSTANDING PERCEPTIONS OF POPULATION

PREVALENCE AND PERSONAL RISK

Jennifer K. Scheideler, BA1, Jennifer M. Taber, PhD.2, Rebecca A. Ferrer, PhD. 1, Emily Grenen,
MPH3, William Klein, PhD.4

1
National Cancer Institute, Bethesda, MD; 2Kent State University, Kent, OH; 3ICF, Washington,
DC; 4National Cancer Institute, Rockville, MD

Americans are more likely to be diagnosed with and die from heart disease than cancer.
However, many believe that cancer is the more common disease, and may regard their
personal risk of cancer to be higher than their heart disease risk. Using the nationally
representative 2013 iteration of the National Cancer Institutes Health Information National
Trends Survey (HINTS 4 Cycle 3), we examined the prevalence of both of these relative risk
beliefs (i.e., relative population prevalence estimates and disease-relative personal risk
perceptions). Additionally, we examined sociodemographic (i.e. race, gender, age, education)
and psychosocial (i.e., perceptions of ambiguity and fatalism) factors, as well as health
behaviors (i.e., physical activity, colorectal cancer screening, fruit/vegetable consumption,
soda consumption, sunscreen use, and mammography) associated with these beliefs.
Consistent with past work, 42.81% of participants rated cancer as more common than heart
disease. Moreover, 78.45% rated their own cancer risk as equal to or exceeding their heart
disease risk. These beliefs were only modestly correlated (r = 0.10, p = .013), and correlates of
the two beliefs differed. Specifically, beliefs about population prevalence were associated
with fatalism (i.e., thinking about cancer leads to thoughts of death; B =.16, p < .001) and
perceived ambiguity of cancer messages (i.e., perceive cancer recommendations as
ambiguous; B =.022, p < .001), whereas perceived personal risk was not. Less educated (r = -
.14, p < .001) and non-White individuals (Mnon-White = 2.50, Mwhite = 2.22, p < .001) rated
cancer as more common. Individuals who thought their personal risk of cancer was greater
consumed more fruit (B = .13, p = .002) and vegetables (B = .09, p = .04), whereas those who
believed cancer to be more common consumed more soda (B = .12, p = .039) and engaged in
less physical activity (B = .20, p = .013). Overall, accuracy in beliefs about cancer and heart
disease relative risk and prevalence is low. Future longitudinal research should explore the
antecedents and consequences of these beliefs and explore why some people think their risk
of cancer is greater than their risk of heart disease.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1849

CORRESPONDING AUTHOR: Jennifer K. Scheideler, BA, National Cancer Institute, Bethesda,

MD, 20892; [email protected]
S1850 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C039 6:00 PM-7:00 PM

INCREASING PHYSICAL ACTIVITY AND CHANGES IN COGNITIVE FUNCTIONING IN BREAST

CANCER SURVIVORS: A RANDOMIZED CONTROLLED STUDY

Sheri J. Hartman, PhD1, Sandahl H. Nelson, MS2, Emily Myers, MPH2, Loki Natarajan, PhD1,
Dorothy Sears, PhD2, Barton Palmer, PhD2, Barbara Parker, MD2, Ruth Patterson, PhD2

1
Department of Family Medicine and Public Health, UC San Diego, La Jolla, CA, USA, La Jolla,
CA; 2University of San Diego, California, La Jolla, CA

Introduction: Increasing physical activity can improve cognition in healthy and cognitively
impaired adults. However, the benefits of physical activity for cancer patients cognition have
not been previously reported.

Methods: Sedentary breast cancer survivors with cognitive difficulties were randomized to an
Exercise intervention (n=43) or Control arm (n=44). At baseline and 12-weeks participants
wore an accelerometer to measure activity and completed the NIH Cognitive Toolbox
providing 9 scores (composite fluid and crystalized ability scores and scores for seven
component tests). Intervention effects for change in each score over 12 weeks were tested
using linear mixed effects regression models.

Results: Participants (n=87) were a mean of 57 years (SD=10.4), diagnosed with breast cancer
2.5 years ago (SD=1.3); 53% had chemotherapy and 70% were taking hormone therapy. At 12
weeks, 60% of the Exercise arm met the study goal of 150 min/week of moderate to vigorous
physical activity (MVPA). Relative to Control, the Exercise arm had greater increases in MVPA
(mean min/day increase 14.2 vs. -0.7, b=7.24, p < .001) and total activity (mean min/day
increase 27.4 vs. 4.9, b=10.05, p=.02).

Of the NIH Toolboxs 9 scores, only the Oral Symbol Digit (a measure of processing speed) had
significant improvement in the Exercise vs. Control (b=2.01;p < .05). The Exercise arm
increased an average of 7 points while the Control arm increased an average of 3 points.
There appeared to be an interaction between time since diagnosis and Oral Symbol Digit
scores (p=.07). Participants 2 years from diagnosis had significantly more improvement in the
Exercise vs. Control (b=4.00, p < .01); there was no significant improvement in patients who
were >2 years post-diagnosis (b=-1.19, p=.40). No modification effect was found for BMI,
years since diagnosis, chemotherapy or hormone therapy.

Within the exercise arm there was a significant dose response with greater increases in MVPA
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1851

associated with greater increases in Oral Symbol Digit (b=2.0, p=.02). Also, greater total
activity was associated with greater increases in the Fluid Composite score (b=0.03, p=.04)
and Picture Sequence (memory; b=0.04, p=.04).

Conclusion. Among breast cancer survivors, the exercise intervention only significantly
improved processing speed. Moderator and dose-response analyses suggest that future trials
should aim for greater increases in physical activity and target women closer to diagnosis to
impact cognition.

CORRESPONDING AUTHOR: Sheri J. Hartman, PhD, Department of Family Medicine and Public
Health, UC San Diego, La Jolla, CA, USA, La Jolla, CA, 92093; [email protected]
S1852 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C040 6:00 PM-7:00 PM

INFLAMMATORY BIOMARKERS ARE ASSOCIATED WITH CHANGE IN SKELETAL MUSCLE MASS

AND STRENGTH IN EXERCISING PROSTATE CANCER PATIENTS

Calvin L. Cole, PhD1, Ian Kleckner, PhD2, Po-Ju Lin, PhD, MPH, RD3, Charles Kamen, PhD, MPH2,
Luke J. Peppone, PhD, MPH4, Michelle Janelsins, PhD, MPH2, Charles E. Heckler, Ph.D., M.S. 2,
Karen Mustian, PHD, MPH, MS2, Sarah Kerns, PhD MPH2, Matthew Asare, PhD, MBA, CHES1

1
University of Rochester, Rochester, NY; 2University of Rochester Medical Center, Rochester,
NY; 3University of Rochester Medical Center, Department of Surgery-Cancer Control,
Rochester, NY; 4University of Rochester, Wilmot Cancer Institute, Rochester, NY

BACKGROUND: Decreases in skeletal muscle mass (SMM) and muscular strength are
debilitating side effects of cancer and its treatments and are independently predictive of
mortality. Recent studies suggest a negative association of inflammatory cytokines with SMM
and strength. Exercise has been shown to positively regulate pathways that govern SMM and
strength. This research study attempted to identify any associations among declines in SMM
and strength and inflammatory biomarkers and to determine if exercise has an effect on these
associations. PURPOSE: The purpose of this study was to assess the relationship among
inflammatory biomarkers, indices of SMM and strength, and engagement in exercise among
prostate cancer patients. METHODS: A secondary data analysis was performed on a two-arm
randomized clinical trial examining the influence of a 6-week home-based aerobic and
resistance exercise program. Participants included 57 sedentary prostate cancer patients
aged 67 8.05 years, who were undergoing cancer treatment or had been previously treated
with hormone therapy and/or radiation therapy. Each participant was randomly assigned to
usual care or usual care with exercise. SMM and muscular strength were measured along
with biomarkers of inflammation (IL-6, IL-1b, IFN, IL-10, IL-8, and TNFR1 proteins) at baseline
and post-intervention. Pearson correlations and linear regressions were used to analyze the
relationship among these variables. RESULTS: In the exercise group, changes in IFN and IL1b
showed a moderate association with changes in SMM (IFN r =.442, p = .016, IL-1b, r = .543, p
= .002) where increases in IFN (p =.016) and IL-1b (p = .002) explained 20% and 30% of the
change in SMM, respectively. A trend towards a moderate relationship between IL-10 and leg
muscle strength (r = -.348, p = .081) was found where decreases in IL10 explained 12% of the
change in leg strength. No significant associations among any of these variables were found in
the usual care group. CONCLUSION: Changes in IFN, IL-1b and IL-10 may be associated with
changes in SMM and strength in exercising prostate cancer patients.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1853

CORRESPONDING AUTHOR: Calvin L. Cole, PhD, University of Rochester, Rochester, NY, 14642;
[email protected]
S1854 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C041 6:00 PM-7:00 PM

INFORMATION AND SUPPORT NEEDS AMONG YOUNG ADULT RELATIVES OF BRCA 1/2
MUTATION CARRIERS

Chalanda N. Evans, BS1, Rebekah J. Hamilton, RN PhD CNL FAAN2, Kenneth P. Tercyak, PhD3,
Beth N. Peshkin, MS, CGC4, Sue Friedman, DVM5, Claudine Isaacs, MD4, Suzanne O'Neill, PhD4

1
Georgetown University/ Lombardi Comprehensive Cancer Center, Washington, DC; 2Rush
University College of Nursing, Chiacgo, IL; 3Georgetown University Medical Center, Lombardi
Comprehensive Cancer Center, Washington, DC; 4Georgetown University, Washington, DC;
5
Facing Our Risk of Cancer Empowered (FORCE), Tampa, FL

Background: Young women from hereditary breast and ovarian cancer (HBOC) families face a
unique set of challenges in managing their HBOC risk where obtaining essential information to
inform decision making is key. Previous work suggests that the need for specific health
information related to risk also comes at a time of heightened distress and greater
individuation from family. Our study sought out to assess the information needs in this
population and correlates of greater need.

Method: One hundred young women (M age= 25 years) who were first or second degree
relatives of BRCA1/2 carriers participated. Participants ages 18-30 were recruited through
research registries and a national non-profit organization for a telephone interview. Using
multivariable linear regression, we examined the association between clinical (age, tested
status, number of cancer-affected relatives and age of youngest affected relative) and
psychosocial (familial disruption, peer support, cancer-related distress and perceived cancer
risk) variables and our outcome of information need.

Results: One third of the sample endorsed wanting HBOC information that they did not know
how to obtain and 63% endorsed unmet HBOC information needs. Multivariate regression
model results revealed that greater information need was associated with being untested
(=.30, p=.004) and greater familial disruption (=.41, p=.002) and cancer-related distress
(=.21, p=.04).
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1855

Conclusion: Our results suggest that young adult women from HBOC families have unmet
cancer genetic information needs. Untested women, as well as those with higher levels of
distress and poorer familial connections, are specifically in need of support. Social support can
mitigate distress and supplement familial relationships. Our results suggest that young adult
relatives could benefit from a structured peer-led intervention that provides needed support
and targets psychologic distress. Trained peers can offer structured social support guidance
and assist with objectively weighing the pros/cons of personal HBOC risk management choices
and lifecycle issues relevant to women in their 20's and 30's.

CORRESPONDING AUTHOR: Chalanda N. Evans, BS, Georgetown University/ Lombardi

Comprehensive Cancer Center, Washington, DC, 20007; [email protected]
S1856 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C042 6:00 PM-7:00 PM

INTEREST IN WEIGHT LOSS AMONG FEMALE CANCER SURVIVORS WITH OVERWEIGHT OR

OBESITY

Christine May, PhD1, Molly E. Waring, PhD2, Martinus Evans, B.S., M.S.2, Sherry Pagoto, PhD2

1
University of Massachusetts Medical School, Fiskdale, MA; 2University of Massachusetts
Medical School, Worcester, MA

Obesity increases risk of cancer recurrence and decreases survival, yet many female cancer
survivors are overweight or obese. We surveyed adult female cancer survivors with
overweight or obesity via the internet on their awareness of the link between obesity and
cancer recurrence and preferences in various intervention modalities (in-person, telephone,
Facebook, Twitter, private website) and interest in 12 components of a Facebook-delivered
intervention. Responses dichotomized as not at all/a little bit, neutral vs quite a bit/extremely
interested. Survivors also reported preferred group composition (any cancer type, same
cancer type, not focused on survivors, no preference). An open-ended question assessed
interest in other weight loss group features. We conducted a content analysis of responses
with two coders (kappa=0.84). Survivors (N=96) were on average 54.3 (SD: 9.6) years old, 89%
white, with average BMI of 32.7 (SD: 5.8) kg/m2; 87% tried to lose weight in the last year. Just
over half (54%) reported that their weight at least somewhat affects their likelihood of cancer
recurrence. Most (81%) reported being quite a bit/extremely interested in a program
delivered via Facebook or private website (71%), while fewer were interested in telephone
(37%) or in person delivery (34%). About three-quarters were interested in having a weight
loss counselor lead the group (78%), healthy recipe links (73%), and free exercise videos
(72%); interest in other program components ranged from 35%-69%. Survivors suggested
features including information on effects of cancer treatment on weight (25%), a calorie
tracker (21%), and exercise modifications for cancer survivors (17%). Almost half (43%)
preferred a group with cancer survivors (19% same cancer type, 3% not focused on survivors,
33% no preference). Nearly half of survivors were not aware of the link between obesity and
cancer. Survivors were interested in a weight loss program that addresses their specific weight
loss needs, and the majority liked the idea of a Facebook-delivered program. Social media
delivered weight loss interventions for female cancer survivors should incorporate
information on the obesity-cancer link and features that address specific needs of this
population.

CORRESPONDING AUTHOR: Christine May, PhD, University of Massachusetts Medical School,

Fiskdale, MA, 01518; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1857

C043 6:00 PM-7:00 PM

IS ANTICIPATED REGRET A SEPARATE AND UNIQUE THEORETICAL CONSTRUCT FROM

MOTIVATION IN THE CONTEXT OF HPV VACCINATION?

Catherine Rochefort, MS1, Austin S. Baldwin, PhD1, Deanna Denman, MA1, Amy McQueen,
PhD2, Jasmin A. Tiro, PhD3

1
Southern Methodist University, Dallas, TX; 2Washington University, School of Medicine, St.
Louis, MO; 3University of Texas Southwestern Medical Center, Dallas, TX

Objective: Anticipated regret, a negative emotion stemming from the ability to reason
counterfactually, has been shown to be a key predictor of various health behaviors (Brewer et
al., 2016), but no studies have demonstrated that it is a unique construct from motivation.
Anticipated regret is conceptually similar to introjected motivation from self-determination
theory (i.e., motivation to avoid guilt and shame). In the context of HPV vaccination, we
hypothesized that anticipated regret about not vaccinating a child would load more strongly
on a factor with introjected motivation than autonomous motivation (i.e., motivation to
engage in behaviors that reflect personal values). Methods: We collected data via telephone
questionnaires from 223 parents of unvaccinated adolescents attending safety-net clinics. We
assessed parents anticipated regret, introjected and autonomous motivation, and intentions
to vaccinate. Results: Using confirmatory factor analyses, a four-factor model including
autonomous motivation (4 items), introjected motivation (2 items), external motivation (2
items), and 1 anticipated regret item provided a good fit to the data (TLI= .971, CFI= .983,
RMSEA= .053, 90% CI: .013, .085). The anticipated regret item did not load significantly onto
introjected motivation (.26) or autonomous motivation (.49). In a univariate regression model,
anticipated regret predicted intentions to vaccinate (= .243, p < .001). A multivariable model
containing introjected motivation and anticipated regret showed both to be significant
predictors of intentions, but when tested simultaneously with autonomous motivation,
anticipated regret was no longer a significant predictor of intentions. Conclusion: Anticipated
regret is a distinct construct from motivation, and loads more strongly on a factor with
autonomous motivation than introjected motivation for HPV vaccination. In addition,
anticipated regret did not explain any variance in HPV vaccination intentions that is not
shared with autonomous motivation. Implications for these findings will be discussed.

CORRESPONDING AUTHOR: Catherine Rochefort, MS, Southern Methodist University, Dallas,

TX, 75206; [email protected]
S1858 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C044 6:00 PM-7:00 PM

KNOWLEDGE, ATTITUDES, AND BELIEFS ABOUT HPV, GARDASIL, AND ACCESS DIFFER AMONG
LATINOS IN A SMALL RURAL NEW DESTINATION SETTING

Jason Daniel-Ulloa, PhD, MPH1, Laura Seegmiller, MPH1, Barbara Baquero, PhD, MPH1,
Gregory Steckel, MDiv2, Marln Mendoza, HS1, Ana Duarte, HS1, Christine Harb, BA1

1
University of Iowa College of Public Health, Iowa City, IA; 2St. Joseph Catholic Church, West
Liberty, IA, West Liberty, IA

The target vaccine coverage rate for HPV set by Healthy People 2020 is 80%. However, few if
any communities meet this goal, yet. Latino populations in rural new destination settlements
(areas in which Latino communities are growing rapidly) present unique challenges for local
providers and public health staff (e.g., language capacity, cultural competence). To provide
better information on the knowledge, attitudes and beliefs about HPV, Gardasil, and access to
vaccines among Latinos in a new destination area in a rural Midwestern state, we conducted
72 intercept interviews with representatives of Latino households. Surveys were conducted
at a local Catholic Church, a small Latino owned grocery store, a large grocery store, two
Latino Festivals, a bakery and a restaurant. The survey included items assessing age, gender
and vaccine status for all household members 9-26 years old, knowledge of HPV and Gardasil,
best places to get vaccinated, trusted sources for health education and the best way to
contact individuals for reminders. The sample was mostly female (67%), born in Mexico (59%)
or the US (24%). Mean age was 39 (13.1), and average number of years lived in the US was 24
(13.2). A majority of households (57%) included at least one vaccine eligible female and 61%
included an eligible male. A higher percentage of females (80%) reported knowing about HPV
than males (57%). Of those that were familiar with HPV, 89% of females and 82% of males
were familiar with Gardasil. Access to the vaccine was rated highest at the mobile clinic
overall. The nearest pharmacy, (one town over) received significantly higher access ratings (p
< .05) by individuals that were born outside the US but immigrated as minors ( < 18). Churches
and beauty salons received highest ratings as places to provide education. Beauty salons were
rated higher by US born than minor (p < .001) and adult (p < .003) immigrants. Participants
reported similar ratings for calls, texts, emails and postcards. Reaching and educating Latino
populations in rural areas will be important in achieving vaccine coverage objectives in the
Midwest. However, in predominantly white states, local efforts may focus more on larger non-
Latino communities, further complicated by a lack of information about how these new
communities access information about cancer and cancer prevention. This exploratory study
can help inform useful strategies and tactics to reach these Latinos for HPV vaccine education
and initiation.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1859

CORRESPONDING AUTHOR: Jason Daniel-Ulloa, PhD, MPH, University of Iowa College of Public
Health, Iowa City, IA, 52246; [email protected]
S1860 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C045 6:00 PM-7:00 PM

LESSONS LEARNED: ASSESSING GRIP STRENGTH AND MOBILITY FUNCTIONS IN A BREAST

CANCER SUPPORT GROUP SETTING

Maria Swartz, PhD, MPH, RD, LD1, Rachel R. Deer, PhD2, Eloisa Martinez, BS, CCRP3, Zakkoyya
Lewis, BS4, H. Colleen Silva, MD5, Elizabeth J. Lyons, PhD5

1
UTMB, Missouri City, TX; 2UTMB, Galveston, TX; 3University of Texas Medical Branch-
Galveston, Texas City, TX; 4University of Texas Medical Branch, Bellflower, CA; 5The University
of Texas Medical Branch, Galveston, TX

Background: Functional impairments are associated with reduced quality of life and survival in
breast cancer survivors. Currently, self-reported measures are most often used to screen
survivors for functional deficits, and community-based measurement is rare. However,
objective measures may better capture complex constructs, and community-based
assessment would potentially allow earlier discovery of deficits in a broader population than is
currently screened. Thus, the aim of this study was to evaluate the feasibility of using
objective functional measures within a community setting of a breast cancer support group.

Methods: Baseline information from a 12-week pilot breast cancer support group
intervention was used to evaluate participants function. Grip strengthassessing muscular
strengthwas evaluated using the Jamal dynamometer and mobility function was assessed
using the Short Physical Performance Battery (SPPB). Feasibility was evaluated based on
number of participants completing both assessments and time required to conduct both
assessments. Descriptive statistics were conducted.

Results: Nineteen survivors completed the grip strength and SPPB assessments. The majority
of the participants were White (68%) with a mean age of 5512 years and mean BMI of
32.47.9 kg/m2. All participants completed three grip strength evaluations for both right and
left hands despite 32% reporting peripheral neuropathy. Seven participants had grip strength
below norms. All attempted the entire SPPB assessments with a mean score of 101.3 out of
12, indicating high functioning. However, 9 of 19 participants completed the 10 seconds
tandem balance test with some difficulty. Approximately 25 minutes per person were
required to complete both assessments using standardized protocols.

Discussion: Overall, we found that it is feasible to conduct grip strength and SPPB
assessments in a breast cancer support group setting. Despite 47% of participants
experiencing difficulty during the 10 seconds tandem balance test, the mean SPPB score was
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1861

high. This indicated a ceiling effect. A protocol modification to increase hold time from 10
seconds to 30 seconds may eliminate the ceiling effect we found and to help capture possible
balance issues.

CORRESPONDING AUTHOR: Maria Swartz, PhD, MPH, RD, LD, UTMB, Missouri City, TX, 77459;
[email protected]
S1862 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C046 6:00 PM-7:00 PM

LIFESTYLE BEHAVIORS AND PSYCHOSOCIAL OUTCOMES AMONG LONG-TERM CANCER

SURVIVORS

Errol J. Philip, PhD1, Thomas V. Merluzzi, PhD2, Steven Grapperhaus, BBA3, Brian Lewis, BS4

1
Laboratory for Psychooncology Research at the University of Notre Dame, Brea, CA;
2
University of Notre Dame, Notre Dame, IN; 3California State University - Fullerton, Fullerton,
CA; 4California State University - Fullerton, San Diego, CA

Purpose: Cancer survivors represent a vulnerable population at increased risk of recurrence

and poor physical and psychosocial long-term effects. These individuals are recommended to
adopt healthy lifestyle behaviors that may reduce the risk of recurrence of certain cancers and
improve overall health status. Further, it is often believed that a cancer diagnosis can
represent a teachable moment and prompt individuals to adopt healthy behavior changes.
Despite this, there exists a paucity of research examining adoption of potentially protective
lifestyle behaviors among long-term survivors, and how such behaviors may be associated
with psychosocial outcomes.

Method: 407 long-term cancer survivors (73.3% female, average age 63.3 years, average time
since diagnosis 14.6 years, 77.1%, White/European American, 19.1% African American)
reported adoption of recommended lifestyle behaviors (regular exercise, non-smoking status,
non-obese weight status) and measures of quality of life (FACT-G), and depression/anxiety
(HADS). Survivors were further divided into groups based on their degree of adoption of
protective health behaviors.

Results: Among a sample of long-term cancer survivors, 80.2% reported regularly exercising,
91.5% were non-smokers and 60.8% were not obese. Overall, 25.6% engaged in all three
recommendations, while 3.9% engaged in none. Importantly, nearly 40% of the sample were
obese and 8.5% reported currently smoking. Further analyses revealed that survivors who
engaged in two or more of the recommended lifestyle behaviors had significantly better
quality of life (114.36 vs. 106.49 p < .001) and fewer symptoms of anxiety (3.99 vs. 5.43, p <
.001 and depression (2.86 vs. 4.33, p < .001) compared to those who engaged in one or zero
behaviors. Further preliminary analyses revealed weight status and exercise to be primarily
responsible for observed differences in psychosocial outcomes.

Conclusion: The importance of health promotion in cancer survivorship is becoming

increasingly recognized. Among a sample of long-term survivors, approximately one-quarter
of those surveyed had adopted general health recommendations, with greater engagement
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1863

associated with improved quality of life and lower rates of anxiety and depression. While
encouraging, a number of challenges remain; nearly 10% of the sample still smoked and 40%
were obese, with weight management continuing to be a critical challenge among survivors
and the general population alike. Further research is needed to maximize engagement in
healthy behaviors and enable professionals to actively and effectively promote lifestyle
change among cancer survivors.

CORRESPONDING AUTHOR: Errol J. Philip, PhD, Laboratory for Psychooncology Research at

the University of Notre Dame, Brea, CA, 92821; [email protected]
S1864 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C047 6:00 PM-7:00 PM

LIFESTYLE INTERVENTION EFFECTS ON MOBILITY AND BODY COMPOSITION IN PROSTATE

CANCER PATIENTS ON ANDROGEN DEPRIVATION THERAPY

Brian Focht, N/A1, Alexander R. Lucas, PhD2, Elizabeth Grainger, PhD3, Christi Simpson, BA4,
Ciaran FAIRMAN, M.S.1, Jennifer M. Thomas-Ahner, PhD1, Steven K. Clinton, MD, PhD1

1
The Ohio State University, Columbus, OH; 2Wake Forest School of Medicine, Winston Salem,
NC; 3The Ohio State University Comprehensive Cancer Center, Columbus, OH; 4The Ohio State
University Comprehensive Cancer Center, Dayton, OH

Although exercise consistently results in significant improvements in clinically relevant

outcomes in prostate cancer (PCa) patients undergoing androgen deprivation therapy (ADT),
the synergistic benefits of promoting concomitant change in both exercise and dietary
behavior could represent an optimal lifestyle intervention approach for offsetting the adverse
effects of ADT on mobility performance and body composition experienced by PCa patients.
The purpose of the single-blind, randomized controlled Individualized Diet and Exercise
Adherence-Pilot (IDEA-P) trial is to evaluate the preliminary efficacy of a combined exercise
and dietary (EX+D), implementing a group-mediate cognitive behavioral (GMCB) approach,
relative to standard of care (SC) treatment upon change in mobility performance and body
composition among PCa patients undergoing ADT. A total of 32 PCa patients (M age = 65
years) on ADT were randomly assigned to the EX+D (n = 16) or SC (n = 16) interventions.
Measures of mobility and body composition were obtained at baseline and 2 and 3 month
follow-up assessments. Results of intention to treat ANCOVA analysis of residualized change
scores yielded a significant Treatment main effects for 400M Walk (p < 0.01) and stair climb (p
< 0.01) performance and body composition (p < 0.05). Results of intention to treat analysis
yielded a significant Treatment main effects for 400M Walk (p < 0.01) and stair climb (p <
0.01) performance and body composition (p < 0.05). The EX+D intervention resulted in
superior improvements in mobility performance relative to SC at 2 and 3 months. Additionally,
the EX+D intervention yielded significantly greater improvements in body weight, body fat
percentage, and fat mass relative to the SC intervention at 3 month follow-up. Findings from
the IDEA-P trial suggest that the EX+D intervention, implementing a GMCB approach designed
to promote adoption and adherence to lifestyle behavior change, resulted in superior changes
in mobility performance and body composition relative to a SC approach. These results
underscore the utility of promoting concomitant change in both exercise and dietary behavior
for enhancing mobility and body composition among PCa patients undergoing ADT.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1865

CORRESPONDING AUTHOR: Brian Focht, N/A, The Ohio State University, Columbus, OH,
43210; [email protected]
S1866 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C048 6:00 PM-7:00 PM

LONGITUDINAL ANALYSIS OF THE PERCEIVED STRESS OF A NEW CANCER DIAGNOSIS AND

PTSD-LIKE SYMPTOMS AT ONE YEAR FOLLOW-UP

Hannah-Rose Mitchell, MPH1, Amanda Ting, BS1, Melanie McKenna, N/a2, William F. Pirl,
M.D., MPH3, Youngmee Kim, PhD2

1
University of Miami, Miami, FL; 2University of Miami, Coral Gables, FL; 3Sylvester Cancer
Center, University of Miami Miller School of Medicine, Miami, FL

Cancer patients are at risk for psychological distress, including post-traumatic stress disorder-
like (PTSD) symptoms reaching clinical levels of impairment. There is a lack of prospective
studies examining early predictors of PTSD in cancer patients. The current longitudinal study
examined the relationship between the appraisal of stress around the time of diagnosis and
the onset of PTSD symptoms around the time of completion of treatment.

Colorectal cancer patients (N=54; stages I-IV, 61% female, 59% Hispanic, M=55 years old)
completed questionnaires at around 3 and 12-months post-diagnosis. At the baseline
assessment, patients completed the Perceived Stress Scale (PSS) measuring their perception
of stress. The Impact of Event Scale (IES) measure of PTSD was administered at both time
points and served as the outcome. Age, gender, ethnicity, and cancer stage were covariates.

Relatively small-to-moderate number of patients reported clinical levels of PTSD symptoms at

3-months and 12-months (13% at both times) post-diagnosis. On the other hand, the majority
reported normative levels of perceived stress around the time of diagnosis (M=14.2, SD=5.94).
Hierarchical linear regression revealed that after controlling for covariates, higher perceived
stress around the time of diagnosis related to higher levels of PTSD symptoms around the
completion of treatment (B=0.90, p < .01).

These longitudinal findings suggest that appraisal and perceptions of stress around the time of
a cancer diagnosis may manifest into sub-clinical or clinical levels of psychological disorder in
patients by the time they complete treatment. The results suggest the need for early
psychosocial assessment to identify patients with high stress appraisals who could benefit
from interventions aimed at stress management and cognitive restructuring.

CORRESPONDING AUTHOR: Hannah-Rose Mitchell, MPH, University of Miami, Miami, FL,

33146; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1867

C049 6:00 PM-7:00 PM

MEDICAL TOURISM: A GOOD THING OR A BAD THING

Linda Ko, PhD

Fred Hutchinson Cancer Research Center, Seattle, WA

Background:Colorectal cancer (CRC) remains the most commonly diagnosed cancer among
Korean Americans (KAs) in part due to low screening rates. Recent studies suggest that some
KA patients engage in medical tourism and receive medical care in their home country. The
impact of medical tourism on CRC screening is unknown. The purpose of this paper was to

1) investigate the frequency of medical tourism, 2) examine the association between medical
tourism and CRC screening, and 3) characterize KA patients who engage in medical tourism.

Methods: This is a community-based, cross-sectional study involving self-administered

questionnaires conducted from August 2013 to October 2013. Data was collected on 193KA
patients, ages 50-75, residing in the Seattle metropolitan area. The outcome variable is up-to-
date with CRC screening, defined as having had a stool test (FOBT or FIT) within the past year
or a colonoscopy within 10 years. Predictor variables are socio-demographics, health factors,
acculturation, knowledge, financial concerns for medical care costs, and medical tourism.

Results:In multi-variate modeling, medical tourism was significantly related to being up-to-
date with colorectal screening. Participants who engaged in medical tourism had 8.91 (95% CI:
3.89-23.89) greater odds of being up-to-date with CRC screening compared to those who did
not travel for healthcare. Factors associated with engaging in medical tourism were lack of
insurance coverage (P=0.008), higher levels of education (P=0.003), not having a usual place of
care (P=0.002), older age at immigration (P=0.009), shorter years-of-stay in the US (P=0.003),
and being less likely to speak English well (P=0.03).

Conclusions: This study identifies the impact of medical tourism on CRC screening and
characteristics of KA patients who report engaging in medical tourism. Healthcare providers in
the US should be aware of the customary nature of medical tourism among KAs and consider
assessing medical tests done abroad when providing cancer care.

CORRESPONDING AUTHOR: Linda Ko, PhD , Fred Hutchinson Cancer Research Center, Seattle,
WA, 98115; [email protected]
S1868 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C050 6:00 PM-7:00 PM

METABOLIC SYNDROME AMONG OVERWEIGHT/OBESE AFRICAN-AMERICAN BREAST CANCER

SURVIVORS VS. MATCHED NON-CANCER CONTROLS

Patricia Sheean, PhD, RD1, Huifang Liang, PhD, MD2, Linda Schiffer, MS, MPH3, Claudia Arroyo,
MPH3, Melinda Stolley, MA, PhD4

1
Loyola University Chicago, Maywood, IL; 2Takeda Development Center Americas, Deerfield,,
IL; 3University of Illinois at Chicago, Chicago, IL; 4Medical College of Wisconsin, Milwuakee, WI

Background: Metabolic Syndrome (MetS) predicts multiple chronic diseases and is

predominant in those who are overweight/obese, as well as minority populations. Emerging
data now suggests MetS may also predict breast cancer (BC) recurrence and overall decreased
survival; however, African-American (AA) BC survivors have been analyzed in aggregate.

Purpose: Therefore, the primary aim of this study was to establish the prevalence of MetS
among an exclusive cohort of AA BC survivors and to compare these estimates with matched
non-cancer population controls. This approach allowed us to assess the impact of BC
treatment on the occurrence of MetS and to generate prevalence estimates affording
comparisons with other BC populations.

Methods: We employed a case-control design. For the cases, we utilized baseline data from a
behavioral weight loss intervention conducted in overweight/obese AA BC survivors
(R01CA154406). To represent non-cancer population controls, we used cross-sectional data
from the National Health and Nutrition Examination Survey. Overweight/obese AA women
with treated Stage I-IIIa BC were matched 1:1 on age, race, sex and body mass index (BMI)
category with non-cancer population controls (n=444 or 222 pairs). All measurements were
procured by trained study personnel and certified clinical laboratories. Three of the following
conditions were used to define MetS: HDL cholesterol < 50 mg/dL (1.3 mmol/L), serum
triglycerides 150 mg/dL (1.7 mmol/L), blood glucose 100 mg/dL (or on treatment), waist
circumference 88 cm, or 130 mm Hg systolic or 85 mm Hg diastolic blood pressure (or on
treatment). Matched-pairs analyses were conducted.

Results:For BC cases, most women had self-reported Stage I (n=76) or Stage II (n=91) disease
and were 6.9 ( 5.2) years post-diagnosis. Compared to non-cancer population controls, BC
cases were more educated (p < 0.001) and had greater access to health insurance, specifically
private insurance ( < 0.001). MetS was significantly less frequent among BC survivors vs. non-
population cancer controls (43.2% vs. 51.4%, respectively; p < 0.05) and did differ significantly
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1869

across any BMI category (Overweight 25.0% vs. 37.9%; Class 1 Obesity 38.1% vs. 44.8%; Class
2 Obesity 64.3% vs. 53.6% and Class 3 Obesity 39.0% vs. 73.2%, respectively.) However, a
lower prevalence of 2 risk factors was observed for all BC cases vs. non-cancer population
controls (80.2 % vs. 85.6%, p < 0.05).

Conclusions: While MetS occurred less frequently in our BC cases vs. non-cancer population
controls, these estimates are nearly two times those reported in other BC investigations,
suggesting important racial/ethnic differences that merit attention and greater intervention
efforts. The prognostic implications of MetS among AA BC survivors remain unknown.

CORRESPONDING AUTHOR: Patricia Sheean, PhD, RD, Loyola University Chicago, Maywood, IL,
60153; [email protected]
S1870 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C051 6:00 PM-7:00 PM

MINDFULNESS MEDIATES ASSOCIATION BETWEEN AGE AND DEPRESSIVE SYMPTOMS IN

EARLY STAGE BREAST CANCER PATIENTS

Melissa Janson, BA1, Kate R. Kuhlman, PhD2, Julienne Bower, PhD3

1
UCLA, Cousins Center for Psychoneuroimmunology, Northridge, CA; 2UCLA, Cousins Center
for Psychoneuroimmunology, Los Angeles, CA; 3UCLA Departments of Psychology and
Psychiatry/Biobehavioral Sciences; Cousins Center for Psychoneuroimmunology, Los Angeles,
CA

Mindfulness mediates association between age and depressive symptoms in early stage
breast cancer patients

BACKGROUND: Diagnosis with breast cancer is highly stressful and is associated with elevated
risk for depression, with negative implications for quality of life, treatment adherence, and
other outcomes. Younger women are at elevated risk for depression in the aftermath of a
breast cancer diagnosis. However, the underlying mechanisms that contribute to age
differences in depressive symptoms in breast cancer patients have not been determined.
Research suggests that acceptance of negative emotional experiences increases with age and
may contribute to lower negative affect. Acceptance is closely related to mindfulness, the
state of being attentive to and aware of ones present moment experience in a
nonjudgmental manner. In the present study, we examined whether dispositional
mindfulness mediated the link between age and depression in women recently diagnosed
with early stage breast cancer.

METHODS: 271 women diagnosed with Stage 0-III breast cancer (mean age=56.2 years,
SD=11.5) were assessed after diagnosis but before starting adjuvant treatment (radiation,
chemotherapy, endocrine therapy). Participants completed the Center for Epidemiologic
Studies Depression Scale (CESD) to assess depressive symptoms in the past week, and the
Mindful Attention Awareness Scale (MAAS) to assess trait levels of mindfulness. We used
multiple linear regression models to examine the association between age and depression, as
well as whether mindfulness mediated this relationship. All analyses controlled for potential
confounds including disease characteristics, ethnicity, education, and income.

RESULTS: Older age was associated with lower depressive symptoms (B=-.24, SE=.06, p < .001)
and with higher levels of dispositional mindfulness (B=.016, SE=.01, p=.002). Higher levels of
mindfulness were associated with lower depressive symptoms (B=-5.00, SE=.70, p < .001),
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1871

and mindfulness partially mediated the association between age and depressive symptoms,
Sobel Test =2.87, SE=0.025, p=.004.

CONCLUSION: The findings suggest that mindfulness may increase with age and therefore be
protective against depressive symptoms following breast cancer diagnosis. Mindfulness
interventions may be aptly targeted toward younger breast cancer patients.

CORRESPONDING AUTHOR: Melissa Janson, BA, UCLA, Cousins Center for

Psychoneuroimmunology, Northridge, CA, 91326; [email protected]
S1872 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C052 6:00 PM-7:00 PM

PATIENT-REPORTED AND GENETIC RISK FACTORS FOR CHEMOTHERAPY-INDUCED NAUSEA IN

CANCER PATIENTS

Kelly A. Hyland, B.A.1, Ashley M. Nelson, MA2, Gillian Bell, PharmD3, Jhanelle Gray, MD4,
Richard Kim, MD4, Paul B. Jacobsen, PhD5, Howard L. McLeod, PharmD 4, Heather Jim, PhD4

1
University of South Florida/Moffitt Cancer Center, Tampa, FL; 2University of South Florida &
Moffitt Cancer Center, Tampa, FL; 3Mission Health, Asheville, NC; 4Moffitt Cancer Center,
Tampa, FL; 5National Cancer Institute, Bethesda, MD

Purpose: Chemotherapy-induced nausea (CIN) is common among cancer patients even when
anti-emetic prophylaxis is provided in accordance with clinical practice guidelines. Early
identification of patients at risk for CIN may help physicians to better personalize anti-emetic
therapies and reduce this undesirable side effect. The current study examined patient-
reported and genetic risk factors for CIN.

Methods: Eligible individuals were a) age > 18, b) English or Spanish-speaking, c)

chemotherapy nave cancer patients, and d) scheduled to receive highly- or moderately-
emetogenic intravenous chemotherapy for solid tumors at Moffitt Cancer Center. Participants
completed a baseline sociodemographic questionnaire and provided a blood sample to
measure genetic variants in ABCB1 (rs1045642) and HTR3B (rs45460698)as well as CYP2D6
activity score. Participants also completed a self-report risk factor questionnaire the morning
of the first infusion and the MASCC Antiemesis Tool (MAT) 24 hours and 5 days post-infusion
to assess the severity of acute and delayed CIN.

Results: Participants were 107 patients (50% male, M=61 years). Twenty-three percent of
participants experienced acute CIN and 55% experienced delayed CIN. In univariate analyses,
younger age, fewer hours slept the night prior to infusion, and a history of morning sickness
with pregnancy were associated with more severe acute CIN; fewer hours slept the night prior
to infusion and receipt of highly emetogenic chemotherapy were associated with more severe
delayed CIN (ps < .05). Of genetic variants assessed in 77 patients, only that in ABCB1 was
associated with acute CIN, with heterozygotes reporting more severe acute CIN. In
multivariate analyses, ABCB1 added an additional 4% predictive value beyond the 10%
variance explained by patient-reported risk factors (p = .067). No genetic variants were
associated with delayed CIN.

Conclusions: The current study identified patient-reported and genetic factors that may place
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1873

patients at risk for acute CIN despite receipt of anti-emetic prophylaxis. Additional studies
examining other genetic variants are needed.

CORRESPONDING AUTHOR: Kelly A. Hyland, B.A., University of South Florida/Moffitt Cancer

Center, Tampa, FL, 33612; [email protected]
S1874 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C053 6:00 PM-7:00 PM

PATTERNS OF EXPERIENTIAL AVOIDANCE AND MISCARRIED HELPING IN PARENTS OF

CHILDREN WITH CANCER

Leah McDiarmid, B.A.1, Casiana Warfield, B.S. 2, Kristina Brookshire, M.S.3, Rachel Kentor,
M.S.3, Michelle Byrd, Ph.D., L.P.4, Megan N. Perez, M.S., M.B.A.5, Lauren Ostarello, M.S.3,
Kristin Kullgren, Ph.D.6, Flora Hoodin, Ph.D.3, Gregory Yanik, M.D.6

1
Eastern Michigan University, Waterford, MI; 2Eastern Michigan University, Belleville, MI;
3
Eastern Michigan University, Ypsilanti, MI; 4Department of Psychology, Eastern Michigan
University, Ypsilanti, MI; 5Oklahoma State University, Guthrie, OK; 6University of Michigan
Health System, Ann Arbor, MI

Family expressiveness and conflict are known to affect child psychosocial adjustment to
cancer. Two potentially pertinent interactional constructs yet to be examined in pediatric
cancer populations are Experiential Avoidance in Parenting (EAP) and Miscarried Helping
(MH). Higher EAP, parental inability to tolerate negative child affect, predicts poorer
psychosocial outcomes in general clinical and pediatric samples. Higher MH, the more well-
intentioned parental efforts to assist their ill children are met with child resistance, reflects
more family conflict. This study of parent-child communication aimed to explore whether
groups of caregivers differ meaningfully in patterns of EAP and MH.
Caregivers (n=161, M age=41, 86% female) of children with cancer (M age=9, 58% male)
completed an electronic survey of self-reported medical background, the Parental Acceptance
and Action Questionnaire (PAAQ; Cheron et al., 2009) and the Helping for Health Inventory
(HHI; Harris et al., 2008), assessing EAP and MH, respectively. A two-step cluster analysis was
conducted to create clusters based on PAAQ and HHI z-scores. Group differences were
examined in follow-up one-way ANOVAs.
The average silhouette measure of cohesion and separation was 0.5, demonstrating that
four clusters were a good fit. Parents in cluster 1 (n=28) were high in both MH (M=1.16,
SD=.68) and EAP (M=1.28, SD=.65). Parents in cluster 2 (n=45) demonstrated higher MH
(M=.47, SD=.54) than EAP (M=-.43, SD=.52). Parents in cluster 3 (n=56) exhibited higher EAP
(M=.42, SD=.42) than MH (M=-.55, SD=.62). Parents in cluster 4 (n=31) were low in both MH
(M=-.89, SD=.44) and EAP (M=-1.33, SD=.47). One-way ANOVA(F(3, 156)=6.60, p=.000) and
post-hoc analyses indicated parents in cluster 1 perceived the probability of cure for their
child as less likely than parents in cluster 2 (p=.002) and 4 (p=.018). Parents in cluster 3, too,
perceived their child as less likely to be cured than parents in cluster 2 (p=.008).
This exploratory cluster analysis yielded 4 groups of caregivers. The group perceiving the
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1875

worst medical outcome for their child had least tolerance for child negative affect and most
parent-child conflict. The group perceiving the best probability for good outcome was
moderately low in intolerance of child negative affect and moderately high in parent-child
conflict. Future research into EAP and MH and their configuration may suggest productive
new directions for interventions in clinical care.

CORRESPONDING AUTHOR: Leah McDiarmid, B.A., Eastern Michigan University, Waterford,

MI, 48328; [email protected]
S1876 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C054 6:00 PM-7:00 PM

POSITIVE AFFECT AND SLEEP QUALITY MEDIATE THE RELATIONSHIP BETWEEN STRESS AND
DEPRESSION AMONG LUNG CANCER PATIENTS

Alyssa K. Choi, B.A. (in progress)1, Timothy J. Williamson, M.A., MPH1, Julie C. Kim, B.A.1,
Annette L. Stanton, Ph.D.2

1
University of California, Los Angeles, Los Angeles, CA; 2UCLA, Los Angeles, CA

Background: Stress and depression evidence a robust association, and poor sleep and low
positive affect (PA) have been hypothesized as mechanisms through which stress predicts
depression. Although some researchers posit that poor sleep more strongly predicts low PA
than the reverse, no studies have examined these variables as sequential mediators explaining
the association between stress and depression. Lung cancer patients report high levels of
stress, sleep disturbance, and depressive symptoms, and it is important to identify putative
mechanisms, which can be harnessed for preventive efforts to reduce depression in this
population. The present study tested two models of serial mediation to illuminate the role of
poor sleep and low PA in explaining the relationship between stress and depression. Method:
Participants (n=33) undergoing treatment for lung cancer completed a daily diary at home for
seven days. Participants reported perceived stress and PA (i.e., happy, calm) at the end of
each day, sleep quality each morning, and depressive symptoms at the end of the week. Serial
mediation was conducted to compare two models: 1) Model 1 tested whether high stress
predicts poor sleep quality, which in turn predicts low PA, which then predicts more
depressive symptoms; 2) Model 2 assessed whether the results differed if low PA preceded
poor sleep quality rather than vice-versa. Results: Low PA significantly mediated the
association between stress and depressive symptoms (b = 5.68, SE = 1.79, 95% CI [2.42, 9.55]),
but sleep quality did not (b = -.79, SE = 1.02, 95% CI [-3.55, .83]). Furthermore, low PA
explained significant variance in the relationship between stress and sleep quality (b = -14.74,
SE = 5.99, 95% CI [-29.07, -5.66]). Results from serial mediation analyses demonstrated that
Model 1 (b = 2.14, SE = 1.16, 95% CI [.32, 4.97]), but not Model 2 (b = -.77, SE = 1.06, 95% CI [-
3.50, .85]), significantly explained the relationship between stress and depressive symptoms.
Unexpectedly, poor sleep quality was unrelated to depressive symptoms. Conclusion: Low PA
mediated the relationship between stress and depressive symptoms and between stress and
sleep quality. Future research should target PA to mitigate the effects of stress on depressive
symptoms in adults diagnosed with lung cancer.

CORRESPONDING AUTHOR: Alyssa K. Choi, B.A. (in progress), University of California, Los
Angeles, Los Angeles, CA, 90095; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1877

C055 6:00 PM-7:00 PM

POSTOPERATIVE QUALITY OF LIFE IN PATIENTS AFTER BOTH TOTAL GASTRECTOMY AND

ESOPHAGECTOMY

Shin Saito, MD1, Yoshinori Hosoya, MD2, Misuzu Nakamura, Ph.D3, Takashi Ui, MD4, Kentaro
Kurashina, MD5, Shiro Matsumoto, MD2, Rihito Kanamru, MD6, Joji Kitayama, MD7, Alan
Lefor, Ph.D.8, Naohiro Sata, MD2

1
Departmet of surgery, Jichi medical university, Shimotsuke, Tochigi, Japan; 2Department of
surgery, Jichi medical university, shimotsuke, Tochigi, Japan; 3Department of nursing, Jichi
medical university, simotsuke, Tochigi, Japan; 4Department of surgery, Jichi medical
university, Shimotsuke, Tochigi, Japan; 5Department of surgery, Jichi medical university,
simotsuke, Tochigi, Japan; 6Department of surgery, jichi medical university, shimotsuke,
Tochigi, Japan; 7Department of surgery, JIchi medical university, shimotsuke, Tochigi, Japan;
8
Department of surgery, Jichi Medical University, Shimotsuke, Tochigi, Japan

Background: Patients with esophageal cancer and a history of previous gastrectomy or

synchronous gastric cancer have neither an esophagus nor stomach after surgery. Few studies
have evaluated quality of life (QOL) in these patients objectively. Health-related QOL includes
physical, social, and psychological aspects, and can be assessed quantitatively. The
Dysfunction After Upper Gastrointestinal Surgery (DAUGS) score was used to evaluate
postoperative QOL in patients undergoing both esophagectomy and gastrectomy.

Methods: From January 1999 to December 2015, 41 patients underwent esophagectomy with
previous or concurrent total gastrectomy at our institution. The proximal jejunal flap was
brought up in a Roux-en-Y fashion subcutaneously together with a vascular pedicle and the
divided vessels anastomosed to the internal thoracic vessels. Of these, 25 patients were
surveyed because they were without recurrence and did not receive chemotherapy within
three months of the survey. Patients were sent a DAUGS questionnaire, rating 20 items in
seven categories related to postoperative dysfunction on a scale of 0 to 5, with higher scores
indicating more dysfunction.

Results: A total of 19 patients (76%) completed and returned the questionnaire. Of these, 11
patients had a previous gastrectomy (mean 20.8 years before), and 8 patients underwent
simultaneous esophagectomy and gastrectomy. The mean (standard deviation) DAUGS20
score in these 19 patients is 25.114.0. Weight loss after surgery was 10.66.5 (%).
S1878 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusion: Previous reports describe a mean DAUGS20 score after gastrectomy alone of 27.8
and after esophagectomy alone of 36.1. Patients in this study had lower DAUGS scores (less
dysfunction), despite resection of both esophagus and stomach, for which there are several
possible explanations. The reconstruction used may reduce reflux symptoms. The majority
(11/19) of these patients had undergone gastrectomy and were already impaired when they
underwent esophagectomy. In addition, some patients may be satisfied with their
postoperative QOL and dysfunction, simply by surviving their disease.

CORRESPONDING AUTHOR: Shin Saito, MD, Departmet of surgery, Jichi medical university,
Shimotsuke, Tochigi, 329-0498; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1879

C056 6:00 PM-7:00 PM

PREDICTORS OF HPV VACCINE INITIATION AMONG LOW INCOME, ETHNIC MINORITY

ADOLESCENTS IN CALIFORNIA

Beth Glenn, PhD1, Roshan Bastani, PhD1, Rita Singhal, MD, MPH2, Catherine M. Crespi, PhD1,
Jennifer Tsui, PhD, MPH3, Alison Herrmann, PhD, MS1, Narissa Nonzee, MA1, Li-Jui Cindy.
Chang, MS1, Victoria Taylor, MD, MPH4

1
University of California Los Angeles, Los Angeles, CA; 2LA County Department of Public
Health, El Monte, CA; 3Rutgers Cancer Institute of New Jersey, New Brunswick,, CA;
4
University of Washington, Seattle, WA

Prophylactic HPV vaccines have been widely heralded as a breakthrough in cancer prevention.
However, uptake of the vaccine in the US remains disappointingly low. As part of a
randomized trial to improve HPV vaccination rates among low income, ethnic minority
adolescents in Los Angeles County, we conducted telephone interviews with caregivers of
never-vaccinated adolescents (10-17 years) at baseline and 9-months post intervention (n =
211). The intervention, targeted to caregivers of adolescents and delivered through a
telephone hotline operated by the county health department, demonstrated no significant
effect on HPV vaccine initiation. Therefore, we examined baseline characteristics of caregivers
(demographics, health care access, HPV-related knowledge/attitudes) and adolescents
(demographics, health care access) that might predict vaccine initiation over the study period.
The following caregiver factors predicted a greater likelihood of vaccine initiation at follow-up
in bivariate analyses: Latino ethnicity, less than college education, Medicaid managed care
insurance, having a usual source of care, higher perceived HPV risk, stronger beliefs about
vaccine effectiveness, lower cost concerns, feeling sufficiently informed to make a decision
about vaccination, higher self-efficacy to have child vaccinated, previous provider
recommendation for HPV vaccination, and intention to have child vaccinated within 6 months
(p < .05). No child-level factors were associated with HPV vaccination. These results suggest
subgroups of caregivers that might be targeted for intervention (e.g., non-Latino caregivers).
In multivariate analyses including study condition, demographics, and significant bivariate
factors, only intention to have child vaccinated in the next 6 months (OR = 4.8, p < .01) and
having received a previous provider recommendation for vaccination (OR = 4.5, p < .01) were
significant. Results underscore the critical influence of behavioral intentions and provider
recommendation on receipt of preventive health services. Only 18% of caregivers in our
sample had received a provider recommendation for HPV vaccination for their child despite
the fact that 90% reported their child had a usual source of care. These findings suggest that
interventions targeting health care providers to reduce missed opportunities during health
S1880 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

care encounters may be an especially important strategy in efforts to improve HPV vaccine
uptake.

CORRESPONDING AUTHOR: Beth Glenn, PhD, University of California Los Angeles, Los Angeles,
CA, 90095; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1881

C057 6:00 PM-7:00 PM

PRELIMINARY FEASIBILITY OF A WEB-BASED COGNITIVE BEHAVIORAL THERAPY FOR

INSOMNIA INTERVENTION DURING CANCER TREATMENT

Amma Agyemang, PhD, MPH

Virginia Commonwealth University Medical Center, Richmond, VA

BACKGROUND: Insomnia is common in cancer, especially around the time of cancer diagnosis
and at the onset of cancer treatment. Yet it is often unrecognized and untreated. Cognitive
Behavioral Therapy for Insomnia (CBT-I) is gold standard for treating insomnia in cancer.
Previous studies have largely used homogenous samples of White, well-educated women with
breast cancer who are years post-treatment. CBT-I dissemination in oncology is further
limited by a lack of available providers and resources needed to implement standard face-to-
face treatment. A solution is to abbreviate standard CBT-I into low-intensity modalities that
can be self-administered online. The current pilot study examined the feasibility,
acceptability, and efficacy of implementing a 6-module online self-administered CBT-I
intervention, Sleeping Healthy Using the Internet (SHUTi), with optional clinician support,
during cancer treatment.
METHODS: Twenty-eight individuals who were on average 2 months post-cancer diagnosis
(36% breast, 25% gastrointestinal, 21% head and neck) and met DSM-IV criteria for insomnia
were randomized to treatment at usual (TAU) (n=14) or SHUTi (n=14). Participants were
undergoing chemotherapy and/or radiation with curative intent, at an NCI-designated cancer
center that is also a safety net hospital (mean age = 47; 79% female; 64% White, 32% Black;
54% reported income < $30,000; 18% were uninsured/indigent, 18% on Medicaid; 25% had
only completed high school or less ). Pre- and post-sleep diaries, and self-reported global
sleep disturbances, psychological distress, and health-related quality of life were collected.
RESULTS: Over an 8-month period, 83% of individuals approached about the study were
screened and 100% of individuals meeting study eligibility consented to participate. The
overall retention rate was 84% with similar attrition for both the TAU (20%) and SHUTi (13%)
groups. On average, participants completed the study in four months, due to delays before
(13%) and during (37%) cancer therapies. In the SHUTi group, 91% of participants who
completed all six modules accessed clinician support; Black participants and those who were
unemployed or on disability more readily accepted clinician support. SHUTi participants rated
the program moderately acceptable (M=3.6; 1-5 scale), easy to use (M=4.1) and easy to
understand (M=4.8). Compared to TAU, SHUTi participants reported greater decrease in
global insomnia severity (p < 0.05; d=0.60), subjective time awake after sleep onset (p < 0.05;
d=0.49), and depressive symptoms (p < 0.05; d=0.43) post-intervention.
CONCLUSIONS: There is preliminary support for the feasibility and efficacy of addressing
S1882 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

insomnia during cancer treatment with a web-based CBT-I program. Some form of clinician
support may be critical for patients, especially those from underserved groups, as they
navigate unexpected delays during cancer treatment.

CORRESPONDING AUTHOR: Amma Agyemang, PhD, MPH, Virginia Commonwealth University

Medical Center, Richmond, VA, 23223; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1883

C058 6:00 PM-7:00 PM

PREVALENCE AND CORRELATES OF SKIN CANCER SCREENING AMONG INDOOR TANNERS AND
NON-TANNERS

Carolyn Heckman, PhD1, Elizabeth Handorf, PhD2, Melissa Auerbach, PhD in Social and Health
Psychology3

1
Fox Chase Cancer Center, Philadelphia, PA; 2Fox Chase Cancer Center, philadelphia, PA; 3Fox
Chase Cancer Center/Temple University Health System, Williamstown, NJ

Introduction:

Millions of Americans who have indoor tanned are at much higher risk of melanoma than the
general population. The FDA recommends indoor tanners be screened regularly for skin
cancer. Very little research has examined the association between indoor tanning and skin
cancer screening.

Methods:

This study used data from the 2015 National Health Interview Survey (NHIS). NHIS
methodology is described in detail elsewhere (https://www.cdc.gov/nchs/nhis/). Potential
correlates of skin cancer screening included demographic factors, skin cancer risk factors, and
health care access variables. Multiple logistic regressions were conducted separately for
indoor tanners and non-tanners, with potential correlates included as independent variables
and skin cancer screening as the outcome. Analyses used survey weights to provide results
representative of the US population.

Results:

The sample included 30,352 participants, excluding individuals with prior skin cancer. 16%
reported ever having indoor tanned, and 22% of these had indoor tanned in the last year. 30%
of ever indoor tanners, 20% of never indoor tanners, and 21% of the overall sample had ever
been screened for skin cancer. Based on the multivariate logistic regressions, significant
correlates of skin cancer screening for both ever indoor tanners and non-indoor tanners were
older age, non-Hispanic ethnicity, higher income, looking up health information on the
internet in the last year, sunscreen use, and receipt of a spray-on tan at a salon in the last
year. In non-indoor tanners only, white race, health insurance, not worrying about paying
bills, and sun-burning easily were also significantly related to skin cancer screening.
S1884 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Discussion:

The likelihood of having been screened for skin cancer was higher among indoor tanners than
non-indoor tanners. Screeners in both groups were older and more likely to engage in sun
protection using sunscreen than non skin cancer screeners. Some never indoor tanners who
are at risk for skin cancer have insurance/financial barriers to skin cancer screening. Findings
suggest that: 1) high-risk individuals who do not use sunscreen need to be targeted for skin
cancer screening, and 2) insurance, financial, and information barriers to screening may also
need to be addressed.

CORRESPONDING AUTHOR: Carolyn Heckman, PhD, Fox Chase Cancer Center, Philadelphia,
PA, 19111; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1885

C059 6:00 PM-7:00 PM

PREVALENCE AND PREDICTORS OF HEALTH ANXIETY IN LYNCH SYNDROME PATIENTS

Tae Hart, PhD1, Jenna Albiani, PhD1, Lindsey Torbit, MA1, Crystal Hare, BA, BSc1, Melyssa
Aronson, MSc2, Kara Semotiuk, MSc2, Spring Holter, MSc2

1
Ryerson University, Toronto, ON, Canada; 2Sinai Health Care System, Zane Cohen Centre for
Digestive Diseases, Toronto, ON, Canada

Background: Health anxiety (HA) is characterized by fear or worry about ill health, ranging
from a sense of indifference to intense health-related concerns (Salkovskis & Warwick, 1986).
In particular, elevated HA has been frequently reported in individuals diagnosed with cancer
(e.g., Grassi et al., 2004). However, little data exist on those at high-risk for developing cancer,
such as Lynch syndrome, which is a genetic predisposition to adult onset cancers. Individuals
with Lynch syndrome have a 60-80 percent chance of developing colorectal cancer in their
lifetime (Aarnio et al., 1999). This study examined the severity of and predictors associated
with health anxiety in Lynch syndrome patients.

Methods: Participants (N=209) were recruited from a genetic cancer registry and completed
self-report measures assessing health anxiety (Short Health Anxiety Inventory; Salkovskis et
al., 2002), depression and anxiety (Hospital Anxiety and Depression Scale; Zigmond & Snaith,
1983), cancer worry interference (Trask et al., 2001) and coping (Coping with Health Problems
Inventory; Endler et al., 1998). Hierarchical multiple regression analyses were conducted to
examine predictors of elevated HA.

Results: Results indicated that 30% of participants reported clinically significant levels of
health anxiety. After controlling for gender, time since LS diagnosis, prior cancer diagnosis,
and number of first-degree relatives with cancer, regression analyses indicated younger age,
=-.12, p=.046, increased use of the coping strategy of emotional preoccupation, =.28,
p=.001, more anxiety, =.25, p=.001, more depression, =.18, p=.02, and more worry
interference, =.20, p=.002, were all significantly associated with greater HA, F(8, 180)=20.28,
p=.001.

Discussion: Findings suggest that health anxiety is of clinical significance for individuals with
Lynch syndrome. Moreover, our results suggest that risk factors for high HA include younger
patients who are struggling with general anxiety, depression, and cope with their problems
using worry and rumination strategies. People with hereditary cancer conditions often feel
like health care providers do not understand the negative impact of living with their condition
S1886 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

(e.g., Sheppard et al., 2013). Disseminating these findings to physicians and genetic counselors
working with Lynch patients is critical, which will help to identify and health anxiety among
this population. We will discuss avenues to disseminate information to both providers and
patients, as well as future directions aimed at reducing the distress among Lynch syndrome
carriers.

CORRESPONDING AUTHOR: Tae Hart, PhD, Ryerson University, Toronto, ON, M5B 2K3;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1887

C060 6:00 PM-7:00 PM

PREVENTIVE HEALTH BEHAVIOR PERCEPTIONS AMONG PARENTS OF YOUNG CHILDHOOD

CANCER SURVIVORS

Justin Wilford, PhD

University of California, Irvine, Santa Ana, CA

Background: Childhood cancer survivors (CCS) face a lifetime of significantly elevated risks for
a variety of chronic and life-threatening illnesses. A consensus has emerged on the need for
preventive health behavior education to reduce long-term health risks among CCS, and yet
little is known about preventive health behavior promotion upon completion of treatment
and in follow-up care, especially regarding parents of young and adolescent CCS.

Methods: A cross-sectional survey of 114 parents of young and adolescent (< 13 years)
childhood cancer survivors was used to assess perceptions of preventive survivorship health
behavior promotion. Recruitment was conducted through the online social network
communities of two childhood cancer non-profits that focus on parent wellbeing. The novel
online survey given to parents included 15 questions about receiving survivorship-related
preventive health behavior promotion from medical professionals, the extent to which
parents perceive specific health behaviors to be important for survivorship preventive health,
and the degree to which parents want help with these specific health behaviors.

Results: The percentage of parents reporting receiving any specific preventive health behavior
advice from medical professionals (doctors, nurses, or other) ranged from 54% (promotion of
healthy diet) to 20% (promotion of healthy sleep habits). Nevertheless, parents strongly
endorsed specific health behaviors as being very important for survivorship health, ranging
from 75% of parents for healthy diet to 63% for healthy sleep habits. However, fewer
parents wanted quite a bit or a lot of help with changing specific preventive health
behaviors, ranging from 44% for healthy diet to 33% for physical activity. Parents who
reported receiving survivorship-related healthy diet promotion from a medical professional
were more likely to report that each specific health behavior is important for survivorship
health (for healthy diet, p=.002; for physical activity, p=.009; for healthy sleep, p=.059;
for stress management, p=.035; for social support, p < .000). Parents who reported
receiving survivorship-related health behavior promotion were no more likely to want help
changing any of the specific health behaviors.

Conclusions: Despite inordinate and lifelong chronic disease risks among childhood cancer
survivors, parents do not report receiving uniform preventive health behavior promotion from
medical professionals. When parents receive such promotion, they may be more likely to
S1888 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

believe specific preventive health behaviors to be important for their childs long-term
health.

CORRESPONDING AUTHOR: Justin Wilford, PhD, University of California, Irvine, Santa Ana, CA,
92705; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1889

C061 6:00 PM-7:00 PM

PROMOTING CANCER LIFE MANAGEMENT VIA INTEGRATIVE HEALTHCARE: EFFECTIVENESS IN

IMPROVING EMOTIONAL HEALTH AND IMMUNE FUNCTION

Rebecca L. Shorter, M.A.1, Kerry M. Cannity, M.A.1, Heather C. Zapor, M.A.1, Jerika C. Norona,
M.A.1, Patricia N.E. Roberson, Ph.D.1, John L. Bell, MD2, Gregory L. Stuart, Ph.D.1

1
University of Tennessee-Knoxville, Knoxville, TN; 2University of Tennessee Medical Center,
Knoxville, TN

Up to 50% of cancer patients experience significant depression, anxiety, and stress (Hopko et
al., 2016). Depressed and anxious cancer patients experience decreased quality of life and
immune system functioning, more rapid progression of cancer, more pain, and possibly
increased mortality relative to non-depressed patients (Reddick et al., 2005; Spiegel & Giese-
Davis, 2003). Among novel treatments developed to ameliorate the psychosocial impact of
cancer is integrative healthcare (IH), which combines conventional Western medicine with
adjunctive, complementary treatments (Deng & Cassileth, 2005). There is some evidence to
indicate short-term benefits of IH usage for cancer patients (Helyer et al., 2006); however, the
long-term efficacy of IH has not been systematically studied. The current study examines the
effectiveness of IH for cancer patients, relative to treatment as usual (TAU), in reducing
emotional and physical symptoms and improving immune functioning. Patients were
recruited from the University of Tennessee Medical Centers Cancer Institute (n = 106) and
were eligible to participate irrespective of cancer diagnosis and stage. Patients opted to
continue TAU or engage in an IH program (services include: psychotherapy, massage therapy,
reiki, acupuncture, yoga, and mind-body skills training). A subset of the sample (29%; n = 31)
was randomized to a blood draw condition. Outcome variables included Patient Health
Questionnaires (PHQ-9, GAD-7, and PHQ-15) and plasma cytokine levels (C-Reactive Protein
[CRP] and Interleukin-6 [IL-6]). Patients were assessed at three intervals (baseline, 3-month
and 12-month follow-up). 15% of the sample utilized IH services between baseline and 12-
month follow-up, attending IH services an average of 4.31 times (SD = 6.81, range: 1 29).
Latent growth curve models (LGCM) revealed patients utilizing IH services (0 = no usage; 1 =
1+ times) generally presented with higher anxiety and depression (p < .05). Reductions in
depression occurred more rapidly with IH usage (r = -.05, p < .05), while change in anxiety was
only marginally associated with IH usage (r = -.05, p = .07). IH usage was not significantly
correlated with change in physical/somatic symptoms. Preliminary analyses did not offer
support for change in plasma cytokines over time. Overall, findings offer preliminary support
for long-term benefits of IH in reducing depression and anxiety in cancer patients. Study
limitations and implications will be discussed.
S1890 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Rebecca L. Shorter, M.A., University of Tennessee-Knoxville,

Knoxville, TN, 37996; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1891

C062 6:00 PM-7:00 PM

PSYCHOLOGICAL DISTRESS AFTER PELVIC EXENTERATION FOR RECURRENT GYNECOLOGIC

CANCER

Kristine A. Donovan, PhD, MBA1, Ashley M. Nelson, MA2, Alyssa L. Fenech, BA1, Alexandra
Albizu-Rivera, MPH1, Hye Sook Chon, MD1, Robert M. Wenham, MD1

1
Moffitt Cancer Center, Tampa, FL; 2University of South Florida & Moffitt Cancer Center,
Tampa, FL

Background Pelvic exenteration (PE) has been performed historically in women with
gynecologic cancer for palliation of recurrent disease. With advances in chemotherapy agents
and operative methods, PE in the well-selected patient now produces 5-year survival rates as
high as 60%. Thus, a greater proportion of women are living longer with a wide range of
physical and psychological issues. Despite the recent increase in rate of PE performed for
curative intent, there is a paucity of data on the long-term effects of PE. In the current study,
we sought to characterize womens long-term psychological health after PE for recurrent
gynecologic cancer and examine its associations with overall quality of life (QOL).

Methods As part of a larger mixed-methods study of the effects of PE on QOL, 21 women who
underwent PE between 6 months and 10 years previously completed the Impact of Event
Scale-Revised (IES-R), Center for Epidemiologic Studies-Depression Scale (CES-D), and EORTC-
Quality of Life questionnaire.

Results Participants ranged in age from 43 to 75 years (mean=58) and were a mean of 4 years
(range=0.7-8.8) post-PE. Mean IES-R total score was 21.2+19 and ranged from 0 to 69. Nearly
one-third exceeded the cut-off for clinically significant distress. Mean subscale scores were
5.8+6 for hyperarousal, 6.7+6 for avoidance and 8.3+8 for intrusion. On the CES-D, 71%
exceeded the cut-off for clinically significant depressive symptoms (CES-D total= 19.8+9;
range=3-39). Participants reported above average physical functioning on the EORTC-QLQ
(69+23; range=27-100). Self-reported global QOL was below average (42.5+22; range=0 to
66.7). In regression analysis, psychological health accounted for 49% of the variability in QOL.

Conclusions Findings suggest womens psychological health after PE is relatively poor and
accounts for considerable variability in overall QOL. Findings suggest targets for clinical
interventions to improve psychological well-being and enhance QOL in long-term survivors.
This study highlights the need for clinicians to attend not just to womens physical symptoms
following this life-altering surgery but to monitor and attend to womens psychological health
S1892 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

as well.

Funding Acknowledgment: Leslie Charitable Trust

CORRESPONDING AUTHOR: Kristine A. Donovan, PhD, MBA, Moffitt Cancer Center, Tampa, FL,
33612; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1893

C063 6:00 PM-7:00 PM

PSYCHOMETRIC PROPERTIES OF A 47-ITEM DISTRESS SCREENING TOOL FOR CAREGIVERS OF

CANCER PATIENTS IN A COMMUNITY-BASED SAMPLE

M Longacre, PhD1, M F. Miller, PhD, MPH2, M Golant, PhD2, V Kennedy, LCSW2, A Zaleta,
PhD2, J Buzaglo, PhD2

1
Arcadia University (affiliation was CSC at time of work), Washington, DC; 2Cancer Support
Community, Washington, DC

Background. The landscape of cancer care is changing with patients living longer and
healthcare trending towards outpatient treatment. Consequently, family caregivers face their
own distress. While validated distress screening for patients is essential to quality care,
theres an absence of validated measures to identify and address unmet psychosocial needs of
caregivers. We developed and tested the psychometric properties of CancerSupportSource-
Caregiver (CSS-CG), a 47-item web-based distress screening and referral instrument, including
a 4-item depression subscale. Methods. Family caregivers rated their concerns about self-care
(27 items), caregiving tasks (11 items), and patient care (9 items). Concurrent validity was
determined by correlations with the Distress Thermometer (DT), the Center for Epidemiologic
Studies Depression Scale (CES-D), the SF-12 mental and physical component scores (MCS,
PCS), the Zarit Burden Interview (ZBI), and the Caregiver Reaction Assessment Scale (CRA).
ROC curve analysis was used to determine sensitivity and specificity for a 4-item depression
subscale compared to the CES-D. Test-retest reliability was assessed in a subsample (n=115)
using the intra-class correlation coefficient (ICC). Results. Participants included 246 caregivers
(median age 54 y) from 10 affiliates of a community-based organization. The 47-item tool and
depression subscale demonstrated high internal consistency (Cronbach's alpha=0.96 and 0.79,
respectively). The total distress score created using summed item scores correlated
substantially (pConclusion. The CSS-CG shows strong psychometric properties and can be
considered a valuable community-based instrument to screen for distress. Next steps include
reducing the number of screening items to facilitate implementation and dissemination.

CORRESPONDING AUTHOR: M Longacre, PhD, Arcadia University (affiliation was CSC at time
of work), Washington, DC, 20036; [email protected]
S1894 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C064 6:00 PM-7:00 PM

PUTTING EMOTIONS IN CONTEXT: FEASIBILITY, ACCEPTABILITY, AND VALIDITY OF AN

EMOTION REGULATION ASSESSMENT FOR CANCER PATIENTS

Claire C. Conley, MA, Barbara L. Andersen, PhD

The Ohio State University, Columbus, OH

Objective: Authors have called for the study of emotion regulation (ER) in health psychology
and psycho-oncology. However, self-report measures of ER may not be appropriate for cancer
patients. ER assessment methodologies that incorporate specific contextual demands may
perform better in this population. Thus, the purpose of the present study was to assess the
feasibility and acceptability of a frequently-used assessment paradigm from the affective
science literature in a sample of cancer patients.

Methods: Cancer patients (N=13) completed an ER task frequently used among samples of
healthy young adults. Participants recalled 4 specific situations in which they had experienced
emotions: (1) non-cancer, moderate intensity emotions; (2) non-cancer, high intensity
emotions; (3) cancer-specific, moderate intensity emotions; and (4) cancer-specific, high
intensity emotions. Participants then rated their use of 11 ER strategies (i.e., reappraisal,
problem-solving, suppression, acceptance, etc.) in each of these situations. Following
completion of the task, participants were asked for feedback on its acceptability (e.g., ability
to comprehend instructions, ease of completion, etc.).

Results: For 52 prompts, 51 examples were generated (98% completion rate), indicating that
the task is feasible and easy to complete. All patients reported that the task was pleasant.
Minimal changes (e.g., providing instructions prior to session to facilitate situation recall) were
suggested. Use of suppression differed between cancer and non-cancer situations (t=1.7,
df=49, p < .05), such that participants suppressed emotions less frequently in cancer-related
situations. Use of acceptance differed between moderately and highly intense situations (t=-
2.3, df=49, p < .05), such that participants accepted emotions more frequently in highly
intense situations.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1895

Conclusions: Use of a context-specific ER task among cancer patients is feasible. Patients

found the task acceptable and, consistent with validity hypotheses, demonstrated differences
in ER strategy use depending on situation type and intensity. Future research should establish
the reliability of this task.

CORRESPONDING AUTHOR: Claire C. Conley, MA, The Ohio State University, Columbus, OH,
43210; [email protected]
S1896 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C065 6:00 PM-7:00 PM

QUALITY OF LIFE OF POSTMENOPAUSAL WOMEN WITH BREAST CANCER TREATED WITH

ADJUVANT ENDOCRINE THERAPY

Keith M. Bellizzi, Ph.D., MPH1, Pamela Taxel, M.D.2, Kate E. Dibble, M.A.1, Elizabeth J.
Siembida, Ph.D.3, Linda S. Pescatello, Ph.D., FACSM1, Elizabeth D. Schifano, Ph.D.1, Samantha
Guarneri, B.A.4, Susan H. Tannenbaum, M.D.2

1
University of Connecticut, Storrs, CT; 2UConn Health, Farmington, CT; 3University of
Connecticut, Bethesda, MD; 4University of Connecticut, Stamford, CT

Background: Breast cancer is the most common cancer in women with an estimated 246,660
new cases to be diagnosed in 2016, joining the over 3 million women living with the disease
(ACS, 2016). To reduce cancer recurrence, postmenopausal women (highest incidence and
prevalence of breast cancer) who test positive for hormone receptors in the tumors, are
candidates for adjuvant endocrine therapy (i.e., aromatase inhibitors (AI)). Despite the
benefits of AIs in the treatment for breast cancer, many women are at risk for complications,
including osteoporosis and fractures all of which can adversely affect health-related quality of
life (HRQOL). Methods: Postmenopausal women with breast cancer (n=174), ages 50 95
(M= 68.5), diagnosed with stage 1-3 disease, forty one percent on AIs, were recruited from an
academic medical center in the Northeast. Demographic and quality of life data (Functional
Assessment of Cancer Therapy-Breast (FACT-B) was collected via patient self-report and
clinical data (AI use, bone health, comorbid conditions) was abstracted from patient medical
records. Results: Preliminary analyses suggest women with a history of adjuvant AI therapy
(compared to those without history of adjuvant therapy use) report better physical and
emotional well-being. Discussion: Despite the well-known clinical benefits of AI therapy for
disease recurrence, it is possible that particular QOL outcomes (i.e., physical, emotional) may
also be positively affected. However, a diagnosis of osteopenia or osteoporosis and bone
fractures, potential side effects of AI use, may put patients at higher risk for worse functional
well-being. The benefits and risks (potential side effects that may affect QOL) of AI use should
be discussed with patients.

CORRESPONDING AUTHOR: Keith M. Bellizzi, Ph.D., MPH, University of Connecticut, Storrs, CT,
06269-1058; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1897

C066 6:00 PM-7:00 PM

RELATIONSHIP OF WORRY ABOUT CANCER TO PERCEIVED RISK AND POSITIVE AND NEGATIVE
EFFECTS OF CANCER

Salene Jones, PhD1, Rebecca Ziebell, BS2, Rod Walker, MS2, Larissa Nekhlyudov, MD, MPH3,
Borsika Rabin, Phd, PharmD, MPH4, Stephanie Nutt, MA, MPA5, Monica Fujii, MPH2, Jessica
Chubak, PhD2

1
Group Health Research Institute, Steilacoom, WA; 2Group Health Research Institute, Seattle,
WA; 3Harvard Medical School, Department Population Medicine, Boston, MA; 4Kaiser
Permanente Colorado, Denver, CO; 5University of Texas Austin, Dell Medical School, Austin,
TX

Background: Worry is associated with negative effects of cancer, such as worse physical
functioning and higher perceived risk of disease progression, but associations with positive
changes post-cancer, such as benefit-finding, have been studied less. We examined the
relationship of worry about cancer to negative and positive effects of cancer and perceived
risk of progression to inform whether worry may lead to negative effects after cancer.

Methods: We invited 1,376 long-term survivors (510 years post-diagnosis) to complete the
survey and 615 (46%) were eligible and completed the survey. Of the 615, 594 survivors of
cancer (breast, prostate, colorectal, lung, melanoma) completed the relevant questions.
Respondents answered questions regarding worry about cancer recurrence/progression,
negative effects of cancer on mental and physical function, perceived risk of
recurrence/progression and benefit-finding (being a stronger person, coping better, making
positive changes). Multivariable regressions estimated associations of worry with negative
effects of cancer, perceived risk and benefit-finding.

Results: Worry was related to greater odds of negative effects of cancer (odds ratio (OR)
range: 1.40 to 1.46, all all ps < .01) and to higher perceptions of recurrence/progression risk
(b=0.69, p < .01). Worry about cancer was unrelated to benefit-finding (all ps >.10).

Conclusions: Worry about cancer is related primarily to negative, but not necessarily positive,
effects of cancer and is related to greater perceptions of disease progression. More research
on effective treatments for worry about cancer is needed.
S1898 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Salene Jones, PhD, Group Health Research Institute, Steilacoom,
WA, 98388; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1899

C067 6:00 PM-7:00 PM

SEDENTARY LIFESTYLE AND GREATER SYMPTOM BURDEN IN PATIENTS DIAGNOSED WITH

ADVANCED CANCERS

Jennifer Steel, PhD1, David Geller, M.D.1, Michael Antoni, Ph.D.2, Lauren Terhorst, PhD3, Frank
Penedo, PhD4, Lora Burke, PhD, MPH5, Allan Tsung, MD1

1
University of Pittsburgh School of Medicine, Pittsburgh, PA; 2University of Miami, Coral
Gables, FL; 3University of Pittsburgh, pittsburgh, PA; 4Northwestern University, Chicago, IL;
5
University of Pittsburgh, Pittsburgh, PA

Background: Physical activity has been shown to improve symptoms such as depression,
fatigue, insomnia, and pain. The aims of this study were to examine the link between physical
activity and symptom burden in patients diagnosed with cancers affecting the hepatobiliary
and pancreatic system.

Methods: Participants were administered a battery of questionnaires that included the Center
for Epidemiological Studies-Depression (CESD) Scale, Brief Pain Inventory, Pittsburgh Sleep
Quality Index, Functional Assessment of Cancer Therapy (FACT)-Fatigue, FACT-Hepatobiliary,
and the International Physical Activity Questionnaire (IPAQ). The IPAQ can be categorized
into sedentary, moderate and vigorous physical activity. Analyses included descriptive
statistics and Analysis of Variance was performed to test hypotheses.

Results: Of the 395 patients enrolled in the study, the mean age at cancer diagnosis was 62
(SD=11); the majority were male (61.7%) and Caucasian (91.1%). Patients who reported
sedentary lifestyle reported the highest levels of depressive symptoms (M=16.61, SD=11.2) vs
those with moderate (M=13.64, SD=8.4) or vigorous physical activity (M =11.83,
9.9); F(2,394)=6.267, p=0.002. Patients who were sedentary also reported significantly higher
levels of fatigue (M= 23.60, SD=13.1) when compared to those who were moderately active
(M=19.70, SD=11.3) or vigorously active (Mean=16.95, SD=12.0), F(2,370)=9.390, p < 0.001.
Levels of physical activity were not associated with pain (p=0.770) or sleep quality (p=0.524).
Sedentary patients reported poorer overall quality of life (M=70.9, SD=17.2) vs those who
were moderately (M=75.01, 14.3) or vigorously physically active (M=80.0, SD=16.6), F(2,
386)=10.607.

Conclusions: Sedentary lifestyle after a diagnosis of cancer was associated with increased
symptom burden that was both statistically and clinically meaningful. Dissemination and
S1900 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

implementation of effective interventions to increase physical activity in cancer patients are

warranted to decrease symptoms burden and improve quality of life

CORRESPONDING AUTHOR: Jennifer Steel, PhD, University of Pittsburgh School of Medicine,

Pittsburgh, PA, 15213; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1901

C068 6:00 PM-7:00 PM

SUICIDAL IDEATION AMONG VETERANS LIVING WITH CANCER

Candice C. Johnson, Psy.D., Kristin M. Phillips, Ph.D., Stephanie N. Miller, Ph.D.

James A. Haley Veterans Hospital, Tampa, FL

Background: Subgroups of Veterans (e.g., those with PTSD, depression, or pain) are at
increased risk for suicide. Suicide rates are also higher among individuals living with cancer
compared to the general population. A recent study of people with metastatic cancer found
factors associated with suicidal ideation (SI) include race, PTSD, Panic Disorder, lack of
spirituality, limited support, not feeling well physically, and seeking psychotherapy. However,
little is known about SI among Veterans with cancer. Objectives: Identify factors associated
with SI among Veterans with cancer. Method: Veterans with cancer (N = 101) who were
referred for mental health evaluation completed a clinical interview and questionnaires,
including Pain Numeric Rating Scale (PNS), Insomnia Severity Index (ISI), Patient Health
Questionnaire (PHQ-2), Generalized Anxiety Disorder (GAD-7), and PTSD Checklist (PCL). SI
was defined by endorsing Thoughts that you would be better off dead or of hurting yourself
in some way on paper-and-pencil questionnaire or thoughts of killing oneself during clinical
interview. Results: 29% reported SI. Compared to those without SI, Veterans with SI had
higher ratings on PNS (t= 2.46, p= 0.016), PHQ-2 (t= 5.36, p< .001), GAD-7 (t= 4.41, p< .001),
and PCL (t= 2.61, p= .011); they were also more likely to have a history of suicide attempt ( 2 =
3.85, p= .050). There were no group differences on disease stage, ISI, race, social support, or
religious prohibition against suicide (p-values >.05). To determine which factors account for
the greatest proportion of variance in SI, a logistic regression was performed with history of
attempts, PHQ-2, GAD-7, PCL, and PNS entered as predictors in separate blocks. The omnibus
model was significant (p < .001). However, only PHQ-2 had a statistically significant
contribution to the model ( = 0.56, p= .008). Conclusion: Although previous studies of SI
among Veterans have emphasized the importance of pain and PTSD, this study is consistent
with the literature showing the link between pain and suicide and PTSD and suicide disappear
when depressive symptoms are considered. Results suggest that pain, PTSD, history of suicide
attempts, and anxiety (including worry and fear) are important to consider when treatment
planning for Veterans with cancer. However, anhedonia and hopelessness explain the greatest
proportion of variance in thoughts of being better off dead or wanting to die and should be
the focus of intervention.

CORRESPONDING AUTHOR: Candice C. Johnson, Psy.D., James A. Haley Veterans Hospital,

Tampa, FL, 33612; [email protected]
S1902 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C069 6:00 PM-7:00 PM

ACTIGRAPHY-ASSESSED SLEEP EFFICIENCY IS ASSOCIATED WITH AMBULATORY BLOOD

PRESSURE IN A COMMUNITY SAMPLE

Caroline Doyle, BA1, Melissa Flores, MS2, Matthew Allison, M.D., MPH3, Timothy Smith, PhD4,
Joshua Smyth, Ph.D5, Daniel J. Taylor, Ph.D.6, Bert N. Uchino, PhD7, John M. Ruiz, Ph.D.8

1
University Of Arizona, Tucson, AZ; 2The University of Arizona, Tucson, AZ; 3University of
California at San Diego, San Diego, CA; 4University of Utah, Park City, UT; 5The Pennsylvania
State University, University Park, PA; 6University of North Texas, Denton, TX; 7University of
Utah, Salt Lake City, UT; 8University of Arizona, Tucson, AZ

High blood pressure is a leading risk factors for overall disease burden, including
cardiovascular disease. Likewise, emerging evidence suggests that various dimensions of sleep
health are associated with cardiovascular health. However, less is known about the
relationship between measures of sleep quality and blood pressure with much of the current
literature reliant on self-reported sleep ratings. The aim of this study was to examine the
association of actigraphic sleep efficiency and both ambulatory day and night-time blood
pressure.

Participants were a community sample of 300 healthy adults (150 men, 150 women) ages 21
to 70 years enrolled in the North Texas Heart Study. The sample was stratified by age within
gender and race/ethnicity, and the mean age at enrollment was 42.44 years (SD=12.76).

Actigraphy using AW Spectrum Actiwatches and sleep diary data were collected over a 48-
hour period. Actigraphic sleep efficiency was defined as the ratio between total sleep time
and time in bed. Ambulatory blood pressure (ABP) was collected at random times during 45-
minute intervals over the corresponding 48-hour period, including one night. Person/day
averages were calculated within day (7:00am to 10:00pm) and night (12:00am to 5:00am).

Higher average sleep efficiency over the two nights was associated with lower average
nighttime systolic and diastolic ABP on the second night, all bs < -0.44, t(208) = -4.05, p <
.001. A temporal trend was also observed where higher sleep efficiency on night one
predicted lower average systolic and diastolic ABP during the subsequent day, all bs< -.17,
t(245) = -2.29 , p < 05. All models were adjusted for age, race/ethnicity, gender, and BMI.

Although this study only provides a snapshot of sleep and blood pressure over a very limited
period, the findings suggest that objectively assessed sleep efficiency may be related to
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1903

cardiovascular disease risk through its effects on daily blood pressure. Studies using objective
measures of sleep health and blood pressure are needed to further explore the role of sleep
in cardiovascular disease risk.

CORRESPONDING AUTHOR: Caroline Doyle, BA, University Of Arizona, Tucson, AZ, 85719;
[email protected]
S1904 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C070 6:00 PM-7:00 PM

ANGER STYLE AS A PREDICTOR OF 10-YEAR HEART ATTACK RISK

Ashley Peak, B.A., Tatiana Richard-Kassar, B.A., Lyssette Rodriguez, B.A., Sarah Schall, B.A.,
Luci Martin, Ph.D.

University of La Verne, la verne, CA

The connection between anger and cardiovascular disease has been inconclusive; however,
demographic and moderating factors may explain these inconsistencies. The current study
investigated whether varying styles of anger expression predict heart attack risk in a sample of
university students. Data was collected across two phases: Phase 1 included a self-report
survey (n = 505; 71% female; mean age 21.39, SD = 4.78) and Phase 2 included a physiological
assessment (n = 258). Fifty-eight percent of the sample self-identified as European American,
19% African American, 12% Latino(a), 8% Asian American, and 4% other ethnicity. The State
Trait Anger Inventory-2 was used to assess anger styles. A 10-year heart attack risk score was
determined using age, gender, total and HDL cholesterol, smoker status, and systolic blood
pressure. Age was entered as 45 years for males and 55 years for females. Anger controlled
out (t = 2.93, p = .004) predicted 10-year heart attack risk (R2 = .05, F(6, 251) = 2.31, p = .03).
Ethnicity was associated with the type of anger expression employed by participants, Pillais
Trace = .93, F(18, 1403) = 1.93, p = .01. European Americans reported the lowest levels of
both state and trait anger in comparison to African Americans and Latinos, respectively.
Though females reported greater overall expression of anger, males reported greater anger
control in and out. Mindfulness skills and coping styles were examined as moderators of anger
style to inform clinical recommendations aimed at prevention.

CORRESPONDING AUTHOR: Ashley Peak, B.A., University of La Verne, la verne, CA, 91750;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1905

C071 6:00 PM-7:00 PM

ANXIETY SENSITIVITY AND MEDICATION NONADHERENCE IN PATIENTS WITH SUSPECTED

ACUTE CORONARY SYNDROME

Othanya Garcia, B.A.1, Marin Kautz, B.A.1, Zunelly Velazquez, B.A.1, Jenny Lee, MPH1, Ian
Kronish, MD, MPH2, Carmela Alcantara, Ph.D.

1
Center for Behavioral Cardiovascular Health, Columbia University Medical Center, New York,
NY; 2Columbia University Medical Center, Center for Behavioral Cardiovascular Health, New
York, NY

Background: Anxiety sensitivity - distress over the physical, social, or cognitive consequences
related to the awareness of anxiety has been associated with increased risk for
cardiovascular disease. Poor adherence to cardioprotective medications may underlie this
association. We examined whether anxiety sensitivity was independently associated with self-
reported medication nonadherence after discharge for a suspected ACS.

Methods: Between 2013 and 2015, a consecutive cohort of patients was enrolled from the
emergency department of a tertiary care hospital in New York City. One month after
discharge, patients were contacted by telephone to have anxiety sensitivity and medication
nonadherence assessed. Anxiety sensitivity was assessed using the Anxiety Sensitivity Index
(ASI-3); a score > 23 was classified as high anxiety sensitivity. Medication adherence was
assessed by asking How much of the time do you take your heart medications? Patients who
reported missing their medications at least some of the time were categorized as
nonadherent. Logistic regression was used to determine the association between anxiety
sensitivity and adherence after adjusting for age, sex, race, education, comorbidity (Charlson
score), prior history of established coronary artery disease, discharge diagnosis (ACS vs non-
ACS), and depressive symptoms (8-item Patient Health Questionnaire score 10).

Results: Among 434 patients with data on anxiety sensitivity and adherence, the mean age
was 62.2 (SD 12.7) years, 44.7% were women, 58.2% Hispanic, and 25.0% had elevated
depressive symptoms (PHQ>9). 39.2% of patients had a confirmed ACS hospitalization and
36.2% had a prior history of CAD. A greater proportion of patients with high anxiety sensitivity
were nonadherent than patients without high anxiety sensitivity (20.3% vs 11.8%; p=.02). In
the adjusted model, anxiety sensitivity was associated with increased odds of medication
nonadherence (AOR 2.1, 95%CI 1.2, 3.9; p=.02). Education was the only other variable
associated with nonadherence in this model, patients with less than high school education
were less likely to be nonadherent (p=0.03). The same pattern was observed when entering
anxiety sensitivity as a continuous measure (adjusted =1.02; 95%CI 1.00, 1.04; p=.03).
S1906 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusion: High anxiety sensitivity was associated with nonadherence to cardiac medications
among patients recently discharged for a suspected ACS. Our results suggest that educating
patients on strategies to reduce their anxiety sensitivity may be useful for improving their
medication adherence, and their elevated cardiovascular risk.

CORRESPONDING AUTHOR: Othanya Garcia, B.A., Center for Behavioral Cardiovascular Health,
Columbia University Medical Center, New York, NY, 10032; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1907

C072 6:00 PM-7:00 PM

ARE PATIENTS WHO RULE OUT FOR ACUTE CORONARY SYNDROME AT RISK FOR
POSTTRAUMATIC STRESS DISORDER?

Zunelly Velazquez, B.A.1, Donald Edmondson, PhD, MPH2, Nathalie Moise, M.D., M.S.1,
Bernard Chang, MD, PhD2, Ying Wei, PhD3, David Lopez Veneros, M.A.1, William Whang, MD,
MS1, Ian Kronish, MD, MPH4

1
Center for Behavioral Cardiovascular Health, Columbia University Medical Center, New York,
NY; 2Columbia University Medical Center, New York, NY; 3Columbia University Mailman
School of Public Health, New York, NY; 4Columbia University Medical Center, Center for
Behavioral Cardiovascular Health, New York, NY

Background: Many patients diagnosed with acute coronary syndromes (ACS; unstable angina
and myocardial infarction) develop posttraumatic stress disorder (PTSD) due to the
experience. Less is known about risk for PTSD in patients who are ruled out for ACS during
their hospital admission.

Methods: A consecutive sample of 500 patients presenting to an urban hospital emergency

department (ED) with symptoms of a probable ACS were enrolled into an observational
cohort study of PTSD after ACS. The reason for ACS symptoms at hospital discharge was
determined through chart review. PTSD was assessed by telephone 1 month after discharge
using the PTSD Checklist (PCL-S). Logistic regression was used to determine the odds of PTSD
(PCL-S32), adjusting for demographics, coronary artery disease history, comorbidities, pre-
existing psychiatric and psychosocial variables, and perceived threat in the ED.

Results: Compared to patients with a confirmed ACS, patients who ruled out for ACS had no
significant differences in the presence of typical ACS somatic symptoms (p>0.1) or in
perceived threat from the ACS symptoms (p=0.87). There were also no differences in the
proportion of patients with and without a confirmed ACS who developed elevated PTSD
symptoms (16.8% versus 19.3%; p=0.53), and risk for PTSD was no different in the adjusted
model (adjusted OR 0.90, 95%CI 0.45, 1.78; p=0.77).

Conclusion: ED presentation with ACS symptoms is sufficient to trigger PTSD whether or not
the etiology was truly due to a life-threatening ACS event. Patients ruled out for ACS should
be targeted for interventions to decrease risk of PTSD.
S1908 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Zunelly Velazquez, B.A., Center for Behavioral Cardiovascular

Health, Columbia University Medical Center, New York, NY, 10032;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1909

C073 6:00 PM-7:00 PM

ASSOCIATIONS OF GENDER IDENTITY AND SEX WITH INFLAMMATION LEVELS IN

HYPERTENSIVE PATIENTS

Nicola J. Paine, PhD1, Kim L. Lavoie, PhD2, Andre Arsenault, MD3, Simon L. Bacon, PhD4

1
Concordia University & Hpital du Sacr-Cur de Montreal, Montreal, PQ, Canada;
2
University du Quebec a Montreal, Montreal, PQ, Canada; 3Montreal Behavioural Medicine
Centre, Montreal, PQ, Canada; 4Concordia University, Montreal, PQ, Canada

Background: Elevated inflammation, as evidenced by increased C-reactive protein (CRP), has

been associated with a greater risk of developing hypertension. Sex (a biological construct) is
also a risk factor for hypertension, with men generally being at greater risk than women.
However, women have been found to have higher levels of CRP compared to men. The
aetiology of this difference is not well understood. It is possible that certain aspects of gender
identity (a sociological construct), may explain this. We investigated the specific role of gender
identity and sex on CRP levels in hypertensive patients.

Methods: A total of 798 hypertensive patients (54.3% women, mean SD age (yrs) =67.9
8.72) undergoing pharmacological stress testing with self-reported, physician diagnosed
hypertension participated in the study. Patients completed a series of questionnaires,
measuring socio-demographics, health behaviours and gender identity, using the Bem Sex
Role Inventory, to determine masculine and feminine gender identity. Blood samples were
taken to assess CRP levels.

Results: Adjusting for age, waist circumference, alcohol consumption, smoking status,
education level and physical activity, separate general linear model analyses revealed a
significant main effect of sex with men having lower CRP levels (F=24.51, p < .001) but no
effect of masculine gender identity (F=0.19, p=.66). There was no effect of feminine gender
identity (F=0.28, p=.60) on CRP levels despite a main effect of sex with lower CRP in men
(F=24.98, p < .001). Inclusion of a sex by gender identity interaction term revealed a significant
masculine gender identity by sex interaction (F=4.20, p=.041), but not a feminine gender
identity by sex interaction (F=0.76, p = .38). Post-hoc analyses revealed that for women,
higher masculinity may be related to lower inflammation levels, which was not found in men -
indicating that lower levels of masculine gender identity were associated with higher levels of
CRP in women but not men. There was no relationship between feminine gender identity and
CRP
S1910 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusion: Results indicate that in women, higher levels of masculine gender identity might
be related to lower inflammation in patients with hypertension. However, given the complex
nature of defining and assessing gender, and that our study only examined the role of gender
identity, our results need to be interpreted with care. Others, who have used a more
comprehensive assessment of gender by assessing gender roles and/or norms, rather than
solely gender identity, have shown femininity to be predictive of CVD events. Our results need
further investigation and replication with a more comprehensive assessment of gender, and
highlight the need to consider all aspects of gender, not just gender identity.

CORRESPONDING AUTHOR: Nicola J. Paine, PhD, Concordia University & Hpital du Sacr-
Cur de Montreal, Montreal, PQ, H4B1R6; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1911

C074 6:00 PM-7:00 PM

CARDIOVASCULAR EMOTIONAL DAMPENING OF THREAT PERCEPTION: A PRELIMINARY STUDY

OF BLOOD PRESSURE AND FINANCIAL RISK-TAKING

James McCubbin, PhD1, Kenneth Whitesides, N/A1, Jessica Tucker, N/A1, Kyla Davis, N/A1,
Caitlyn Priddy, N/A1, Rachel Basiura, N/A1, Brooke Haile, N/A1, Daniel H. Wood, PhD2

1
Clemson University, Clemson, SC; 2Federal Trade Commission, Washington, DC

Elevated resting blood pressure is associated with emotional dampening, including reduced
responses to painful stimuli, emotionally evocative photographic scenes, and reduced
perception of affect in facial expressions and written narratives. Several current models of
health behavior suggest that perception of threat is a critical motivator in avoidance of risky
health-damaging behavior. We hypothesize that that blood pressure-associated dampening
of threat assessment may influence decision-making, and may be especially important in
propensity for various risk-taking behaviors. Work from our lab has recently shown that
increased resting blood pressure is associated with: 1) increased self-report of risky behavior
from a modified National College Health Risk Behavior Survey, and 2) risky driving behavior
assessed in a high fidelity driving simulator. We seek to build upon these results by
assessment of the relationship between blood pressure and financial risk-taking using a
simulated paired-choice lottery game.

Forty-one healthy, normotensive adults (11 men, 30 women) 18-43 years of age were
recruited for a laboratory study of resting blood pressure and financial risk-taking behavior
using a simulated paired-choice lottery task. Blood pressures were determined over a 10
minute rest period using a calibrated Dinamap V100 monitor. Participants then played a
simulated lottery task involving ten paired-choice decisions to assess propensity for financial
risk-taking.

Average resting systolic blood pressure was 110.7 +/- 1.41 mmHg (mean +/- SE) and diastolic
blood pressure was 66.5 +/- .68. As in previous studies, systolic blood pressure was higher in
men than in women (p=.001), but there were no other significant sex differences in blood
pressure or risk behaviors. Multiple regression predicting lottery risk-taking from systolic and
diastolic blood pressures indicated that higher resting systolic blood pressure was associated
with increased risk-taking (B= -.075, t= -2.215, p= .033).

These data suggest that cardiovascular emotional dampening may be associated with
increased finanacial risk-taking. The body of data suggests that CNS mechanisms associated
S1912 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

with resting blood pressure levels are possibly associated with dampened perception of
threat. This dampened threat perception may reduce motivation to avoid risk-taking behavior
and thus lead to increased risk-taking behavior in multiple domains. Better understanding of
the integration of CNS circuits that influence visceral, emotional, motivational and, hence,
cognitive function may give more insight into individual differences in propensity to engage in
risk-taking behavior. This approach to understanding risk-taking may provide new strategies
for basic and translational research to reduce health-damaging behaviors, and thus reduce
risk for multiple chronic diseases related to unhealthy lifestyle choices.

CORRESPONDING AUTHOR: James McCubbin, PhD, Clemson University, Clemson, SC, 29631;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1913

C075 6:00 PM-7:00 PM

DEPRESSION IN CARDIAC PATIENTS AND RESULTING THREAT PERCEPTIONS IN THE

EMERGENCY DEPARTMENT

Tara B. St. Onge, BA1, Beatrice Konrad, BA1, Gabriel Sanchez, B.A.1, Amberle Cusmano, MS1,
Donald Edmondson, PhD, MPH2

1
Center for Behavioral Cardiovascular Health, Columbia University Medical Center, New York,
NY; 2Columbia University Medical Center, New York, NY

Background: Patients that present to the emergency department (ED) with depression are
more likely to develop posttraumatic stress disorder (PTSD) after an acute coronary syndrome
(ACS) than their nondepressed counterparts. Threat perceptions during a traumatic event are
associated with the severity of subsequent PTSD. Prior research, for example, suggests that
depressed ACS patients are at increased risk of developing PTSD when they present to an
overcrowded ED. However, no study has addressed the mechanism by which these ED threat
perceptions influence PTSD risk. We tested whether depressed cardiac patients were more
likely to perceive the emergency department as stressful during ED evaluation than
nondepressed patients.

Methods: 581 patients being evaluated for acute coronary syndrome were enrolled in the
Reactions to Acute Care and Hospitalization (REACH) study, an observational cohort study in
the emergency department (ED), and provided complete data. Participants completed a
baseline interview containing an Emergency Room Perceptions questionnaire that assessed
perceptions of emergency room crowding, stress, and length of stay. They also completed the
eight-item Personal Health Questionnaire Depression Scale (PHQ 8) to measure depression.

Results: Participants were 60.75 0.6 years, 47% women, and 37.06% were depressed.
Though perceptions of emergency room crowding did not differ [t= -1.292, p=.197], depressed
patients perceived the ED as more stressful [t= -3.927, p=.000], perceived a one hour longer
length of stay [t= -2.530, p=.012], and a 2.3 hour longer wait for a care plan [t= -2.701,
p=.008].

Conclusion: Emergency Departments can be incredibly stressful environments and cardiac

patients with depression are particularly vulnerable to the effects of ED perceptions. In this
study, even though depressed patients did not perceive the ED as more crowded, they did
perceive it as more stressful, their length of stay to be longer, and their wait for a care plan to
be longer. This may be why previous research has found differential effects of ED crowding on
depressed patients.
S1914 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Tara B. St. Onge, BA, Center for Behavioral Cardiovascular Health,
Columbia University Medical Center, New York, NY, 10032; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1915

C076 6:00 PM-7:00 PM

DEPRESSIVE SYMPTOMS AND MORTALITY IN HEART FAILURE: THE ROLE OF MEDICATION

NON-ADHERENCE

Emily C. Gathright, M.A.1, Mary A. Dolansky, RN, PHD2, John Gunstad, PhD3, Richard
Josephson, MS, MD, FACC, FAHA, FACP, FAACVPR4, Shirley M. Moore, PhD2, Joel W. Hughes,
PhD3, Joseph Redle, MD5

1
Kent State University, Providence, RI; 2Case Western Reserve University, Cleveland, OH;
3
Kent State University, Kent, OH; 4University Hospitals Health System, Cleveland, OH; 5Summa
Health System, akron, OH

Introduction

Approximately 20% of individuals with heart failure (HF) experience depression or depressive
symptoms. Depression is detrimental to prognosis in HF, roughly doubling all-cause mortality
risk. However, the mechanisms underlying this relationship are not fully known. Medication
non-adherence may contribute, as depressed patients are less likely to adhere to medication
regimens.

Purpose

The present study evaluated the role of medication non-adherence in the relationship
between depressive symptoms and mortality risk in adults with HF.

Method

Participants were 303 HF patients in a longitudinal, observational study of self-care behaviors.

Participants completed 21 days of medication monitoring using an electronic pillbox.
Medication adherence was treated as a continuous variable. Depression was defined as
Patient Health Questionnaires scores 5. Mortality data were obtained from the Centers for
Disease Control and Preventions National Death Index. Cox proportional hazards regression
was used to assess the relationship between depressive symptoms and mortality, with and
without adjustment for demographic and medical covariates, and medication non-adherence.

Results

After adjustment for covariates, depression was associated with increased all-cause (HR: 2.07;
95% CI: 1.02-4.17), but not cardiovascular, mortality risk. When medication non-adherence
S1916 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

was added to the model, non-adherence (HR: 1.01; 95% CI: 1.002 1.02), but not depression,
was associated with increased mortality risk.

Conclusion

Medication non-adherence appears to contribute to increased risk of all-cause mortality

conferred by depressive symptoms in HF. Depression and medication adherence represent
modifiable risk factors for poor prognosis. Future research is needed to understand whether
interventions that concomitantly target these factors can improve outcomes.

CORRESPONDING AUTHOR: Emily C. Gathright, M.A., Kent State University, Providence, RI,
02906; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1917

C077 6:00 PM-7:00 PM

ELIGIBILITY CRITERIA FOR HYPERTENSION-RELATED BEHAVIORAL INTERVENTIONS: ARE

BLACKS INCLUDED?

Aisha T. Langford, PhD, MPH1, Zhe He, PhD, MS2

1
New York University School of Medicine, New York, NY; 2Florida State University, Tallahassee,
FL, FL

Background: Despite higher rates of hypertension (HTN) in blacks compared to other racial
groups, blacks have historically been underrepresented in HTN-related clinical trials and
health research studies. Behavioral interventions have been shown to improve HTN
management among Americans in addition to, and sometimes in lieu of, drug therapy.
However, it is not clear how often blacks are specifically mentioned in protocols for HTN-
related behavioral interventions.

Objective: To quantify and describe HTN-related behavioral interventions registered with

ClinicalTrials.gov that explicitly include black or African American as part of the eligibility
inclusion criteria.

Methods: Using a novel database developed by Zhe He et al. that summarizes key attributes
of ClinicalTrials.gov, Commonalities in Target Populations of Clinical Trials (COMPACT), we
queried COMPACT for variables of interest. Broadly, there were 323 interventional HTN
studies using behavioral interventions between 01/2005 and 08/2015. Interventional HTN
studies using behavioral interventions between 01/2005 and 08/2015 with black or African
American in the eligibility criteria yielded 28 studies. After qualitative review of the studies,
two were dropped, leaving a total of 26 studies. Descriptive statistics were run in IBM SPSS 24
for gender, whether blood pressure was included as a primary outcome, minimum and
maximum age for recruitment, focus of the behavioral intervention, and range of enrollment
for the studies.

Results: With regard to gender, 22 (85%) of the studies recruited both men and women, while
four studies (15%) only recruited men. Blood pressure was included a primary outcome in 21
studies (81%), while five studies (19%) had other primary outcomes (e.g., insulin sensitivity,
dietary change, medication adherence, and feasibility assessment). The most common
minimum age was 18 (50%), followed by 21 (23%), and 50 (8%). Other age minimums included
19, 35, 40, and 60. Regarding age maximums, 23 studies (89%) did not have an upper age
limit, while two studies had 79 as the cutoff and one study had 80 as the cutoff. The focus of
the interventions included six motivational interviewing (23%), three lifestyle
S1918 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

change/counseling (12%), two DASH diet (8%), two health/patient education or support (8%),
two nursing (8%), and two multimedia/mhealth including DVDs and text messaging (8%). The
nine other studies (34%) included blood pressure screenings, personal health records, group
visits, self-affirmation, community health workers, health literacy, and pharmacy
interventions. The enrollment targets ranged from 25 to 3698.

Conclusion: More research is needed on how eligibility criteria and the focus of behavioral
interventions may affect black enrollment in HTN-related research studies.

CORRESPONDING AUTHOR: Aisha T. Langford, PhD, MPH, New York University School of
Medicine, New York, NY, 10016; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1919

C078 6:00 PM-7:00 PM

EMOTION REGULATION AND ILLNESS PERCEPTIONS: IMPLICATIONS FOR ADHERENCE AND

NEGATIVE CARDIAC EVENTS

Kelly L. Wierenga, PhD, MSN, RN

Case Western Reserve University, Holt, MI

African American women with heart failure (HF) have compounded stressors that negatively
impact HF self-management and heighten the occurrence of negative cardiac events (visits to
emergency departments, hospitalizations, and death for cardiac reasons). Perceptions of
illness coherence and controllability and emotion regulation (the experiencing, processing,
and modulating of emotional responses) are known to facilitate self-management in the face
of stressors. Scant information exists to understand how cognitive and emotional
components of illness may impact health disparities of African American women with HF. This
prospective study examined the hypothesis that difficulties with emotion regulation and
negative perceptions of illness coherence and controllability are detrimental to 1) HF self-
management regimen adherence and 2) negative cardiac events in African American women
with HF. African American women (n = 54) with HF ages 49 to 84 years participated in this
longitudinal descriptive correlational study. Patients recruited from hospitals and heart failure
clinics using convenience sampling completed interviews at intake and at 30 days, with
medical records reviewed at 90 days. Linear and logistic regression models were used to
assess predictors of HF regimen adherence and negative cardiac events. Of the 54 patients, 28
experienced a negative event for any reason over the 90 days, 57% of these being cardiac
related. The only clear predictor of these events was greater New York Heart Association
functional classification ( = 1.47, p = .03). No associations were found between predictors
(emotion regulation, controllability, coherence, age, education) and HF regimen adherence.
Despite the lack of significance with illness perceptions and emotion regulation, the individual
impact on the outcomes prompts further inquiry. Emotion regulation showed a possible
greater impact on negative cardiac events (b = .80, p = .07) than on HF regimen adherence (b
= -.10, p = .59). Alternatively, perceived illness coherence showed less impact on negative
cardiac events (b = -.27, p = .51) than on HF regimen adherence (b = -.32, p = .08). Findings
indicated that a high percentage of women in this sample experienced negative cardiac
events, with the hypothesized predictive factors remaining unexplained.

CORRESPONDING AUTHOR: Kelly L. Wierenga, PhD, MSN, RN, Case Western Reserve
University, Holt, MI, 48842; [email protected]
S1920 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C079 6:00 PM-7:00 PM

EXAMINING THE RELATIONSHIPS AMONG INFORMATION PRACTICES AND NEUROPROCESSING

IN AFRICAN AMERICANS WITH PREHYPERTENSION

Lenette M. Jones, PhD, RN, Kathy Wright, PhD, RN, Shirley M. Moore, PhD, Jared Friedman,
BA, Anthony I. Jack, PHD

Case Western Reserve University, Cleveland, OH

Title: Examining the Relationships Among Information Practices and Neuroprocessing in

African Americans with Prehypertension

Background: Self-management of blood pressure requires patients to find, understand, and

use information to set goals and act on them. Understanding information practices (seeking,
sharing, use) in which African Americans with prehypertension participate and associated
underlying neurobiological mechanisms may be key to designing more effective self-
management interventions to reduce blood pressure. Previous studies have shown that
African American women who participate in information sharing, a social activity where one
patient gives information about blood pressure management to another peer or peers, were
more likely to report using information to self-manage their blood pressure. Studies have also
shown those who are able to switch between two neurocognitive networks the task-positive
network (hypothesized to be involved with using analytic skills, knowledge for problem-
solving and goal-directed action) and the default mode network (hypothesized to be involved
with socio-emotional and motivational cognition) are better self-managers.

Purpose: To explore associations among information practices and the neuroprocesses

associated with the task-positive network (TPN) and the default mode network (DMN) in
African Americans with prehypertension.

Methods: Participants (N = 19; mean age = 53; 53% males) completed questionnaires
assessing information practices related to self-management of blood pressure and underwent
brain imaging (fMRI) to: (1) evaluate ability to switch between the TPN and DMN and (2)
identify activation in the ventromedial prefrontal cortex (associated with emotional
processing).

Results: Findings reveal that information sharing was positively associated with activation in
the ventromedial prefrontal cortex (r = .49, p = .03) and ability to task switch (r = .38, p = .10).
Information seeking and use were not associated with the neuroprocessing variables
assessed.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1921

Conclusions: Uncovering this brain mechanism associated with information sharing suggests
that interventions based on information sharing by African Americans at risk for elevated
blood pressure may be a fruitful blood pressure self-management strategy.

CORRESPONDING AUTHOR: Lenette M. Jones, PhD, RN, Case Western Reserve University,
Cleveland, OH, 44106; [email protected]
S1922 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C080 6:00 PM-7:00 PM

HOSTILITY, DEFENSIVENESS, AND TELOMERE LENGTH IN INDIVIDUALS WITH AND WITHOUT

CARDIOVASCULAR DISEASE

Louisia Starnino, B.Sc1, Gilles Dupuis, PhD2, Bianca D'Antono, PhD3

1
University of Quebec in Montreal/Montreal Heart Institute, Vaudreuil, PQ, Canada;
2
University of Qubec at Montral, Montral, PQ, Canada; 3Montreal Heart
Institute/Universit de Montral, Montreal, PQ, Canada

Background: Shorter telomere length (TL) may indicate premature cellular aging and
increased risk for disease. While individuals diagnosed with depression have been shown to
possess shorter telomeres, data for potentially maladaptive personality traits related to
cardiovascular disease (CVD), such as hostility and defensiveness, are sparse.

Objective: The purpose of this study is toevaluate the cross-sectional association of TL with
hostility and defensiveness in individuals with and without CVD. Whether chronological age,
sex and health status moderates the associations observed is also examined.

Methods: 781 men and women (Mage = 65.5 6.8 yrs) with and without (N = 284) CVD
completed the Marlowe-Crown Social Desirability Scale and the Cook-Medley Hostility Scale.
Relative TL was measured via quantitative polymerase chain reaction of total genomic DNA
samples. Hierarchical regressions were performed separately for hostility and defensiveness,
and included interactions of hostility/defensiveness with age, sex, and CVD status (entered
stepwise) in Block 3. Of the participants without CVD, only 37 were without other illnesses
(i.e, were very healthy). These participants were examined separately in secondary analyses.

Results: As expected, women and younger individuals had longer TL compared to their
counterparts (all ps < .05).In the primary analyses, defensiveness was not associated with TL.
While no main effect of hostility emerged, a 3-way interaction of hostility with age and sex
emerged ( = -0.102, t=-2.820, p=0.005). More specifically, greater hostility was associated
with longer TL in older men and younger women, while in older women, the opposite was
true. This effect remained significant after controlling for behavioural and physiological risk
factors (e.g. smoking, exercise, IL6, systolic blood pressure).In healthy individuals, greater
defensiveness was associated with shorter TL, though this did not reach significance due to
lack of power ( = -0.270, t = -1.701, p = 0.098). This relation was not influenced by sex or age.
Independence from covariates could not be examined given the sample size.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1923

Conclusion: The association of hostility and defensiveness with telomere length was
influenced by age, sex, and health status. Surprisingly, only a small fraction of the variance in
TL was explained by the predictors and covariates included in the analyses. Whether this
reflects disease spectrum bias or other factors remain to be determined, as does the
implication of these relations for the progression of disease.

Key words: telomere length, hostility, defensiveness, age, sex , cardiovascular disease

CORRESPONDING AUTHOR: Louisia Starnino, B.Sc, University of Quebec in Montreal/Montreal

Heart Institute, Vaudreuil, PQ, J7V 2L6; [email protected]
S1924 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C081 6:00 PM-7:00 PM

'I BELIEVE HIGH BLOOD PRESSURE CAN KILL ME:' PATIENTS PERCEPTIONS OF AN
INTERVENTION TO CONTROL HYPERTENSION IN GHANA

Sarah Blackstone, MPH1, Juliet Iwelunmor, PhD1, Joyce Gyamfi, MSc.2, Nana Kofi Quakyi,
MPH2, Gbenga Ogedegbe, MD2

1
University of Illinois Urbana-Champaign, Champaign, IL; 2New York University Langone
Medical Center, New York, NY

Objectives: In Ghana, hypertension represents the second leading cause of out-patient

morbidity and mortality. Regular screening for and early detection of hypertension are
necessary for controlling the burden of the disease, yet are made difficult by the severe
shortage of health care workers. Task-shifting strategy is one such approach to address
hypertension in light of an under resourced health system. Currently in Ghana, there is an on-
going TAsk- Shifting Strategy for Hypertension (TASSH) program in which nurses are trained in
hypertension management. While this study will provide useful information on the viability of
this approach, it is not clear how patients in the intervention perceive hypertension, the task-
shifting strategy, and its effects on blood pressure management. The objective of this paper is
to examine patients perceptions of hypertension and hypertension management in the
context of an on-going task-shifting intervention to manage hypertension in Ghana.

Design: Forty-two patients participating in the TASSH program (23 males, 19 females, and
mean age 61. 7 years) completed in-depth, qualitative interviews. Interviews were
transcribed, and key words and phrases were extracted and coded using the PEN-3 cultural
model as a guide through open and axial coding techniques, thus allowing rich exploration of
the data.

Results: Emergent themes included patients perceptions of hypertension, which

encompassed misperceptions of hypertension and blood pressure control. Additional themes
included enablers (e.g. patients degree of commitment to intervention) and barriers (e.g.
financial difficulties in obtaining antihypertensive medication) to hypertension management,
and how the intervention nurtured lifestyle change associated with blood pressure control.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1925

Conclusions: This study offers a unique perspective of blood pressure control by examining
how patients view an on-going task-shifting initiative for hypertension management. The
results of this study will shed light on factors that can help and hinder individuals in low-
resource settings with long-term blood pressure management.

CORRESPONDING AUTHOR: Sarah Blackstone, MPH, University of Illinois Urbana-Champaign,

Champaign, IL, 61820; [email protected]
S1926 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C082 6:00 PM-7:00 PM

LIFESTYLE BEHAVIORS, TYPE D PERSONALITY AND PARASYMPATHETIC FUNCTIONING

FOLLOWING A SOCIAL AND COGNITIVE TASK

Kevin Tajii, M.S., Mariam Ter-Petrosyan, M.S., Luci Martin, Clinical Psychology, PhD

University of La Verne, La Verne, CA

Heart rate variability has a well-documented relationship with negative health outcomes
including coronary heart disease and all-cause mortality. Lifestyle behaviors, such as diet,
exercise and sleep may contribute to the development of negative health outcomes. Stable
personality traits suggest a means to assess chronic attributes of individuals that may be at
risk for disease. The purpose of this study was to examine whether lifestyle behaviors such as
sleep, caffeine consumption, alcohol consumption and eating habits as well as the trait Type D
Personality were associated with parasympathetic functioning in a diverse sample of young
adults. Participants (n = 95, mean age = 20.77, SD = 4.97) completed self-report measures
including a Lifestyle Behaviors questionnaire, the Type D Scale-14 and a physiological
assessment to compute heart rate variability during a baseline imagery task and following a
social and cognitive condition. Participants (80% female) self-identified as European American
(40%), Latino/a (36%), African American (7%), and other (17%). Descriptive statistics were
revealed for average hours of sleep (7-8 hours = 42.1%), quality of sleep (fairly good = 65.3%),
alcoholic beverages consumed each day (M = .17, SD =.613), average daily caffeine
consumption (M = 1.09, M = 1.203), eats 3-5 servings of fruits/vegetables a day (yes = 38.9%)
and Type D personality (participants with Type D personality = 24.2%). Bivariate correlations
revealed that following a regular exercise routine was associated with greater
parasympathetic function at baseline (r = .204 p 2 =.09, F (3, 85) = 5.18, p < .01). Diet, sleep,
alcohol consumption and Type D personality were not significantly associated with
parasympathetic functioning within this sample; however, Type D was associated with lower
quantity of sleep (r = -.295, p < -.01), lower quality of sleep (r = -.43, p < .001) as well as lower
average exercise per day (r = -.216, p < .05). These results suggest that it is important to
continue to examine the complex relationships between lifestyle behaviors, personality traits
and heart rate variability, as these relationships may lead to better health outcomes.

CORRESPONDING AUTHOR: Kevin Tajii, M.S., University of La Verne, La Verne, CA, 91750;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1927

C083 6:00 PM-7:00 PM

PSYCHOSOCIAL FACTORS AND AMBULATORY BLOOD PRESSURE INDEXES

Amanda K. Small, B.A., Matthew J. Zawadzki, PhD.

University of California, Merced, Merced, CA

Ambulatory blood pressure (ABP) methods provide the opportunity to obtain robust BP data,
including how BP may vary over the course of day and during sleep periods. However, in
general ABP research has utilized largely mean measures, ignoring the oscillations in BP across
the day. Given that both BP mean and variability independently predict morbidity and
mortality, there is need to better what relates to ABP variability. The purpose of this study is
to test the relationships between psychosocial factors and ABP measurements, comparing
ABP mean with ABP variability and separately analyzing daytime and nighttime values.
Community participants (n=291; 52.9% Male; age M=40.98; 59.4% Black/African America,
20.6% White, 7.3% Asian/Indian, 16.1% Latino/Hispanic; 63.3% Never married, 14.5% Married,
18.7% Divorced or Separated) completed measures of personality types, depression,
rumination, anxiety, anger, and hostility. Afterward participants completed two 24-hour
sessions of noninvasive ABP monitoring using Spacelabs 90207 ABP monitor spaced 2-4
months apart. Bivariate correlations tested the relationships between psychosocial factors
and ABP measures (mean, standard deviation, and average real variability for systolic and
diastolic BP). Mean measures (both day and night) related to more psychosocial factors than
either variability measure, although the relationships were weak. During the day, openness to
experience related to all three indexes although the relationship was inverse for mean and
standard deviation but not average real variability. Hostility also related to both mean and
standard deviation. Extraversion and Agreeableness related to mean only. Similar patterns
were found in night time measures, although average real variability and mean measures
showed more relationships to traits than during the day, included depression, rumination and
anger. Several relationships exist between psychosocial traits and ABP indexes with mean
measures related to the largest number and variety of traits. This is consistent with prior work
and helps increase understanding about which personality traits relate to BP mean. Future
work should address BP variability. Strong relationships between psychosocial factors and ABP
variability were not identified. It is possible that while ABP mean does related to stable
personality aspects, ABP variability may be better understood when considering considering
context and potential state based psychosocial states rather than traits.

CORRESPONDING AUTHOR: Amanda K. Small, B.A., University of California, Merced, Merced,

CA, 95343; [email protected]
S1928 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C084 6:00 PM-7:00 PM

SOCIAL SUPPORT AND HEALTH BEHAVIORS IN CARDIAC REHABILITATION: ADHERENCE AND

DEPRESSION AS MEDIATORS

Jillian Clark, M.A.1, Kadie M. Harry, M.A.2, Kalon R. Eways, M.A.2, Alisha Howarter, PhD3,
Elizabeth J. Wilson, B.A.2, Kymberley K. Bennett, Ph.D.2

1
University of Missouri-Kansas City, San Diego, CA; 2University of Missouri-Kansas City, Kansas
City, MO; 3Remington College, Kansas City, MO

Limited social support and psychological distress (i.e., depression) have been linked to
negative outcomes, such as recurrence and mortality, following a cardiac event. Further,
these constructs have been associated with health behaviors in both the general population
and within samples of patients with cardiovascular disease (CVD). Using a conceptual model
as a framework, the purpose of this study was to examine the relationship between social
support and health behaviors in individuals with CVD, and the mediating roles adherence to
cardiac rehabilitation (CR) and depressive symptoms may play in this relationship.

CR patients (n = 83) completed questionnaires at Time 1 (beginning of CR), Time 2 (end of CR),
and Time 3 (18 months following the end of CR). Participants completed self-report measures
assessing emotional social support (Time 1), depressive symptoms via the Beck Depression
Inventory-II (Time 1 and 2), and adherence to recommended health behaviors (Time 3).
Adherence to CR was measured by number of sessions completed. The mean age of
participants was 63.9 years, the majority were male (63.9%) and European-American (92.7%)

A path analysis was estimated in MPlus, with social support as the exogenous variable, CR
session and depressive symptoms as mediators, and health behavior adherence as the
endogenous variable. The model controlled for Time 1 depressive symptoms, Time 1 health
behaviors, and risk stratification. Overall, fit indices suggested poor model fit (SB2 [6] =
79.36, p .001; RMSEA = .39, 90% RMSEA CI [.31, .46]; CFI = .73; SRMR = .07). Social support
and adherence to CR were not significantly associated with any of the variables in the path
model. Results did suggest that depressive symptoms at Time 2 was negatively associated
with adherence to healthy behaviors at Time 3 ( = -.29, p = .001). There was no evidence of
mediation. The model in its entirety explained 24.5% of Time 3 health behavior adherence.

Results suggest that CR practitioners should identify individuals who are experiencing greater
depressive symptoms following a cardiac event and promote intervention methods to address
their heightened psychological distress. This approach would be complementary to the
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1929

ultimate goal of CRlifestyle modification to adhere to recommended health behaviors to

reduce the likelihood of cardiac event recurrence and cardiac-related mortality.

CORRESPONDING AUTHOR: Jillian Clark, M.A., University of Missouri-Kansas City, San Diego,
CA, 92122; [email protected]
S1930 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C085 6:00 PM-7:00 PM

THE ASSOCIATION OF RELIGIOSITY AND BLOOD PRESSURE AMONG SEXUAL MINORITIES:

RESULTS FROM A NATIONALLY REPRESENTATIVE SAMPLE

Kalina M. Lamb, B.A.1, Kelsey A. Nogg, B.A.2, Benjamin M. Rooney, B.A.2, Aaron J. Blashill,
Ph.D. 2

1
San Diego State University, San Deigo, CA; 2San Diego State University, San Diego, CA

Objective: Hypertension is a major public health concern, given prevalence and morbidity.
Among the general population, higher levels of religiosity have been associated with lower
blood pressure (BP) compared to nonreligious individuals. However, no known studies have
examined the relationship between religiosity and BP among sexual minorities (e.g., gay,
lesbian, and bisexual individuals). Given that many world religions promote doctrines that
include negative views toward sexual minorities, it is plausible that elevated religiosity among
sexual minorities may be associated with increased BP.

Methods: The current study analyzed data collected from Wave IV of the National
Longitudinal Study of Adolescent to Adult Health (Add Health), a publically available,
nationally representative dataset. Participants were identified as a sexual minority if they
identified themselves as 100% gay, mostly gay, or bisexual, or reported same sex attraction (N
= 297, 25% Male, Mean age = 28.55, SD = 1.68 years). A religiosity variable was created by
standardizing and averaging responses of two items assessing attendance of religious services
and religious activities. BP was measured as systolic (SBP) and diastolic blood pressure (DBP)
each measured with three 30-second interval readings, with scores equaling the average of
the second and third reading. Relevant covariates were controlled for in the models (i.e., age,
gender, BMI, ethnicity, income, education, smoking and alcohol use).

Results: Complex Samples within SPSS (v23) was employed to account for the weighting,
clustering, and stratification inherent to Add Health. Within Complex Samples, two separate
General Linear Models were conducted, with SBP and DBP entered as outcome variables.
Whilst controlling for relevant covariates, elevated religiosity was associated with increased
SBP (b = 1.29, SE = .83, 95% CI: 0.35, 2.54, t = 2.045, p = .04) and DBP (b = 1.01, SE = 0.51, 95%
CI: 0.003, 2.02, t = 2.00, p = .05).

Conclusions: Previous research has found religiosity to be a protective factor from

hypertension among the general population. However, the results from the current study are
the first to test and find an inverse effect among sexual minority individuals. There is a paucity
of research examining the intersection of promoting health as a function of sexual orientation
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1931

and religion. Further research may benefit from examining the interaction of sexual
orientation and religiosity in predicting hypertension.

CORRESPONDING AUTHOR: Kalina M. Lamb, B.A., San Diego State University, San Deigo, CA,
92108; [email protected]
S1932 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C086 6:00 PM-7:00 PM

NEW MOTHERS USE OF NEW MEDIA FOR HEALTH INFORMATION & SUPPORT

Lorraine Walker, MPH, EdD1, Michael Mackert, PhD1, Misha Vaughan, PhD2, Bobbie S.
Sterling, RN, MSN, PhD3, Sherry Hendrickson, PhD, APRN, CNS-BC4, Sarah Guy, MSN, RN5,
Jisoo Ahn, MA4

1
University of Texas at Austin, Austin, TX; 2Oracle, Fremont, CA; 3University of Texas at Austin,
Georgetown,, TX; 4The University of Texas at Austin, Austin, TX; 5The University of Texas at
Austin, Liberty Hill, TX

Electronic media are an important means for new mothers to obtain health information and
support. Uses and Gratifications theory posits that needs of subgroups lead to differing media
use. Understanding mothers use of new media is key to patient education about electronic
resources. The aim of this study was to determine new mothers media use in 6 areas, and
correlates (race/ethnicity, income, stress, emotional support, and number of children) of
specific media use. Data for this study are from a random stratified survey sent to new
mothers in Texas. Race/ethnicity and income level were stratifying variables. The adjusted
response rate was 32%. Respondents had a mean age 31.4 y, SD = 5.4), were about 9 months
postpartum, were African American (n = 35), Hispanic (n = 58), or White (n = 75), and most
had between 1-3 children. The survey included items on demographic factors, perceived
stress, emotional support, and mother- or baby-related media use. Data were analyzed by
percentages and non-parametric tests (Kendalls tau, Kruskal-Wallis, and Friedman). Mothers
use of specific media was as follows: search for baby-related information (95%), search for
mother-related information (87%), email to family or friends about baby (83%), read online
sites for new mothers (78%), socialize on Facebook (77%), and watch YouTube videos on
mother/baby care (46%). Types of media use differed significantly (p < .001). More often
searching for mother- or baby-related information was correlated with having a lower number
of children (-.30 and -.16, respectively). More frequent email use was significantly correlated
with higher income (.33), lower stress (-.17), and lower number of children (-.17). Use of
YouTube, Facebook, and online health reading had no significant correlates, indicating usage
was relatively similar on correlate variables. Also, race/ethnicity and support were not related
to media usage. In conclusion, these findings indicated that mothers top media use was in
searches about babies or mothers, with searches more frequent among those with fewer
children. Race/ethnicity was not a factor differentiating use. These findings support patient
education to identify credible sources (especially for first-time mothers) during health-related
searches.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1933

CORRESPONDING AUTHOR: Lorraine Walker, MPH, EdD, University of Texas at Austin, Austin,
TX, 78701; [email protected]
S1934 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C087 6:00 PM-7:00 PM

PARENT VERSUS CHILD PERCEPTIONS OF FEEDING CONTROL MECHANISMS

Caitlyn Murphy, Student Researcher1, Marla Schroeder, Student Researcher2, Allison Kiefner-
Burmeister, Ph.D.1

1
The University of Findlay, Findlay, OH; 2University of Findlay, Sidney, OH

The parental feeding practices that contribute to childhood obesity are both important and
commonly investigated (Goyder 2007; Musher-Eizenman & Kiefner, 2013). One understudied
area of this line of research that may relate to food preoccupation is within the differentiation
of overt and covert feeding practices (Ogden 2006). This descriptive study sought to
investigate the differences in perception of control practices and eating attitudes of parents
and their children. It was hypothesized that parental unhealthy eating attitudes would relate
to the unhealthy attitudes of their children, and these attitudes would differ depending on
parental control mechanisms. Participants were 25 parent-child dyads (92% White, 60% had
an income of $100k or over, child age ranged from 12-14 years). Children participated in their
sixth-grade classroom and parents filled out paper surveys at home. Measures included the
EAT, the CHeat, the EEQ, the CFPQ, and the Overt and Covert feeding scales (Birch et al. 2001;
Arnow, Kenardy, & Agras 1995; Musher-Eizemann & Holub 2008; Norwika, Flodmark, Hales, &
Faith 2014).

Parents' unhealthy eating attitudes were related to their childs unhealthy eating attitudes.
There was a positive relationship between parent-reported covert feeding control and both
parental dieting behaviors and restriction of their childrens food for weight control (dieting r
= .45, p < .05; restriction r = .57, p < .05), such that the more parents control feeding covertly,
the more they also show dieting behaviors themselves and restricted their childs food based
on their childs weight. Parent-reported covert feeding was positively related to childrens
preoccupation with food (r = .54, p < .05). These findings were not found with overt control.
These findings suggest that covertly controlling practices may relate to childrens unhealthy
relationships with food if the child perceives the covert behaviors. However, not all forms of
covert practices related to negative attitudes. For example, there is a negative relationship
between children who use food for emotional regulation and parents who use covert control
(r = -.42, p < .05).

Children who perceived that parents encouraged a balanced diet with a variety of healthier
food choices also reported parents who used more overt control behaviors (r = .42, p < .05).
Overt control was also related to childrens reports that parents wanted them to be actively
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1935

involved in food preparation and food choices in the household (r = .41, p < .05) and reports
that parents wanted to teach them the reasoning behind overt control (r = .42, p < .05).
Our preliminary study suggests that parent and child perceptions of the different types of
control mechanisms used in a household may be just as important in relation to child
attitudes toward food, dieting, and weight as the actual control mechanisms.

CORRESPONDING AUTHOR: Caitlyn Murphy, Student Researcher, The University of Findlay,

Findlay, OH, 45840; [email protected]
S1936 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C088 6:00 PM-7:00 PM

PARENTAL MONITORING IN THE LINK BETWEEN EARLY DATING INITIATION AND DATING
VIOLENCE AMONG DIVERSE ADOLESCENTS

Ashley C. Baker, B.S.1, Patricia Cabral, M.A.2, Jan Wallander, PhD3

1
University of California Merced, Merced, CA; 2University of California, Merced, North
Hollywood, CA; 3Univ. California, Merced, Los Gatos, CA

Dating violence (i.e. emotional, physical, and sexual) is prevalent among adolescents and can
differ across race/ethnicity. Unhealthy romantic relationships can start early and have long-
term influence on adolescent outcomes (e.g. depression and suicidal behaviors). However,
positive parenting practices, such as parental monitoring, may play a role in reducing the
likelihood of dating violence (DV) among adolescents. Yet, the association between early
dating initiation and subsequent DV across race/ethnicity has not been examined, nor has the
role of parental monitoring in this association. Using data from 3,295 (51% female) Black
(26%), Latino (54%), and White (20%) adolescent participants from the Healthy Passages
longitudinal cohort study, we first examined the prevalence of DV experienced and committed
across racial/ethnic groups by 10th grade. Next, we explored longitudinal associations
between early dating behaviors by 5th grade and subsequent DV by 10th grade, followed by
the moderation of this association by parental monitoring. Among boys and girls, Latino
adolescents reported significantly more physical, emotional, and sexual DV than Black
adolescents, who in turn reported significantly more physical and emotional DV experiences
than White adolescents (ps < .01). Also, adolescents who initiated dating early were more
likely to experience physical DV among Latino boys (OR = 2.28), Latina girls (OR = 2.28), Black
boys (OR = 2.91), and Black girls (OR = 3.19). Among all groups, early dating initiation was
significantly related to a greater likelihood of committing physical DV (OR = 2.11 - 6.59). Early
dating initiation was also related to a greater prevalence of emotional DV experiences among
Latino boys (OR = 2.34), Black boys (OR = 4.11), and Black girls (OR = 4.64), as well as a greater
likelihood of committing emotional DV among Black boys (OR = 4.68) and girls (OR = 2.22).
Whereas White adolescent boys who initiated early dating were more likely to experience
sexual DV (OR = 1.65) and White girls were more likely to commit sexual DV (OR = 1.53).
Moreover, parental monitoring moderated this association regarding physical DV committed
and experiences of physical DV, emotional DV, and sexual DV (ps < .05), but in different ways
for different racial/ethnic and gender groups. Findings may have important implications for
parental interventions and approaches to child dating.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1937

CORRESPONDING AUTHOR: Ashley C. Baker, B.S., University of California Merced, Merced, CA,
95340; [email protected]
S1938 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C089 6:00 PM-7:00 PM

PARENTAL RELATIONSHIP QUALITY AND QUALITY OF LIFE AMONG YOUTH WITH AND
WITHOUT CHRONIC MEDICAL CONDITIONS

Lauren E. Wisk, PhD, Elissa R. Weitzman, ScD, MSc

Boston Children's Hospital and Harvard Medical School, Boston, MA

Background: Family functioning is an important contributor to quality of life (QoL),

particularly for families of youth with chronic conditions (YCC), as chronic conditions increase
stress and distress for the affected individual and their family members. As family functioning
and relationship quality may be particularly dynamic during adolescence and young
adulthood, we evaluated trajectories of parental relationship quality (RQ) during this period
for YCC compared to their healthy peers, and evaluated how these trajectories impacted
individual QoL over time. Methods: Longitudinal data are from 3038 adolescents and young
adults (ages 8-28 years) interviewed for the Panel Study of Income Dynamics, a national,
population-based survey, between 2002 and 2013. We utilized multivariate generalized linear
mixed models with natural cubic splines to model trajectories in RQ (MSALT scale, range 1-4)
and their effect on QoL (MIDUS scales for emotional, social and psychological well-being,
range 1-6) for youth with and without chronic conditions, adjusting for confounders. Results:
Compared to their healthy peers, YCC (55% of sample) reported lower RQ with their mothers
and fathers, particularly during late adolescence (2.86 vs 2.91, p < 0.01) and after age 20 (2.64
vs 2.73, p < 0.01) respectively. YCC also reported lower QoL than their peers across emotional,
social and psychological domains (all p < 0.01); these difference persisted after adjustment for
confounders. Better relationships with both mothers and fathers resulted in gains in all three
domains of QoL and differences in QoL by parental RQ diminished with age. Nevertheless,
disparities between YCC and their peers widened over time (particularly during emergent
adulthood) for youth with lower parental RQ (e.g., average emotional QoL at age 16 was 0.24
(a quarter-SD) lower for YCC with low maternal RQ and 0.14 lower for those with high
maternal RQ, but 0.36 lower at age 26 with low maternal RQ and 0.15 lower with high
maternal RQ). Conclusions: Parent-child relationship quality appears to exert a significant
effect on youth QoL well into young adulthood, especially so for YCC who may rely on support
from their parents over a longer period of time than their peers. Practitioners and therapists
may want to assess parent-child relationship quality to buttress resilience resources among
patients with pediatric onset chronic disease.

CORRESPONDING AUTHOR: Lauren E. Wisk, PhD, Boston Children's Hospital and Harvard
Medical School, Boston, MA, 02115; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1939

C090 6:00 PM-7:00 PM

PARENTING STYLES AND BODY MASS INDEX: A SYSTEMATIC REVIEW OF PROSPECTIVE STUDIES
AMONG CHILDREN

Rebeccah L. Sokol, B.S. 1, Bo Qin, PhD2, Jennifer Poti, PhD3

1
Gillings School of Public Health, UNC Chapel Hill, Carrboro, NC; 2Rutgers Cancer Institute of
New Jersey, New Brunswick, NJ; 3University of North Carolina at Chapel Hill, Chapel Hill, NC

Parenting style is correlated with outcomes that impact eating and physical activity, and its
classification schema provides a childhood obesity intervention target. Parenting style can be
categorized into groupsauthoritative, authoritarian, permissive, and neglectfulbased on
the response and demand dimensions. Reviews that evaluate the relationship between
parenting style and weight have been largely based on cross-sectional studies, resulting in the
inability to describe the interplay between parenting styles and obesity risk as children age. To
address this, we systematically searched PubMed, Embase, and PsychInfo for studies
published between 1995-2016 that evaluated the prospective relationship between parenting
style and subsequent weight outcomes. We aimed to answer: 1) Is authoritative parenting
associated with less body mass index (BMI) gains; and 2) Are there critical ages where
parenting style has a more pronounced association with later weight outcomes?

We identified 11 prospective cohort studies. Among the 8 studies that categorized parenting
style into groups, 5 provided evidence that authoritative parenting was associated with lower
BMI gains. The 4 studies that conceptualized parenting styles along the response and demand
dimensions offered less consistent evidence. Among the 6 highest quality studies, 4 suggested
a protective role of authoritative parenting style against adverse weight outcomes. However,
only one study controlled for a complete set of confounders. The small number of studies
conducted within certain age groups precluded our ability to ascertain critical periods when
parenting style is most strongly related to child weight. We found parents from minority
populations were underrepresented, and all studies were conducted in the United States or
Australia. Moreover, analyses in the included articles did not attend to the nuances of family
structure, such as one-parent households, non-biological caregivers, or dual mother or father
parenting, and only 4 studies measured paternal parenting style.

Authoritative parenting may be protective against later overweight and obesity, but to better
define this relationship, future studies should consider paternal styles both independently and
in combination with maternal parenting style, conduct studies of a longer duration to allow
S1940 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

the assessment of the critical period to intervene, more carefully control for child-related and
parent-related confounding factors, and account for the nuances in family structure.

CORRESPONDING AUTHOR: Rebeccah L. Sokol, B.S. , Gillings School of Public Health, UNC
Chapel Hill, Carrboro, NC, 27510; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1941

C091 6:00 PM-7:00 PM

PARTICIPANTS EXPERIENCE AND ENGAGEMENT IN CHECK IT!, A POSITIVE PSYCHOLOGY

INTERVENTION FOR ADOLESCENTS WITH TYPE 1 DIABETES

Erin M. Bergner, MPH, MA1, Niral J. Patel, MPH1, Kimberly L. Savin, BA1, Emily R. Hamburger,
MEd1, Robin Whittemore, PhD, APRN, FAAN2, Sarah S. Jaser, PhD1

1
Vanderbilt University Medical Center, Nashville, TN; 2Yale University, School of Nursing,
Orange, CT

Problems with adherence are common among adolescents with type 1 diabetes (T1D).
Adolescents and their families are responsible for managing a complex treatment regimen,
which includes administering daily insulin, and monitoring blood sugar, diet, and activity
levels. Positive psychology interventions are shown to improve health outcomes; however,
they have not been used widely in pediatric populations. We tested the acceptability and
feasibility of Check It!, a positive psychology intervention to improve adherence in
adolescents with T1D.
Adolescents with T1D and their parents were randomized to a positive psychology
intervention (via phone or text message) or an attention control (education) group. The
intervention used positive psychology exercises, small gifts, and parent affirmations to
increase positive affect (PA). Adolescents in both groups received education packets. Exit
interviews and satisfaction surveys were conducted with adolescents and parents (n=58) to
assess the acceptability and feasibility of Check it! from a representative sample of each
group. Participants rated the acceptability of Check It! using a 5-point Likert scale, with higher
scores indicating more favorable ratings. Interview transcripts were independently coded, and
then reviewed to reach consensus. Adolescents were 15.4 1.6 years old; 50% female; 8.6%
non-White; 53.4% on insulin pump; with a T1D duration of 6.6 4.1 years; and A1c of 9.1
.90%. Parents were 41 9.7 years old; 85% female; 8.3% non-White; and 42% had annual
family income >$80,000. Participants rated the helpfulness (3.4 1.1) and enjoyment (3.5
1.0) of Check It! favorably. Parents and adolescents indicated that the intervention provided
support for the adolescent, offering the chance to discuss emotional topics in an age-
appropriate way. Parents appreciated the reminders to provide affirmations to their children,
as they offered an opportunity for self-reflection and helped the parent-child relationship.
Adolescents also enjoyed the parental affirmations, and the opportunity to set adherence
goals and learn about diabetes. However, feasibility of the phone delivery was low: fewer
adolescents in the phone group completed PA phone interviews, as compared to those in the
text group (14% vs. 67%; t=7.97, p < .001).Parents and adolescents reported that technical
issues with cell phones, timing of calls and text messages, life events, and diabetes-specific
S1942 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

factors presented challenges to participation. Participants rated Check It! favorably, but some
feasibility problems were identified, particularly in the phone group. Adolescents and their
parents enjoyed Check it! and found the intervention helpful. Findings indicate that text
messages are more feasible than phone calls for interventions with adolescents, but
determining optimal times to send text messages is important.

CORRESPONDING AUTHOR: Erin M. Bergner, MPH, MA, Vanderbilt University Medical Center,
Nashville, TN, 37203; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1943

C092 6:00 PM-7:00 PM

POSTTRAUMATIC STRESS AND DEPRESSION PREDICTING PEDIATRIC TRANSPLANT OUTCOMES:

RESULTS FROM A MULTISITE PROSPECTIVE STUDY

Claire Dunphy, BA1, Christina Supelana, EdM, MA1, Ravinder Anand, PhD2, Estella M. Alonso,
MD3, George V. Mazariegos, MD4, Robert S. Venick, MD5, John Bucuvalas, MD6, Margaret L.
Stuber, MD7, Eyal Shemesh, MD8, Rachel A. Annunziato, PhD1

1
Fordham University, Bronx, NY; 2The Emmes Corporation, Rockville, MD; 3Lurie Children's
Hospital, Chicago, IL; 4Children's Hospital Pittsburgh, Pittsburgh, PA; 5Mattel Children's
Hospital at UCLA, Los Angeles, CA; 6Cincinnati Children's Hospital Medical Center, Cincinnati,
OH; 7David Geffen School of Medicine at UCLA, Los Angeles, CA; 8Icahn School of Medicine at
Mount Sinai, New York, NY

Introduction

Caregivers of children with chronic medical illnesses often experience posttraumatic stress
symptoms (PTSS) as well as symptoms of depression in the wake of their childs illness or its
treatment. The potential effects of these symptoms on child medical outcomes have not been
investigated prospectively. We evaluated whether caregiver PTSS and depression symptoms
predict robust behavioral (adherence to medications) and medical (rejection) outcomes in a
prospective, multisite study: Medication Adherence in Children who had a Liver Transplant
(MALT).

Methods

Pediatric liver transplant recipients (n=400) ages 1-17 and their primary caregivers (n=386)
were recruited from five pediatric liver transplant centers in the United States. Parents
completed the Impact of Events Scale (IES) to measure caregiver distress and the Center for
Epidemiologic Studies Depression Scale (CES-D) to measure caregiver depression. Patient
medical variables and outcomes were followed for a two-year period. Medication Level
Variability Index (MLVI), which is the level of variability of immunosuppressant levels in the
blood, was used to determine adherence while centrally-read, biopsy-confirmed organ
rejection was the primary medical outcome.

Results
S1944 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

There were no significant associations between IES (PTSS) scores and either adherence or
rejection outcomes. In contrast, there were significant correlations between higher CES-D
(depression) scores and higher MLVI (lower medication adherence); r=.13, p < .01). A trend
towards higher scores on the CES-D was found among those whose children had experienced
rejection (mean=12.4, SD=10.9) in comparison to those who did not (mean=9.1, SD=8.6;
p=.07).

Conclusions

Caregivers PTSS were not a risk factor for poor child outcomes in this cohort, whereas
caregivers depression symptoms were associated with nonadherence. These findings suggest
that symptoms of depression as opposed to PTSS in caregivers confer greater risk for child
well-being and therefore may be a priority for identification and intervention.

CORRESPONDING AUTHOR: Claire Dunphy, BA, Fordham University, Bronx, NY, 10458;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1945

C093 6:00 PM-7:00 PM

PREGNANT AND POSTPARTUM WOMENS AWARENESS OF EXISTING COMMUNITY SUPPORT

PROGRAMS

Danae Dinkel, Ph.D., Kailey Snyder, MS, Sofia Jawed-Wessel, Ph.D.

University of Nebraska at Omaha, Omaha, NE

Background: Research consistently provides evidence that mothers have a critical role in the
well-being of their children and the family as a whole. Specifically, a mothers mental and
physical health is an integral component to her familys well-being. In order to ensure
communities are providing programs that meet the mental and physical needs of mothers, it
is imperative to understand perceptions of available programs for mothers as well as their
desire for additional programs.

Purpose: The purpose of this study was to determine the awareness of existing community
programs that pregnant and postpartum women are aware of and their desire for related
programs.

Methods: Women who were pregnant or up to one-year postpartum (n=527) were recruited
primarily through social media to complete a 32-question survey within a Midwestern
community. The questionnaire contained several open-ended questions regarding use of
current programs, awareness of other programs they did not use, and desired information
they felt could help them to become better parents. Data were analyzed using a content
analysis approach.

Results: Less than have of all mothers (43.8%) reported participating in a program that
provides assistance or support to mothers who are pregnant or up to one year postpartum. Of
those who participated, mothers primarily took part in governmental programs such as WIC
(41.1%) or breastfeeding support groups (17.7%). A portion of women (26.1%) noted they
were familiar with other programs (35.8%) but did not use them because their services were
not needed (23.4%), they had no time (15.2%) or they didnt qualify (13.9%). Importantly,
46.8% of women reported not participating in any programs and 32.1% of women reported
not being aware of any programs that provide support for this population. Almost a quarter of
all women reported their primary needs for more information were in the areas of parenting
concerns (e.g., discipline, starting solid foods) and mental or emotional support (e.g., stress
management, support groups for single or working moms).

Discussion: A third of women were not aware of programs that could support them and their
S1946 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

primary need for more information (parenting concerns and mental/emotional support) did
not align with the programs that women were primarily aware of. Efforts are needed to
ensure that 1) community programs are available that meet womens needs and/or 2) women
are aware that these programs exist.

CORRESPONDING AUTHOR: Danae Dinkel, Ph.D., University of Nebraska at Omaha, Omaha,

NE, 68106; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1947

C094 6:00 PM-7:00 PM

PRELIMINARY EFFICACY OF A PILOT INTERVENTION FOR EMOTIONAL OVEREATING AND BINGE

EATING AMONG ADOLESCENTS

Rebecca C. Kamody, M.S.1, Idia B. Thurston, PhD2, Elvin T. Burton, PhD, MPH3

1
The University of Memphis, Memphis, TN; 2University of Memphis, Memphis, TN; 3Le
Bonheur Children's Hospital / University of Tennessee Health Science Center, Memphis, TN

Background: Overeating to cope with stress develops in adolescence and may progress into
binge eating disorder (BED) in early adulthood. Emotional overeating (i.e., eating in excess to
soothe emotions) can be conceptualized as a subclinical symptom of BED that can lead to
physical and mental health risks and thus it merits intervention among youth. The affect
regulation model, in which overeating occurs in an attempt to regulate emotions, is an
explanatory model for emotional overeating and binge eating (i.e., eating large amounts of
food in a short time with loss of control). The model is consistent with a Dialectical Behavior
Therapy (DBT) approach. In this study, we examined a DBT skills intervention for emotional
overeating in adolescents and young adults (AYA).

Method: AYA aged 14-18 were recruited from a pediatric hospital to participate in a 10 week
outpatient DBT skills group which introduced mindfulness, distress tolerance, emotion
regulation, and interpersonal effectiveness, all in the context of emotional overeating. Three
cycles were implemented (final post-attrition N = 9 [n per group = 3]; 100% Black, 67%
Female; Mage = 15.22 years). AYA completed the Emotional Eating Scale for Children and
Adolescents (EES-C), Eating Disorder Examination Questionnaire (EDE-Q), and Binge Eating
Scale (BES) at baseline and post intervention.

Results: Baseline to post intervention means and standard deviations were compared. Mean
scores on the EES-C decreased from 48.33 (SD = 16.45) to 44.33 (SD = 19.58; Cohens d =
0.22), the EDE-Q Global Score decreased from 3.28 (SD = 1.20) to 2.71 (SD = 0.84; d = 0.55),
and the BES decreased from 11.67 (SD = 10.10) to 8.33 (SD = 7.16; d = 0.38). Mean binges per
month decreased from 6.00 (SD = 6.36) to 4.00 (SD = 6.10; d = 0.32), and the number of
participants meeting BED criteria decreased from five to three.

Discussion: Findings support the preliminary efficacy of an outpatient DBT skills group for
decreasing binge eating and emotional overeating in AYA. This streamlined modality provides
opportunities for individuals with fewer economic resources to pursue needed treatment and
S1948 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

thus may increase scalability. Future directions include conducting a well-powered

randomized controlled trial to determine intervention efficacy.

CORRESPONDING AUTHOR: Rebecca C. Kamody, M.S., The University of Memphis, Memphis,

TN, 38103; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1949

C095 6:00 PM-7:00 PM

PRENATAL CARE AND PREGNANCY-RELATED OUTCOMES AT CHULA VISTA MEDICAL PLAZA

Nicole Villa, MPH

San Diego State University, San Diego, CA

ABSTRACT

Background: The CenteringPregnancy model is a nationally recognized prenatal care program

that emphasizes education, risk assessment, and supportive care in a group environment.
Assessing outcomes between women in CenteringPregnancy and women in traditional care
will shed light upon the possible benefits of group prenatal classes for a population of
predominantly Latina women. Previous research has found the CenteringPregnancy model to
be associated with lower rates of preterm delivery, lower rates of low birth weight babies, and
higher breastfeeding rates, when compared to traditional care. However, these findings have
been mixed among Latina women.

Objective: To determine if women in the CenteringPregnancy program at Chula Vista Medical

Plaza (CVMP) and women in traditional care at CVMP have different pregnancy outcomes.

Methods: Through an audit of electronic medical records (EMRs), outcomes between women
in two different types of prenatal care were evaluated. Indicators and outcomes that were
evaluated include: maternal weight gain, preterm delivery, infant birth weight, Apgar scores,
feeding type at discharge, and feeding type at follow-up. Chi-squared and t-tests were used to
test for differences between the two groups.

Results: Women in the two groups differed significantly in terms of risk level (ppp=.083), the
percentage of low birth weight (LBW) infants (0.0% for Centering versus 10.2% for traditional
care, p=.028), and exclusive breastfeeding at discharge (71.2% for Centering versus 39.0% for
traditional care, p=.014). No significant differences were found between the two groups for
preterm birth, Apgar scores, or breastfeeding rates at follow-up.

Conclusion: The CenteringPregnancy program may offer certain benefits to mothers and their
infants. Future directions in this area should include a qualitative component that explores
the opinions of the mothers and the providers involved in the CenteringPregnancy program.

Keywords: prenatal care, group prenatal care, CenteringPregnancy, pregnancy

S1950 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Nicole Villa, MPH, San Diego State University, San Diego, CA,
92108; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1951

C096 6:00 PM-7:00 PM

MINDFULNESS, YOGA PRACTICE, AND WELL-BEING: PREDICTING SELF-REGULATION OVER A

STRESSFUL TRANSITION

Lucy Finkelstein-Fox, B.A.1, Crystal L. Park, PhD2, Kelly Stoldt, Undergraduate Student1

1
University of Connecticut, Storrs, CT; 2University of Connecticut, Storrs Mansfield, CT

Mindfulness has been associated with well-being in many domains, and may be especially
adaptive during periods of high stress and adjustment, such as the transition to college
(Jimenez et al., 2010). Yoga practice, which often includes aspects of mindfulness, has also
been suggested as a means of promoting well-being during periods of stress (Netz & Lidor,
2003). However, few studies have examined the roles of either mindfulness or yoga practice
in reducing distress during stressful transitions. In order to better understand the connections
between mindfulness, well-being, and experience with yoga in the context of stress, the
present study examined these variables in a sample of first-year undergraduates.

Online surveys were administered to 157 undergraduates at the beginning and end of their
first semester (M age=17.8; 80% female; 79% White). At semester start, students completed
self-report measures of mindfulness (CAMS-R, Feldman et al., 2007), experience with yoga
practice (yes/no and how often), and emotional (depression, anxiety, and stress; DASS-21,
Lovibond & Lovibond, 1995) and spiritual (meaning/peace and faith; FACIT-Sp, Peterman et
al., 2002) well-being. At semester end, well-being was re-assessed.

On average, emotional well-being did not change significantly across the semester, but
average levels of meaning/peace and faith both decreased (ts (128) =2.20 and 2.16, ps= .03).
Regression analysis revealed that mindfulness was negatively associated with individual
increases in depression, anxiety, and stress (s = -0.17, -0.17, and -0.16, ps< .05) over the
semester, such that greater person-level mindfulness predicted smaller increases in
symptomatology. In contrast, mindfulness was not associated with changes in spiritual well-
being. Prior experience with yoga, reported by 55% of participants, was not associated with
mindfulness or measures of well-being other than depression (= -0.15, p < .05). Just 2% (N=3)
of participants reported a current weekly yoga practice.

Despite its well-established association with well-being in the literature, mindfulness

appeared to protect only against reductions in emotional, and not spiritual, well-being in our
sample. Experience with yoga practice also appeared to protect against increases in
depression, but was not strongly associated with mindfulness; however, very few participants
had a regular weekly yoga practice, which may yield the greatest benefits for self-regulation.
S1952 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

The finding that mindfulness and experience with yoga were both somewhat predictive of
emotional well-being over a period of stressful adjustment suggests that both may be helpful
to include in interventions and programming targeted at new college students. These findings
did not extend to spiritual well-being, however, and future research is needed to examine
other variables that may promote resilience in spiritual domains.

CORRESPONDING AUTHOR: Lucy Finkelstein-Fox, B.A., University of Connecticut, Storrs, CT,

06269; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1953

C097 6:00 PM-7:00 PM

NEUROPHYSIOLOGICAL INVESTIGATION ON DIFFERENTIAL EFFECTS OF COMPUTER-

GENERATED-MODAL MUSIC IMPLICATION FOR MUSIC THERAPY

Tadashi Kato, Ph.D., Timothy S. Rice, Shane M. Spencer, Dylan J. Thomas-Rogers, Kayla M.
Kerns

Fairmont State University, fairmont, WV

Objective: The present study examined the neurophysiological effects of seven modal music
(Ionian, Dorian, Phrygian, Lydian, Mixolydian, Aolian, & Locrian). Method: Thirty-three healthy
adults (12 men & 21 women) listened to seven modal music generated by FractMus software
(2-minute-long each) and their 16-channel-electroencephalogram (EEG), pulse wave,
electrodermal activity, and respiration rates were measured. Fast Fourier Transformation
(FFT) was applied to each of the EEG channels, leading to the mean power computation at
alpha (8-13 Hz) and beta (13-30 Hz) frequency bands. FFT was also applied to pulse-wave
interbeat intervals (IBIs) for the evaluation of autonomic nervous system, while pulse rate
itself was also submitted to further analyses. Subjects rated their six basic emotions on 10-
point-Likert scales (happiness, sadness, anger, surprise, fear, & disgust) before and after each
music and the differences between pre- and post- ratings were computed. Mixed Design
ANOVAs ('gender' as between and 'measurement stages' as within factors) were applied to
each parameter. Results: Mixed Design ANOVAs for the mean power of alpha frequency
bands elicited significant main effect on F7 (left frontal lobe) while listening to Ionian and its
post-hoc-t-test with Bonferroni correction suggested that females had higher alpha power
than males (p < .05). Mixed Design ANOVAs for the mean power of beta frequency band
demonstrated a stage difference on T5 (left temporal lobe) where beta power was
significantly higher during Locrian than during Mixolydian (p < .05). A Mixed Design ANOVA
for pulse rate demonstrated a significant interaction between gender and measurement
stages (p < .05) and its post-hoc t-tests with Bonferroni correction suggested that males had
significantly lower pulse rate than females during both Phrygian and Aolian (p < .05). Further,
Mixed Design ANOVAs for Basic Emotions demonstrated the main effects of 'Happiness' and
'Disgust' ratings (p < .05) and their post-hoc-t-tests suggested that Happiness was higher in
females than in males during Phrygian (p < .05) while Disgust was lower in males than in
females during Aolian (p < .05). Discussion: The results suggested differential physiological
response depending on the musical modes, which also interacted with gender. Higher alpha
power in females during Ionian suggested that women were more relaxed while listening to
Ionian, whereas higher beta power during Locrian than during Mixolydian suggested that
Locrian produced higher mental tension, regardless of gender. Higher pulse rate in females
during Phrygian suggested higher excitement in women during Phrygian, while lower pulse
S1954 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

rate in males during Aolian implied the increased calmness in men during Aolian. Further
investigation will be the key to a deeper understanding of the mechanism behind the effective
music therapy.

CORRESPONDING AUTHOR: Tadashi Kato, Ph.D., Fairmont State University, fairmont, WV,
26554; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1955

C098 6:00 PM-7:00 PM

YOGA OFF THE MAT: A DAILY DIARY STUDY EXAMINING THE EFFECTS OF YOGA TO ENHANCE
SELF-COMPASSION AND SOCIAL CONNECTEDNESS

Mo Kishida, MS1, Steriani Elavsky, PhD2

1
The Pennsylvania State Univeristy, State College, PA; 2Masaryk University, State College, PA

While the yoga literature has certainly proliferated over the past decades demonstrating a
wide array of salutary benefits, there is a lack of research examining the relational outcomes
that may be derived from yoga practice. Moreover, the majority of previous research has
been conducted in lab-based contexts; hence, whether the benefits of yoga translate off of
the mat in the context of ones day-to-day life has yet to be determined. Through a daily
diary approach, the overarching aim of the present study was to examine the daily effects of
yoga practice on the relational variables of self-compassion and social connectedness.
Community-dwelling yoga practitioners (N = 110; age range 18-76 years) with a yoga practice
of at least once a week were recruited for a 21-day daily diary study. At the end of the day,
practitioners were asked to complete daily Internet surveys which included questions in
regards to ones yoga practice and relational items of self-compassion and social
connectedness. Multilevel analyses indicated substantial variability on self-compassion (ICC=
.38) and social connectedness (ICC= .39) on a day-to-day basis at the within-person level. In
concurrent models that controlled for weekend effects, mindfulness, and yoga experience,
yoga practice days were associated with greater self-compassion (B=2.59, p< .05). A within-
person association with the duration of yoga practice and self-compassion also emerged such
that practitioners reported greater self-compassion on days in which one practiced more yoga
(minutes) than what is compared to ones usual (B=1.41, p < .05). In multilevel models
predicting social-connectedness, yoga practice days and the duration of yoga practice
demonstrated no effects on ones daily social connectedness (B=.19, p=.76 ; B=.04, p=.92;
respectively). In this community-based sample of yoga practitioners, these results suggest
yoga practice has positive influences in boosting ones self-compassion on a day-to-day basis,
which may lead to enhanced health and wellbeing outcomes.

CORRESPONDING AUTHOR: Mo Kishida, MS, The Pennsylvania State Univeristy, State

College, PA, 16801; [email protected]
S1956 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C099 6:00 PM-7:00 PM

GOAL-SETTING IN DAFNE: CONTENT AND FIDELITY EXPLORATION OF THE GOAL-SETTING

COMPONENT IN A TYPE 1 DIABETES EDUCATION PROGRAMME.

Milou Fredrix, BA, MsC1, Molly Byrne, BA, MsC, PhD2, Jenny McSharry, B.A., H.Dip., M.Sc.,
Ph.D2

1
National University of Ireland, Galway, Galway, Ireland, Galway, Ireland; 2National University
of Ireland, Galway, Galway, Galway, Ireland

BACKGROUND
Type 1 diabetes (T1DM) is one of the most challenging public health issues (Guariguata et al.,
2014). T1DM treatment includes continuous self-management by the patient (Ahola & Groop,
2013). To provide patients with knowledge, and skills, diabetes self-management programmes
such as The Dose Adjustment For Normal Eating (DAFNE) have been arising. The DAFNE
programme is quickly becoming part of routine care globally, and is now available in Europe
and Australia while looking into expanding to the United States. The programme includes 5
days of structured education in T1DM self-management, with an emphasis on insulin dose
adjustment using flexible dietary intake and detailed estimation of dietary carbohydrate
intake as the basis for insulin dosing. However, while DAFNE is well received, it often fails to
help patients instil key self-management practices into their lives on a long term basis (Speight
et al., 2010). A core component of DAFNE is goal setting. Goal setting can help facilitate long
term behaviour change (Miller & Bauman, 2014). However, little is known about how the
DAFNE goal setting component is delivered and received. To improve and further implement
DAFNE, putting this goal setting session under a microscope and exploring its inner
mechanisms could be beneficial. A first step in exploring the goal setting session, is assessing
its content and levels of intervention fidelity (the extent to which an intervention is
implemented as intended). Since DAFNE is delivered by different educators, variation in
fidelity is expected. Therefore, we aimed to explore the content and the fidelity of delivery of
the goal-setting session in DAFNE.

METHOD
The content of the goal setting session, as defined in the DAFNE handbook for delivering the
intervention, was coded using a reliably taxonomy of behavioural change techniques (BCTs;
Michie et al., 2013). A checklist was developed, listing the identified BCTs. To assess fidelity of
delivery of the BCTs, a set number of DAFNE goal-setting sessions were observed and audio-
recorded. Patient participation was observed and qualitatively described.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1957

RESULTS
Problem solving and action planning were the most frequently identified BCTs within the
goal setting session. Preliminary results reveal different levels of intervention fidelity across
various educators. Furthermore fidelity per BCT varied as certain BCTs had higher levels of
fidelity than others.

CONCLUSIONS
This study showed that the content of the goal-setting component of DAFNE is relatively
limited and could benefit from additional behaviour change techniques. Furthermore, the
importance of fidelity checks for the delivery of interventions is highlighted, as differences in
quality of intervention delivery increase the possibility of inconsistent intervention effects
(Michie, Jochelson, Markham, & Bridle, 2009).

CORRESPONDING AUTHOR: Milou Fredrix, BA, MsC, National University of Ireland, Galway,
Galway, Ireland, Galway, 0000; [email protected]
S1958 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C100 6:00 PM-7:00 PM

COMPARING DIABETIC & HEALTHY OLDER WOMEN ON DESIRE TO IMPROVE HEALTH, HEALTH
SCREENING BEHAVIORS, AND COGNITIVE FUNCTIONING

Luciana Lagana, Ph.D.1, Kimberly Arellano, Abigail Pajulas, none2, Alberto Zacarias, Melissa
Martinez, N/A3

1
CSUN, Chatsworth, CA; 2California State University Northridge, Glendale, CA; 3California State
University Northridge, Canyon Country, CA

The incidence of breast cancer peaks in late life; thus, breast cancer screening is important for
older women, and particularly for those already living with a serious health condition such as
diabetes. The latter increases the risk of breast cancer and breast cancer mortality, thus
adequate breast cancer screening in older women with this condition is important. Moreover,
establishing a significant relationship between diabetes and cognitive impairment would have
important clinical and public health implications. The findings of published research suggest
that diabetic as well as pre-diabetic older women often experience impaired cognitive
performance and are at greater risk for developing cognitive impairment: this represents a
significant threat to their ability to follow medical advice and to keep their diabetes in check.
To our knowledge, there are no studies on diabetic versus healthy older women concerning
receiving preventative medical services. To fill a gap in the literature, we compared 36
diabetic older women to 36 healthy older women (i.e., respondents who reported having no
physical illnesses) on: desire to improve their health (hypothesized as being higher in the
diabetes group), receiving mammograms and regular health screenings (analyzed without any
hypotheses, due to the lack of evidence on this topic), as well as cognitive functioning
(hypothesized as lower in the diabetes group, based on prior research findings). Participants
(N=72, Mean age= 69.29, SD= 6.579, Age Range=50-90) were multiethnic, non-
institutionalized women over the age of 50 residing in Los Angeles County who completed our
research packet. The latter contained the first authors demographics List and the Older
Womens Health - Qualitative Protocol, as well as the well-known MiniCog (Borson, 2000). The
results of an ANOVA showed that, as hypothesized, diabetic women desired to improve their
health more than the women in the control group [F (1,70)=11.87, p < .05, 2=.15].
Additionally, upon implementing Chi-square analyses, we found that diabetic respondents
were significantly more likely to receive mammograms [X2(1)=5.87, p < .05] and general
health screenings [X2(1)=4.51, p < .05] than healthy respondents. Surprisingly, in contrast with
prior literatures findings, cognitive health in the diabetic group obtained marginal significance
in an ANOVA as being better than the cognitive health of the control group [F(1,68)=3.30, p <
.10, 2=.05].
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1959

CORRESPONDING AUTHOR: Luciana Lagana, Ph.D., CSUN, Chatsworth, CA, 91313;

[email protected]
S1960 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C101 6:00 PM-7:00 PM

DEPRESSION AND DIABETES SELF-CARE AMONG BLACK AMERICAN ADULTS: MODERATING

EFFECTS OF REASONS FOR LIVING AND SPIRITUAL COPING

Rheeda Walker, Ph.D.1, Mary O. Odafe, M.A.2, Judy H. Hong, B.A.1, David Talavera, MA1, Chris
Barr, Ph.D.1, Aanand Naik, MD3

1
University of Houston, Houston, TX; 2University of Houston, TX, Houston, TX; 3Baylor College
of Medicine, Houston, TX

African Americans are significantly more likely than European Americans to be diagnosed with
Type 2 Diabetes Mellitus (T2DM) and to experience severe diabetes-related complications;
this is true also when SES is taken into consideration (Tull & Roseman, 1995). Compared to
other ethnic groups, African Americans are significantly less likely to meet recommended
blood pressure and LDL cholesterol goals (Casagrande et al., 2014).

Compared to non-diabetic adults, adults diagnosed with T2DM are also twice as likely to be
depressed (Anderson et al., 2001; Nouwen et al., 2010). Importantly, higher levels of
depression symptomatology are associated with decreased self-care behavior and
considerably higher health care costs (Egede, Zheng, & Simpson,2002; Lustman et al., 2000).
Among African Americans, depression is more debilitating in part because African American
adults are among those least likely to seek clinical care for depression-related pathology
(Williams et al., 2007).

Though there have been some efforts to address the dual need for integrated depression and
diabetes care (Bogner et al., 2012), unique cultural factors associated with self-care among
African Americans diagnosed with T2DM are understudied. We hypothesize that having
reasons for living (RFL) buffers the association for depression symptomatology and self-care
among African American adults.

Data were collected for 211 community-based, African American men and women
(Mage=55.07, SD=8.79) who were prescribed medication or taking insulin for T2DM.
Participants completed the Summary of Diabetes Self-Care Activities measure (SDSCA) as well
as self-reported, well-validated measures of depression symptoms and reasons for living. The
majority of participants (n=118; 51%) reported having graduated from high school or having
some college or specialized training. Female participants were the majority (n=154; 73%) of
the sample.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1961

We found that higher ratings of depression symptoms were associated with poorer diabetes
self-care for the total sample of African American women and men [b=-.31, p=.010]. Follow-up
stratified analyses showed that, for women, there was a significant main effect for depression
[b=-.32, p=.047] and a depression X RFL interaction predicting self-care that approached
significance [b=.25, p=.074]. We followed up on this potential interaction effect by plotting
self-care and depression symptoms for women who reported high and low RFL. Based on that
plot, it appeared that depression was related to self-care for women who reported low [r=-
.57, p < .001] but not high [r=-.06, p=.625] RFL. In stratified analyses, RFL did not moderate
the depression - self-care association for African American men perhaps due to the limited
sample size.

Future studies should examine how cognitive bias can impact self-care behavior particularly
for African Americans who are diagnosed with T2DM.

CORRESPONDING AUTHOR: Rheeda Walker, Ph.D., University of Houston, Houston, TX, 77204;
[email protected]
S1962 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C102 6:00 PM-7:00 PM

EAT, MOVE, LIVE: A FEASIBILITY STUDY MEASURING BEHAVIORAL AND BIOLOGICAL

OUTCOMES IN MEDICALLY VULNERABLE

Mayra Serrano, MPH1, Marisela Huerta, CHES, BS 1, Kimlin Ashing, Ph.D2, Katty Nerio, BS 1

1
City of Hope, Duarte, CA; 2City of Hope National Medical Center, Duarte, CA

Background: Despite the tremendous need and enormous efforts to develop public health
recommendations for appropriate exercise levels, especially among families at risk for
diabetes, we still have no exact prescription to give the public. Obesity, unhealthy diet and
physical inactivity are linked to increased diabetes and cancer risk. Our Eat Move Live (EML)
program was implemented in collaboration with the City of Duarte and the Duarte Unified
School District (DUSD). The goal of the program focuses on diabetes and chronic disease
prevention through lifestyle, nutrition and physical activity. The program aimed to assess the
possible positive effects of a nutrition and exercise intervention on metabolic components
among underserved, Latino women at high-risk of diabetes.

Methods: The EML intervention is a series of weekly, culturally and community informed
nutrition and physically activity workshops delivered by health educators. The target
population is low-income Latina mothers who have elementary-aged children in the DUSD.
Demographically matched schools were assigned to the intervention (n=2) or wait-list control
(n=2). Participants were recruited through passive (fliers, home mailer) and active recruitment
strategies. The intervention approach was community engaged, and school-based. The
intervention design was a feasibility trial with a pre-test post-test, wait-list control.
Participants weight and A1c was measured at baseline and following the 3month
intervention.

Results: A total of 40 participants were enrolled in the study, 20 in the intervention and 20 in
the wait list control conditions; there was just 1 attrition from the control-group. All
participants reported elevated A1c and were at least over weight. As a result of participating
in the EML program, 69% of intervention group saw a significant reduction of A1c compared
to 28% control group (p

Conclusion: Our study shows that a community responsive approach implementing cultural
exercise, nutrition, shopping and food preparation knowledge and skills combined with social
support enhanced intervention acceptability and effectiveness. The results support our
hypotheses that increasing level of exercise to CDC recommendations and providing nutrition
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1963

and physical activity skill-building workshops can improve A1c and BMI status among Latinas
at high-risk of diabetes.

CORRESPONDING AUTHOR: Mayra Serrano, MPH, City of Hope, Duarte, CA, 91010;
[email protected]
S1964 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C103 6:00 PM-7:00 PM

EFFECT OF DOSE OF BEHAVIORAL OBESITY TREATMENT ON HBA1C IN ADULTS WITH

PREDIABETES

Vivian Bauman, B.S.1, Aviva H. Ariel-Donges, M.S.2, Eliza L. Gordon, M.S.2, Kathryn Ross, PhD,
MPh2, Michael Perri, PhD, ABPP2

1
Universtiy of Florida, Gainesville, FL; 2University of Florida, Gainesville, FL

Introduction: Standard behavioral weight-loss interventions typically produce an 8-10%

reduction in body weight and are associated with reduced diabetes risk. The optimal intensity
(i.e., dose) of behavioral treatment for achieving maximum glycemic improvements remains
unclear. Previous findings from the Rural LITE Trial indicated that moderate (16 sessions) and
high (24 sessions) doses of treatment produced significantly larger weight losses at 6 months
than low dose (8 sessions) treatment and a diet education control group (8 sessions). In this
study, we examined the effect of dose of behavioral treatment on changes in HbA1c (A1c) and
explored the mediating effect of weight loss.

Methods: The study included 315 adults with obesity (female=77%, Caucasian=78%; MSD
age=5511 years, BMI=36.23.9 kg/m2) who exhibited baseline A1c levels in the prediabetes
range (5.7-6.4%). Intervention content was the same for the high, moderate, and low dose
treatments. Participants in the control group received diet education without behavioral
strategies. A1c and weight were measured at baseline and month 6. The effect of treatment
dose on A1c was evaluated by a one-way ANOVA with Bonferroni adjusted post-hoc testing.
The mediating effect of weight loss on A1c change was evaluated using the Preacher and
Hayes model.

Results: There was a significant main effect of treatment dose on weight loss (p < .001) and
on A1c (p=.005). A1c reductions were significantly greater in the high (-.131%) vs. the control
(-.003%) conditions (p=.004). The association between treatment dose and A1c (p=.003)
became non-significant when weight change was added to the model (p=.493). The indirect
effect of dose through weight loss was significant (95% CI: .019-.043), suggesting full
mediation.

Discussion: We found that a high dose of behavioral weight-loss treatment was required to
achieve significant reductions in A1c compared to a diet education control group and that the
relationship between treatment dose and A1c was fully mediated by weight loss. While most
lifestyle interventions implement a moderate dose, these findings suggest that a higher dose
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1965

of treatment resulting in greater weight loss may be necessary to maximize improvements in

diabetes risk.

Support: R18 HL 087800

CORRESPONDING AUTHOR: Vivian Bauman, B.S., Universtiy of Florida, Gainesville, FL, 32607;
[email protected]
S1966 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C104 6:00 PM-7:00 PM

EXPLORATORY DEVELOPMENT OF A METHOD TO CLUSTER ADULTS WITH TYPE 2 DIABETES

BASED ON SELF-MANAGEMENT STYLE

Bryan Gibson, DPT, PhD1, Leah Yingllng, BS2, Yue Zhang, PhD1, Jorie Butler, PhD3

1
University of Utah, Salt Lake City, UT; 2University or Utah Department of Biomedical
Informatics, Salt Lake City, UT; 3University of Utah, SALT LAKE CITY, UT

Background: Adoption and engagement with Type-2 Diabetes (T2DM) self-management

interventions vary across individuals. This may be related to the degree to which the
interventions mode of delivery and functions match the individuals self-management style.

The goal of this exploratory study was to determine whether a combination of a modified q-
sort survey and k-means analysis could be used to develop a brief questionnaire that would
cluster adults with Type 2 Diabetes according to their self-management style.

Methods: We conducted a web-based survey among adults with T2DM. Participants sorted 31
statements using a modified q-sort methodology. The first two sets of statements addressed
intentions to change current self-management behaviors and individuals' desire for external
support. The next five sets of statements addressed self-management functions that
interventions often target: Teaching, Goal-setting and Planning, Feedback, Reminders, and
Social Support. Participants sorted each statement to represent the degree to which the
statement reflected their way of thinking on a bipolar scale (i.e., strongly agree to strongly
disagree). For example, in the set of statements related to social support, participants sorted
statements such as, I am more likely to attend a support group to meet other people with
diabetes; and I am more likely to join an online social network to meet other people with
diabetes.

A k-means cluster analysis was performed to extract homogenous groups for each self-
management function. The Big 5 Personality Traits and Regulatory Focus (collected at the start
of survey) were assessed as cluster discriminators.

Results: For each self-management function, the k-means algorithm divided the cohort
(n=1783, age= 5910 years; 64% female; mean Hemoglobin A1c=7.41.6%) into two clusters:
Techie and Old style." Individuals were varied in their style: of the 64 possible
permutations of Techie vs. Old style for each of 5 self-management functions, the ten most
frequent clusters only accounted for 55% (n=977) of the sample. Interestingly, age,
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1967

socioeconomic status, personality traits and regulatory focus did not differ significantly
between clusters.

Conclusions: A modified q-sort methodology combined with a k-mean clustering algorithm

appears to capture individual differences in T2DM self-management style. Future work to test
the resulting brief questionnaire will be discussed.

CORRESPONDING AUTHOR: Bryan Gibson, DPT, PhD, University of Utah, Salt Lake City, UT,
84148; [email protected]
S1968 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C105 6:00 PM-7:00 PM

SLEEP DURATION, POSTPRANDIAL METABOLIC FUNCTION, AND THE ROLE OF INSULIN

RESISTANCE IN NONDIABETIC INDIVIDUALS

Karin Garcia, M.S.1, William Wohlgemuth, Ph.D.2, Ele Ferrannini, M.D.3, Andrea Mari, Ph.D.4,
Alex Gonzalez, B.A.1, Theresa Dimascio, B.A.5, Armando Mendez, Ph.D.1, Ronald Goldberg,
M.D.1, Neil Schneiderman, Ph.D.1, Barry Hurwitz, Ph.D.1

1
University of Miami, Miami, FL; 2Miami VA Healthcare System, Miami, FL; 3University of Pisa,
Pisa, N/A, Italy; 4Institute of Neuroscience- National Research Council, Padua, N/A, Italy;
5
University of Miami, Coral Gables, FL

In preclinical diabetic dysmetabolism, there is an insulin secretion upregulation during

postprandial periods to compensate for diminished insulin sensitivity. It is not known whether
there is an association of shortened sleep duration with elevated postprandial insulin
secretion in nondiabetic persons, or whether this relationship is moderated by insulin
sensitivity. This study examined these interrelationships, while controlling for age, sex and
cardiovascular risk factors (smoking status, visceral adiposity volume, systolic blood pressure,
triglycerides, total cholesterol/HDL ratio), and postprandial metabolic indices (-cell glucose
sensitivity, early insulin secretion rate sensitivity, and insulin secretion potentiation). The 143
healthy, nondiabetic men and women, aged 18-55 years, had no diagnosed sleep, psychiatric
or cardiometabolic conditions. Habitual sleep function was assessed using actigraphy for 1
week. Insulin sensitivity was derived using a euglycemic hyperinsulinemic clamp. Postprandial
metabolic indices were derived using a 2-day in-patient evaluation, wherein serial blood
sampling was performed before and during 4 mixed-meal tests/day. Meals had identical fat
and protein content, but carbohydrate content was manipulated. Hence, in randomized order,
meals included standard carbohydrate load (300 kcal) on one day, and double carbohydrate
load (600 kcal) on the other day. Metabolic indices were derived using quantitative
decomposition modeling. Regression analyses assessed carbohydrate loading separately,
examining the prediction of total daily postprandial insulinemia by sleep duration (SDur),
insulin sensitivity (IS), and the interaction of these factors (SDur x IS). For the 300 kcal/meal
day, the analysis indicated a significant SDur x IS interaction (=0.149, p=.031) and a
significant main effect for IS (=-0.394, p < .001). For the 600 kcal/meal day, both the SDur x IS
interaction (=0.161, p=.034) and main effect for IS (=-0.486, p < .001) remained significant.
In sum, these findings show that the association of shorter sleep duration with more elevated
postprandial insulinemia is linked with poorer insulin sensitivity in otherwise healthy persons,
independent of age, sex, traditional cardiovascular risk factors and other metabolic measures.
Thus, preclinical diabetic risk, when poorer insulin sensitivity emerges, may not only include a
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1969

compensatory adaptation of postprandial insulin regulation, but also a decrease in sleep

duration.

CORRESPONDING AUTHOR: Karin Garcia, M.S., University of Miami, Miami, FL, 33183;
[email protected]
S1970 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C106 6:00 PM-7:00 PM

USING THE TRANSTHEORETICAL MODEL TO PROMOTE BEHAVIORAL CHANGE WITH DIABETES

SELF MANAGEMENT

Kerry Littlewood, Ph.D., MSW1, Robert Sherman, na2, Frances Russell, RN2, Kerry Littlewood,
Ph.D., MSW3, Abhishek Pandey, MD4

1
AAJ Reseaarch & Evaluation, Inc., Seminole, FL; 2St. Anthony's Hospital, St. Petersburg, FL;
3
AAJ Research and Evaluation, Seminole, FL; 4University of Arizona, Seminole, FL

Background: Success in chronic disease self management education programs require lifestyle
change and not all participants may find themselves ready or willing to modify their behavior.
Few studies have examined the relationship between insufficient sleep and motivation for
change in behavior for persons with type 2 diabetes mellitus.

Purpose: This study uses data from a Diabetes Self-Management Education Program to
explore the relationship between insufficient sleep and the participants stage of behavioral
change as identified by the transtheoretical model. This study will use descriptive and anova
analysis to examine these relationships.

Results: One hundred and one persons with T2DM (mean HbA1c = 8.79 2.42) participating in
a DSME Program at a southern urban community non profit hospital enrolled in this study.
The majority of these patients were obese (mean BMI=38.568.20) . Only 11% reported
normal sleep, with 41% reporting short sleep ( < 6 hours). Pre-contemplative stages of change
were related to more insufficient sleep (69%) [x2(3, N=97=8.67, p < .05].

Conclusions: Participants in a DSME program were more likely to be motivated to change their
behavior when they reported adequate sleep. Future studies should more closely examine
sleep quality and quantity to more discriminately articulate this relationship as it relates to
people desiring to better control their diabetes. Future interventions could integrate healthy
sleep hygiene education into existing DSME programs to improve participants motivation.
This could potentially improve the successful uptake of life style modifications for persons
with T2DM.

CORRESPONDING AUTHOR: Kerry Littlewood, Ph.D., MSW, AAJ Reseaarch & Evaluation, Inc.,
Seminole, FL, 33774; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1971

C107 6:00 PM-7:00 PM

MERITORIOUS AWARD WINNER

A DEFAULT OPTION TO ENHANCE DIETARY QUALITY IN PARTICIPANTS IN THE

SUPPLEMENTAL NUTRITION ASSISTANCE PROGRAM

Jaime Coffino, M.P.H., Julia M. Hormes, Ph.D.

University at Albany, SUNY, Albany, NY

Introduction: Diets consumed by participants in the Supplemental Nutrition Assistance

Program (SNAP) are of relatively lower quality, compared to eligible non-participants, leaving
them at increased risk for adverse health outcomes such as type II diabetes. The effectiveness
of psychoeducation or use of monetary incentivizes to enhance diets in SNAP participants
remains limited and cost can be prohibitive. In behavioral economics, the default option
refers to the option a consumer selects if s/he does not make an active choice (e.g., opting out
vs. opting in to organ donation programs). We sought to assess if use of a default option
would result in healthier food purchases within the constraints imposed by SNAP, compared
to monetary incentives and psychoeducation. Methods: Female undergraduates (n=53, mean
BMI=24.26, SD=5.12, age 20.24, SD=5.49) filled out a baseline questionnaire and were then
given a budget of $48.50 and instructed to purchase nutritious, affordable, and tasty food for
a week using the online shopping service of a local grocery store (T1). Before completing the
task again (T2) they were then randomized to: (1) receive a $10 incentive if they selected
groceries that met recommended nutritional guidelines, (2) read an educational brochure
adapted from materials currently utilized by SNAP, and (3) be presented with a default pre-
filled online shopping cart containing a nutritionally balanced selection of groceries and told
they could make changes as they wished. Results: There was a significant difference among
the 3 groups in average daily calorie consumption (F (2, 49 = 3.76, p = .03). Average daily
caloric consumption decreased in all three groups, but the default group had the largest
decrease and was significantly different from the education group. There was a significant
difference among the three groups in average daily sodium intake (F (2, 49) = 3.43, p = .04),
and average daily cholesterol intake (F (2, 49) = 3.92, p = .03). The default group showed
significantly greater decreases in calories and sodium compared to the education group and a
significantly greater decrease in cholesterol compared to the reinforcement group.
Discussion: Although the initial focus of SNAP was on providing sufficient quantities of food to
program participants, the importance of the quality of the food made accessible to benefit
recipients is thus becoming increasingly clear. SNAP currently uses education to promote
healthy eating even though it appears that education may not be very effective. This data
shows that research regarding default menus is necessary as it is the least costly intervention
S1972 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

option. Preliminary data suggests this option may be the most effective intervention strategy
in reducing unhealthy decisions surrounding food choices.

CORRESPONDING AUTHOR: Jaime Coffino, M.P.H., University at Albany, SUNY, Albany, NY,
12210; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1973

C108 6:00 PM-7:00 PM

A COMPARISON OF DIET-RELATED RISK FACTORS FOR CHRONIC DISEASE BETWEEN BLACK

AND WHITE MEN: FINDINGS FROM NHANES, 2007 - 2012

Terry Thompson, DHA, Chelesea Singleton, PhD, Sparkle Springfield, PhD, Angela
Odomsyoung, PhD

University of Illinois @ Chicago, Chicago, IL

Compared to men of different racial/ethnic backgrounds, African American men experience

higher rates of chronic disease such as cardiovascular disease and type 2 diabetes. Diet quality
has been shown to be a key risk factor for chronic disease. While gender differences in dietary
quality are well documented, the research literature is limited on evidence-based results
targeting dietary quality in men. Furthermore, literature on the diet quality of specifically
African American men is sparse. Using cross-sectional data from the National Health and
Nutrition Examination Survey (NHANES) years 2007-2012, this study aimed to examine
difference in diet quality, nutrient intake, and household food security status between African
American and White men. Data on 5,050 male participants (31.3% African American; 68.7%
White) over 20 years old were extracted and analyzed. Information on demographics and
household food security was self-reported by study participants. Information on dietary intake
was collected via 24-hour recall. Healthy Eating Index (HEI) 2010 total and component scores
were calculated for each participant. Multivariable adjusted linear regression models were
used to determine if African American race was associated with HEI-2010 total score.
Compared to White men, a higher percentage of African American men were obese according
to their body mass index (37.8% vs. 34.0%; p < 0.0001. A lower percentage of African
American men were food secure (67.1% vs. 80.6%; p < 0.0001. Mean dietary fiber intake
(18.6g ve. 14.7g; p < 0.0001) and sodium intake (4,126.3mg vs. 3,725.9mg; p < 0.0001) for
White men was significantly higher than African American men. Mean HEI-2010 total score
was significantly lower among African American men compared to White men (46.4 vs. 48.9; p
= 0.0003). After adjusting for demographics (e.g. age, education level, marital status) and
body mass index, African American race was found to not be associated with HEI-2010 total
score ( = 0.35; SE = 0.60; p = 0.57). In conclusion, while household food security may
significantly differ between Africian American and White men, diet quality was similar after
accounting for factors such as education level. Future research on this subject will highlight
the importance of developing interventions that target nutrition education among African
American men.
S1974 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Terry Thompson, DHA, University of Illinois @ Chicago, Chicago,

IL, 60615; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1975

C109 6:00 PM-7:00 PM

WEIGHING THE OPTIONS: ENHANCED SELF-AWARENESS OF BODY WEIGHT IMPACTS FOOD

CHOICE

Katie Hodgin, MPH, MS, Dan J. Graham, PhD

Colorado State University, Fort Collins, CO

Though enhanced self-awareness has been shown to reduce transgressive behaviors like
cheating, few studies have examined how self-awareness of a health indicator, body weight,
impacts food selection. The present study investigated how short-term self-awareness
enhancement of body weight influences the nutritional content of foods chosen and
hypothesized that those who weighed themselves prior to (vs. after) food selection would
choose foods lower in calories, saturated fat, and sugar. Two hundred and twenty students
participated in the study at a laboratory featuring a mock grocery aisle with a range of slow-
to-perish food items including fruit juice, cereals, chips, and pasta. Participants were
randomly assigned to weigh themselves on a calibrated scale either before or after making a
selection of four food items that they could have had the opportunity to take home with them
by winning a raffle drawing. After selecting their four food items, all participants completed
an electronic questionnaire assessing their demographic information and some basic
information about the food items chosen. T-tests revealed that participants chose foods
approximately 9% lower in calories (t(217) = -2.50, p = 0.013) and 18% lower in saturated fat
(t(217) = -2.61, p = 0.010) when they were asked to weigh themselves before rather than after
selecting their food items from the mock grocery aisle. Further results indicated that average
sugar content in the food items chosen did not significantly differ as a function of when
participants weighed themselves. Such findings are consistent with self-awareness theory,
and hold important intervention implications by suggesting that consumers who are reminded
of their body weight before a food choice is made may make more healthful decisions related
to that choice.

CORRESPONDING AUTHOR: Katie Hodgin, MPH, MS, Colorado State University, Fort Collins,
CO, 80528; [email protected]
S1976 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C110 6:00 PM-7:00 PM

APPLYING PROSPECT THEORY TO EATING: ENDOWED FOOD IS WORTH MORE

David Ede, Jr., BS1, Misty AW Hawkins, Ph.D.2, John Gunstad, PhD3, Joel W. Hughes, PhD3

1
Kent State University, RAVENNA, OH; 2Oklahoma State University, Stillwater, OK; 3Kent State
University, Kent, OH

Applying Prospect Theory to Eating: Endowed Food is Worth More

David Ede, BS1, Misty A.W. Hawkins, PhD2,John Gunstad, PhD1, Joel W. Hughes, PhD1

1
Department of Psychological Sciences, Kent State University, Kent, OH

2
Department of Psychology, Oklahoma State University, Stillwater, OK

Introduction

Dual process theories of decision making, such as Prospect Theory, have rarely been used to
study eating behaviors. The endowment effect from Prospect Theory is evidenced when
people place a higher value on something if they own it compared to if they do not own it,
reflecting loss-aversion. The endowment effect has been demonstrated with non-food objects
like mugs and pens, but it has not been applied to food items. Risk-sensitive Foraging Theory
suggests that foods with greater energy density are preferred to foods with lower energy
density. Deriving from Prospect Theory and Risk-sensitive Foraging Theory, we hypothesized
that participants would rate owned / endowed snacks as higher in value, and that this effect
would be stronger for energy dense foods.

Method

The present study comprised four conditions based on two factors: whether the presented
food was endowed to the participant or not (ownership), and whether the food had high
energy density (candy) or low energy density (vegetables). A total of 59 undergraduates
(age=19.5; 72% women) were randomized to evaluate either a pint of candy or vegetables in a
clear container, which were of equal retail cost. They were told they either get to keep or
dont get to keep the snack when they were done with the evaluation. Participants rated
how much they believed the food to be worth and how much they would personally pay for
their food item. ANOVA was used to examine the main effects and interaction of endowment
and energy density on perceived food value, using body mass index as a covariate.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1977

Results

Endowed items were valued more highly than non-endowed items, F(1,54)=4.188, p=.046.
Participants ranked the vegetables as more valuable than the candy, F(1,54)=5.668, p=.021.
Endowment and food type did not interact, F(1,54)=.027, p=.870.

Discussion

Our results are consistent with the endowment effect, such that participants placed a greater
value on endowed food items. In contrast to our hypothesis, vegetables were consistently
rated as more valuable than candy. Dual process theories such as prospect theory can be
applied to eating behaviors, which may yield insights relevant to the obesity epidemic.

CORRESPONDING AUTHOR: David Ede, Jr., BS, Kent State University, RAVENNA, OH, 44266;
[email protected]
S1978 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C111 6:00 PM-7:00 PM

ATTENTIONAL MECHANISMS UNDERLYING LEARNING TO READ NUTRITION LABELS

Lisa S. Miller, PhD1, Tanja N. Gibson, MS1, Kara DeSouza, PhD2

1
University of California, Davis, Davis, CA; 2UC Davis, American River College, Sacramento, CA

Individuals who use nutrition labels tend to eat healthier diets (Ollberding et al., 2010), still,
many individuals avoid using labels because they are too difficult or time consuming to read
(Guthrie et al., 2015). Eye tracking methodology has been used to shed light on the
attentional mechanisms underlying nutrition label comprehension (e.g., Bialkova et al., 2014;
Miller et al., 2015), however, relatively less work has focused on eye movements while
learning to read nutrition labels. In the present study, we examined changes in attention as
individuals completed a nutrition label reading training tool.

40 college students (75% female) participated in the study. The training began with a 7-
minute slide presentation on general nutrition information, followed by a label reading task
requiring participants to select the more healthful product using two nutrition labels,
presented side-by-side on a computer screen. Participants used a mouse to select their
choice, and were told if their choice was correct or incorrect before moving on to the next
comparison. There were 24 comparisons in each of 3 blocks of practice. Eye movements were
monitored using a video-based eye tracker to determine if training is accompanied by a
reduced number of sweeps between choices prior to making a decision, reflecting greater
efficiency. Participants provided usability ratings on the training tool at the end of the session.

Individuals improved their accuracy across practice blocks, t(35)=7.1, p < .001, with the
greatest improvement from block 1 to 2, t(35)=5.8, p < .001, but significant improvement also
evidence from block 2 to 3, t(35)=2.2, p < .05. Importantly, eye tracking data showed
significant decreases across the 3 blocks in the number of sweeps between the labels for both
the to-be-limited nutrients (e.g., saturated fat), t(35)=3.4, p < .01, and the non-limited
nutrients (e.g., protein), t(35)=4.5, p < .001. Usability ratings indicated that participants
generally agreed that the tool was very easy or easy to use (76%), very useful or useful (77%),
and was fun (72%). Participants also indicated that it was very likely or likely (88%) that they
would continue to improve their skills if they continued practicing. In general, label training
can increase accuracy and reduce attentional demands associated with label reading.
Additional research is needed to determine whether increased efficiency can translate into
sustained, healthier food choices.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1979

CORRESPONDING AUTHOR: Lisa S. Miller, PhD, University of California, Davis, Davis, CA,
95616; [email protected]
S1980 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C112 6:00 PM-7:00 PM

DEVELOPMENT OF LA VIDA INC, A MOBILE-ENABLED INTERACTIVE COMIC TO PROMOTE

HEALTHY EATING AMONG LOW-INCOME HISPANIC YOUTH

Katrina F. Mateo, MPH1, Jennifer R. Furman, MS2, Olivia Barata-Cavalcanti, MPH, MIA 3, May
May Leung, PhD, RD4

1
City University of New York, Graduate School of Public Health and Health Policy, New York,
NY; 2City University of New York, Hunter College Nutrition Program, Brooklyn, NY; 3City
University of New York, Graduate School of Public Health and Health Policy, Brooklyn, NY;
4
City University of New York, Hunter College Nutrition Program, New York, NY

Background: Childhood obesity continues to be a serious clinical and public health issue in the
US, with increased prevalence particularly among low-income Hispanics. Effective, yet
innovative interventions are needed to capture the attention of children in a technology and
media-dense environment. Digital health interventions have potential to reach and engage
with culturally diverse and hard-to-reach communities. We aimed to: 1) identify factors
influencing child dietary behaviors, 2) assess technology use, and 3) identify preferred
character profiles based on popular youth media.

Methods: Hispanic parent-child dyads were recruited using convenience sampling in East
Harlem, New York City. Participants completed a survey and participated in a semi-structured
interview. Question domains included: knowledge, attitudes and practices around
eating/health and technology, popular media/storylines, preferred character profiles, and
possible intervention components. Audio-recordings were transcribed and translated, and
analyzed by two coders using Dedoose software.

Results: Thirteen dyad and two triad interviews were conducted (parent: n=15, mean age 37.4
5.1; child: n=17, mean age 10.4 1.3). Most child participants were female, self-identified
Mexican and bilingual; all were US-born. All reported their mother being primarily responsible
for food shopping and preparation. Family food-related practices, culture, and food
taste/texture strongly impacted child eating behaviors and food preferences. Nearly all
children and parents identified eating fruits/vegetables and drinking water as the most
important habits/behaviors for being healthy. All children used mobile apps regularly mainly
for homework, games, and watching videos. Children preferred characters who evolve over
time to develop traits or powers to help others, fight evil, or solve problems. Children
expressed willingness to receive information to help make healthy choices and understand
the benefits of different foods through interactive gameplay. They preferred built-in
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1981

motivation (e.g. unlocking game elements) as opposed to external/monetary incentives to

encourage continual health tool use.

Conclusions: Findings will inform the design of a mobile-enabled interactive nutrition comic
prototype that will be used in a 2-group pilot randomized study targeting behaviors to reduce
childhood obesity risk. Our user-centered approach will help develop culturally-tailored
material and may increase potential intervention adoption, engagement, and sustained use.

CORRESPONDING AUTHOR: Katrina F. Mateo, MPH, City University of New York, Graduate
School of Public Health and Health Policy, New York, NY, 10019;
[email protected]
S1982 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C113 6:00 PM-7:00 PM

DIETARY INADEQUACY AMONG SMOKELESS TOBACCO USER HOUSEHOLDS IN BANGLADESH

Mandeep Virk-Baker, PhD, MPH, MSc, RDN1, Muhammad J. Husain, PhD2, Mark Parascandola,
PhD, MPH3

1
NCI/NIH, Clovis, CA; 2CDC, Atlanta, GA; 3NCI/NIH, Rockville, MD

Tobacco users on average have less adequate diet as compared to non-users. While the
majority of the studies are from developed countries, less is known about the influence of
tobacco on dietary composition in developing countries where malnutrition is a major public
health challenge. Additionally, the effect of smokeless-tobacco use on dietary composition are
unknown. We present a household-level analyses that compares diet using the nationally
representative Household Income Expenditure Survey (HIES-2010) from Bangladesh. Overall,
71% of the households reported expenditure on tobacco (smoking and/or smokeless), and
were considered tobacco users. Out of 12240 households, 2061 used smoking tobacco only
(16.8%), 3284 used smokeless tobacco only (26.8%), and 3348 were dual-users (27.4%). After
controlling for household expenditure, household size, place of residence, and education,
smokeless-tobacco user households consumed significantly lower mean per capita daily total
calories ( = - 342.88; p < 0.0001) as compared to non-users. Dietary analyses revealed
smokeless-tobacco users consumed significantly lower daily mean per capita of vegetables (
= -19.65 g/day; p 0.0001), milk and dairy ( = -9.81 g/day; p 0.0001), fish ( = -9.84 g/day;
p 0.0001), meat ( = -10.9360 g/day; p 0.0001), legumes ( = -3.23 g/day; p 0.0001), eggs
( = -1.60 g/day; p 0.0001) as compared to non-users. However, mean per capita daily
intakes of cereal products ( = 39.26 g/day; p 0.0001) was significantly higher among
smokeless-tobacco users as compared to non-users. Corresponding to these profound dietary
differences, the intake of total dietary protein ( = - 10.01 g/day; p 0.0001), dietary fat ( = -
27.55 g/day; p 0.0001) were significantly lower, and dietary carbohydrate ( = 94.32 g/day;
p 0.0001) was significantly higher among smokeless-tobacco users as compared to non-
users. The study provides evidence to inform policy for addressing dietary inadequacy and
malnutrition burden among smokeless-tobacco user households in Bangladesh.

CORRESPONDING AUTHOR: Mandeep Virk-Baker, PhD, MPH, MSc, RDN, NCI/NIH, Clovis, CA,
93619; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1983

C114 6:00 PM-7:00 PM

EATING FREQUENCY AND CHRONIC DISEASE RISK AMONG POSTMENOPAUSAL WOMEN

Cynthia Thomson, PhD, RD1, Betsy Wertheim, MS; 1, Melanie D. Hingle, PhD, MPH, RD1, Tracy
Crane, BS: MS; PhD1, Marian Neuhouser, BS; PhD2

1
University of Arizona, Tucson, AZ; 2Fred Hutchinson Cancer Research Center, Seattle, WA

Background/ Objective: Eating frequency, defined as eating events at least 30 minutes apart,
has been associated with risk for chronic disease including obesity, CVD, and cancer. However,
the findings are sparse and inconsistent. Efforts to expand on these results in order to develop
guidance in regard to this modifiable behavior are warranted.

Methods: 24-hour dietary recall data collected at year 3 of study from a sub-sample of
postmenopausal women (age 50-80 y) enrolled in the Womens Health Initiative Diet
Modification Trial (n=2460) provided the eating frequency exposure data for this analysis.
Adjudicated disease outcomes included incident cardiovascular disease (stroke and coronary
heart disease combined), myocardial infarction, and cancer; self-reported diabetes was also
evaluated. Risk for disease was assessed using Cox proportional hazards regression models,
adjusted for potential confounders.

Results: Most women reported 3 (31%), 4 (33%), or 5 (20%) eating events over a 24-hour
period. Higher eating frequency was reported in younger women, those with higher
neighborhood socioeconomic status, non-Hispanic whites, and women with lower body mass
index. Eating frequency was not associated with a higher risk for cardiovascular disease,
myocardial infarction, or cancer. A significant association between eating frequency and
diabetes was demonstrated such that women reporting eating 4 times per day showed 36%
greater risk for developing diabetes than women reporting eating 13 times per day (HR, 1.36;
95% CI, 1.041.79); however, there was no difference in risk between women eating 13
versus 5 times per day.

Conclusion: An inverse U-shaped association between eating frequency and diabetes was
demonstrated in this cohort of postmenopausal women; relationships between eating
frequency and other obesity-related chronic diseases of aging were not found.

CORRESPONDING AUTHOR: Cynthia Thomson, PhD, RD, University of Arizona, Tucson, AZ,
85714; [email protected]
S1984 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C115 6:00 PM-7:00 PM

EFFECTS OF HEALTHY-EATER SELF-SCHEMA AND NUTRITION LITERACY ON HEALTHY-EATING

BEHAVIORS

Chia-Kuei Lee, PhD1, Li-Ling Liao, PhD2

1
University of Rochester, Rochester, NY; 2I-Shou University, Taiwan, Kaohsiung City, N/A,
Taiwan (Republic of China)

Unhealthy-eating behaviors contribute to obesity and chronic illness. Researchers have

increasingly recognized the role of healthy-eater self-schema (reflects the healthy-eating as an
important aspect of the self) and nutrition literacy (reflects the competence of healthy-eating)
in healthy-eating behaviors. However, the relative contribution of healthy-eater self-schema
and nutrition literacy is unknown. This study aimed to determine the unique effects of
healthy-eater self-schema and nutrition literacy on healthy-eating behaviors in undergraduate
students

Participants were 1,216 students from three national and three private universities in Taiwan.
Healthy-eating behaviors were measured with a 13-item frequency rating scale. Healthy-eater
self-schema was measured by rating three key phrases healthy eater, someone who eats in
a nutritious manner, and someone who is careful about what I eat according to self-
descriptiveness (1-11) and importance (1-11). Ratings of at least two of the three descriptors
as both very descriptive (811) and very important (811) were considered evidence of a
healthy-eater self-schema. Nutrition literacy was measured with an 8-item scale. Known
determinants of eating behaviors (nutrition-related information, health status, nutrition
knowledge needs, gender, and residence) were covariates. Hierarchical multiple regression
was conducted.

Results showed that healthy-eater self-schema ( = 0.27, p < .001) and nutrition literacy ( =
0.37, p < .001) were significant predictors of healthy-eating behaviors. A healthy-eater self-
schema explained 9% of the variance in healthy-eating behaviors and nutrition literacy
explained additional 12% of the variance in healthy-eating behaviors.

Findings suggest that a healthy-eater self-schema and nutrition literacy have additive effects
on healthy-eating behaviors. Nutrition literacy has relatively stronger effect on healthy-eating
behaviors than a healthy-eater self-schema. Intervention to improve healthy-eating behaviors
may be more effective if they incorporate the enhancement of both personal value of healthy
eating and nutrition literacy at the same time.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1985

CORRESPONDING AUTHOR: Chia-Kuei Lee, PhD, University of Rochester, Rochester, NY,

14642; [email protected]
S1986 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C116 6:00 PM-7:00 PM

EMPOWERING INDIVIDUALS TO MAKE HEALTHY FOOD CHOICES THROUGH NUTRITION LABEL

READING SKILLS

Lisa S. Miller, PhD1, Jacqueline Bergman, PhD2, Tanja N. Gibson, MS1, Kara DeSouza, PhD3

1
University of California, Davis, Davis, CA; 2Nutrition Department at University of California,
Davis, Davis, CA; 3UC Davis, American River College, Sacramento, CA

Nutrition labels are an important tool for navigating healthful food choices. However, labels
tend to be underutilized, partly because they are difficult to interpret (Cowburn & Stockley,
2005). Research has shown that label reading skills can be improved (Jay et al., 2009),
however, it is unclear whether skill improvement is accompanied by changes in attitudes
toward healthy food choices (Dukeshire et al., 2014). That is, do consumers feel more
empowered to make healthy food choices after nutrition label training? We examined
changes in self-perceptions and motivation before and after using a web-based nutrition label
reading training tool.

Participants were 44 young adults (ages 18-29, 59% female, with an average of 18 years of
education). The training tool included a 20-minute nutrition overview followed by 3 blocks of
intense self-paced label reading practice with feedback. For the practice task, participants
compared two nutrition labels, presented side-by-side on a computer screen, to determine
which was more healthful. Feedback was provided after each comparison, as well as at the
end of each of the 3 blocks of comparisons (24 in each block). Attitudes toward healthy food
choices were assessed using the following 4 items: My ability/willingness to select healthful
foods when Im offered many, many choices (large grocery store, online shopping, restaurant)
is ___: and My ability/willingness to interpret nutrition information on food labels in order to
select healthful foods is ___: Participants used a slider bar to select a value between 0 (very
low) and 100 (very high).

Results showed significant increases in objectively assessed label reading skills from the first
to last practice block, t(43)=2.4, p < .05, Moreover, we found improvements in self-
perceptions of both ability and willingness to use food labels and to select healthy foods, from
pre- to post- training, p < .05, for all four t-tests. These data are consistent with the notion
that label reading skills empower individuals to make healthy food choices. Future work in this
area is needed to determine if label reading skills and feelings of empowerment can be
sustained with additional training and can translate into more healthful food choices in the
long run.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1987

CORRESPONDING AUTHOR: Lisa S. Miller, PhD, University of California, Davis, Davis, CA,
95616; [email protected]
S1988 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C117 6:00 PM-7:00 PM

HEALTHY DIETS IN THE AFRICAN AMERICAN COMMUNITY: APPLYING THE THEORY OF

PLANNED BEHAVIOR IN FAITH-BASED SETTINGS

Alexandria G. Booker, M.A.1, Jannette Berkley-Patton, PhD2, Kymberley K. Bennett, Ph.D.3,

Delwyn Catley, Ph.D.4, Carole Bowe Thompson, BS5, Sheila Lister, B.S.3, Kelsey N. Christensen,
M.A. 6

1
University of Missouri-Kansas City, Raytown, MO; 2University of Missouri-Kansas City, Kansas
City, MS; 3University of Missouri-Kansas City, Kansas City, MO; 4Children's Mercy Hospital &
University of Missouri-Kansas City, Kansas City, MO; 5Community Health Ressearch
Group/School of Medicine/University of Missouri-Kansas City, Kansas City, MO; 6University of
Missouri - Kansas City, Kansas City, MO

Diabetes and heart disease are two major health concerns for African Americans, who tend to
have worse nutritional intake than Whites. Black churches, which serve primarily African
Americans, are influential institutions in the African American community that can assist with
promotion of healthy eating behaviors. Although church involvement has been associated
with healthier dietary beliefs and behaviors, church settings can also impede healthy eating
due to social pressure to consume unhealthy foods. To contribute to better design of church-
based dietary intervention studies, there is a need to understand psychosocial influences,
particularly beliefs and intentions to eat a healthy diet. Yet, there is a gap in the literature on
studies that have examined this key predictor of dietary behaviors among church-affiliated
African American populations. The Theory of Planned Behavior (TPB) is an established theory
that has been demonstrated to explain the relationships between behavioral beliefs,
normative beliefs, control beliefs, intentions to perform a health behavior and the behavior
itself. The purpose of the current study is to examine intentions to eat a healthy diet and
dietary behaviors (i.e., F/V and fat intake) in a sample of African-American church and
affiliated community members using the TBP. Participants (N = 352) were recruited from six
Black churches in the Kansas City metropolitan area and completed surveys exploring TPB
constructs and health behaviors. Participants were primarily female (68%), with an average
age of 54 years. Results showed that path models testing fat intake and F/V intake had good
overall fit (RMSEA = .058, PCLOSE = .334; and RMSEA = < .001, PCLOSE = .963, respectively).
Findings indicated behavioral ( = .240, p < .001) and normative beliefs ( = .171, p < .05)
predicted intentions to eat a healthy diet, even when controlling for covariates (i.e., age, sex,
BMI, religiosity, baseline fat or F/V intake). Control beliefs predicted fat intake ( = -.011, p <
.05). Intentions predicted fat ( = -.217, p < .05) and F/V intake ( = .313, p < .001). These
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1989

findings provide support for the use of the TPB in examining dietary behaviors among church-
affiliated African Americans and suggest church-based healthy diet interventions should
emphasize promotion of behavioral and normative beliefs. This study represents an
opportunity to understand dietary beliefs and behaviors in the African-American faith
community, with potential to inform dietary interventions in key African American faith-based
settings.

CORRESPONDING AUTHOR: Alexandria G. Booker, M.A., University of Missouri-Kansas City,

Raytown, MO, 64138; [email protected]
S1990 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C118 6:00 PM-7:00 PM

IMPACT OF FRONT OF PACKAGE NUTRITION LABEL TYPE AND EXPLANATION ON PERCEPTION

OF PRODUCT HEALTHINESS AND FOOD CHOICE

Pamela Lundeberg, M.A., Dan J. Graham, PhD

Colorado State University, Fort Collins, CO

Background: Front-of-package (FOP) nutrition labels are increasingly used to present

nutritional information to consumers. A variety of FOP nutrition schemes exist, each of which
presents condensed nutritional profiles in a different manner. The present study was designed
to directly compare two policy-relevant symbolic FOP labeling systems traffic light and star-
based labeling schemes with specific regard to healthiness perception and purchase
intention for a variety of food products. Additionally, we sought to determine which method
of message framing (gain, loss, gain+loss) would best enable customers to effectively utilize
the FOP labeling schemes.

Method: Participants (n = 306) viewed food packages featuring either the star or traffic light
FOP labels and rated both the healthiness of each product and their likelihood of purchasing
the product. Within each label type, participants were presented with differing instructions
regarding how to make use of the labels (i.e., differently-framed instructions).

Results: Participants who viewed the star labels rated products with the lowest healthiness as
significantly less healthy and rated products with the highest healthiness as significantly
healthier compared to participants who viewed those same products with traffic light labels.
Purchase intention did not differ by label type. Additionally, including any type of framing
(gain, loss, or gain+loss) of the labeling systems assisted consumers in differentiating between
foods with mid-range nutritional value and foods with little nutritional value.

Conclusions: In our study, the star-based labeling system led to healthy foods being seen as
even healthier and to unhealthy foods being seen as even less healthy compared to the same
foods with traffic light labels. Additionally, our results indicate a benefit of including framing
information for FOP nutrition label use instructions, however our study did not provide clear
support for a particular method of framing. While ratings of product healthiness were
influenced by the framing and the label type, purchase intention was not impacted by either
of these factors.

CORRESPONDING AUTHOR: Pamela Lundeberg, M.A., Colorado State University, Fort Collins,
CO, 80525; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1991

C119 6:00 PM-7:00 PM

IMPULSIVITY AND ITS ASSOCIATION WITH FOOD PURCHASING BEHAVIORS AND THE HOME
FOOD ENVIRONMENT IN SOUTHEAST ALABAMA

Kimberly Garza, PharmD, MBA, PhD1, Christiana Datubo Brown, MS2, Philippe Gaillard, PhD3,
Ruth Jeminiwa, BPharm1

1
Auburn University Harrison School of Pharmacy, Auburn, AL; 2Abuurn University Human
Development and Family Studies, Auburn, AL; 3Auburn University Mathematics and Statistics,
Auburn, AL

Background: Food purchasing behaviors such as the emphasis on certain food characteristics
and the use of food labels while grocery shopping have been associated with purchasing
decisions. Home food environment has also been shown to influence dietary behaviors, which
in turn is associated with level of impulsivity. However, little is known about the relationship
between impulsivity, food purchasing behaviors, and healthfulness of the home food
environment.
Objective: To determine the relationships between impulsivity, food purchasing behaviors,
and home food environment among employed adults.
Methods: A cross-sectional, online survey of employees of a southeastern university was
conducted with participants recruited through mass electronic mailing. Main Measures:
Impulsivity, consideration of food quality (nutrition, taste) and grocery characteristics (price,
ease of preparation, shelf life) while shopping (5 items), use of food labels (2 items), and
home food environment (4 items). Impulsivity is assessed by the area under the delay
discounting curve (AUC), which is estimated using a binary choice delay discounting task
incorporating hypothetical monetary rewards. Greater AUC reflects lower impulsivity.
Structural equation modeling (SEM) was used to allow impulsivity to simultaneously predict
purchasing behaviors and home food environment. SEM were conducted using Mplus.
Results: A total of 477 participants completed the online survey. The sample population were
mostly female (n = 329), married (n = 332), and held at least an associate's degree (n = 407).
More than half (55%) of participants reported a household income of $70,000 or more a year.
AUC was positively associated with food label use and emphasis on nutrition and taste when
grocery shopping (p < 0.001), meaning that individuals who were less impulsive were more
likely to use food labels and placed greater emphasis on these factors. In the final model, AUC
was shown to be positively associated with healthful foods in the home and emphasis on food
quality and grocery characteristics when purchasing foods. About 33% of the variance in
consideration of food characteristics, 5% of the variance in food label use, 7% of the variance
in availability of healthful foods, and 4% of the variance in the availability of unhealthful foods
S1992 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

was explained by impulsivity.

Conclusions: Individuals with low impulsivity appear to be more thoughtful in making food
purchasing decisions and perhaps as a result, have healthful foods available in the home more
frequently compared to individuals with higher levels of impulsivity.

CORRESPONDING AUTHOR: Kimberly Garza, PharmD, MBA, PhD, Auburn University Harrison
School of Pharmacy, Auburn, AL, 36849; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1993

C120 6:00 PM-7:00 PM

SNACK IT UP FOR PARENTS: A PILOT TRIAL TO IMPROVE CHILDRENS SNACKS

Sara Folta, PhD1, Namibia Lebron-Torres, MNSP2, Stephanie Anzman-Frasca, PhD3, Megan
Halmo, MPH, MSW4, Heather Angstrom, MS4, Alyssa Koomas, MPH, MS, RD4, Sarah Johnson,
MS2, Nicole Smith, MS2, Shenjia Shi, BS2, Delphine Van Roosebeke, MS2, Sean Cash, PhD2

1
Tufts University, Boston, MA; 2Friedman School of Nutrition Science and Policy, Tufts
University, Boston, MA; 3University at Buffalo, Buffalo, NY; 4ChildObesity180, Friedman School
of Nutrition Science and Policy, Tufts University, Boston, MA

Purpose. Nearly one-quarter of childrens daily energy intake comes from snacks, typically
energy-dense and nutrient-poor sweet and salty foods. The objective of the Snack It Up for
Parents study was to test the feasibility and preliminary effectiveness of a parent intervention
to improve snacks for children. Intervention strategies were designed to address the
frequently-cited barriers of time and cost.

Methods. In this quasi-experimental, 7-week pilot study, a 3-arm intervention design was
used. Four parent-child dyads received a $5 per week per child incentive for a local grocery
store; seven received a brief nutrition education video (viewable on any device), with
accompanying tip sheets with information on providing palatable snacks at low-cost and with
minimal time; and five received both. Parents of children ages 8-12 were included in the study
and were recruited based on enrollment of their child in the American SCORES Soccer for
Success program for under-resourced youth. Outcomes included psychosocial factors
(measured by survey) and snack quality (measured using web-based 24-hour recalls, ASA24).
Key informant interviews with parents were also conducted to assess perceptions of change in
childrens snacking habits, parents shopping habits, and experiences with the intervention
activities.

Results. Across study arms, there was no increase in vegetable snacks from pre to post.
Consumption of fruit as a snack increased in the video and incentive arms (p < 0.05). Only the
video arm showed improvement in self-efficacy for providing healthier snacks (p=0.003). In
key informant interviews, parents described an increase in the variety of snacks provided.
Other themes were that the intervention activities provided an opportunity to evaluate
household eating habits, consider healthier options, and experiment with new ideas for
snacks. However, some parents identified on-going barriers to providing healthy snacks
including limited income, food waste, food prices, and time constraints.
S1994 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Discussion. Results suggest promise of these low-burden intervention strategies to reach

parents to improve snacking in children, although barriers of time and cost remained an issue
for some parents.

CORRESPONDING AUTHOR: Sara Folta, PhD, Tufts University, Boston, MA, 02111;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1995

C121 6:00 PM-7:00 PM

THE EFFECT OF ADHERENCE TO DIETARY TRACKING ON WEIGHT LOSS: USING HLM TO MODEL
WEIGHT LOSS OVER TIME

J. Spencer Ingels, M.A., Jay Stewart, M.A., Samantha Shawley, MS, Brandon Lucke-Wold, MS,
Ranjita Misra, PhD

West Virginia University, Morgantown, WV

The role of dietary tracking in weight loss has received little research attention despite being
part of multiple diabetes and weight management programs. The Diabetes Prevention and
Management (DPM) program was a 22-session, yearlong program that asked participants to
track what they ate for the whole program. Participants were provided booklets to record
food and beverage intake, and a Calorie King book was given to look up fat, carbohydrate, and
fat content. A participant was considered to have tracked if they wrote down at least one food
or beverage item for the day. This minimal requirement for dietary tracking provided data to
explore the impact of a bare minimum effort to track dietary intake. A scatterplot of days
tracked versus total weight loss revealed a non-linear relationship, to fit this relationship the
total number of possible days to track was divided by 3, creating the following 3-groups: rare
trackers ( < 114 total days tracked), inconsistent trackers (114-228 total days tracked) and
consistent trackers (>228 total days tracked). The tracking groups were split and dummy
coded with the rare tracking group as the reference group. Multiple models were built to first
best fit the weight loss data and then to determine the best controls control variables. A cubic
model provided the best fit for weight loss change over time at level-1. At level-2, a model
with initial BMI and A1c, gender, and week-1 weight loss was the best fit for the data,
explaining 5.29% of within person variance and 6.26% of between person variance in total
weight loss. Adding tracking further improved the model, explaining an additional 2.15% of
within person variance and 15.31% of between person variance in total weight loss. After
controlling for initial BMI, initial diabetes status, week-1 weight loss, and gender only the
group that consistently tracked lost a statistically significant amount of weight (-10.58
pounds), the weight change for rare (-2.99 pounds) and inconsistent trackers (1.98 pounds)
did not statistically differ from zero. None of the control variables were significant predictors
of total weight loss. In addition, the weight loss trend for the rare and inconsistent trackers
followed a non-linear path, with the holidays slowing weight loss and the onset of summer
increasing weight loss. However, the consistent trackers weight loss trend more closely
followed a linear relationship, with consistent loss throughout the year. These results show
the importance of dietary tracking in long-term weight loss success. The implications and
limitations of these results will be discussed.
S1996 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: J. Spencer Ingels, M.A., West Virginia University, Morgantown,

WV, 26505; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1997

C122 6:00 PM-7:00 PM

EXAMINING MOBILE IMAGING, PERVASIVE SENSING, SOCIAL MEDIA AND LOCATION

TRACKING (MISST) RESEARCH USING NIH REPORTER DATABASE

Sarah Dunseath, B.S.1, Camille Nebeker, Ph.D.2

1
UC San Diego School of Medicine, Coronado, CA; 2University of California - San Diego, La
Jolla, CA

Background: Researchers are using Mobile Imaging, pervasive Sensing, Social-media and
location Tracking (MISST)technologies to intervene with and/or observe personal health
behaviors. Studies using MISST methods/tools raise new challenges for Institutional Review
Boards (IRBs) charged with protecting research participants. This study explores the MISST
landscape by identifying the nature and scope of studies supported by the National Institutes
of Health (NIH).

Objectives: Due to the novelty of MISST use in research, we sought to: 1- examine the extent
to which the NIH is supporting this research and, 2- identify how these tools are being used in
currently funded research. A high level aim is to conduct a systematic evaluation of how IRBs
carry out their obligations to protect participants and facilitate high quality scientific research.

Methods: MISST research supported by the NIH and its agencies and centers were identified
through a search of the NIH Research Portfolio Online Reporting Tools (RePORTER) database
using 43 key words and phrases during 2005, 2010 and 2015. Relevant abstracts were coded
by the technology used, theme of the research, funding institute/center, amount of funding,
and funding recipient.

Results: Results reveal MISST use in research is relatively novel as funded projects using these
devices accounted for only approximately 1% of the total NIH budget in 2015. However, the
number of awards allocated to MISST research increased 384% from 2005 to 2015.
Additionally, the types of MISST devices used in research has shifted from geographic
information systems (GIS) being the most widely used tool in 2005 to mobile
devices/application being the most widely used tools in 2015. The number of institutes,
agencies, and centers supporting MISST research has also increased by roughly 50% and the
scope of MISST research has widened, with its increasing use in diverse fields ranging from
tracking disease transmission via analysis of social media content to personalizing medicine
through mobile health applications.
S1998 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions: Evolving research practices and methods enabled by technological advances are
rapidly changing how research is conducted. This study sheds light on studies using MISSTE
technologies and the institutes supporting this research.

CORRESPONDING AUTHOR: Sarah Dunseath, B.S., UC San Diego School of Medicine,

Coronado, CA, 92118; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S1999

C123 6:00 PM-7:00 PM

ANALYSIS OF A MOBILE HEALTH APPLICATION TO INCREASE WILLINGNESS TO ASK FOR A

LIVING KIDNEY DONATION IN BLACK RENAL PATIENTS

John C. Sieverdes, PhD1, Lynne S. Nemeth, PhD, RN, FAAN1, Frank A. Treiber, PhD2, Brenda M.
Brunner-Jackson, MPH1, Sachin K. Patel, MSc1, Shawntel M. Parker, BS1, Prabhakar K. Baliga,
MD, FACS1

1
Medical University of South Carolina, Charleston, SC; 2Medical Univ of South Carolina,
charleston, SC

Rates of living donor kidney transplantation (LDKT) are greater in White renal patients while
the burden of renal disease disproportionately affects the Black population. The Living Organ
Video Educated Donors program (LOVED) was designed to address this disparity using a
distanced-based education platform delivered on 10 tablet computers to aid Black renal
patients who had difficulty approaching others to ask for a living kidney donation (LKD). The 8-
week culturally tailored mobile health (mHealth) education program consists of short weekly
video education modules and online video chat sessions led by a Black former patient
navigator who has had a LDKT. A parallel convergent mixed method approach was used to
triangulate quantitative data from pre-post surveys and qualitative data from end of trial
focus group transcriptions. Four LOVED groups (n=25, 11 men; mean age = 48.6 years
(SD14.6)) using 2 navigators were conducted between 2014 and 2016. Analysis showed an
increase in LDKT knowledge (p=0.02), lower LDKT concerns (p=0.03), and higher self-
efficacy/willingness to speak with others about LDKT (p < 0.001). Thematic analysis of
qualitative data using NVivo 10 (QSR International Ltd.) showed that social support and self-
efficacy were two key constructs. Results showed: 1) Video chat sessions were viewed
positively as an opportunity to develop social support, LDKT knowledge, and behavior skills to
seek a potential LKD. The co-learning that occurred through shared conversations with other
participants increased a sense of self-efficacy, which converged in agreement with the survey
results. 2) The positive opportunities for interaction outweighed the barriers of connectivity
and adaptation to video chat as a mechanism for communication. However, some individuals
needed technical support or stronger cellular or network signals to achieve consistently good
connectivity to outweigh live group interactions. 3) Participants viewed the content of the
video modules as informative, helpful, and conducive to generate requests for a kidney
donation. This increase in knowledge lessened barriers to ask others for a LDKT 4) A sense of
increased empowerment resulted from the interactions between group members through
mentoring activities and practicing asking others for a LDKT. Overall, the video education
modules and video chat sessions were inherently interdependent with the results and findings
S2000 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

were positive. Refinement of content and presentation material was made and LOVED is
undergoing a small scale efficacy RCT. Further study would be needed to determine if self-
efficacy can be maintained without the video chat sessions to support asking for a LDKT.

CORRESPONDING AUTHOR: John C. Sieverdes, PhD, Medical University of South Carolina,

Charleston, SC, 29425; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2001

C124 6:00 PM-7:00 PM

COLLEGE STUDENTS USE OF MOBILE APPLICATIONS FOR WEIGHT LOSS

Matthew C. Whited, PhD, Licensed Psychologist1, Jordan Ellis, MA1, Ansley Taylor Corson,
M.A.1, John T. Freeman, B.A.1, Emily P. Midgette, Psychology, BA (Expected May 2017)2,
Sherry Pagoto, PhD3, Molly E. Waring, PhD3, Christine May, PhD4, Kristin Schneider, Ph.D.5,
Andrew Busch, PhD6, Brad Appelhans, PhD7, Eliza Varju, BS (Expected May, 2019)1, Laura K.
Nye, BA in Psychology, May 20171

1
East Carolina University, Greenville, NC; 2East Carolina University, Vanceboro, NC; 3University
of Massachusetts Medical School, Worcester, MA; 4University of Massachusetts Medical
School, Fiskdale, MA; 5Rosalind Franklin University of Medicine & Science, North Chicago, IL;
6
The Miriam Hospital, Rumford, RI; 7Rush University Medical Center, Chicago, IL

Understanding college students use of and preferences for health-related mobile applications
is needed to inform the development of mhealth interventions for this population.

As part of an annual survey of college student health, students were asked: have you ever used a mobile application (app)
for weight loss? Users, those who reported using an app for weight loss (WL) reported on their most and least liked feature
of the WL app they used the most; Non-users, who never used a WL app, reported on desired features for a WL app.
Responses to these open-ended questions were categorized by 3 coders (percent agreement 78-88%).

Of all participants (n=1906; 67% women; 30% overweight/obese), almost one-third (31%) were users. Among
overweight/obese students 44% were WL app users compared to 26% of normal weight students (p

Of the users (n=586), 43% reported using a WL app for both exercise and diet, 23% for exercise only, and 32% for diet only.
Three in 10 (29%) reported using the app for 1-2 months; 39% for less than one month; 16% for 3-4 months; and 15% for 5
months or longer. Most users (57%) reported daily use. Users reported that tracking/self-monitoring (64%) and
tutorials/instructions (9%) were their most liked features. Least-liked features included tracking/self-monitoring (38%) and
financial/time cost (6%). Non-users (n=1320) reported needing features related to tracking/self-monitoring (18%) and
tutorial/instruction (12%). Users and non-users infrequently reported preferences for other features such as goal setting,
motivational features, social networking.

For the two most liked features among users, more men liked tutorial features (26% of men) than women (10% of women;
p=.002). Most liked feature did not differ based on weight status. No differences for gender or weight status were observed
for least liked or needed features.

Almost one third of college students have used an app for WL, although most have used them for less than 2 months.
Tracking is important to this population, and satisfaction with tracking appears important to attract and maintain users. Other
features that are part of empirically-supported weight loss treatment were reported infrequently and may not feature
prominently into users experiences (i.e., goal setting and motivational features to maintain engagement). College men are
underrepresented among app users.
S2002 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Matthew C. Whited, PhD, Licensed Psychologist, East Carolina

University, Greenville, NC, 27858; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2003

C125 6:00 PM-7:00 PM

FEASIBILITY AND ACCEPTABILITY OF A SMARTPHONE APPLICATION FOR MEASURING TIME USE

(LIFE IN A DAY) IN BREAST CANCER SURVIVORS

Matthew C. Ainsworth, MPH1, Dori Pekmezi, PhD1, Heather R. Bowles, PhD2, Edward
McAuley, PhD3, Kerry S. Courneya, PhD4, Laura Q. Rogers, MD, MPH1, Diane K. Ehlers, PhD5

1
University of Alabama at Birmingham, Birmingham, AL; 2National Cancer Institute, Rockville,
MD; 3University of Illinois at Urbana Champaign, Urbana, IL; 4University of Alberta, Edmonton,
AB, Canada; 5University of Illinois at Urbana-Champaign, Urbana, IL

Background: Advancements in mobile technology allow innovative data collection techniques

such as measuring time use (i.e., how individuals structure their time) for the purpose of
improving health behavior change interventions. To advance technology-based measurement
of time use, the current study examined the acceptability of a five-day trial of the Life In A Day
(LIAD) smartphone app measuring time use in breast cancer survivors. Methods: Feasibility
and acceptability data were collected from participants (N=40; 100% response rate) using a
self-administered survey after 5 days of LIAD use. Results: Overall, participants had a mean
age of 55 years (SD=7.96) and completed 16 years of school (SD=2.08). Participants generally
agreed that learning to use LIAD was easy (82.5%, N=33) and preferred to log activities using
LIAD over paper and pencil diary (72.5%, N=29). Most felt that completing LIAD for 5
consecutive days was not too much (60%, N=24) or overly time consuming (67.5%, N=27).
LIAD was rated as easy to read (87.5%, N=35)) and navigate (70%, N=32). Participants also
agreed that it was easy to log activities using the activity timer at the start and end of an
activity (89.7%, N=35). Only 12.5% (N=5) downloaded the app on their personal phone while
63.3% of the remaining 35 would have preferred to use their personal phone. Overall,
participants felt that the LIAD app was good or very good (76.9%, N=30). Those who
agreed it was easy to edit activities were significantly more likely to be younger when
compared with those who disagreed (M=53 vs 58 years, p < .05). Similarly, those who agreed
that it was easy to remember to log activities were more likely to be younger (M=52 vs 60
years, p < .05). Qualitative coding of two open-ended survey items yielded three common
themes for LIAD improvement (i.e., convenience, user interface, and reminders). Conclusion:
A smartphone app is a feasible and acceptable time use measurement modality. Improving
convenience, user interface, and memory prompts while addressing the needs of older
participants are needed to enhance app utility.
S2004 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Matthew C. Ainsworth, MPH, University of Alabama at

Birmingham, Birmingham, AL, 35205; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2005

C126 6:00 PM-7:00 PM

MFIT (MOTIVATING FAMILIES WITH INTERACTIVE TECHNOLOGY) STUDY: A RANDOMIZED

PILOT TO PROMOTE PHYSICAL ACTIVITY & HEALTHY EATING

Danielle E. Jake-Schoffman, PhD1, Brie Turner-McGrievy, PhD, MS, RD2, Sara Wilcox, PhD2,
James Hussey, PhD2, Justin B. Moore, PhD, MS, FACSM3, Andrew Kaczynski, PhD4

1
University of Massachusetts Medical School, Worcest, MA; 2University of South Carolina,
Columbia, SC; 3Wake Forest School of Medicine, Winston-Salem, NC; 4University of South
Carolina, Colubmia, SC

BACKGROUND: The purpose of the Motivating Families with Interactive Technology (mFIT)
study was to test the feasibility, acceptability, and effectiveness of two remotely-delivered
family-based health promotion programs for improving physical activity (PA) and healthy
eating (HE).

METHODS: Thirty-three parent-child (child age 9-12 years) dyads were randomized to one of
two 12-week mobile interventions to increase PA and HE, which included weekly email
newsletters and the use of pedometers. The programs differed in focus of content (individual
vs. family) and method of tracking (paper vs. mobile website). At baseline and 12 weeks
height and weight were measured, and participants completed questionnaires about dietary
intake (BRFSS and YRBSS), and feedback on and engagement in the mFIT program. Descriptive
statistics and linear mixed effects models were used to analyze program data; all analyses
were conducted in SAS 9.4.

RESULTS: Of the 33 randomized dyads (parents: 43+6 years, 88% female, 70% white, BMI
31.1+8.3 kg/m2; children: 11+1 years, 64% female, 67% white, BMI 77.6+27.8 percentile), 31
(94%) had follow-up data. There were no between-group differences for PA or HE, but there
was an overall significant increase in average daily steps and servings of fruit during the
intervention and excellent adherence to self-monitoring protocols. Most parents (97%) and
children (86%) would recommend the program to a friend.

CONCLUSION: The mFIT program showed excellent feasibility and acceptability as a low-cost,
remotely-delivered family intervention for PA and HE promotion, and could serve as a
disseminable model for public health interventions.
S2006 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Danielle E. Jake-Schoffman, PhD, University of Massachusetts

Medical School, Worcest, MA, 01609; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2007

C127 6:00 PM-7:00 PM

MYLIFE: USING DIGITAL HEALTH TECHNOLOGY TO ENCOURAGE HEALTHY BEHAVIORS IN

PATIENTS AT RISK FOR LIFESTYLE-RELATED DISEASES

Daniel M. Croymans, M.D., M.B.A., M.S.1, Paul Bixenstine, M.D.2, Ian Hurst, High School
Diploma3, Carlos Casillas, B.S.4, Lisa Gantz, M.D.2, Sung Hyun Kim, M.P.H.4, Sally Elliott, B.S.4,
Sarah Tan, High School Diploma3, Alice Kuo, M.D., Ph.D.2

1
Internal Medicine, UCLA Health, Los Angeles, CA; 2Internal Medicine and Pediatrics, UCLA
Health, Los Angeles, CA; 3UCLA, Los Angeles, CA; 4UCLA David Geffen School of Medicine, Los
Angeles, CA

Background:
Many patients at risk for lifestyle-related chronic disease struggle to adopt and maintain
healthy behaviors despite health education and lifestyle recommendations from their
physicians. Digital health technology could improve patient-provider communication and
patient health outcomes if integrated into the health delivery workflow. MyLife is a 16-week
digital health program that uses wearable activity trackers, mobile messaging, and data
analytics to encourage healthy behaviors in patients at risk for lifestyle-related chronic
diseases.

Methods:
UCLA primary care providers (PCPs) referred patients for enrollment. Upon intake, study
personnel gave patients a Jawbone UP3 tracker and instructions for using the mobile
application. Patients set exercise goals, identified barriers and discussed solutions. Patients
received follow-up calls every 2 weeks and 4-6 text messages per week for 16 weeks. PCPs
were given periodic reports on patient progress and helped promote patient engagement.
PDSA cycles were used to evaluate and refine the MyLife program based on changes in patient
bodyweight, physical activity (PA), motivation and feedback. The Behaviour Change Wheel
and Behavior Change Techniques (BCT) Taxonomy were consulted to guide iterative
improvements.

Results:
Participant baseline characteristics (cycle 1, n=12; cycle 2, n=10) included mean age 42 11
years, 50% female, mean BMI 35.4 7.5, mean PA 34.538 min/wk, and mean patient
activation measure 42 5 (raw score, stage 3 of 4). Patients who completed the 16-week
program increased their baseline PA by more than 200% and lost an average of 5% of their
bodyweight. Based on results and patient feedback from the first cycle, which focused on
S2008 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

increasing PA, the second cycle will also address patient dietary choices and patients will have
a designated health coach. In total over 100 unique, branching logic messages that leverage
several key BCTs including Goal Setting + Behavioral Contract, Problem Solving + Graded
Tasks, Action Planning, and Review Behavior Goals were developed.

Conclusions:
Our pilot digital health intervention demonstrated significant improvement in PA and weight
among participants after 16 weeks. Digital health technology may be a powerful tool to
prevent lifestyle-related chronic diseases when integrated into the health system. In the
future, we hope to improve our program by using more interactive messaging and real-time
analytics to leverage proven behavior change techniques.

CORRESPONDING AUTHOR: Daniel M. Croymans, M.D., M.B.A., M.S., Internal Medicine, UCLA
Health, Los Angeles, CA, 90095; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2009

C128 6:00 PM-7:00 PM

REGIONAL CHALLENGES OF USING DATING APPS AMONG MSM IN RURAL AREAS

Anne Marie Schipani-McLaughlin, MPH1, Danielle N. Lambert, MPH, CHES2, Natalia

Truszczynski, MPH1, Carolyn Lauckner, PhD1, Nathan Hansen, PhD1

1
University of Georgia, Athens, GA; 2University of Georgia, Atlanta, GA

Background: Little is known about the challenges of using dating apps to meet potential
partners among MSM residing in different geographical locations of the United States. This
study explored MSMs experiences of using dating/hookup apps to meet potential sexual and
romantic partners within different US regions and areas of varying population density.

Methods: Semi-structured qualitative interviews were conducted with 20 MSM between the
ages of 18-60 years currently residing in rural areas. Eligibility criteria included identifying as a
man who dates or has sexual relations with other men, being at least 18 years of age, and
residing in a non-metropolitan area of the United States. Interviews were transcribed
verbatim and analyzed in NVivo 11 by a diverse coding using grounded theory approaches.

Results: Results reveal a number of challenges MSM face when using dating/hookup apps to
meet partners in rural areas, specifically regarding regional characteristics, comfort being
out, and population size. Population density and sprawling regional characteristics limit the
number of openly identifying MSM who use dating/hookup apps in rural areas. This leads to a
limited number of partner matches on the apps and respondents therefore cycle through
multiple dating/hookup apps to increase their dating pool. Rural areas with populations
predominantly comprised of university staff or students (college towns) also present issues
for finding potential matches online. Respondents indicated that they did not feel comfortable
being out within the context of their rural community, which deterred them from meeting
partners in that area. This was not perceived to be an issue in large urban areas where
respondents had more anonymity and many preferred to use dating/hookup apps when they
were visiting metropolitan cities with larger communities of MSM.

Conclusion: These findings reveal that the regional environment plays a key role in the
success of using dating and hookup apps to meet potential partners among MSM. The
regional environment also impacts MSMs comfort openly identifying as gay within their
community and seeking partners in rural areas. Health professionals should therefore
consider regional characteristics and partner-seeking behavior using dating/hookup apps
when designing interventions for MSM in rural areas.
S2010 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Anne Marie Schipani-McLaughlin, MPH, University of Georgia,

Athens, GA, 30606; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2011

C129 6:00 PM-7:00 PM

SELF-EFFICACY FOR EXERCISE-SPECIFIC MEMORY AND PROSPECTIVE MOBILE HEALTH

APPLICATION USE

Daniel Palac, MA1, Sean P. Mullen, PhD2

1
University of Illinois at Urbana-Champaign, Champaign, IL; 2University of Illinois at Urbana-
Champaign, Urbana, IL

As smart phone devices become more prevalent in today's society, so do mobile health
(mHealth) applications (apps). Tens of thousands of mHealth apps are commercially available
that target health and fitness, yet minimal research has identified antecedents associated
with continued use. Research has consistently associated self-efficacy and memory with
physical activity behavior. Exercise-specific memory is a construct that has received little
attention, yet there is reason to believe high levels of self-efficacy for exercise-specific
memory (SEFESM, i.e., confidence in one's ability to remember public health
recommendations, modifications, and how to evaluate progress with respect to aerobic,
strength, and flexibility training) might contribute to the frequency and continued use of
exercise-related mHealth apps. To test this hypothesis, a survey was administered on two
occasions: baseline (126 initiated; n=94 completed; Mage=38 years; 64% female) and one-
month follow-up (n=76). Path modeling with robust maximum likelihood estimation was
conducted using MPlus (version 7.4) to assess theorized direct and indirect effects of SEFESM
on weekly app use ("...to assist me with my exercise program;" assessed via an 8-point ordinal
scale ranging from 0 to 30+ times) at baseline and at one-month follow-up, while statistically
adjusting for age, gender, education, access to technology (e.g., "Do you own a
smartphone/tablet/wrist device?"), and self-reported moderate-to-vigorous physical activity
levels (derived from the Godin Leisure-Time Exercise Questionnaire). The overall model fit the
data (2=20.908 (19), p=.34, RMSEA=.03, CFI=.99, SRMR=.07). Results indicated that SEFESM
and access to technology had significant (p < .05) direct effects on app use at baseline (=.235
and =.177, respectively). App use at baseline also had a direct effect at one-month follow-up
(=.667); also found was a marginal indirect effect of SEFESM on weekly app use at follow-up
via app use at baseline (=.157, p=.06). These results support our hypothesis that self-efficacy
for exercise-specific memory contributes to an individual's use of exercise-related apps and
should be an important consideration when designing mobile interventions.

CORRESPONDING AUTHOR: Daniel Palac, MA, University of Illinois at Urbana-Champaign,

Champaign, IL, 61820; [email protected]
S2012 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C130 6:00 PM-7:00 PM

THE EHEALTH ACTIVITY ASSESSMENT: PRELIMINARY FINDINGS AND IMPLICATIONS FOR

CANCER SURVIVORSHIP AND USER-CENTERED DESIGN

Hayley S. Thompson, Ph.D.1, Nicole Senft, PhD2, Anne Katz, PhD3, Tara Eaton, PhD4, Deborah
H. Charbonneau, PhD5, Charity A. Barnes, B.S.5, Jennifer L. Beebe-Dimmer, Ph.D., M.P.H.6, Ke
Zhang, PhD5, Judith Abrams, PhD7, Elisabeth Heath, MD8

1
Karmanos Cancer Institute - Wayne State University, Detroit, MI; 2Karmanos Cancer
Institute/Wayne State University, Ann Arbor, MI; 3Wayne State University/Karmanos Cancer
Institute, Detroit, MI; 4Levine Cancer Institute, Charlotte, NC; 5Wayne State University,
Detroit, MI; 6Wayne State University School of Medicine/Karmanos Cancer Institute, Detroit,
MI; 7Wayne State university/Karmanos Cancer Institute, Detroit, MI; 8Karmanos Cancer
Institute, Detroit, MI

There are an increasing number of eHealth interventions targeting cancer survivors but little is
known about eHealth preferences and experiences in this group, areas that are central to
user-centered digital design. The current study presents early data from an ongoing,
population-based study of eHealth activity among African American (AfAm) and white cancer
survivors that will establish survivor-centered design principles to be translated in the
development of a survivor resource app. Investigators created the eHealth Activity
Assessment (eAA), a mixed methods interview examining 17 eHealth activities across 5
domains: informational, communal, self-care, expert care, and transactional. Dimensions of
activity examined included direct versus surrogate engagement, frequency, the extent to
which an activity served a cancer-related need, and future interest. Participants were 96
survivors (mean age= 61 years; 64% AfAm; 78% breast cancer) enrolled through the
Metropolitan Detroit Cancer Surveillance System - SEER. Quantitative findings are presented
here. In this sample of internet users and non-users, 89% reported at least one eHealth
activity but only 6% reported any surrogate engagement (i.e., others conducted activity on
their behalf). The median number of activities was 6 (min=1; max=14). The most common
activities were online search for health information related to symptoms, treatments, or
procedures (81%), online search for healthy lifestyle information (62%), and use of technology
to reduce stress (60%). In terms of frequency, a substantial proportion of those who used
technology to manage medications (100%), reduce stress (54%), and track weight, diet, or
exercise (50%) did so on a daily basis. On average, 16% of reported activities addressed
cancer-related needs post-treatment. Across the sample, future interest was highest in online
search for ways to reduce healthcare costs (37%) and looking at medical records or test
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2013

results online (31%). T-test results showed that, compared to white survivors, AfAm survivors
engaged in fewer eHealth activities (4 vs. 8, p < .0001). Fisher's exact tests showed that a
significantly lower proportion of AfAms engaged in 9 of the 17 eHealth activities, most notably
in the communal and informational domains. AfAms also reported fewer activities related to
cancer needs post-treatment (11% vs. 22%, p < .0003). Finally, a significantly greater
proportion of AfAms indicated strong interest in connecting with similar others online (p <
.001) and using technology to reduce stress (p < .04) if they had never done so. Findings
suggest that in-depth assessment of eHealth activity may reveal trends that can inform and
enhance the user-centeredness of new and existing digital interventions for cancer survivors,
especially among specific subgroups, thereby increasing the likelihood of adoption and
meaningful impact.

CORRESPONDING AUTHOR: Hayley S. Thompson, Ph.D., Karmanos Cancer Institute - Wayne

State University, Detroit, MI, 48201; [email protected]
S2014 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C131 6:00 PM-7:00 PM

USABILITY TESTING OF AN MHEALTH INTERVENTION TO IMPROVE ADOLESCENTS' HEADACHE

SELF-MANAGEMENT

Kelly A. Manser, B.A.1, Lisa Sawyer, BFA2, Kimberlee J. Trudeau, Ph.D.1

1
Inflexxion, Inc., Waltham, MA; 2QuickBase, Cambridge, MA

Adolescents who experience recurring headaches report lower quality of life as headaches
become more frequent and severe (Donovan et al., 2011; Sillanpaa, 1983). Self-management
skills training has been shown to increase self-efficacy for healthful behaviors as part of
multidisciplinary care for chronic health conditions (Burckhardt, 2005). To facilitate
adolescents self-management of recurring headaches, we developed the Headzup App for
the iPhone. Features of Headzup include a daily headache tracker, evidence-based self-
management strategies that were developed with clinician and end-user input, and a stats
view showing recent tracking trends (e.g., pain level, suspected headache etiology).

Nine adolescents who experience >4 bothersome headaches per month (mean age = 14.33
years) were recruited via online resources. They were asked to use Headzup daily for 14 days,
then participate in a one-hour interview about their use of the app. At the end of the
interview participants were asked to complete the 10-item System Usability Scale (Brooke,
1996). The SUS is a standard measure to assess ease of use with new programs/products.
Participants were paid $50 for participating in the interview.

All participants completed the headache tracker on > 11 days (M = 14.33; range from 11 to 21
days). SUS scores >68 indicate above-average usability; the mean SUS score for the Headzup
App was 88.6 on a 100-point scale. All but one of the participants reported that using
Headzup helped them to cope with headache pain. When asked which specific feature was
most worth keeping, 6 teens (66.7%) advocated for the tracker. While 7 participants did not
notice the stats feature until they were directed to it during the interviews, all said it was
helpful and really interesting. Common wish list items included the ability to provide
free-text answers to a tracker question about suspected headache cause and stats reports
that display a longer (>7 days) tracking period.

The present study suggests that the Headzup app has good usability and includes features
that teens find useful for headache self-management. Developers of future mHealth
interventions for adolescents should consider including highly visible features that provide
data that helps users identify patterns in the co-occurrence of certain health behaviors and
chronic disease symptoms. The relationship between longer-term use of Headzup and
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2015

adolescents headache frequency, coping, and self-efficacy is currently being evaluated in a

clinical trial (Clinical Trials # NCT02475005).

CORRESPONDING AUTHOR: Kelly A. Manser, B.A., Inflexxion, Inc., Waltham, MA, 02451;
[email protected]
S2016 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C132 6:00 PM-7:00 PM

AN ASSESSMENT OF PARTICIPANTS EXPERIENCES WITH TWO EHEALTH INTERVENTIONS

Marie A. Sillice, Ph.D.1, Patricia J. Morokoff, Ph.D2, Ginette Ferstz, Ph.D.3, Timothy Bickmore,
Ph.D.4, Beth C. Bock, PhD5, Wayne F. Velicer, Ph.D.6

1
Department of Psychiatry and Human Behavior, the Miriam Hospital and Alpert School of
Medicine at Brown University, Providence, RI; 2University of Rhode Island, Kingston, RI;
3
College of Nursing, University of Rhode Island, Kingston, RI; 4College of Computer and
Information Science Northeastern University, Boston, MA; 5Brown Medical School,
Providence, RI; 6Cancer Prevention Research Center Behavioral Science, University of Rhode
Island, Kingston, RI

Objective: During the last 30 years, computer-based intervention for changing health
behaviors have been developed and demonstrated to be effective in clinical trials. However,
the effect sizes have been small due to a lack of engagement by a significant proportion of the
population. Relational Agents (RAs) represent one of the recent innovative eHealth
approaches to increasing engagement. RAs are computational figures designed to engage
participants in the change process. A recent study tested the effectiveness of RAs, combined
with existing computer-based interventions, to increase regular exercise and sun protection
behaviors. The study demonstrated that these interventions can be effective but need further
development. This study focuses on examining the participants experiences of using the RA,
using mixed methods approaches.

Method: Thirty-four participants completed a standardized semi-structured open-ended

interview. A 25-question interview guide assessed different components of participants
experiences with the intervention, including motivation, engagement, satisfaction/
dissatisfaction, quality of their interaction with the RA, and behavior change. Quantitative
assessment of satisfaction was based on a scale of 1 to 10, with 1 representing least satisfied
and 10 representing most satisfied. A summative analytic approach was used to assess
individuals qualitative responses. An ANOVA examined levels of satisfaction by gender.

Results: The results showed individuals were motivated by and satisfied with the intervention,
viewed the RA as supportive, informative, and caring, and reported positive behavior change
in both exercise and sun protection. Some participants felt that the RA was less judgmental
and less overbearing compared to a human counselor, others said that the interaction was
sometimes repetitive or overly general. The majority of participants viewed the RA as an
important contributor to their behavior change for exercise, sun protection, or both. Levels of
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2017

satisfaction ranged between 7 and 10. Gender differences were not noted for levels of
satisfaction.

Conclusion: RAs provide an innovative and attractive platform to increase exercise and sun
protection behaviors. This component can serve as an effective medium to increase other
health behaviors. Nevertheless, improvement of this component to include a broader
database of tailored messages to decrease receptiveness of information will further increase
its appeal.

CORRESPONDING AUTHOR: Marie A. Sillice, Ph.D., Department of Psychiatry and Human

Behavior, the Miriam Hospital and Alpert School of Medicine at Brown University, Providence,
RI, 02906; [email protected]
S2018 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C133 6:00 PM-7:00 PM

AN UPSTREAM APPROACH TO ADDRESS ALCOHOL & SUBSTANCE USE: INTEGRATING SBIRT IN

HEALTH PROFESSIONAL TRAINING PROGRAMS

Diane King, PhD, Alexandra Edwards, MA, Bridget Hanson, PhD, Jodi Barnett, MA, Amanda
Zold, M.S.

University of Alaska Anchorage, Anchorage, AK

Misuse of alcohol and drugs is the third leading cause of preventable death in the United
States. Despite strong research evidence that standardized screening, brief intervention and
referral to treatment (SBIRT) changes behavior and improves patient outcomes in diverse
practice settings, its adoption lags best practice recommendations. The Substance Abuse and
Mental Health Services Administration (SAMHSA) has identified a need to move SBIRT
implementation approaches further upstream, including the integration of SBIRT within
medical residency and multi-disciplinary health professional training programs. We studied
the implementation of SBIRT training into psychology, social work, nursing and family
medicine residency programs at a state University. A standardized SBIRT curriculum was
adapted to allow implementation within core and elective courses, in conjunction with
opportunities to practice skills with standardized patients through the Universitys Simulation
Center. Over eight academic terms, 704 students and 30 residents received the course
curriculum. Student surveys administered before and after receiving the SBIRT curriculum,
indicated significant increases in curriculum knowledge, confidence to apply SBIRT skills with
patients related to alcohol and drug use, and perceived level of responsibility to use SBIRT
skills with patients related to alcohol and drug use (p < .001 for all). Effect sizes were 2 = .00
for responsibility to use SBIRT, 2 = .04 for confidence, and 2 = .25 for knowledge.
Recommendation from students included more hands-on opportunities to practice SBIRT
skills, which may also improve the effectiveness of the training in increasing student
confidence to apply SBIRT with patients. Recommendations from faculty included more
discipline-specific SBIRT skills demonstration videos for students, and resources, such as
articles and discussion materials, that would help instructors fit the SBIRT curriculum content
into courses already constrained by limited classroom timea key implementation challenge.
While use of videos and readings was reported to improve implementation feasibility,
maintaining SBIRT within program curricula may require changes to professional accreditation
and credentialing standards, as well as dissemination of SBIRT within practice settings, to
ensure upstream training of healthcare professionals is necessitated by its midstream use in
clinical practice as a standard of integrated evidence-based care.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2019

CORRESPONDING AUTHOR: Diane King, PhD, University of Alaska Anchorage, Anchorage, AK,
99508; [email protected]
S2020 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C134 6:00 PM-7:00 PM

THE GROW! PARENTING PROGRAM: A HYBRID TYPE III DESIGN STUDY

Jennifer M. DiNallo, Ph.D.1, Melina T. Czymoniewicz-Klippel, Ph.D.2, Ryan P. Chesnut, Ph.D.2,

Daniel F. Perkins, Ph.D.2

1
Pennsylvania State University, Boalsburg, PA; 2Pennsylvania State University, State College,
PA

Background: The development and dissemination of evidence-based parenting programs is a

major focus of researchers and practitioners in the social and behavioral sciences.
Traditionally, these programs have focused on enhancing specific parenting skills, such as
assertive discipline and labeled praise, in an attempt to improve parent and child mental
health outcomes. The Grow! parenting program (Grow!), for parents of five-to-eight year olds,
emphasizes these same parenting skills, but is unique in that it also underscores evidence--
informed stress management and health promotion skills (e.g., child feeding, physical activity,
and screen time behaviors) in an effort to impact parent and child health holistically.

Methods: In 2015, a Hybrid Type III design study was conducted in two Pennsylvanian
communities. Implementation data were collected from facilitators and participants
throughout program delivery and at program completion. Treatment data were collected
from participants at pretest, posttest, three-month, and six-month follow-up using measures
assessing positive parenting practices, stress management, and health promotion. Responses
from 15 participants who completed measures at all four time points were analyzed using
Friedmans ANOVAs and appropriate follow-up tests.

Findings: Overall, participants and facilitators were satisfied with the program. Participants
were engaged during the sessions and indicated the material was useful. Facilitators also
delivered the program with a high degree of fidelity (i.e., adherence rating = 94%). Repeated
measures analyses showed statistically significant (p < .05) improvements on several stress
management and health promotion measures with some parenting measures approaching
significance (p < .10). Effect sizes (i.e., r) for significant outcomes ranged from .29 to .48.

Implications for D&I Research: Dissemination and implementation research seeks to translate
knowledge into practice. This study contributes to that mission by expanding the focus of
parent programming to include evidence-informed stress management and health promotion
strategies. The findings demonstrate that implementing Grow! with fidelity in community
populations is feasible and well received. In addition, the results indicate Grow! has the
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2021

potential to produce changes in stress management and health promotion outcomes. These
findings will help to stimulate greater thinking about how best to develop, disseminate, and
evaluate parenting programs that influence all aspects of childrens health.

CORRESPONDING AUTHOR: Jennifer M. DiNallo, Ph.D., Pennsylvania State University,

Boalsburg, PA, 16827; [email protected]
S2022 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C135 6:00 PM-7:00 PM

ONLINE EDUCATION TO ENHANCE DISCUSSION ABOUT COMPLEMENTARY THERAPIES

BETWEEN HEALTH CARE PROVIDERS AND CANCER PATIENTS

Kirsti Toivonen, HBA1, Greg Levin, PhD1, Aven Sidhu, BSc2, Erin Zelinski, PhD1, Michelle J.
Flynn, BA Hons, Psychology1, Jennifer White, PhD3, Lynda Balneaves, PhD4, Linda Carlson,
PhD1

1
University of Calgary, Calgary, AB, Canada; 2Alberta Health Services, Calgary, AB, Canada;
3
Hunter Medical Research Institute, New South Whales, N/A, Australia; 4University of
Manitoba, Winnipeg, MB, Canada

Background: Nearly 50% of cancer patients report using complementary therapies (CTs) since
diagnosis and evidence supports the use of CTs in conjunction with usual cancer treatment.
However, both patients and health care practitioners (HCPs) report a lack of knowledge about
the benefits and potential adverse effects of CTs, and research indicates there is little
discussion between HCPs and cancer survivors about CT use. The present study examines
whether an online CT education program delivered to HCPs can improve attitudes toward CTs
as well as increase discussion about CTs between patients and providers.

Methods: All HCPs affiliated with the Tom Baker Cancer Centre in Calgary, Canada (approx.
300) were invited to enroll in a 3-module online course, which included information about
patient CT use, how to talk to patients about CTs, and a summary of evidence for CTs.
Attitudes towards CTs and CT-specific clinical practice behaviours with patients over the past
30 days were assessed at baseline and two months later following completion of online
modules. HCPs were assessed at follow-up regardless of whether they completed all modules.
Quantitative and qualitative data were examined.

Results: Of the HCPs who were invited, 105 expressed interest, 83 enrolled in the study, 73
completed all assessments and 61 completed all three modules. Both those who completed
all modules and those who did not reported an increase (80% and 10%, respectively) in asking
their patients about CT use. However, only completers provided more information about CTs
(+11.5% vs -16.7%), made more recommendations for CTs (+14.7% vs 0%), and discussed the
evidence supporting CTs (+18.1% vs -8.3%). While HCPs saw the value in knowing about CTs to
help counsel interested patients, they felt that they did not have sufficient depth of
knowledge and expressed interest in having a support staff with specialized CT knowledge.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2023

Conclusions: The online education course appeared to increase the frequency of HCPs asking
their patients about CT use, but only those who completed the course were able to engage in
more meaningful discussions about CTs. Interested HCPs may benefit from more advanced
seminars about CTs. Further, a designated position for a CT practitioner in a Cancer Centre
may be valuable given the great patient interest in CT use.

CORRESPONDING AUTHOR: Kirsti Toivonen, HBA, University of Calgary, Calgary, AB, T2N 1N4;
[email protected]
S2024 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C136 6:00 PM-7:00 PM

A MIXED METHODS APPROACH TO DESIGNING A SOCIAL MEDIA INTERVENTION FOR

MOTHERS OF ADOLESCENT GIRLS IN TENNESSEE

Katie Baker, DrPH, MPH1, Sherry Pagoto, PhD2, David Buller, PhD3, Barbara Walkosz, PhD4,
Joel Hillhouse, PhD1, Julia Griffith, Master of Social Work5, Oluwatosin Ariyo, MPH1

1
East Tennessee State University, Johnson City, TN; 2University of Massachusetts Medical
School, Worcester, MA; 3Klein Buendel, Inc., Golden, CO; 4Klein Bendel, Inc., Wheat Ridge,
CO; 5Klein Buendel Inc, Golden, CO

Reducing indoor tanning (IT) is critical to preventing melanoma among young women.
Mothers strongly influence whether daughters engage in IT, and teen girls often initiate IT
with their mothers. Furthermore, girls who first experience IT with their mothers begin at an
earlier age, become more habitual tanners, and are more resistant to change. Thus, mothers
of teen girls are a significant target for IT interventions. Currently, 27 states, including
Tennessee (TN), require parental permission for minors to IT. To ensure these laws have
maximum prevention benefits, we developed a social media health intervention aimed at
motivating mothers in TN to withhold IT permission. Because mothers who use IT might not
be motivated to join a program designed to change their attitudes about tanning, we designed
an approach in which tanning messages are embedded in an intervention about health topics
of high interest to moms. To identify high interest topics, trained facilitators conducted key
informant interviews with TN coordinated school health (CSH) personnel and focus groups
with mothers of teen girls. Interviews and focus groups were recorded and transcribed and
content analyzed. CSH interviews (n=19, representing 40 high schools) indicated that mental
health (stress, bullying), substance abuse (tobacco, alcohol, prescription drugs), and obesity
(physical activity, nutrition) were the topics of greatest interest. Results were confirmed in the
focus groups with mothers (n=10), who also cited body image and healthy peer relationships.
Intervention content was then designed using theory-driven messages to address IT plus these
other health topics to maintain mothers interest throughout the social media campaign. An
initial prototype of a Facebook group was used by mothers (n=8) for 1 week, who then
attended a focus group to provide feedback. Mothers reacted positively to the intervention,
appreciating the diversity of topics delivered over a popular platform, felt more aware of
important teen health issues, and used the posts as a checklist that often prompted mother-
daughter communication. Using 5-point Likert-type agreement scales, participants rated the
IT posts on interest (avg. rating = 3.99), clarity (avg. rating = 4.70), credibility (avg. rating =
3.81), and appropriateness for social media (avg. rating = 4.43). Results were used to refine
the intervention, which was launched in fall 2016.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2025

CORRESPONDING AUTHOR: Katie Baker, DrPH, MPH, East Tennessee State University, Johnson
City, TN, 37614; [email protected]
S2026 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C137 6:00 PM-7:00 PM

ITS NOT JUST SKIN CANCER: UNDERSTANDING THEIR CANCER EXPERIENCE FROM
MELANOMA SURVIVOR STORIES SHARED ONLINE

Smita Banerjee, Ph.D.1, Thomas A. D'Agostino, PhD1, Mallorie Gordon, MA2, Jennifer L. Hay,
Ph.D.1

1
Memorial Sloan Kettering Cancer Center, New York, NY; 2The New School for Social Research,
New York, NY

Cancer survivors narrate their stories in unique ways, articulating different aspects of cancer
experience. The purpose of this study was to analyze the content of cancer stories that
melanoma survivors share online in order to present the ways that survivors narrate their
cancer experience, to identify survivors motivations for sharing, and to better understand the
ways in which survivors are impacted by and cope with the diagnosis and treatment of cancer.
In order to accomplish the objective, we conducted a qualitative analysis of melanoma
survivor stories shared online. The sample consisted of 95 unique melanoma survivor stories,
accessed from the Melanoma Research Foundation in November 2015, that were inductively
and deductively coded for key themes and sub-themes. Emergent themes described different
aspects of the melanoma experience from pre-diagnosis (identification of self-phenotype,
putative behavioral causes of melanoma, suspicious findings, delay in diagnosis), diagnosis
(communication of diagnosis, emotional responses), transition from diagnosis to beginning
treatment (second opinion), treatment (positive reframing of attitude, proactive cancer
management, side effects), and post-treatment phases (social support, vigilance behaviors
post-treatment). Two themes that cut across all phases of the cancer journey were
recognizing and dealing with uncertainty and survivors motive for sharing stories. The time
period between undergoing a diagnostic test and receiving results is rife with uncertainty and
anxiety, and the stories described uncertainty across three time periods: between diagnostic
test and receiving results, while undergoing treatment (uncertainty about efficacy of a new
drug used/clinical trial), and the period after finishing treatment and getting the next PET scan
to assess the effectiveness of the treatment in shrinking tumor/eliminating cancer. In
addition, survivors expressed interest in sharing their stories for a variety of reasons,
frequently related to lessons they had learned through the trajectory of their illness and a
desire to pass on their knowledge. Many survivors additionally expressed a strong desire to
improve public awareness of melanoma and draw attention to the difficulty, and often
deadliness, of the disease. These findings have implications for understanding how melanoma
survivors may benefit personally from sharing their cancer experience online as well as the
potential for survivor narratives to motivate behavior change and facilitate coping among
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2027

readers. Finally, this study presents immense potential to inform melanoma prevention health
communication campaigns and interventions, and to help inform supportive models of care
for other melanoma survivors

CORRESPONDING AUTHOR: Smita Banerjee, Ph.D., Memorial Sloan Kettering Cancer Center,
New York, NY, 10022; [email protected]
S2028 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C138 6:00 PM-7:00 PM

DO US HOSPITAL HOMEPAGES PROVIDE ACCESSIBLE CONTACT MODES FOR PEOPLE WITH

HEARING LOSSES? A NATIONALLY REPRESENTATIVE STUDY

Liza S. Rovniak, PhD, MPH1, Laura Penny, BS1, Sean Nutting, BS1, Christina Li, BS1, Jessica Dick,
BS1, Khaled Iskandarani, MS1, Allen R. Kunselman, MA1, Melbourne F. Hovell, PhD, MPH2,
Jennifer L. Kraschnewski, MD, MPH1, Christopher N. Sciamanna, MD, MPH1

1
Penn State College of Medicine, Hershey, PA; 2San Diego State University, San Diego, CA

Background: A hospitals homepage is like a virtual front door that can function to either
welcome diverse patients or discourage entry. Patients with hearing losses are a medically
underserved priority population who may use hospital homepages to identify non-hearing
dependent contact modes for accessing medical or mental health services. No prior study, to
our knowledge, has assessed the extent to which a nationally representative sample of US
hospital homepages provide accessible contact modes for patients with hearing losses.

Methods: From an initial database of 3,484 US hospitals that accepted Medicare in 2015
(www.medicare.gov), we used stratified cluster random sampling (stratified by hospital type
and clustered by US state) to obtain a representative sample of 526 hospitals. The homepages
and immediate relevant linking pages (e.g., contact information pages) of these hospitals were
coded for the presence vs. absence of five contact modes: telephone, email (address or online
email form), text message, TDD/TTY, and sign language services. Coding was done by two
independent raters who were trained to use a scripted search strategy that simulated how a
typical patient might search for contact modes. Discrepancies between the initial two raters
were resolved by a third independent rater.

Results: Cohens kappa across all five patient contact modes averaged 0.82, with a range of
0.73 to 0.91, ps < .0001, indicating good to excellent inter-rater agreement. Of the 526
hospitals, 4 had no website and 4 were closing. Of the remaining 518 hospitals, almost all
(99.4%) provided a phone number on their homepages. Among contact modes accessible to
both deaf and hard-of-hearing populations, 74.9% of hospitals provided email contact
information (62.7% had an online email form, and 18.5% had an email address), while just
0.4% provided text messaging information. Among contact modes typically used by deaf
populations, only 9.1% of hospital homepages provided a TDD/TTY number, and just 0.6%
described the availability of sign language services. One in five (20.3%) of the hospital
homepages failed to provide any non-hearing dependent contact mode that would be
accessible to deaf or hard-of-hearing populations.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2029

Conclusions: Many hospital homepages provide limited contact information for people with
hearing losses. The limited diversity of contact modes suggests a need to update existing
website design guidelines in the Americans with Disabilities Act and related policies. Some of
the contact modes studied, such as email and text messaging, could also offer collateral
benefits to those with language barriers or those with limited availability during working
hours. As hospital homepages provide an initial access point to healthcare for many patients,
providing more diverse contact modes could help reduce public health disparities.

CORRESPONDING AUTHOR: Liza S. Rovniak, PhD, MPH, Penn State College of Medicine,
Hershey, PA, 17033; [email protected]
S2030 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C139 6:00 PM-7:00 PM

EXPLORING IDENTITY RECONSTRUCTION IN PEOPLE LIVING WITH MULTIPLE SCLEROSIS

Rebecca Floyd, Ph.D.1, Shirin Khazvand, B.A.2, Rebecca Howell, B.A.1, Lara Stepleman, Ph.D.1

1
Medical College of Georgia at Augusta University, Augusta, GA; 2N/A, Augusta, GA

A person's identity shapes one's perspective of life's possibilities and aspirations for the
future. Quality of life can be adversely impacted by changes in indentity that lead
to downshifts in beliefs about future possibilities and in engagement in participatory roles in
life. Understanding how a person's identity is reshaped by experience of chronic disease is
therefore important to exploring attitudes that may impact treatment interventions and
restrict rehabilitation outcomes.

Ninety people living with Multiple Sclerosis (MS) and receiving services from a clinic in the
southeastern United States responded to the query of how, if at all, has MS changed the way
you think about yourself? Responses were analyzed using grounded theory and a number of
themes emerged. People indicated that their identity had been impacted by loss of focus on
the present, including lack of continuity between ones present self and ones past or future
selves. Additionally, sense of self was affected by changes in social relationships and threats
to ones integrity. Being forced to slow down, even when core components of self were
preserved, was enough to disturb ones identity, forcing identity reconstruction. These effects
may be moderated by the role of faith, spirituality, or religiosity in influencing the persons
perspective on the valence of altering ones identity. The salience of these themes was
also compared across demographic groups such as age and race/ethnicity and these
findings will be discussed.

Findings suggest that people may have difficulty focusing on the present due to a narrowing of
perspective that either overemphasizes past abilities or riddles a person with anxiety about
the future. Being unable to concentrate on the present or to perceive continuity between
present and past, or present and future, may likely stymy the rehabilitation objective of
meeting people where they are to move them forward.

Additionally, people living with MS appear to view being forced to slow down or being
interdependent as steps backward in their treatment trajectory, escalating the perception of
lost identity. While engagement of important others in one's care is often a vital source of
support, it is important then for the clinician or provider to recognize that involvement of
others may exacerbate a person's sense of identity disturbance rather than facilitating identity
sustainment or reconstruction. Realizing how people with MS communicate barriers to
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2031

identity reconstruction, in addition to disease management, will likely provide an

interdisciplinary rehabilitation team with better understanding of obstacles to psychosocial
functioning and overall quality of life.

CORRESPONDING AUTHOR: Rebecca Floyd, Ph.D., Medical College of Georgia at Augusta

University, Augusta, GA, 30912; [email protected]
S2032 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C140 6:00 PM-7:00 PM

HEALTH INFORMATION SEEKING AMONG SEXUAL MINORITY PEOPLE IN THE HEALTH

INFORMATION NATIONAL TRENDS SURVEY (HINTS)

Jennifer M. Jabson, PhD, MPH1, Joanne G. Patterson, MSW, MPH1, Charles Kamen, PhD, MPH2

1
University of Tennessee, Knoxville, TN; 2University of Rochester Medical Center, Rochester,
NY

Background. Seeking health information about health outcomes is associated with a reduction
in health risk behaviors, more informed healthcare decision making, and better health
outcomes. The general population increasingly uses the internet as a source for health and
cancer-related information. However, all subgroups within the population may not use the
internet at the same rate. Sexual minority people (SMP; i.e., lesbian, gay, and bisexual
people) are considered early adopters of technology and therefore may be more likely to
use the internet as a source of health and cancer-related information. However, there is no
published evidence concerning health information seeking behaviors among sexual minority
people as compared to heterosexual people (HP). The current project addresses this gap.

Methods. Data from the 2015 Health Information National Trends Survey (HINTS) FDA Cycle
were used to describe and summarize health and cancer-related information seeking among
SMP (n=103) and HP (n=1443). Differences in health and cancer-related information seeking
were tested using chi-squared tests.

Results. More SMP in this sample reported having access to the internet (92.4%) than HP
(79.4%; p=.001). SMP people were equally as likely as HP (59% vs 58%) to report
unintentional exposure to health information while online (p=.21). SMP used the internet for
a broad range of health information seeking activities. SMP used the internet most often to
seek health information for themselves (88.1%) and least often to participate in online
support groups (4.2%). The only statistically significant difference between SMP and HP was
the use of health related videos on YouTube; SMP (37.2%) were more likely than HP (22.5%)
to watch health-related videos on YouTube (p=.001). Multi-variable logistic regression did not
reveal differences by sexual orientation in information seeking.

Conclusion. SMP are a difficult-to-reach population that experience significant health

disparities. Our findings confirm that SMP access the internet at high rates and seek out
health information online. This provides empirical evidence that online interventions could be
a valuable way to address and reduce health disparities that affect SMP.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2033

CORRESPONDING AUTHOR: Jennifer M. Jabson, PhD, MPH, University of Tennessee, Knoxville,

TN, 37996; [email protected]
S2034 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C141 6:00 PM-7:00 PM

HOW ARE NON-NUMERICAL PROGNOSTIC STATEMENTS INTERPRETED AND ARE THEY

SUBJECT TO POSITIVE BIAS?

Samantha Siess, MA1, Anne Moyer, Ph.D.1, Sumeet Bhinder, BS2

1
Stony Brook University, Stony Brook, NY; 2Stony Brook University, Hopewell Junction, NY

Background: Frank, clear communication with surrogate decision makers for terminally-ill or
incapacitated patients has important implications for satisfaction with care and sound
decision making. Negative numerical prognostic statements have been found to be
interpreted in a positively-biased manner. Less precise non-numerical statements, particularly
those using threatening terms, may be even more subject to such biases.

Objective: This study examined whether there is an optimistic bias in the interpretation of
non-numerical prognostic statements and the extent to which this might vary by the severity
of the prognosis portrayed and the language used to convey them.

Methods: 200 participants interpreted 14 non-numerical prognostic statements about a

hypothetical patient with varying levels of risk framed in terms of dying or surviving. They
indicated what they believed the likelihood of survival to be, on a scale from a 0% to a 100%.

Results: The most negative prognostic statement, They will definitely not survive, was
interpreted significantly more optimistically (18.82%), in that it was significantly further from
its reference point of 0% survival, than the most positive prognostic statement, They will
definitely survive, (92.77%) was from its reference point of 100% survival, t(199) = 6.59, p <
.001. A 2 x 2 within-group ANOVA was used to examine the main effects of framing and
prognostic outcome in response to the statements It is very likely that they will die, It is
very unlikely that they will die, It is very likely that they will survive, and It is very unlikely
that they will survive. There was no main effect of framing, F(1, 199) = .119, p = .731, and no
interaction of framing and prognostic outcome, F(1, 199) = .66, p < .001. However, there was a
main effect of outcome, again suggesting that statements that conveyed a poorer prognosis
were more likely to be interpreted in a positively biased manner, F(1, 199) = 61.01, p < .001.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2035

Conclusions: In addition to using suggested strategies, such as inquiring about how much
individuals wish to know about prognoses, practitioners should be aware of the ways in which
commonly-used non-numeric language will be understood in numeric terms during prognostic
discussions, and check recipients understanding for potential positive bias.

CORRESPONDING AUTHOR: Samantha Siess, MA, Stony Brook University, Stony Brook, NY,
11794-2500; [email protected]
S2036 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C142 6:00 PM-7:00 PM

ILLUSTRATING CANCER RISK: DEVELOPMENT OF BREAST CANCER RISK COMMUNICATION

MATERIALS FOR A NOVEL BRCA1/2 GENETIC MODIFIER TEST

Margaux C. Genoff, B.S.1, Joy S. Westerman, B.A.1, Elyse Shuk, M.A.1, Jennifer L. Hay, Ph.D.1,
Kenneth Offit, M.D. , M.P.H.1, Mark E. Robson, M.D.2, Jada G. Hamilton, Ph.D., M.P.H.1

1
Memorial Sloan Kettering Cancer Center, New York, NY; 2Memorial Sloan Kettering Cancer
Center, New York City, NY

Background: Breast cancer risk estimates by age 70 range from 31-78% for BRCA1/2 carriers.
Recent research has identified a set of genetic variants that modify breast cancer risks among
female BRCA1/2 carriers. These variants will be used as part of a clinical genetic modifier test
(GMT) to help healthcare providers more accurately predict BRCA1/2 carriers breast cancer
risk. To prepare for the future availability of GMT, It is critical to develop effective strategies
for communicating this risk information to BRCA1/2 carriers.

Method: We conducted in-person interviews with 30 female BRCA1/2 carriers about their
perceptions and preferences regarding sample GMT risk communication materials.
Participants (aged 25-80 years; 87% Caucasian; numeracy M=4.8 on a scale ranging from 1-6)
were randomized to receive one of three sample GMT results (indicating low, moderate, or
high breast cancer risk). Participants were shown results in four different formats (panel of
genes, social comparison icon array, verbal risk estimate, and graphical risk estimate) and
asked how likely they would be to get breast cancer in their lifetime (rated from 1=very
unlikely to 5=very likely) in response to each format.

Results: The preferred format across the three breast cancer risk levels was the graphical
estimate, alone or in combination (often with the verbal estimate). Participants described
positive, neutral, and negative emotional responses to the formats; however, the icon array
evoked only negative emotions. The perceived lifetime risk of developing breast cancer
significantly differed among participants in the three groups in response to the verbal risk
estimate (Low M=3.06; Moderate M=3.75; High M=4.65; p < .001) and graphical risk estimate
formats (Low M=2.85; Moderate M=3.55; High M=4.33; p=.002); no differences across risk
groups were observed in response to the panel of genes or icon array formats. Overall,
interest in obtaining GMT was high, and the majority of participants reported that GMT
results would make their cancer risk management decisions easier.

Conclusion: The minimal amount of necessary information to include in GMT test results
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2037

appears to be graphical and verbal estimates of risk. Information presented in these formats is
acceptable to patients, influences their breast cancer risk perceptions, and may assist in
decision making about risk management. These findings will inform future efforts to translate
GMT into care consistent with patients preferences.

CORRESPONDING AUTHOR: Margaux C. Genoff, B.S., Memorial Sloan Kettering Cancer Center,
New York, NY, 10022; [email protected]
S2038 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C143 6:00 PM-7:00 PM

OVERCOMING BARRIERS TO A SCHOOL BMI SCREENING PROGRAM

Bryce Abbey, PhD, Kate Heelan, PhD, Todd Bartee, PhD

University of Nebraska at Kearney, Kearney, NE

Body Mass Index (BMI) screening in schools has been scrutinized due to privacy issues and
costs associated with communicating BMI screening results to parents. To overcome these
barriers a web application was developed to decrease the burden of staff training, easily
generate BMI Report Cards (BMI RC) and send them electronically to parents via email. The
current research compared parental satisfaction of BMI screening in schools and utilization of
BMI RC. Methods: In 2011, paper copies of BMI RC were sent via a sealed envelope addressed
to the parents of 2557 elementary school children. In 2015, emails were sent to parents of
2718 elementary school children that included their childs BMI RC. Parents responded to a
survey that was sent approximately ten days after delivery of the BMI RC. In 2011, 971
parents completed either a paper or email survey. In 2015, the survey was administered only
via email with 503 parents responding. The survey consisted of thirteen closed-ended
questions regarding BMI RC content, parent perceived importance of BMI screening in school,
and parent intention for behavior change. Results: The response rate in 2011 was 38.0% and
18.5% in 2015. Nearly all respondents (95%) reported they read all or most of the BMI RC in
both 2011 and 2015. More than three-quarters (76.4%) of respondents reported it was
somewhat or very important to include height and weight measurement as part of their
childs yearly health screening in 2015, compared to 88.4% in 2011. In 2011 and 2015,
respectively, parents with children equal to the 85th percentile or above reported that after
they had received the BMI RC that they had already or intended to create a better eating plan
at home (41.7%, 64.7%), decrease TV/video time (59.9%, 56.9%), and increase exercise or
physical activity (33.8%, 72.5%). The electronic calculation of BMI percentile decreased staff
time and electronic generation and distribution of BMI RC decreased printing costs by $0.15
per student. Discussion: Using email to send BMI RC allowed schools to address concerns
through protecting children and peers from seeing the BMI RC. In both survey years, parents
reportedly read all or most of the BMI RC, which included nonjudgmental language about
student BMI percentiles and their use as a screening tool, importance of lifestyle changes for
better health, suggested actions for parents, as well as, local and national resources on
healthy lifestyles. Parents also continued to respond that it is somewhat or very important to
include height and weight as part of yearly health screenings. These findings are consistent
with past research suggesting when schools communicate BMI screening results with parents
they are more likely to support healthy lifestyle changes for their children. Schools who utilize
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2039

a BMI RC as part of their BMI Screening program should consider safeguards and cost saving
methods such as email-based communication.

CORRESPONDING AUTHOR: Bryce Abbey, PhD, University of Nebraska at Kearney, Kearney,

NE, 68849; [email protected]
S2040 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C144 6:00 PM-7:00 PM

PATIENT REPORTS ABOUT COMMUNICATION WITH THEIR PHYSICIAN AND ASSOCIATIONS

WITH PATIENT ACTIVATION

Jorie Butler, PhD1, Bryan Gibson, PhD2, Man Hung, PhD2, Marjorie Carter, MSPH1, Candace
Haroldsen, MS3, Matthew Samore, MD2

1
University of Utah, SALT LAKE CITY, UT; 2University of Utah, Salt Lake City, UT; 3University of
Utah, Salt Lake ity, UT

Patient activation involves 4 developmental stages 1) patient recognition of the importance of

their vital role in health management 2) patient development of confidence and acquisition of
information necessary to be an active participant in health management 3) actively
participating in health actions, and 4) maintenance of active health management despite
challenges, such as stress. The relationship and communication between physicians and
patients may play an important role in supporting patients in the process of becoming
activated: patients with positive interactions with their providers may be more likely to
develop the necessary skills and tenacity to maintain active involvement in health
maintenance. Understanding patient-provider communication factors that contribute to
activation is important to support patients in developing greater activation. A survey of 810
patient participants receiving care in primary care clinics in rural communities was conducted
to understand patient experience in primary care clinics. Patients with at least one chronic
illness were more heavily recruited to ensure that participants had a need for regular follow-
up care and interaction with their provider. The participants were likely to report a chronic
illness (79%) and were largely female (61%, 39% male) with 42% over age 65 and mean
education level above high school graduation. In a telephone survey, participants were asked
to relate their experiences with their primary care physician over the preceding twelve
months. Participants reported on two communication areas - how often their physician
listened carefully to them and whether their physician was helpful when talking with the
patient about problems or concerns. Participants also completed the 13-item Patient
Activation Measure (PAM). A regression model examining patient reports of activation
demonstrated that age, and two items about communication with their provider were related
to patient activation whereas patient gender was not. Patient reports that the doctor really
listened ( = .10; p .05) was not ( F (4, 600) = 13.10; p < .01). Results suggest that sensitive
communication style from a primary care physician may be an important contributor to
patient activation. Implications include interventions to support providers in developing
communication skills to foster patient activation particularly for older patients.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2041

CORRESPONDING AUTHOR: Jorie Butler, PhD, University of Utah, SALT LAKE CITY, UT, 84102;
[email protected]
S2042 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C145 6:00 PM-7:00 PM

THE PROVIDER-PATIENT RELATIONSHIP AND SATISFACTION WITH CARE AMONG HIGH AND
LOW ACCULTURATED LATINA BREAST CANCER SURVIVORS

Dara H. Sorkin, PhD1, Senxi Du, Undergraduate Student2, Kelly Biegler, PhD3, Yunan Chen,
PhD3, Alfred Kobsa, PhD3, Alice Police, MD3, Mienah Sharif, PhD3, Dana B. Mukamel, PhD3,
Sadeeka Al-Majid, PhD4

1
University of California Irvine, Irvine, CA; 2University of Chicago, Irvine, CA; 3UC Irvine, Irvine,
CA; 4California State University, Fullerton, Fullerton, CA

Due to advances in breast cancer detection and treatment, the average 5 year survival rate is
89%. Thus, efforts to improve health outcomes during survivorship are a high priority.
Converging evidence suggests that the quality of the provider-patient relationship is not only
associated with satisfaction with care, but also associated with improved health outcomes
and better quality of life.

Important factors in the provider-patient relationship have been shown to include

participatory decision making (PDM), trust in the physician, and being treated as an equal
partner. Several recent studies have identified disparities in PDM with Latina patients
exhibiting the lowest rates of PDM and poorest physician-patient communication; however,
this group generally has the highest ratings of satisfaction with care. This paradoxical finding
may be attributed to the fact that prior studies have not considered acculturation factors,
including English language proficiency, which are likely to impact physician-patient
communication and patients confidence in initiating participation in treatment decision
making.

Objective: This study aimed to address this gap in the literature by examining: 1) Associations
among three factors important to the provider-patient relationship: PDM, trust in physician,
and being an equal partner, 2) Associations between these qualities and satisfaction with
care, and 3) Whether acculturation differences moderate the association between these
qualities of the provider-patient relationship and satisfaction with care.

Method: Data were derived from 70 surveys completed by Latina breast cancer survivors.

Results: PDM scores were positively correlated with trust (r = 0.48, p < 0.001) and satisfaction
with care (r = 0.39, p < 0.001), irrespective of acculturation status. In higher acculturated
women, lower trust ratings were associated with lower satisfaction with care; however,
associations between trust and satisfaction with care were not observed in women with lower
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2043

acculturation status (F (1,66) = 5.43, p=0.02). Being treated as an equal partner was not
significantly associated with ratings of satisfaction with care in either of the acculturation
groups of our sample.

Discussion: High-acculturated patients may have greater familiarity with the healthcare
system, and thus have adapted expectations that are in-line with those of non-white Hispanic
women. Low-acculturated patients, in contrast, may not have the same expectations of trust
for contributing to their ratings of care. Implications for providing high-quality care for
diverse women will be discussed.

CORRESPONDING AUTHOR: Dara H. Sorkin, PhD, University of California Irvine, Irvine, CA,
92697; [email protected]
S2044 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C146 6:00 PM-7:00 PM

THE RELATIONSHIP BETWEEN FORGIVENESS, WELL-BEING AND HEALTH-INFORMATION

SEEKING BEHAVIOR IN A SAMPLE OF LGBTQ INDIVIDUALS

Victoriya Vakhotina, M.S1, Asani H. Seawell, PhD2

1
Pacific University Oregon, West Hollywood, CA; 2Pacific University, Hillsboro, OR

Forgiveness has been found to be beneficial to human health in previous research, suggesting
that this construct may be protective of health. Well-being also appears to promote health as
those who are lower on life satisfaction tend to have poor to fair general health and that life
satisfaction was related to positive health outcomes. Forgiveness and well-being both tap into
the health-information seeking behavior (HISB) construct as part of the overall health
construct. Previous studies regarding HISB have mostly looked at the general population;
therefore, the purpose of this study was to see if a relationship exists between forgiveness,
well-being and HISB in a minority (LGBTQ) population.

Participants (N = 136) for the study were recruited through social media and the Q Center a
safe space for Portlands LGBTQ community - and asked to take an online survey through
Qualtrics, a survey-building website. The survey consisted of a consent form, demographic
information, Heartland Forgiveness Scale, Flourishing Scale, Krantz Health Opinion Survey and
a thank you note. Participants were asked to complete the survey in one sitting to ensure
privacy and anonymity and took less than 30 minutes to complete.

Multiple regression analysis found marginal significance between well-being and behavioral
HISB (b = .044, SE = .023, p = .059) with higher well-being scores indicating less behavioral
involvment in healthcare needs. Situational forgiveness and informational HISB was significant
(b = .135, SE = .059, p < .05) as situational forgiveness increased, information seeking declined.
Finally, well-being and informational HISB was significant (b = -.086, SE = .20, p < .01)
indicating that LGBTQ individuals who scored higher on well-being were more likely to be
actively involved in seeking out information regarding their healthcare needs.

These findings reveal that a relationship exists between forgiveness, well-being and HISB in
this sample; therefore, it would be prudent to utilize this connection in future studies as tools
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2045

to better combat health disparities in the LGBTQ population, as previous studies denote that
this community does not seek out health information and treatment on par with the general
population due to prejudice and discrimation.

CORRESPONDING AUTHOR: Victoriya Vakhotina, M.S, Pacific University Oregon, West

Hollywood, CA, 90046; [email protected]
S2046 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C147 6:00 PM-7:00 PM

TOBACCO INFORMATION SEEKING AND SOURCES OF HEALTH INFORMATION AMONG

RACIAL/ETHNIC GROUPS

Anh Nguyen, PhD, MPH1, Joelle Robinson, MPH2, Erin Keely. OBrien, Ph.D.3, Xiaoquan Zhao,
Ph.D.4

1
Center for Tobacco Products/ FDA, Silver Spring, MD; 2U.S. Food and Drug Administration
Center for Tobacco Products, Silver Spring, MD; 3Center for Tobacco Products, U.S. Food and
Drug Administration, Silver Spring, MD; 4George Mason University, Fairfax, VA

Introduction

Racial/ethnic differences in tobacco use may be explained by different levels of access and use
of tobacco-related health information. The objectives of the study are to describe the
different sources of health information and the different types of tobacco-specific information
that are sought after among specific racial/ethnic groups.

Methods

We analyzed cross-sectional data from HINTS-FDA 2015, a nationally representative survey of

non-institutionalized U.S adults. Weighted bivariate analyses assessed differences in first
sources of health information and types of tobacco information sought among racial/ethnic
groups. Racial/ethnic groups examined included non-Hispanic (NH) White, NH-Black,
Hispanics, NH-Asian and Pacific Islander, and NH-Other.

Results

A significantly higher proportion of NH-Whites (13%) compared to NH-Asians and Pacific

Islanders (4%) reported ever searching for information related to cost/coupons for tobacco
products.

Among NH-Other, a higher proportion of respondents who sought health information from
print media and traditional sources (59%) or the Internet (34%) reported searching for
information related to the health effects of tobacco compared to those who sought
information from interpersonal health sources ( < 1%).

Among NH-Blacks, respondents who sought health information from print media and
traditional sources had a significantly lower proportion of those who ever searched for
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2047

information related to quitting (1%) compared to those who sought information from the
Internet (12%) or the interpersonal health sources (30%). A similar same pattern was found
among Hispanics with regard to information related to cost/coupons: print media and
traditional sources ( < 1%), Internet (7%), and interpersonal health sources (26%).

Among NH-Blacks, non-current tobacco users had a significantly higher proportion of those
who sought health information from the Internet (70%) compared to those who sought
information from print media or traditional sources (12%) or interpersonal health sources
(18%).

Conclusions

Information seeking behavior varies by information type and source of the information across
racial/ethnic groups. The current research suggests that different strategies, such as culturally
tailoring messaging and leveraging the use of preferred information channels, could more
effectively reach specific racial/ethnic groups.

CORRESPONDING AUTHOR: Anh Nguyen, PhD, MPH, Center for Tobacco Products/ FDA, Silver
Spring, MD, 20903; [email protected]
S2048 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C148 6:00 PM-7:00 PM

PENNSYLVANIA POLICYMAKERS ATTITUDES & KNOWLEDGE REGARDING WATERPIPE

TOBACCO SMOKING & ELECTRONIC NICOTINE DELIVERY SYSTEMS

Beth L. Hoffman, BSc1, Megan C. Tulikangas, MPP2, Ariel Shensa, MA1, A. Everette James, JD,
MBA1, Jason B. Colditz, MEd1, Brian A. Primack, MD, PhD1

1
University of Pittsburgh, Pittsburgh, PA; 2Allegheny County Health Department, Pittsburgh,
PA

Introduction. While youth smoking of traditional cigarettes is decreasing, waterpipe tobacco

smoking (WTS, also known as hookah smoking) and use of electronic nicotine delivery systems
(ENDS, such as e-cigarettes) are rapidly increasing. However, most current nicotine and
tobacco regulations, do not apply to WTS and ENDS. Furthermore, legislatures have been slow
to update policy measures related to them. Therefore, we sought to assess knowledge of and
attitudes toward WTS and ENDS among members of the Pennsylvania legislature.

Methods. We approached all standing members of the key health and welfare committees for
both the Pennsylvania House and Senate. Recognizing that this is a notoriously difficult
population to assess, we developed a complex protocol that allowed completion via mail,
email, or fax. To assess knowledge, we asked participants to self-report familiarity with these
products using a visual analogue scale. We also asked 14 factual items related to WTS and
ENDS; items were based upon recent data reports including the 2014 National Youth Tobacco
Survey. To assess attitudes, we included 5 items based upon current pending legislation
concerning substance regulation. Additionally, participants ranked different substances of
abuse from highest to lowest legislative priority. We also included four open-ended items, and
three independently working researchers analyzed responses to these items using thematic
analysis.

Results. We received responses from 13 of 25 eligible individuals (52%). On a scale from 0 (not
at all familiar) to 100 (very familiar), participants ranked their familiarity with WTS and ENDS
as 28 (SD=15) and 50 (SD=13), respectively. Participants answered a mean of 27% (SD=20) of
knowledge items correctly. With regard to legislative priority of substance abuse, WTS ranked
eighth (least urgent) and ENDS ranked fifth. Thematic analysis revealed that the most
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2049

common sentiment was that policy makers were eager for additional information and
guidance around these substances.

Conclusion. Despite the fact that tobacco use remains a leading cause of preventable death
worldwide, lawmakers exhibit a lack of knowledge concerning newer forms of tobacco and
nicotine and consider them to be relatively low legislative priorities. However, respondents
did express a desire for more information about these substances, suggesting the potential for
public health entities to promote effective policy development via improved dissemination of
information.

CORRESPONDING AUTHOR: Beth L. Hoffman, BSc, University of Pittsburgh, Pittsburgh, PA,

15213; [email protected]
S2050 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C149 6:00 PM-7:00 PM

DEVELOPING A BIOBEHAVIORAL HIV INTERVENTION FOR HIGH-RISK DRUG USERS: TAKING

INTO ACCOUNT IMPACT OF NEUROCOGNITIVE IMPAIRMENT

Roman Shrestha, MPH1, Pramila Karki, RN2, Frederick Altice, PhD3, Michael Copenhaver, PhD2

1
University of Connecticut Health Center, Willimantic, CT; 2University of Connecticut, Storrs,
CT; 3Yale University, New Haven, CT

Introduction: There is an urgent need for innovative approaches to the prevention of HIV
transmission. Innovative combination approaches can be comprised of both evidence-based
biomedical and behavioral interventions while incorporating strategies to accommodate a
wide range of neurocognitive deficits (NCI) for many people who use drugs (PWUDs). We
therefore conducted formative research aimed at systematically adapting and combining
evidence-based interventions that integrates HIV risk reduction, strategies to accommodate
cognitive deficits, and encourage uptake and adherence to pre-exposure prophylaxis (PrEP) to
implement among drug users in treatment.

Methods: Between May and June 2016, we conducted structured focus group sessions with
treatment providers (n=10) and members of the target population (n=20) who were enrolled
in a community-based methadone maintenance program in New Haven, Connecticut. Data
analysis followed a thematic analysis approach utilizing several qualitative data analysis
techniques, including inductive analysis and cross-case analysis.

Results: Given the parameters that were suggested, it seemed appropriate to refine the CHRP
intervention an evidence-based behavioral intervention that has demonstrated efficacy in
reducing drug- and sex-related HIV risk behaviors by focusing primarily on a range of
relevant topics pertaining to reducing sex- and drug-related HIV risk behaviors. We also
concluded that greater emphasis would be given to certain strategies to accommodate
participants with NCI in order to help them to better concentrate, learn, and remember
details. Additionally, it was noted that the revised intervention should cover content specific
to PrEP, including basics about PrEP, potential motivation driving PrEP use (e.g., pros and cons
of achieving high levels of adherence to PrEP), problem solving (e.g., improving strategies for
identifying and overcoming obstacles to adherence), facilitators to PrEP adherence (e.g.,
learning memory aids for improving adherence), enhancing decision-making related to PrEP,
and overcoming stigma related to being on PrEP.

Conclusion: The current study details the formative research process in preparation to adapt
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2051

and to enhance an EBI taking into account both published empirical evidence and input from
target population and treatment providers for use with high risk PWUDs and are in drug
treatment. The resulting bio-behavioral intervention, the CHRP-NCI, is designed to address the
HIV-related risk behaviors and PrEP uptake and adherence as experienced by many the drug
users in treatment. We hope that the process and outcome of this formative research will
help to inform similar work in the future as a growing number of EBIs have become widely
available, but may not yet be in optimal form for implementation among individuals with
neurocognitive issues.

CORRESPONDING AUTHOR: Roman Shrestha, MPH, University of Connecticut Health Center,

Willimantic, CT, 06226; [email protected]
S2052 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C150 6:00 PM-7:00 PM

FOOD INSECURITY, DEPRESSION, AND ALCOHOL USE IN PERSONS LIVING WITH HIV

Dominica Hernandez, PhD1, Seth Kalichman, PhD2

1
University of Connecticut, Storrs, CT; 2University of Connecticut, Institute for Collaboration
on Health, Intervention, and Policy (InCHIP), Storrs, CT

Background: Food insecure persons living with HIV/AIDS (PLWHA) suffer desperate economic
circumstances and are at a higher risk for poorer mental health outcomes, which in turn may
lead to maladaptive coping behaviors, such as alcohol use. Due to a compromised immune
system and the need to be on anti-HIV medication in order to suppress HIV replication, the
use of any alcohol places PLWHA at greater health risks. Furthermore, the association of food
insecurity and Alcohol use may be co-facilitated by depression.

Aims: The aim of this study was to assess the impact of depression on alcohol use in food
insecure PLWHA.

Methods: Analyses were carried out on a 171 PLWHA in Atlanta, Georgia. Participants were
recruited as part of an intervention trial aimed at decreasing alcohol use in PLWHA. Baseline
data was collected at enrollment where participants were consented and asked to complete
an anonymous survey that included the Center for Epidemiologic Studies Short Depression
Scale (CESD-10) and the Alcohol Use Disorders Identification Test (AUDIT) as well as a measure
of food security.

Results: Our sample was primarily Black (94%), 64% were male and 36% female, mean age
was 47.2 years, SD = 9.3. A mediation model was constructed to assess the affects of
depression on food insecurity and alcohol use, with AUDIT score as the continuous outcome
variable, CESD-10 score as the continuous mediator variable, and food insecurity as a
dichotomous predictor. The model displayed a statistically significant indirect effect as
expressed by a 95% confidence interval that is above zero (CI = .352 1.82). The indirect
effect showed that relative to participants that were food secure, participants with food
insecurity scored, on average, .97 units higher on the AUDIT as a result of the effect of food
insecurity on depression, which, in turn, affected their alcohol use. The direct effect was not
statistically significant, which suggests that when accounting for depression there was no
effect on food security in their AUDIT score.

Conclusion: Our findings suggest depression mediates the use of alcohol in food insecure
PLWHA; however, results may not be generalizable to other PLWHA. Even so, these results
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2053

suggest important findings that should be further explored and considered in future
intervention development for PLWHA living in poverty.

CORRESPONDING AUTHOR: Dominica Hernandez, PhD, University of Connecticut, Storrs, CT,

06237; [email protected]
S2054 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C151 6:00 PM-7:00 PM

PATIENT REPORT VS OBJECTIVE MEASURES IN PREDICTING EVERYDAY FUNCTIONING

OUTCOMES IN HIV+ ADULTS

Anne Heaton, N/A1, Raeanne C. Moore, PhD2, Emily Paolillo, B.S.3, Alexandra Rooney, B.A.4,
Dilip J. Jeste, MD5, David J. Moore, Ph.D.6

1
UCSD HIV Neurobehavioral Research Program, Del Mar, CA; 2University of California, San
Diego, San Diego, CA; 3SDSU/UCSD Joint Doctoral Program in Clinical Psychology, San Diego,
CA; 4University of California San Diego, San Diego, CA; 5UCSD Stein Institute for Research on
Aging, La Jolla, CA; 6UCSD HIV Neurobehavioral Research Program, San Diego, CA

Functional disability is a common and growing problem in older HIV+ adults. Due to the new
generation of antiretrovirals, HIV+ individuals are living longer, producing an emerging need
for researchers to identify and treat real-world functioning problems. This study assesses
ways to predict such problems in older HIV+ adults by evaluating functional capacity and
cognitive complaint constructs as they relate to everyday functioning. One hundred HIV+
adults (Mean age=50 years; range: 35-65 years) completed a comprehensive
neuropsychological and neuromedical battery, including a measure of functional status (IADL
Dependence vs. IADL Independence), an objective measure of functional capacity (UCSD
Performance-Based Skills Assessment-Brief [UPSA-B]) and a self-report measure of cognitive
complaints (Profile of Mood States [POMS]-Confusion/Bewilderment subscale). Higher UPSA-
B scores were related to better global neuropsychological ability, but unrelated to functional
independence (p>0.05). Conversely, while POMS scores were unrelated to global
neuropsychological ability or the UPSA-B, lower POMS scores were significantly associated
with functional independence (pp < 0.001) than the area of no information (AUC of 0.50). A
cutoff score of 7.5 best discriminated between functional dependence vs. functional
independence, with a sensitivity of 73.1% and specificity of 71.3%. This study found that the
POMS demonstrates good support as a brief screening tool for functional dependence among
adults living with HIV. Functional disability among adults HIV+ is still not well understood and
is likely multifactorial in nature. Self-appraisals of cognitive status may be just as important (or
perhaps even more so) than actual cognitive or functional ability in understanding real-world
problems in everyday functioning.

CORRESPONDING AUTHOR: Anne Heaton, N/A, UCSD HIV Neurobehavioral Research Program,
Del Mar, CA, 92014; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2055

C152 6:00 PM-7:00 PM

PERCEIVED RISK OF HIV TRANSMISSION APPRAISAL AMONG MSM WHO ARE PREP AWARE
VERSUS PREP UNAWARE

Redd Driver, BA1, Devon Price, MS1, Stephanie D. Finneran, MA2, Seth Kalichman, PhD1

1
University of Connecticut, Storrs, CT; 2University of Connecticut, Vernon, CT

New diagnoses of HIV in Men who have sex with men (MSM) are 44 times that of men who do
not have sex with men (CDC, 2015). Current HIV treatments, daily-dosed antiretroviral
therapy (ART), reduce the amount of HIV in the body preserving immune functions.
Additionally, ART serves as a prevention strategy, treatment as prevention (TasP), preventing
transmission to others (CDC, 2015). Although TasP is efficacious in preventing transmission, it
remains unclear how MSM perceive the relative risks of condomless receptive anal sex with a
partner who is using TasP. Moreover, the perceived risk of sex with an HIV positive partner
using TasP may be associated with awareness of using ART as pre-exposure prophylaxis
(PrEP). PrEP is the use of a daily dose of ART in HIV negative individuals to prevent infection if
exposed to HIV. Previous research has indicated that a central predictor of attending to health
information and engaging in health behavior is perceived risk (Gerard et al., 1996). Therefore,
perceptions of sexual risks with a partner using Tasp may be related to knowledge of PrEP.
The present study examined whether perceptions of risk associated with engaging in
condomless anal sex with a TasP-using partner generalizes across MSM who are aware of PrEP
and those who have no prior knowledge of PrEP.

Participants (267 MSM) were surveyed during the 2015 Atlanta Pride Festival. Surveys
assessed demographics, sexual behaviors, and health care experiences. Risk perceptions were
assessed through four 5-point measures under two condition; (1) condomless anal with an
HIV positive partner using TasP, and (2) condomless anal sex with an HIV positive partner not
receiving treatment. For each condition participants rated risks for both receptive and
insertive anal sex.

Results demonstrated that MSM who were aware of PrEP reported significant differences in
receiving health care in the past year, and being open about their sexual orientation
Additionally, PrEP aware MSM did not report a significant difference in engaging in
condomless receptive anal sex compared to PrEP unaware MSM, p =.96. Regarding perceived
risk of engaging in condomless anal sex, PreP aware MSM reported significantly greater risk
perceptions of engaging in the highest risk behavior, condomless receptive anal sex with an
HIV positive partner not receiving treatment, than those who were not aware of PrEP p = .02.
However, risk perceptions of engaging in lower risk behavior, condomless insertive anal sex
S2056 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

with an HIV positive partner using TasP, were not significantly different for PrEP aware MSM
compared to PrEP unaware MSM p = .09. These findings suggest that PrEP awareness is
associated with HIV TasP risk perceptions reflecting a realistic view of risks, whereas
unawareness of PrEP may be related to misperceptions of risk related to TasP. These results
have implications for educating MSM more generally about HIV prevention treatments to
increase PrEP uptake.

CORRESPONDING AUTHOR: Redd Driver, BA, University of Connecticut, Storrs, CT, 06269-
1020; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2057

C153 6:00 PM-7:00 PM

SELF-EFFICACY MEDIATES THE ASSOCIATION OF DEPRESSION TO ANTIRETROVIRAL

MEDICATION ADHERENCE AMONG HIV-POSITIVE OUTPATIENTS

Jacklyn Babowitch, B.S.1, Peter A. Vanable, PhD1, Shannon M. Sweeney, PhD2

1
Syracuse University, Syracuse, NY; 2Le Moyne College, Kirkville, NY

Adherence to prescribed antiretroviral therapy (ART) is a central feature in the clinical

management of HIV. Although several studies have documented the negative effects of
depression on ART adherence, little is known about the mechanisms underlying this
association. In this study, we test the hypothesis that medication-taking self-efficacy will
mediate the relationship of depression to medication adherence utilizing a sample of HIV-
positive outpatients recruited during routine care (N = 206). Participants (M age = 46.5; 62%
male) completed the Center for Epidemiological Studies Depression Scale (CES-D), as well as
self-report measures of ART adherence for the previous week. Across the sample, 62%
endorsed clinically significant depressive symptomatology on the CES-D. Nevertheless,
participants reported missing few antiretroviral doses in the past week (M = 0.10, SD = .26).
Mediation analyses confirmed that depressive symptoms were negatively associated with
medication-taking self-efficacy (b = -1.20, SE = .29, p < .001), and there was a significant
correlation of medication-taking self-efficacy with the number of missed antiretroviral doses
for the past week (b = -.06, SE = .01, p < .001). Although depressive symptoms were not
directly associated with ART adherence (b = -.03, SE = .04, p = .38), a test of the indirect effect
of depressive symptoms on ART adherence via medication-taking self-efficacy was significant
(b = .07; bootstrap 95% CI: .03, .14). Thus, findings support the hypothesis that medication-
taking self-efficacy mediates the association between depressive symptoms and ART
adherence assessed for the previous week. Behavioral interventions to enhance self-efficacy
may benefit from the inclusion of content focusing on alleviating depressed mood as a means
of enhancing ART adherence.

CORRESPONDING AUTHOR: Jacklyn Babowitch, B.S., Syracuse University, Syracuse, NY, 13244;
[email protected]
S2058 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C154 6:00 PM-7:00 PM

VARIATION IN THE FORMS AND SOURCES OF SOCIAL SUPPORT FOR CARE-SEEKING ACROSS
THE HIV CARE CONTINUUM IN THE RURAL SOUTH

Miranda Iverson Hill, MPH1, Amber Huff, PhD2, Neale Chumbler, PhD3

1
University of Georgia, Athens, GA; 2Institute of Development Studies, University of Sussex,
Brighton, East Sussex, England, United Kingdom; 3Western Kentucky University, Bowling
Green, KY

In spite of progress towards understanding the importance of social support for improved
health outcomes in Persons Living with HIV (PLWH), more remains to be known about which
sources and forms of support are most beneficial at each stage of HIV treatment. In this study,
we use a qualitative analytic approach to investigate the mechanisms of social support that
have been deemed most integral to the HIV diagnosis, care linkage, engagement, and
retention behaviors of a diverse sample of rural PLWH (N=18). The in-depth interviews used
for this study were collected during the qualitative phase of a larger mixed methods needs
assessment for the Northeast Georgia health district. A deductive-inductive analysis of
participant narratives revealed variation in the perceived importance of particular forms and
sources of social support during the initial versus advanced stages of HIV care. Participants
identified the emotional, informational, and appraisal support provided by family as critical
for emotional stability, coping, and care linkage during the initial stages of diagnosis and
treatment. Once in care, participants emphasized informational and instrumental forms of
support from care providers and appraisal support from peers as key influences in care
engagement and retention. Increased understanding of how social support contributes to the
HIV treatment behaviors of PLWH can fill important knowledge gaps in research and inform
the efforts of health care providers seeking to leverage HIV patients social networks towards
improving treatment and health.

CORRESPONDING AUTHOR: Miranda Iverson Hill, MPH, University of Georgia, Athens, GA,
30602; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2059

C155 6:00 PM-7:00 PM

BEHAVIORAL HEALTH NEEDS ASSESSMENT OF AN UNDERSERVED URBAN FAMILY MEDICINE

CLINIC

Zach Radcliff, M.S.1, Allison Baylor, B.A.1, Sarah Braun, MS1, Mark Ryan, MD2, Bruce
Rybarczyk, PhD1

1
Virginia Commonwealth University, Richmond, VA; 2Department of Family Medicine and
Population Health, Virginia Commonwealth University School of Medicine, Richmond, VA

Title: Behavioral Health Needs Assessment of an Underserved Urban Family Medicine Clinic

Objective: Family medicine clinics are a common access point for mental health care for
underserved populations. These patient groups often have limited access and other barriers
to receiving adequate mental health services, which can place the burden on physicians to
provide extensive behavioral healthcare. In 2015, integrated behavioral healthcare (IBHC)
services were added to a safety net family medicine clinic to address these needs. Currently,
psychology doctoral students provide pro-bono, IBHC sessions at this clinic. The present study
describes this population and delineates behavioral health needs of this vulnerable
population.

Methods: As part of regular patient monitoring, English or Spanish versions of the Patient
Health Questionnaire-9 (PHQ-9), the Generalized Anxiety Disorder-7 (GAD-7), and Pediatric
Symptom Checklist-17 (PSC-17) are administered to screen for depression and anxiety in
adults and externalizing/internalizing behaviors in children, respectively. Presenting problems
and session length/type are also documented daily.

Results: The family medicine clinic provided services to more than 3,500 unique patients
resulting in 10,000 medical visits annually. The population was 58.3% African American, 24.2%
Caucasian, 22.9% Hispanic/Latino, and 17.5% other race/ethnicity. Further, 28.8% of patients
were

In the initial year of IBHC clinic, 687 unique patients (19.4% of the clinic population) were
provided IBHC services for 1560 sessions. The PHQ-9 and GAD-7 for an initial sample of 105
patients (64% African American, 22% Caucasian, 7% Hispanic/Latino, 7% other) indicated that
those who are referred to the IBHC team have moderately severe levels of depression
(m=15.2) and moderate levels of anxiety (m=13.1). Descriptive statistics for presenting
problems, length of sessions, and session interventions will be provided.
S2060 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusion: This sample provides a unique view of the needs of a largely minority urban
population with limited access to mental healthcare. Early indications show that patients
referred to the IBHC team have significant behavioral healthcare needs that were likely unmet
prior to initiation of these critical services.

CORRESPONDING AUTHOR: Zach Radcliff, M.S., Virginia Commonwealth University, Richmond,

VA, 23226; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2061

C156 6:00 PM-7:00 PM

PATIENT RETENTION: THE KEY TO SUCCESSFUL SHARED MEDICAL APPOINTMENTS.

Ruth Nutting, MA

Via Christi, Wichita, KS

Shared medical appointments (SMAs) optimize patient and provider time through a team
based, interprofessional approach. SMAs have been shown to reduce hemoglobin HbA1C,
blood pressure, and LDL cholesterol; boost adherence to ADA standards of care; and decrease
hospitalization rates. Although research has demonstrated that SMAs improve health related
outcomes, literature also indicates that poor retention rates of patients have affected health
related outcomes. We undertook a mixed-methods study to explore why our SMA program
had a high rate of retention and achieved success.

This study took place at a Joint Commission certified Primary Care Medical Home. Patients
with poorly controlled Type 2 diabetes, determined as HbA1C greater than 8.0% were invited
to participate in the project. A total of 21 patients chose to participate. Four SMA group
cohorts were formedparticipants joined whichever meeting time was convenient for them.
Ultimately, group sizes ranged from 4-6 participants. In order to maintain cohesion, new
patients were added to these groups once per year.

At the end of year one, 18 of the 21 participants (86%) who attended the first SMA continued
to participate. At this time, we performed a quantitative pre/post study of clinical outcomes
and survey responses. Quantitative tools included the Patient Health Questionnaire-9, the
Diabetes Lifestyle Survey, and the Full Circle Diabetes Program. A1C and other biometric
screenings were analyzed. Qualitative investigation included four semi-structured interviews
that assessed: barriers patients faced before joining SMAs; how patients increased control
over diabetes; changes in mood; patients overall experiences of SMAs; perceptions of the
value of SMAs; and patients confidence in sustaining learned coping behaviors. These
interviews were held at the conclusion of the year two.

Quantitative and qualitative findings depicted patients experiencing increased psychosocial

wellbeing and improved adherence to treatment regimens. Qualitative findings highlighted
that SMAs reduced barriers to self-management, through the accessibility of resources
provided by the inter-professional team.

First, the presenter will highlight the implications of SMAs. Second, the presenter will discuss
S2062 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

the findings of this study. Third, the presenter will highlight the importance of retention in
SMAs and how to achieve this.

CORRESPONDING AUTHOR: Ruth Nutting, MA, Via Christi, Wichita, KS, 67202;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2063

C157 6:00 PM-7:00 PM

PROVIDER PRACTICES AND PATIENT PERSPECTIVES REGARDING THE MANAGEMENT OF

MULTIMORBIDITY AMONG OVERWEIGHT AND OBESE VETERANS

Laura J. Buchholz, Ph.D.1, Jennifer S. Funderburk, PhD2

1
VA Center for Integrated Healthcare, Buffalo, NY; 2VA Center for Integrated Healthcare,
Syracuse, NY

Roughly 77% of veterans are overweight or obese (Kahwati, Lance, Jones, & Kinsinger, 2011).
Among primary care patients, overweight and obesity is associated with a greater likelihood
of multimorbidity (Booth, Prevost, & Guilford, 2014), defined as the prevalence of one or
more conditions (Valderas et al., 2009). Because multimorbidity is associated with increased
cost of healthcare expenses, increased mortality, and greater healthcare utilization (Wolff,
Starfield, & Anderson, 2002), it is imperative for research to identify facilitators and barriers to
reduce the prevalence and consequences associated with multimorbidity. To date, existing
research has not examined primary care provider practices in the management of
multimorbidity among patients who are overweight and obese, in conjunction with patients
perspectives regarding these practices. The purpose of this study is to examine primary care
provider referral practices for multimorbidity among a sample of overweight and obese
veterans. A secondary goal of this study is to examine patients perspectives regarding the
management of multimorbidity. We examined chart data and performed telephone-based
interviews for overweight and obese primary care patients (N = 45) who screened positive for
four annual behavioral health screens (i.e., at-risk drinking, tobacco use, depression, and
PTSD). On average, veterans had 3.6 (SD = 1.34) comorbid conditions, and were prescribed
3.4 (SD = 2.77) medications. Results indicated primary care providers were more likely to
place behavioral treatment referrals for smoking cessation (60%), depression (44.4%), weight
management (44.4%), and PTSD (26.7%), as compared to alcohol use (15.6%) and sleep
(8.9%). Meanwhile, veterans were more likely to attend referrals for PTSD (50%), depression
(45.0%), alcohol use (28.6%), and sleep (25.0%), as compared to weight management (10.0%),
and smoking cessation (3.7%). These findings highlight the need for using a multidisciplinary
treatment approach amongst members of the primary care team in the management and
treatment of patients with multimorbidity. Limitations and future directions will be discussed.

CORRESPONDING AUTHOR: Laura J. Buchholz, Ph.D., VA Center for Integrated Healthcare,

Buffalo, NY, 14215; [email protected]
S2064 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C158 6:00 PM-7:00 PM

SELF-MANAGEMENT STRATEGIES FOR STRESS AND ANXIETY USED BY NON-TREATMENT

SEEKING PRIMARY CARE PATIENTS

Robyn L. Shepardson, PhD, Jennie Tapio, MA, Jennifer S. Funderburk, PhD

VA Center for Integrated Healthcare, Syracuse, NY

The majority of individuals with anxiety disorders and subthreshold symptoms do not seek or
receive formal mental health treatment. One of the most common reasons for this is a
preference to manage emotional concerns on ones own. Self-management refers to the
strategies or techniques that individuals use on their own (i.e., without professional guidance)
to manage their symptoms. If promoted as the first step in a stepped care model of primary
care treatment for subthreshold and mild anxiety symptoms, self-management holds great
potential for prevention and early intervention. However, little research has examined self-
management for anxiety. Therefore, the objectives of this study were to describe patients
anxiety self-management strategies, identify which types of strategies were perceived to be
effective, and explore potential correlates. Participants were 182 non-treatment seeking
Veteran primary care patients (M age=58.3 years, SD=14.9) who reported current anxiety
symptoms (8 on the Generalized Anxiety Disorder-7, M=14.2, SD=3.7). Self-management
strategies were assessed via telephone and coded by two independent raters into one of 7
categories (kappa=.85) and 23 subcategories (M kappa=.82). Participants reported nearly
universal (98%) use of self-management, with an average of 2.96 (SD=1.2) strategies used in
the past three months, and 91% of all strategies perceived as effective. Self-care (37.0%),
cognitive (15.8%), and avoidance (15.1%) strategies were reported most commonly; more
specifically, the most prevalent strategies were exercise (11.0% of all strategies), redirecting
thoughts (9.1%), and family/friends (8.1%). Sex, age, and depression screen status were
associated with self-management strategy use. This study demonstrates the ubiquity and high
perceived effectiveness of self-management for anxiety among primary care patients. Further
research should explore the utility of promoting evidence-informed self-management for
anxiety in primary care, given its potential as a wide-reaching initial step in a population-
based approach to stepped care. Primary care providers and integrated behavioral health
providers can encourage patients to use healthy, effective self-management strategies to
cope with anxiety symptoms.

CORRESPONDING AUTHOR: Robyn L. Shepardson, PhD, VA Center for Integrated Healthcare,

Syracuse, NY, 13210; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2065

C159 6:00 PM-7:00 PM

THE ELEPHANT IN THE EXAM ROOM: MEDICAL PROVIDER PERCEPTIONS OF AND COMFORT IN
DISCUSSING WOMEN'S BEHAVIORAL HEALTH NEEDS

Kerri Garruba, PsyD, Laura Maphis, PhD, Julie Hergenrather, PhD, ABPP

Geisinger Health System, Danville, PA

Primary care has become the defacto mental health system in the country, with 70-90% of all
visits entailing a psychosocial or behavioral component. Similarly, OB/GYN practices address
womens health needs across the lifespan. Research is limited on behavioral health
integration into specialty care clinics, like OB/GYN. This grant funded study aimed to 1)
ascertain womens health specialty provider perceptions of patient behavioral health needs
and 2) understand attitudes and perceived barriers to referral to and collaboration with
behavioral health.

Methods: Two distinct, 90-minute focus groups were conducted and recorded by an
independent moderator trained in focus group facilitation. Focus group discussions centered
on participant awareness of the mental and behavioral health needs of patients, barriers to
behavioral health referral and access, and strategies for remediating problems. The groups
consisted of 1) physicians and advanced practitioners and 2) nursing staff and medical
assistants. A trained research assistant transcribed the recordings, and the research team
coded the transcript for themes.

Results: Analysis revealed three main themes: Patient needs, barriers to access and solutions.
Within the main themes, subthemes were evident. Patient needs were identified according to
life transition (ante- and post-natal periods, menopause, infertility, sexual health, cancer
diagnosis), health behaviors (smoking, obesity, pain management), traditional mental health
and substance abuse, and domestic violence and relationship conflicts. Barriers to access
included: provider experience with and knowledge about behavioral health integration; issues
around patient access (cost, transportation, time commitment, stigma/reluctance to consult);
provider access to behavioral health during the office day (lack of warm hand offs and same
day appointments, limited presence in clinic, uncertainty about outcome of referrals).
Solutions included improving workflow (referral process, office space, billing, creating
opportunities for warm hand offs and same day appointments), use of telemedicine to
increase patient access and increase the presence of behavioral health in the clinic.

Discussion: Results suggest that focus group participants are aware of patients psychological
distress and problematic health behaviors yet lack knowledge of the role of behavioral health
in integrated care settings. At the same time, participants enthusiastically welcomed the
S2066 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

integration of behavioral health into the specialty womens setting. While barriers to access
were identified, participants also generated several possible solutions to enhance integration
efforts. Future research will focus on how provider education impacts patient acceptance of
behavioral health referrals, and whether services improve overall patient health and
wellbeing.

CORRESPONDING AUTHOR: Kerri Garruba, PsyD, Geisinger Health System, Danville, PA, 17822;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2067

C160 6:00 PM-7:00 PM

THE IMPACT OF URGENT VISITS ON DELIVERING HOME-BASED PRIMARY CARE TO

HOMEBOUND SENIORS

Gregory Norman, PhD1, Kristann Orton, BA1, Julie Beecher, BA2

1
West Health Institute, La Jolla, CA; 2Medstar Washington Hospital Center, Washington DC,
DC

Background: Home-based primary care (HBPC) is a multidisciplinary care delivery model that
provides comprehensive longitudinal care to medically complex seniors in the home. Recent
studies demonstrate HBPC can result in better quality of care at a lower cost for homebound
older adults. The over four million homebound seniors in the US represent a significant
proportion of the most costly Medicare beneficiaries. These seniors are typically frail with
multiple chronic conditions, behavioral health conditions, and functional disabilities. Without
easy access to primary care, homebound seniors resort to the emergency department and
hospitalizations when they experience exacerbations of their chronic conditions. In fee-for-
service Medicare, the number of patients visited per day by the provider can impact the
sustainability of the practice.

Methods: A clinical practice quality improvement project was undertaken at one mid-Atlantic
HBPC practice to determine potential efficiency solutions. The nature and impact of urgent
visits on nurse practitioners (NPs) daily workflows was examined to determine ways to
improve how the practice managed these unscheduled appointments. GPS software on
mobile tablets tracked patient visits of 5 NPs from September 29 th to December 24th in 2015.

Results: During the data collection period NPs visited 416 of the 589 patients in the practice
(71%) at least once. A total of 711 visits were tracked and matched out of 1448 visits (49%)
logged in the practices arrival database. The majority of house call visits were scheduled
regular visits (79%), 6% were follow-up visits after hospitalization, and 15% were urgent visits.
Patients with a greater number of visits had a greater proportion of urgent visits (X2(5) = 32.7,
p 2(4) = 30, p < .001), but urgent visits did not vary by time of day. Travel time and visit-
duration at patients homes did not vary by visit type. The top three most frequently used
ICD-9 codes, regardless of visit type, were for hypertension, constipation, and immunization.
When ICD-9 codes were used for billing a home visit, 70% of urinary tract infection, 53% of
insomnia, and 50% of Type 2 diabetes with peripheral angiopathy codes were used for urgent
visits.
S2068 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions: Although NPs reported they reserve time in their schedules in anticipation of
urgent visits, these visits only occurred 15% of all house call visits. In many respects, urgent
visits did not differ in nature from routine house call visits. The commonly used ICD-9 codes
across all visit types suggest some urgent appointments might not need a same-day visit.
Changing the work flow and management of urgent visits across NPs in the practice might
allow for an increase in the volume of house call visits/day without decreasing quality of care
delivery.

CORRESPONDING AUTHOR: Gregory Norman, PhD, West Health Institute, La Jolla, CA, 92037;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2069

C161 6:00 PM-7:00 PM

WILLINGNESS TO ENGAGE IN TOBACCO AND ALCOHOL INTERVENTIONS DEPENDS ON FOCUS

BUT NOT FORMAT OF THE INTERVENTION

Jennifer M. Wray, PhD1, Jennifer S. Funderburk, PhD2, Julie C. Gass, PhD3

1
Ralph H. Johnson VA Medical Center, Buffalo, NY; 2VA Center for Integrated Healthcare,
Syracuse, NY; 3VA Center for Integrated Healthcare, Buffalo, NY

Background. Integrated primary care is an optimal setting for the delivery of interventions for
tobacco use and at-risk drinking, but rates of implementation are low. Patient perspectives on
the focus and format of interventions they would be willing to engage in may help to improve
the rates at which these types of interventions are received. Objective. We examined the
extent to which willingness to engage in an intervention was influenced by the described (1)
focus and (2) format of the intervention. Design. This was a one-time telephone survey.
Participants. Patients who had recently attended a primary care appointment and who
screened positive for tobacco use and at-risk drinking were eligible to participate. Data from
patients who used cigarettes and met criteria for at-risk drinking in the last 30 days (N=53)
were included in analyses. Main Measures. Participants rated a number of scenarios based on
their willingness to engage in different services in primary care on a scale from 1 (highly
unlikely) to 5 (highly likely). Key Results. Participants reported that they would be more
willing to engage in services focused on helping them reduce their risk of medical problems
than in services focused on discussing cigarette or alcohol use. They did not indicate a
preference related to whether the intervention was delivered by a behavioral health provider
during a primary care appointment, immediately following a primary care appointment, or as
a scheduled follow up. Conclusions. Framing services for tobacco use and at-risk drinking as a
way to reduce patients risk of medical problems may increase willingness to attend these
interventions. Patients reported being equally receptive to receiving brief interventions in
several different formats available in integrated primary care settings.

CORRESPONDING AUTHOR: Jennifer M. Wray, PhD, Ralph H. Johnson VA Medical Center,

Buffalo, NY, 14215; [email protected]
S2070 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C162 6:00 PM-7:00 PM

WOMENS PELVIC EXAM EXPERIENCE: A PATIENT-CENTERED VIEW

Lisa R. Rubin, PhD1, Christina Shenko, MD2, Saskia Schuman, MHS2, Margaux C. Genoff, B.S.1,
Katrina S. Hacker, BA1, Lara Dhingra, PhD3

1
New School for Social Research, New York, NY; 2Institute for Family Health, New York, NY;
3
MJHS Institute for Innovation in Palliative Care, New York, NY

Background: Women commonly describe the pelvic exam (PE) as a negative experience. Yet it
is also a routinely conducted and important examination. Given these factors, researchers in
medicine, psychology, and public health have explored factors that influence womens pelvic
exam experiences to inform patient-centered best practices for this routine procedure.
However, little of this research is focused on urban low-income and ethnically diverse groups
within the United States.

Methods: In depth interviews were conducted with 11 cisgender women, recruited from a
federally-qualified community based health center network providing integrative care in NYC.
Participants reflected the centers diversity (46% African-American/Black; 55% Latina; 64%
heterosexual). Forty-six percent reported at least some college or post secondary vocational
training). Interviews explored participants understanding of the purpose of the PE; early and
recent experiences of the PE; descriptions of PEs that went well and PEs that went poorly;
and recommendations for improving the PE. Interviews were transcribed verbatim, and
transcripts were evaluated using a thematic content analysis approach consisting of coding
and interpretation of the data by team consensus.

Results: Among participants, first-time PEs often were experienced in the context of a
stressful medical/reproductive circumstance. The PE remained (to varying degrees) a stressful
experience, but for some participants, it was also described as a self-care activity that
allowed them to exercise agency in relation to their body/health. Study findings highlight
emotional (e.g., anticipatory anxiety) and behavioral (e.g., avoidance) responses to the PE,
and participants PE coping strategies. Strategies for improving the PE suggested by
participants include improved communication/explanation and acknowledging the patient
experience. Several participants expressed a preference for a female provider. Findings are
contextualized in relation to participants self-reported trauma history.

Conclusions: Although many women view the PE as a stressful medical experience, these
findings highlight both negative and positive dimensions of the PE among study participants.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2071

Further, these results highlight strategies for improving practitioner training, and support the
approach of trauma-informed care.

CORRESPONDING AUTHOR: Lisa R. Rubin, PhD, New School for Social Research, New York, NY,
10011; [email protected]
S2072 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C163 6:00 PM-7:00 PM

A SYSTEMATIC REVIEW OF MENTAL HEALTH AND WELLNESS INTERVENTIONS FOR MEDICAL

STUDENTS

Breanne Dixon, Current student1, Andrea T. Kozak, PhD2

1
Oakland University, Lake Orion, MI; 2Oakland University, Rochester, MI

Background: Medical students have some of the highest rates of depression, anxiety
diagnoses, suicide, stress, and burnout. Programs developed to prevent or treat these issues
are being conducted and evaluated. Methods: A systematic review of programs/interventions
was conducted. PsychInfo, Pubmed, SCOPUS, and EMBASE were searched for articles from Jan
1990 to June 2016 based upon predetermined inclusion criteria. The search yielded a total of
1743 titles/abstracts. Thirty-eight articles were coded after reviewing the titles/abstracts and
removing duplicates. Results: Of the 38 articles coded, 27 were intervention studies (9 were
randomized controlled trials [RCTs]) and 11 were focused on the medical school curriculum.
Most of the programs involved first year students only or students from all four years of
medical school. The types of programs were categorized as course/curriculum modification
(e.g., new course or program change), psychological intervention (e.g., mindfulness), wellness
program, mental health referral program, and other. The majority of the programs were
psychological interventions (n=15), followed by course/curriculum modification (n=11). The
duration of the program was highly dependent on the type of intervention provided to the
students, such that curriculum changes were over a semester or ongoing, whereas RCTs
ranged from 4 to 8 weeks. The number of participants ranged from 5-1,008. Fourteen
psychological intervention studies (n=7 RCTs) provided results; twelve of the fourteen
demonstrated significant improvement on measures of stress, anxiety, depression, and/or
mindfulness. For the course/curriculum modification studies, all but one study had significant
findings which showed a positive change in stress, anxiety, and/or depression. Conclusion:
This systematic review suggests that psychological interventions and course/curriculum
modifications benefit medical students on important mental health outcomes. However,
many of the studies did not include a comparison group, which limits ones confidence that
the intervention lead to the improved mental health outcomes. The long-term impact of these
interventions/programs is not well documented given the minimal amount of follow-up
evaluations. Further research should be conducted to address these limitations.

CORRESPONDING AUTHOR: Breanne Dixon, Current student, Oakland University, Lake Orion,
MI, 48359; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2073

C164 6:00 PM-7:00 PM

ACTIVE DUTY: THE ROLE OF INCOME, DEPRESSION, AND STRESS ON HEALTH BEHAVIORS
AMONG MILITARY SPOUSES

Aaron Swank, B.S. Human Nutrition; B.S. Kinesiology

Kansas State University, Manhattan, KS

Active Duty: The role of income, depression, and stress on health behaviors among military
spouses

Aaron Swank, Emily Mailey, Brandon Irwin

Kansas State University

Introduction

Military spouses report high levels of stress and depression and low participation in health
behaviors. Barriers such as frequent moves, isolation from family and friends, uncertainty of
the spouses well-being, inability to find work, and sole caregiving responsibilities contribute
to poor health in this population. Interventions to improve health in military spouses are
needed and could focus on reducing financial burdens or improving access to mental health
services. The purpose of this investigation was to enhance our understanding of factors
related to health behaviors among military spouses by examining the relationships between
income, stress and depression and physical activity (PA) and nutrition behaviors.

Methods

Military spouses (N=300) completed questionnaires online. Surveys included the PA and
nutrition subscales of the Health Promoting Lifestyle Profile, the Perceived Stress Scale, and
the Hospital Anxiety and Depression Scale. Participants also reported their annual household
income. A series of multiple linear regression analyses were conducted with PA and nutrition
as the dependent variables. Each analysis included income and one mental health outcome
(i.e., perceived stress or depression) as an independent variable.

Results

Income was significantly assoicated with PA (B=.122, p=.03), but the relationship was no
longer signifcant when depression (B=-.300, p < .001) was added to the model (B=.092, p=.11).
Similarly, perceived stress was more strongly associated with PA (B= -.266, p < .001) than
S2074 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

income (B=.080, p=.16). Income remained significantly associated with nutrition when
depression (B= -.404, p < .001) and perceived stress (B=-.315, p < .001) were added to the
models.

Discussion

Both income and mental health status are significantly associated with health behaviors such
as PA and nutrition, however mental health variables (e.g., depression and perceived stress)
are more strongly associated with PA than income. Interventions to improve health among
military spouses should address individual, social, and environmental barriers to PA and
healthy eating, and should explicitly include strategies for actively managing stress and
depression. These results support the position that investing in the mental health of military
service members and their families should be a higher priority in military populations.

CORRESPONDING AUTHOR: Aaron Swank, B.S. Human Nutrition; B.S. Kinesiology, Kansas State
University, Manhattan, KS, 66502; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2075

C165 6:00 PM-7:00 PM

ACUTE STRESS DISORDER SYMPTOMS AFTER EVALUATION FOR ACUTE CORONARY SYNDROME
PREDICT 30-DAY READMISSION

Jennifer A. Sumner, PhD1, Ian Kronish, MD, MPH1, Bernard Chang, MD, PhD2, Ying Wei, PhD3,
Donald Edmondson, PhD, MPH2

1
Columbia University Medical Center, Center for Behavioral Cardiovascular Health, New York,
NY; 2Columbia University Medical Center, New York, NY; 3Columbia University Mailman
School of Public Health, New York, NY

Introduction: Both costly and common, 30-day readmissions are a major concern for U.S.
hospitals. Numerous readmission risk prediction models have been developed, but their
performance has been modest, and few predictors are modifiable. Potentially life-threatening
medical events that prompt patients index hospitalizations, along with experiences during
hospitalization, can trigger psychological stress, including posttraumatic stress reactions.
Stress may increase risk of adverse post-hospitalization outcomes. We examined whether
posttraumatic stress days after evaluation for acute coronary syndrome (ACS)termed acute
stress disorder (ASD) symptomswas associated with 30-day all-cause emergency
department (ED) and hospital readmission.

Methods: The REactions to Acute Care and Hospitalization (REACH) study is an observational
cohort study that enrolls patients during ED evaluation for ACS. During inpatient stay or by
phone after discharge (median 3 days after enrollment), participants reported their ASD
symptoms with respect to the event that brought them to the ED using 14 items from the
Acute Stress Disorder Scale (range=14-70). Scores 28 suggest probable ASD. ED and hospital
readmissions in the month after discharge were identified by contacting patients and
reviewing electronic health records. Readmissions were confirmed by research nurse record
review. Participants were excluded if they were readmitted before ASD symptom assessment
or died within 30 days of discharge, resulting in an analytic sample of 974 adults (mean
age=60.913.2 years; 45.7% female; 55.7% Hispanic; 20.7% black). Logistic regression was
used to determine the association between ASD symptoms and 30-day ED/hospital
readmission with and without adjustment for demographics, clinical characteristics, ACS
discharge diagnosis, and depression.

Results: There were 123 ED/hospital readmissions in the 30 days after discharge. A 10-point
ASD symptom increase was significantly associated with increased 30-day readmission risk in
the unadjusted model, odds ratio (OR)=1.32 (95% CI, 1.12-1.56), P=.001, and adjusted model,
S2076 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

OR=1.24 (95% CI, 1.02-1.51), P=.03. Results were consistent when using a dichotomous
probable ASD variable based on the 28+ cutoff score: unadjusted OR=1.92 (95% CI, 1.29-
2.87), P=.001; adjusted OR=1.65 (95% CI, 1.04-2.63), P=.04.

Conclusions: The psychological stress that often follows ACS evaluation was associated with
greater odds of 30-day ED/hospital readmission, independent of demographics, clinical
characteristics, and ACS discharge diagnosis. Research is needed to examine if preventing or
treating ASD symptoms may reduce readmission risk.

CORRESPONDING AUTHOR: Jennifer A. Sumner, PhD, Columbia University Medical Center,

Center for Behavioral Cardiovascular Health, New York, NY, 10032;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2077

C166 6:00 PM-7:00 PM

ASSOCIATIONS BETWEEN NEUROPSYCHIATRIC SYMPTOMS, COGNITIVE FUNCTION, AND DAILY

FUNCTIONING IN ELDERLY WITH EARLY DEMENTIA

Rainbow T. H. Ho, PhD1, Ted C. T. Fong, MPhil1, Tiffany Hon, BSc(Psy)2, Bing-yu Li, M.Ed.2, Wai
Chi Chan, MRCPsych (UK)3, Joseph SK. Kwan, MPhil MD FRCP FHKCP FHKAM2, Patrick K. C.
Chiu, MBBS4, Linda C. W. Lam, MD5

1
Centre on Behavioral Health, Hong Kong, N/A, Hong Kong; 2The University of Hong Kong,
Hong Kong, N/A, Hong Kong; 3The University of Hong Kong, Pokfulam, N/A, Hong Kong;
4
Queen Mary Hospital, Hong Kong, N/A, Hong Kong; 5The Chinese University of Hong Kong,
Hong Kong, N/A, Hong Kong

Background: Elderly with early dementia suffer from progressive deteriorations in cognitive
functioning such as memory, spatial navigation, movement and language, leading to
interferences in daily functioning. The present study aims to examine the associations
between neuropsychiatric symptoms, psychosocial well-being, cognitive functioning, and daily
functioning in elderly with early dementia.

Methods: The study sample was 86 elderly who were diagnosed with early dementia via a
clinical dementia rating rating of 0.5 to 1. The mean age of the sample was 81.1 years old and
the majority of them were female (83%), widowed (58.8%), and had at most primary
education (75%). Data on neuropsychiatric symptoms, psychosocial well-being, cognitive
functioning, and daily functioning were collected from the participants via cognitive tests
(Digit Span Test, Fuld Object Memory Evaluation and Trail Making Test) and self-reported
scales (Geriatric Depression Scale, Neuropsychiatric Inventory, Loneliness Scale, and
Instrumental Activities of Daily Living).

Results: Male participants showed greater neuropsychiatric symptoms and older participants
showed worse digit spans, memory storage and recall, and daily functioning. Controlling for
gender and age, neuropsychiatric symptoms were negatively correlated with backward digit
span, memory storage and recall. Loneliness was positively associated with depression but not
with any other variables. Digit spans, particularly backward digit span, were correlated with
memory storage, memory recall, trail making time and daily functioning. Higher memory
storage and recall and shorter trail making time were associated with better daily functioning
among the participants.

Discussions: The study findings suggest that better episodic memory, executive functioning,
S2078 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

and visuospatial construction are linked to fewer neuropsychiatric symptoms and better
functioning in daily lives. Further studies could explore the effects of non-pharmacological
interventions on ameliorating cognitive impairments and improving prognosis of dementia.

CORRESPONDING AUTHOR: Rainbow T. H. Ho, PhD, Centre on Behavioral Health, Hong Kong,
N/A, 852; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2079

C167 6:00 PM-7:00 PM

DEPRESSIVE SYMPTOMS IN PATIENTS WITH COPD PARTICIPATING IN A PRAGMATIC TRIAL OF

A PHYSICAL ACTIVITY COACHING INTERVENTION

Ellen Rippberger, MPH1, Mayra Macias, MS1, Erika Estrada, BS1, Janet Lee, MS1, Leah
Maddock, MPH1, Huong Nguyen, PhD, RN2

1
Kaiser Permanente Southern California, Pasadena, CA; 2Kaiser Permanente Southern
Calofornia, Psadena, CA

Depressive Symptoms in Patients with COPD Participating in a Pragmatic Trial of a Physical

Activity Coaching Intervention

Ellen Rippberger, MPH, Erika L. Estrada, BS, Janet Lee, MS, Mayra Macias, MS, Leah Maddock,
MPH, Huong Q. Nguyen, PhD, RN

Kaiser Permanente Southern California, Department of Research and Evaluation Pasadena,

CA, USA.

Background: Depression is common in chronic obstructive pulmonary disease (COPD) with

prevalence rates ranging from 20-50%.

Aim: We aimed to compare depressive symptoms between patients randomly assigned to

standard care (SC) vs. a 12-month physical activity coaching intervention (Walk On!).

Methods: Walk On! is a pragmatic individual-level randomized trial testing the effectiveness
of a 12-month physical activity coaching intervention compared to SC on the primary
composite outcome of hospitalizations, emergency department visits, observational stays and
death.

Patients assigned to Walk On! were contacted to participate in study activities, whereas SC
patients were contacted to complete only the surveys without knowledge of the intervention.
Participants completed an assessment of depressive symptoms using the modified Patient
Health Questionnaire (PHQ-8) either during a baseline visit (Walk On!) or via postal
mail/telephone (SC).

Results: A total of 235 patients completed the PHQ-8 (Walk On!=158, SC=77). Mean age of the
sample is 72 (10) with 56% females and 25% non-Caucasian. Mean PHQ-8 scores were higher
for SC 7.3 (6.4) compared to Walk On! 5.7 (5.0) (p=.05). The overall prevalence of PHQ-8
scores >10 was 33% vs. 15% in SC and Walk On! When patients were grouped into severity of
S2080 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

depressive symptoms (19=severe), more SC patients were severely depressed compared to

Walk On! (8% vs. 3%, p

Conclusions: Standard care patients reported higher depressive symptoms compared to Walk
On! patients. Patients who consent to participate in a 12-month physical activity coaching
program are likely more activated and have better psychological well-being than patients who
are responding to a mailed survey that does not ask for a high level of engagement.

Funding: Patient-Centered Outcomes Research Institute, PCORI 1403-14117, 2015-2018

CORRESPONDING AUTHOR: Ellen Rippberger, MPH, Kaiser Permanente Southern California,

Pasadena, CA, 91101; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2081

C168 6:00 PM-7:00 PM

DISPARITIES IN HEALTHCARE UTILIZATION AND FUNCTIONAL LIMITATIONS IN U.S. ADULTS

WITH SERIOUS PSYCHOLOGICAL DISTRESS

Judith D. Weissman, PhD, JD

NYU, New York, NY

Objective: The studys purpose was to compare healthcare utilization in adults age 18 years
and over with and without serious psychological distress. Approximately 3.4% of U.S. adults
have serious psychological distress, defined as having mental health problems causing
moderate to serious impairment in social, occupational, or school functioning, requiring
treatment. Serious Psychological Distress is associated with substantial comorbidity and up to
30 years premature mortality compared to the general population, representing the greatest
U.S. health disparity. Methods: The 2006-2014 National Health Interview Survey was used to
compare nine indicators of healthcare and functional impairments: no healthcare coverage,
delays in healthcare, insufficient money for healthcare, change in usual place of healthcare,
change in usual place of healthcare due to insurance, more than 10 doctor visits in 12 months,
insufficient money to buy prescription medications, having seen a mental health provider,
insufficient money for mental healthcare, limitations in ability to work and activities of daily
living. Results: Compared to adults without serious psychological distress, adults with serious
psychological had diminished healthcare utilization and functioning after controlling for other
health conditions and sociodemographic characteristics. From 2006-2014, we observed an
increasing trend in the proportion of adults with serious psychological distress who lacked
healthcare coverage, experienced delays in healthcare, lacked mental healthcare and/or
money to buy prescription medications. Conclusions: Our study suggests a need for greater
attention to factors underlying healthcare utilization in the mentally ill, and to identify
mechanisms for improving access to care.

CORRESPONDING AUTHOR: Judith D. Weissman, PhD, JD, NYU, New York, NY, 10028;
[email protected]
S2082 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C169 6:00 PM-7:00 PM

EFFECT OF INDIVIDUAL DIFFERENCES ON RELATIONSHIP BETWEEN SUBJECTIVE DISTRESS AND

HEART RATE DURING 7.5% CO2 CHALLENGE

Rachel Wallace, B.S. Psychology1, Charles Calderwood, Ph. D.2, Roxann Roberson-Nay, Ph. D.3,
Scott Vrana, Ph. D.4

1
Clinical Psychology/Department of Psychology/Virginia Commonwealth University,
Richmond, VA; 2Department of Social Psychology Virginia Commonwealth University,
Richmond, VA; 3Department of Psychiatry Virginia Commonwealth University, Richmond, VA;
4
Virginia Commonwealth University Departments of Psychology and Psychiatry, Richmond, VA

Literature suggests that an individual's self-reported experience during psychological distress

is associated with their psychophysiological response. However, studies examining synchrony
and concordance in these measures have found inconsistent results. Individual differences
have been found that affect the relationship between physiological response and self-report.
This study examines the effects of individual differences in Anxiety Sensitivity (AS) and
baseline heart rate variability (HRV) on the relationship between heart rate (HR) and
Subjective Units of Distress (SUDS) ratings within an undergraduate sample (n=179,
Mage=20.1, 55.3% female, 27.4% African-American) during CO2 challenge. The CO2 challenge
reliably elicits changes in physiological responding and self-reported distress. Participants
breathed a steady state of 7.5% CO2 gas mixture for 8 minutes, preceded and followed by 5-
minutes of breathing room air. Average HR and SUDS were collected every 2 minutes. A series
of multilevel models were employed to assess whether SUDS, AS, and baseline resting HRV (a
physiological measure of emotion regulation) predicted HR. An unconditional model revealed
that 76% of the criterion variance in heart rate was at the between-participants level. Entry of
SUDS, AS, and HRV in the model led to an improvement in model fit, difference of -2 log
likelihood = 141.48, p < .05. The coefficient relating SUDS to HR was positive and statistically
significant ( = 0.12, p < .001), indicating that higher heart rate covaried with higher SUDS
ratings. The coefficient relating AS to HR ( = 0.05, p = 0.58) was positive but not statistically
significant and HRV to HR ( = -5.03, p < .001) was negative and statistically significant,
indicating that HR during C02 challenge was lower for people with higher baseline HRV. The
addition of a random slope term in the next step led to an improvement in model fit,
difference of -2 log likelihood = 16.76, p < .05. However, a cross-level moderator analysis
found no evidence to suggest that this slope variability was explained by AS (p = 0.57) or HRV
(p = 0.09). Thus, while individual differences in HRV, but not AS, predicted HR during a CO2
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2083

challenge, neither variable significantly predicted synchrony between HR and self-reported

distress.

CORRESPONDING AUTHOR: Rachel Wallace, B.S. Psychology, Clinical Psychology/Department

of Psychology/Virginia Commonwealth University, Richmond, VA, 23220;
[email protected]
S2084 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C170 6:00 PM-7:00 PM

FORGIVENESS AND PSYCHOLOGICAL WELL-BEING OF ADULT MALE SURVIVORS OF

CHILDHOOD SEXUAL ABUSE BY CATHOLIC PRIESTS.

charlene Niemi, RN, PHN, MSN, PhD

California State University Channel Island, Newbury Park, CA

Forgiveness and psychological well-being of adult male survivors of childhood sexual abuse
by Catholic priests.

Purpose: The purpose of this study was to (a) explore the mediating effects of forgiveness
with the psychological well-being of adult male survivors of childhood sexual abuse by
Catholic priests and (b) elucidating relationships may enhance interventions. Characteristics of
abuse include age of first occurrence, frequency, length of time abuse continued, and types of
abuse.

Methods: A non-experimental, cross sectional design was used to examine the relationship
between forgiveness and psychological well-being among adult male survivors of childhood
sexual abuse by Catholic priests. This quantitative, descriptive, correlational study employed
self-administered questionnaires that were distributed through the Internet to obtain
information from the 67 participants. Descriptive statistics were used to summarize data.
Additionally, t test and analysis of variance was used to compare mens abuse characteristics
across forgiveness and psychological well-being. Pearsons correlation was conducted
between forgiveness and psychological well-being. Simultaneous multiple regression was
performed to ascertain the effects of the various characteristics of abuse and forgiveness on
prediction of psychological well-being.

Results: The results found the highest number of men were first abused between 11 and 14
years of age; the abuse continued over a period of few years; frequency was three to 10
times; and a total of 355 types of abuse noted, for an average of 5 to 6 types of abuse per
participant. The only t-test results that were statistically significant were forgiveness (absence
of negative) and age of first occurrence, with the second category (11 to 17 years) having
higher forgiveness (absence of negative). Pearsons correlation showed moderate to high
significance. The multiple regression revealed forgiveness had the highest variance in
predicting psychological well-being.

Conclusion: While this study identified facilitation of forgiveness as a possible intervention to

improve psychological well-being in male survivors of child sexual abuse by Catholic priests,
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2085

the small size limited the possibility of discovering the role the other factors of abuse may
have in well-being. Of importance is that while the characteristics of abuse cannot be altered,
forgiveness was the only variable that can be modified.

CORRESPONDING AUTHOR: charlene Niemi, RN, PHN, MSN, PhD, California State University
Channel Island, Newbury Park, CA, 91320; [email protected]
S2086 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C171 6:00 PM-7:00 PM

HIGHER PERCEIVED THREAT UNDERLIES GREATER POSTTRAUMATIC STRESS SYMPTOMS IN

YOUNGER ADULTS WITH ACUTE CORONARY SYNDROME

Laura Meli, M.S.1, Melinda Chang, M.S.1, Faith Parsons, B.S.1, Donald Edmondson, PhD, MPH2,
Jennifer A. Sumner, PhD3

1
Center for Behavioral Cardiovascular Health, Columbia University Medical Center, New York,
NY; 2Columbia University Medical Center, New York, NY; 3Columbia University Medical Center,
Center for Behavioral Cardiovascular Health, New York, NY

Introduction

Acute coronary syndrome (ACS) is a traumatic event that can trigger posttraumatic stress
symptoms (PSS). Younger age is linked to elevated risk for PSS after ACS. Younger individuals
may perceive an ACS event as particularly threatening as it may be their first life-threatening
illness, and thus may be more vulnerable to develop post-ACS PSS. We examined associations
between age and PSS in the acute aftermath of evaluation for ACS in the emergency
department (ED) and whether perceived threat in the ED mediated this association. We
hypothesized that younger age would be associated with greater ED perceived threat and PSS
and that perceived threat would mediate, in part, the age-PSS relation.

Methods

The Reactions to Acute Care and Hospitalization (REACH) study is an observational cohort of
patients presenting to the ED with suspected ACS. Demographics and perceived threat were
assessed in the ED. PSS were measured upon hospital admission. The analytic sample
comprised 871 participants (mean age=60.912.9 years; 45.8% women; 56.6% Hispanic;
19.7% black). Regression analysis was used to test if perceived threat mediated the
association between age and PSS; bootstrapping with percentile-based confidence intervals
(CIs) was used to test the indirect effect.

Results
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2087

Each year of age was associated with lower PSS (b=-0.12, p < .001), adjusting for demographic
and medical covariates. Older age was also associated with lower perceived threat (b=-0.05, p
< .001). Greater perceived threat predicted greater PSS (b=0.94, p < .0001). The indirect effect
was negative (-0.04) and statistically different from zero (95% CI, -0.07, -0.02).

Conclusions

Compared to older individuals, younger individuals were at risk for greater PSS after ED
evaluation for ACS, and elevated perceived threat in the ED partially mediated this
association. These findings begin to explain why younger age predicts PSS after ACS.
Understanding age differences and the potential impact of age on threat perceptions may
help inform treatment in the ED.

CORRESPONDING AUTHOR: Laura Meli, M.S., Center for Behavioral Cardiovascular Health,
Columbia University Medical Center, New York, NY, 10032; [email protected]
S2088 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C172 6:00 PM-7:00 PM

HOW DEPRESSED PATIENTS AND THEIR CAREGIVERS INTERACT AROUND SELF-CARE

Rashmi Risbud, MA1, Cindie Slightam, MPH2, James E. Aikens, Ph.D.3, Tim Guetterman, PhD3,
Erin D. Bouldin, MPH, PhD4, Ranak B. Trivedi, PhD5

1
VA Palo Alto Health Care System, Menlo park, CA; 2VA Palo Alto Healthcare System, Menlo
Park, CA; 3University of Michigan, Ann Arbor, MI; 4Appalachian State University, Boone, NC;
5
Stanford University/VA Palo Alto, Menlo park, CA

Objective: To characterize enhanced collaboration between depressed patients and their

caregivers.

Methods: Patients were recruited after receiving a yearlong research intervention involving
interactive voice response (IVR) system for depression self-care. Each patient participated
with an in-home caregiver and a non-cohabitating caregiver (CarePartner). All participants
underwent individual semi-structured interviews. Interviews were professionally transcribed,
coded using Atlas.ti, and analyzed via rapid analyses techniques using grounded theory.

Results: We interviewed 10patients (mean age 39.616.2 years; 60% female), 8 CarePartners
(56.312.1 years; 100% female), and 6 in-home caregivers (46.815.7 years; 33.3% female).
Emergent themes were: (1) Communication, (2) Roles, and (3) Barriers to self-care. Patients
collaborated with both caregivers, yet the caregivers seldom collaborated with each other.
Both types of caregivers provided emotional and instrumental support, but differences were
noted in their roles. For instance, CarePartners often encouraged outside activities,
socializing, and engaging in hobbies, whereas in-home caregivers were more involved with
daily management such as household chores, medication and appointment reminders, and
monitoring symptoms. All respondents noted at least one barrier to depression self-care.
Patients barriers tended to be logistical, motivational, or mood related. Caregivers were
concerned about violating the patients privacy and noted that this was a barrier to initiating
and maintaining discussions about depression care. Participating in the intervention program
gave caregivers explicit permission to discuss patients depression more openly with the
patient. For all participant types, intervention appeared to increase open communication,
leading to increased mutual awareness of the patients mood status.

Conclusions: Patients and caregivers often play unique roles in depression self-care.Assigning
a role to a close family member or friend improves communication with the patient about
their depression. Given the lack of communication between the two types of caregivers, the
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2089

patients role remains central to depression self-care.

Impact: In depression, the role of an outside caregiver may be critical in ensuring optimal self-
care. Specific aspects of social interaction among the patient and caregivers can be bolstered
to better impact patient health outcomes.

CORRESPONDING AUTHOR: Rashmi Risbud, MA, VA Palo Alto Health Care System, Menlo park,
CA, 94025; [email protected]
S2090 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C173 6:00 PM-7:00 PM

INDIRECT EFFECTS FROM NEUROLOGICAL SIGNS TO FUNCTIONAL OUTCOMES VIA

PSYCHIATRIC SYMPTOMS IN CHINESE SCHIZOPHRENIA PATIENTS

Ted C. T. Fong, MPhil, Rainbow T. H. Ho, PhD, Adrian H.Y. Wan, MSW

Centre on Behavioral Health, Hong Kong, N/A, Hong Kong

Background: Neurological soft signs (NSS) in motor coordination and sequencing of complex
motor acts occur in the majority of schizophrenia patients and are an intrinsic sign of the
underlying neural dysfunctions of schizophrenia. The present study explored the longitudinal
relationships among NSS, psychiatric symptoms, and functional outcomes in patients with
chronic schizophrenia.

Methods: The present study recruited a sample of 151 Chinese patients with chronic
schizophrenia from a residential care hostel in Hong Kong. The participants completed the
Neurological Evaluation Scale for NSS at Time 1, Positive and Negative Syndrome Scale for
psychiatric symptoms at Time 1 and Time 2 (3-month follow-up), Activities of Daily Living
(ADL) index and Instrumental Activities of Daily Living (IADL) scale for daily functioning at Time
3 (6-month follow-up). Direct and indirect effects of NSS on functional outcomes via
psychiatric symptoms were evaluated using path analysis in Mplus 7.2 under bootstrapping.

Results: The path model showed an adequate model fit to the data (2(38) = 45.11, p = 0.20,
CFI = 0.99, TLI = 0.96, RMSEA = 0.04, SRMR = 0.05). Time 1 motor coordination and motor
sequencing NSS did not have any significant direct effects on Time 3 functional outcomes in
the path model. However, motor coordination NSS exerted significant and negative indirect
effects on ADL via positive and disorganized symptoms and on IADL via negative and
disorganized symptoms.

Discussions: The current results contribute to a better understanding of the determinants of

functional outcomes by demonstrating significant indirect pathways from motor coordination
NSS to functional outcomes via psychiatric symptoms. The finding that motor sequencing NSS
did not affect functional outcomes either directly or indirectly may be explained by their trait
marking features.

CORRESPONDING AUTHOR: Ted C. T. Fong, MPhil, Centre on Behavioral Health, Hong Kong,
N/A, 852; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2091

C175 6:00 PM-7:00 PM

PSYCHOLOGICAL FUNCTIONING IN LUNG TRANSPLANT CANDIDATES

Lauren Z. Rynar, M.S.1, Mekhala Samsi, Ph.D.2

1
Rosalind Franklin University of Medicine and Science, Chicago, IL; 2Loyola University Medical
Center, Maywood, IL

Background: Lung transplant candidates are in a highly critical state of physical health and
experience a marked reduction in their vitality and quality of life. Existing literature suggests
that up to one-fourth of patients with end-stage pulmonary disease awaiting lung
transplantation meet diagnostic criteria for a mood or anxiety disorder. Understanding the
interrelationships between depressive symptoms and other biopsychosocial variables relevant
to lung transplant candidacy may help to identify patients who are high-risk from a
psychological standpoint.

Methods: 81 patients with end-stage lung diseae completed a packet of clinical

questionnaires prior to their pre-transplant psychosocial evaluation. Scores on the PHQ-9, a
multipurpose depression screening tool, were summed and categorized into depression
severity. One-way ANOVA was used to examine differences across categories.

Results: 56.8% of the sample (N=46) did not endorse depression, 29.6% (N=24) endorsed mild
depression, and 9.9% (N=8) endorsed moderate depression. Patients with moderate
depression exhibited higher anxiety (p=.003), higher use of denial (p=.050) and substance use
(p=.003) as coping strategies, more family dissatisfaction (p=.034) and poorer medication
adherence (p < .001) than patients with no depression. Endorsing mild or moderate
depression was associated with higher stress (p < .001), worse shortness of breath (p=.009),
more fatigue (p < .001), worse quality of life (p=.006), and higher use of disengagement
(p=.037) and blaming (p=.004) styles of coping when compared to patients with no
depression.

Conclusions: Depression predicts reduced quality of life, poor medication adherence, and
maladaptive coping among a sample of patients with ESLD, each of which has been shown to
predict poor outcomes pre- and post-lung transplantation. This finding highlights the
importance of referring depressed candidates for behavioral health treatment as early as
possible in the pre-transplant period. The PHQ-9 is a user-friendly depression screening tool
that can be used in multiple clinical settings to identify patients endorsing depressive
symptoms, enabling transplant teams to intervene proactively to improve patients' physical
S2092 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

and psychological well-being, engagement in health behaviors, and overall candidacy for
transplantation.

CORRESPONDING AUTHOR: Lauren Z. Rynar, M.S., Rosalind Franklin University of Medicine

and Science, Chicago, IL, 60622; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2093

C176 6:00 PM-7:00 PM

PSYCHOLOGICAL OUTCOMES OF HEALTH ANXIETY IN CHRONIC ILLNESSES: A SYSTEMATIC

REVIEW OF THE LITERATURE

Georden Jones, B.A. , Viviane Ta, Christina Tomei, M.A. , Caroline Sguin-Leclair, B.Sc.,
Danielle Petricone-Westwood, B.A. , Brittany Mutsaers, B.Sc., Sophie Lebel, Ph.C., C.Psych

University of Ottawa, Ottawa, ON, Canada

Purpose. Health anxiety (HA) is a prevalent, highly preoccupying belief that is apparent in indi
viduals living with chronic medical conditions, such as diabetes, cancer, cardiac disease, and m
ultiple sclerosis. Despite its debilitating effects on chronically ill patients, HA remains a vastly u
nderstudied topic. This study was part of a larger systematic review conducted to comprehen
sively examine the literature on HA in individuals living with a chronic illness, with a specific fo
cus on psychological correlates of HA. Methods. A systematic review of the literature was und
ertaken on HA in individuals living with a chronic illness. Relevant studies were identified via P
sycINFO, PubMED, CINAHL, Web of Science, and SCOPUS, from January 1996 to October 2014.
A total of 244 articles and abstracts were retained for data extraction. A standardized data sp
readsheet was designed and utilized to extract and summarize data. Results. The most preval
ent illnesses included cancer (n =
95), Parkinsons disease (n=47), diabetes (n=43), and cardiac illnesses (n=27). Results indicated
strong evidence of a relationship between HA and depression (r=0.21 to 0.67), general anxiet
y (r=0.29 to 0.71), trait anxiety (r=0.25 to 0.58), state anxiety (r=0.22 to 0.56), mental health (r
= -0.25 to -0.67) and quality of life (r= -0.25 to -
0.74). Moderate evidence was found in relation to HA and mood, and limited evidence was
found in stress, and post-
traumatic stress disorder symptoms. A moderate relationship between HA and avoidance stra
tegies was found (r=0.17 to 0.60). Mixed and limited results were found in other coping strate
gies and HA. Conclusions. These results suggest that HA is associated to negative psychologica
l functioning including symptoms of depression, general anxiety, state and trait anxiety, lower
mental health and lower quality of life across diverse chronic illnesses. It also appears that avo
idance may be a coping strategy utilized by this population. This emphasizes the importance o
f screening for and addressing HA in patients living with chronic illnesses to promote psycholo
gical wellbeing.

CORRESPONDING AUTHOR: Georden Jones, B.A. , University of Ottawa, Ottawa, ON, K1N 6N5;
[email protected]
S2094 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C177 6:00 PM-7:00 PM

SOCIAL MEDIA USE DURING CRISIS: THE UTILIZATION OF FACEBOOK IN RESPONSE TO CAMPUS
VIOLENCE

Holly Rus, B.A. , Jitske Tiemensma, PhD

University of California, Merced, Merced, CA

Social media has become an important resource during and after crises. Platform affordances
such as real-time information sharing and broad network reach have the potential to facilitate
previously unprecedented communication among victims, witnesses, and their associated
social network members. However, little is known about how social media is used in response
to trauma, and how it may influence (long-term) well-being. Given the reliance on virtual
communication during crises, as well as the importance of social support and emotional
expression in promoting recovery after trauma, the current study assessed Facebook use for
the purposes of seeking information, seeking social support, and expressing emotions in
response to the acts of violence carried out at the University of California, Merced in
November of 2015. Students (n = 552) reported their uses of Facebook on the day of the
attacks and their current depressive and post-traumatic stress disorder (PTSD) symptoms five
months after the attacks. Results showed that many students were experiencing psychological
distress even after five months. The majority of students reported using Facebook more, and
in different ways than normal in response to the events, with most reporting using it to seek
information. Although students reported short-term affective change, using the site to seek
information, seek social support, or express emotions was not significantly associated with
depressive or PTSD symptoms five months later. Findings highlight the lasting impact of
campus violence on psychological well-being as well as the function and impact of social
media use during crisis.

CORRESPONDING AUTHOR: Holly Rus, B.A. , University of California, Merced, Merced, CA,
95340; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2095

C178 6:00 PM-7:00 PM

THE IMPACT OF A MIND-BODY PROGRAM ON MULTIPLE DIMENSIONS OF RESILIENCE AMONG

PATIENTS WITH NEUROFIBROMATOSIS

Catherine Pierre-Louis, BA1, Emily Zale, MS2, Eric Riklin, BS3, Eric A. Macklin, PhD4, Jessica N.
Coleman, BA5, Ana-Maria Vranceanu, PhD2

1
Massachusetts General Hospital, Boston, MA; 2Massachusetts General Hospital/Harvard
Medical School, Boston, MA; 3Fordham University, Boston, MA; 4Massachusetts General
Hospital & Harvard Medical School, Boston, MA; 5Massachusetts General Hospital,
Cambridge, MA

Introduction: The neurofibromatoses (NF) comprise of a group of genetic disorders (NF1, NF2,
and schwannomatosis) that predispose patients to develop nerve sheath tumors, which often
cause: significant morbidity, including disfiguring cutaneous tumors (NF1); complete hearing
loss, facial weakness, and poor gait (NF2); and chronic disabling pain (schwannomatosis).
There is currently no cure for NF, treatments are primarily focused on symptom management
(e.g., via surgery and palliative measures), and patients with NF have been found to report
psychological distress and impaired quality of life. Resiliency the ability to bounce back when
faced with chronic stressors is a multidimensional construct that may be particularly
relevant to patients with NF, who have to adjust to leaving with a chronic incurable condition.
A recent randomized controlled trial (Vranceanu et al., 2016) demonstrated that the 8 session
live video Relaxation Response Resiliency Program adapted for patients with NF (3RP-NF)
produced sustained increases in both physical and psychological quality of life, relative to an
attention placebo control The purpose of this secondary analysis is to further examine the
effects of the 3RP-NF versus control on multiple dimensions of resilience.

Method: Sixty-three patients (46 female, 56 White) were randomized to 3RP-NF (N = 32,
Mage = 43) or control (N = 31, Mage = 40), and completed pre- and post-intervention
measures. Of these, 52 completed a 6-month follow-up assessment. The multidimensional
assessment of resilience included measures of coping, social support, gratitude, optimism,
spirituality, and mindfulness.

Results: Repeated measures ANOVA with linear contrasts indicated that the 3RP-NF produced
sustained increase in multiple measures of resilience. Participants randomized to 3RP-NF
S2096 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

demonstrated greater improvements from pre- to post-intervention in coping (6.68; 95% CI:
1.7811.58; p = .008), social support (9.16; 95% CI: 0.82-17.50; p = .032), and mindfulness
(2.23; 95% CI: 0.164.29; p = .035), relative to control, and these improvements were
maintained at 6-month follow-up. Although, on average, participants in 3RP-NF showed
improvements at in optimism and spirituality, they were similar to those in HEP-NF.

Conclusion: The 3RP-NF improved multiple dimensions of resilience, and produced increases
in coping, social support, and mindfulness that were over and above those observed in the
control condition. Improvements were sustained at six-month follow-up. Results suggest that
psychosocial intervention can promote resilience among patients with NF and that 3RP-NF
may be efficacious in targeting multiple dimensions of resilience.

CORRESPONDING AUTHOR: Catherine Pierre-Louis, BA, Massachusetts General Hospital,

Boston, MA, 02114; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2097

C179 6:00 PM-7:00 PM

THE PREVALENCE OF PSEUDOBULBAR AFFECT IN A MENTAL HEALTH CLINIC

Kishen Bera, Undergraduate, Bachelor of Science

University of California, Riverside, Irvine, CA

Objective: Pseudobulbar Affect (PBA) is a neurological condition affecting the brain,

characterized by frequent uncontrollable laughing/crying episodes often unrelated to mood
or social context. Our main objective was to determine if the Center for Neurologic Study-
Liability Scale (CNS-LS), the only scrrening tool for PBA, could be a valuable primary tool for
clinicians to use in an outpatient mental health clinic to screen for PBA.

Methods: We administered the CNS-LS scale to 223 outpatients who came in for their
regularly scheduled outpatient mental health appointment. A score of 13 or higher on the
scale is felt to coordinate with a high likelihood that PBA may exist.

Results: It was found that 44.39% of the patients had a score of 13 or greater, indicating that
further questioning for possible PBA was warranted.

Conclusion: To my knowledge this is the first screening for PBA utilizing the CNS-LS Scale in an
outpatient psychiatric clinic. The high prevalence of positive screening that was found
suggests that PBA may be more common than previously recognized in the psychiatric setting.
With newly approved pharmacologic treatments now avaialble for PBA, regular assessments
for PBA in the mental health clinic may lead to improved outcomes for patients.

CORRESPONDING AUTHOR: Kishen Bera, Undergraduate, Bachelor of Science, University of

California, Riverside, Irvine, CA, 92618; [email protected]
S2098 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C180 6:00 PM-7:00 PM

USING THE TRANSTHEORETICAL MODEL TO PREDICT PHYSICAL ACTIVITY AMONG

OVERWEIGHT ADULTS WITH SERIOUS MENTAL ILLNESS

Ahmed Jerome Romain, PhD, Amal Abdel-Baki, MD, Msc, FRCPC

University of Montreal Hospital Research Centre, Montreal, PQ, Canada

Introduction: While physical activity (PA) is widely recommended among adults with serious
mental illness (SMI), most of them are inactive or insufficiently active. In this context, the
transtheoretical model (TTM) could provide a theoretical framework to understand the
motivational factors associated with PA practice. However, the TTM has been rarely used
among adults with SMI.

Objectives: Analyze whether TTM components are related to PA practice among overweight
adults with SMI

Methods: Cross-sectional study. Participants completed questionnaires assessing TTM

components: stages of change (SOC), decisional balance (pros and cons), self-efficacy and
processes of change (POC; experiential and behavioural). PA was evaluated using the GPAQ
questionnaire.

Results: 43 participants with SMI (27 males; mean age of 29.0 6.64; mean body mass index:
31.91 5.10 kg/m; 30% (n = 13) with schizophrenia, 23% with bipolar disorder) were
included. Among them, 44% (n = 19) were physically inactive and 81% (n = 35) reported no
vigorous PA at all. PA was positively correlated to SOC (p = 0.008), experiential POC (p = 0.01),
behavioural POC (p = 0.002) and perceived Pros (p = 0.01). SOC, behavioural POC and age
explained 47% of PA variance. Physically active participants had higher score in behavioural
POC (p = 0.04) and perceived Pros (p = 0.0001).

Discussion: Our results confirm the validity and interest of taking into account the TTM in PA
promotion planning. Future studies should create PA plans using behavioural POC (e.g., social
support, counter-conditioning) and perceived Pros (enlist advantages to exercise).

CORRESPONDING AUTHOR: Ahmed Jerome Romain, PhD, University of Montreal Hospital

Research Centre, Montreal, PQ, H2X0A9; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2099

C181 6:00 PM-7:00 PM

I DONT KNOW MY RISK: ADDITIONAL EVIDENCE THAT DONT KNOW RESPONDING TO RISK
PERCEPTION QUESTIONS IS DUE TO KNOWLEDGE DEFICITS

Heather Orom, PhD1, Jennifer L. Hay, Ph.D.2, Caitlin Biddle, MA, MS1, Marc T. Kiviniemi, BA,
PhD1, Erika A. Waters, PhD, MPH3, Yuelin Li, PhD.2

1
University at Buffalo, Buffalo, NY; 2Memorial Sloan Kettering Cancer Center, New York, NY;
3
Washington University in St. Louis, St. Louis, MO

Purpose: There are competing explanations for why survey takers respond dont know to
survey questions. Correspondingly, there is debate over whether dont know options should
be included in surveys and how these responses should be treated in analyses. The purpose of
this study is to test whether people respond dont know to perceived risk questions because
they lack knowledge about risk factors or because such a response is less time consuming and
effortful, i.e., satisficing.

Methods: Participants were 280 MTURK workers (median education = some college; 80%
White; mean age = 34 (SD = 9)) who completed a one-time on-line survey. We used logistic
regression to test whether actual knowledge (number of correct responses to risk knowledge
questions) or perceived knowledge about risk was associated with lower dont know
responding (DKR) for both absolute risk and comparative risk for diabetes and colorectal
cancer. We controlled for satisficing (time to complete the survey), race/ethnicity (white vs.
minority), level of education, income and age in all four models.

Results: Participants perceived knowledge (mean=2.8 vs. 2.2), and actual knowledge
(mean=4.1 vs. 3.8) about risk factors were higher for diabetes than colorectal cancer (ps <
.002). They were more likely to indicate that they did not know their risk for colorectal cancer
(31.8% for absolute and 18.2% for comparative risk) than for diabetes (14.3% for absolute risk
and 5.4% for comparative risk). The hypothesized association between lower risk factor
knowledge and higher DKR was confirmed for both colorectal cancer models (ps.01). Only
lower actual knowledge was associated with higher odds of DKR for comparative risk for
diabetes (p=.03) and lower perceived knowledge of risk factors was associated with higher
DKR for absolute diabetes risk (p>.12).

Discussion: Evidence is mounting that knowledge deficits result in uncertainty about personal
risk and DKR to perceived risk questions. These responses appear to be meaningful and should
be assessed and analyzed. Furthermore, despite decades of risk messaging about both
S2100 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

diseases, the public has persistent uncertainty and incorrect beliefs about risk factors for
common diseases, even among relatively well-educated individuals. Novel approaches to risk
communication are warranted.

CORRESPONDING AUTHOR: Heather Orom, PhD, University at Buffalo, Buffalo, NY, 14214;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2101

C182 6:00 PM-7:00 PM

DEVELOPING AND VALIDATING AN INSTRUMENT MEASURING THEORY-BASED

DETERMINANTS OF SUGARY DRINK CONSUMPTION OF OVERWEIGHT ADULTS

Maria Collado-Rivera, MS, CHES, Paul Branscum, PhD, RD

The University of Oklahoma, Norman, OK

Interventions directed towards obesity prevention and treatment are greatly needed, and
planning for such interventions should include community participation. The purpose of this
study was to develop and validate a new theory-based instrument with overweight and obese
adults to assess theory-based determinants of the behavior To stop drinking regular soda and
other sugary drinks for the next 6 months. The theoretical underpinnings of this study were
housed within the Integrative Model of Behavioral Prediction (IMB). According to the IMB,
individual behavior is predicated upon having sufficient intentions to behave in certain ways,
barring any deficiency in skills and abilities to perform the behavior, and environmental
constraints. Behavioral intentions are further influenced by three factors: attitudes towards a
behavior, perceived norms (PN), for which there are two types (injunctive and descriptive
norms) and perceived behavioral control (PBC). The IMB is a Value Expectancy Theory, and as
such, each construct of the model is measured via direct methods (i.e. asking respondents to
evaluate whether engaging in a behavior is Good/Bad or Pleasant/Unpleasant), and indirect
methods (i.e. asking respondents to judge the likelihood an outcome of a behavior (behavioral
beliefs) and value associated with each value (outcome evaluation). In order to operationalize
all of the direct and indirect measures of the IMB, this study was conducted over 2 phases.
During Phase 1, an elicitation of behavioral, injunctive normative, descriptive normative and
control beliefs was accomplished, using a brief survey with open-ended items to a sample of
overweight and obese adults attending a weight loss clinic (n=30) in a Southwestern city.
Results were then thematically analyzed into codes, and the top codes were used for
instrument development and validation in Phase 2. For Phase 2, the newly developed 65-item
instrument was first tested for face and content validity by a panel of 6 experts, and pilot
tested with a small sample within the target population (n=10). Next, the survey was
administered to 410 patients. Psychometric properties next tested the instrument for
construct validity (confirmatory factor analysis(CFA)), and internal consistency reliability
(Cronbachs alpha). Initial results from CFA models suggested some scales required
respecification by removing weak items. After model respecification, all scales appeared to be
valid and reliable. Results from this study can be used to facilitate other researchers in survey
development when using behavioral prediction models.
S2102 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Maria Collado-Rivera, MS, CHES, The University of Oklahoma,

Norman, OK, 73071; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2103

C183 6:00 PM-7:00 PM

ETHNIC DIFFERENCES IN SOCIAL NETWORKS AND SUPPORT: MEASUREMENT

CONSIDERATIONS, AND IMPLICATIONS FOR DISPARITIES RESEARCH

Melissa Flores, MS1, John M. Ruiz, Ph.D.2, Christian R. R. Goans, M.S.3, Emily Butler, Ph.D.2,
Michiyo Hirai, Ph.D.4, Timothy Smith, PhD5

1
The University of Arizona, Tucson, AZ; 2University of Arizona, Tucson, AZ; 3University of
North Texas, Lewisville, TX; 4University of Texas Rio Grande Valley, Edinburg, TX; 5University
of Utah, Park City, UT

Social relationships, and the quality of these relationships are associated with health.
Racial/ethnic, and cultural differences in valuing social relationships are increasingly
hypothesized as important moderators of these effects. However, there is little published
research documenting the size and quality of social networks across different ethnic groups.
The aims of this study were to 1) test whether Hispanics and non-Hispanics (NH) respond
similarly to classic measures of social networks and support, and 2) examine whether there
are significant between group differences on these measures.

A sample of 3,283 undergraduate participants (37.3% Hispanic) from 3 population bases

(Texas, Utah, Georgia) completed the Social Network Index (SNI; Cohen, et.al., 1997), Social
Provisions Scale (SPS; Russell, et.al., 1984), and the Interpersonal Support Evaluation List (ISEL;
Merz et al., 2014) as part of a larger online survey study.

To assess whether groups responded similarly to the measures, a confirmatory factor analytic
model was used to test for the invariance of factor covariance and mean structures in three
latent constructs: SNI, SPS, and ISEL. Configural, weak factorial, and strong factorial invariance
was found using the change in comparative fit index < .01, and model fit indices (Little, 1997).
These findings indicate that Hispanics and NHs respond in similar ways to the measures and
that a reasonable between groups comparison can be pursued.

To test for ethnic differences in network size and support, a multiple indicator, multiple
causes (MIMIC) model was used to determine mean differences in these latent constructs.
Model fit indices indicated adequate overall model fit, 2(60, N = 3283) = 1208.63, p < .001;
RMSEA = .076 (90% CI: .073, .080); CFI = .96; SRMR = .04. Contrary to expectations about
Hispanics and high support, we found that NHs reported larger social networks (p < .001), and
greater social provisions (p < .01).
S2104 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Results are limited to the specific measures employed here and other measures emphasizing a
particular source of support (e.g., family) may yield different findings. This study represents an
important first step to understanding how size and quality of social networks and social
support vary across ethnicity with implications for studying racial/ethnic health disparities.

CORRESPONDING AUTHOR: Melissa Flores, MS, The University of Arizona, Tucson, AZ, 85721;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2105

C184 6:00 PM-7:00 PM

FACTORIAL INVARIANCE OF DECISIONAL BALANCE AND SELF-EFFICACY MEASURES FOR BLOOD

DONATION INTENTIONS AMONG BLACK ADULTS

Alexandra N. Murray, MA1, Mark Robbins, PhD2

1
University of Rhode Island, New London, CT; 2Cancer Prevention Research Center, University
of Rhode Island, Kingston, RI

Black adults in the United States donate blood at lower rates than non-Hispanic Whites
despite significant need for antigen-matched blood to treat medical conditions that
disproportionately affect Blacks (e.g., sickle cell disease). Measures and tailored interventions
were previously developed, guided by the Transtheoretical Model (TTM) of behavior change
with all Black samples to increase blood donation. This study aimed to establish factorial
invariance of these measures to support their generalizability and the validity of the items as
measuring the same underlying constructs across population subgroups. Participants were
716 Black adults recruited from the northeastern U.S. in two studies conducted to develop
and pilot TTM measures for decisional balance (DB) and self-efficacy (SE) to donate blood.
Measurement invariance was conducted using three sequentially more restrictive models:
configural (unrestricted), metric/weak (factor loadings restricted), and strong (factor loadings
and error variances restricted). Invariance of measures was evaluated across gender
(male/female) and age groups (18-25; 26-39; 40-54; 55-75). For DB, strong factorial invariance
was achieved for gender with good model fit [CFI = .960, RMSEA = .063 (90% CI = .053-.072)].
Metric invariance was not achieved for DB across age groups as this model did not provide
better fit than the unrestricted model [CFI = .018]; though the DB measure did show
adequate model fit across all age groups [CFI = .964, RMSEA = .07 (90% CI = .06-.08)].
Invariance of SE measures was not achieved across gender or age groups, and configural
invariance showed poor model fit for gender [CFI = .886, RMSEA = .17 (90% CI = .16-.19)] and
age [CFI = .875, RMSEA = .18 (90% CI = .17-.20)]. The existing measure of blood donation DB
was found to have strong invariance across gender groups and weak invariance across age
groups. Results of this study indicate that potential differences between gender and age
groups may exist for confidence to donate blood that are not captured by this scale. Future
efforts aimed at exploring blood donation SE across subgroups and multicultural populations
could ensure that measures reflect meaningful behavior change. This study will inform future
TTM interventions aimed at examining DB and SE for blood donation in minority groups.

CORRESPONDING AUTHOR: Alexandra N. Murray, MA, University of Rhode Island, New

London, CT, 06320; [email protected]
S2106 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C185 6:00 PM-7:00 PM

INTERACTIVE GROUP-BASED ORIENTATION SESSIONS TO IMPROVE ADHERENCE AND DATA

COMPLETION IN PRAGMATIC CLINICAL TRIALS

Allison Bonifay, MA, LPC1, Meghan Mayhew, MPH1, Lynn DeBar, PhD, MPH2, Michaela
Kiernan, PhD3

1
Kaiser Permanente Center for Health Research, Portland, OR; 2Kaiser Permanente, Center for
Health Research, Portland Oregon, Portland, OR; 3Stanford University School of Medicine,
Stanford, CA

Recruitment and retention are challenging for intensive behavior change intervention trials
requiring high patient engagement. Specifically, pragmatic clinical trials typically do not use
incentives and rely on patient-reported outcomes, underscoring the need for successful
recruitment and retention strategies. Guided by methods employed with success in
behavioral weight-loss (Goldberg & Kiernan, 1995) and chronic disease self-management
studies, we developed an interactive group-based orientation session to improve participant
understanding of our chronic pain intervention, address ambivalence about behavior change,
and emphasize the importance of assessing patient-reported outcomes for pragmatic trials in
general. During the orientation session attendees learn about the intervention components,
randomization process, and methodological impact of not completing assessments. They
engage in a small group motivational enhancement-based activity, identifying pros and cons
of being or not being in the study. The facilitator uses reflective listening and remains neutral
regarding study participation.

The Pain Program for Active Coping and Training (PPACT) is a pragmatic trial conducted in
Kaiser Permanente Northwest (NW), Georgia, and Hawaii that evaluates the integration of a
primary care-based intervention versus usual care among chronic pain patients. The
intervention, delivered by an interdisciplinary team, includes comprehensive intake and end-
of-treatment evaluations, 12 weekly cognitive behavioral therapy skills training group
sessions, individualized coaching calls, and physician consult. The primary outcome assesses
patient-reported pain-related severity and functioning using the 4-item Brief Pain Inventory-
short form (BPI) collected quarterly over a 1-year period. Given initial trial adherence, we
conducted interviews with non-adherent participants (attended 6 sessions) that revealed
confusion about intervention and trial requirements. Thus, we began requiring orientation
session attendance prior to enrollment.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2107

Before orientation session implementation (3/2014-12/2015), 234 patients enrolled in NW.

88% completed the3-month BPI assessment and 86% the 6-month; completion varied greatly
between intervention and usual care groups. 115 NW patients were randomized to the
intervention, 30% attended 0 sessions and 34% attended 9-12 sessions. After orientation
session implementation (1/2016), 110 enrolled; 97% completed the 3-month BPI and 98% the
6-month with no difference between study groups. 47 were randomized to intervention, 12%
attended 0 sessions whereas 56% attended 9-12 sessions. The interactive orientation sessions
show promise as a method to improve adherence and data completion for pragmatic trials.

CORRESPONDING AUTHOR: Allison Bonifay, MA, LPC, Kaiser Permanente Center for Health
Research, Portland, OR, 97227; [email protected]
S2108 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C186 6:00 PM-7:00 PM

PREDICTING POST-TRAUMATIC STRESS DISORDER: A MACHINE LEARNING APPROACH

Dejan Magoc, Ph.D.1, Tanja Magoc, Ph.D.2, Joe Tomaka, Ph.D.3, Stormy Monks, Ph.D.4

1
Stetson University, DeLand, FL; 2Jacksonville University, Jacksonville, FL; 3New Mexico State
University, Las Cruces, NM; 4Texas Tech University Health Sciences Center El Paso, El Paso, TX

Introduction: Post-Traumatic Stress Disorder (PTSD) is an anxiety disorder that involves a

specific set of symptoms which develop after experiencing, witnessing, or confronting
stressful and traumatic events. With repeated exposure to such traumatic events, firefighters
are at great risk for developing post-traumatic-stress (PTS) symptoms and related problems
such as alcohol misuse, especially if they use alcohol as means of coping with stress. The
purpose of our study was to build an automated predictor of PTSD in municipal firefighters,
using three machine learning algorithms.

Methods: 740 municipal firefighters completed assessments of PTS symptoms, alcohol

consumption, alcohol problems, drinking motives, and coping with stress as part of a larger
study. We used data on PTS symptoms and alcohol related outcomes to train and test the
machine learning algorithms, including Neural Network (NN), Nave Bayes Method, and
Decision Tree, to build and validate the automated predictor of PTSD in municipal firefighters.

Results: The results of this study indicated that the automatic predictors can successfully
predict PTSD with the accuracy of 88.65% using Nave Bayes and 91.76% using both NN and
Decision Tree.

Conclusions: Even though the results are not 100% perfect, they are highly promising and
show a great potential for quick and early identification of firefighters susceptible to PTSD and
potentially alcohol related problems, which could help focus on less intense interventions
among firefighters, such as education and simple advice rather than counseling, diagnostic
evaluation, and treatment in more severe cases when firefighters with health related
problems are identified in later stages.

CORRESPONDING AUTHOR: Dejan Magoc, Ph.D., Stetson University, DeLand, FL, 32723;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2109

C187 6:00 PM-7:00 PM

PROMIS MEASURES FOR PATIENT CENTERED CARE

Jerry Bounsanga, BS1, Andrew R. Tyser, MD2, Tom Greene, PhD3, Yushan Gu, BS4, Maren
Wright. Voss, MS4, Charles L. Saltzman, MD2, Wendy C. Birmingham, PhD5, Man Hung, PhD4

1
University of Utah Orthopaedics Department, Salt Lake City, UT; 2Universtiy of Utah
Orthopaedic Surgery Operations, Salt Lake City, UT; 3Universtiy of Utah Study Design and
Biostatistics Center, Salt Lake City, UT; 4University of Utah, Salt Lake City, UT; 5Brigham Young
University, Provo, UT

INTRODUCTION:

Measuring treatment response is an indicator of not just an objective treatment outcome but
also the patient-perceived value of treatment. The responsiveness of patient-reported
outcome (PRO) instruments, their ability to detect change over time, is particularly
meaningful when linked to the patients report of meaningful change. This study evaluated
the responsiveness of several Patient-Reported Outcome Measurement Information System
(PROMIS) instruments pre to post treatment in patients with hand and upper extremity
disorders and provided comparisons with the legacy qDASH instrument.

METHODS:
Patients presenting at an orthopaedic hand clinic completed the PROMIS and qDASH
PROsprior to the doctor visit. The responsiveness of the PROMIS Upper Extremity (UE) v1.2,
the PROMIS Physical Function (PF) v1.2, the PROMIS Pain Interference (PI) v1.1, and the
qDASH was assessed. Changes in functional outcomes from baseline visit to follow-up were
evaluated by paired-sample t-test, effect size, and standardized response mean (SRM) at four
different follow-up periods (3-6 months or more) after initial assessment. Institutional Review
Board approval was obtained prior to study commencement.

RESULTS:

The study sample consisted of 255 patients (131 females and 124 males) with an average age
50.75 years (SD = 15.84). Effect sizes for all measures were large and ranged from 0.80-1.48
S2110 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

across the four different follow-up time-points. Responsiveness was high, with SRMs ranging
from 1.05 to 1.63. Changes in patient scores were significant for all instruments (p p=0.253).

DISCUSSION:

The three PROMIS instruments and the qDASH demonstrated the ability to detect patient-
reported health changes at various follow-up time points. This high level of responsiveness in
an orthopaedic hand population provides the necessary information to assess treatment
outcomes in clinical and research settings. Having instruments that can effectively measure a
patients response to treatment is critical for informing clinical care. When treatment
response is adequately measured and studied it becomes a valuable aspect of patient-
centered care, allowing doctors to help patients set expectations for treatment outcomes.

CORRESPONDING AUTHOR: Jerry Bounsanga, BS, University of Utah Orthopaedics

Department, Salt Lake City, UT, 84108; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2111

C188 6:00 PM-7:00 PM

PSYCHOMETRIC PROPERTIES OF A SELF-EFFICACY SCALE AMONG VETERANS WITH

HYPERTENSION

Maria Antonia Rodriguez, PhD, Matthew Zhao, N/A, Binhuan Wang, PhD, Sundar Natarajan,
MD

Department of Veterans Affairs, NY, NY

Background: Hypertension in adults remains a major risk factor for cardiovascular disease, the
leading cause of death in the United States. We conducted a 6-month RCT that followed 533
adult veterans with uncontrolled hypertension, which demonstrated that a tailored stage-
matched intervention was more effective at improving blood pressure control than non-
tailored health education intervention and the usual care. The aim of the current study is to
examine the psychometric properties of the self-efficacy scale utilized in the original study.

Methods: This self-efficacy instrument is comprised of three subscales that assess diet self-
efficacy (DSE), exercise self-efficacy (ESE), and medication adherence self-efficacy (MSE). We
examined the psychometric properties of this self-efficacy instrument in order to determine
its validity and reliability. Cronbachs coefficients were calculated in order to determine
internal consistency. Spearmans Rho correlation coefficients were used in the determination
of convergent and divergent validity, test-retest reliability, as well as exploratory factor
analysis.

Results: The self-efficacy instrument demonstrated strong internal consistency with relatively
high values for Cronbachs alpha: DSE (=.81), ESE (=.82), and MSE (=.74). Between
baseline and 6-month follow up, the instrument showed only some test-retest reliability.
Analysis showed that the instrument possesses both convergent and discriminant validity.
Factor analysis and the Skree plot demonstrated three distinct factors, which correspond to
the three subscales in our self-efficacy instrument.

Conclusion: The results of this investigation have demonstrated that the self-efficacy
instrument utilized in our parent hypertension study for adult male veterans in the United
States suffering from uncontrolled hypertension is a valid and reliable instrument.

CORRESPONDING AUTHOR: Maria Antonia Rodriguez, PhD, Department of Veterans Affairs,

NY, NY, 10010; [email protected]
S2112 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C189 6:00 PM-7:00 PM

REDUCING "DON'T KNOW" RESPONSES AND MISSING SURVEY DATA: IMPLICATIONS FOR
MEASUREMENT

Deanna Denman, MA1, Austin S. Baldwin, PhD1, Emily G. Marks, MS2, Amy McQueen, PhD3,
Jasmin A. Tiro, PhD2

1
Southern Methodist University, Dallas, TX; 2University of Texas Southwestern Medical
Center, Dallas, TX; 3Washington University, St Louis, MO

Dont know (DK) responses are common in health behavior research among underserved
populations (Hay et al., 2015; Waters et al., 2013). These responses are typically treated as
missing data which can limit interpretation of a studys findings (Waters et al., 2013). Effects
of methodologic approaches to minimize DK responses are not well understood.

Methods: As part of a larger study, parental responses to psychosocial variables about

adolescent HPV vaccination (intentions, perceived susceptibility, perceived benefits) were
collected through a telephone survey with DK responses coded as missing data. After one
year, to reduce the large number DK responses, survey procedures were changed and
research assistants were instructed to remind participants of survey response options and
encourage use of the Likert scale (i.e., We just want to know your opinion. Does your opinion
match with any of the following choices?)

Analyses: To assess methodologic effects, survey data were compared before and after
procedural changes in three ways: (1) number of participants providing DK responses; (2)
magnitude of correlations between psychosocial variables; and (3) invariance of the factor
loadings for these latent constructs.

Findings: The percentage of participants responding DK 1 or more times declined from 66.5%
pre- to 22.5% post-procedure change. Mean number of total DK responses per survey was 3.6
vs. 1.3, respectively, out of 41 possible responses. The magnitude of correlations between
scales consistently increased post-change (e.g., for intentions and benefits the correlation
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2113

increased from .404 to .589). Confirmatory factor analyses comparing participants pre- and
post-change showed invariance in the factor loadings for all factors ( 2 = 14.88, p= .317) and
Cronbachs alpha >.70.

Implications: Reminding participants of survey response options appears to be a useful

approach to reduce dont know (missing) responses. Although the measurement properties
of latent variables were consistent over time, having less missing data increased observed
correlations between variables. Our findings suggest that simple reminders of survey response
options can reduce missing data, improve measurement reliability and concurrent validity,
and thus reduce Type II errors.

CORRESPONDING AUTHOR: Deanna Denman, MA, Southern Methodist University, Dallas, TX,
75206; [email protected]
S2114 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C190 6:00 PM-7:00 PM

RESPONSIVENESS OF THE PROMIS INSTRUMENTS FOR PAIN AND FUNCTION

Maren Wright. Voss, MS1, Charles L. Saltzman, MD2, Yushan Gu, BS1, Jerry Bounsanga, BS3,
Tom Greene, PhD4, Wendy C. Birmingham, PhD5, Man Hung, PhD1

1
University of Utah, Salt Lake City, UT; 2Universtiy of Utah Orthopaedic Surgery Operations,
Salt Lake City, UT; 3University of Utah Orthopaedics Department, Salt Lake City, UT;
4
Universtiy of Utah Study Design and Biostatistics Center, Salt Lake City, UT; 5Brigham Young
University, Provo, UT

INTRODUCTION:

Healthcare is increasingly concerned with providing treatment that meets the patients
expectations and needs. Patient reported outcomes (PRO)s are a way to measure treatment
outcomes from the patient perspective. A PRO that can measure the level of response to
treatment must be sensitive enough to detect change over time. This study reported the
responsiveness to change of two Patient-Reported Outcomes Measurement Information
System (PROMIS) instruments PROMIS Physical Function (PF) v1.2 and PROMIS Pain
Interferences (PI) v1.1, in an orthopedic sample with foot and ankle ailments pre- and post-
treatment.

METHODS:

Patients aged 18 and older presenting to an orthopaedic foot and ankle clinic between the
years 2014 and 2016 responded to the PROMIS PF and PROMIS PI while in the clinic waiting
area at the initial clinic visit as well at follow-up visits at 3-6 months and beyond. Analysis of
responsiveness to changes in functional and pain outcomes were analyzed by paired sample t-
test, effect size, and standardized response mean. In order to assess change levels that were
considered meaningful from the patient perspective, analyses were anchored by patient
responses to the question: How much relief and/or improvement do you feel you have
experienced as a result of your treatment? Institutional Review Board approval was obtained
for this study.

RESULTS:
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2115

A total of 1,715 patients completed follow-up measures either at three months and/or six
months (1,010 females and 704 males). Their average age was 52.01 years (SD = 16.54).
Effect sizes for the PROMIS PF ranged from 0.92 and 1.14 and for the PROMIS PI ranged from
0.89 and 1.02. PROMIS PF and PIs standardized response means had a narrow range across
time points of 1.04-1.28. Paired sample t-tests indicated these changes from pre to post
treatment were significant (p < 0.05) for the PROMIS PF and PI at all four different follow-up
periods.

DISCUSSION:

As the use of the PROMIS instruments becomes more widespread in health outcomes
research, it is important that these instruments are responsive to medical treatment. The
PROMIS PF and the PROMIS PI were sensitive to patient-reported health changes in this lower
extremity orthopaedic population. Clinical access to information about treatment response
levels made available by research conducted with responsive instruments will assist clinicians
in setting expectations with patients about likely treatment outcomes.

CORRESPONDING AUTHOR: Maren Wright. Voss, MS, University of Utah, Salt Lake City, UT,
84108; [email protected]
S2116 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C192 6:00 PM-7:00 PM

LIFE AND TREATMENT GOALS OF VETERANS WITH GULF WAR ILLNESS

Lisa McAndrew, Ph.D.1, Nicole Anastasides, M.S.2, Carmelen Chiusano, MPA3, Melanie
Chelenza, LMHC4, Fiona Graff, Psy.D.5, Christina G. Gonzalez, B.A.2, Drew Helmer, MD6,
Lauren M. Greenberg, Ph.D.7, David R. Litke, Ph.D.5, Shou-En Lu, PhD8, Erica R. McDonald,
B.S.9, Beth Ann Petrakis, MPA10, Wilfred R. Pigeon, Ph.D.11, Jennifer Presnall-Shvorin, Ph.D.2,
Karen Quigley, Ph.D.12, Joseph F. Rath, PhD13

1
University at Albany and War Related Illness and Injury Study Center, Veterans Affairs New
Jersey Healthcare System, East Orange, NJ; 2War Related Illness & Injury Study Center
(WRIISC), VA New Jersey Health Care System, East Orange, NJ; 3VA New Jersey Health Care
System, War Related Illness & Injury Study Center, East Orange, NJ; 4Department of Veterans
Affairs, VISN 2 Center of Excellence for Suicide Prevention, Canandaigua, NY; 5War Related
Illness and Injury Study Center (WRIISC), VA New Jersey Health Care System, East Orange, NJ;
6
Veterans Affairs, East Orange, NJ; 7War Related Illness and Injury Study Center, VA New
Jersey Health Care System, East Orange, NJ; 8Rutgers School of Public Health, Piscataway, NJ;
9
War Related Illness & Injury Study Center (WRIISC), VA New Jersey Health Care System,
Washington, NJ; 10Edith Nourse Rogers Memorial VA Hospital, Bedford, MA; 11VISN 2 Center
of Excellence for Suicide Prevention, VHA, Canandaigua, NY; 12Edith Nourse Rogers (Bedford,
MA) Memorial VA Hospital, Bedford, MA; 13NYU Rusk Institute of Rehabilitation Medicine,
New York, NY

Gulf War Illness is characterized by multiple medically unexplained symptoms. Current

guidelines for Gulf War Illness focus on improving quality of life through behavioral
treatments (e.g., Cognitive Behavioral Therapy, Graded Exercise). There is a need to better
tailor behavioral treatments for Gulf War Veterans to improve adherence and outcomes. Life
and treatment goals impact motivation for treatment. Identifying Gulf War Veterans goals for
their life and treatment is important as they could be used to tailor treatments.

Methods: We recruited 117 Gulf War Veterans with Gulf War Illness for a randomized
controlled trial comparing Problem-Solving Therapy to Health Education. At baseline,
Veterans were asked their goals for life and their goals for treatment. We developed codes
based on Veterans responses.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2117

Results: The most prevalent life goals were to improve their quality of life (interpersonal
relationships (21%), career (13%), a good life (13%), live longer (9%) and spirituality (6%)) and
to improve heath (both physical (20%) and emotional (14%)). The most common treatment
goals were to improve health (both physical (31%) and emotional (20%)), improve quality of
life (15%), to engage in treatment (10%), and understand health (10%).

Conclusions: Current treatment recommendations for Veterans with Gulf War Illness are
primarily to improve quality of life. This is consistent with Gulf War Veterans own primary life
goals (to improve quality of life (47%)). Gulf War Veterans goals for treatment are more
focused on improving (51%) and understanding (10%) their health. Current treatment
recommendations for Gulf War Illness are more consistent with Veterans goals for their life
than their goals for treatment. Behavioral treatments should focus on explicitly connecting
treatments to Veterans goals for their life, and addressing inconsistencies between Veterans
goals for treatment and what behavioral treatments can be expected to achieve.

CORRESPONDING AUTHOR: Lisa McAndrew, Ph.D., University at Albany and War Related
Illness and Injury Study Center, Veterans Affairs New Jersey Healthcare System, East Orange,
NJ, 07018; [email protected]
S2118 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C193 6:00 PM-7:00 PM

POST-TRAUMATIC STRESS DISORDER AND DEPRESSION ARE ASSOCIATED WITH TINNITUS

HANDICAP IN A SAMPLE OF MILITARY VETERANS

Matthew Valente, B.S.1, Michelle Sisson, M.A.1, Sumner Sydeman, Ph.D.1, Samantha Russell,
M.A.2, O'neil Guthrie, Ph.D.1

1
Northern Arizona University, Flagstaff, AZ; 2University of Montana, Missoula, MT

Tinnitus is the most common disabling condition among veterans, with an estimated two
billion dollars spent annually by the U.S. government in disability compensation. Psychological
distress is frequently co-morbid with tinnitus, however, few studies have examined tinnitus,
PTSD, and depression in veterans. Thus, the purpose of the current project was to explore the
relationship of PTSD and depression with tinnitus handicap among United States veterans
with tinnitus. Data were retrospectively extracted from patient medical records at a tinnitus
clinic for ninety-five veterans ranging from 31-84 years of age (M = 62.00) at a veterans
administration hospital and included audiometric information, data on tinnitus, and measures
on PTSD and depression. PTSD was considered present if the veteran had an ICD-9 code of
Post-Traumatic Stress Disorder in their medical records. Furthermore, depressive
symptomatology was measured using the Hospital Anxiety and Depression scale depression
scale (HADS-D). Tinnitus was assessed using the Tinnitus Handicap Inventory (THI) total score.
On average, veterans with tinnitus and PTSD reported significantly higher levels of tinnitus
handicap (M = 59.75, SD = 22.12) than those with tinnitus alone (M = 39.86, SD = 26.78), t(93)
= 3.01, p = .003, 95% CI [-33.02, -6.76], d = 0.74. A simple linear regression analysis regressing
tinnitus handicap on depression found that a significant proportion of the total variation in
tinnitus handicap was predicted by the HADS-D, R2 = .319, F(1, 93) = 43.50, p < .001, 95% CI
[2.04, 3.79]. Lastly, a multiple linear regression identified that a significant amount of the
variance in THI scores was accounted for by the combination of PTSD and depression; R2adj =
.321, F (2, 92) = 23.24, p < .001, 95% CI [13.90, 36.91] with depression (b = .52, p < .001), but
not PTSD (b = .14, p = .129) being a significant predictor. Findings suggest that both PTSD and
depression are associated with tinnitus handicap among military veterans. Future research
needs to explore effective strategies (e.g. SSRIs, CBT, sound therapies, etc.) to reduce the risk
of tinnitus severity among veterans experiencing psychological distress.

CORRESPONDING AUTHOR: Matthew Valente, B.S., Northern Arizona University, Flagstaff, AZ,
86001; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2119

C194 6:00 PM-7:00 PM

THE ROLES OF PAIN AND INJURY IN THE CURRENT STRESS-PTSD RELATIONSHIP AMONG
OEF/OIF/OND VETERANS

Sharon Y. Lee, MS, MA1, Crystal L. Park, PhD2, Rani Hoff, MPH, PhD3

1
University of Connecticut, Vernon, CT; 2University of Connecticut, Storrs, CT; 3Yale School of
Medicine, West Haven, CT

Pain and PTSD co-occur at prevalence rates as high as 16.5% among OEF/OIF/OND veterans.
To explain the high comorbidity, Sharp and Harvey (2001) proposed the mutual maintenance
model of pain and PTSD, in which properties of one condition feed into the escalation of
symptoms for the other. The relationship is maintained, in part, because pain triggers
reminders of the trauma, which may be particularly salient for veterans injured during
combat. Psychological states also alter pain perceptions, such that current stress increases
pain perceptions. However, the mediating role of pain in the current stress-PTSD relationship
has yet to be tested. The present study tested a moderated mediation model to examine if
sustaining a combat-related injury moderates the pain-PTSD relationship, and if the strength
of the indirect effect of current stress on PTSD is conditional on whether an injury is
sustained. The current study is limited to the 729 veterans from the Survey of the Experiences
of Returning Veterans study who completed phone surveys at baseline, 3-months, and 6-
months. Consistent with our hypotheses, pain partially mediates the effect of current stress
on PTSD. Further, greater pain predicted greater PTSD (B=4.07, p < .001), particularly for
veterans with a combat-related injury (B=.92, p < .001). Overall, the moderated mediation
model explains more of the effect of current stress on PTSD (B=2.85, p < .001) than did the
direct effect of current stress on PTSD (B=2.64, p < .001). Our study shows that pain is a
mechanism by which current stress affects PTSD, which may be particularly important for
veterans with combat-related injuries. Our findings about the pain-PTSD relationship support
and build upon the mutual maintenance model by demonstrating that the relationship can
unfold over several months. Given that many veterans perceive mental health care as
stigmatized in military culture, veterans may be more amenable to targeting pain as an
indirect way of treating PTSD. Treatments targeting pain may provide a less stigmatized and
more comprehensive approach to improving well-being.

CORRESPONDING AUTHOR: Sharon Y. Lee, MS, MA, University of Connecticut, Vernon, CT,
06066; [email protected]
S2120 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C195 6:00 PM-7:00 PM

ACCULTURATION AND WEIGHT CHANGE IN ASIAN-AMERICAN CHILDREN: EVIDENCE FROM

THE ECLS-K:2011

Cassandra S. Diep, PhD1, Tom Baranowski, PhD2, Rachel T. Kimbro, PhD3

1
The University of Texas MD Anderson Cancer Center, Houston, TX; 2Baylor College of
Medicine, Houston, TX; 3Rice University, Houston, TX

Background: Despite relatively low rates of overweight and obesity among Asian-American
children, disparities exist based on acculturation, socioeconomic status, and Asian ethnicity.
Purpose: The purpose of this study was to examine the association between acculturation and
weight change in Asian-American children. Secondary aims were to compare changes by Asian
ethnic group and acculturation x socioeconomic status.
Methods: Participants included 1,200 Asian-American children from the Early Childhood
Longitudinal Study, Kindergarten Class of 2010-11, a longitudinal study of U.S. children
attending kindergarten in 2010-2011. Multinomial logistic regressions were conducted to
predict weight change, which was categorized based on body mass indices in kindergarten
(spring 2011) and second grade (spring 2013): consistently healthy weight, consistently
overweight/obese, healthy weight change, and unhealthy weight change. Models included
demographic, household, socioeconomic status, and acculturation measures, specifically
mothers English proficiency and percentage of life spent in the U.S.
Results: Overall, 72.3% of children were at healthy weights in kindergarten and in second
grade. Filipino children, when compared to Chinese children, had higher odds of being in the
consistently overweight/obese category than consistently healthy weight (OR: 2.26, p <
.05). In addition, mothers English proficiency significantly predicted unhealthy weight change
(OR: 0.83, p < .001) and healthy weight change (OR: 0.75, p < .001), relative to consistently
healthy weight. English proficiency also predicted being consistently overweight/obese but
only for children with less educated mothers.
Conclusion: Findings enhance our understanding of obesity disparities within Asian Americans
and highlight the need to disaggregate the population. Specifically, obesity interventions are
needed for Filipino children and families with low socioeconomic status but high English
proficiency.

CORRESPONDING AUTHOR: Cassandra S. Diep, PhD, The University of Texas MD Anderson

Cancer Center, Houston, TX, 77230-1402; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2121

C196 6:00 PM-7:00 PM

MERITORIOUS AWARD WINNER

ADOLESCENT DEPRESSION MEDIATES LONG-TERM IMPACT OF CUMULATIVE MATERNAL

DEPRESSION EXPOSURE ON ADOLESCENTS BODY MASS INDEX

Amber Cordola Hsu, MPH1, Zhongzheng Niu, MS, MD2, Emily Kiresich, MS, RD3, Bin Xie, PhD2

1
Claremont Graduate University, Marina del Rey, CA; 2Claremont Graduate University,
Claremont, CA; 3Claremont Graduate University, Lakewood, CA

Introduction

Prior research shows that maternal depressive symptoms may influence childs risks of
both depression and obesity. However, the potential mediating role of childs depression in
the linkage of maternal depression to adolescents obesity remain unclear in the literature. In
this study, we use existing data from the Eunice Kennedy Shriver National Institute of Child
Health and Human Development Study of Early Child Care and Youth Development (SECCYD)
to evaluate the mediation effects of adolescents depression on the pathway from cumulative
maternal depression exposure to adolescents BMI at age 15.

Method

The analysis sample size includes 844 children and their families from childs age 1 month
to 15 years old. Weight and height measured at age 15 years were used to calculate BMI and
define overweight/obesity status. Maternal depression reported by mother of the study child
as repeated assessed from childs age 1 month to grade 6 by the Center for Epidemiological
Studies Depression Scale (CES-D). Cumulative measures of exposure to maternal depression
were developed to capture the dynamic variation across time and were used to prospectively
link to adolescents BMI and overweight/obesity status. Childrens Depression Inventory (CDI)
at age 15 years was used as the mediator in the analysis. Multivariate linear regressions were
employed and mediational effects were estimated with bias-corrected bootstrapping
resampling approach to obtain 95% confidence intervals for indirect effects. Covariates
included childs gender, mother's delivery age, mother's minority status (dichotomous),
mother's educational level (years), and income-to-need ratio at grade 5 were also included in
the analysis.

Results
S2122 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Prospective analysis linking cumulative exposure to maternal depression by grade 6 of the

study child to childs BMI percentile at age 15 revealed a statistically significant positive
relationship (total effect = 0.35, p = 0.03, 95% CI: 0.02, 0.67). The childs depression at age 15
was included as a mediator and a statistically significant mediation effect was observed
(indirect effect = 0.06, 95% CI: 0.01, 0.12). The direct effect of cumulative maternal depression
by grade 6 on childs BMI percentile at age 15 was not statistically significant (direct effect =
0.29, 95% CI: -0.03, 0.61).

Conclusion

Our findings revealed a prospective positive effect of cumulative exposure to maternal

depression from age 1 month to grade 6 on adolescents BMI percentile at age 15, which was
fully mediated by adolescents depression. The observed significant pathway from maternal
mental health to childs mental health status, which in turn has impact on childs obesity risk.
Our findings highlight the needs to incorporate mental health efforts in pediatric obesity
prevention.

CORRESPONDING AUTHOR: Amber Cordola Hsu, MPH, Claremont Graduate University, Marina
del Rey, CA, 90292; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2123

C197 6:00 PM-7:00 PM

AN EXAMINATION OF DIETARY COMPOSITION PRECEDING LAPSES FROM A WEIGHT CONTROL

DIET

Cara Dochat, B.A.1, Stephanie Goldstein, M.S.1, Brittney Evans, BA1, Evan M. Forman, PhD1,
Alexis Wojtanowski, N/A2, Gary Foster, Ph.D.3

1
Drexel University, Philadelphia, PA; 2Weight Watchers Intl, New York, NY; 3Weight Watchers,
New York, NY

Traditional behavioral weight loss programs instruct participants to consume less than a
target number of daily calories as the primary mode of weight loss. Though programs typically
recommend a low-fat, high-fiber diet, they rarely address absolute and relative intake of
macronutrients, sugar, and fiber. In these programs, frequency of lapsing from ones diet plan
(e.g., eating a forbidden food, eating when one had not planned, or eating more than
planned) predicts less weight loss. Preliminary evidence suggests that feelings of hunger,
cravings, and deprivation increase likelihood of lapsing, yet little is known about whether and
how the quality of ones diet predicts lapsing independently, which is examined in the current
study for overweight individuals following a self-directed Weight Watchers (WW) plan.
Adults (n=12) completed six weeks of dietary self-monitoring using the WW mobile phone
app, and responded 6 times per day to EMA prompts from another app regarding whether a
lapse had occurred since the last prompt (90.9% of prompts completed). Separate generalized
estimating equations were used to examine whether dietary intake at one time of day
(morning or afternoon) prospectively predicted lapse occurrence at the following time of day
(afternoon or evening). Reported here are results from study Phase I; Phase II (n=30) will also
be reported for presentation. Results indicated that WW PointsPlus (=-.04, p=.04) and fiber
(=.032, p=.07)but not calories, fat, carbohydrate, protein or sugar (s < .01, ps>.05)
negatively predicted lapsing at the following meal time, such that greater points consumption
(but lower fiber consumption) predicted lower likelihood of lapsing. Because WW points are a
composite score (reflecting macronutrient composition) which represent the balance of
calories and macronutrients in a meal, meals with relatively higher WW points may be more
satiating, thus decreasing likelihood of future lapse. An alternative explanation for our
findings is that participants lapsed less frequently because they were simply less hungry;
however, null findings for the predictive validity of calorie intake alone suggest that lapsing is
not necessarily a result of hunger or low energy, and that higher energy intake alone is not
protective against lapsing. Future work in this domain may help weight loss programs
construct dietary plans that best facilitate adherence and weight loss.
S2124 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Cara Dochat, B.A., Drexel University, Philadelphia, PA, 19130;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2125

C198 6:00 PM-7:00 PM

MERITORIOUS AWARD WINNER

BEHAVIORAL PHENOTYPES IN OVERWEIGHT AND OBESE ADULTS: PATHWAYS TO THE ONSET

AND MAINTENANCE OF OBESITY

Daniel Flack, M.A.1, Stephanie M. Manasse, MS1, Brittney Evans, BA1, Ross D. Crosby, PhD2,
Meghan L. Butryn, PhD1, Evan M. Forman, PhD1

1
Drexel University, Philadelphia, PA; 2Neuropsychiatric Research Institute, Fargo, ND

Even our best behavioral weight loss treatments produce sub-par outcomes in a large
proportion of those receiving them. Better understanding the heterogeneity within obese
populations, especially in the differential pathways to obesity, could inform enhancements to
standard behavioral treatments. Identification of distinct pathways comprised of potentially
malleable psychological factors associated with obesity (such as food reward sensitivity,
susceptibility to internal and external cues, depressive symptoms, and impulsivity) would be
especially relevant for treatment development. As such, the current study utilized latent
profile analysis to identify obesity phenotypes within a treatment-seeking sample of
overweight and obese individuals (n=190).

Indicator variables included self-report measures of susceptibility to external cues, emotional

eating, depression, food reward sensitivity, and a behavioral measure of impulsivity. A five-
class solution emerged from latent profile analysis (LPA), comprised of (1) an unknown
vulnerability class (low food and emotional responsivity, low impulsivity, low hedonic
hunger); (2) an impulsive class (high impulsivity, but otherwise better than average across
other measures); (3) an impulsive plus class (high impulsivity, elevated other measures
relative to the impulsive class); (4) a cue reactive class (high food and emotion
responsivity, low general impulsivity); and (5) a disinhibited-hedonic class (high food and
emotion responsivity, poor impulsivity, high hedonic hunger). Additionally, post-hoc analyses
revealed that classes differed on both baseline physical activity levels and weight loss
following a behavioral weight loss intervention.

To our knowledge, this was the first study to utilize LPA to identify phenotypes of obesity
within a treatment-seeking sample of obese adults. The findings that these behavioral
phenotypes may effectively discriminate between both baseline and post-treatment
measures of weight-related behaviors strongly support the hypothesis that combinations of
reward sensitivity, susceptibility to cues, depression, and impulsivity do relate to the
S2126 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

maintenance and etiology of obesity. Future research should incorporate these ideas when
attempting to create and refine refine obesity treatment and prevention programs.

CORRESPONDING AUTHOR: Daniel Flack, M.A., Drexel University, Philadelphia, PA, 19146;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2127

C199 6:00 PM-7:00 PM

CONSCIENTIOUSNESS AND BODY MASS INDEX: MEDIATION BY EATING BEHAVIORS

Natalie Keirns, B.S., Misty AW Hawkins, Ph.D.

Oklahoma State University, Stillwater, OK

Background: Over 30% of American adults are obese. This high prevalence combined with the
adverse consequences of excess adiposity support the need for continued identification of risk
factors in obesity development and progression. Such research is essential given that obesity
is resistant to treatment in the long term. Personality factors are now being examined in
relation to health outcomes, and conscientiousness has emerged as the most robust
personality predictor of obesity. Relatedly, conscientiousness may also predict food choices. It
is less clear which facets of conscientiousness are related to obesity and whether disordered
eating accounts for observed relationships. Objective: Thus, the objective of our study was to
discover which facets of conscientiousness predict body mass index (BMI), and if any observed
relationships are mediated by eating behaviors. Methods: Participants were 523 young adults
(age: 19.9 1.8 years, female: 400 (76.5%), non-white: 113 (21.6%), BMI: 25.0 5.8 kg/m2)
who completed online self-report questionnaires assessing personality and eating-related
variables. Conscientiousness (CN) was assessed with the M5-120. Excess adiposity was
estimated continuously using BMI. Eating behaviors were measured with the Three Factor
Eating Questionnaire (TFEQ) and its subscales of Cognitive Restraint, Disinhibition, and
Perceived Hunger. Data were analyzed using PROCESS regression-based mediation macros in
SPSS. Covariates included age, gender, race, parents education, parents occupation, and
English as primary language. Results: Linear regression results indicated that total domain
score of CN did not significantly predict BMI (b =-.07, p=.096). When the facets were examined
Orderliness and Achievement Striving did predict BMI (b =-.15, p < .001; b=.13, p=.037).
Mediation analysis of the Orderliness facet revealed that Disinhibition (b =-.08, CI[-.13, -.04], p
< .001) mediated the relationship. The relationship between Achievement Striving and BMI
was not mediated by eating behaviors. Conclusions: Results indicate that orderliness and
achievement striving predict BMI in young adults. For orderliness, this relationship was
mediated by disinhibited eating. As personality is an enduring, lifelong trait, it may be
beneficial to provide tailored obesity prevention and intervention programs for individuals
low in orderliness. Further research into the relationship between achievement striving and
BMI is needed.
S2128 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Natalie Keirns, B.S., Oklahoma State University, Stillwater, OK,
74078; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2129

C200 6:00 PM-7:00 PM

DIETARY OUTCOMES WITHIN THE STUDY OF NOVEL APPROACHES TO WEIGHT GAIN

PREVENTION (SNAP) RANDOMIZED CONTROLLED TRIAL

Jessica G. LaRose, PhD1, E. Whitney Evans, PhD, RD2, Rebecca Neiberg, MS3, Letitia Perdue,
MS3, Karen Hatley, MPH4, Deborah F. Tate, PhD5, Mark Espeland, PhD3, Amy A. Gorin, Ph.D.6,
Cora E. Lewis, MD7, Erica Robichaud, MSW, RD8, Rena R. Wing, PhD2

1
Virginia Commonwealth University School of Medicine, Richmond, VA; 2The Miriam Hospital
/ Brown Medical School, Providence, RI; 3Wake Forest University School of Medicine,
Winston-Salem, NC; 4UNC Chapel Hill / Lineberger Cancer Center, Chapel Hill, NC; 5University
of North Carolina at Chapel Hill, Chapel Hill, NC; 6University of Connecticut; Institution for
Collaboration on Health, Intervention, and Policy, Storrs, CT; 7University of Alabama at
Birmingham, Birmingham, AL; 8The Miriam Hospital, Providence, RI

Background: Young adulthood has been identified as a high-risk period for unhealthy dietary
behaviors, as well as significant weight gain. Previously we reported that two innovative self-
regulation approaches were effective in reducing this weight gain over 2 years. Here we
examine changes in key dietary behaviors and their association with weight change.

Methods: Participants were 599 young adults, age 18-35 years with BMI of 21.0 to 30.0 kg/m2
(27.44.4 years; 25.42.6 kg/m2; 22% men; 73% non-Hispanic White), who were randomized
to self-regulation with Small Changes (SC), self-regulation with Large Changes (LC) or Control
(C). Both SC and LC emphasized frequent self-weighing to cue behavior changes. SC promoted
daily small decreases in intake and increases in physical activity to produce a 200 kcal/day
change. LC focused on making larger changes to produce initial weight losses of 2.3-4.5 kg to
buffer against anticipated weight gains. Both approaches directly targeted high-risk dietary
behaviors associated with weight gain in young adulthood, including fat, fast food, sugared
beverages, alcohol, and breakfast consumption. Dietary behaviors were assessed at baseline,
4 months (i.e., post initial intervention) and 2 years using the Block Food Frequency
Questionnaire and another questionnaire on dietary patterns. Weight was assessed
objectively at baseline, 4 months, 1 and 2 years.

Results: Although the 3 groups all differed in weight gain over time (C>SC>LC), there were no
significant differences between groups for changes in fast food or other meals away from
home, or overall dietary quality as measured by the HEI-2010. However, compared to Control,
both LC and SC trended toward greater decreases in alcohol intake (p=.08) and more frequent
S2130 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

breakfast consumption (p=.07) at 4 months. Moreover, the relationships of weight change at

4 months with improvements in breakfast consumption (p=.02), alcohol intake (p=.05) and
overall dietary quality (p=.005) differed across the 3 groups, with significant associations with
weight change in LC relative to C.

Conclusions: This study suggests that decreased alcohol intake, increased breakfast
consumption, and improved dietary quality might be important dietary targets within weight
gain prevention programs for high-risk young adults.

CORRESPONDING AUTHOR: Jessica G. LaRose, PhD, Virginia Commonwealth University School

of Medicine, Richmond, VA, 23219; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2131

C201 6:00 PM-7:00 PM

GRIT'S EFFECT ON WEIGHT LOSS THROUGH INCREASED TREATMENT ADHERENCE: A PILOT

STUDY

Caitlin Smith, B.S., Misty AW Hawkins, Ph.D.

Oklahoma State University, Stillwater, OK

Background: Obesity is a public health crisis affecting 37% of U.S. adults. Although existing
interventions have proven successful in the short-term, long-term weight loss is rare
highlighting the need to identify factors that foster long-term success. One such factor is grit,
defined as the perseverance/persistence for long-term goals. Lower levels of grit are
associated with higher body mass index, suggesting grit may be linked to weight regulation
efforts. Given that grit has been associated with greater adherence in the academic literature,
higher levels of grit may also predict adherence and success to weight loss treatment.

Objective: To examine the predictive ability of grit on treatment adherence and percentage
weight loss (%WL) among overweight/obese persons in a larger, ongoing NIH-funded
behavioral weight-loss intervention. We hypothesized that grit would have a positive effect on
%WL through increased treatment adherence.

Method: The pilot sample was 26 adults (BMI27 kg/m2) enrolled in a randomized controlled
trial for behavioral weight-loss (aged 41.8111.13 years; 81% female). Baseline grit was tested
using the 12-item GRIT scale, and treatment adherence was indexed by the percentage of
treatment sessions attended. %WL was calculated as follows: (baseline weight post
treatment weight/baseline weight)*100. The hypothesized indirect effect was tested using
1,000 bootstrapped resamples.

Results: Results indicated a non-significant direct effect for grit (B=.3919, SE=2.6415, 95% BC
[-5.2085, 5.9922]) on %WL. A significant indirect effect was found for grit on greater %WL
through increased treatment adherence (B=2.6513, SE=1.5122, 95% BC [0.2014, 6.1875].
Participants with higher levels of grit had nearly 10% points higher adherence to treatment
when compared to participants with lower levels of grit (79% vs. 70% adherence,
respectively). Additionally, participants who attended 80% or more treatment sessions had a
mean %WL of 8.91, while participants who did not had a mean %WL of 5.31.

Conclusions: Grit predicted a persons ability to adhere to weight-loss treatment or not, which
in turn predicted weight loss success or not. Though replication is clearly needed, grit may be
a promising target for screening at-risk clients and improving outcomes. If the findings are
S2132 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

replicated (data collection is on-going), the next steps include determining whether enhancing
grit is possible in an effort to foster greater weight loss success and maintenance.

Funding Sources: K23DK103941-01A1 awarded to MAWH & Oklahoma Shared Clinical and
Translational Resources NIH-Funded Summer Research Training Program awarded to CES.

CORRESPONDING AUTHOR: Caitlin Smith, B.S., Oklahoma State University, Stillwater, OK,
74078; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2133

C202 6:00 PM-7:00 PM

HOW ACCURATE ARE RECALLS OF SELF-WEIGHING FREQUENCY? DATA FROM A 24-MONTH

RANDOMIZED TRIAL

Melissa M. Crane, PhD, Jennifer A. Linde, PhD

University of Minnesota, Minneapolis, MN

Background: Recommendations for frequency of self-weighing during weight loss have not
been determined but are needed. In order to properly assess the behaviors needed to create
recommendations, methods of measuring self-weighing need to be compared and validated.
The purpose of this analysis is to compare self-reported self-weighing frequency to objective
self-weighing recorded via wireless scales.

Methods: Data come from the 24-month Tracking Study. All participants in the study received
12 months of a group-delivered behavioral weight loss program and were randomly assigned
to receive 1 of 3 different self-weighing recommendations: 1) do not weigh at home, 2) weigh
weekly at home (WEEKLY), or 3) weigh daily at home (DAILY). Those in the WEEKLY and DAILY
conditions received Wi-Fi enabled scales that automatically recorded weights to a web-based
platform. Participants were asked on a seven-point scale how frequently they weighed
themselves (1=never, 7=more than once a day) during assessment visits every 6 months. Data
from the WEEKLY and DAILY conditions at baseline, 12-month, and 24-month assessments are
included here.

Results: Participants (N=223; age 46.610.2 years; BMI 33.03.6; 85% white; 66% female)
were assigned to the DAILY or WEEKLY groups. At baseline, 57% of participants reported
weighing themselves less than once per week while 9% reported weighing daily. At 12 and 24
months, 87% and 43% of DAILY participants reported weighing themselves daily while 91%
and 62% of WEEKLY participants reported weighing themselves weekly (ps < 0.001). During
the intervention (from baseline to 12 months), DAILY participants weighed themselves an
average of 4.81.8 times per week while WEEKLY participants weighed themselves 0.90.3
times per week (p < 0.001). Post-intervention (from 12 to 24 months), DAILY participants
weighed themselves an average of 1.71.9 times per week compared to 0.60.8 in the
WEEKLY condition (p < 0.001). At the 12 month assessment, regardless of assigned frequency,
those who reported weighing themselves daily had weighed themselves an average of 5.31.6
times per week over the first year vs. 1.20.8 times for those who reported weekly self-
weighing (p < 0.001). At 24 months, individuals who self-reported daily weighing weighed
themselves 2.72.3 times per week between the 12 and 24 month assessments compared to
0.91.0 times per week in individuals who reported weekly weighing (p < 0.001).
S2134 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions: Self-reported weighing frequency is able to correctly order frequency of self-

weighing but may overestimate daily self-weighing. This overestimation was greater during
the follow-up period without continued intervention contact. Future studies should assess the
concordance of self-reported and objective weighing in the absence of specific self-weighing
recommendations.

CORRESPONDING AUTHOR: Melissa M. Crane, PhD, University of Minnesota, Minneapolis,

MN, 55454; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2135

C203 6:00 PM-7:00 PM

HOW DO DIETARY LAPSES IMPACT WEIGHT LOSS? EXAMINATION OF LAPSES AND ENERGY
INTAKE

Brittney Evans, BA1, Stephanie P Goldstein, M.S.2, Cara Dochat, B.A.1, Evan M. Forman, PhD1,
Alexis Wojtanowski, N/A3, Gary Foster, Ph.D.4

1
Drexel University, Philadelphia, PA; 2Drexel University, Philadelphi, NJ; 3Weight Watchers
Intl, New York, NY; 4Weight Watchers, New York, NY

Failure to successfully adhere to a weight loss diet can be attributed to discrete instances in
which an individual violates a dietary recommendation, i.e., dietary lapses. Research on the
characteristics of lapses and the magnitude of their contribution to excess caloric intake is still
in its infancy. To better understand if and how to target lapses in the context of a behavioral
weight loss program, it is important to clarify the nature of lapses (e.g., frequency, size,
macronutrient content) and investigate whether lapse frequency relates to broader weight
loss failures. Participants (n=12) tracked dietary intake through Weight Watchers mobile app
for six weeks and used another app that repeatedly assessed lapses from the programs
dietary prescription. Lapse frequency, nutritional content of lapse vs. non-lapse eating
occasions, and the associations between lapse frequency, daily caloric intake, and dietary
goals were examined. On average, participants lapsed 3.47 times per week (SD=2.41) and the
number of daily lapses ranged from 0-2. Lapse frequency decreased over time spent in the
program (p < .01). As expected, caloric intake at eating occasions when a lapse was
reported (M=742.49, SD=536.26) was greater than occasions when a lapse was not reported
(M=579.61, SD=551.20), indicating that lapse reports represented substantially higher caloric
intake (p < .001). Converting macronutrient grams to calories, lapse eating occasions were
higher in fat (mean difference=81.18, p < .01), protein (mean difference=22.77, p < .01),
and sugar (mean difference=2.93, p < .01) than non-lapse occasions. Frequency of lapses was
correlated with the degree to which participants exceeded their daily dietary goals (r=.26, p <
.001) and total number of daily calories consumed (r=.14, p< .001). Overall, results confirmed
expected findings that lapses are related to excess caloric intake. Findings underscore the
importance of interventions to reduce frequency of lapses in order to promote success in
weight loss programs.

CORRESPONDING AUTHOR: Brittney Evans, BA, Drexel University, Philadelphia, PA, 19104;
[email protected]
S2136 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C204 6:00 PM-7:00 PM

IMPACT OF ADVERSE CHILDHOOD EXPERIENCES ON RISK FOR FOOD ADDICTION AND

SUBSTANCE USE IN OBESITY TREATMENT SEEKING INDIVIDUALS

Allison Holgerson, Ph.D.1, Matthew Clark, Ph.D.1, Gretchen Ames, Ph.D.2, Maria Collazo-
Clavell, M.D.1, Todd Kellogg, M.D.1, Karen Graszer, M.A.1, Sarah Kalsy, M.A.1, Karen Grothe,
Ph.D.1

1
Mayo Clinic, Rochester, MN; 2Mayo Clinic, Jacksonville, FL

Personal history of adverse childhood experiences (ACE) is a known risk factor for
psychopathology (Caslini et al., 2015; Clark et al., 2007) and substance misuse (Dube et al.,
2003). In addition, there is a strong association between ACE and disordered eating (Moulton
et al., 2015), and lifetime risk for developing obesity (Danese & Tan, 2014; Grilo et al., 2005).
This is an important area of study as some survivors of ACE will also experience psychiatric
symptoms after weight loss surgery (Clark et al., 2007). Food addiction has also been shown to
have a positive relationship with ACE (Imperatori et al., 2016; Mason et al., 2013). The present
study evaluated the risk for food addiction, alcohol abuse, and tobacco and drug use
associated with a history of ACE. Participants were 1,609 individuals seeking obesity
treatment (73.6% female) with (mage= 48.48, SD = 13.07) and (mBMI = 45.57, SD = 9.17).
Analyses included binary logistic regressions of three types of ACE (abuse, neglect, and family
dysfunction) as well as total number of ACE and the risk for food addiction, alcohol abuse, and
tobacco and drug use. Results indicated no significantly increased risk for alcohol abuse yet
found an increased risk for tobacco and drug use with exposure to ACE, with odds ratio (OR)
of Exp(B) = 1.25 (95% CI = 1.12-1.39), p = .000. Moreover, all types of ACE corresponded to
increased risk for tobacco and drug use: abuse (Exp(B) = 1.43 (CI = 1.10-1.87), p = .008),
neglect (Exp(B) = 1.79 (CI = 1.21-2.67), p = .004), and family dysfunction (Exp(B) = 1.57 (CI =
1.27-1.93), p = .000). There was also an increased risk for food addiction associated with ACE,
with OR of Exp(B) = 1.14 (CI = 1.05-1.24), p = .002, and for all types of ACE: abuse (Exp(B) =
1.32 (CI = 1.09-1.60), p = .005), neglect (Exp(B) = 1.54 (CI = 1.14-2.08), p = .005), and family
dysfunction (Exp(B) = 1.19 (CI = 1.02-1.40), p = .031). The present study demonstrated that a
history of ACE, in general, as well as specific types corresponds to increased risk for tobacco
and drug use and food addiction. Of note, those who endorsed neglect in childhood had the
greatest risk for both tobacco and drug use and food addiction. Individuals seeking treatment
for obesity should be carefully evaluated for history of ACE. For some patients, it might be
important to provide evidence-based treatment for the impact of childhood trauma prior to
or during with weight management intervention to enhance obesity treatment outcomes.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2137

CORRESPONDING AUTHOR: Allison Holgerson, Ph.D., Mayo Clinic, Rochester, MN, 55902;
[email protected]
S2138 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C205 6:00 PM-7:00 PM

INVESTIGATING THE RELATIONSHIP BETWEEN PERCEIVED STRESS AND COPING SKILLS ON

ADIPOSITY IN PRE-TEEN GIRLS

Biagina-Carla Farnesi, MSc1, Nicola J. Paine, PhD2, Kim L. Lavoie, PhD3, Simon L. Bacon, PhD4

1
Concordia University & Hpital du Sacr-Cour de Montreal, Brossard, PQ, Canada; 2Concordia
University & Hpital du Sacr-Cur de Montreal, Montreal, PQ, Canada; 3University du
Quebec a Montreal, Montreal, PQ, Canada; 4Concordia University, Montreal, PQ, Canada

Background: Elevated levels of acute and chronic stress have been hypothesized to be a risk
factor for the development of childhood obesity. However, it remains unclear whether
children who have better coping skills can mitigate these risks. The purpose of this research
was to examine the relationship between perceived stress and coping skills on adiposity in
girls.
Methods: Baseline data from the NIHs National Growth and Health Study (NGHS), an
American longitudinal study, was used for these cross-sectional analyses. Measures of acute
stress were taken from the Perceived Stress Scale, while coping was represented using the
total score from the Coping Strategies Inventory. Measures of adiposity were: body mass
index (BMI), BMI z-score, minimum and maximum below waist circumference, and % fat by
skinfolds. GLM-like linear regressions examined the main and interaction effects of coping and
stress on adiposity. Analyses were adjusted for age, race, maximum parental education, and
household income.
Results: A total of 2287 girls were included in the study (mean age = 11.0 0.57 yrs; white =
48%). Increased stress was found to be significantly related to higher BMI (=0.03 0.01,
F=4.59, p=.033), minimum waist circumference (=0.07 0.03, F=5.89, p=.015) and maximum
below waist circumference (=0.06 0.03, F=4.71, p=.030). When both stress and coping total
score were entered into the models, only stress was significant for BMI (=0.03 0.01, F=4.24,
p=.040), minimum waist circumference (=0.07 0.02, F=5.64, p=.018), and maximum below
waist circumference (=0.06 0.03, F=4.40, p=.036). There were no relationships between
coping and any measures of adiposity, with no evidence of a stress by coping interaction on
adiposity.
Conclusion: Though more longitudinal and intervention work needs to be done, our findings
indicate the potential importance of focusing on perceived stress when considering ways to
prevent or manage obesity in girls.

CORRESPONDING AUTHOR: Biagina-Carla Farnesi, MSc, Concordia University & Hpital du

Sacr-Cour de Montreal, Brossard, PQ, J4X1R7; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2139

C206 6:00 PM-7:00 PM

IS SOCIAL MEDIA HARMFUL FOR THE BODY IMAGE OF YOUNG ADULTS WITH
OVERWEIGHT/OBESITY? EXAMINING THE ROLE OF APPEARANCE COMPARISON

Dean Lim, M.A.1, Amanda F. Suplee, PhD2, Sylvia Herbozo, Ph.D.3

1
Department of Psychology, Loma Linda University, Rancho Cucamonga, CA; 2Children's
Hospital Colorado, Riverside, CA; 3Loma Linda University, Loma Linda, CA

Previous research has established that social media usage has increased significantly over the
past decade. Approximately 10% of young adults reported using social media in 2005 and 90%
of young adults reported using social media in 2015. Research has also shown that the
relationship between social media use and body dissatisfaction is mediated by appearance
comparison. However, this research has primarily focused on average weight young adult
women, despite the prevalence of overweight/obesity in young adult women and men. The
current study investigated appearance comparisons as a mediator between social media use
and body dissatisfaction in young adults with overweight/obesity. Participants were 385
young adults (63.50% female) aged 18 to 29 (M = 19.40, SD = 1.67) with a body mass index
(BMI) of 25 or greater (M = 30.03, SD = 5.89). They completed the Physical Appearance Scale,
Eating Disorder Inventory-Body Dissatisfaction, and a social media questionnaire. Results
indicate that approximately 23.9%, 18.8%, 22.6%, and 13.7% of young adults use Facebook,
Twitter, Instagram, and other forms of social media for about one hour per day, respectively.
Appearance comparisons was found to significantly mediate the relationship between the
amount of time spent on social media (i.e., Facebook, Instagram, Twitter, or other) and body
dissatisfaction (ab = .40, 95%CI [.26, .60], p < .05). As the amount of time spent on social
media sites increases, body dissatisfaction increases via the effect of appearance
comparisons. Findings suggest that individuals with overweight/obesity who spend greater
amounts of time on social media websites may experience greater body dissatisfaction if they
engage in more appearance comparisons. Further research examining the effects of social
media on body image is needed, especially as social media usage has increased among young
adults in recent years.

CORRESPONDING AUTHOR: Dean Lim, M.A., Department of Psychology, Loma Linda

University, Rancho Cucamonga, CA, 91737; [email protected]
S2140 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C207 6:00 PM-7:00 PM

PRIOR BEHAVIORAL WEIGHT LOSS TREATMENT EXPERIENCE IS ASSOCIATED WITH BETTER

OUTCOMES

Stephanie Kerrigan, MS, Evan M. Forman, PhD, Meghan L. Butryn, PhD

Drexel University, Philadelphia, PA

Introduction: Repeated attempts at behavioral weight loss (BWL) may be common given that,
at the end of treatment, many individuals remain overweight and weight regain is typical.
Prior BWL experience may be advantageous, as individuals may be more familiar with lifestyle
changes required of treatment or may have greater self-efficacy for change given earlier
success. Alternatively, repeated BWL efforts may be a marker of inability or low motivation to
make or sustain adequate lifestyle change for weight control. No prior research has evaluated
differences pre-treatment in psychological processes often associated with outcome (such as
self-efficacy or motivation for behavior change) or treatment outcome between these groups.

Methods: Participants (n=281) entering a BWL program completed measures assessing self-
efficacy for weight control, motivation, and prior weight loss experiences at baseline.
Participants wore an accelerometer for seven days and were weighed at baseline, mid-
treatment (6 months), and post-treatment (12 months).

Results: At baseline, individuals with prior BWL experience (60.8% of sample) had lower self-
efficacy for weight loss when confronted with negative emotions (p=.02) and physical
discomfort (p<.01). Additionally, individuals with prior BWL experience had greater weight
losses at 6 (9.79% vs. 7.47%; p < .01) and 12 (10.75% vs. 8.09%; p < .01) months. Although at 6
months, groups had similar levels of physical activity (102.47 minutes vs. 90.66
minutes; p=.42), only those with prior BWL sustained this level (102.19 minutes vs. 65.96
minutes; p=.01) at 12 months.

Discussion: Individuals with prior BWL experience, as opposed to those without, had lower
self-efficacy in several domains, but higher autonomous motivation. Additionally, they had
better weight losses and higher physical activity than those without BWL experience.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2141

Individuals with prior BWL may better anticipate and have higher motivation for behavioral
changes required of treatment, and thus make greater behavioral commitment throughout
treatment. Thus, experience with BWL may confer advantages for individuals entering new
BWL attempts. Repeated exposure to BWL techniques and messages may be necessary for
individuals to make sufficient lifestyle modifications to induce and sustain behavior change.

CORRESPONDING AUTHOR: Stephanie Kerrigan, MS, Drexel University, Philadelphia, PA,

19104; [email protected]
S2142 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C208 6:00 PM-7:00 PM

RACE, PLACE AND OBESITY IN NEW YORK CITY

Stephanie A. Grilo, N/A1, Shayna Cunningham, PhD2, Carlos M. Grilo, PhD3

1
Columbia University Mailman School of Public Health, New York, CT; 2Yale School of Public
Health, New Haven, CT; 3Yale University School of Medicine, New Haven, CT

Race, Place and Obesity In New York City

Objectives: This study examined the relationships among race and ethnicity, neighborhood
factors (place), and selected behavioral factors (nutrition and physical activity) associated
with energy regulation in relation to obesity.

Methods: This study utilized data from 8,665 participants living in 34 neighborhoods who
participated in the 2010 New York City Community Health Survey. Chi-square and ANOVAs
were used to test univariate categorical and continuous variables. Linear and logistic
regression models were used to examine the independent and joint effects of neighborhood,
neighborhood factors (distance to fresh produce, neighborhood safety), and individual factors
relevant to weight regulation (overall diet quality, meeting physical activity guidelines) on
body mass index (BMI) and obesity (BMI > 30 kg/m2).

Results: Chi-square analyses revealed that the 34 neighborhoods considered in the NYC
Community Health Survey differed significantly in rates of obesity (c2 (33) = 303.83, p <
.0001). Obesity rates ranged from a low of 8.1% (Upper East Side) to a high of 38.2% (East
Harlem).ANOVAs revealed significant differences in BMI across ethnic/racial groups (F (3,
8131) = 140.72, p < .0001) as did chi-square analyses (p < .001) indicating higher rates of
obesity in black and Hispanic groups relative to white and Asian groups. Nominal logistic
regressions using neighborhood to predict obesity while controlling for race and also using
race/ethnicity to predict obesity while controlling for neighborhood demonstrated that both
were significant in predicting obesity. Linear regression analyses revealed neighborhood
factors (distance to produce, neighborhood safety), individual factors (overall diet quality,
meeting physical activity guidelines), neighborhood, and demographic (race, income)
variables were significant predictors of BMI, together accounting for 9.8% of the variance.

Conclusions: There exist striking disparities regarding obesity in NYC. Rates of obesity in NYC
vary markedly by race and place, with significantly higher rates occurring in persons of color
and in certain neighborhoods. The two neighborhoods with the lowest and highest rates of
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2143

obesity respectively were directly adjacent to each other. Moreover, both race/ethnicity and
place were significant independent predictors of obesity.

CORRESPONDING AUTHOR: Stephanie A. Grilo, N/A, Columbia University Mailman School of

Public Health, New York, CT, 10025; [email protected]
S2144 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C209 6:00 PM-7:00 PM

SELF-MONITORING BEHAVIORAL CHARACTERISTICS AS PREDICTORS OF PHYSICAL ACTIVITY

AND WEIGHT LOSS MAINTENANCE

Kara L. Gavin, MPH, Jennifer A. Linde, PhD

University of Minnesota, Minneapolis, MN

Despite moderately positive weight outcomes from behavioral weight loss interventions,
behavior change maintenance resulting in sustained weight loss continues to be a challenge.
Although high rates of physical activity (PA) are associated with better weight loss
maintenance, strategies to improve PA continuation, and to identify individuals who show
signs of behavioral disengagement are still lacking. In order to improve intervention efficacy
and identify individuals in need of supplemental intervention support, interventionists may be
able to use information from behavior tracking logs. This analysis aims to define and examine
the effect of PA self-monitoring gaps during an active weight loss intervention on PA
maintenance and weight loss maintained at 24-month follow-up.

Data were collected from 339 adults participating in a weight loss intervention study in
Minneapolis, MN between July 2012 and August 2013. Participants were instructed to track
diet and PA behaviors for the 12-month active weight loss intervention using electronic or
paper logs. Tracking gaps were defined from these logs as 1 week of no activity logging.
Short gaps were defined as gaps lasting 1-3 weeks, medium gaps were defined as 4-7 weeks,
and long gaps were defined at 8-12 weeks of no behavior tracking. PA was self-reported at
baseline, 6, 12, 18, and 24 months via the Paffenbarger Physical Activity Questionnaire.
Weight was measured at the same time points by trained staff. The association between gaps
in tracking and PA level and weight outcomes at 24 months was examined. All models were
adjusted for gender, race, marital status, education level, age, baseline self-efficacy, and
treatment arm.

Increases in total number of gaps in PA tracking (RR: 0.95 p < 0.01), average length of gaps in
PA tracking (RR: 0.99 p = 0.01), and medium length gaps in PA self-monitoring (RR: 0.89 p
=0.02) were associated with lower PA level at 24 months. Late onset of PA tracking (RD: 6.31 p
< 0.01), increased total gaps in PA tracking (RD: 0.77 p < 0.01) and increased gaps in dietary
tracking (RD: 0.90 p < 0.01) were associated with higher weight at 24 months. Finally higher
numbers in short gaps in both PA self-monitoring (RD: 0.58 p 0.04) and dietary self-monitoring
(RD: 0.80 p=0.02) were associated with higher weight at 24 months. This analysis offers
information about self-monitoring behavior during active weight loss interventions that may
be utilized to improve intervention outcomes.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2145

CORRESPONDING AUTHOR: Kara L. Gavin, MPH, University of Minnesota, Minneapolis, MN,

55454; [email protected]
S2146 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C210 6:00 PM-7:00 PM

THE IMPACT OF HEDONIC CAPACITY ON FOOD INTAKE AND NEGATIVE MOOD

Laura Aylward, B.S.1, Kristin Schneider, Ph.D.2, Sherry Pagoto, PhD3

1
Rosalind Franklin University of Medicine & Science, Chicago, IL; 2Rosalind Franklin University
of Medicine & Science, North Chicago, IL; 3University of Massachusetts Medical School,
Worcester, MA

Background: Low hedonic capacity, a reduced ability to experience pleasure from things that
are typically rewarding, could increase the risk of emotional eating. Low hedonic capacity is
associated with increased substance use and similar reward mechanisms may influence food
intake. We hypothesized that low hedonic capacity would predict greater food intake
following negative mood induction relative to neutral mood condition and more rapid mood
recovery following food consumption during negative mood states.

Methods: In study 1, participants (n=61; 74% female, 39% obese) underwent neutral, anxiety,
and anger mood inductions and received 2400 kcal of palatable food. In study 2, participants
(n=54; 67% female, 41% obese) underwent anxiety and anger mood inductions; after each
they consumed a fixed amount of palatable food (1/5 of their total energy requirement). They
completed the Profile of Mood States questionnaire at post-mood induction, post food
consumption, and 15 and 30 minutes post food consumption. Hedonic capacity was assessed
using the Fawcett Clark Pleasure Capacity Scale.

Results: For study 1, the interaction between mood condition and hedonic capacity did not
predict food intake. Hedonic capacity predicted food intake, F(1, 177)=6.149, p=.014, such
that participants with lower hedonic capacity consumed more calories. For study 2, a
significant time by hedonic capacity interaction was observed for the anxiety condition [F(3,
192)=5.122, p=.002] and the anger condition [F(3, 191)=21.622, p < .001]. Simple slopes
analyses showed that the lower the participants hedonic capacity, the slower the recovery of
negative mood.

Discussion: Those with lower hedonic capacity consumed more calories than those with
higher hedonic capacity, regardless of mood, but negative mood recovered more slowly for
them after consuming palatable foods. Future research should examine if low hedonic
capacity is a risk factor for obesity.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2147

CORRESPONDING AUTHOR: Laura Aylward, B.S., Rosalind Franklin University of Medicine &
Science, Chicago, IL, 60640; [email protected]
S2148 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C211 6:00 PM-7:00 PM

THE NEIGHBORHOOD ENVIRONMENT IN THE CONTEXT OF A CHILDHOOD OBESITY

INTERVENTION.

Fabiana Brito, BSN, MSPH, PhD1, Jennie L. Hill, PhD2, Korine Kolivras, Ph.D.3, Jamie Zoellner,
PhD, RD3, Wen You, Ph.D.4, Paul A. Estabrooks, PhD2

1
University of Nebraska Medical Center, La Vista, NE; 2University of Nebraska Medical Center,
Omaha, NE; 3Virginia Tech, Blacksburg, VA; 4Virginia TEch, Blacksburg, VA

This study examined potential relationships between neighborhood environmental factors

and the reach and effectiveness of a family-based childhood obesity intervention. In this
mixed method study reach and individual level data were collected as part of the intervention
assessments and census data and maps were used to create a geodatabase for the region.
Families addresses were georeferenced and a 400 and 1600-meter buffer was used to
determine existing food and physical activity resources (i.e. NEMS and PARA audits). Family
participant demographic characteristics were compared with census data by block group.
Spatial distribution was evaluated by location, distance, proximity and multiple ring buffer
analyses. Multiple regression analysis was used to investigate environmental factors related
to change in child outcomes. Results. Enrolled families (n=99) were representative based on
census data with 36% from low income categories, with the exception of African American
families who were over represented in the sample (74% vs 48% census). The majority of
enrolled families (81%) were located 10 miles from the intervention location and a high
proportion of those that declined participation (77%) were from more than 10-miles away.
Physical activity resources were not within 400m of home for 70% of families and one third
had no resources within 1600m. PA resources that were available were rated as having poor
features (0.43.14) and limited amenities (1.110.27). Families traveled an average 3.7 miles
for their grocery shopping. Similar to PA resources, 74% of families did not have any type of
food outlet within 400m of home and over half did not have any grocery store within 1600m.
Further, grocery stores that were closer to participant homes were rated significant lower in
availability of healthy items (27.056.46) when compared to those where families reported
actually shopping (34.694.43). None of the investigated food and PA environmental factors
were associated with changes in child outcomes. Conclusions. Spatial analysis indicated that
distance from intervention site may be an important factor for reach and recruitment of
families in this rural region. Enrolled families did not shop at the closest grocery storebut
rather shopped at stores that were further away but had higher quality options. This is an
important finding considering food access is most often conceptualized as distance in the food
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2149

environment literature. Expanding conceptual definitions of food access beyond distances an

important next step.

CORRESPONDING AUTHOR: Fabiana Brito, BSN, MSPH, PhD, University of Nebraska Medical
Center, La Vista, NE, 68128; [email protected]
S2150 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C212 6:00 PM-7:00 PM

USING TELEHEALTH TO ENHANCE FAMILY-CENTERED CARE IN A CHILDHOOD OBESITY

INTERVENTION

Sarah Price, MPH, Christine Horan, MPH, Lauren Fiechtner, MD, MPH, Monica Gerber, MPH,
Caitlin Perry, MS, Gabriella Puente, B.S., Elsie M. Taveras, MD, MPH

MassGeneral Hospital for Children, Boston, MA

Background: Family-centered interventions to address childhood obesity have been found to

be effective. Incorporating digital technologies to provide parents with tailored behavior
change support outside the primary care visit could increase family engagement in
interventions.

Objective: To describe the use of telehealth to enhance engagement in a family-centered,

randomized trial to improve weight-related outcomes in 2-12 year olds with a BMI 85th
percentile.

Methods: Health coaches provided tailored behavior change support to families in the
intervention arm of Connect for Health, a two-arm RCT. Coaching tools considered family
preferences, needs, and residential location to improve their experience and study outcomes.
Telehealth modalities used included: interactive text messaging to assess behavioral goals and
provide counseling; video calls as an alternative option to in-person visits or phone calls with
health coaches; and a GIS-mapping tool to identify neighborhood resources that support
healthy living (i.e. healthy eating, physical activity).

Results: Between July 2014-March 2015, 360 children were randomized to the intervention
arm of Connect for Health. From that group, 336 (93%) parents were sent text messages, with
99% responding to at least 1 text message question and 61% responding to more than half of
the 46 text questions over a 1-year period. Among parents who completed the 1 year follow-
up survey, 72% (226/314) were very satisfied and 23% (72/314) were satisfied with the text
message content. Health coaches used a GIS-mapping tool to identify neighborhood resources
for families at their first visit and 314 (87%) completed this visit; 95 (30%) parents completed
at least one visit by video call, and at follow-up 96% (93/95) of them indicated they would
recommend video calls for health coach visits, with 66% citing time saved on travel, 54% the
face-to-face interaction, and 38% the money saved on babysitter/parking/travel as reasons to
use this modality.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2151

Conclusions: Telehealth enhanced engagement in a family-centered intervention to address

childhood obesity. Follow-up analyses will assess how levels of engagement impacted study
outcomes.

CORRESPONDING AUTHOR: Sarah Price, MPH, MassGeneral Hospital for Children, Boston, MA,
02114-1101; [email protected]
S2152 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C213 6:00 PM-7:00 PM

WEIGHT STIGMA AND WEIGHT CHANGE OVER THE FRESHMAN YEAR

Irene van Woerden, MS1, Marc A. Adams, PhD, MPH2, Alexandra Brewis, PhD3, Stephanie
Brennhofer, MS, RDN2, Meg Bruening, PhD, MPH, RD2

1
ASU, Tempe, AZ; 2Arizona State University, Phoenix, AZ; 3Arizona State University, Tempe, AZ

OBJECTIVE

The purpose of this study was to assess how weight stigma predicted changes in BMI over the
course of the academic year among college freshmen in the SPARC (Social impact of Physical
Activity and nutRition in College) study.

METHODS

Freshman students (n=527; 72% female; 49% white), residing on campus at a southwestern
university had height and weight measured by trained research assistants at the start, and up
to three additional times during, the 2015-2016 academic year. Participants completed
surveys measuring weight stigma by 1) hesitating to have a romantic relationship with a
person who is obese 2) being too embarrassed to participate in physical activity in public
places because of your weight 3) rating who is most responsible when someone is obese.
Response options for 1 and 2 were a four point likert scale and dichotomized to
agree/disagree. Response options for 3 were 'The individual' and 'Other factors'. Growth
curve models examined the relation between levels of weight stigmaand level of BMI change
over the year, after adjusting for gender, race/ethnicity, Pell grant status, initial weight status,
and clustering of students within dormitories.

RESULTS

On average, participants BMI at time one was 24.1 kg/m2 and increased by 0.7 kg/m2 over
the academic year. The majority of BMI change occurred in the first semester. A significant
non-linear effect of BMI change was observed (p2,95% CI: 0.58 to 2.13, p=0.001). Among
those who were embarrassed, BMI change over the academic year was 0.32 kg/m2 less than
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2153

their counterparts (95% CI: -0.65 to 0.01, p=0.061). Participants who stated the individual was
the most responsible for being obese had a BMI change of 0.37 kg/m2 less than their
counterparts over the year (95% CI: -0.62 to -0.11, p=0.005). Hesitating to have a romantic
relationship was not related to BMI change.

CONCLUSIONS

Participants embarrassed about their weight status and who believed they were responsible
for their own weight had significantly smaller changes in their BMI over the academic year.
This suggests that freshman's attitudes towards obesity is related to weight gain, however,
this finding needs to be interpreted with extreme caution. Future research should address
potential mechanisms behind these changes.

CORRESPONDING AUTHOR: Irene van Woerden, MS, ASU, Tempe, AZ, 85281;
[email protected]
S2154 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C214 6:00 PM-7:00 PM

AN EXAMINATION OF FACTORS THAT ACCOUNT FOR RACE-RELATED DIFFERENCES IN PAIN

INTENSITY AMONG INNER-CITY WOMEN

Teresa A. Lillis, PhD1, John Burns, PhD1, Frances Aranda, PhD, MPH1, Yanina A. Purim-Shem-
Tov, MD2, Helen Burgess, Ph.D.1, Stevan E. Hobfoll, PhD2

1
Rush University Medical Center, Chicago, IL; 2Rush University, Chicago, IL

Objective: Painful conditions disproportionately affect minority populations as compared to

Non-Hispanic/Whites. However, little is known about the mechanisms underlying these race-
related differences. The current study investigated 3 potential mediators of the relationship
between Race and pain intensity from a diverse sample of women presenting to an inner-city
Emergency Department (ED) for treatment of an acute pain complaint.

Methods: As part of an ongoing, longitudinal study examining the relationship between pain
and trauma among inner-city women presenting to the ED, baseline ratings of pain
catastrophizing [Pain Catastrophizing Scale; (PCS)], posttraumatic stress symptoms [PTSD
CheckList; (PCL)], subjective sleep quality [PROMIS-Short Form; (SQ)] and pain intensity were
collected from 177 women (M Age = 28.91 years; 65% Black, 20% Latina, 15% White). Three
series of bootstrapped mediation models with contrast coding of racial status were conducted
to examine whether PCS, PCL and SQ scores mediated links between Race and pain intensity

Results: Across all races, PSC, PCL, SQ and pain intensity scores were significantly correlated
(all ps < .05). Significant relative indirect effects were observed for PCL on pain intensity for
Black participants relative to Non-Black participants [Latina and Whites combined;
(a1b=0.233)] and Black participants relative to White participants (a2b=.326). No significant
relative indirect effects were observed for PCS or SQ for any race-related comparison on pain
intensity.

Conclusion: Results suggest that Black participants reported higher pain intensity as compared
to Non-Black and White participants, respecitvely. These effects may be partly accounted for
by Black participants having the tendency to report greater PTSD symptoms. Significant race-
related differences in pain intensity were not accounted for by race-related differences in Pain
Catastrophizing or Sleep Quality. Instead, results imply that greater pain intensity among
Black participants when reporting to an ED for acute pain may be due in large measure to
their disproportionately greater exposure to traumatic events.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2155

CORRESPONDING AUTHOR: Teresa A. Lillis, PhD, Rush University Medical Center, Chicago, IL,
60612; [email protected]
S2156 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C215 6:00 PM-7:00 PM

ARE PHYSICAL ACTIVITY RECOMMENDATIONS PROVIDED TOO LATE IN THOSE WITH ARTHRITIS
AND OBESITY?

Madelyn Ruggieri, B.A.1, Taylor L. Stallings, B.S.2, E. Amy Janke, PhD2

1
Center for Weight and Eating Disorders, University of Pennsylvania, Kennett Square, PA;
2
University of the Sciences, Philadelphia, PA

Background: Physical activity is a key component of evidence-based guidelines to manage

many forms of arthritis-related pain. However, implementation of guideline
recommendations is suboptimal. Patient characteristics that may influence guideline
implementation are not well examined.

Objective: To evaluate patient factors associated with receiving a health professionals

recommendation for physical activity to help manage arthritis symptoms in a nationally-
representative sample.

Method: Data from the CDC Behavioral Risk Factor Surveillance System 2011 were used.
Binary logistic regression models examined the impact of weight status, symptom severity,
sociodemographic factors, and health risk behaviors on likelihood of receiving a physician or
health professionals recommendation to engage in physical activity to aid symptom
management in overweight or obese individuals with arthritis (n=15,004).

Results: Among those with arthritis and BMI 25, 59.6% (n= 6,555) reported receiving a
recommendation for physical activity. Individuals who met criteria for obesity were 1.5 times
more likely to receive a recommendation compared to overweight peers (95% CI=1.4, 1.6).
Pain and symptom severity were also associated with a recommendation for physical activity.
Those reporting increased joint limitation (OR=1.2, 95%CI=1.1, 1.4), negative impact of pain
on social function (OR=1.5, 95%CI=1.3, 1.8), and mild (OR=1.8, 95% CI=1.5, 2.1) to maximum
pain (OR=1.6, 95%CI=1.3, 2.0) were all more likely to receive a recommendation. Several
sociodemographic variables increased likelihood of receiving a recommendation for physical
activity including female sex, Black race, and inability to work (ps .05)

Conclusions: Recommendations for physical activity in those with arthritis were associated
with higher BMIs and more severe symptom presentations, and only occur in about half of
those who may benefit from physical activity. Given barriers to physical activity for those with
more persistent pain and the potential for even small amounts of physical activity to positively
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2157

impact arthritis outcomes across time, it is possible recommendations are being given too late
in the disease course for maximal benefit.

CORRESPONDING AUTHOR: Madelyn Ruggieri, B.A., Center for Weight and Eating Disorders,
University of Pennsylvania, Kennett Square, PA, 19348; [email protected]
S2158 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C216 6:00 PM-7:00 PM

COMORBIDITY OF CHRONIC PROSTATITIS AND FUNCTIONAL DISORDERS, A TWIN STUDY.

Marianna Gasperi, Ph.D.1, John N. Krieger, M.D.2, Christopher W. Forsberg, MS3, Jack
Goldberg, PhD4, Dedra Buchwald, N/A5, Niloofar Afari, PhD6

1
UC San Diego, La Jolla, CA; 2University of Washington, Seattle, WA; 3VA Puget Sound HCS,
Portland, OR; 4Vietnam Era Twin Registry, Seattle VA ERIC, Carlsborg, WA; 5Washington State
University, Seattle, WA; 6University of CA, San Diego & VA San Diego Healthcare System, La
Jolla, CA

Chronic prostatitis/chronic pelvic pain syndrome (CP/CPPS) is characterized by pelvic and

perineal pains of unclear etiology. CP/CPPS frequently occurs with non-urological associated
chronic somatic syndromes (NUAS) of unknown etiology. As part of the MAPP Research
Network, we conducted a discordant twin study with men to investigate whether the
comorbidity between CP/CPPS and NUAS may reflect a shared familial liability. Data from
1,947 male monozygotic and dizygotic twin pairs participating in the Vietnam Era Twin
Registry were used. The twins were middle aged adults (mean age = 61.1 years; SD = 3.1) and
self-reported lifetime physician diagnoses of chronic prostatitis (CP of which approximately
90% is CP/CPPS), fibromyalgia, chronic fatigue syndrome, irritable bowel syndrome, and
temporomandibular disorder. The overall association of CP with NUAS was examined with a
series of mixed effects logistic regression models controlling for twin status and age, resulting
in ORs. To test for familial liability, prevalence rates in the non-CP member of discordant pairs
were compared to rates in the non-CP control twins and ORs were calculated. A total of 59
twin pairs were discordant for CP and 1,886 pairs were classified as non-CP control pairs.
Twins with CP were significantly more likely to also report all of the examined NUAS,
confirming the comorbidity of these conditions. With the exception of the association with
temporomandibular disorder (OR = 4.73; 95% CI = 1.93-11.58), none of the associations
remained significant in analyses examining familial liability, possibly as a result of limited
statistical power. There is substantial comorbidity between CP and a number of NUAS in men,
but the comorbidity may reflect limited shared familial liability. Larger studies are needed to
elucidate familial and environmental factors underlying the association of CP/CPPS with
NUAS, especially with temporomandibular disorder.

CORRESPONDING AUTHOR: Marianna Gasperi, Ph.D., UC San Diego, La Jolla, CA, 92093;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2159

C217 6:00 PM-7:00 PM

COPING PROFILES AND HEALTH OUTCOMES AMONG INDIVIDUALS WITH SYSTEMIC

SCLEROSIS: A LATENT PROFILE ANALYSIS APPROACH

Shelley Condon, MA1, Scott Roesch, Ph.D2, Michael H. Weisman, MD3, Suzanne Kafaja, MD4,
Philip J. Clements, MD4, Daniel E. Furst, MD4, Vanessa L. Malcarne, Ph.D.2

1
San Diego State University, Tuscaloosa, AL; 2San Diego State University, San Diego, CA;
3
Cedars-Sinai Medical Center, Los Angeles, CA; 4UCLA School of Medicine, Los Angeles, CA

Systemic sclerosis (SSc) is a severe, chronic, and progressive rheumatic disease with an
extensive impact on quality of life. Although coping has been identified as a mechanism by
which individuals can exercise control over serious health challenges and adapt more
successfully, to date most studies of coping with illness have examined the relationship of
individual coping strategies to adjustment, rather than examining profiles of coping. Latent
profile analysis (LPA) allows the identification of coping profiles that incorporate multiple
strategies, and that can be used to distinguish subgroups of patients. The primary aims of this
study were to (1) identify patient groups with distinct coping profiles using LPA, and (2)
compare identified coping profile groups on physical and mental health outcomes.

Participants (N = 94) filled out psychological measures including the revised Ways of Coping
Checklist (WCCL-R), Health Assessment Questionnaire (HAQ), Psychological Adjustment to
Illness Scale (PAIS), and Modified Rodnan Skin Score (mRSS). Two LPAs were done to create
profiles derived from participants raw and relative scores on the eight coping subscales from
the WCCL-R including problem-focused (PF), wishful thinking (WT), seeking social support (SS),
avoidance (AV), self-blame (SB), blaming others (BO), counting ones blessings (CYB) and
religiosity (RG).

A three-profile solution was supported using participants raw scores. Low Copers (n = 7)
exhibited below-average coping across all subscales; High Copers (n = 23) exhibited above-
average coping across all subscales, and Mixed Copers (n = 64) exhibited relatively infrequent
use of SB, WT, AV, and BO strategies versus frequent use of the PF, WT, CYB, and RG
strategies. Two ANCOVAs examined group differences in physical health (HAQ) controlling for
age and disease severity, and psychological health (PAIS) controlling for age. Significant
differences were found on the PAIS, with High Copers exhibiting greater levels of
psychological distress when compared with the other two groups.

A two-profile solution was supported using participants relative scores. Adaptive Copers (n =
62) emphasized PF, SS, WT, and RG while Maladaptive Copers (n = 32) emphasized SB AV, CYB,
S2160 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

and BO. The ANCOVA comparing the groups on psychological health revealed that Adaptive
Copers were significantly less distressed than Maladaptive Copers. The findings suggest that
patient groups with profiles for coping with illness can be identified. These profiles can further
impact patients mental and physical adjustment to the disease, thus underscoring the
importance of examining coping in a multidimensional way.

CORRESPONDING AUTHOR: Shelley Condon, MA, San Diego State University, Tuscaloosa, AL,
35401; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2161

C218 6:00 PM-7:00 PM

EVALUATION OF PAIN DISABILITY AMONG VETERANS: A TALE OF TWO MEASURES

Jane A. Skoyen, PhD1, Anya V. Kogan, PhD2, Sarah A. Palyo, PhD3, Timothy P. Carmody, PhD4

1
None, Broadmoor, CA; 2VA Palo Alto Health Care System, Palo Alto, CA; 3San Francisco VA
Health Care System, San Francisco, CA; 4San Francisco VA Medical Center, San Francisco, CA

Purpose

A large proportion of Veterans experience chronic pain; however, studies assessing

performance of different chronic pain measures in this population are lacking. The present
study evaluates psychometric properties and responsiveness of two frequently used measures
of functioning with chronic pain, the Oswestry Disability Inventory (ODI) and Pain Disability
Questionnaire (PDQ), including both the Functional Status Component (FSC) and Psychosocial
Component (PC).

Methods

In Study I, 1988 Veterans with chronic pain who were enrolled in a VA facility and seeking pain
management services (10.5% female, mean age = 57.5, SD = 14.0) reported on locations of
pain sites and typical pain levels over the past week, and also completed the ODI and PDQ.
Correlations were examined between the ODI, PDQ, and the Veterans typical pain levels.

In Study II, a subsample of these Veterans (n = 71; 22.5% female, mean age = 55.3, SD = 9.5)
reported on the above measures before and after completing a 12-week Intensive Pain
Rehabilitation Program, which included psychological, physical, and medical treatment
modalities. A series of t-tests were completed to evaluate these measures responsiveness to
change among Veterans in treatment for chronic pain.

Results

Study I. The ODI demonstrated good internal consistency ( = .85). PDQ and its component
scales demonstrated good to excellent internal consistency ( = .92 for total PDQ, = .91 for
FSC, = .82 for PC). The ODI, PDQ-total, PDQ-FSC, PDQ-PC, and typical pain were all
significantly, positively correlated (e.g., r = .74 for ODI and PDQ, r = .41 for ODI and typical
pain, and r = .42 for PDQ and typical pain, all ps < .001).

Study II. Paired-samples t-tests comparing pre- and post-treatment scores revealed significant
change in the ODI (t = 3.1, p = .003), PDQ-total (t = 5.4, p < .001) and both its subscales (t =
S2162 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

5.0, p < .001 for FSC, t = 4.4, p < .001 for PC), as well as typical pain (t = 3.7, p < .001) for
Veterans who completed pain treatment.

Conclusions

Results suggest that both ODI and PDQ measures are useful in measuring pain-related
disability among US Veterans with chronic pain. Both measures demonstrated good to
excellent internal consistency, and were highly positively correlated with each other and
moderately positively correlated with typical pain. Both scales detected change over a 12-
week pain rehabilitation program. Future research is needed to better understand the
differences in how these measures perform in samples with specific pain conditions.

CORRESPONDING AUTHOR: Jane A. Skoyen, PhD, None, Broadmoor, CA, 94015;

[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2163

C219 6:00 PM-7:00 PM

EXERCISE IS ASSOCIATED WITH RESILIENCE TO PAIN IN PEOPLE WITH SPINAL CORD INJURY

Kristen E. Riley, Ph.D. 1, Michelle M. Rodriguez Diaz, B.A., Doctoral Candidate2, Kim Anderson,
PhF3, Eva Widerstrom-Noga, DDS, PhD4

1
Memorial Sloan Kettering Cancer Center, New York, NY; 2University of North Carolina at
Charlotte, Miami, FL; 3University of Miami, Davie, FL; 4Miller School of Medicine, University of
Miami, Miami, FL

Severe chronic pain is common among people with spinal cord injury (SCI), and interferes with
medical treatment, quality of life, and adjustment after injury. Exercise has been broadly
shown to help manage chronic pain. However, only a few studies have been conducted about
exercise and coping with pain in the context of SCI, and these have focused on the efficacy of
pain management programs. No studies using a mixed methods approach have examined
personal perspectives regarding the association between exercise and resilience to chronic
pain in people with SCI. The primary purpose of the present study was to examine the
association between utilizing exercise to relieve pain and resilience to chronic pain.

Method

This mixed method study focused on positive (facilitators) and negative (barriers) contributors
to individual experiences of living with chronic pain after SCI. The study consisted of 2 parts:
(1) individual open-ended qualitative interviews (n=35) to identify common
perspectives/themes, and convert them into statements; and (2) online survey (n=491) to
quantify the agreement with each statement using Numerical Rating Scales (NRS). The present
study utilized part of the data obtained from the survey, specifically regarding pain
characteristics and associated psychosocial factors, resilience to pain, and utilizing exercise to
relieve pain.

Results and Discussion

Participants (56.0%) reported severe pain (NRS > 6) in the past week. We performed two sets
of stepwise regression analyses controlling for age, time since injury, gender, and average
pain level. Utilizing exercise was shown to be positively and significantly related to resilience
to pain ( = .182). Gender ( = -.122), perceived control over ones life ( = .191), and
preferring treatment other than medication ( = .155) predicted utilizing exercise. Men
reported engaging in exercise to relieve pain more than women ( = .201).
S2164 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Overall, individuals with SCI who reported utilizing exercise to relieve pain also had higher
agreement ratings on resilience to chronic pain. Utilizing exercise was predicted by gender,
having greater life control, and preferring alternative treatments to pain medication. Clinical
implications include encouraging women with SCI to utilize exercise to relieve their pain,
offering additional non-pharmacological treatments, including psychotherapy focused on
increasing perceived life control, and increasing physical activity. Study limitations include
limited demographic and geographic variability, which may not be generalizable.

CORRESPONDING AUTHOR: Kristen E. Riley, Ph.D. , Memorial Sloan Kettering Cancer Center,
New York, NY, 10022; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2165

C220 6:00 PM-7:00 PM

PAIN CATASTROPHIZING IN VETERANS: A PILOT STUDY INVESTIGATING PREVALENCE, CLINICAL

CORRELATES, AND TREATMENT OUTCOMES

Nidhi Anamkath, B.A., Sarah A. Palyo, PhD, Irina Strigo, Ph.D

San Francisco VA Health Care System, San Francisco, CA

Pain catastrophizing (PC) is a psychological construct comprised of three dimensions

(rumination, magnification, helplessness) that is known to heighten chronic pain intensity and
severity; however, research on this topic has been mostly limited to Civilian populations. To
date, Veteran-specific investigations of PC have found that higher levels of catastrophizing are
linked to increases in negative clinical outcomes (depression, pain severity) and that
decreases in catastrophizing via Cognitive-Behavioral Therapy (CBT) explains decreases in
these very outcomes. However, ongoing evaluation of this construct in the Veteran
population is critical as results may better inform pain treatments. The current pilot study
aims to fill a gap in the literature by examining 1) the prevalence and 2) the relationship of PC
to clinical outcomes in Veterans with chronic pain in an intensive pain rehabilitation program
(IPRP) (n=35; 71.4% male; age (36-68, M=56.2); pain duration in years (0.5-46.0, M=18.7)).
Unlike prior Veteran studies, the IPRP is an integrated program which includes CBT but also
contains Acceptance and Commitment Therapy and physical therapy. A retrospective study
design evaluated pre- and post- treatment assessments of PC (using the 13-item Pain
Catastrophizing Scale) in relation to pain intensity, depressive symptoms, pain disability, and
kinesiophobia. Results indicated the prevalence of PC was similar to previous Veteran
samples, and on average catastrophizing was not clinically significant (M =23.8, SD = 12.2).
Pre-treatment Spearmans correlations indicated greater overall PC was related to greater
depressive symptoms (r = 0.57, p < .01), fear of movement (r = 0.56, p < .01), and pain
disability (r = 0.46, p < .01). In addition, Wilcoxon signed-rank tests comparing pre- and post-
treatment measures indicated significant decreases in depression (z = -3.61, p < .001), pain
disability (z = -3.69, p < .001), and participants subjective acceptable level of pain (z = -3.80,
p < .001), but not significant changes in current pain intensity (z = -1.59, p > .05). This suggests
the administered treatment affected cognitions and emotions associated with chronic pain
and subjects perceived disability level, but had minimal effects on self-reported pain levels.
These findings warrant further clarification on the role of PC in the experience and treatment
of chronic pain in Veterans, as well as an investigation into mechanisms of change during
treatment.

CORRESPONDING AUTHOR: Nidhi Anamkath, B.A., San Francisco VA Health Care System, San
Francisco, CA, 94121; [email protected]
S2166 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C221 6:00 PM-7:00 PM

PAIN IN CHILDBIRTH

Lana Kinney, BSc, Anne Howarth, PhD, Nicola Swain, PhD

University of Otago, Dunedin, Otago, New Zealand

Pain in childbirth receives little research attention. Although it may be seen as a biological
imperative, pain experienced in childbirth is a predictable acute pain episode that has the
potential to be well managed. Just like initial reluctance to use anesthesia in surgery, there
seems some opinion that childbirth pain should be tolerated rather than treated. Previous
research suggests anxiety is the strongest predictor of childbirth pain. A sample of 137 first
time mothers were recruited nationwide in New Zealand. Data from 3 questionnaires is
included in this analysis: 24 weeks gestation, pre-birth and post-birth. Caesarian and
complicated births were excluded from analysis. A general linear model analysis of covariance
was constructed to predict labour pain from both dichotomous and continuous variables.
Socio-economic status, the use of hot/warm water, walking, and movement as pain
management techniques, antenatal yoga for pregnancy, Plunket antenatal classes, how
painful the woman expected her labour to be at 32 weeks, and a womans birth expectations
at 32 weeks were found to significantly predict experienced labour pain. This data has
implications for both birth preparation and managment and may be helpful for various
practitioners involved in childbirth. Childbirth is both an emotional and painful experience and
much is to be learned from examining the intersection of the various contributions to the pain
experience.

CORRESPONDING AUTHOR: Lana Kinney, BSc, University of Otago, Dunedin, Otago, 9010;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2167

C222 6:00 PM-7:00 PM

RANDOMIZED CONTROLLED TRIAL OF A NURSING-DELIVERED CBT VERSUS SUPPORTIVE CARE

TELEPHONE INTERVENTION FOR CHRONIC BACK PAIN

Thomas Rutledge, PhD1, Rachael Holloway, PsyD2

1
VA San Diego, San Diego, CA; 2VA San Diego Healthcare System, San Diego, CA

Objective: To evaluate the efficacy of a nursing-delivered, telephone-based intervention of

cognitive behavioral therapy versus supportive care for chronic low back pain.

Methods: Participants (N=57) were patients with chronic back pain (experiencing pain on a
daily basis 6 months at a minimum intensity 4/10 on a 10-point scale) randomized to
either an 8-week, 8-session, Cognitive-Behavioral Self-Management Skills Training (CBSST) or
to a Supportive Care condition (SC) matched for contact frequency, format, and time, with
each treatment delivered by a trained nurse. Primary outcomes included changes in Roland
Morris Disability Questionnaire (RMDQ) and changes in pain and function as self-rated on the
Patient Global Impressions Scale (CGI).

Results: Participants in the CBSST condition (n=29) showed significant improvements on the
RMDQ (mean=11.3[5.6] at baseline versus mean=8.2[6.1] at post-treatment, p < .05) and on
the CGI (83.6% reporting improvement). SC participants (n=28) also showed improvement on
the RMDQ (mean=11.1[5.5] at baseline versus mean=9.0[5.3] at post-treatment, p < .05), with
48.1% reporting improvement on the CGI. Between groups comparisons showed no
differences in changes on the RMDQ but significantly better outcomes for CBSST participants
on the CGI (p < .05).

Conclusions: Results from this clinical trial suggest that telephone-based, nurse-delivered
CBSST and SC treatments for patients with chronic back pain each offer potential benefits and
may warrant further research for applications to hospital settings.

CORRESPONDING AUTHOR: Thomas Rutledge, PhD, VA San Diego, San Diego, CA, 92161;
[email protected]
S2168 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C223 6:00 PM-7:00 PM

SELF-REPORTED RATINGS OF PAIN: DIFFERENCES AT HOME AND AT THE HOSPITAL

Julie Kangas, PhD1, Peter Link, MA2, Colin Depp, PhD3

1
UC San Diego & San Diego VA Healthcare System, Oak Leaf, TX; 2Veterans Affairs San Diego
Healthcare System, San Diego, CA; 3University of California San Diego, La Jolla, CA

Background. Little work has been done to monitor how social-contextual environment
influences subjective pain rating. In other vital signs, such as blood pressure, hospital-based
recordings are often exacerbated compared to home-based assessments. We conducted an
analysis comparing pain scores collected via home telehealth device to pain scores that were
collected at the hospital by a medical provider.

Methods. Data was obtained for a two-year period from a home telehealth device designed to
help Veterans with chronic illness monitor and manage their health on a daily basis. Veterans
had an opportunity to rate their pain daily on the standard 0-to-10 pain scale. We also
retrieved pain scores from each Veterans electronic medical record that were collected at
outpatient hospital appointments during the same timespan. Veterans (N = 33) each
completed an average of 94 pain ratings (range = 2-546) at home through the home
telehealth system, without a provider present (n = 3102 total ratings). They rated their pain to
a medical provider at an average of 7.88 outpatient medical appointments (range = 2-32)
during the same timespan (n = 260 total ratings). Patient-averaged scores for pain were
calculated and paired T-tests were conducted, and a discrepancy scores (home minus hospital
pain scores) were associated with selected clinical characteristics.

Results. We found that pain ratings at home were significantly higher than those taken at the
hospital, t(32) = 4.16, p < .001, with average at-home pain being rated as 5.42 (SD = 2.15) and
average at-hospital pain being 3.79 (SD = 2.33), d = .73. We also found that discrepancy scores
were associated with hospitalizations, such that higher pain at home than at the hospital was
related to a lower likelihood of medical hospitalization, r = -.38, p=.03; note, 13 of the 33
Veterans were hospitalized. Discrepancy scores were also associated with suicide behaviors,
such that higher at-home pain than at-hospital pain was related to a higher likelihood of
suicide behaviors, r = .40, p=.02; note, 4 of the 33 Veterans had suicide behaviors during the
two-year period.

Implications. In contrast to what might be expected from other home- versus hospital- based
vital sign comparisons, results indicate that pain ratings are reported as being lower when
taken at the hospital than at home. The findings suggest that there may be important
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2169

environmental cues and psychological processes that impact either actual perceptions of pain
or willingness to disclose true pain that may differ according to environmental context.
Systems for home-based telemonitoring of pain may need to consider these potential
discrepancies in designing clinical response alerts. Future research directions regarding the
associations of pain score reporting with hospitalizations and suicide behaviors will also be
discussed.

CORRESPONDING AUTHOR: Julie Kangas, PhD, UC San Diego & San Diego VA Healthcare
System, Oak Leaf, TX, 75154; [email protected]
S2170 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C224 6:00 PM-7:00 PM

THE PAIN RELATED SELF-EFFICACY SCALE: A SHORT FORM

Dagmar Amtmann, PhD, Fraser D. Bocell, PhD, Kendra Liljenquist, PhD, Mark Jensen, PhD,
Dennis Turk, PhD

University of Washington, Seattle, WA

Introduction: Pain-Related Self-Efficacy (PRSE) is a persons condence that she or he can

engage in activities despite chronic pain (CP). Self-reported PRSE is an important therapeutic
target because it is more predictive of disability than measures of biomedical domains.
Interventions that target and increase PRSE have been found to improve pain, mental health,
and quality of life of people with CP. Brief, flexible, and psychometrically sound measures of
PRSE facilitate development of effective strategies for managing CP. This study developed a
short form (SF) for measuring PRSE.

Methods: The full PRSE item bank contains 35 items calibrated to Item Response Theory (IRT).
Development followed rigorous psychometric methodology. CP experts and people with CP
provided feedback. All items were administered to a sample of people with CP (n=583)
including those with low back pain, multiple sclerosis, osteoarthritis, spinal cord injury,
amputations, and diabetic neuropathy. Items for the SF were selected from the full item bank
using the IRT parameters and representation of the subdomains. Validity of the IRT-based t-
score based on the SF was examined by correlations with related constructs. Test-retest
reliability was tested 40 to 80 hours after the first administration.

Results: A correlation between the 5-item SF and the full item bank score (r=.95) and the test-
retest reliability (ICC=.90) were both excellent. The correlations of the SF score with related
constructs were all in the expected direction and magnitude. The SF five items ask about a
persons confidence in, despite having pain, maintaining personal hygiene, ability to do things
she/he wants and enjoys doing, ability to cope with pain, and socialize with friends. The SF can
be administered on the computer or by paper and pencil questionnaires. A lookup table that
converts a simple summary score to an IRT-based t-score was developed.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2171

Conclusions: A 5-item SF measures well across the PRSE continuum and covers important
subdomains of PRSE. The results support the validity and reliability of the 5 item PRSE SF. The
SF score is on the same metric and directly comparable to the t-score based on the whole
item bank. Because the SF is brief it lends itself well for inclusion in clinical practice and
research. The SF and the full item bank are publicly available to researchers and clinicians.

CORRESPONDING AUTHOR: Dagmar Amtmann, PhD, University of Washington, Seattle, WA,

98195; [email protected]
S2172 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C225 6:00 PM-7:00 PM

VIEWING A DOCUMENTARY ON BEHAVIORAL MEDICINE APPROACHES TO PAIN REDUCTION

INCREASES EMPATHY AMONG FUTURE HEALTH PROFESSIONALS

Luciana Lagana, Ph.D.1, Larisa Gavrilova, David Grewe, Alberto Zacarias, Melissa Martinez,
N/A2, Abigail Pajulas, none3

1
CSUN, Chatsworth, CA; 2California State University Northridge, Canyon Country, CA;
3
California State University Northridge, Glendale, CA

In this experimental study, we tested whether viewing an original documentary created by the
first author could reduce ageism and increased empathy among future health professionals
such as psychology students. The films duration is one and a half hours; it features several
interviews with pain experts who are members of the American Pain Society and who discuss
in simple terms: the findings of their behavioral medicine research in areas such as breast
cancer-related pain; health disparity topics, including why pain medications may not work
well for patients from ethnic minority backgrounds; as well as pain prevention and treatment
strategies, emphasizing the difficulties concerning achieving effective pain management for
older patients. Moreover, the movie contains accounts of living in chronic pain by several
people age 50 and older. To our knowledge, no published literature is available on the effects
of viewing a documentary covering prevention and treatment issues related to pain in older
age on young peoples empathy and ageism scores. We hypothesized that only those research
participants who viewed our documentary would have reported lower ageism and higher
empathy scores, not the control group participants, who watched a neutral/nature and park
documentary of the same duration. Eighty psychology students (age 18 to 29) were
randomized via tossing a coin to either the experimental or the control condition, with 40
people per group. Measures included a demographic list, the 17-item Jefferson Scale
EmpathyHealth Profession Students version, and the 20-item Prescriptive Intergenerational-
Tension Ageism Scale. An independent samples t-test approach was used to compare pre-test
and post-test scores between the control and the experimental conditions. For empathy
before viewing a film, scores of the control (M=87.44, SD=12.22) and the experimental
(M=89.5, SD=13.84) groups were not statistically different, t(75) = -.069, p>.05. For post-
viewing empathy, there was a significant difference in scores between the control (M=88.46,
SD=13.43) and the experimental (M=95.66, SD=9.72) groups, t(75) = -2.69, p < .05, suggesting
that viewing the experimental documentary influenced empathy scores. Ageism scores,
contrary to our predictions, were not influenced by viewing our film, as we achieved non-
significant findings on this variable. The studys limitations and clinical implications are
discussed.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2173

CORRESPONDING AUTHOR: Luciana Lagana, Ph.D., CSUN, Chatsworth, CA, 91313;

[email protected]
S2174 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C226 6:00 PM-7:00 PM

IMPACT OF PROLONGED CLASSROOM SITTING TIME ON STUDENT PERCEPTIONS OF PHYSICAL

DISCOMFORT AND ALERTNESS. HOW MUCH IS TOO MUCH?

Katie Hosteng, M.S., Roberto Benzo, M.S., Alison Phillips, Ph.D., Lucas Carr, Ph.D.

University of Iowa, Iowa City, IA

Background: Prolonged sitting has been shown to increase fatigue and risk of low back pain
and injury in sedentary office workers. Conversely, regularly breaking up workplace sitting
with standing bouts every 30 minutes has been shown to improve fatigue and musculoskeletal
discomfort in office workers. It is not known whether prolonged classroom sitting similarly
impacts fatigue and discomfort in college students and/or how much sitting is too much.

Objective: To explore the impact of prolonged bouts of classroom sitting on students

perceived physical discomfort and alertness. To determine the minimum duration of
classroom sitting time needed to induce meaningful impairments in students perceived
physical discomfort and alertness. We hypothesize a positive association between sitting time
and perceived discomfort, and a negative association between sitting time and alertness.

Methods: We recruited 58 undergraduate college students enrolled in a single 2.5-hour

lecture. Students were asked to remain seated the entire lecture. Students completed the
Stanford Sleepiness Scale (SSS; 1=feeling active, vital, alert, wide awake; 8=asleep) to measure
altertness and the Shackel General Comfort Scale (SGCS; 0 = I feel completely relaxed; 10 = I
feel unbearable pain) to measure physical discomfort every 15 minutes throughout the
lecture. Pearson correlations were used to determine relations between sitting time,
alertness, and comfort. Regression analysis was used to determine the time it took for
students to report a 4 (I feel uncomfortable) on the SGCS and a 3 (Awake, but relaxed;
responsive but not fully alert) on the SSS as these values represent critical thresholds for
discomfort and alertness.

Results: A positive association was observed between sitting time and perceived discomfort (r
= 0.31, P < 0.01) while a negative association was observed between sitting time and alertness
(r = 0.47, p < 0.01). Regression analyses indicate students reported a 4.0 on the SGCS after
88.4 minutes and a 3.0 on the SSS after 26.8 minutes.

Conclusion: These findings suggest students alertness and discomfort deteriorated over the
duration of the seated class period. These findings also suggest students may also benefit
from brief activity breaks every 30 minutes during class to minimize impairments in alertness.
Future sedentary classroom interventions are needed to confirm these findings.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2175

CORRESPONDING AUTHOR: Katie Hosteng, M.S., University of Iowa, Iowa City, IA, 52242;
[email protected]
S2176 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C227 6:00 PM-7:00 PM

ACCELEROMETER-ASSESSED PHYSICAL ACTIVITY AND SEDENTARY TIME IN LUNG CANCER

SURVIVORS: ASSOCIATIONS WITH QUALITY OF LIFE

Adrijana D'Silva, BSc1, Gwyn Bebb, MD, PhD2, Terry Boyle, PhD3, Steve Johnson, PhD4, Jeff
Vallance, PhD4

1
Athabasca University, Calgary, AB, Canada; 2University of Calgary, Calgary, AB, Canada;
3
Curtin University, Perth, Western Australia, Australia; 4Athabasca University, Athabasca, AB,
Canada

Background: Few studies have examined physical activity and sedentary time among lung
cancer survivors. These studies self-reported these behaviors, which may lead to biased
results and incorrect conclusions. This study set out to determine associations of
accelerometer-assessed moderate to vigorous intensity physical activity (MVPA) and
sedentary time with health-related quality of life (HRQoL) and fatigue among lung cancer
survivors. Methods: Lung cancer survivors in Southern Alberta (N=527) were invited to
complete a mailed survey that assessed HRQoL (FACT-L: Functional Assessment of Cancer
Therapy-Lung) and fatigue (FS: Fatigue Scale). Consenting participants wore an Actigraph
GT3X+ accelerometer on their hip for seven days. Average daily MVPA and sedentary time
were derived from the accelerometer data, processed using 60-second epochs, and examined
as quartiles. To examine associations between accelerometer outcomes and HRQoL, we used
linear regression adjusting for wear time, demographic, and clinical variables. Results: A total
of 127 lung cancer survivors participated (Mean age=71 years; Mean time since diagnosis=75
months) for a 24% response rate. MVPA was significantly associated with FACT-L (B=3.35, 95%
CI, 0.37 to 6.32, p=.028) and FS (B=1.83, 95% CI, 0.17 to 3.5, p=.031). Total sedentary time was
associated with FACT-L (B=-5.03, 95% CI, -8.45 to -1.61, p=.004) and FS (B=-2.63, 95% CI, -0.46
to -0.72, p=.007). Sedentary time accrued in 30-minute bouts was associated with FACT-L (B=-
4.03, 95% CI, -6.97 to -1.09, p=.008). Conclusion: Objectively measured MVPA was positively
associated with HRQoL and fatigue, while sedentary time was negatively associated with
HRQoL and fatigue in lung cancer survivors. Differences in HRQoL and fatigue met thresholds
for determining clinically meaningful differences. This is the first study to examine
associations of accelerometer-assessed physical activity and sedentary time with HRQoL and
fatigue in lung cancer survivors.

CORRESPONDING AUTHOR: Adrijana D'Silva, BSc, Athabasca University, Calgary, AB, T2N 4N2;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2177

C228 6:00 PM-7:00 PM

BODY IMAGE AND SELF-EFFICACY AS PREDICTORS OF ACHIEVING PHYSICAL ACTIVITY

GUIDELINES IN UNIVERSITY STUDENTS

Jamie Faro, MS1, Jessica Whiteley, PhD2, Laura L. Hayman, PhD, RN, FAAN, FAHA3, Samuel J.
Simmens, PhD4, Melissa Napolitano, PhD5

1
University of Massachusetts Boston, Newton, MA; 2UMass Boston, Boston, MA; 3University
of Massachusetts Boston, Boston, MA; 4George Washington University, Washington, DC; 5The
George Washington University, Washington DC, DC

Declines in moderate-to-vigorous physical activity (MVPA) have been observed during

developmental transitions particularly in young adults entering college. Negative body image,
which also tends to become more negative in young adults, has been associated with lower
levels of exercise self-efficacy and physical activity (PA). PURPOSE: To examine the role of
body image and other predictor variables on meeting MVPA guidelines among university
students at 2 campuses. METHODS: N=128 college students (28 males and 84 females,
meansd: age, 21.63.1 yrs; BMI 31.63.7 kg.m-2) completed demographic questions (age,
gender, race, BMI), the International Physical Activity Questionnaire (IPAQ), the Body Image
Quality of Life Inventory Scale (BQLI) and physical activity self-efficacy (PA-SE). MVPA was
calculated for each participant based on the days and minutes of self-reported activity, with
meeting guidelines defined as 150 minutes/week of combined MVPA. RESULTS: When
entered separately into logistic regression models, higher levels of PA-SE (pppCONCLUSION:
Among young adults with excess weight, physical activity self-efficacy was an important factor
in predicting MVPA. While body image might play a role in the initiation of PA, further work
should be conducted to determine the importance of the confidence young adults have to
continue PA despite barriers (including negative body image).

Supported by: NIH Grant R01DK100916

CORRESPONDING AUTHOR: Jamie Faro, MS, University of Massachusetts Boston, Newton,

MA, 02458; [email protected]
S2178 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C229 6:00 PM-7:00 PM

DOES TIME MATTER? ASSOCIATIONS BETWEEN REPORTED PARENT-CHILD TIME TOGETHER

AND OBJECTIVELY MEASURED CHILD PHYSICAL ACTIVITY

Helena H. Laroche, MD1, Brad R. Julius, MD2, Amy O'Shea, PhD 1, Shelby L. Francis, M.A.1,
Kathleen F. Janz, EdD1

1
University of Iowa, Iowa City, IA; 2University of Iowa Hospitals & Clinics, North Liberty, IA

Background: Active parents tend to have active children but how much of this is due to
encouragement and facilitation of activity compared with actual time spent being active with
their children?US guidelines note that there is little research to fully substantiate how and
how much parents influence child physical activity (PA).

Methods: We examined the effects of 1) daily waking parent-child time together (parent
reported average over past month) and 2) parent-child time doing PA (parent-reported over
the past week) on both moderate-vigorous intensity physical activity (MVPA) and sedentary
time (min/d) of children in the following week. All variables were separated by weekend days
(WE) and weekdays (WD). A low-income parent with obesity and her child aged 6-12 years
wore accelerometers simultaneously for a week (minimum of 3 WD and 1 WE day). Linear
regression was used to determine the effect of self-reported parent-child time together on
objectively measured child PA adjusting for child age, child sex, parent BMI, child BMI %, and
parent education.

Results: The initial 47 parent-child dyads with complete data from an ongoing study of 240
families were included. 96% of adult and 51% of children were female. On average, children
performed MVPA for 17.6 (SD =10.9) and 17.7 (SD = 11.3) min/d for WD and WE days. Parents
averaged 1.4 (SD = 1.44) hours total time participating in PA with their children on weekends
(2 days) and 2.63 hours (SD= 2.61) over 5 WD. Parent MVPA (accelerometry) during the same
week showed a significant effect on child MVPA on both WD ( = 0.246, p = 0.044) and WE
days ( = 0.428, p = 0.005). However self-reported average daily waking time spent with
children had no effect on sedentary time (WD = 1.821, p=0.46; WE =0.111, p=0.98) or
MVPA (WD = 0.158, p=0.63; WE =0.255, p=0.60) in children. Neither did time spent in PA
with children in the week prior affect sedentary time (WD = -5.626, p=0.27; WE = -0.784,
p=0.95) or MVPA (WD = 0.658, p=0.30; WE =1.209, p=0.36).

Conclusions: Measurement of MVPA simultaneously with accelerometers showed an

association between parent and child MVPA. However parent reports of waking time spent
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2179

with children and time spent in PA with children did not affect objectively measured child
MVPA or sedentary time. Self-reported measures of time children and parents spend
together may not be sufficient. Technology tracking movement and proximity to another
person may be needed to understand these effects.

CORRESPONDING AUTHOR: Helena H. Laroche, MD, University of Iowa, Iowa City, IA, 52245;
[email protected]
S2180 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C230 6:00 PM-7:00 PM

ENHANCING INTERNAL VALIDITY IN WORK PERFORMANCE ASSESSMENTS WITH ACTIVE

WORKSTATIONS: RECOMMENDATIONS FROM A CASE STUDY

Liza S. Rovniak, PhD, MPH1, Jay Cho, BS2, Mari Greecher, PhD1, Erica Mellinger, BS1, Jason
Stanton, Student2, Thomas Hartmann, Student2, Melissa Bopp, PhD3, Christopher N.
Sciamanna, MD, MPH1, Andris Freivalds, PhD, CPE2

1
Penn State College of Medicine, Hershey, PA; 2Pennsylvania State University, State College,
PA; 3Pennsylvania State University, University Park, PA

Background: As lack of time is a common barrier to exercise, many studies are investigating if
adults can concurrently use active workstations (e.g., pedaling or treadmill desks) while
accomplishing productive office work. Experimental studies that assess if concurrent physical
activity affects work performance should be set up with high internal validity. However, there
is a lack of established protocols to maximize internal validity in studies assessing the effects
of active workstations on work performance.

Methods and Results: We present a case study of efforts to maximize internal validity in an
ongoing lab-based experiment with approximately 90 sedentary adults that is comparing the
effects of using a compact under-the-desk pedaling device at 17 and 25 watts vs. a standard
seated condition on office-based work performance. First, to ensure a consistent pedaling
speed at each wattage level, we designed a pedaling speed monitor that uses Hall effect
sensors with a microcontroller to transmit participants pedaling data to a computer in real-
time. Participants substantively deviating from the targeted speed during our experiment
have received real-time visual and auditory prompts, yielding typical adherence rates of over
85% to the pedaling speed protocol. Furthermore, on four separate trials, the 60 observed
pedaling rotations corresponded 100% to the actual-measured pedaling rotations recorded by
the speed monitor, indicating reliable and valid measurement. Second, to facilitate valid
assessment of work performance at different pedaling speeds, we selected typing, reading,
and phone/logic tasks with non-significant learning curves/improvement after performing two
initial practice trials, and we developed scripted task instructions. Third, based on pilot-
testing, to control confounding ergonomic/environmental factors we used an adjustable-
height desk, a non-swivel and non-wheeled adjustable-height chair, non-slip equipment mats,
and 30 second to 5 minute rest/hydration breaks between randomly-sequenced experimental
tasks.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2181

Conclusions: Assessing the effects of active workstations on work performance requires

controlling measurement errors related to physical activity, work performance, and
extraneous ergonomic and environmental factors. Grant proposals should budget funds for
developmental work to maximize internal validity and journal editors should require
describing this work, so that efforts to integrate active workstations into workplaces are
based on valid and reproducible science.

CORRESPONDING AUTHOR: Liza S. Rovniak, PhD, MPH, Penn State College of Medicine,
Hershey, PA, 17033; [email protected]
S2182 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C231 6:00 PM-7:00 PM

EXERCISE-RELATED ATTITUDINAL CHANGES AFTER 20 HOURS OF COGNITIVE TRAINING

Tiffany Bullard, B.S. 1, Amanda L. Reber, MS, PhD2, Jason D. Cohen, B.S.1, Daniel Palac, MA3,
Aaron L. Wichman, Ph.D.4, Edward McAuley, PhD5, Sean P. Mullen, PhD1

1
University of Illinois at Urbana-Champaign, Urbana, IL; 2Central Queensland University,
Rockhampton, Queensland, Australia; 3University of Illinois at Urbana-Champaign,
Champaign, IL; 4Western Kentucky University, Bowling Green, KY; 5University of Illinois at
Urbana Champaign, Urbana, IL

Explicit and implicit attitudes can influence motivation for engagement and maintenance of
physical activity (PA) behavior. To date, no randomized controlled trial has been designed to
change PA-related attitudes through cognitive training (CT). The aim of the NHLBI-funded
Cognitive Regulation and Exercise (CORTEX) trial was to test the efficacy of a 20-hour CT
program to increase exercise adherence for middle-aged adults by enhancing exercise self-
efficacy and self-regulatory strategy-use. The CT intervention emphasized dual task training
and self-as-exerciser priming while the control group watched health education videos. All
participants were then enrolled in a 4-mo exercise program. In a secondary data analysis, path
modeling was used to test mechanisms of exercise-specific cognitive change. Explicit attitude
was assessed via a composite score from the Exercise Self-Schema Questionnaire. Implicit
attitude was assessed via a PA-specific Single-Category Implicit Association Task (S-CIAT). Pre-
post CT change scores were computed for both mediators. The total number of electronically-
recorded fitness facility visits was the dependent PA behavioral outcome. The model
(2 (5)=10.93, p=.053, CFI=0.93, SRMR=0.03) adjusted for age, gender, education, race,
baseline PA, injury, and pre-exercise dropout, fit the data with group having a significant
direct effect on total visits (=0.30, p= < 0.05), favoring the CT group. The results show
significant (p= < 0.05) direct effects of group on changes in self-schema (=-0.45), which had a
direct effect on total visits (=0.13) and a marginally significant indirect effect of group on
total visits via exercise self-schema (=-0.13, p=0.08). Group and total visits were not
predicted by implict attitudes. In a simple regression model, baseline Implicit Assoication Task
(S-CIAT) did predict lifestyle PA (=0.22, p= < 0.05) assessed via Fitbit steps at program end.
These findings are consistent with prior research, indicating that explicit self-schema
influences goal-directed behavior and lifestyle PA may be more sensitive to effects from
implicit attitudes. Changing explicit and implicit attitudes towards PA can increase adherence
to exercise programs.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2183

CORRESPONDING AUTHOR: Tiffany Bullard, B.S. , University of Illinois at Urbana-Champaign,

Urbana, IL, 61801; [email protected]
S2184 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C232 6:00 PM-7:00 PM

FEASIBILITY AND EFFICACY OF A PHYSICAL ACTIVITY INTERVENTION FOR THE PREVENTION OF

POSTPARTUM DEPRESSION: A RANDOMIZED TRIAL

Beth Lewis, PhD, Katie Schuver, PhD, Dwenda Gjerdingen, MD, MS, Carrie Terrell, MD, FACOG,
Melissa Avery, PhD, CNM

University of Minnesota, Minneapolis, MN

Approximately 10-13% of women experience postpartum depression. Furthermore, one-third

of women who have a history of depression report depression following childbirth. Therefore,
it is important to examine the efficacy of low cost, innovative interventions designed to
prevent postpartum depression among women at risk. Research indicates that physical
activity interventions prevent depression among the general population; however, few
studies have examined the efficacy of these interventions for postpartum women. The
purpose of this study was to examine the efficacy of a six-month physical activity intervention
on the prevention of postpartum depression. Specifically, 450 sedentary postpartum women
(participants were less than eight weeks postpartum) who had a history of depression prior to
pregnancy and were not taking an antidepressant were randomized to one of three groups:
(1) Telephone-based exercise intervention (based on social cognitive theory and self-
determination theory); (2) telephone-based wellness/support intervention (topics such as
sleep, stress, and healthy eating); or (3) usual care. Participants were recruited via online,
email, and print advertisements. Participants were mostly married (72%), a majority had a
college education (68%), and several were employed (65%). Forty-one percent of the
participants identified as a racial or ethnic minority. Depression was assessed using the
Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I) and the Edinburgh Postnatal
Depression Scale (EPDS). Physical activity was assessed using the 7-Day Physical Activity
Recall Interview and the ActiGraph. Of the 423 participants due for their six-month
assessments, 91% have completed this assessment. The assessments are currently ongoing
and the six-month assessments will be available at the time of the conference presentation.
Based on a 3-item questionnaire administered at 9 months, preliminary data indicate that
91% of the participants reported enjoying the program and found it useful. Our study
indicated that a physical activity intervention is feasible and acceptable among postpartum
women at risk for depression.

CORRESPONDING AUTHOR: Beth Lewis, PhD, University of Minnesota, Minneapolis, MN,

55455; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2185

C233 6:00 PM-7:00 PM

FEASIBILITY OF A CLIMATE-BASED AFTER SCHOOL PHYSICAL ACTIVITY INTERVENTION FOR

MIDDLE SCHOOL YOUTH; CONNECT THROUGH PLAY

Nicole Zarrett, PhD1, Dawn K. Wilson, Ph.D., FSBM1, Bethany A. Bell, PhD, MPH1, Brittany S.
Cook, M.A2, Michelle Abraczinskas, M.A.1, Stephen Taylor, B.S.1, Alex Roberts, B.A.1

1
University of South Carolina, Columbia, SC; 2University of south Carolina, West Columbia, SC

The proposed study tested the feasibility of the Connect through Positive Leisure Activities
for Youth (Connect) afterschool physical activity (PA) program, a climate-based intervention
that expands on the social-motivational constructs highlighted by Self-Determination
Theoryand Achievement Goal Theory, to increase the PA of underserved (i.e., low income,
minority status) middle school youth within afterschool programs (ASPs). The Connect
intervention aims to improve the PA social motivational climate of pre-existing ASPs by
emphasizing social benefits of PA (i.e., developing friendships, peer connectedness, staff-
youth connections, and group belonging), as the primary goal, purpose and achievement of
PA. The two primary components of the program involved Get to know you small group
sessions aimed at providing youth guided social opportunities that facilitate peer acceptance,
cooperation, and friendship, and a novel socially-oriented PA curriculum infused within ASPs
free play sessions. For the current study, intervention feasible was evaluated by examining: 1)
youths greater perceived connectedness and affiliation goal orientation toward PA from
baseline to post-intervention (e.g., The Need for Relatedness Scale; Social Motivational
Orientations Scale for Sport) using youth survey; 2) changes in targeted social climate
components from baseline to post-intervention as measured by systematic observations
(SOCARP and MCOT supplement) and youth surveys (e.g., Perceived Teacher Support Scale;
Peer Motivational Climate in Youth Sport Questionnaire), and; 3) whether youth in the social
climate program (vs. comparison) demonstrate greater increases in positive PA affect, and
cognitions (e.g., The Intrinsic Interest Scale), and moderate-to-vigorous PA (accelerometer
and SOCARP) from baseline to post-intervention. Comparison of ASPs receiving the Connect
intervention (N=3 ASPs) with ASPs providing the typical curriculum serving as the control (N=3
ASPs) yield preliminary data on the effectiveness of social mediators for improving PA among
at-risk youth that can inform future intervention design and youth programming policy. For
example, preliminary analyses indicated changes in critical social climate components (e.g.,
positive peer interactions [t(59)=3.60, p=.000]; peer PA social support [t(47)=2.46, p=.001]),
youth PA motivations (e.g., PA regulatory motives [t(47)=2.12, p=.05]; PA engagement
[t(59)=6.73, p=.000] ) and youth PA ([t(59)=4.65, p=.000]).
S2186 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Nicole Zarrett, PhD, University of South Carolina, Columbia, SC,
29229; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2187

C234 6:00 PM-7:00 PM

HOW LONG DOES THE POSITIVE AFFECTIVE RESPONSE TO EXERCISE LAST? AN EMA STUDY.

Lauren Connell, Ph.D.1, Shira Dunsiger, PhD2, Jessica Emerson, MS3, Courtney J. Stevens, MA4,
David Williams, Ph.D.5

1
The Miriam Hospital and Alpert School of Medicine at Brown University, Providence, RI; 2The
Miriam Hospital and Brown University, Portsmouth, RI; 3Brown University, East Providence,
RI; 4Alpert Medical School of Brown University, Providence, RI; 5Brown University School of
Public Health, Providence, RI

Background/Aim: Prior research has shown that many people experience a positive shift in
affective valence from pre- to post-exercise. This study aims to describe the duration of this
positive affective response in a natural setting.

Methods: Participants included 59 overweight or obese insufficiently active adults enrolled in

a 6-month field-based exercise (i.e., walking) promotion intervention. Participants were asked
to self-report via ecological momentary assessment (EMA) the beginning and end of each
exercise session, and were then prompted to report their affective valence on the 11-point
Feeling Scale (FS; from -5 [very bad] to +5 [very good] with zero as neutral) immediately prior
to exercise, immediately following exercise, and 15 minutes after the end of exercise.
Participants were also prompted to respond to the FS at random intervals throughout the
day. The responses to the FS were used to characterize how long it took for the positive shift
in affective valence from immediate pre- to immediate post-exercise to return to pre-exercise
levels among those who experienced a positive shift from pre- to post-exercise.

Results: The data set included 2788 sessions of exercise of at least 10 minutes over the 6-
month period from the 59 participants. Of these, 1169 (42%) displayed a positive shift in
affective valence from pre- to post-exercise, 1135 (41%) showed no change, and 480 (17%)
showed a negative shift. For 17% of the exercise sessions for which there was a positive shift
in affective valence, participants affect returned to pre-exercise levels within 15 minutes after
exercise. Affective valence returned to pre-exercise levels in 25%, 31%, 41%, 47% and 51% of
sessions by 1 hour, 2 hours, 3 hours, 4 hours, and 5 hours post-exercise, respectively,
representing a significant dissipation of affective valence over time (Z=4.02, p < 0.001).

Discussion: The commonly experienced positive shift in affective valence from pre- to post-
exercise dissipated over time, but nonetheless persisted in nearly 50% of exercise sessions by
5 hours post-exercise, suggesting a lasting positive effect of acute exercise on affective
S2188 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

valence in a significant proportion of exercise sessions. In future research we plan to identify

predictors of the persistence versus dissipation of the positive effect of exercise on affective
valence.

CORRESPONDING AUTHOR: Lauren Connell, Ph.D., The Miriam Hospital and Alpert School of
Medicine at Brown University, Providence, RI, 02912; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2189

C235 6:00 PM-7:00 PM

IDENTIFYING UNIQUE PATTERNS OF DAILY ACTIVITY AMONG LATINAS ENROLLED IN A 12-

MONTH PHYSICAL ACTIVITY INTERVENTION

Lucas Carr, Ph.D.1, Shira Dunsiger, PhD2, Bess H. Marcus, Ph.D.3

1
University of Iowa, Iowa City, IA; 2The Miriam Hospital and Brown University, Portsmouth, RI;
3
University of California, San Diego, La Jolla, CA

Background: Traditional randomized controlled trial physical activity interventions have relied
on measuring physical activity during brief time points over the course of the intervention
(baseline, 6 months, 12 months) to determine effectiveness. Adding a continuous measure of
activity (pedometer) would allow interventionists to better understand the impact of
intervention exposures on patterns of physical activity through the use of latent class
analyses.

Objective: The purpose of the current study is to use a latent class model (LCM) to identify
unique patterns of daily step counts over 12 months, among Latinas enrolled in a print-based
physical activity intervention.

Methods: Participants were low-active females (N=122, mean age=41.1, SD=10.0) randomized
to an individually tailored and linguistically adapted physical activity intervention. Objective
daily steps were recorded using Omron HJ720ITC pedometers which were given to
participants as part of the intervention. Steps were recorded daily for 12 months. Self-
reported physical activity (as collected via the 7-day PAR) was the primary study outcome, and
was collected at baseline, 6 and 12 months.

Results: Results suggested a 5-class model was best supported by the data. Based on the
observed patterns, classes were labeled as non-engaged (26%), steady adopters (29%),
early attempters (18%), late attempters (12%), and unstable engagers (15%). Analyses
suggested a significant association between class and mean min/week of self-reported
moderate to vigorous physical activity (MVPA), p < .01, such that "steady adopters" reported
significantly more MVPA at 6 and 12 months compared to all other classes. Furthermore, "late
attempters" reported significantly more MVPA at 12 months compared to "non-engaged",
"early attempters", and "unstable engagers", p < .05. In addition, between-class differences
were found in baseline demographics (BMI, household make-up), and changes in psychosocial
constructs (self-efficacy, enjoyment, process variables) over time.
S2190 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions: These findings identified five unique patterns of activity uptake among
participants of a 12 month intervention. This information could be used to further enhance
future interventions by identifying participants who are at risk of failing the intervention (e.g.,
non-engaged, unstable engagers) early in the process and providing those individuals with
enhanced or additional treatment.

CORRESPONDING AUTHOR: Lucas Carr, Ph.D., University of Iowa, Iowa City, IA, 52242; lucas-
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2191

C236 6:00 PM-7:00 PM

LONGITUDINAL RELATIONSHIPS BETWEEN SELF-CONCEPT AND NEIGHBORHOOD SOCIAL LIFE

AS PREDICTORS OF PHYSICAL ACTIVITY

Allison Sweeney, PhD1, Dawn K. Wilson, Ph.D., FSBM1, M Lee Van Horn, PhD2

1
University of South Carolina, Columbia, SC; 2U of new mexico, Abq, NM

People are more likely to engage in physical activity (PA) when PA engagement is seen as a
central to their self-concept (SC). Self-Determination Theory and Social-Cognitive Theory
propose that volition (i.e., self-determined-motivation; SDM) and competency (i.e., self-
efficacy; SE) are critical for PA, but these constructs may fluctuate across time as motivation
and barriers change. SC, which is grounded more centrally in ones self-identity, should be a
more stable predictor of PA. Little research has directly compared how SC, SE, and SDM
predict longitudinal PA, with few studies testing these relationships among underserved
groups. Furthermore, it remains unclear how well such individual-level factors predict PA
when accounting for broader ecological factors including perceived neighborhood social life
(SL) and neighborhood satisfaction (NS). We hypothesized that SC would be a stronger
predictor of longitudinal PA (relative to SDM or SE) and that SL and NS would also predict PA.
Data were collected from African Americans (N = 417, Mage = 51.65, 62%
female, Mincome <= $25,000/yr) enrolled in the Positive Action to Todays Health randomized
trial. PA was assessed using 7-day accelerometry-estimates and psychosocial data were
collected at baseline, 12- and 24-months. Multilevel growth modeling was used to test for
within- and between-person effects. Sex, age, time, BMI, income, and community were
included as covariates (no community effects were observed). There were no significant
within-person effects. At the between-person level, only average SC was significant, (b = .58,
SE = .24, p= .018), which was not moderated by time. At baseline people with a higher average
SC engaged in greater PA than those with a lower average SC, and the positive association
between average SC and PA at baseline remained constant across time. A significant
interaction between time and average SL (b = .29, SE = .11, p = .010) also showed that people
with a higher average SL increased their PA over time compared to people with a lower SL.
These results suggest that SC and SL are critical predictors of PA in African Americans, which
should be considered in future interventions. These factors may be particularly relevant to
long-term changes and maintenance of PA in underserved populations.

CORRESPONDING AUTHOR: Allison Sweeney, PhD, University of South Carolina, Columbia, SC,
29201; [email protected]
S2192 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C237 6:00 PM-7:00 PM

MORE THAN JUST A GAME? A RANDOMIZED CONTROLLED TRIAL OF POKMON GO ON

PHYSICAL ACTIVITY HABITS

Allene Gremaud, B.S., Jacob Simmering, PhD, Sarah Coe, B.S., Nicholas Evans, B.A., B.S., Linnea
Polgreen, PhD, James Cremer, PhD, Alberto Segre, PhD, Philip Polgreen, M.D., Lucas Carr,
Ph.D.

University of Iowa, Iowa City, IA

Background:With only 20% of Americans meeting the federal physical activity guidelines,
there is a need for scalable and effective interventions. Pokmon Go is a free, augmented
reality game that encourages users to explore/walk to real locations to complete the games
objectives. While extremely popular (100 million downloads, 20 million active users daily), the
efficacy of Pokmon Go for increasing physical activity habits has not yet been tested.

Objective: The purpose of this study was to test the efficacy of Pokmon Go for increasing
daily steps as compared to a control group that did not receive Pokmon Go over eight weeks.

Methods: We recruited 100adults who owned a smart phone and reported having never
played Pokmon Go previously. Upon enrollment, all participants were provided a Fitbit Zip to
wear daily throughout the intervention as an objective measure of physical activity.
Participants were randomized (1:1 ratio) to either: 1) Fitbit only (FB; N=50; 80% female, 37.3
14.0 years; BMI=29.1 6.6 kg/m2); or 2) Fitbit + Pokmon Go (FB+P; N=47; 70% female; 39.1
16.0 years; BMI=28.7 6.5 kg/m2). Average daily steps and number of daily active minutes
(>100 steps/minute) over the 8 week intervention were analyzed with linear mixed models.

Results: The FB+P arm walked more daily steps/day (+539 steps/day; SE=630, t = 0.86) and
completed more active steps (+5.0; SE=3.56; t = 1.14) than FB participants on average over the
entire eight-week intervention but neither effect was statistically significant. However, by the
end of eight-weeks, the FB+P arm completed significantly more active minutes than the FB
group.

Conclusions: Pokmon Go appears to have had a modest, but positive, impact on physical
activity levels among these participants. Given its widespread adoption and use, this
innovative game and others like it hold significant potential for improving physical activity
levels on a large scale. Future collaborative relationships between public health researchers
and game developers are encouraged.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2193

CORRESPONDING AUTHOR: Allene Gremaud, B.S., University of Iowa, Iowa City, IA, 52245;
[email protected]
S2194 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C238 6:00 PM-7:00 PM

OBESE ADULTS' FORECASTED AND ACTUAL EXPERIENCES DURING PHYSICAL ACTIVITY: IS

THERE A DISCREPANCY?

Leah M. Schumacher, M.S., Margaret Clark, B.S., Alexandra Convertino, B.S., Evan M. Forman,
PhD, Meghan L. Butryn, PhD

Drexel University, Philadelphia, PA

Affective experiences during physical activity (PA) are related to current and future PA
behavior. Limited research has examined individuals predictions of how they will feel during
PA, despite the potential of such forecasts to affect PA decisions. The present study examined
forecasted and actual affective (i.e., excitement, contentment) and physical (i.e., energy,
strength, pain) experiences during PA, as well as the accuracy of forecasts based on current
PA level, among 311 obese adults beginning weight loss treatment. Participants (MBMI=35.6
kg/m2, 78.5% female) completed a half-mile walk test on a treadmill in the laboratory. Prior to
the task, participants forecasted how they would feel during the walk along five bipolar scales
(bored-enthusiastic, irritable-content, exhausted-energetic, weak-strong, great pain-no pain).
Negative ratings were anchored at 0; positive ratings were anchored at 100. Mid-walk,
participants used the same scales to report on their actual experiences. After the walk, PA was
measured for 7 days using accelerometers, and participants were classified as currently
engaging in no PA (no bouted moderate-to-vigorous PA/week) or in some PA (10 min. of
moderate-to-vigorous PA/week). Results revealed participants generally made favorable
forecasts (Ms = 64.3 to 82.8). Contrary to past research, participants experienced enthusiasm
(M=60.7) was lower than forecasted (M= 64.3; p=.01, partial eta squared=.02). However,
there was a trend for experienced energy (M=70.7) being greater than forecasted (M=68.9,
p=.07). Differences also were observed in forecasting accuracy for strength and pain based on
7-day PA level (no vs. some). While individuals engaging in some PA forecasted slightly greater
strength and less pain than was experienced, individuals engaging in no PA forecasted lower
energy and more pain than was experienced (ps < .03). Of note, however, all effect sizes and
differences in forecasted and actual ratings were small. These findings suggest obese adults
make generally favorable forecasts about PA experiences, though forecasts may not be
completely accurate. Additionally, inactive individuals may make greater, more negative
forecasting errors for some aspects of PA. As this is the first study to examine PA forecasting
among obese individuals, additional research on this topic is needed. Future studies should
examine whether forecasting errors impact PA decisions to help clarify the clinical significance
of these findings.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2195

CORRESPONDING AUTHOR: Leah M. Schumacher, M.S., Drexel University, Philadelphia, PA,

19104; [email protected]
S2196 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C239 6:00 PM-7:00 PM

PHYSICAL ACTIVITY AND MULTIPLE SCLEROSIS: AN UPDATED META-ANALYSIS

Dominique Kinnett-Hopkins, B.S. 1, Brynn Adamson, MS1, Kate Rougeau, M.S.1, Robert W.
Motl, PhD2

1
University of Illinois at Urbana-Champaign, Urbana, IL; 2University of Alabama at
Birmingham, Birmingham, AL

Introduction: Approximately 2.5 million people worldwide are living with multiple sclerosis
(MS). Evidence suggests that persons with MS can manage the manifestations of the disease
with physical activity. This meta-analysis examines physical activity levels in persons with MS
compared with non-diseased and other clinical populations. Methods: We searched PUBMED,
PsycINFO and Web of Science using the key words physical activity, exercise and physical
fitness in conjunction with MS. Manual searches of bibliographies of the retrieved papers
were included. We conducted a quantitative synthesis of the difference in physical activity
levels among persons with MS compared with non-diseased and other clinical populations and
then examined moderators that explain variation in the overall difference of physical activity
using meta-analytic procedures. Results: Overall, 21 studies were included involving 5,303
persons with MS and yielded a mean effect size (ES) of -0.567 (95% CI= -0.763,-0.371). The
weighted mean ES was heterogeneous (Q=443.811, df=31, p < 0.001). The magnitude of the
ES increased when comparing non-diseased populations versus the MS population but
decreased when comparing clinical populations with the MS population. Likewise, the ES was
larger for objective measures of physical activity versus subjective measures of physical
activity. Conclusion: The cumulative evidence suggests that persons with MS are less
physically activity than non-diseased, but not clinical populations.

CORRESPONDING AUTHOR: Dominique Kinnett-Hopkins, B.S. , University of Illinois at Urbana-

Champaign, Urbana, IL, 61801; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2197

C240 6:00 PM-7:00 PM

QUALITY PARTICIPATION AND DELIVERY STRATEGIES OF PHYSICAL ACTIVITY PROGRAMS FOR

VETERANS WITH A PHYSICAL DISABILITY

Celina H. Shirazipour, PhD(cand), MHK, BA, Amy E. Latimer-Cheung, PhD

Queen's University, Kingston, ON, Canada

Introduction: The number of physical activity programs for veterans with a physical disability
has increased due to improved knowledge of the physical and psychosocial benefits of
physical activity participation for veterans post-injury. When developing physical activity
programs for veterans post-injury, it is important that program developers and staff focus not
only on whether their participants are showing up (i.e. the quantity of a veterans
participation) but whether they are providing a quality program that fosters quality physical
activity experiences. A quality program or experience can be conceptualized as including the
following elements: autonomy; belongingness; challenge; engagement; mastery; and
meaning. While the elements of quality participation have been identified, an important
research gap exists when building evidence for how to develop and deliver quality
experiences for veterans, particularly the need to understand what program delivery
strategies result in quality physical activity experiences. The purpose of the current study is to
identify strategies used to deliver physical activity programs to veterans with a physical
disability, and interpret these strategies in relation to their potential utility for fostering
quality participation.

Methods: Semi-structured interviews were conducted with program staff from three veteran
physical activity programs, and program documentation collected. Each program represented
a different country, with programs for veterans from the United States of America, United
Kingdom, and Canada represented. Interviews and documents were first analyzed using a
thematic analysis to identify delivery strategies. Strategies were then interpreted in relation
to their potential links to fostering quality elements of participation.

Results: Four themes were identified as program delivery strategies: (1) foster social
connections; (2) challenge participants; (3) tailor programs and outcomes to match participant
needs; and (4) include knowledgeable coaches/instructors. All strategies linked to elements of
quality participation.

Discussion/Conclusion: This study provides a first look at physical activity program delivery
for veterans with a physical disability. The findings, evidence of real-world strategies used to
S2198 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

deliver programming to veterans, provide indications of how future programs may potentially
create and deliver quality physical activity experiences.

CORRESPONDING AUTHOR: Celina H. Shirazipour, PhD(cand), MHK, BA, Queen's University,

Kingston, ON, K7L3N6; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2199

C241 6:00 PM-7:00 PM

SOCIO-DEMOGRAPHIC MODERATORS OF ENVIRONMENT-PHYSICAL ACTIVITY ASSOCIATIONS:

THE INTERNATIONAL PREVALENCE STUDY RESULTS

Lilian Perez, MPH1, Terry L. Conway, PhD2, Adrian Bauman, PhD3, Heather R. Bowles, PhD4,
James F. Sallis, PhD2

1
UC San Diego/San Diego State University, San Diego, CA; 2University of California, San Diego,
San Diego, CA; 3University of Sydney School of Public Health, Sydney, N/A, Australia; 4National
Cancer Institute, Rockville, MD

Background: Neighborhood environmental factors are related to physical activity (PA).

Associations between environmental factors and PA may vary across socio-demographic
factors, but evidence has been limited to single-country studies with little variability in
geographic and population characteristics. We investigated socio-demographic moderators of
associations between perceived environmental factors and self-reported PA using data from a
multi-country study of adults PA.
Methods: Questionnaires assessed socio-demographic factors, 7 perceived neighborhood
environmental factors, and total physical activity as measured by the IPAQ short from 11,144
adults from 9 countries. Logistic models, weighted and adjusted for country site, examined
associations of the perceived environment factors with meeting PA guidelines (PAG). To
assess moderator effects, we tested interactions of age, sex, and education with each
perceived environmental factor.
Results: 3 out of 21 interactions tested were significant (p < .05). Sex moderated associations
of safety from crime and the presence of transit stops with meeting PAG. Perceived safety
from crime was positively related to meeting PAG only among women (OR= 1.24, 95% CI:
1.07-1.45). Perceived presence of transit stops was positively related to meeting PAG only
among men (OR=1.29; 95% CI: 1.03-1.62). Education moderated associations between
perceived presence of transit stops and meeting PAG. Perceived presence of transit stops was
positively related to meeting PAG among the high education group (OR=1.27; 95% CI: 1.04-
1.56) but inversely related in the low education group (OR=0.71, 95% CI: 0.53-0.94).
Conclusions: For women, safety from crime was the most important perceived environmental
factor for meeting PAG while for men it was the presence of transit stops. In the high
education group, the presence of transit stops was important for meeting PAG. Among those
with lower education, the presence transit stops may be a barrier to meeting PAG. Disparities
in global PA levels may be partly explained by interactions between neighborhood
environmental and socio-demographic factors.
S2200 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Lilian Perez, MPH, UC San Diego/San Diego State University, San
Diego, CA, 92104; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2201

C242 6:00 PM-7:00 PM

SUNBURN AND SUN PROTECTIVE BEHAVIOR IN RELATIVELY ACTIVE AND INACTIVE TEENS

Frank Perna, EdD, PhD1, Kasey Morris, PhD2

1
National Cancer Institute, Rockville, MD; 2National Cancer Institute, Bethesda, MD

Physical activity (PA) has been strongly related to melanoma in adults, ostensibly via increased
sun exposure. Surveillance data also documents similarly strong positive associations between
PA and sunburn, the best clinically relevant proximal intervention target for melanoma.
However, we lack national data regarding these associations in teens. These data are vital
since children typically have more outdoor time, are more physically active, and have greater
sensitivity to ultraviolet radiation than adults. Using data from the National Cancer Institutes
Family Life, Activity, Sun, Health, and Eating (FLASHE) national survey, our purpose was to
compare incidence of sunburn and sun-safe behaviors among relatively physically active and
inactive teens (N = 1661) aged 12 -18. More girls (47%) than boys (39%)reported sunburn in
the last year (2 (1) = 8.03, p = .003). In logistic regressions stratified by gender and controlling
for age and race, total PA was associated with slightly increased odds of sunburn (at least one
in last 12 months) for girls (OR = 1.003, p = .009), but not for boys. However, out-of-school PA
was associated with increased odds of sunburn (OR = 1.005, p = .005) for both boys and girls.
In multiple regressions controlling for age, race, and gender, PA was positively associated with
sunscreen and hat use (bs = .003 & .004, ps < .0001, respectively), but not use of other
clothing or sun-safe practices. Unfortunately, frequent sunscreen use (often or always on
warm sunny days) was uncommon (34%) and positively associated with sunburn (OR = 1.14,
p = .012), whereas clothing (except hat) use was unrelated to sunburn. These data are
consistent with prior reports of high sunburn rate among teens, but extend knowledge by
indicating that sunburn incidence may be heightened, albeit minimally, by PA participation,
especially out-of-school PA. The association of PA with both sunburn and some sun-safe
behaviors suggests several possibilities, including, reverse causation between sunburn and
sun protection or the influence of a third unmeasured variable; less than optimal use of
sunscreen or pairing of sunscreen with sun-protective clothing among physically active teens
(given likely higher sun exposure accompanying PA); or possible strategic intent for sunscreen
use gone awry among the physically active (i.e., to tan, but not burn). We discuss these
prospects and implications for future research that may explore the extent to which motives
for PA (e.g., appearance) impact sun-safe behavior, and intervention research opportunity
with youth sports and recreation organizations who may leverage their infrastructure and rule
making authority to infuse sun-safe policies and practices.
S2202 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Frank Perna, EdD, PhD, National Cancer Institute, Rockville, MD,
20850; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2203

C243 6:00 PM-7:00 PM

TRACKING PATTERNS OF FATIGUE AND THE SYMPTOM BURDEN IN ACUTE LEUKEMIA - PACE-
AL RANDOMIZED CONTROLLED TRIAL

Mary Jarden, Ph.D. 1, Tom Mller, RN, MPH, Ph.D.2, Lars Kjeldsen, MD, DMSc3, Henrik Birgens,
MD, DMSc4, Lis Adamsen, Ph.D.5

1
University Hospitals Center for Health Research, Copenhagen University Hospital,
Rigshospitalet, Copenhagen, Hovedstaden, Denmark; 2Copenhagen University, Faculty of
Health Sciences, Copenhagen O, Hovedstaden, Denmark; 3Copenhagen University Hospital,
Rigshospital, Copenhagen, Hovedstaden, Denmark; 4Herlev Hospital and University of
Copenhagen, Herlev, N/A, Denmark; 5Universitets Hospitalernes Center for Sundhedsfaglig
Forsksning, Rigshospitalet, Copenhagen, Hovedstaden, Denmark

Background: Exercise and counseling interventions, widely used as rehabilitation strategies in

cancer patients to remedy disease and treatment-related sequel is novel in patients with
acute leukemia due to the life threatening nature of the disease and its intensive treatment.
In a two-center, randomized controlled trial (PACE-AL; clinicaltrials.gov identifier:01404520),
we assessed whether a multimodal intervention could modulate treatment-related symptoms
and physical and functional debilitation of patients with acute leukemia during outpatient
consolidation chemotherapy.

Objective: To evaluate the effect of a multimodal intervention on the type, severity and
pattern of commonly experienced treatment-related symptoms in patients with acute
leukemia during consolidation chemotherapy.

Material and methods: Patients were randomized to usual care or a 12-week, 3 day/week
supervised multimodal program: stationary cycle, resistance training, dynamic exercises,
relaxation training, nutritional support and a goal-setting counseling intervention. 70 patients
(mean 53.1 years) were randomly assigned to the intervention (n=32) or control group (n=34),
and 62 completed study requirements (88.6%). All patients scored their symptoms weekly
using the MD Anderson Symptom Inventory and Brief Fatigue Inventory during the 12 week
study period.

Results and conclusion: Fatigue and symptom patterns were identified throughout the 12-
week study period. The intervention group reported experiencing greater fatigue during the
study period, however there was no difference in symptom interference in daily life. At end-of
study, most symptoms were lessened in both groups, though there was a clear trend towards
S2204 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

reduced fatigue intensity experienced in the intervention group. Understanding the patterns
of fatigue and the symptom burden during consolidation chemotherapy treatment for
patients with acute leukemia can assist in the development of important strategies in
symptom management. The role of exercise and counseling integrated in the clinical setting
has the potential of optimizing the management of acute leukemia and may facilitate
resumption of everyday activities.

CORRESPONDING AUTHOR: Mary Jarden, Ph.D. , University Hospitals Center for Health
Research, Copenhagen University Hospital, Rigshospitalet, Copenhagen, Hovedstaden, 2100;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2205

C244 6:00 PM-7:00 PM

THE RELATIONSHIP BETWEEN CHRONOTYPE AND PHYSICAL QOL IN COLLEGE STUDENTS: THE
MEDIATING EFFECT OF FATIGUE AND ACADEMIC BURNOUT

Hae lim Noh, MA, Donghee Jeong, BA, Hyeju Ha, BA, Jungmin Yoon, BA, Eun-Jung Shim, Ph.D

Pusan National University, Busan, N/A, Republic of Korea

Objective: Evening chronotype has been previously reported to be associated with decreased
overall health of college students. According to the social jetlag theory, social time demands
evening type students to accordingly adjust their daily rhythm, which potentially results in
poor health outcomes. Given that regular schedule in college appears to be more workable
for morning type students, evening type students might be more vulnerable to fatigue and
burnout due to the inconsistency between their daily life and personal chronotype. The
present study examined whether fatigue and academic burnout mediate the relationship
between chronotype and physical quality of life (QOL).

Methods: The participants (N=500, mean age=21.3, 56.2% male) completed Composite Scale
for Mornings, Chalder Fatigue Scale, Maslach Burnout Inventory-Student Survey, and World
Health Organization Quality of Life Instrument-Short Version. PROCESS macro for SPSS (Hayes,
2013) was used to examine the mediating effects

Results: Chronotype, fatigue, and academic burnout accounted for 38.6% of variance in the
physical QOL. The direct effect of chronotype on physical QOL was not significant (b=.007,
p=.574). By contrast, the indirect effects of chronotype on physical QOL through fatigue
(b=.014, [95% bias-corrected bootstrap CI .005, .026]) and academic burnout (b=.007, [.002,
.014]) were significant. Furthermore, the dual mediating effect of fatigue and academic
burnout between chronotype and physical QOL was also significant (b=.001, [.000, .003]).

Discussion: Our results suggest that the chronotype (evening type) negatively affects physical
QOL through its impact on fatigue and academic burnout. Evening type students may benefit
from interventions aiming at reducing fatigue and academic burnout in order to improve their
physical QOL.

CORRESPONDING AUTHOR: Hae lim Noh, MA, Pusan National University, Busan, N/A, 617-
764; [email protected]
S2206 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C245 6:00 PM-7:00 PM

A QUALITATIVE STUDY OF PREPAREDNESS FOR END-OF-LIFE CARE AMONG FAMILY

CAREGIVERS CONFRONTING ADVANCED CANCER

Allen C. Sherman, PhD, Stephanie Simonton-Atchley, PhD, E Anderson, MS, Cindy W. Mikeal,
MS, Konstantinos Arnaoutakis, MD, Laura F. Hutchins, MD, Fade Mahmoud, MD, Natasha
Milojkovic, MD, Sarah E. Harrington, MD, James Y. Suen, MD

Winthrop P. Rockefeller Cancer Institute, University of Arkansas for Medical Sciences, Little
Rock, AR

End-of-life care presents families with myriad challenges; however, it remains unclear what
family caregivers need to feel prepared to manage this phase of illness. Perceptions of
preparedness may have considerable implications for caregiver burden and quality-of-life, as
well as for the cancer patients course of treatment and healthcare utilization. This study
sought a deeper understanding of preparedness for end-of-life care, from the caregivers
perspective.

Family members caring for cancer patients with end-stage disease participated in individual
open-ended interviews. (Independent interviews were conducted with patients as well and
are reported separately.) Sixty-four percent of caregivers were spouses and 72.7% were
white. A variety of malignancies were represented. Transcribed audiotapes were coded by 3
raters using thematic analysis, and data were organized using NVivo 11 software; recruitment
continued through data saturation.

Findings were classified into 6 major themes that caregivers regarded as salient aspects of
preparedness: (1) emotional well-being (e.g., managing caregiver stress, grief, resentment,
etc.); (2) spiritual well-being (e.g., religious strain or reassurance); (3) financial well-being (e.g.,
coping with healthcare costs and estate planning); (4) relationships with healthcare providers
(e.g., questions about prognosis, symptom management, and resources); (5) relationships
with family/friends (e.g., obtaining support, managing conflict); and (6) end-of-life decisions
(e.g., choices about intensity of care, utilization of hospice or palliative care, funeral plans).

Findings illuminate dimensions of preparedness that are important to family caregivers.

Readiness to accommodate these particular demands was viewed as a central challenge of
end-of-life care. Delineation of these core concerns provides a critical step toward construct
definition, and helps pave the way toward creation of useful measures and relevant screening
and intervention services.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2207

CORRESPONDING AUTHOR: Allen C. Sherman, PhD, Winthrop P. Rockefeller Cancer Institute,

University of Arkansas for Medical Sciences, Little Rock, AR, 72205; [email protected]
S2208 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C246 6:00 PM-7:00 PM

AN IDIOGRAPHIC MEASURE OF MEANINGFUL ACTIVITY PARTICIPATION IS LONGITUDINALLY

ASSOCIATED WITH GREATER WELL-BEING

Stephanie A. Hooker, M.S., M.P.H.1, Kevin Masters, PhD1, Jean A. Wood, MSW2

1
University of Colorado Denver, Denver, CO; 2University of Colorado, Arlington, VA

Background: Engagement in personally meaningful activities is associated with greater well-

being. However, most studies use cross-sectional or recall methods, and the researchers pre-
determine which activities are meaningful. Yet, individuals differ in the extent to which they
rate activities as meaningful; thus, personalized assessment of meaningfulness of activity may
enhance measurement precision. This study examined an idiographic measure of meaningful
activity participation in relation to well-being.

Methods: Participants (N=160; M age=43.3 years; 77% female) were recruited into a 4-week
longitudinal study. At baseline, they rated the meaningfulness of 48 common daily activities.
During the next 4 weeks (4 weekdays and 4 weekend days), participants received emailed
surveys on 8 random days asking them to recall their activities for the last 24-hours.
Participants also completed surveys at baseline and follow-up of life satisfaction, vitality,
meaning, purpose, and depressive symptoms. Meaningful activity participation scores were
calculated and entered into multilevel models and regression analyses to predict well-being at
4-weeks.

Results: Greater engagement in meaningful activities was positively related to thoughts of

meaning, b=.28, p < .0001, and positive affect, b=.19, p=.005, on that same day. After
controlling for relevant covariates, the average engagement in meaningful activities over eight
days was positively associated with subjective vitality, b=.30, p=.0004, life satisfaction, b=.22,
p=.0078, and purpose in life, b=.29, p=.0007, at 4-week follow-up. Engaging in personally
meaningful activities was not significantly associated with global ratings of meaning in life,
b=.15, p=.087.

Discussion: Results suggest that an idiographic measure of meaningful activity participation

may be useful for furthering understanding of relationships between daily engagement in
meaningful activity and greater well-being. Relationships between meaningful activity
participation and well-being were stronger than in previous studies using nomographic ratings
of meaningfulness. Future research should examine interventions targeting participation in
activity that is personally meaningful as a way to enhance well-being.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2209

CORRESPONDING AUTHOR: Stephanie A. Hooker, M.S., M.P.H., University of Colorado Denver,

Denver, CO, 80217; [email protected]
S2210 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C247 6:00 PM-7:00 PM

DEVELOPMENT OF A BILINGUAL PEDIATRIC CRANIOFACIAL QUALITY OF LIFE OUTCOMES

MEASURE

Julia Drizin, MA1, Viridiana J. Tapia, MPH2, Vanessa L. Malcarne, Ph.D.3, Amanda Gosman,
MD4

1
San Diego State University, San Marcos, CA; 2UCSD School of Medicine, Department of
Surgery, Chula Vista, CA; 3San Diego State University, San Diego, CA; 4UCSD, san diego, CA

Craniofacial conditions (CFCs) can have a severe impact on health-related quality of life
(HRQoL) for children. Yet, little is known about how diverse CFCs can differentially impact
HRQoL, and there is limited research on HRQoL in less prevalent CFCs. Presently, there are no
suitable instruments that measure HRQoL of patients with diverse CFCs from the perspective
of children and parents. Current measures either have not been validated for use with the
population of interest or focus on one specific CFC, precluding cross-group comparisons.
Additionally, many HRQoL measures are self-report in format, a challenge for pediatric use as
some children are too young to express themselves. Finally, most existing measures were
developed for use in one culture and language, thus limiting their utility. The present study
describes the development of a comprehensive, bicultural patient- and parent-reported CFC
HRQoL outcomes measure. A literature review and expert opinion were used to develop in-
depth interview questions. Interviews were conducted with 139 subjects: 89 parents of
patients ages 018+ and 50 patients ages 718+. English and Spanish speakers were
represented. CFCs included were cleft lip/palate, craniosynostosis, craniofacial microsomia,
microtia, and dermatological conditions. Qualitative analysis of interviews revealed six
domains that represent the conceptual framework of CFC-associated HRQoL. Themes that
emerged within each domain were operationalized into items in English and Spanish that
represent HRQoL issues for both patient and parent versions. Six bilingual, bicultural subscales
were developed based on the following domains: (1) Social Impact; (2) Psychological Function;
(3) Physical Function; (4) Family Impact; (5) Appearance; and (6) Finding Meaning. The new
scales will now undergo reliability and validity testing at multiple sites in the United States and
Mexico.

CORRESPONDING AUTHOR: Julia Drizin, MA, San Diego State University, San Marcos, CA,
92078; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2211

C248 6:00 PM-7:00 PM

EXPLORING INDIVIDUAL DISEASE BURDEN IN PATIENTS WITH ALOPECIA, ATOPIC DERMATITIS

AND PSORIASIS

Melissa M. Matos, Ph.D.

California School of Professional Psychology at Alliant International University, Artesia, CA

Exploring Individual Disease Burden in Patients with Alopecia, Atopic Dermatitis and Psoriasis

Melissa M. Matos, Ph.D., Peter Theodore, Ph.D., Ron Duran, Ph.D., and Erin OCallaghan, Ph.D.

California School of Professional Psychology

at Alliant International University, Los Angeles

The concept of individual disease burden measures individual disability by encompassing

psychological, physical, social and economic factors. The factors that contribute specifically to
the individual burden of skin disease remain unclear. The present investigation examined a
model in which emotional distress, body image change, social discomfort, perceived
discrimination, social support stress, unpredictable disease course, and financial stress were
expected to load onto individual skin disease burden (ISDB) as a latent construct. The
relationships between threat appraisals, illness perceptions, and self-efficacy with ISDB were
subsequently examined. A sample of 207 adults with alopecia, atopic dermatitis and/or
psoriasis participated in this study (n=175 participated online; n=31 participated at a local
dermatology office). The results supported emotional distress, social discomfort, and social
support stress broken into two distinct factorspartner and parental-based support stress as
components of ISDB. Body image change, perceived discrimination, unpredictable disease
course, and financial stress were not included in the latent construct. In the final model,
general self-efficacy was found to covary inversely with ISDB (r=-.471, pp2/df = 2.071, CFI=.91,
RMSEA=.07, SRMR=.06). This study offers preliminary evidence for emotional distress, social
discomfort, partner support stress, and parental support stress as important sources of
individual skin disease burden. Higher levels of general self-efficacy beliefs may decrease the
experience of individual skin disease burden. These findings serve to alert healthcare
providers to the possible components of individual skin burden that may affect treatment
planning, expectations, and barriers in patients with skin disease. Clinical implications also
extend to mental health treatment approaches that promote self-efficacy, as well as
S2212 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

psychological flexibility and acceptance, which may serve to reduce the individual burden of
skin disease.

CORRESPONDING AUTHOR: Melissa M. Matos, Ph.D., California School of Professional

Psychology at Alliant International University, Artesia, CA, 90701; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2213

C249 6:00 PM-7:00 PM

PERCEIVED EPILEPSY STIGMA MEDIATES RELATIONSHIPS BETWEEN PERSONALITY AND SOCIAL

WELL-BEING IN A DIVERSE EPILEPSY POPULATION

Seth Margolis, Ph.D.1, Luba Nakhutina, Ph.D.2, Sarah Schaffer, Ph.D.3, Arthur Grant, M.D.,
Ph.D.2, Jeffrey Gonzalez, Ph.D.4

1
Rhode Island Hospital, Alpert Medical School of Brown University, Providence, RI; 2SUNY
Downstate Medical Center, Brooklyn, NY; 3Northwell Health, Formerly North Shore-LIJ Health
System, Great Neck, NY; 4Yeshiva University, Bronx, NY

Perceived epilepsy stigma and reduced social well-being are prevalent sources of distress in
people with epilepsy. Yet, research on patient-level correlates of these difficulties is lacking,
especially among underserved groups. Racially/ethnically diverse adults with intractable
seizures (N=60, 62% female; 79% Black, 20% Hispanic/Latino, 8% White) were recruited and
completed validated measures of personality (NEO Five Factor Inventory, NEO-FFI-3),
perceived epilepsy stigma (Epilepsy Stigma Scale), and quality of life (Quality of Life Inventory
in Epilepsy, QOLIE-89). Controlling for covariates, ordinary least-squares regression and path
analysis evaluated the direct and indirect effects of NEO-FFI-3 neuroticism and extraversion
scores on epilepsy-related social well-being (i.e., combination of QOLIE-89 Social Functioning
and Social Isolation subscales, =.87), mediated through perceived epilepsy stigma. In
separate models, higher levels of neuroticism (N) and lower levels of extraversion (E) were
significantly and independently associated with greater perceived stigma (N path a=.71,
p=.005; E path a=-1.10, p < .005); stigma, in turn, was significantly and independently
associated with poorer social well-being (N path b=.23, p < .001; E path b=-.23, p < .001). Bias-
corrected bootstrap confidence intervals showed that neuroticism and extraversion were
indirectly associated with social well-being through their respective associations with
perceived stigma (N path ab=-.16, 95%CIs [-.347,-.044]; E path ab=.25, 95%CIs [.076,.493]).
These relations are consistent with a conceptual model wherein elevated neuroticism and
reduced extraversion may predispose patients with epilepsy towards feeling stigmatized,
which in turn leads to reduced social well-being. Future research should examine these
associations longitudinally. Intervention studies may benefit from promoting cognitive re-
appraisal strategies to reduce perceived stigma.

CORRESPONDING AUTHOR: Seth Margolis, Ph.D., Rhode Island Hospital, Alpert Medical School
of Brown University, Providence, RI, 02903; [email protected]
S2214 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C250 6:00 PM-7:00 PM

PREDICTORS OF SUBJECTIVE HAPPINESS IN OLDER ADULTS

Katie E. Greenawalt, BS, Elizabeth M. Orsega-Smith, PhD, Steve Goodwin, PhD, Jennie L.
Turner, BS, Erica Rathie, Student

University of Delaware, Newark, DE

Predictors of Subjective Happiness in Older Adults

K. Greenawalt, E. Orsega-Smith, S. Goodwin, J. Turner, E. Rathie

Department of Behavioral Health & Nutrition, University of Delaware

Introduction:

Research has shown that positive psychology interventions can enhance subjective wellbeing
and reduce depression, however, efficacy on older adult populations has not been widely
examined. Specifically, subjective wellbeing, life satisfaction, optimism, and positive emotions
have been shown to improve health and longevity. The present study sought to develop and
evaluate an intervention with the goals of enhancing happiness levels and overall mental
wellbeing in the older adult population.

Methods:

The Art of Happiness is an 8-week intervention that was conducted at 3 senior centers in the
state of Delaware. Each 90 minute class examined a different topic including; (1) defining
happiness, (2) stress management, (3) reflecting on happiness, (4) compassion and human
connection, (5) forgiveness, (6) transforming suffering, (7) mindfulness, and (8) humor. Pre
and post questionnaires assessed participant subjective happiness, stress, gratitude, life
satisfaction, depression, mindfulness, arousal states, and general demographic and health
information.

Results:

The 32 participants who completed the course were mostly married (43.8%), female (87.5%)
and Caucasian (90.6%), with an age range of 53-84 years of age (mean age= 70 years). Results
revealed that the intervention significantly increased participants subjective happiness (p <
.0001), with life satisfaction as the strongest predictor of happiness (r=.84, p < .0001).
Subjective happiness scores were also significantly correlated with gratitude (r=.69, p <
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2215

.0001), stress (r=-.50, p < .003), depression (r=-.46, p < .005), mindfulness (r=.34, p < .045),
and the arousal state of energy (r=.47, p < .006).

Conclusion:

These results suggest that older adults subjective happiness can be influenced or predicted
by variables such as life satisfaction, gratitude, stress levels, depression, mindfulness, and
energy levels. Programs like The Art of Happiness, which emphasize these topics may have
the potential to buffer the older adult population against poor mental health by improving
subjective happiness and overall mental well-being. This may lead to the development of
incorporating these types of programs into senior centers as part of their healthy lifestyle
curriculum.

CORRESPONDING AUTHOR: Katie E. Greenawalt, BS, University of Delaware, Newark, DE,

19716; [email protected]
S2216 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C251 6:00 PM-7:00 PM

QUALITY OF LIFE IN CHRONIC HEMODIALYSIS PATIENTS: RESULTS FROM THE PATIENT

CENTERED MEDICAL HOME FOR KIDNEY DISEASE STUDY

Denise M. Hynes, MPH, PhD, RN1, Anna Porter, MD2, Michael Berbaum, PhD3, Ifeanyi
Chukwudozie, MPH4, Linda Schiffer, MS, MPH5, Michael Fischer, MD, MHS2, Marian
Fitzgibbon, PhD6, Jose Arruda, MD7

1
ACADEMIC INTERNAL MEDICINE/UNIVERSITY OF ILLINOIS & US DEPT OF VETERANS AFFAIRS,
CHICAGO, IL; 2Nephrology Division/Department of Medicine/University of Illinois, Chicago, IL;
3
Institute for Health Research and Policy, Chicago, IL; 4Cancer Center/University of Illinois,
Chicago, IL; 5University of Illinois at Chicago, Chicago, IL; 6Pediatrics/University of Illinois,
Chicago, IL; 7Nephrology/Department of Medicine/COllege of Medicine/University of Illinois,
Chicago, IL

Background:

In the U.S., more than 400,000 individuals with end-stage renal disease require chronic
hemodialysis (HD). Chronic HD patients experience a high burden of morbidity, mortality,
healthcare use, and poor quality of life (QOL). Under current care models, chronic HD patients
receive fragmented care from multiple providers. The Patient-Centered Medical Home
(PCMH) is a team approach to provide coordinated care across the healthcare continuum.
While the PCMH model demonstrated early benefits for complex chronic diseases such as
diabetes, this study is among the first applied to chronic HD patients.

Methods:

Using a non-randomized quasi-experimental intervention trial design, we implemented a

PCMH model for kidney disease patients (PCMH-KD), which extended the existing HD care
team (comprising a nephrologist, nurse, social worker, and dietitian) by adding a general
internist, pharmacist, nurse coordinator, and health promoter. The PCMH-KD was
implemented over 18 months at two sites in Chicago: a non-profit, university-affiliated dialysis
facility and an independent for-profit dialysis facility operated by a national corporation. The
RE-AIM (Reach, Effectiveness, Adoption, Implementation, Maintenance) framework was our
organizing model for process evaluation. Patients were recruited and consented on a rolling
basis at each site beginning in January 2015; the intervention ended in August 2016. Sample
size estimates were based on self-reported QOL using the Kidney Disease QOL (KDQOL).
Additional patient and caregiver reported outcomes, clinical measures, healthcare use, and
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2217

staff perceptions were assessed every six months. We examined QOL as measured by the
KDQOL and used a mixed model to adjust for covariates.

Results:

Of 248 eligible patients, 175 (71%) patients ultimately consented, completed baseline
assessments and continued in the intervention. Components of the KDQOL related to mental
health (MCS) and kidney disease effects (KDE) were significantly improved in the intervention
group compared to usual care group over 18 months (MCS 48.9 to 51.7 (5.7%) respectively;
p=0.02; KDE 72.5 to 76.2 (5.1%), respectively; p=0.04). Covariates accounted for included
sociodemographics, dialysis vintage, site, self-reported diabetes, prior access to a primary care
provider, and time varying covariates for biochemical parameters for dialysis adequacy and
overall health. Other QOL components increased early in the observation period or varied
throughout. Improvements in depression, knowledge about kidney disease, and satisfaction
with primary care were also noted in the intervention group.

Conclusions:

Compared to the usual dialysis care team model, chronic HD patients in the PCMH-KD model
experienced improved QOL in mental health and kidney disease effects domains. Additional
factors that may affect QOL and other outcomes of the PCMH-KD warrant further exploration.

CORRESPONDING AUTHOR: Denise M. Hynes, MPH, PhD, RN, ACADEMIC INTERNAL

MEDICINE/UNIVERSITY OF ILLINOIS & US DEPT OF VETERANS AFFAIRS, CHICAGO, IL, 60608;
[email protected]
S2218 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C253 6:00 PM-7:00 PM

ROLES OF ADULT ATTACHMENT AND CAREGIVING STRESS IN DEPRESSION IN HISPANIC VS.

NON-HISPANIC CANCER CAREGIVERS

Amanda Ting, BS1, Hannah-Rose Mitchell, MPH2, Melanie McKenna, N/a2, Youngmee Kim,
PhD2

1
University of Miami, Miami, FL; 2University of Miami, Coral Gables, FL

Cancer diagnosis is life-changing also for the family caregivers. Resulting caregiving stress can
be manifested in depression outcomes. Lesser known is the extent to which self-reported
stress and adult attachment relate to caregivers depression around the time of the patients
diagnosis and treatment, and whether ethnicity moderates the associations. This study aimed
to extend current knowledge by investigating them.

Caregivers of recently diagnosed colorectal cancer patients (Stage I to IV) participated in the
study (N=102; 73% female; 49 years old; 55% Hispanic; 3 months post diagnosis). Self-
reported Hispanic vs non-Hispanic ethnicity, caregiving stress (stress overload subscale of the
Pearlin Stress Scale), depression (CES-D), as well as adult attachment orientations (MAQ:
secure, avoidant, anxious-ambivalent worry, and anxious-ambivalent merger) were measured.
Age and gender were covariates.

Results showed caregivers (40%) reported clinical levels of depression. Hierarchal regression
analysis revealed that female caregivers, greater caregiving stress, lower attachment security,
and higher attachment anxiety-merger related to greater depressive symptoms (s>.40,
ps=.54, p=.13, ns).

Findings provide evidence that gender, ethnicity, relationship quality, and perceived
caregiving stress played a significant role in family caregivers depression symptoms. These
results suggest that sociocultural factors and relationship quality to the patient may play an
influential role on the caregivers depression during the period of initial caregiving stress with
diagnosis and treatment. Findings warrant further investigation for specific factors at an
individual, dyadic, and cultural level in close relationships relating to promoting adaptive
outcomes of lower depression with caregiving stress.

CORRESPONDING AUTHOR: Amanda Ting, BS, University of Miami, Miami, FL, 33155;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2219

C254 6:00 PM-7:00 PM

THE PROTECTIVE ROLE OF SOCIAL SUPPORT IN CHRONIC LYMPHOCYTIC LEUKEMIA: A

CONDITIONAL PROCESS ANALYSIS

Travis Westbrook, M.A.1, Thomas Valentine, B.A.1, Eleshia Morrison, Ph.D.2, Barbara L.
Andersen, PhD1

1
The Ohio State University, Columbus, OH; 2Adult Pain Rehabilitation Center, Mayo Clinic,
Columbus, OH

Background: Chronic lymphocytic leukemia (CLL) accounts for one-third of all leukemia
diagnoses in the United States. Early stage patients are typically monitored without
treatment until disease progression, a process known as active surveillance (AS). During this
time, patients may experience fatigue, weight loss, enlarged lymph nodes, and night sweats.
These physical symptoms have been linked to decreased quality of life among AS patients,
which is impaired relative to the general population (Shanafelt et al., 2007). We sought to
examine whether CLL symptoms impact quality of life via an increase in depressive
symptoms. Further, as social support has been linked to psychological and physical outcomes
in patients with CLL (Morrison et al., 2016), we wanted to examine whether social support
may moderate the pathway from CLL symptoms to depressive symptoms.

Methods: A cross-sectional design was used. AS CLL patients (N=67; 55% male, mean
age=60.4 years) completed questionnaires assessing CLL symptoms, quality of life (SF-36
MCS), depressive symptoms (CESD), and social support (ISEL total score). Mediation,
moderation, and moderated mediation were conducted using conditional process analysis
(Hayes, 2014). Analyses controlled for gender.

Results: Quality of life scores were variable across the sample (range=9.43-62.14 [possible 0-
100]). The direct effect of depressive symptoms on quality of life was significant (b = -.654,
95% CI = [-.935, -.373]), such that a one-unit increase in depressive symptoms was associated
with a .654 unit decrease in quality of life. The effect of CLL symptoms on depressive
symptoms was significant only at low and moderate levels of social support. CLL symptoms
did not influence depressive symptoms and quality of life among patients with high levels of
social support, reflecting a protective effect.

Conclusions: CLL patients with high levels of social support experience lower levels of
depressive symptoms in response to CLL symptoms. This protective effect is associated with
enhanced quality of life for CLL patients with high levels of support. Psychosocial
S2220 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

interventions with CLL patients focused on improving quality of life should take into
consideration patients level of social support.

CORRESPONDING AUTHOR: Travis Westbrook, M.A., The Ohio State University, Columbus, OH,
43220; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2221

C255 6:00 PM-7:00 PM

TRAUMA HISTORY, RELIGIOUS SOCIAL SUPPORT & RESILIENCE: CORRELATES OF ROLE

LIMITATIONS IN OLDER GAY MEN AND LESBIANS

Sarah M. White, N/A

University of North Texas, Wylie, TX

Trauma History, Religious Social Support & Resilience: Correlates of Role Limitations due to
Emotional Distress in Older Gay Men and Lesbians

Sarah White, Tosha Griggs & Mark Vosvick, Ph.D.

University of North Texas

Older gay men and lesbians (LG) with a history of trauma may experience role limitations due
to emotional distress, especially if the trauma is a violent crime (Min, Tracy & Park, 2014;
Herek, Gillis & Cogan, 1999). Resilience is the result of personal growth in spite of trauma
history (Lyons, 2015). Historically, churches have condemned sexual minorities (Halkitis et al.,
2009). However, some churches welcome LGBT individuals and older LGs may benefit from
the social support of these establishments (Halkitis et al., 2009). Through a minority stress
(Meyer, 2003) lens, we hypothesize crime related trauma history is positively associated with
role limitations due to emotional distress. Religious social support and resilience are
negatively associated with role limitations due to emotional distress.

After IRB approval, we recruited 50 gay men and 50 lesbians in DFW, (Mage = 58.99, SD= 6.48).
Participants were 68% European American, 17% African American and 15% other ethnicity. A
hierarchical regression analysis (F[9,90]=12.38, p < .001) indicated, in addition to crime related
trauma (=.16, p < .05) and resilience (=.40, p < .001), being single (=.40, p < .001) was also
a significant predictor, accounting for approximately 51% of the variance in role limitations
due to emotional distress (adjusted R2=.51, p < 001).

In our sample, single LGs were more likely to experience role limitations due to mental
distress possibly from lack of emotional support from a romantic partner (Lyons, 2016). Older
LGs that scored higher on crime related trauma reported higher role limitations due to mental
distress this may suggest a lasting emotional impact from the crime related trauma (Herek,
Gillis & Cogan, 1999). Also, resilience was negatively associated with role limitations due to
emotional distress possibly due to the strengthening of effective coping strategies (Morrow,
S2222 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

2001). Interventions should include techniques to strengthen resilience and provide

emotional support, while addressing the unique issues of single older LGs.

CORRESPONDING AUTHOR: Sarah M. White, N/A, University of North Texas, Wylie, TX, 75098;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2223

C256 6:00 PM-7:00 PM

ASIAN ETHNIC GROUPS PERCEPTIONS AND UTILIZATION OF HEALTHCARE IN THE U.S.: A

COMMUNITY-BASED PHENOMENOLOGICAL APPROACH

Jennifer Bordon, MA, Hung-Bin Sheu, PhD, Angela Lui, MA

University at Albany - SUNY, Albany, NY

Asian populations in the U.S. are underserved and understudied in health disparities research.
Specifically, little is known about the diverse perceptions of the American healthcare system
and the various statuses of mental and physical health within and among Asian ethnic groups.
Thus, with a community-based sample, this study was aimed to investigate the particular
experiences three Asian ethnic groups (Chinese, Filipino, and Korean) have with utilizing the
American healthcare system and the respective groups general health maintenance
approaches. The sample included six focus groups (Ntotal = 33; male = 11; mean age = 47
years). Using the phenomenological approach, transcripts were analyzed to identify emergent
themes. Findings show different and similar perceptions of the American healthcare system
and general healthcare within and across the three ethnic communities. All groups utilized
both American healthcare and indigenous methods to a degree to treat and maintain health.
Some shared themes include an agreement of major barriers to accessing American
healthcare (e.g., understanding insurance, difficulties navigating the healthcare systems),
satisfaction and reliance on American healthcare for acute and chronic conditions, and using
indigenous methods to treat and maintain health. Some themes unique to different ethnic
communities include self-advocacy and being proactive for Korean and Filipino groups, the
importance of patient-doctor communication for Chinese groups, and the emphasis of social
support to treat and maintain health for the Filipino groups exclusively. Findings help
illuminate the perceptions of the barriers and supports that may help inform policies and
outreach programs to promote access and utilization of American healthcare for the
understudied Asian populations. Considering the community-based approach, results are
useful to better understand the unique needs from these different ethnic groups, and to
further the development of measurement tools for health disparities research.

CORRESPONDING AUTHOR: Jennifer Bordon, MA, University at Albany - SUNY, Albany, NY,
12208; [email protected]
S2224 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C257 6:00 PM-7:00 PM

COPING STRATEGIES AMONG OLDER LATINO CAREGIVERS

Guilherme Balbim, MS, Isabela Marques, BS, Melissa Magallanes, MS, Judith Rocha, LCSW,
David Marquez, PhD

University of Illinois at Chicago, Chicago, IL

The number of Americans living with Alzheimer's disease (AD) is growing fast. An estimated
5.4 million Americans of all ages have AD in 2016. Latinos are the fastest-growing population
in the United States and are about 1.5 more times more likely to develop AD than Whites.
Caregivers of people with AD report high levels of stress and burden, impacting their health.
The purpose of this study was to analyze the coping strategies used by Latino caregivers.
Participants were 16 Latino caregivers, aged 50 years and older, who had been caring for a
relative with AD or related disorder (6 were caring for mother, 4 caring for husband), at least 4
hours per day, for at least 6 months. The majority of caregivers were female (12). Interviews
were conducted in English and Spanish about the caregiver role, family, and social support,
and coping strategies. For purpose of this study coping strategies answers were analyzed.
Interviews were transcribed, translated, coded, and directed content analysis was conducted.
The results depicted that Latino caregivers reported use of nine different coping strategies: (1)
Rationalize: efforts to explain or justify care recipients misbehaviors to plausible reasons, as
the disease manifesting itself; (2) Avoid: refraining from noticing or recognizing stressful
situations; (3) Repress: attempts of keeping away from burden-suppressing desires, feelings,
actions, tears; (4) Social interactions: attempts to interact with individuals or groups about
their situation; (5) Self-care: intentional actions to take care of ones own physical, mental and
emotional health;(6) Spirituality: focus on faith (God, church) and the things to try to make a
connection with spirituality (prayer, sacraments); (7) Leisure activities: actions in free time, far
from work and duties, to relieve burdens, such as reading, watching TV and movies, listening
to music, and dancing; (8) Medications: taking any form of medication to cope with
stress/burden; (9) Physical activity, any body movement that works ones muscles and
requires more energy than resting to cope with stress, such as walking, running and dancing.
Also was reported the use of any coping strategy, showing no efforts to solve stressful
situations.It is concluded that Latinos caregivers seek a broad range of strategies to cope with
stress and burden. Future studies should focus on strategies to highlight sustainable coping
strategies in their routine as caregivers.

CORRESPONDING AUTHOR: Guilherme Balbim, MS, University of Illinois at Chicago, Chicago,

IL, 60612; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2225

C258 6:00 PM-7:00 PM

DEVELOPMENT OF A CULTURALLY MODIFIED BEHAVIORAL WEIGHT LOSS MANUAL FOR OBESE

AND OVERWEIGHT LATINA WOMEN WITH EATING DISORDERS.

Phoutdavone Phimphasone, M.A.1, Fary M. Cachelin, PhD2, Alyssa Vela, M.A.2, Brooke Palmer,
M.A.2

1
University of North Carolina, Charlotte, New Haven, CT; 2University of North Carolina
Charlotte, Charlotte, NC

Obesity remains one of the greatest public health threats in the 21 st century, increasing the
risk for many chronic diseases, including Type 2 diabetes and CVD (CDC, 2014). Economic
costs associated with treatments for obesity are also significantly high, with total medical
costs reaching $147 billion/year in 2008 for inpatient and outpatient treatments and
prescription drugs. Despite an increase in prevention and intervention efforts to address
obesity in the past decade, the rates are still disproportionate among women and
racial/ethnic minority groups. In particular, Latina women who are obese/overweight with
eating disorders are less likely to seek and receive healthcare services and have a higher risk
of type 2 diabetes and CVD. Therefore, there is still a need to explore sociocultural factors
that influence Latina womens engagement in evidence-based treatment for obesity, as
previous research demonstrates that culturally adapted interventions are four times more
effective than general interventions.
The aim of this study was to identify culturally appropriate modifications of an evidenced-
based behavioral weight loss program, The LEARN Program for Weight Management,10th
Edition (LEARN; Brownell, 2004), using focus groups and qualitative analysis. Through constant
comparative analysis, participants (n = 20, MAGE = 30, SD = 9.87, MBMI = 34.40, SD = 6.64,
Mbinge eating episodes= 4, SD = 4.92) identified surface and deep level changes needed for the
LEARN manual. Surface level changes included: redundancy of manual, positive remarks and
critiques of manual content, need for culturally relevant examples, and structure of food logs.
Deep level changes included a consideration of diversity factors, Latino health beliefs and
expectations, environmental facilitators and barriers, cultural meanings of food and health
behaviors, and unique individual factors (e.g., family dynamics, past experiences with health
monitoring and chronic illness management, and responsibilities as a caregiver).
With cultural modifications, participants reported that the LEARN manual could be an
effective and feasible method to address obesity and eating disorders. Suggestions to improve
adherence to such programs are offered. In order to address health disparities and reduce the
prevalence rates of obesity among Latina women, it is crucial that treatments emphasize a
S2226 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

healthy lifestyle change approach while considering key cultural factors to improve reach and
effectiveness.

CORRESPONDING AUTHOR: Phoutdavone Phimphasone, M.A., University of North Carolina,

Charlotte, New Haven, CT, 06511; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2227

C259 6:00 PM-7:00 PM

LATINA ADOLESCENTS' GENERATIONAL STATUS, FAMILISMO, SOCIO-EMOTIONAL

COMPETENCE, AND NEIGHBORHOOD QUALITY IN RISK BEHAVIORS

Patricia Cabral, M.A.1, Jan Wallander, PhD2

1
University of California, Merced, North Hollywood, CA; 2Univ. California, Merced, Los Gatos,
CA

Objective: We tested a longitudinal model among Latina (female) adolescents examining

resilience related to sexual risk behaviors and substance use associated with familismo in the
context of neighborhood quality, as mediated bysocial-emotional competence (SEC), and how
these associations vary with acculturation.

Method: Using data from 896 Latina adolescent participants from the Healthy Passages study,
a multi-group structural equation model was analyzed across 1st (19%), 2nd (55%), and 3rd
generation (26%) Latinas. Latent constructs representing neighborhood quality and familismo
values were measured at 5th grade (M age = 10.6), SEC at 7th grade (M age = 13.1), and sexual
risk behaviors and substance use at 10th grade (M age = 16.1).

Results: A partially invariant multi-group structural equation model across the three
generational status groups showed that, among 2nd generation Latina adolescents only,
familismo mediated the link between neighborhood quality and substance use. Moreover, SEC
mediated the resilience provided by familismo against substance use. Thus, both familismo
and SEC were negatively associated with substance use among 2nd generation Latina
adolescents. These associations were not found among 1st or 3rd generation Latinas, nor for
sexual risk behaviors.

Conclusions: Findings illuminate the role of social-cultural factors that may support resilience
against certain health risk behaviors. Specifically, among 2nd generation Latinas, familismo
and SEC appear to confer resilience against substance use. These findings may guide
strategies in reducing substance use among Latina youth.

CORRESPONDING AUTHOR: Patricia Cabral, M.A., University of California, Merced, North

Hollywood, CA, 91605; [email protected]
S2228 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C260 6:00 PM-7:00 PM

MINORITY BREASTFEEDING DURATION: RESULTS OF THE BREASTFEEDING OPINIONS,

OUTCOMES, BEHAVIORS, & SERVICES STUDY

Sarah Rhoades-Kerswill, Ph.D., Emily D. Fry, Undergraduate Student, Misty AW Hawkins, Ph.D.

Oklahoma State University, Stillwater, OK

Background: The American Academy of Pediatrics recommends breastfeeding infants for at

least 12 months post-partum. However, many factors impact a womans intent and ability to
breastfeed for the recommended timeframe, such as obesity, delivery method, the delivery
term, age, and race. The current study focused on breastfeeding duration among minority
women.

Objective: This study sought to further explore breastfeeding duration among a diverse
sample of minority women who had their first child between 1985 and 2016. We
hypothesized that the race of women, the delivery method, and obese status during
pregnancy would predict breastfeeding duration of the first child after adjusting for age and
delivery term.

Methods: The sample included 313 minority women who self-identified as African American,
American Indian/Alaska Native (AI/AN), Asian, Indian, Latina, Middle Eastern, Pacific Islander,
multi-racial, and other. Their ages ranged from 20-59 years-old. Snowball sampling was used
to recruit these participants who completed an online survey of demographic and
breastfeeding behaviors: The Breastfeeding Opinions, Outcomes, Behaviors, and Services
(BOOBS study).

Results: Forty-four percent of the participants reported that they did not meet the
breastfeeding recommendation of at least 12 months. Age of the mother was found to predict
breastfeeding duration ( = .15, SE = .07 p = .01) with women aged 35+ breastfeeding longer
(13 months) than younger women (11 months) on average. Obesity status and delivery
method did not predict breastfeeding duration. However, being AI/AN did predict
breastfeeding duration ( = -.23, SE = 1.78, p < .05). AI/AN women tended to have the shortest
breastfeeding duration (on average 9 months and 53% reported breastfeeding less than 12
months). No other race was found to influence a womans breastfeeding duration.

Conclusions: It appears that belonging to certain racial groups, particularly American

Indian/Alaskan Native may impact how long women breastfeed even when controlling for
age and delivery term. There may be cultural and other factors (such as socioeconomic status)
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2229

that influence an AI/AN womans breastfeeding duration. Given the extensive benefits of
breastfeeding for both mothers and infants, culturally sensitive interventions may be needed
to enhance the outcomes of certain age and race-ethnic groups.

CORRESPONDING AUTHOR: Sarah Rhoades-Kerswill, Ph.D., Oklahoma State University,

Stillwater, OK, 74075; [email protected]
S2230 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C261 6:00 PM-7:00 PM

RACIAL/ETHNIC DIFFERENCES IN WEIGHT LOSS STRATEGIES AMONG U.S. ADULTS

Becky Marquez, PhD MPH1, Rosenda Murillo, PhD MPH2

1
University of California San Diego, La Jolla, CA; 2University of Houston, Houston, TX

Background: Dieting, exercising, and seeking professional help have been associated with
intentional weight loss among adults.

Purpose: We examined the use of diet (e.g. ate less, ate less fat, switched to low calorie
foods), exercise, diet and exercise, and professional help (e.g. weight loss program, prescribed
diet pills) for weight loss among non-Hispanic Whites, Mexican-Americans, and non-Hispanic
Blacks trying to lose weight.

Methods: We used cross-sectional data from 7,424 male and female participants in the 2007-
2010 National Health and Nutrition Examination Survey (NHANES) aged 20-65 years. Logistic
regression models were used to estimate associations of race/ethnicity with strategies to lose
weight.

Results: Lower proportions of Mexican-Americans (34%) and non-Hispanic Blacks (35%) than
non-Hispanic Whites (39%) reported trying to lose weight. Among those who tried to lose
weight, non-Hispanic Blacks (OR: 0.78, 95% CI: 0.66-0.92) were less likely than non-Hispanic
Whites to use diet for weight loss. Mexican-Americans (OR: 0.66, 95% CI: 0.47-0.94) and non-
Hispanic Blacks (OR: 0.73, 95% CI: 0.56-0.95) were also less likely than non-Hispanic Whites to
use professional help for weight loss.

Conclusion: Targeted efforts are needed to address racial/ethnic disparities in weight loss
attempts and use of recommended strategies.

CORRESPONDING AUTHOR: Becky Marquez, PhD MPH, University of California San Diego, La
Jolla, CA, 920393; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2231

C262 6:00 PM-7:00 PM

THE EFFECTS OF EXPRESSIVE WRITING ON HEALTHCARE UTILIZATION, PHYSICAL SYMPTOMS,

AND PSYCHOLOGICAL DISTRESS IN ASIAN INDIANS

Jagruti Bhakta, PhD

Naval Center for Combat & Operational Stress Control (NCCOSC), San Diego, CA

An accumulating body of evidence has shown that expression of emotions associated with
stressful experiences by writing or talking about them results in better physical health and
psychological well being. Traditionally, Asian Indians encourage the suppression of emotional
conflicts and discourage the full expression of emotions. Additionally, Asian Indians are often
reluctant to talk about their personal problems to others, as they fear it will bring shame to
the family. Therefore, written emotional disclosure may provide a more culturally sensitive
way for Asian Indians to express emotions about stressful events and traumas without being
stigmatized, as it assures privacy and anonymity.

This randomized, controlled study examined the effects of a written emotional expression
intervention on healthcare utilization, physical symptoms, and psychological distress in Asian
Indians who reported stress related to intergenerational family conflict. Seventy participants
were recruited from Asian Indian community centers, professional and student organizations,
and community temples in Southern California. Thirty-eight women and thirty-two men were
randomly assigned to either write about their deepest thoughts and feelings related to their
intergenerational family conflict (n = 35) or to a time-management control group (n = 35).
Participants wrote on three separate days over one week and assessments were completed at
baseline, one-week post-writing (post), and six-weeks after writing (follow-up).

A significant group by time interaction (p = .002) was revealed on the depression subscale of
the Brief Symptom Inventory (BSI). On this measure of psychological distress, participants in
the expressive writing group showed improvement over time while participants in the control
group remained the same from baseline to post-writing and follow-up. A significant effect of
time (p < .0005) where both groups showed improvement from baseline to post-writing and
follow-up was revealed on three global indices of psychological distress on the BSI, the anxiety
subscale of the BSI, and on the Impact of Events Scale (IES). There were no significant findings
revealed on outcome measures of physical symptoms, healthcare utilization, or on
exploratory variables of acculturation and ambivalence over emotional expression.

The results from this study imply that a brief expressive writing intervention may help
improve symptoms of depression in Asian Indians. This was one of the first studies to
S2232 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

systematically investigate the effects of expressive writing in a homogenous sample

comprised of Asian Indians. While differences observed between groups were suggestive,
they were not conclusive and require further examination. Future research recommendations
include the use of more stringent inclusion criteria, objective measures, long-term follow-up
assessments, and essay content analyses.

CORRESPONDING AUTHOR: Jagruti Bhakta, PhD, Naval Center for Combat & Operational
Stress Control (NCCOSC), San Diego, CA, 92131; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2233

C263 6:00 PM-7:00 PM

EVALUATION OF PATIENT OUTCOMES FOLLOWING ATTENDANCE AT A SIX-WEEK

BIOPSYCHOSOCIAL MULTIDISCIPLINARY SEXUAL WELLNESS PROGRAM

Patricia Mumby, Ph.D.1, Mary Lynn, D.O.2, Michael Hakimi, Psy.D.1, Corey Bolton, B.S.3

1
Department of Psychiatry and Behavioral Neurosciences, Loyola University Medical Center,
Maywood, IL; 2Department of Obstetrics and Gynecology, Loyola University Medical Center,
Maywood, IL; 3Department of Psychiatry and Behavioral Neurosciences, Loyola University
Medical Center, Oak Park, IL

BACKGROUND: The purpose of the present study is to prospectively assess sexual functioning
and relationship satisfaction among patients who attend a multidisciplinary program for the
diagnosis and treatment of sexual dysfunction. There is some research to suggest that
women benefit more than men from sexual therapy programs and therefore, additional
analyses were conducted to examine for gender differences in response to
treatment. METHODS: Subjects include 21 men and 23 women who attended the sexual
treatment program. Patients were 18 years of age and older, in a committed relationship, and
attended the program with their partner. The program takes a biopsychosocial approach to
the assessment and treatment of sexual dysfunction and includes a multidisciplinary team of
gynecology, urology, psychology, psychiatry, physical therapy and nutrition specialists.
Couples meet for 6 consecutive weeks with a therapist dyad and receive a psycho-social-
sexual assessment, physical examination (i.e., sexological exam) and weekly cognitive
behavioral sex therapy and psychoeducational lectures. Patients completed the following
validated self-report measures pre- and post-treatment (pre-tx, post-tx): Patient Health
Questionnaire-9 (PHQ-9; pre-tx only), PROMIS Sexual Function Profile v1.0, Dyadic Adjustment
Scale (DAS), and the International Index of Erectile Function (IIEF). RESULTS: Paired-samples t
test were conducted to evaluate pre-tx to post-tx sexual function/satisfaction, relationship
adjustment, and erectile function. The results indicated that sexual function/satisfaction
(t(34) = -3.259, p < .01) and relationship adjustment (t(38) = -2.641, p < .05) significantly
improved pre-tx to post-tx. However, when analyzed by gender, females showed greater
improvement in scores on the PROMIS (Females: t(18)= -2.650, p < .05; Males: t(15) = -1.905,
p = .076) and the DAS (Females: t(19) = -2.925, p < .01; Males: t(18) = -1.111, p = .281). There
was no significant change reported for erectile functioning pre-tx to post-tx. CONCLUSION:
The study found statistically significant improvement pre-tx to post-tx on select aspects of
sexual well-being for attendees of the program, particularly for women participants. The
findings provide preliminary support for the efficacy of this multidisciplinary program and also
highlight the need to further our understanding of gender differences. Future research should
S2234 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

evaluate the causes of the gender differences in treatment response with the ultimate goal of
improving sexual treatment and outcomes for men and women.

CORRESPONDING AUTHOR: Patricia Mumby, Ph.D., Department of Psychiatry and Behavioral

Neurosciences, Loyola University Medical Center, Maywood, IL, 60153; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2235

C264 6:00 PM-7:00 PM

A TECHNOLOGY ASSISTED BEHAVIORAL INTERVENTION FOR EXTENDING SLEEP DURATION

AMONG SHORT SLEEPERS

Kelly Glazer. Baron, PhD, MPH, CBSM1, Jennifer Duffecy, PhD2, Kathryn J. Reid, PhD3, Lauren
A. Caccamo, BS4

1
Rush University Medical Center, Chicago, IL; 2University of Ilinois at Chicago, Chicago, IL;
3
Northwestern University Feinberg School of Medicine, Chicago, IL; 4Northwestern University,
Chicago, IL

It is estimated that 29% of adults sleep < 6 hours and sleep duration may be a modifiable risk
factor for the development of chronic illnesses, including hypertension and diabetes. Studies
demonstrate health benefits of extending sleep duration among short sleepers but to date
there are no theory based behavioral interventions. The rapidly expanding popularity of
wearable sleep tracking devices provides an opportunity to engage users in their own sleep
behavior change. We developed at tested the Sleep Bunny sleep extension intervention
including a smartphone application, wearable sleep tracker and brief telephone coaching. The
goals of this program were to increase sleep duration by one hour and increase bedtime and
wake time regularity. We conducted user testing and collected feedback from 10 participants
and evaluated adherence and sleep duration outcomes of two participants who completed
the four-week intervention. Results suggested that users enjoyed the use of the wearable
sleep tracker and found the application visually pleasing. Sleep duration improved by 1.5
hours per night from week 1 to week 4 in adults with short sleep duration who engaged in the
intervention (n=2). The brief telephone coaching was viewed as helpful and feasible.
Improvements in notifications and reminders may improve the experience of the intervention.

CORRESPONDING AUTHOR: Kelly Glazer. Baron, PhD, MPH, CBSM, Rush University Medical
Center, Chicago, IL, 606012-3833; [email protected]
S2236 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C265 6:00 PM-7:00 PM

BEDTIME E-DEVICE USE AND SLEEP QUALITY AMONG YOUNG ADULTS: USING THE THEORY OF
PLANNED BEHAVIOR

Xiaohui Zhao, MS1, Xue Feng, MPH2, rahul garg, M Pharm3, Kimberly M. Kelly, PhD, MS2

1
School of Pharmacy, West Virginia University, Morgantown, WV; 2West Virginia University,
Morgantown, WV; 3west virginia university, morgantown, WV

Background: Understanding the factors affecting bedtime e-device use is important to

improve the sleep quality of young adults in the US. Objectives: Our study examined: (1) the
association of attitudes, subjective norms, and perceived behavioral control with bedtime e-
device use and intentions to reduce e-device use; and (2) the efficacy of a brief theory-driven
intervention to increase the intentions to reduce e-device use. Methods: We used the Theory
of Planned Behavior (TPB) to conduct a cross-sectional online survey among young adults
aged 18-30 years (N=212). We designed two scales to measure attitudes and perceived
behavioral control and used a single item to measure subjective norms for bedtime e-device
use. After feasibility testing (N=53), we randomly administered the intervention (N=159) to
affect participants attitudes, subjective norms, and perceived behavioral control and
analyzed the participants intentions to reduce bedtime e-device use before and after the
intervention. We used Spearman's correlation for bivariate and logistic regression for
multivariate analyses. Results: The scales for attitudes and perceived behavioral control had
good internal consistency (Cronbachs >0.65). Participants mean age was 24.9 (SD=3.0), and
majority were females (77.3%), White (91%), had > high school degree (98.6%), and annual
household income > $24,300 (90.5%). The majority (73.3%) reported frequent (>10 days)
bedtime e-device use in the past two weeks and more than half (59.1%) spent >30 min on e-
device daily. We found relatively high levels of behavioral control (3.820.72) and subjective
norms (3.750.94) for bedtime e-device use. Higher positive attitude and lower perceived
behavioral control were significantly associated with higher e-device use and higher intentions
to reduce e-device use. Participants receiving the intervention were more likely to increase in
their intentions to reduce bedtime e-device use (AOR=2.23, 95%CI= [1.09-4.57]). Conclusion:
Our brief educational intervention was efficacious in increasing the intentions to reduce
bedtime e-device use among young adults. Future studies are recommended to use the TPB
based interventions to reduce bedtime e-device use among young adults.

CORRESPONDING AUTHOR: Xiaohui Zhao, MS, School of Pharmacy, West Virginia University,
Morgantown, WV, 26506; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2237

C266 6:00 PM-7:00 PM

COLLEGE SLEEP HABITS: SLEEP QUALITY & DISTRESS DURING THE FIRST SEMESTER

Lauren Boddy, BA1, Cora Burgoyne, BA anticipated May 20172, Natasia Adams, MA & MPH 1,
Nancy A. Hamilton, PhD1

1
University of Kansas, Lawrence, KS; 2The University of Kansas, Lawrence, KS

Poor sleep quality is highly prevalent and has been linked with poor health outcomes in all age
groups. It is known that college students have poor sleep habits; however, less is known about
changes in sleep quality across the first semester for freshmen students. The purpose of this
study was to examine changes in self-reported sleep quality, as well as biopsychosocial
correlates of disturbed sleep during the first semester of college in a sample of female,
freshmen students. We hypothesized that mood and stress symptoms, social support, specific
health indicators, and hostility would co-vary with poor sleep quality at multiple time points.
Freshmen students, enrolled at large, Midwestern university (N=76, 77% Caucasian), were
assessed during the first month (T1), at midterm (T2), and during the last two weeks of the
Fall semester (T3). In this sample, average Pittsburgh Sleep Quality Index (PSQI) scores at T1
(M=8.47, SD=3.11), T2 (M=8.92, SD=3.25), and T3 (M=9.62, SD=3.44) exceeded a cut off score
of PSQI5 at all time points. Moreover, results demonstrate a linear increase in sleep
disturbances across the semester (F(1,48)=4.25, pr=.41), perceived stress (Perceived Stress
Scale; r=.38), and hostility (Cook-Medley Hostility Scale; r=-.74) co-varied with poor sleep
quality. At T2, perceived stress (r=.39), stressful life events (Student Life Event Scale; r=.29),
somatic and physical symptoms (Cohen Hoberman Inventory of Physical Symptoms; r=.51),
and social support (Interpersonal Support Evaluation List; r=-.52) significantly co-varied with
PSQI scores. Finally, at T3, perceived stress (r=.49), stressful life events (r=.28), somatic and
physical symptoms (r=.53), and social support (r=-.53) were significantly associated with PSQI
scores. Evidence for a prospective, longitudinal relationships was also found, such that poor
sleep quality at T2 predicted physical and somatic symptoms at T3, controlling for T2 physical
symptoms (=1.28, p < .05). The results of this study demonstrate that sleep becomes more
disrupted across the first semester of college and it co-varies with multiple indices of
psychological distress. Further research is needed to determine the predictive relationships
between these risk factors and sleep quality for female, freshmen college students.

CORRESPONDING AUTHOR: Lauren Boddy, BA, University of Kansas, Lawrence, KS, 66044;
[email protected]
S2238 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C267 6:00 PM-7:00 PM

EFFECTS OF SCHOOL START TIME ON STUDENTS SLEEP: A META-ANALYSIS

Jennifer Bowers, B.A.1, Anne Moyer, Ph.D.2

1
Stony Brook University, Port Jefferson, NY; 2Stony Brook University, Stony Brook, NY

Introduction

Research finds that few adolescents adhere to recommended sleep guidelines. Lack of sleep is
a predictor of a number of serious physical and mental health consequences. This meta-
analysis examines the effects of school start time (SST) on sleep duration of students by
aggregating the results of 5 longitudinal studies, and separately, 14 cross-sectional studies.

Methods

Database searches and manual searches for citations in recently published reviews of the
literature were used to identify 19 records for inclusion. Data for the following moderator
variables were extracted and coded: publication status of article, delay of SST in minutes,
students grade in school, and months between Time 1 and Time 2.

Analyses were conducted using Version 2 of Comprehensive Meta-Analysis (CMA) software.

Once all study effect sizes were calculated, CMA produced an aggregate effect size using
weights and the random effects model. A positive effect size represents longer sleep
durations for the later SST group.

Results

Longitudinal studies

The effect sizes (d) for sleep duration ranged from 0.24 to 1.07, with an aggregate effect size
of d = 0.38, 95% CI 0.20 0.57, p < 0.001. Significant heterogeneity was found among the
effect sizes, Q(4) = 9.99, p < 0.05, I2 = 59.97.

Meta-regression was used to analyze moderator variables. Longer delay predicted longer
sleep duration, Q(1) = 7.56, p < 0.01. The length of time between T1 and T2 was not found to
be significant, Q(1) = 0.61, p = 0.435.

Cross-sectional Studies
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2239

The effect sizes (d) for sleep duration ranged from 0.05 to 0.96, with an aggregate effect size
of d = 0.41, 95% CI 0.25 0.57, p < 0.001. Significant heterogeneity was found among the
effect sizes, Q(13) = 323.83, p < 0.01, I2 = 95.99.

Grade in school and publication status moderators were analyzed using analog-to-ANOVA.
Longer delay predicted shorter sleep duration, Q(1) = 63.36, p < 0.01. Neither grade in school,
Q(2) = 2.04, p = 0.90, nor publication status was significant, Q(1) = 1.72, p = 0.19.

Discussion

Results indicate that later starting school times are associated with longer sleep durations.
Further investigation is needed to understand the discrepancy in directionality between cross-
sectional and longitudinal studies for the delay length moderator. Although more longitudinal
primary research is needed and there was unexplained heterogeneity in the cross-sectional
studies, overall, this systematic analysis of SST studies suggests that delaying SST is beneficial
for students sleep, and thus their general well-being.

CORRESPONDING AUTHOR: Jennifer Bowers, B.A., Stony Brook University, Port Jefferson, NY,
11777; [email protected]
S2240 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C268 6:00 PM-7:00 PM

ETHNIC/RACIAL DIFFERENCES IN SLEEP DURATION, SLEEP QUALITY AND HEALTH

Sheena Mirpuri, PhD

Memorial Sloan Kettering Cancer Center, New York, NY

Ethnic/racial health inequalities are well-established (Williams et al., 2007), and research has
more recently suggested the role of sleep in mental and physical health outcomes (e.g.,
Altman et al., 2012). Using a subsample from the American Time Use Survey dataset, a large,
nationally representative sample (n=36,942 participants, Mage=47.63, 15-85 years old, 55.7%
female), the current study investigates sleep and health disparities based on ethnicity/race
and the relationships between sleep indicators and health outcomes. The majority of
participants were White (65.8%), followed by Black (14.6%), Hispanic (14.2%), and Asian
(3.7%) participants. Other/Mixed Ethnicity/Race (1.7%) were removed for ease of
interpretation.

Addressing the first goal, significant ethnic/racial differences in sleep duration (i.e., the
longest period of sleep over 24 hours) were found. White individuals reported the lowest
sleep duration in minutes (M=523.60) in comparison to Hispanic (M=552.35), Black
(M=544.53), and Asian (M=536.27) individuals. A larger proportion of White individuals
reported obtaining the recommended amount of sleep of 7-9 hours (43%) in comparison to
Asian (37.5%), Hispanic (34.2%), and Black participants (32.5%). There were also significant
differences in sleep quality ("How well-rested did you feel yesterday? on a scale from 1-4) by
ethnic/racial group, such that Black (M=3.19) and Hispanic (M=3.19) individuals reported
feeling more well-rested than White (M=3.12) individuals. In terms of general health ("Would
you say your health in general is excellent, very good, good, fair, or poor?"), Asian (M=3.63)
and White (M=3.57) individuals reported significantly better health than Black (M=3.27) and
Hispanic (M=3.28) individuals. Almost half of Black participants reported high blood pressure
in the past five years (43%) in comparison to other groups. Addressing the second goal,
significant associations were found between general health and sleep quality (r=.24) and sleep
duration (r=-.10). Sleep quality and sleep duration were only slightly related (r=.07), although
the strength of this association differed by ethnic/racial group.

Our findings are consistent with prior research suggesting that sleep quality may be more
predictive of overall health in comparison to sleep quantity (Pilcher et al., 1997). This study
also highlights ethnic/racial differences in sleep and health indicators, emphasizing the need
for future work to consider these mechanisms.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2241

CORRESPONDING AUTHOR: Sheena Mirpuri, PhD, Memorial Sloan Kettering Cancer Center,
New York, NY, 10036; [email protected]
S2242 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C269 6:00 PM-7:00 PM

INSOMNIA, IMMUNITY AND INFECTIONS IN CANCER PATIENTS TREATED WITH

CHEMOTHERAPY

Sophie Ruel, Ph.D. student, B.Sc. Inf, B.A.1, Jose Savard, Ph.D.2, Hans Ivers, Ph.D.3

1
Laval University, Levis, PQ, Canada; 2Laval University, Qubec, PQ, Canada; 3Laval University,
Quebec, PQ, Canada

Background: Research has shown that insomnia symptoms, which are very common in cancer
patients, especially in those with breast (42-69%) or gynecological cancer (50-68%), influence
the immune system. However, the impact on health resulting from such immune changes has
rarely been studied, especially during chemotherapy, when the immune system is already
compromised. The aim of this study was to evaluate the relationships between insomnia,
immunity and the occurrence of infections in women treated with chemotherapy for breast or
gynecological cancer. It was hypothesised that patients with insomnia (INS) would present
more immune alterations during and after chemotherapy, in addition to reporting more
infections, as compared to good sleepers (GS).

Methods: 52 patients were assessed before chemotherapy (T1), on four occasions during the
first two cycles of chemotherapy (i.e., immunosuppression and recovery weeks; T2-T5), at the
end of the treatment (T6), and at 3- (T7) and 6-month (T8) follow-ups. Sleep was assessed
using the Insomnia Severity Index, a daily sleep diary and a clinical interview, which also
assessed the presence of infections. Finally, blood samples were collected at each time point
(T1-T8) to measure several immune parameters (e.g., neutrophils, lymphocytes).

Results: Linear mixed-model ANCOVAs revealed significant group-by-time interactions on

infection (ps.04) and immune variables (e.g., neutrophils, T-cells; ps.04). Moreover, INS
reported more infectious symptoms than GS at T2 (p=.004).

Conclusions: These results partially confirm our hypotheses and indicate that the risk of
infections and immune variables evolve differently in INS as compared to GS during and after
chemotherapy.

CORRESPONDING AUTHOR: Sophie Ruel, Ph.D. student, B.Sc. Inf, B.A., Laval University, Levis,
PQ, G6Z7S6; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2243

C270 6:00 PM-7:00 PM

RESTLESS SLEEP IS ASSOCIATED WITH WEIGHT GAIN: A WITHIN-SUBJECT ANALYSIS

Tom Quisel, BS1, John Torous, MD2, Luca Foschini, PhD3

1
Evidation Health, Santa Barbara, CA; 2Beth Israel Deaconess Medical Center /Harvard
Medical School, Brookline, MA; 3Evidation Health, San Mateo, CA

Background: The rising popularity of digital health trackers presents the opportunity to
understand the relationship between personal behaviors and health outcomes outside of the
lab or clinic. In this study, we use longitudinal weight data and minute-level sleep data from
over 6,000 users of digital health trackers across the United States to examine the relationship
between sleep patterns and a clinical outcome: weight gain.

Methods: We analyzed minute-level sleep and weight data collected from fitness trackers for
11,552 users of a proprietary platform that incentivizes healthier lifestyles (Achievemint.com)
in the period 4/1/20154/1/2016. We considered users with at least 5 weight measurements
spanning two or more months. Users with 5 or more nights of sleep recorded in at least two
calendar months spanned by their weight measurements were included. We modeled
monthly percent weight change for each user using fixed-effects panel regression analysis,
performing separate regressions for each gender. We included average nightly sleep start
time, sleep duration, number of naps during the day, time in bed until sleeping, number of
restless episodes, as well as the standard deviation of their time asleep during the month as
explanatory variables. We further included variables to control for seasonality and variation in
sleep/weight measurement device and frequency.

Results: 1,087 male and 5,207 female users met the inclusion criteria. The strongest effect in
the regression was an association between restless episodes and weight gain. Each additional
restless episode per night was associated with a 0.058 percentage point monthly increase in
weight for females (95% CI [0.046, 0.072]; p < .001) and a 0.052 percentage point increase for
males (95% CI [0.024, 0.081]; p < .001). On average, users had 7.8 3.2 restless episodes per
night.

Conclusions: There is a significant association between frequency of restless episodes and

weight gain in the population studied. By individually identifying those with sleep difficulties,
our results suggest it is feasible to offer personalized care that focuses on preventative risk
factors. Identifying poor sleep is the first step towards offering tailored sleep therapies such
as insomnia CBT, sleep hygiene, psychological, and medical care. Our results also represent a
S2244 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

scalable and accessible means of generating clinically actionable health data given that many
across the world increasingly own and use trackers and wearables.

CORRESPONDING AUTHOR: Tom Quisel, BS, Evidation Health, Santa Barbara, CA, 93101;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2245

C271 6:00 PM-7:00 PM

SINGLE ITEM MEASURES FOR THE DETECTION OF SLEEP PROBLEMS IN UNITED STATES
MILITARY VETERANS

Jaime M. Hughes, PhD, MPH, MSW1, Christi S. Ulmer, PhD2, Jennifer M. Gierisch, PhD, MPH3,
Mid-Atlantic VA MIRECC Workgroup, .4, Matthew O. Howard, PhD5

1
Center for Health Services Research in Primary Care, Durham VA Health Care System, Cary,
NC; 2Center for Health Services Research in Primary Care, Durham VA Health Care System,
Durham, NC; 3Duke Univeristy, Durham, NC; 4Durham VA Health Care System, Durham, NC;
5
UNC School of Social Work, Chapel Hill, NC

Chronic sleep problems may contribute to higher rates of healthcare utilization, incident
mental health disorders, and increased risk of suicide. Although sleep problems are reported
by as many as two-thirds of U.S. military Veterans who served in Iraq and Afghanistan, sleep
problems remain overlooked in clinical settings. To date, there are no brief sleep screeners or
clinical guidelines addressing how to best screen for sleep problems in Veterans Health
Administration (VHA) primary care settings. Using data from the Study of Post-Deployment
Mental Health, a volunteer-research registry housed at the Mid-Atlantic VAs Mental Illness
Research, Education, and Clinical Center, the objective of this research was to examine the
utility of single sleep items embedded in existing psychological assessments. Three such items
are embedded into the Symptom Checklist-90 and include distress related to trouble falling
asleep, awakening in the early morning and sleep that is restless or disturbed. Participants
were U.S military Veterans who served since September 11, 2001, had one or more
deployments, and did not meet past-month criteria for a DSM-defined Axis I psychiatric
disorder (N = 1,118). Participants completed a battery of questionnaires assessing
demographics, physical and mental health, and sleep. The performance of the three items
noted above was evaluated in relation to two reference outcomes: poor sleep (Pittsburgh
Sleep Quality Index total score > 5) and DSM-defined probable insomnia disorder. Using an
response score of one (a little bit distressing), each of the three items demonstrated
moderate sensitivity values and acceptable rates of each false positives (FP) and false
negatives (FN) in predicting poor sleep (sensitivity: 0.70 to 0.78; FP: 0.23 to 0.30; FN: 0.26 to
0.42) and probable insomnia (sensitivity: 0.81 to 0.89; FP: 0.42 to 0.48; FN: 0.11 to 0.32).
Although additional research is needed to test the performance of these items across diverse
Veteran groups, our initial findings suggest that existing items may serve as a first step in
screening for sleep problems.
S2246 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Jaime M. Hughes, PhD, MPH, MSW, Center for Health Services
Research in Primary Care, Durham VA Health Care System, Cary, NC, 27519;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2247

C272 6:00 PM-7:00 PM

THE EFFECT OF SCHOOL START TIMES ON SLEEP AND ACADEMIC PERFORMANCE AMONG
ELEMENTARY AND MIDDLE PUBLIC SCHOOL STUDENTS IN THE US

Tracy Trevorrow, Ph.D., Sakiko Sugihara, BA Psychology, Stephanie Lum, N/A

Chaminade University of Honolulu, Honolulu, HI

Most K-12 students in the US are not getting enough sleep. Early school start times appear to
contribute to this problem, resulting in unwanted outcomes in health, psychosocial
development and learning. Only 20% of US schools start after 8:30 a.m. as recommended by
the American Academy of Pediatrics.

The purpose of this study was to determine whether US school start times were associated
with academic performance. We hypothesized that later school start times would be
associated with improved academic performance. Only elementary and middle schools
students were included due to insufficient reporting of high school outcomes.

Participants included 4th and 8th grade public school students from 49 States for the 2011-12
academic year. Student data and school start times were obtained from the National Center
for Educational Statistics (NCES). A total of 50,310 elementary schools, 13,990 middle schools
and 33, 269,172 students were included. The National Assessment for Educational Progress
(NAEP) was used to assess academic proficiency for Grades 4 and 8 in mathematics, reading,
and science.

Pearson correlations were computed to determine associations between school start

times by state and academic proficiency. Simple linear regression was used to predict
academic proficiency based on school start times.

For Grade 4, school start times were significantly correlated with NAEP scores for math
proficiency, r2 = .48, p2 = .40, p2 = .50, p The data supported our hypothesis regarding
school start times and academic achievement for Grade 4 students but not for those in Grade
8. Elementary school students may be at particular risk for poorer academic achievement due
to sleep loss from early school start times. There is a need for studies that include sleep
assessment and allow for the evaluation of causal relationships between school start times
and academic achievement.
S2248 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Tracy Trevorrow, Ph.D., Chaminade University of Honolulu,

Honolulu, HI, 96822; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2249

C273 6:00 PM-7:00 PM

THE RELATIONSHIP BETWEEN SLEEP HEALTH AND WEIGHT LOSS HISTORY

Allison Engwall, BS, Michael Mead, BS, Leah A. Irish, Ph.D.

North Dakota State University, Fargo, ND

Obesity is a significant public health concern for U.S. adults, with approximately 27%
considered to be obese and an additional 35% considered to be overweight. Obesity is a
leading risk factor for a broad range of mental and physical health problems. Accordingly,
weight loss attempts are common, though these attempts typically yield limited long-term
success. Sleep is a crucial factor to consider in the context of weight loss attempts because
poor sleep may undermine positive changes to diet and exercise, reducing the likelihood of
successful weight loss. The present study aimed to evaluate the relationship between sleep
health and a history of weight loss attempts. Participants included 153 adults enrolled in a
commercial weight loss program who completed an online survey. On average, participants
had a history of 3.15 weight loss attempts (SD=4.98). The number of weight loss attempts was
negatively correlated with sleep health (r = -.26, p = .001), suggesting that better sleep health
was associated with fewer weight loss attempts. These findings support the hypothesis that
poor sleep may undermine weight loss attempts, resulting in lower success rates and
increased number of weight loss attempts over time. Moving forward, longitudinal research
would allow for improved understanding of the mechanisms underlying this relationship, and
additional consideration should be given to individual characteristics that may moderate
these effects (e.g., age, stress). A broader understanding of this relationship could inform
interventions to improve the chances of weight loss success, thus potentially reducing the
prevalence of obesity in todays society.

CORRESPONDING AUTHOR: Allison Engwall, BS, North Dakota State University, Fargo, ND,
58104; [email protected]
S2250 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C274 6:00 PM-7:00 PM

TREATING SLEEP COMPLAINTS: A STEPPED CARE APPROACH

Kelsey C. Utter, Ph.D.1, Delilah O. Noronha, Psy.D. 2

1
VA Palo Alto Healthcare System, Department of Veterans Affairs, Seaside, CA; 2Department
of Veterans Affairs, san jose, CA

The VA Palo Alto Healthcare System utilizes a stepped care approach for insomnia in its
Primary Care-Mental Health Integration (PCMHI) program, in order to increase access and
provide care that matches patient need. The model relies on treating patients under the least
restrictive methods and utilizing self-correction through outcome monitoring. The following is
a program evaluation that describes the PCMHI stepped protocol for insomnia, patient
utilization rates, and clinical outcome data, gathered over 12 months in two community
based outpatient clinics in the healthcare system. This stepped care protocol includes sleep
assessments, interdisciplinary collaboration, psychoeducation, web-based self-management
tools, shared medical appointments (SMAs), and Cognitive Behavioral Therapy for Insomnia
(CBTI). The steps provide interventions that are matched to the patients current need,
ability, and motivation. Self-report measures, including the Insomnia Severity Index (ISI) and
sleep diaries, were gathered throughout treatment to aid in planning.

During the 12 months, a total of 118 consults were received for sleep concerns; after initial
psychoeducation and consultation, 69% of patients were referred to a higher level of care in
the Sleep SMA (N=82). Twelve of these patients (14.6%) attended SMAs more than once, as
these are designed to be an open access tool for self-mangement. After completion of at
least one SMA, 7 patients were referred to CBTI as the next step of care. Analysis conducted
for patients who completed the CBTI protocol (N=6) indicated statistically significant
improvements in sleep efficiency (t(5)=-2.94, p< .05), improvements in total sleep time (t(5)=-
3.3, p < .05), and improvements in sleep latency (t(5)=2.76, p < .05). There was no significant
change in time awake after sleep onset nor ISI scores. When 2 patients who elected to
terminate CBTI due to pre-existing conditions (PTSD and sleep apnea) were removed from the
data, a statistically significant decrease in ISI scores (t(3)=3.37, p < .05) was also found.
Average length of CBTI treatment was 9.33 sessions.

The program evaluation highlights that a stepped care approach is useful in PCMHI for
treatment of insomnia, allowing treating providers to customize care based on acuity of need.
Outcome analysis demonstrates that patients experience overall improvement in sleep
functioning at various levels. Future suggestions for improvement of the model will be
discussed.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2251

CORRESPONDING AUTHOR: Kelsey C. Utter, Ph.D., VA Palo Alto Healthcare System,

Department of Veterans Affairs, Seaside, CA, 93955; [email protected]
S2252 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C275 6:00 PM-7:00 PM

ENVIRONMENTAL FACTORS, HEALTH BEHAVIORS, AND OBESITY: FINDINGS FROM THE

HISPANIC COMMUNITY HEALTH STUDY/STUDY OF LATINO YOUTH

Alan M. Delamater, PhD1, Ashley Marchante, M.S.2, Elizabeth Pulgaron, PhD2, Krista Perreira,
PhD3, Carmen R. Isasi, MD, PhD4, Mercedes Carnethon, Ph.D.5, Guadalupe X. Ayala, PhD,
MPH6, Linda C. Gallo, PhD7, Linda Van Horn, PhD, RD5, Shrikant I. Bangdiwala, PhD3

1
University of Miami Miller School of Medicine, Miami, FL; 2University of Miami, Miami, FL;
3
University of North Carolina, Chapel Hill, NC; 4Albert Einstein College of Medicine, Bronx, NY;
5
Northwestern University, Chicago, IL; 6San Diego State University, San Diego, CA; 7San Diego
State University, Chula Vista, CA

Background and Objective: Evidence suggests that environmental factors are related to
physical activity, dietary intake, and overweight status in youth; however, clear relationships
have not been established in Latino youth. We examined associations of barriers to healthy
lifestyle behaviors in the school, home, and neighborhood environments with physical
activity, diet, and body mass index (BMI) in Hispanic youth.

Method: HCHS/SOL Youth enrolled 1,466 Hispanic/Latino youth (728 boys and 738 girls, 8-16
years) and their parents from four cities (Bronx, Chicago, Miami, and San Diego). Parents
provided demographic information. Standardized scales for assessment of environmental
barriers were completed by youth and parents, and composite scores calculated to reflect
barriers to healthy lifestyle behaviors in the school (e.g., fast food availability), home (e.g.,
unhealthy foods), and neighborhood (e.g., crime,) environments; 24-hour dietary recalls were
quantified and scored with the Healthy Eating Index 2010 (HEI); 7-day accelerometry
measured moderate-vigorous physical activity (MVPA); and height and weight was measured
to obtain BMI %iles of youth. Hierarchical regression analyses were utilized to determine
relationships of environmental barriers to MVPA, HEI, and BMI. Analyses used sample and
survey-weighting procedures; demographic variables (age, sex, parental income and
education) were entered in the first step of each model.

Results: More after-school environmental barriers were associated with less healthy eating (p
< .001) and less MVPA (p < .005). More barriers in the home environment with regard to
availability of workout equipment (p < .01), healthy foods in the home (p < .001), and food
practices associated with TV-watching (p < .001) were associated with less healthy eating;
food practices associated with TV-watching were also related to less MVPA (p < .05). More
neighborhood barriers related to eating away from home were associated with less healthy
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2253

eating (p < .001) and higher BMI (p < .01).

Conclusion: Several youth and parent-reported barriers in the after-school, home, and
neighborhood environments are associated with less healthy eating, decreased physical
activity, and higher BMI in Hispanic youth. Interventions to improve health behaviors and
weight in this population should include and address these environmental factors.

CORRESPONDING AUTHOR: Alan M. Delamater, PhD, University of Miami Miller School of

Medicine, Miami, FL, 33136; [email protected]
S2254 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C276 6:00 PM-7:00 PM

ADAPTATION OF THE CULTURAL FORMULATION INVENTORY TO ASSESS

PSYCHOSOCIOCULTURAL FACTORS ASSOCIATED WITH PRENATAL STRESS

Kendra Campbell, Ph.D.1, Dhara Shah, B.A.1, Hugh Leonard, B.A.1, Loren McKnight, M.S.1,
Kirsten Pickard, B.S. student1, Omar Gomez, B.A.1, Emily Krier, B.A. student2

1
University of Alaska Fairbanks, Fairbanks, AK; 2University of Alaska Anchorage, Fairbanks, AK

Exposure to psychosocial, cultural, or environmental stress during pregnancy has known

adverse impacts on development. Prenatal stress can include general stressors as well as
stressors specific to pregnancy, such as worry about fetal health complications. In order to
optimize prenatal preventive health efforts, it is important to understand the multifaceted
contextual factors affecting the pregnancy experience. We sought to understand
psychosociocultural factors that affect optimal prenatal health, with a specific focus on stress.
We adapted the DSM-5 Cultural Formulation Interview (CFI; American Psychiatric Association,
2013) to assess psychosociocultural barriers and facilitators associated with prevention of
prenatal stress. Our sample included 14 pregnant women (mean age = 30.6; mean gestational
age at CFI = 33.2 weeks) who were part of an ongoing study on prenatal stress. A directed
thematic analysis was used (n = 14) to contextualize participant responses through an
ecological systems model (Bronfenbrenner, 1977). Multiple themes emerged amongst
participant responses. Locus of control appeared to influence participants experience of
stressors during pregnancy; that is, participants who expressed more perceived control
around external stressors indicated less distress. Incongruences between and within multiple
systemic factors (e.g., family cultural values versus participant individual values; perceived
support from family versus perceived support from healthcare provider) emerged as barriers
to optimal prevention of prenatal stress. Conversely, when systemic factors were more
congruent, it appeared that optimal prenatal health was better facilitated. These findings
suggest that psychosociocultural factors play an important role in the pregnancy experience,
and specifically in how pregnant women are affected by or buffered from external stressors.
The current results add to the body of research suggesting the importance of adequately
assessing for and addressing psychosociocultural factors during routine prenatal care.
Additionally, our results suggest that adapting the CFI to fit a specific issue or clinical
population can offer useful clinical and research utility by adding a richer understanding of the
sociocultural context.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2255

CORRESPONDING AUTHOR: Kendra Campbell, Ph.D., University of Alaska Fairbanks, Fairbanks,

AK, 99775; [email protected]
S2256 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C277 6:00 PM-7:00 PM

CURRENT AND IDEAL SKIN TONE: ASSOCIATIONS WITH TANNING BEHAVIOR AMONG SEXUAL
MINORITY MEN

Patrycja Klimek, BA1, Benjamin M. Rooney, B.A.2, Kalina M. Lamb, B.A.3, Kelsey A. Nogg, B.A.2,
Aaron J. Blashill, Ph.D. 2

1
SDSU/UC San Diego Joint Doctoral Program in Clinical Psychology, La Jolla, CA; 2San Diego
State University, San Diego, CA; 3San Diego State University, San Deigo, CA

Background: Skin cancer is the most common form of cancer. Sociocultural models of body
image implicate appearance norms in the formation of skin tone ideals and attitudes toward
tanning. Discrepancies between current and ideal skin tone, hypothesized to relate to body
dissatisfaction and internalization of the tanned ideal, predict tanning behavior. However,
little is known about skin cancer risk behaviors in sexual minority (e.g., gay and bisexual) men.
This is a notable gap, given that sexual minority men report high prevalence of skin cancer
diagnoses. Therefore, it was hypothesized that current skin tone will significantly moderate
the relationship between darker ideals and tanning behaviors among sexual minority men.
Method: Data were collected online from 238 sexual minority men residing in San Diego
County, California, who had a mean age of 24 (SD = 5.43). Independent variables were ideal
and current skin tone ratings on a perceptual scale of 1 (lightest) to 12 (darkest) and the
dependent binary variables were indoor tanning and outdoor tanning (i.e., none vs. 1 or more
times over the past three months).
Results: Two separate logistic regression models were used to test main and interaction
effects on indoor and outdoor tanning, and subsequent simple slope analyses indicated
effects at the mean, and 1 standard deviation (SD) moderator values (current skin tone).
Darker ideals significantly predicted increased indoor tanning (B = 1.02, Odds ratio (OR) =
2.79, z = 2.73, p = .006); however, this effect was qualified by a significant interaction (B = -
0.12, OR = 0.89, z = -2.01, p = .045). The effect of darker ideals was strongest among
individuals with lighter current skin tone (B = 0.79, OR = 2.21, z = 2.85, p = .004). Darker ideals
also significantly predicted increased outdoor tanning (B = 1.34, OR = 3.81, z = 6.07, p < .001);
however, this effect was also qualified by a significant interaction (B = -0.13, OR = 0.87, z = -
4.01, p < .001). The effect of darker ideals was strongest among individuals with lighter
current skin tone (B = 1.06, OR = 2.89, z = 6.36, p < .001).
Conclusion: The results indicate that sexual minority men whose skin tone does not match
their ideal, whether or not they are currently light or dark skin-toned, are at greater risk for
engaging in skin cancer risk behaviors. Future skin cancer prevention programs aimed at
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2257

sexual minority men may wish to consider techniques that modify internalization of skin tone
ideals.

CORRESPONDING AUTHOR: Patrycja Klimek, BA, SDSU/UC San Diego Joint Doctoral Program in
Clinical Psychology, La Jolla, CA, 92092; [email protected]
S2258 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C278 6:00 PM-7:00 PM

ENROLLMENT STRATEGIES OF THE HEALING AND EMPOWERING AND ALASKAN LIVES

TOWARD HEALTHY HEARTS (HEALTHH) STUDY

Mariah Knox, BA1, Nicole Jeffery, BA2, Nicole E. Anzai, BS2, Neal Benowitz, MD3, Matthew
Schnellbacher, MD4, Judith J. Prochaska, PhD, MPH2

1
Alaska Native Medical Hospital, Anchorage, AK; 2Stanford University, Stanford, CA; 3Chief,
Division of Clinical Pharmacology, San Francisco, CA; 4Alaska Native Medical Center,
Anchorage, AK

High rates of smoking, inactivity, hypertension and hypercholesterolemia in Alaska Native

(AN) people contribute to increased risk for cardiovascular disease and death. The HEALTHH
study is a telemedicine-delivered, multibehaviorally targeted, two-group RCT in the Norton
Sound Region of Alaska, with 16 villages and populations from 107 to 3,695 (average=596).
The study interventions target smoking and physical activity vs. medication compliance and a
heart-healthy AN diet. Eligibility include; AN adult, living in the Norton Sound region, smoking
5+ cigarettes/day, with high blood pressure or high cholesterol, English speaking, BMI < 50,
not pregnant, and not in a tobacco cessation program. Study staff travel to the villages with
stays ranging from 1 to 8+ days. In this rural and geographically remote area, we summarize
our recruitment efforts and identify best practices for enrolment.

Since June 2015 the team has made 46 trips to the villages screening 419 individuals. Of the
245 ineligible individuals, 38% had normal blood pressure or cholesterol. Most screenings
occurred on Days 2 (n=98, 25%), 3 (n=102, 26%), and 4 (n=80, 21%), with Day 1 (n=31, 8%)
focused on community outreach, and coordinating with the clinic. To date, 104 participants
enrolled into the study (53% women, 47% men; mean age =51, SD=14, R: 21 to 81).
Enrollment has been greatest on Days 3 (n=21, 20%) and 4 (n=35, 34%), with waning return
Days 5-8+ (15% to 2%). Village size is significantly correlated with enrollment (r=0.76, p

Funding: Grant #3R01HL117736

CORRESPONDING AUTHOR: Mariah Knox, BA, Alaska Native Medical Hospital, Anchorage, AK,
99508; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2259

C279 6:00 PM-7:00 PM

EXPERIENCES OF DISCRIMINATION IN HISPANIC AMERICANS AS A FUNCTION OF

ACCULTURATION AND ENVIRONMENTAL CRIME

Stefani Betos, N/A1, Quinn Wilson, N/A1, Sarah D. Mills, MS, MPH2, Scott Roesch, N/A3,
Georgia Sadler, Ph.D4, Vanessa L. Malcarne, Ph.D.5

1
SDSU, San Diego, CA; 2San Diego State University/University of California, San Diego Joint
Doctoral Program in Clinical Psychology, Los Angeles, CA; 3SDSU and UC San Diego, San Diego,
CA; 4University of California, San Diego Moores Cancer Center, San Diego, CA; 5San Diego
State University, San Diego, CA

Discrimination is widely experienced by people from different communities and has been
linked with negative health outcomes. However, environmental characteristics that provide
the context for experiences of discrimination(EOD) are poorly understood. Crime has been
identified as a potentially important environmental factor that can create an atmosphere of
distrust and diminish social engagement, but crime has received little attention as
a community-leveldeterminant of peoples EOD, including in Hispanic Americans. This study
examined the relationship of violentand non-violent community crime to EOD in 372 Hispanic
American men and women, considering acculturation levels of the participants.
The Experiences of Discrimination (EOD) scale, using the 9-item frequency scoring (Kreiger,
Smith, Naishadham, Hartman, & Barbeau, 2005), and Brief Acculturation Scale for Hispanics
(BASH) were administered to determine the frequency of
discrimination experienced and level of acculturation of the participants. Census data from
2013(the year preceding the administration of the EOD and BASH) were used to determine
rates of violent (aggravated assault) and nonviolent crime (residential burglary) in
participants communities. ANOVAs or t-tests were conducted to determine if
there were EOD differences across gender, income, education, and country of origin groups;
none of the tests yielded significant results (ps were all > .05). Multi-level modeling was used
to examine the relationships among BASH scores, crime variables, and EOD scores, controlling
for age. Violent and non-violent crime were not significant predictors of EOD
scores (aggravated assault: b = .05, p = .73; residential burglary: b= -.07, p = .53). BASH scores
were positively associated with EOD scores (b = .24, p = .04). There was no significant
interaction of BASH and crime scores in predicting EOD scores (aggravated assault: b = .40, p =
.17; residential burglary: b < -.01, p > .99). In sum, results did not support crime as an
environmental characteristic related to personal EOD, but
documented that more highlyacculturated members of the Hispanic American
community reported more frequent EOD. Determining whether there are other
S2260 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

environmental characteristics common to more highly acculturated Hispanic Americans that

contribute to their EOD should be the focus of future research.

CORRESPONDING AUTHOR: Stefani Betos, N/A, SDSU, San Diego, CA, 91942;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2261

C280 6:00 PM-7:00 PM

LATENT CLASSES OF PARENTING STRESS, FAMILY MEALS AND CHILD BODY MASS INDEX IN A
NATIONALLY-REPRESENTATIVE SAMPLE OF MOTHERS

Eleanor T. Shonkoff, PhD1, Chih-Ping Chou, PhD2, Adam Leventhal, PHD3, Ricky Bluthenthal,
PhD4, Mary Ann Pentz, PhD, MA2, Genevieve F. Dunton, PhD, MPH5

1
Tufts University, ChildObesity180, Boston, MA; 2University of Southern California,
Department of Preventive Medicine, Institute for Health Promotion and Disease Prevention
Research, Los Angeles, CA; 3USC, Pasadena, CA; 4Keck School of Medicine, University of
Southern California, Los Angeles, CA; 5University of Southern California, Los Angeles, CA

Purpose: Parenting stress, combined with other stressors, has been linked to poorer child
diet, sedentary behavior, and obesity risk. Yet previous findings are inconsistent and limited
by small sample sizes. This study explored latent subgroups of mothers characterized by stress
(parenting stress, maternal mental health issues, poverty, single parenting) and links to family
meals and child obesity status.

Methods: Nationally-representative cross-sectional data came from mothers of children aged

10 17 years (National Survey of Childrens Health, 2011). Stress measures included items
from the Parental Stress Index & Parental Attitudes about Childrearing ( = .67); maternal
mental health; BMI 95th percentile; family meals per week; % federal poverty level; and
single parent status. Latent Class Analysis identified subgroups of mothers. Multinomial
logistic and linear regressions tested associations with family meals and child obesity risk,
controlling for child minority status and age.

Results: The analytical sample was mothers (n = 31,239, 69% of sample) of children 13.65 yrs.
on average (SD = 2.32), 52% male, and 70% white. Four latent classes emerged, distinguished
by poverty level; maternal stress levels were similar across classes. Class 1 (C1), the second
largest class, (30.8%) had the highest poverty. Class 2 (20.3%) had moderately higher income;
Class 3 (9.6%) had moderately lower. The largest class, Class 4 (C4) (39.2%), had the lowest
poverty. Compared to C1, the mid-higher and higher income classes had family meals fewer
days per week (C2 = -23, SE = .08, p < .01; C4 = -.39, SE = .07, p =

Conclusions: Four subgroups of mothers of US children were identified; the main

distinguishing stressor was poverty. High and mid-high poverty subgroups had more frequent
family meals but higher risk of obesity. Although the literature indicates that family meals can
be protective against child obesity, these nationally representative data suggest this may not
S2262 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

occur in higher poverty families. Perhaps other risk factors, such as lack of access to healthy
food outlets or safe spaces for physical activity, outweigh the potential benefit of frequent
family meals in higher poverty families. Further research is needed to understand components
of frequent family meals, such as nutrient content or feeding practices, and differences across
demographic groups.

CORRESPONDING AUTHOR: Eleanor T. Shonkoff, PhD, Tufts University, ChildObesity180,

Boston, MA, 02111; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2263

C281 6:00 PM-7:00 PM

MINDFULNESS AND POSITIVE AFFECTIVITY: DISPOSITIONAL PROTECTIVE FACTORS OF SELF-

RATED HEALTH IN UNDERGRADUATE STUDENTS

Christine E. Smith, M.A.1, Virginia Gil-Rivas, Ph.D.2

1
University of North Carolina at Charlotte, Charlotte, NC; 2UNC Charlotte, Charlotte, NC

Introduction: This study examined the contribution of mindfulness and positive affect on self-
rated health among young adults. Dispositional mindfulness and positive affect have been
found to moderate the impact of stress on psychological symptoms but relatively few studies
have examined their contribution to self-rate health. In addition, few have examined the
contribution of these factors to self-rated health accounting for psychological distress and
poor sleep quality.

Participants: Undergraduate students (N=201) from diverse backgrounds (63.7% White);

72.6% female; age range 18-46 years (M=19.2; SD=3.55).

Procedure: Self-administered electronic survey. Demographics, self-rated health, positive and

negative affectivity (PANAS-Watson et al., 1988), dispositional mindfulness (FFMQ-Baer et al.,
2006), perceived stress (PSS-Cohen, Karmack, & Mermelstein, 1983), depressive and anxiety
symptoms (HSCL-Derogatis et al., 1974), and sleep quality (PSQI Buysse et al., 1989). All
scales indicated excellent internal reliability in the sample ( = 0.83 - 0.93).

Results: Less than good health was reported by 10% of the sample. Regression analyses
indicated that with the inclusion of covariates (age, gender, education, race, sleep quality, and
symptoms of depression and anxiety) at step 1, depressive symptoms (=.28, p < .01, CI = .136
- .705) significantly predicted poorer self-rated health. With the addition of dispositional and
state psychological factors at step 2, only mindfulness (= -.22, p < .05, CI = -.043 - -.004) and
positive affectivity (= -.19, p < .05, CI = -.018 - -.001) significantly predicted poorer self-rated
health. Interestingly, depressive symptoms were no longer a significant predictor of self-rated
health in step 2. This model explained 26% of the variance in self-rated health.

Conclusion: Findings provide additional evidence that dispositional mindfulness and positive
affect contribute to self-rated health, even in the context of current symptoms of
psychological distress and poor sleep quality. Evidence supports that training can inculcate
changes in dispositional factors such as mindfulness. Thus, identifying protective factors that
are amenable to change may guide the development of programs to increase mindfulness and
promote positive affect, which may influence health in this population.
S2264 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Christine E. Smith, M.A., University of North Carolina at Charlotte,

Charlotte, NC, 28223; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2265

C282 6:00 PM-7:00 PM

PARENT AND TEACHER SELF-PERCEPTION AS A ROLE MODEL FOR HEALTH AND DIET AND
EXERCISE BEHAVIOR

Noel Kulik, PhD CHES, Jeffrey Martin, Ph D, Erin E. Centeio, PhD, Cheryl L. Somers, Ph.D.

Wayne State University, Detroit, MI

Nurses and physicians are expected to be role models of health for adults and children.
However, for children, parents and teachers are often more significant role models for health.
The ways in which parents and teachers view themselves as health-promoting role models,
specifically related to healthy eating (HE) and physical activity (PA), is an important yet
understudied area of research. Parents and teachers self-perceptions as health-promoting
role models can significantly influence their effectiveness when teaching about or advocating
for health. The purpose of this study is to examine the relationship between self-perceptions
as a role model and HE and PA behaviors among parents, school teachers and school
administrators. Parent participants (n=205) were 38.1 (7.1) years old, mostly mothers
(n=174; 84.9%) and of White (n=93; 46.5%), Black (n=85; 42.5%) or Other (n=22; 11.0%)
racial/ethnic background. Teachers/Admins (n=175) were 43.2 (9.5) years old, mostly female
(n=145; 82.9%) and of White (n=121; 69.5%), Black (n=42; 24.1%) or Other (n=11; 6.3%)
racial/ethnic background. Data were collected using the Self as Role Model for Health
Promotion (SARMHEP) scale, which has been designed for use with nurses. It was reduced and
adapted for this study to examine parent/teacher/admin perceptions. Spearman-Brown
prophecy predictions of reliability (if the subscales were expanded to the full scale) was .974
and .924 parent and teacher/admin scales, respectively. Convergent validity was established
with correlations (r = .6, p < .01) with the question: I am a strong role model for health.
Scores were summed and ranged from 5-30 for parents and 5-66 for teachers/admins, with
higher scores indicating more intrinsic motivation to engage in personal health practices. PA
was measured using the Godin Leisure-Time Exercise Questionnaire, and HE was assessed
using the following question: How would you rate the overall nutritional content of your
diet (1: unsure to 5: excellent). Results show that for parents, perceptions of self as a role
model for health is positively associated with PA (r = .43, p < .001) and HE (r = .56, p < .001).
For teachers, perceptions of self as a role model for health is positively associated with PA (r =
.28, p < .001), HE (r = .34, p < .001), and the number of years as a certified teacher (r = .16, p <
.05). Role model perceptions did not differ by age, race/ethnicity or sex for parents or
teachers. Findings from this study suggest that an awareness as a role model for health is
associated with positive healthy eating and physical activity behaviors. This study is significant
because little is known about the health role-modeling beliefs of parents,
S2266 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

teachers/administrators and behavior change, which is important to consider when planning

whole-school health interventions that involve parents and the larger community.

CORRESPONDING AUTHOR: Noel Kulik, PhD CHES, Wayne State University, Detroit, MI, 48202;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2267

C283 6:00 PM-7:00 PM

SOCIAL INTEGRATION AND EMOTIONAL REGULATION

Kimberly Stevens, N/A1, Zachary Gibson, N/A2, Athena Howell, N/A3, Matthew White, N/A4,
Julianne Holt-Lunstad, PhD3

1
Brigham Young University, Poway, CA; 2Brigham Young University, Spanish Fork, UT;
3
Brigham Young University, Provo, UT; 4Brigham Young University, Draper, UT

Objective: Researchers hypothesize greater social support leads to stress buffering,

protecting socially integrated adults from the physiological effects of stress. Our primary aim
was to evaluate if adults with greater social integration would show less activity in ROIs
associated with emotional reactivity (Amygdala) and greater activity in areas associated
emotional regulation (ACC) when asked to regulate negative emotions.

Methods: 31 participants, aged 40-68, completed the social network index (SNI). During the
MRI task, participants were shown a series of photographs. There were three conditions: a
neutral photo that participants viewed naturally (look neutral), a negative photo that
participants viewed naturally (look negative) and a negative photo, shown after participants
were instructed to regulate (or decrease) their negative reaction (decrease negative).

Results: Social Integration scores were not predictive of activity in the ACC or Amygdala. In an
exploratory, whole brain analysis, social network size was predictive of BOLD activation in the
right superior parietal lobe. Participants with a larger social network had a smaller difference
in activity between the look neutral condition and the decrease negative condition, as
compared to adults with a smaller social network.

Discussion: A small difference in BOLD activation in the right superior parietal lobe when the
look decease negative condition is compared to the look neutral condition is indicative of
less reactivity when presented with disturbing images. As the right superior parietal lobe is
associated with visual working memory, it is possible that adults with higher levels of social
integration have less need for vigilance to threats in their environments. This is consistent
with the social support/ buffering hypothesis, indicating a social protection against the
physiological effects of stress that would accompany constant vigilance. This suggests that
those who are more socially integrated have less of a neural reaction to emotionally
provocative stimuli, and therefore may provide potential neurobiological explanations for why
some adults show better age-related trajectories that affect health than others.

References
S2268 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Available upon request

CORRESPONDING AUTHOR: Kimberly Stevens, N/A, Brigham Young University, Poway, CA,
92064; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2269

C284 6:00 PM-7:00 PM

SPOUSAL RETIREMENT COMMUNICATION AND DECISION-MAKING ON RELATIONSHIP

QUALITY AND SELF-REPORTED HEALTH IN OLDER ADULTS

Wendy C. Birmingham, PhD1, Lori Wadsworth, PhD1, Erin Kaseda, N/A2, Maren Wright. Voss,
MS3, Man Hung, PhD3

1
Brigham Young University, Provo, UT; 2Brigham Young University, Draper, UT; 3University of
Utah, Salt Lake City, UT

BACKGROUND: Research has shown a consistent association between marriage and both
physiological and psychological health outcomes. However, it also shows that it isnt marriage
per se that is beneficial; relationship quality (RQ) matters. One indicator of satisfying
relationship quality is the utilization of verbal and nonverbal communication. Poor
communication may result in poorer RQ and thus worse health outcomes. For older couples,
retirement can bring significant changes to their relationship. Thus, retirement affects both
the retiree and their spouse, yet, decisions regarding retirement may be made unilaterally by
the retiring spouse without partner input. Such couples may report lower RQ while those who
communicate effectively and make joint decisions regarding retirement issues may report
better RQ which can lead to better health outcomes.

METHODS: We conducted one-on-one interviews with 240 retired and semi-retired

individuals and administered surveys on RQ using the Marital Adjustment Test (MAT) and self-
reported health measures.

RESULTS: Participants were aged 53-90, with 58% male, mostly white (91.8%), educated
(88.0%), married (79.4%), and of higher income (70.0% >$50,000/year). Most married
participants (96%) indicated they discussed retirement decisions with their spouse and most
(89.6%) reported being either happy or very happy in their marriage. Couple communication
was associated with participants perceived positivity of their relationship (p < .001). When
asked about the secret to happiness in retirement many participants discussed strong
communication and relationship with their spouse: You have to have a good relationship
with your spouse. Individuals who reported better communication with their spouse also
reported significantly higher overall health (p=.04), lower incidences of cardiovascular issues
(97.5%) and diabetic issues (90%). Of particular interest, distressed MAT scores showed a
trend toward reported high blood pressure issues (p=.07).

CONCLUSIONS: Retirement can be a stressful time in older adults lives. RQ could suffer;
S2270 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

however, in our sample of older adults, communication about retirement decisions was
associated with greater marital positivity and better self-reported health. In a population of
healthy older adults, RQ still plays a significant role in health.

CORRESPONDING AUTHOR: Wendy C. Birmingham, PhD, Brigham Young University, Provo, UT,
84602; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2271

C285 6:00 PM-7:00 PM

WEIGHT-RELATED BELIEFS AND BEHAVIORS IN APPALACHIA

Elise L. Rice, PhD1, Minal Patel, PhD, MPH2, Katrina J. Serrano, PhD3, Chan Thai, PhD, MPH4,
Kelly D. Blake, ScD5, Robin Vanderpool, BS, MPH, DrPH6

1
National Cancer Institute, Bethesda, MD; 2National Cancer Institute, Rockville, MD; 3National
Cancer Institute/Behavioral Research Program, Bethesda, MD; 4Santa Clara University, Santa
Clara, CA; 5Behavioral Research Program, Division of Cancer Control and Population Sciences,
National Cancer Institute, National Institutes of Health, Bethesda, MD; 6University of Kentucky
College of Public Health, Lexington, KY

Background: Extensive research has identified health disparities among residents of

Appalachia, a unique 13-state geographic region of the U.S. In particular, Appalachians exhibit
elevated rates of obesity and related behaviors such as physical inactivity. However, the
psychological processes underlying those differences are less apparent.

Objective: The current study aims to clarify how health beliefs differ among Appalachians
relative to non-Appalachians in light of related disparities in health behaviors and conditions.

Methods: The present analyses explored health beliefs in a sample of American adults
(n=14,451) who responded to the Health Information National Trends Survey between 2011
and 2014, 8% of whom (n=1,015) resided within Appalachia. Specifically, respondents
indicated how much they thought both health behaviors and genetics determined whether a
person would develop obesity. They also reported on their typical patterns of physical activity,
sedentary behavior, and fruit and vegetable consumption, as well as their height, weight, and
general health, among other variables. Weighted regression analyses were used to test
whether Appalachians differed from non-Appalachians on measures of calculated BMI,
healthy lifestyles, confidence in their ability to take care of their health, and perceptions of
risk factors for obesity.

Results: The average BMI was higher among Appalachians than non-Appalachians (=0.98,
p=.03). Additionally, Appalachians were less likely to report that health behaviors like diet and
exercise are related to obesity (=-0.38, p=.03), though they were not more or less likely than
non-Appalachians to attribute obesity to genetics (=-0.11, p=.42), nor were they more or less
confident in their ability to take care of their health (=0.09, p=.53). Appalachians reported
engaging in strength training less frequently than non-Appalachians (=-0.35, p=.01), although
S2272 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

there were no differences in physical activity per week (=5.36, p=.88) or consumption of
fruits and vegetables ( =-0.01, p=.97) in the current sample.

Conclusions: The results of this study suggest that Appalachians may undervalue behavioral
determinants of health, potentially complicating efforts to reduce obesity and encourage
healthy activities throughout the region.

CORRESPONDING AUTHOR: Elise L. Rice, PhD, National Cancer Institute, Bethesda, MD, 20892;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2273

C286 6:00 PM-7:00 PM

WILL WORK FOR FOOD: HOURS WORKED AMONG COLLEGE STUDENTS IS LINKED TO POORER
HEALTH BEHAVIORS AND OUTCOMES

Meg Bruening, PhD, MPH, RD1, Irene van Woerden, MS2

1
Arizona State University, Phoenix, AZ; 2ASU, Tempe, AZ

Objective: The transition to college is marked by shifts in health behaviors and outcomes,
including poor nutrition, physical activity (PA), and weight gain. How working for pay
contributes to these shifts has not been studied. As a result, this study sought to examine how
working for pay is related to health behaviors and outcomes among college freshmen.

Methods: During fall of 2015, diverse freshman students (n=1403; 65% female; 50% white)
participating in the SPARC (Social impact of Physical Activity and nutRition in College) study
completed web-based surveys; anthropometrics were measured by trained research
assistants at the same time. Mixed multivariate logistic regression models were used to
examine the relationship between hours worked (0 hrs, 0.5 to 19.5 hours, 20 hrs per week)
and eating behaviors, physical activity, sleep, stress, food insecurity, and measured weight
status, adjusting for sociodemographics, and clustering of students within residence halls.

Results: Most students (73%) did not work for pay; the mean hours worked among working
students was 18.8. As compared to students working 0.5 to 19.5 hrs, students not working at
all had a significantly higher odds of 3 hrs of screen time (OR=1.3; 95% CI: 1.01, 1.9) and
reporting getting enough sleep (OR=1.6; 95% CI: 1.2, 2.2). As compared to students working
0.5 to 19.5 hrs, students working more than 20 hrs reported lower odds of fast food intake
(OR=0.6; 95% CI: 0.4, 0.9), and higher odds of food insecurity (OR=1.7; 95% CI: 1.1, 2.6) and
perceived stress (OR=1.5; 95% CI: 1.1, 2.1). Hours worked was not associated with breakfast,
fruit or vegetable intake, PA, or weight status.

Conclusions: Despite being full time students, many students may need to work more than 20
hours per week, which was related to higher stress and food insecurity. Not working was
related to more sedentary behavior and higher quality of sleep. Interventions are needed to
support students working more than half time and to support less sedentary behavior among
non-working students. Future research should address the long-term health and academic
effects of working for pay among students.
S2274 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Meg Bruening, PhD, MPH, RD, Arizona State University, Phoenix,
AZ, 85004; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2275

C287 6:00 PM-7:00 PM

EFFECTS OF DANCE MOVEMENT THERAPY ON SUBJECTIVE STRESS AND CORTISOL SLOPE

AMONG BREAST CANCER PATIENTS UNDERGOING RADIOTHERAPY

Rainbow T. H. Ho, PhD, Ted C. T. Fong, MPhil, Irene K. M. Cheung, MSW

Centre on Behavioral Health, Hong Kong, N/A, Hong Kong

Background: Breast cancer patients are at risk for psychosocial distress, which may lead to
neuroendocrine dysfunction. Dance Movement Therapy (DMT) is a movement-based
psychosocial intervention that incorporates exercise, artistic, and recreational components.
This study aimed to examine the effectiveness of DMT on perceived stress and diurnal cortisol
slope.

Methods: This clinical trial randomized 121 Chinese breast cancer patients undergoing
radiotherapy treatment into the DMT (n = 63) or control group (n = 58). The intervention
comprised six 1.5-hour group sessions held twice a week over the course of radiotherapy. The
participants completed self-report measures on perceived stress, health and sleep
characteristics at baseline. Salivary cortisol samples were collected at five time points and
perceived stress at the end of the intervention. Moderated mediation analysis was used to
evaluate the DMT effect on the diurnal cortisol slope.

Results: Overall, DMT showed a marginal beneficial effect on diurnal cortisol slope (B = 0.66,
95% CI = 1.36 to 0.06, = 0.17). Baseline perceived stress significantly and negatively
moderated the effect of DMT on diurnal cortisol slope (B = 0.21, 95% CI = 0.36 to 0.05, =
0.35). At one SD above the pss1 mean, the cortisol slope of the DMT group (M = -6.93) was
significantly steeper than the control group (M = -5.35) with 95% CI for the difference of -1.58
= -2.59 to -0.61. The indirect effect of DMT on cortisol slope via Time 2 perceived stress was
not statistically significant.

Discussions: The DMT program had a modest overall effect on diurnal cortisol slope in breast
cancer patients but appeared to be effective in modulating the neuroendocrine response of
those who were distressed. The present findings suggest that perceived stress may play a
moderator but not mediator role in the DMT effect on diurnal cortisol slope.

CORRESPONDING AUTHOR: Rainbow T. H. Ho, PhD, Centre on Behavioral Health, Hong Kong,
N/A, 852; [email protected]
S2276 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C288 6:00 PM-7:00 PM

NEUROTICISM, AND OPENNESS TO EXPERIENCE PREDICTIVE OF CARDIOVASCULAR REACTIVITY

TO A STRESS TASK.

Daniel Stephenson, M.S., Kevin Larkin, Ph.D.

West Virginia University, Morgantown, WV

Numerous studies have demonstrated that various personality factors are related to health
outcomes (e.g., Hampson & Friedman, 2008; Turiano et al., 2011). However, the mechanisms
through which personality affects health are not clear. Our study sought to explore one
possible mechanism by which personality traits may affect health: via increased or decreased
reactivity to a stress task. The Midlife Developmental Inventory (MIDI), which assesses
extraversion, agreeableness, neuroticism, conscientiousness, agency, and openness to
experience, was given to 77 undergraduate students. Cardiovascular reactivity was measured
for all participants during performance of a mental arithmetic task during which participants
were verbally harassed by the experimenter.

To examine which personality traits were related to reactivity to stress, hierarchical

regressions were performed for each cardiovascular parameter (systolic, diastolic, and mean
arterial blood pressure, heart rate). Resting cardiovascular parameters were entered in the
first step, and all personality traits were entered in the second step. Results revealed that
neuroticism was positively related to both iastolic blood pressure (DBP; = .270, p = .015) and
mean arterial pressure (MAP; = .222, p = .031). Openness to experience was inversely
related to MAP ( = -.478, p = .017).

These results show that various personality factors are indeed related to cardiovascular
reactivity to stress. These findings indicate that stress reactivity is one likely mechanism by
which personality traits affect health. Additionally, this finding highlights the importance of
including personality factors as potential covariates when conducting cardiovascular reactivity
research.

CORRESPONDING AUTHOR: Daniel Stephenson, M.S., West Virginia University, Morgantown,

WV, 26501; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2277

C289 6:00 PM-7:00 PM

SOCIAL MEDIA AS A SHIELD: FACEBOOK MAY BUFFER ACUTE STRESS

Holly Rus, B.A. , Jitske Tiemensma, PhD

University of California, Merced, Merced, CA

Background: Facebook is commonly used as a tool in behavioral health interventions.

However, a conflicting body of research suggests that Facebook use may both enhance and
undermine psychosocial constructs related to well-being. Despite these associations, to date
there has been hardly any objective measurement of the potential impact of Facebook use on
health in the context of stress. This pilot study aimed to examine the impact of Facebook use
on stress reactivity. Methods: 48 Facebook users (10 males, mean age 19.82.3) were
randomly assigned to either the experimental group (Facebook use) or the control group
(quietly sitting with digital reading materials) during 18 minutes before experiencing the Trier
Social Stress Test (TSST). All participants completed measures of mood and subjective well-
being. Objective physiological markers (blood pressure, heart rate, and salivary cortisol) were
collected at baseline and several times during recovery. Results: Groups did not differ on any
measure at baseline. During the TSST, the experimental group showed a decrease in heart
rate, whereas the control group showed a steep increase (pConclusion: This exploratory study
showed that participants who used Facebook before experiencing acute social stress showed
a more blunted response to the TSST with regard to heart rate and salivary cortisol compared
to the control group. Results suggest that Facebook use may act as a buffer against acute
stress. Findings hold value for informing the use of social media in behavioral medicine.

CORRESPONDING AUTHOR: Holly Rus, B.A. , University of California, Merced, Merced, CA,
95340; [email protected]
S2278 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C290 6:00 PM-7:00 PM

THE EFFICACY OF INTERNET-BASED INTERVENTIONS FOR STRESS MANAGEMENT: A META-

ANALYSIS

Malwina Tuman1, Anne Moyer, Ph.D.2

1
Stony Brook University, Glen Head, NY; 2Stony Brook University, Stony Brook, NY

Exposure to chronic stress can be deleterious to psychological, somatic, and behavioral health,
and even lead to premature death. In recent years, the Internet has become a platform for
many health-related activities, including becoming a medium for the delivery of psychological
interventions. Such interventions can be delivered on computers, tablets, or mobile phones,
and have been found to be effective in addressing a variety of health-related problems.
Additionally, interventions delivered via the Internet offer the potential to overcome barriers
associated with pursuing in-person psychological care. Many stress management techniques,
such as exercising or meditating, can be time consuming. Given that the lack of time often is a
source of stress, additional solutions, including Internet-based ones, that offer greater time
effectiveness, flexibility and readiness of use are needed. Although much of the research
examining Internet- based interventions has been conducted with clinical populations,
recently, more studies have been devoted to the development and testing of Internet-based
programs for the general public, such as interventions for stress management. This review
quantitatively synthesizes 18 randomized clinical trials (RCTs) examining the efficacy of
interventions for stress management delivered via the Internet. The aggregate effect size was
medium (d = .69, p < .0001) indicating that, on average, individuals assigned to the
intervention showed significantly lower stress levels than controls post-intervention.
Moderator analyses showed that the type of population (employees produced the strongest
effect size relative to students or adults in the general population), the intervention technique
(mixed approach techniques were superior to single intervention techniques), the treatment
length and the number of intervention sessions (the greater the length and the more
interventions sessions, the greater the treatment effectiveness) impacted the strength of the
effect. Overall, Internet-based interventions emerge as a promising tool for targeting stress,
particularly those that are multifaceted, involve a higher dose of intervention, and perhaps
are targeted toward groups experiencing specific types of stressors.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2279

CORRESPONDING AUTHOR: Malwina Tuman, Stony Brook University, Glen Head, NY, 11545;
[email protected]
S2280 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C291 6:00 PM-7:00 PM

MERITORIOUS AWARD WINNER

TRAUMA, PSYCHOLOGICAL DISTRESS, AND CARDIOMETABOLIC HEALTH AMONG

UNDOCUMENTED LATINOS: RESULTS FROM PROJECT VOCES

Angela P. Gutierrez, MS1, Luz M. Garcini, PhD2, Juan M. Pea, B.A.3, Linda C. Gallo, PhD4,
Elizabeth A. Klonoff, PhD5

1
SDSU/UC San Diego Joint Doctoral Program in Clinical Psychology, La Jolla, CA; 2Rice
University, Houston, TX; 3University of New Mexico, Albuquerque, NM; 4San Diego State
University, Chula Vista, CA; 5SDSU/UC San Diego Joint Doctoral Program in Clinical Psychology,
San Diego, CA

Exposure to traumatic events has been associated with increased risk for multiple health
conditions such as cardiovascular disease and type 2 diabetes, and with all-cause mortality. It
remains unclear if it is the exposure to trauma and/or the experience of subsequent
psychological distress (e.g., post-traumatic stress disorder [PTSD], depression) that
deleteriously affects physical health. US Latinos experience a high rate of cardiometabolic
conditions (e.g., type 2 diabetes, hypertension) and are disproportionally affected by poor
outcomes. Undocumented Latinos may be at increased risk for trauma exposure due to
unique stressors (e.g., migration). This cross-sectional quantitative study evaluated the
association between previous trauma exposure, current psychological distress, and self-
reported cardiometabolic conditions, among a sample of undocumented Latinos (n=254)
living in southern California and recruited through respondent driven sampling (69.3% female;
63% 18-40 years old; 97.6% Mexican; mean years living in US=16.48; 64.9% completed
middle school). Structured clinical interviews were conducted to assess for exposure to
potentially traumatic events relevant to immigrants and refugees (Harvard Trauma
Questionnaire, adapted), current PTSD and Major Depressive Disorder (MINI International
Neuropsychiatric Interview), and self-reported cardiometabolic conditions (type 2 diabetes,
hypertension, hyperlipidemia, and heart disease). Approximately 27% of participants reported
at least one cardiometabolic health condition. 49% had experienced confinement/robbery,
48.6% material deprivation, 44.7% a loved one injured due to violence, 45.1% a bodily injury
due to violence, 39.9% war conditions, 22% had experienced a deportation, and 13.4%
witnessed violence towards others. In a logistic regression model adjusting for demographic
covariates (age, sex, income, years in the US), having experienced any traumatic event
significantly increased the odds of reporting at least one cardiometabolic condition (OR=3.42
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2281

[1.20, 9.75]). In a final model that also accounted for PTSD and depression, trauma exposure
remained significantly associated with increased likelihood of cardiometabolic conditions
(OR=3.08 [1.06, 8.95]), while current PTSD and/or depression was not (OR=1.79 [0.82, 3.90]).
Results suggest that trauma exposure may be associated with poor cardiometabolic health
among Latinos, above and beyond the effects of psychological distress, which has important
clinical implications.

CORRESPONDING AUTHOR: Angela P. Gutierrez, MS, SDSU/UC San Diego Joint Doctoral
Program in Clinical Psychology, La Jolla, CA, 92037; [email protected]
S2282 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C292 6:00 PM-7:00 PM

CHANGE TALK, BEHAVIORAL WILLINGNESS, AND COLLEGE ALCOHOL USE: A MEDIATION

ANALYSIS

Benjamin Ladd, Ph.D.1, Kristen Anderson, Ph.D.2

1
Washington State University Vancouver, Vancouver, WA; 2Reed College, Portland, OR

Introduction: Change talk (CT), or self-motivational statements promoting health behavior,

predicts subsequent behavior change and is theorized to be an important mechanism of
change underlying psychotherapies, particularly motivational interviewing. However, the
mechanisms by which in-session language influences health behavior change is understudied.
The current study examined whether the relationship between CT and alcohol use is mediated
by behavioral willingness (BW), a proximal variable in the decision-making process. Method:
Participants (N = 76) were first-year college students assessed at the beginning of the
academic year (M = 10.7, SD = 7.8 days in college). Participants completed the Collegiate
Simulated Intoxication Digital Elicitation (C-SIDE), during which participants view five
simulated drinking scenarios. After each scene, participants reported their BW to accept each
offer of alcohol or food/nonalcoholic beverages (control offers) on a 5-point scale.
Participants verbal behavior was recorded using a free-response open elicitation at
predetermined points throughout the C-SIDE and then transcribed and coded for occurrences
of CT and its counterpart, sustain talk (ST). Percent change talk (PCT) was computed as CT/(CT
+ ST). Participant alcohol use was assessed at baseline and 8-month follow-up. Regression
analyses were conducted using the PROCESS macro in SPSS24. Results: Controlling for
baseline alcohol use, the total effect of PCT on alcohol use eight months later was significant,
B = -0.99, SE = 0.37, 95% CI (-1.73, -0.26). This effect was mediated by alcohol BW, such that
the direct effect of PCT on alcohol use was nonsignificant, B = 0.37, SE = 0.51, 95% CI (-0.64,
1.38) while the indirect effect through BW was significant, B = -1.37, SE = 0.36, 95% CI (-2.17, -
0.74). Discussion: The effect of PCT on subsequent alcohol use was significantly mediated by
BW. Greater PCT was associated with less alcohol use eight months later. This relationship is
explained by BW such that greater PCT led to lower BW which in turn resulted in lower
alcohol use. These results suggest the influence of client language on health behavior may
work through changes in more proximal factors of the alcohol use decision-making process.
This research may help researchers and clinicians begin to understand how client language in
non-therapeutic settings leads to desirable health behavior change in patients lives.

CORRESPONDING AUTHOR: Benjamin Ladd, Ph.D., Washington State University Vancouver,

Vancouver, WA, 98686; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2283

C293 6:00 PM-7:00 PM

EXAMINATION OF THE HEPATITIS C VIRUS CARE CONTINUUM AMONG INDIVIDUALS WITH AN

OPIOID USE DISORDER IN SUBSTANCE USE TREATMENT

Jennifer L. Brown, Ph.D., Theresa Winhusen, PhD, Nicole Gause, M.A., Daniel Lewis, BA

University of Cincinnati, Cincinnati, OH

Background: For individuals with an opioid use disorder (OUD), risk of Hepatitis C Virus (HCV)
transmission and acquisition is elevated, particularly through drug use behaviors including
injection drug use (IDU) or sharing needles and other injection paraphernalia. The CDC, U.S.
Preventive Services Task Force, and Institute of Medicine recommend routine HCV testing for
high-risk individuals, which includes those with a history of IDU. HCV antibody testing
addresses the first step in the HCV treatment care cascade. Uptake and completion of HCV
treatment among individuals with chronic HCV is the optimal endpoint of the care cascade;
however, there is some evidence to suggest that there may be gaps in the HCV care
continuum. Thus, the aim of the current study was to characterize the HCV treatment cascade
including: HCV testing, acute HCV diagnosis, chronic HCV diagnosis, and receipt of HCV
treatment among individuals with an OUD engaged in outpatient methadone or
buprenorphine assisted substance use treatment.

Methods:Individuals receiving methadone or buprenorphine outpatient treatment (N = 202,

67.8% female, M age = 35.0, SD = 8.4) completed a brief, anonymous paper-and-pencil survey
assessing history of HCV testing, diagnosis, and treatment.

Results: 65.1% reported either no prior HCV testing or not being tested for HCV during the
past year. 42.7% of those with a lifetime HCV testing history indicated they had been told they
had HCV (n=67/157), with 21% (n=14/67) of those individuals reporting that they had been
told they had chronic HCV, and 71.4% (n=10/14) of those with chronic HCV having received
HCV treatment.

Discussion: Results underscore gaps in the HCV care continuum among individuals with OUD
engaged in outpatient substance use treatment. Interventions to increase uptake of HCV
testing, linkage to HCV medical care, and uptake and adherence to HCV treatment are
urgently needed, particularly among individuals with an OUD.

CORRESPONDING AUTHOR: Jennifer L. Brown, Ph.D., University of Cincinnati, Cincinnati, OH,

45229; [email protected]
S2284 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C294 6:00 PM-7:00 PM

IMPLEMENTING AN SBIRT CURRICULUM ENHANCEMENT PROGRAM FOR SUBSTANCE USE

DISORDERS AT AN HBCU TEACHING HOSPITAL

Julie A. Dunlap, MSPH1, Heather O'Hara, MD, MSPH2

1
Meharry Medical College, Antioch, TN; 2Meharry Medical College, Nashville, TN

Background: Low-SES and racial/ethnic minority populations in the United States have a lower
probability of receiving needed substance-related services and are disproportionately affected
by health and social consequences of substance use disorders (SUD). Integrating Screening,
Brief Intervention and Referral to Treatment (SBIRT) training and best practices for SUD has
shown positive outcomes in primary care settings, particularly for low-SES populations. This
paper will introduce the research design and preliminary results of a SAMHSA-funded SBIRT
training program for medical students, residents and faculty at Meharry Medical College.

Methods: We have developed a graded level SBIRT training program across Family,
Occupational and Preventive Medicine disciplines and primary care clinics on MMC campus.
Pre and post data on SBIRT knowledge, attitudes, practices and training satisfaction are being
collected and analyzed to determine the impact of SBIRT training on improving clinician skills
and patient care outcomes.

Results: To date, 84 medical students and 26 residents have received a 2-hour didactic SBIRT
training and 24 additional residents (four who also received the 2-hour didactic) are currently
receiving more intensive 8-hour trainings. Results for 2-hour training show improved SBIRT
knowledge scores, 5.39 to 6.16 (p=0.002). Overall, training satisfaction, attitudes towards
working with patients who use drugs and alcohol, importance and confidence in using SBIRT
for substance use were high. Regarding clinical behaviors, participants reported using alcohol
and drug counseling interventions less than Half-the-Time (2.88). Preliminary results for
residents receiving the more intensive 8-hour training also show significantly improved
knowledge scores from 4.82 to 5.50 (p=0.019) and reported use of alcohol and drug
counseling interventions more than Half-the-Time" (3.41) in the clinic. The four participants
who have received continuous SBIRT training have reported the highest increase in use of
SBIRT practices in the clinic from Sometimes to Usually (2.17 to 4.19, p=0.033).

Discussion: This pilot presents a patient-centered and culturally-relevant model to integrate

SBIRT training and services for clinicians in primary care settings, particularly those serving at-
risk and underserved communities. Our results indicate some improvements in SBIRT
knowledge and practices, and overall high training satisfaction and attitudes. Future research
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2285

will analyze the effectiveness of more intensive trainings on SBIRT clinician practices and
patient care outcomes, comparing outcomes based on sex and discipline, and training
"champions" to sustain SBIRT practices across MMC clinics.

CORRESPONDING AUTHOR: Julie A. Dunlap, MSPH, Meharry Medical College, Antioch, TN,
37013; [email protected]
S2286 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C295 6:00 PM-7:00 PM

REDUCING DEPRESSIVE SYMPTOMS AMONG MOTHERS IN TREATMENT FOR OPIOID USE

DISORDER THROUGH MINDFULNESS BASED PARENTING

Vanessa L. Short, PhD, MPH, Meghan Gannon, PhD, MSPH, Diane Abatemarco, PhD, MSW

Thomas Jefferson University, Philadelphia, PA

Introduction: Many women with substance use disorders also have co-occurring mood
disorders, including depression which is associated with poor parenting outcomes and
impaired maternal functioning and bonding. Finding effective approaches to decreasing
depression among mothers with substance use disorders could mitigate its negative effects on
infants and children. Mindfulness has been shown to positively affect a variety of
psychological issues, including preventing relapse for depression, in the general population.
However, no studies have examined the effectiveness of mindfulness among mothers with
substance use disorder. Here we assessed whether a Mindfulness Based Parenting (MBP)
intervention was successful in decreasing self-reported depressive symptoms in a population
of parenting women receiving medication assisted treatment for opioid use disorder.

Methods:Data from 30 women with children less than three years of age who attended a two-
hour MBP session weekly for twelve consecutive weeks were collected. The Becks Depression
Inventory was administered prior to and following the MBP intervention period. Changes in
the total depression score from pre- to post-MBP were investigated using mixed-effects linear
regression models implemented via maximum likelihood accounting for correlations arising
from the repeated-measures. Analyses were also performed to investigate differences by
baseline total depression score.

Results: Among all participants, total depression scores decreased significantly from baseline
to post-intervention (mean difference=-3.57, SD 6.24, p=0.04). Additionally, those with the
highest baseline depression scores experienced the greatest change (mean difference=-8.3,
SD 9.9, p=0.008).

Conclusions:Mindfulness Based Parenting is an innovate program with potential important

application to mothers engaged in treatment for substance use disorder. Participants of an
MBP intervention reported fewer depressive symptoms post-intervention compared to
baseline suggesting that MBP could be used as a mental health-improving strategy for this
population. Such improvements in maternal mental well-being could have significant and
lasting impacts on infant and child health and development, and future MBP interventional
studies are warranted.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2287

CORRESPONDING AUTHOR: Vanessa L. Short, PhD, MPH, Thomas Jefferson University,

Philadelphia, PA, 19107; [email protected]
S2288 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C296 6:00 PM-7:00 PM

TREATMENT MOTIVATION AMONG CAREGIVERS AND ADOLESCENTS WITH SUBSTANCE USE

DISORDERS

Talea Cornelius, M.A., M.S.W.

University of Connecticut, Vernon, CT

Substance use disorders (SUDs) in adolescence have negative long-term health effects,
which can be mitigated through successful treatment. Caregivers play a central role in
adolescent treatment involvement; however, studies have not examined treatment
motivation and pressures to enter treatment in caregiver/adolescent dyads. Research
suggests that internally motivated treatment (in contrast to coerced treatment) tends to lead
to better outcomes. We used Self-Determination Theory (SDT) to examine intersecting
motivational narratives among caregivers and adolescents in SUD treatment. Relationships
between motivation, interpretation of caregiver pressures, adolescent autonomy, and
relatedness were also explored. Adolescents in SUD treatment and their caregivers
(NDyads=15) were interviewed about treatment experiences. Interviews were coded for
treatment motivation, including extrinsic (e.g., motivated by punishment), introjected (e.g.,
motivated by guilt), and identified/integrated motivation (e.g., seeing a behavior as integral to
the self). Internalization of treatment motivation, autonomy support/competence (e.g.,
caregiver support for adolescent decisions), and relatedness (e.g., acceptance and support)
were also coded. Four dyadic categories were identified: agreement that treatment was
motivated by the adolescent (intrinsic); agreement that treatment was motivated by the
caregiver (extrinsic); agreement that treatment was motivated by both, or a shift towards
adolescent control (mixed/transitional); and disagreement (adolescents and caregivers each
claimed they motivated treatment; conflicting). Autonomy support and relatedness were
most prominent in intrinsic dyads, and least prominent in extrinsic dyads. The
mixed/transitional group was also high in autonomy support and relatedness. The extrinsic
group characterized caregiver rules as an unwelcome mechanism for behavioral control;
caregivers in the other groups saw rules as a way to build adolescent competence and repair
relationships, and adolescents saw rules as indicating care rather than control. Adolescents
with intrinsic motivations were the most engaged in treatment. Results suggest the
importance of intrinsically motivated treatment, and highlight autonomy support and
relatedness as mechanisms that might facilitate treatment engagement.

CORRESPONDING AUTHOR: Talea Cornelius, M.A., M.S.W., University of Connecticut, Vernon,

CT, 06066; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2289

C297 6:00 PM-7:00 PM

USING A FAMILY SYSTEMS FRAMEWORK TO ASSESS THE IMPACT OF FAMILIAL OPIOID MISUSE
ON THE OVERALL WELL-BEING OF YOUNG ADULTS

Drishti R. Shah, MS, B.Pharm, Khalid Alhussain, Pharm D, James Douglas. Thornton, PharmD,
Kimberly M. Kelly, PhD, MS

West Virginia University, Morgantown, WV

Objective: Misuse of opioids has grown in epidemic proportions over the last decade in the
US, particularly in the Appalachian region. Considering that misuse of opioids may significantly
affect the family environment, we sought to understand overall well-being of young adults in
the context of familial opioid misuse and misuse of other substances using a family systems
framework (family organization, health-related beliefs, and communication within the family).
Methods: We conducted a cross-sectional, mixed-methods survey of young adults (18-30
years old) (n=194). Consistent with a family systems framework, we developed a scale
assessing health-related beliefs about familial opioid misuse (Cronbachs alpha=0.72). The
associations between the overall well-being of young adults and familys cohesiveness, level
of coping, frequency of communication with immediate family members and knowledge
about opioid misuse in Appalachia were assessed using multivariable linear regression.
Quantitative data was augmented by narrative analysis of responses of familial substance
abuse-related experiences by young adults (n=60).
Results: Our sample was mostly women (77%) with a mean age of 24.9 years, living in the
Appalachian region (82%), who communicated with immediate family members at least
weekly (84%). Nearly one fifth (17.5%) reported having familial opioid misuse, all of whom
reported that opioid misuse affected familial communication, and 43.3% reported having no
familial substance misuse. More than half of the study sample had heard a lot about opioid
misuse in Appalachia (53%) and thought that their family could cope with opioid misuse
(66%). Both adjusted (F (10,194) =3.7, p=.001) and unadjusted analyses showed that health-
related beliefs about familial opioid misuse (=0.31, p=0.01), familial coping (=0.18, p=0.04),
and misuse of any substance by family members (=1.14, p=0.01) were significantly
associated with overall well-being. Qualitative data also indicated the family environment was
more disrupted by familial alcohol misuse rather than opioid misuse. Family members
misusing opioids seem to be more isolated and disconnected from their family than those
with familial alcohol misuse.
Conclusion: While familial opioid misuse did not seem to impact overall well-being, familial
misuse of other substances, especially alcohol, seemed to be more disruptive for the overall
well-being of young adults in our largely Appalachian sample.
S2290 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Drishti R. Shah, MS, B.Pharm, West Virginia University,

Morgantown, WV, 26505; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2291

C298 6:00 PM-7:00 PM

A DESCRIPTIVE ANALYSIS OF UNEMPLOYED, JOB-SEEKING SMOKERS IN CALIFORNIA

Adrienne Lazaro, BS1, Peter Soyster, BS2, Nicole E. Anzai, BS3, Sarah Stinson, BA4, Anne
Michalek, BA4, Judith J. Prochaska, PhD, MPH3

1
Stanford Prevention Research Center, Palo, CA; 2Department of Psychology, University of
California, Berkeley, Stanford, CA; 3Stanford University, Stanford, CA; 4Stanford Prevention
Research Center, Stanford University, Stanford, CA

Background: Our recent research found that job-seeking smokers were half as likely as
nonsmokers to be re-employed after one year. Among those who re-gained employment,
smokers earned, on average, $5 less per hour than nonsmokers. These findings led to a
randomized controlled trial to test a tobacco intervention in unemployment service settings. A
vulnerable, impoverished group, we characterize the study sample at baseline.

Method: Participants were job-seeking smokers from the San Francisco Bay Area, working
under 10 hours/week, unemployed for less than 2 years, currently smoking 1 or more
cigarettes/day with an expired CO sample of at least 7 ppm and having smoked 100 or more
cigarettes in ones lifetime. Intention to quit smoking was not required as the intervention
was tailored to readiness to quit. Of 419 job-seeking daily smokers screened, 217 met
eligibility (52%) and N=206 have enrolled into the trial (5% declined).

Results: The sample is 64% male, 28% female, and 8% transgender, with a mean age of 44
years (SD=11), identifying as 42% Black, 27% non-Hispanic White, 11% Hispanic, and 20%
other. Time out of work averaged 6 months (SD=6); 19% are unhoused. Most have a high
school diploma or GED (84%); 22% have a college degree; 19% have a criminal history
reportable on a job application. Half (51%) interviewed for jobs they did not get in the past 12
months; the most common reported barriers to finding work were advanced age and criminal
history. Tobacco use averaged 12 cigarettes/day (SD=6.2) with 44% reporting past-year use of
cigars/cigarillos, 29% e-cigarettes, and 25% blunts. Most (75%) reported a 24-hr quit attempt
in their lifetime and 49% in the past year. The most common quitting methods were cold
turkey (57%), followed by gradual reduction of use (37%) and nicotine replacement therapy
(24%). At study enrollment, 34% intended to quit in the next 30 days, 36% in the next 6
months, and 30% not in the near future.

Conclusion: With a focus on unemployed smokers, study of novel tobacco treatment outreach
strategies may be especially beneficial given the groups general low intention to quit, poly-
S2292 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

tobacco use, and reliance on cold turkey quitting methods. Additional cessation support and
pharmacotherapy may augment quit efforts and, in turn, aid re-employment.

CORRESPONDING AUTHOR: Adrienne Lazaro, BS, Stanford Prevention Research Center, Palo,
CA, 94305-5411; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2293

C299 6:00 PM-7:00 PM

A NOVEL MOBILE APP FOR YOUNG ADULT SMOKING CESSATION: A PILOT STUDY

Megan Jacobs, MPH, Sarah Cha, MSPH, Amanda Graham, PhD

Truth Initiative, Washington, DC

Young adult (YA) smokers are challenging to reach and less likely to engage with evidence-
based cessation interventions than older adults. Mobile interventions leveraging the
stickiness of social media may be an innovative approach to engage YA smokers. This pilot
study evaluated signals of engagement with such a novel smoking cessation app.

The This is Quitting app was developed based on formative research with YA current and
former smokers, including monitoring social media to determine how YA smokers talk about
quitting; key informant interviews/focus groups; feedback on sketch concepts; beta tests; and
usability testing. Main components of the app are: social media content about quitting
smoking curated from multiple platforms (i.e., Tumblr, Twitter) into an aggregated feed;
interactive tools (training packs) that list challenges to prepare users to stay smoke free in
various situations; and information on evidence-based resources such as NRT and quitlines.
App users were recruited via organic social media, paid online ads, and referral from YA
organizations (e.g., universities). Engagement with the app was tracked via automated
tracking tools and examined with descriptive statistics.

Between 12/28/15-1/17/16, 810 users installed the app. During the first 60-days after
installing the app, 355 (43.8%) users opened the app more than once (M=3.3, SD=2.7 among
return users). Median number of social media posts viewed was 20 (IQR=10-30). About half of
the users viewed at least one training pack (360, 44.4%) and more than a quarter of the users
completed at least one challenge (220, 27.2%). Fewer users accessed static content: 109
(13.5%) read about NRT and 24 (3.0%) viewed quitline information.

This pilot study demonstrated significant demand for a smoking cessation app with over 800
installations in 20 days. Engagement metrics were promising with nearly half of the users
opening the app more than once (previous research showed about 1/3 repeat users for
mobile health apps). Static content was accessed at lower rates than the social media-based
content of the app. Future research should explore increasing adoption/engagement with
mobile apps and whether this modality may promote abstinence.

CORRESPONDING AUTHOR: Megan Jacobs, MPH, Truth Initiative, Washington, DC, 20001;
[email protected]
S2294 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C300 6:00 PM-7:00 PM

DEVELOPMENT OF A SMARTPHONE-BASED MINDFULNESS INTERVENTION WITH

CONTINGENCY MANAGEMENT FOR OUTPATIENT DEPRESSED SMOKERS

Haruka Minami, Ph.D.1, Hannah R. Brinkman, B.A.2, Shadi Nahvi, M.D.3, Julia H. Arnsten, M.D.,
M.P.H.3, Monica Rivera-Mindt, Ph.D., ABPP-CN4, David W. Wetter, Ph.D.5, Lawrence H. Price,
M.D.6, Lauren M. McClain, B.A.4, Carlos Vieira, .7, Remington E. Donnelly, .4, Katherine A.
Kennedy, B.A.4, Erica D'Aquila, MPH8, Jacki Hecht, R.N., M.S.N.9, Erika Litvin Bloom, Ph.D.10,
Richard A. Brown, Ph.D.9

1
Fordham Univeristy, White Plains, NY; 2Fordham University, New City, NY; 3Albert Einstein
College of Medicine, Montefiore Medical Center, Bronx, NY; 4Fordham University, Bronx, NY;
5
University of Utah, Salt Lake City, UT; 6Alpert Medical School of Brown University, Butler
Hospital, Providence, RI; 7Fordham Univeristy, Bronx, NY; 8ICAP at Columbia University, New
York, NY; 9The University of Texas at Austin, Austin, TX; 10Alpert Medical School of Brown
University, Rhode Island Hospital, Providence, RI

Smoking rates among individuals with psychiatric disorders are disproportionately higher than
in the general population and in turn this group bears a large proportion of the tobacco
morbidity and mortality burden. However, less than 20% of outpatient mental health
treatment facilities offer smoking cessation treatments. Development and evaluation of
smoking cessation interventions that could be integrated into psychiatric treatment settings is
needed. In the current pilot study, we developed a smartphone-based mindfulness smoking
cessation intervention with contingency management (MSI-CM) and delivered it to 8 adult
daily smokers who were receiving outpatient psychiatric treatment for depression in Bronx,
NY. Of the 8 participants, 100% were female and racial/ethnic minorities (50% Hispanic),
87.5% had a high school diploma/GED or less, 100% reported household incomes of less than
$24,999, and 75% were on disability. Half of participants reported that they have only one or
no prior serious quit attempts. Over a 38-day period (10 days before and 4 weeks after a
target quit date), participants were prompted to practice mindfulness by listening to audio
recordings (approx. 5 mins each) and to complete ecological momentary assessments using a
smartphone 5 times per day, and to submit carbon monoxide (CO) videos twice a day for 2
weeks after a target quit date. Overall, participants practiced mindfulness an average of 3.4
times/day ( 3 minutes), 68% of requested CO videos were submitted, and 58% of those
submitted earned cash incentives (CO 6 ppm). Participants completed 72.3% of prompted
ecological momentary assessments, with 97% of reports completed within 30 minutes of the
scheduled time. At each follow-up, 1 participant (12.5%) reported biochemically verified 7-day
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2295

point prevalence abstinence. All participants reported reductions in the number of cigarettes
smoked per day from baseline (M=12.5) to 2-week, 4-week, and 3-month post-quit follow-ups
(M=2.0, 2.0, 7.2; 84%, 84%, and 46% reduction, respectively). Participants reported that the
program was helpful overall (M=4.85, on 5-pt scale with 1= not at all and 5= very much), daily
mindfulness practice was enjoyable and helpful for both managing mood and quitting
smoking (M=4.5-4.63), and submitting videos of CO levels aided in the cessation effort
(M=4.75). These findings indicate high levels of acceptability and satisfaction with MSI-CM.
We are currently in the preliminary randomized controlled trial phase.

CORRESPONDING AUTHOR: Haruka Minami, Ph.D., Fordham Univeristy, White Plains, NY,
10603; [email protected]
S2296 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C301 6:00 PM-7:00 PM

DISCOVERING DIMENSIONS OF DEPENDENCY AMONG YOUNG ADULT ELECTRONIC NICOTINE

DELIVERY SYSTEMS USERS

Erica R. Nahin, B.A., B.S.1, Jordan A. Baeker Bispo, MPH2, Maria M. Llabre, PhD1, Alexander W.
Sokolovsky, MA3

1
University of Miami, Coral Gables, FL; 2University of Miami, Miami, FL; 3Brown University,
Providence, RI

Electronic Nicotine Delivery Systems (ENDS) use is becoming increasingly common among
young adults, necessitating an understanding of the salient dimensions of nicotine
dependence that may underscore this novel public health phenomenon. The purpose of this
study was to evaluate dimensions of dependency within the young adult ENDS user
population. Data come from the second phase of an ongoing three phase study which aims to
investigate ENDS use patterns and motives of continued use among young adults (ages 18-
24). Participants in this sample (N = 304), were predominantly male (61.5%) and Caucasian
(69.1%). Each participant self-identified as a regular ENDS user and self-reported using within
a week of data collection. Participants were recruited using Amazon Mechanical Turk, an
online micro-task marketplace, and were asked to self-report their demographics, smoking
patterns, as well as complete a modified 14-item Wisconsin Inventory for Smoking
Dependency Motives (WISDM) adapted to query ENDS dependence motives. We used
confirmatory factor analysis to model six WISDM sub-scales as indicators of dependence,
including automaticity, loss of control, tolerance, craving, cognitive enhancement, and
affective enhancement. Subscale scores were indexed as the mean of the items composing
the scales. Both a one- and two-factor solution for dependence were compared, where
refinements to model specification were made as appropriate upon inspection of factor
loadings, residuals, and modification indices. The two-factor model best fit the data.
Automaticity, loss of control, craving, and tolerance loaded on the first factor and represent
physiological drive. Drive can be conceptualized as the desire to satisfy the urge to use ENDS.
Affective and cognitive enhancement loaded onto the second factor and represent positive
reinforcement. These data demonstrate that there may be two distinct clusters of factors
driving dependence among ENDS users.

CORRESPONDING AUTHOR: Erica R. Nahin, B.A., B.S., University of Miami, Coral Gables, FL,
33134; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2297

C302 6:00 PM-7:00 PM

EFFECTS OF ANHEDONIA ON HOOKAH USE AMONG YOUNG ADULTS

Angela Petersen, PhD, Kristin Brikmanis, BA, Neal Doran, PhD

University of California, San Diego, La Jolla, CA

Introduction: Use of hookah among young adults has increased over recent years, yet little is
known about how depressive constructs, such as anhedonia, are related to hookah use.
Anhedonia predicts poorer outcomes with various aspects of cigarette smoking, particularly
cessation, and may similarly affect other tobacco use. The purpose of the present study was
to examine the association between anhedonia and hookah use over time among young
adults.

Methods: Participants (n = 392, 52% male) aged 18-24 were recruited for a longitudinal study
examining cigarette and alternative nicotine and tobacco product use. Eligibility criteria
included non-daily cigarette smoking for > 6 months and no lifetime history of daily cigarette
smoking for one month. Anhedonia was measured at baseline and participants reported
hookah use at baseline, 3, 6, and 9 months. All measures were completed online or via
mobile app.

Results: Anhedonia was prospectively associated with frequency of hookah use such that
more anhedonic individuals reported more frequent use of hookah tobacco. Hookah use was
reported by 14-32% of participants across the 4 assessments, and frequency of use
significantly decreased over time.

Conclusion: Initial findings suggest that anhedonia predicts increased use of hookah among
young adult non-daily smokers. More research is needed to identify mechanisms that link
anhedonia and hookah use, as well as longer-term outcomes such as dependence and
cessation.

CORRESPONDING AUTHOR: Angela Petersen, PhD, University of California, San Diego, La Jolla,
CA, 92037; [email protected]
S2298 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C303 6:00 PM-7:00 PM

GREATER SUSCEPTIBILITY TO NICOTINE PATCH TOXICITY AMONG ASIAN AMERICANS

COMPARED TO CAUCASIANS

Alina M. Ostrander, B.A., Jennifer Hernandez, Julia B. Yi, B.A. , Larry D. Jamner, Ph.D.

University of California, Irvine, Irvine, CA

Smoking rates for Asian Americans is about half that shown by White Americans with those
Asian American who do smoke also more likely to be light and intermittent cigarette smokers
compared to Whites. A plausible explanation for the reduced consumption of cigarettes
among Asians may be differences in nicotine metabolism. Benowitz et al. (2002) have shown
Chinese-Americans to demonstrate substantially lower CYP2A6 liver enzyme activity than
Whites or Latinos; and it is this slower nicotine metabolism that may place Asians at greater
risk for nicotine toxicity. Few studies have compared nicotine toxicity sensitivity between
Asians and Caucasians in response to fixed doses of nicotine under controlled conditions. In
this study 21mg and 5mg nicotine patches (Nicoderm) were administered to smokers and
non-smokers respectively, as well as a placebo patch, across two laboratory sessions.
Participants were monitored for physical symptoms associated with nicotine toxicity and
sessions were terminated as soon as participants reported becoming sick. The sample
included 48 smokers and 23 non-smokers who were Asian American, and 95 smokers and 47
non-smokers who were Caucasian. Results revealed that among smokers, only 6% of
Caucasian in contrast to 31% of Asian American smokers became sick from the nicotine patch
(Fishers Exact Test, p=.003). Among non-smokers, only 2% of Caucasians reported sickness
compared to 22% of Asian Americans (Fishers Exact Test, p=.013). The similarity of nicotine
side-effects patterns observed in both smokers and non-smokers suggest that the findings
cannot be attributed simply to increased tolerance in smokers. The greater susceptibility of
Asian Americans to nicotine toxicity is consistent with models of reduced nicotine metabolism
in this ethnic group and affords an explanation for the historically lower cigarette smoking
prevalence rates observed in Asian Americans. An important implication of these findings
relate to the relatively recent explosive popularity of electronic nicotine delivery systems
(ENDS). Specifically, the greater titration control of nicotine exposure levels made possible by
ENDS may pose disproportionately higher risks of these devices acting as gateways towards
developing nicotine dependence among Asian Americans.

CORRESPONDING AUTHOR: Alina M. Ostrander, B.A., University of California, Irvine, Irvine,

CA, 92617; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2299

C304 6:00 PM-7:00 PM

LIMITATIONS OF RISK PERCEPTION AND IMPLICATIONS FOR DUAL PROCESS THEORIES OF

BEHAVIOUR.

Ron Borland, PhD1, Bill King, MSc2, Hua Yong, PhD3

1
Cancer Council Victoria, Melbourne, Victoria, Australia; 2Cancer council Victoria, Melbourne,
Victoria, Australia; 3Cance Council Victoria, Melbourne, Victoria, Australia

The limited capacity of people to make accurate quantitative estimates of risk has been
known for some time. However, the extent of such limitations and their implications for
making sensible behaviour choices is less well understood. CEOS theory predicts that
affective factors will swamp more rational ones unless rational reasoning is supported. This
raises the question of how can we strengthen capacity to make sensible choices in the context
of affective forces pointing in the other direction. The aim of this study was to both quantify
and use qualitative interviews to better understand the ways in which smokers understand
the risks of various tobacco and nicotine products and the implications of such beliefs for
making rational choices about their smoking.

Methods. Qualitative interviews and focus groups with smokers or recent quitter and a survey
of two overlapping samples of smokers and young adults (under 35) resulting in
approximately equal groups of young smokers, young non-smokers and older smokers (total
n=1056). The quantitative survey asked for semi-quantitative estimates of risk of 4 aspects .

Results: Only 1.4% of the sample correctly identified the effects of combustion as the primary
cause of harm and did not nominate multiple aspects as contributing to more than half of the
harms (ie arithmetically impossible). Nearly all judgements are qualitative, so a new variable
was created with thos correct plus those indicating smoke was thehighest source of harm
combined giveing 4.5% of the sample. There was a a generalised negative reaction to all
aspects of tobacco, and no differentiation, in particular of the lower harms of the nicotine in
isolation (only 10.8% rated nicotine in the lowest category of harm . The interviews found
that most smokers interviewed could talk about the harms in relatively accurate ways, but
were unable to use this information in making judgements of risk. There were no clear age
effects and minimal effects of smoking status, with small effects suggesting non-smokers were
better informed.

Conclusions. Risk assesments are qualitatie, not empirical, so emotinal responses to risk need
to be taken into account before attempting to change beliefs about extent.This confusion in
S2300 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

understanding may be inhibiting moves to less harmful forms of nicotine. If people are to
make more rational choices about risk, communication about risk needs to focus on a
separation between acceptable and unacceptable risks and educate people to have a better
understand of the core determinants of health risk from various types of agent. CEOS theory
provides a framework for improving communication and enhancing executive capacities to
overcome the identified affective influences and allow more evidence-based judgements to
influence decisions..

CORRESPONDING AUTHOR: Ron Borland, PhD, Cancer Council Victoria, Melbourne, Victoria,
3004; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2301

C305 6:00 PM-7:00 PM

PARENTAL DEPRESSION IS ASSOCIATED WITH LOWER LIKELIHOOD OF SMOKING CESSATION

AND POOR CHILD ASTHMA OUTCOMES: PROSPECTIVE STUDY

Romano Endrighi, PhD1, Elizabeth L. McQuaid, PhD, ABPP2, Kiera Bartlett, PhD3, Ashley H.
Clawson, Ph.D.4, Belinda Borrelli, PhD5

1
Boston University Henry M. Goldman School of Dental Medicine, Boston, MA; 2Alpert
Medical School, Brown University, Providence, RI; 3The University of Manchester,
Manchester, England, United Kingdom; 4Oklahoma State University, Stillwater, OK; 5Boston
University, Henry M. Goldman School of Dental Medicine, Boston, MA

Parents of children with asthma have higher rates of depression than parents of healthy
children. Given that those who are depressed are more likely to smoke and have difficulty
quitting, and that secondhand smoke (SHS) triggers asthma exacerbations, we examined the
effect of parental depression on smoking cessation and child asthma outcomes.
This study is a secondary analysis from a cessation trial involving Motivational Interviewing
(MI) and SHS biofeedback to motivate cessation. Parents did not have to be motivated to quit
to enroll. Parents (n=341, 79.8% female) of a child (Mean age=5.154.5 years) who had an
asthma exacerbation requiring urgent care received 2 home-based asthma education visits
(plus MI and SHS biomarker feedback to motivate cessation). Parents were then randomized
to receive 6 counselling calls (MI plus repeated SHS feedback) or 6 control calls (brief check on
asthma status) for 4 months. Depression status was defined as a score of 22 (Haringsma et
al, 2004) on the Center for Epidemiologic Study-Depression scale. Smoking outcomes were
biochemically verified 7-and 30-day point-prevalence abstinence (ppa). Child asthma
outcomes were past month asthma functional limitation scores, health care utilization (1
urgent care or hospitalization event vs none) and asthma symptom days. Data were obtained
at baseline, 2-4-and-6-months and analyzed with longitudinal generalized estimating
equations models adjusting for covariates.
Depressed parents were less likely to achieve 7-day ppa (OR=.38, 95% CI=.23, .64) and 30-day
ppa (OR=.27, 95%CI=.15, .47) than parents who were not depressed. Effects were consistent
overtime (no group by time interaction). Parental depression predicted greater odds of health
care utilization for the childs asthma exacerbation (OR=1.85, 95%CI=1.00, 3.44), and greater
asthma functional limitation (B=.13, SE=.06, p=.03), even after controlling for smoking status
and SHSe during the study. Depression also predicted a greater number of asthma symptom
days (B=.87, SE=.44, p=.05), but became non-significant after controlling for parent smoking
status (30-day ppa). Results suggest that parent depression plays a greater role that smoking
in some asthma outcomes, but abstinence reduces child asthma symptom days beyond the
S2302 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

effect of depression.
Findings suggest that, among parents who smoke, depression as well as their smoking, should
be a target for treatment aimed at improving pediatric asthma outcome.

CORRESPONDING AUTHOR: Romano Endrighi, PhD, Boston University Henry M. Goldman

School of Dental Medicine, Boston, MA, 02118; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2303

C306 6:00 PM-7:00 PM

PATTERNS AND REASONS FOR ELECTRONIC CIGARETTE USE IN PRIMARY CARE PATIENTS

Sara Kalkhoran, MD, MAS1, Patrick Yuan, B.A.2, Nicholas J. Alvarado, MPH3, Jason Satterfield,
PhD4

1
Massachusetts General Hospital/Harvard Medical School, Boston, MA; 2University of
California, San Francisco, San Francisco, CA; 3Project Manager/Clinic Research Coordinator,
San Francisco, CA; 4UCSF, San Francisco, CA

Introduction: E-cigarette use is rising in both the general population and in clinical
populations, including hospitalized patients and cancer patients. Little is known about e-
cigarette use among patients in other clinical settings, particularly in primary care where
physicians report discussing e-cigarette use with patients. We aimed to identify how and why
smokers in primary care use e-cigarettes.

Methods: We used cross-sectional data from current cigarette smokers (within the past 7
days) enrolled in a randomized controlled trial of a tablet intervention to deliver the 5As for
smoking cessation in primary care clinics in San Francisco, CA (N=718). We recruited patients
from 3 sites: a primary care clinic at a University hospital, and a primary care clinic and HIV
care clinic at a University-affiliated public safety-net hospital. We asked patients about
sociodemographics, cigarette smoking, and ever and current use of e-cigarettes during the
post-intervention assessment. We also asked reasons they have used or would use e-
cigarettes. ICD-9 codes for comorbidities were extracted from electronic medical records.

Results: 57% (n=408) of patients reported ever using an e-cigarette and 21% (n=154) reported
past 30 day use. Ever e-cigarette users were more likely to be younger, white, more educated,
daily smokers, have smoked for more years, and more nicotine dependent than never users.
Ever e-cigarette users had higher prevalence of mental illness than never users. The most and
second most common reasons for e-cigarette use among current and former e-cigarette users
were to cut down on and quit cigarette smoking, respectively. These were also the most
common reasons never e-cigarette users reported for why they would use e-cigarettes. The
mean number of days of e-cigarette use in the past 30 was 7, increasing with duration of e-
cigarette use. Most current e-cigarette users did not know the nicotine content of their e-
cigarettes.

Conclusion: Over half of smokers in primary care use e-cigarettes, with 1 in 5 reporting
current use. E-cigarette use is intermittent in this population, with most reporting non-daily
S2304 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

use. Most patients report using e-cigarettes to cut down or quit cigarettes. Screening of
patients for e-cigarette use may help identify those interested in changing their smoking
habits and provides an opportunity for providing evidence-based smoking cessation treatment
and counseling.

CORRESPONDING AUTHOR: Sara Kalkhoran, MD, MAS, Massachusetts General

Hospital/Harvard Medical School, Boston, MA, 02114; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2305

C307 6:00 PM-7:00 PM

SECONDHAND SMOKE INTERVENTION IN LAKOTA FAMILY HOUSEHOLDS

Janet Thomas, PhD1, Priscilla Nez, BS2, Jeffrey A. Henderson, MD, MPH 2, Dedra Buchwald,
professor, Elson S Floyd College of Medicine3, Carolyn Noonan, MS3, Patricia Nez Henderson,
MD, MPH2, Meredith Schreier, MHA4

1
University of Minnesota, Mpls, MN; 2Black Hills Center for American Indian Health, Rapid
City, SD; 3Washington State University, Seattle, WA; 4University of Minnesota, Eden Prairie,
MN

Secondhand Smoke Exposure Intervention in Lakota Family Households

Background: Nonsmokers from the Lakota tribes in South Dakota disproportionately share the
burden of secondhand smoke exposure (SHS), especially in their homes. With smoking
prevalence as high as 50%,the consequences of SHS include high prevalence of childhood
asthma and sudden infant deaths. The expanded use of commercial tobacco products for
ceremonial and cultural practices creates intervention challenges not typically seen in non-
American Indian communities. Eliminating SHS in a culturally appropriate manner is
fundamental to decreasing morbidity and mortality among the Lakota tribes. A novel
approach to encourage the adoption of home smoking restrictions is to provide parents with
objective, biomarker feedback documenting child exposure to tobacco toxins. This pilot
project is the first to use biomarker feedback with American Indian adult smokers.

Methods: From 2013-2015 we recruited American Indian smokers with potty-trained children
ages 2 to 8 residing in their homes into a two-arm randomized clinial trial (N=110). We
randomized eligible participants to the intervention arm (n=56), consisting of biomarker
feedback and counseling on their childs exposure to NNAL and cotinine. The control group
(n=54) received a brochure on SHS. An in-home survey was administered at baseline and week
12. Child urine cotinine and air nicotine were collected at both time points. Primary outcome
was badge-verified (air nicotine) complete home smoking restrictions at 3-months post-
randomization.

Results: All enrolled participants were American Indian, 81% female, average age 37 (SD=12)
years, 62% at least high school graduates, 65% unemployed, and 81% had an annual
combined family income less than $25,000. The mean number of smokers/household was 2.4
S2306 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

(SD=1.3) and average cigarettes smoked/day was 8.8 (SD=6.7). Median child cotinine level
was 7 ng/mL (interquartile range: 3-22). At week 12, a t-test comparison indicated that the
intervention and control groups had a similar reduction in air nicotine (change score in
intervention = -1.02, 95% CI -1.80, -0.24; change score in control = -0.99, 95% CI -1.81, -0.17),
with no changes in child urine cotinine. However, at week 12, more participants assigned to
biomarker feedback than the control group self-reported implementing complete home
smoking restrictions (41% vs. 20%, p=0.05).

Conclusion: Providing parents or guardians with feedback documenting their childs exposure
to SHS increased the implementation of home smoking bans. Clinical, public health, policy and
research implications will be discussed.

CORRESPONDING AUTHOR: Janet Thomas, PhD, University of Minnesota, Mpls, MN, 55414;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2307

C308 6:00 PM-7:00 PM

SMOKING-RELATED HEALTH BELIEFS AND SMOKING CESSATION IN THE NATIONAL LUNG

SCREENING TRIAL

Annette Kaufman, PhD, MPH1, Laura Dwyer, PhD2, Stephanie Land, PhD3, William Klein, PhD4

1
Tobacco Control Research Branch, Behavioral Research Program, Division of Cancer Control
and Population Sciences, National Cancer Institute, National Institutes of Health, Rockville,
MD; 2Cape Fox Facilities Services, Manassas, VA, Rockville, MD; 3Tobacco Control Research
Branch, Behavioral Research Program, Division of Cancer Control and Population Sciences,
National Cancer Institute, Rockville, MD; 4NCI/NIH, Rockville, MD

Lung cancer is the leading cause of cancer death in the United States and 90% of lung cancers
are caused by cigarette smoking. Understanding the association between smoking-related
health beliefs and smoking cessation in the screening environment is important for prevention
and effective treatment. We conducted a sub-study of the National Lung Screening Trial
(NLST; 2002-2006) to understand how smoking-related health beliefs may be related to
smoking cessation.

Participants (N=2,761) at the time of recruitment were 55-74 years of age, current smokers
with a history of 30 pack-years or more, had no history of lung cancer, and had not been
treated for any cancer within the past 5 years. All trial enrollees at 8 of the 23 NLST study sites
were invited to complete a questionnaire to assess beliefs: absolute risk perception,
comparative risk perception, worry, perceived severity, self-efficacy to quit, perceived
benefits of quitting, and perceived barriers to quitting. The outcome was cessation between
the first and last assessment. Logistic regression analyses examined associations with beliefs,
demographics, and all two-way interactions among beliefs, gender, and age. We applied
backward elimination of terms with p-values =0.5.

Over 37% (N=1028) of baseline smokers quit smoking at last assessment. In simple logistic
regressions, the likelihood of quitting was higher among participants reporting higher
perceived severity (OR=1.16, p=0.05), greater self-efficacy for quitting (OR=1.32, pp=0.05),
fewer perceived barriers (B=-0.22, p=.01), and were older at baseline (B=0.03, p=0.00) were
more likely to quit. Worry interacted with comparative risk perception (B=-0.19, p=0.04), and
perceived benefits (B=0.19, p=0.05). Higher worry attenuated the relationship between
comparative risk perception and quitting, but perceived benefits predicted increased odds of
quitting at the highest level of worry.
S2308 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

These results suggest that, among heavy smokers undergoing lung screening, smoking-related
health beliefs are associated with smoking cessation, and worry is an important moderator of
these associations. Smoking cessation has an impact on lung cancer survival and smoking-
related health beliefs must be considered an integral component of delivering cessation in
screening settings.

CORRESPONDING AUTHOR: Annette Kaufman, PhD, MPH, Tobacco Control Research Branch,
Behavioral Research Program, Division of Cancer Control and Population Sciences, National
Cancer Institute, National Institutes of Health, Rockville, MD, 20850; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2309

C309 6:00 PM-7:00 PM

SOCIAL CONTEXTS, ETHNICITY, AND ADOLESCENT SMOKING BEHAVIOR

Julia B. Yi, B.A. 1, Alina M. Ostrander, B.A.1, Jennifer Hernandez1, Bryan C. Mac, B.A.2, Larry D.
Jamner, Ph.D.1

1
University of California, Irvine, Irvine, CA; 2University of California, Irvine, Seal Beach, CA

Cigarette smoking in adolescence may be motivated by the desire to fit in and avoid negative
evaluation from peers. Adolescents may also feel pressure to smoke if others around them
are smoking. Social stimulus control over smoking has been shown to vary as a function of
culture, with Korean-Americans reporting greater social determinants of cigarette use. The
use of Ecological Momentary Assessment (EMA) methods to evaluate the antecedents,
concomitants, and consequences of tobacco use holds great promise for the study of the
cultural variations in smoking behavior. The aim of this study was to determine the impact of
social environments/contexts on smoking urge and likelihood to smoke in real-time among
Asian American, Latino, and White adolescents. Eighty-five female and 70 male adolescents
(20% Asian American, 11% Latino, 51% White) completed two 4-day daily diary-monitoring
sessions approximately 6 months apart, each year from ninth grade until twelfth grade,
totaling eight monitoring periods. In each monitoring session, participants were signaled
every 30 +10 minutes to complete an e-diary entry. The diary contained 24 items tapping
information about current activities, social interactions, moods, urge to smoke, and cigarette
consumption on both school days and weekends. Participants e-diary data revealed evidence
of social contexts exerting differing patterns of stimulus control over smoking behavior
between ethnic groups. Compared to being alone, cigarette smoking was significantly less
likely while with family (OR=.20; p < .01) or boy/girlfriend (OR=.37; p < .01) among White, but
not Asian-American adolescents. In contrast, Asian-American adolescents were significantly
more likely to smoke across a range of social contexts including friends (OR=2.29; p < .001),
boy/girlfriend (OR=1.97; p < .01), teacher/class (OR=2.45; p < .001) and other adults (OR=1.98;
p < .01) compared to times they were alone. Among Whites, the only social context associated
with a greater likelihood of cigarette smoking was found to be while with friends (OR=1.69; p
< .01). Social contexts were observed to exert intermediate influences on the smoking urges
and behaviors of Latino youth. The results of this study point to significant ethnic differences
in the influence of social contexts/stimulus control on tobacco-use.

CORRESPONDING AUTHOR: Julia B. Yi, B.A. , University of California, Irvine, Irvine, CA, 92612;
[email protected]
S2310 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C310 6:00 PM-7:00 PM

TRAJECTORIES OF RETURN TO SMOKING DURING THE POSTPARTUM YEAR AMONG WOMEN

WHO QUIT SMOKING DURING PREGNANCY

Michele D. Levine, PhD1, Yu Cheng, PhD2, Rachel P. Kolko, PhD3, Rachel H. Salk, PhD4, Lisa J.
Germeroth, phd5, Rebecca Emery, MS2, Rebecca Emery, MS2

1
University Of Pittsburgh, Pittsburgh, PA; 2University of Pittsburgh, Pittsburgh, PA; 3University
of Pittsburgh School of Medicine, Pittsburgh, PA; 4University of Pittsburgh Medical Center,
Pittsburgh, PA; 5University of pittsburgh, Pittsburgh, PA

Although it is well known that most women who quit smoking during pregnancy will return to
smoking during the postpartum year, the time course of relapse after delivery is poorly
understood. Identifying whether there are trajectories of relapse and understanding their
time course may improve efforts to target treatments. Accordingly, we sought to document
patterns of smoking resumption among women who quit smoking during pregnancy and to
examine predictors of different relapse trajectories.

Women (N = 300) who recently had quit smoking participated in postpartum-specific smoking
relapse prevention interventions. Participants were interviewed about smoking and provided
expired air carbon monoxide samples at the end of pregnancy and 12-, 24- and 52-weeks
postpartum. Overall, 39% of the sample maintained biochemically verified smoking
abstinence through 52-weeks postpartum. To examine the time course of relapse, data on
number of cigarettes smoked following delivery were abstracted from time line follow back
interviews. Daily data on number of cigarettes were then dichotomized as any or no smoking.
Trajectory analyses (PROC TRAJ) revealed four unique relapse trajectories: sustained
abstinence (50% of the sample), quick relapse (21%), slow relapse (13%), and
inconsistent smoking (16%). Multinomial logistic regression models identified factors
related to trajectory membership. Predictors considered included demographic variables,
smoking history and psychiatric disorders. Not surprisingly, the number of cigarettes smoked
prior to quitting, previous quit attempt, and years of smoking predicted trajectory
membership. Women with more prior quit attempts were more likely to relapse quickly than
to sustain abstinence OR = 1.1, CI = [1, 1.2]), and women who had started smoking at a
younger age were more likely to be inconsistent smokers than to sustain abstinence OR = 1.1,
CI = [1, 1.1]).

There are four distinct courses of smoking resumption during the postpartum year. Although
most women in an intervention sustained tobacco abstinence, approximately one-fifth of
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2311

women enrolled in a postpartum relapse prevention intervention resumed smoking quickly.

Women who relapse quickly can be identified by their smoking history prior to pregnancy.
Thus, efforts to target interventions for these women may need to be initiated prior to
delivery and can be improved by readily indentifiable smoking history variables.

CORRESPONDING AUTHOR: Michele D. Levine, PhD, University Of Pittsburgh, Pittsburgh, PA,

15213; [email protected]
S2312 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

C311 6:00 PM-7:00 PM

YOUNG ADULT MOTIVES FOR ELECTRONIC NICOTINE DELIVERY SYSTEM USE

Marie Chesaniuk, MA, Alexander Sokolovsky, MA

University of Illinois at Chicago, Chicago, IL

This study examined motives associated with ENDS use and cessation outcomes among a
sample of young adults (N=304; mean age: 22.4; 38.2% female; 62.8% white). We
hypothesized that motives associated with use and cessation outcomes among cigarette
smokers would predict such outcomes among young adult ENDS users. We also hypothesized
that previously observed gender differences (e.g., men are more motivated to smoke to
increase positive affect than are women) would be replicated. Participants who self-identified
as regular ENDS uses and self-reported using within a week of data collection were asked to
complete a survey on the Amazon MTurk platform. Measures included an adapted version of
the 14-item Wisconsin Inventory for Smoking Dependency Motives (WISDM), Smoking
Consequences Questionnaire, and self-reported demographics and smoking patterns. Mean
number of days smoked in the past 30 days was 21.64 with 40.5% of participants reporting
daily use. 37.5% of participants reported at least one prior quit attempt with an average
length of 26.55 days. Average length of non-use excluding quit attempts was 15.55 days. More
men than women attempted quitting. T-tests revealed gender differences across smoking
motives and outcome expectancies. Backward stepwise regression analysis was conducted,
subtracting non-influential motives or outcome expectancies to achieve final models for each
use and cessation outcome. Regression analyses predicting smoking use outcomes found
consistent main effects of loss of control and tolerance. Paradoxically, less loss of control
predicted higher levels of ENDS use. Greater tolerance predicted more use across all use
outcomes. For cessation outcomes, only loss of control predicted quit attempts. Less loss of
control increased odds of attempting to quit. Lower levels of craving and higher levels of
affective enhancement increased length of quit attempts. Only tolerance predicted non-use
days: lower levels of tolerance predicted more non-use days. No gender interactions were
found in any regression analyses. Results suggest an escalation is most associated with
tolerance among young adult ENDS users. One possible explanation for the negative effect of
loss of control on ENDS use is that more frequent use prevents withdrawal symptoms and a
sense of losing control, whereas intermittent use allows gaps in which these symptoms might
appear. While a number of gender differences were expressed, none of them affect use or
non-use outcomes.

CORRESPONDING AUTHOR: Marie Chesaniuk, MA, University of Illinois at Chicago, Chicago, IL,
60607; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2313

C312 6:00 PM-7:00 PM

YOUNG ADULT'S PARTICIPATION AND BELIEFS ABOUT THE IMPORTANCE OF GENETIC

MARKERS ON SMOKING BEHAVIOR

Kathleen Diviak, PhD1, Adriana Perez, BS1, Emily Anderson, PhD2, Robin Mermelstein, PhD1

1
University of Illinois at Chicago, Chicago, IL; 2Loyola University, Maywood, IL

This poster will examine the consent process and participation in a genetic study of tobacco
dependence as well as participant beliefs about the important and influence of genetic
markers on smoking behavior. Participants were part of the Social and Emotional Contexts of
Adolescent Smoking Patterns longitudinal study (cohort N = 1,263). The cohort was recruited
when they were approximately 14-15 years old and is quite diverse in terms of race and
ethnicity with 56% Non-Hispanic White, 17% non-Hispanic African American, 13% Hispanic
White, and 14% all other categories. During the 5- and 6-year follow-up assessments when
participants were between 19 and 23 years old, they were recruited to participate in a genetic
sub-study. Of the 1,112 participants approached, 1,019 provided a saliva sample for
genotyping for a participation rate of 91.6%. Participation did not vary by gender,
race/ethnicity, or current smoking status. Participants who declined to participate in the
genetics sub-study were slightly older (M=21.9, SD=0.95) than participants who provided a
sample (M=21.3, SD=0.79); t(963)=-2.79, p=0.005. Everyone who asked to participate
completed a semi-structured interview about their understanding of the genetic project,
reasons for participation, and their beliefs about genetic influences on smoking behavior.
Reasons for participation were primarily to help the overall project (25.5%), the $20 cash
incentive (23.0%), and interest/curiosity about the overall results (16.6%). Most participants
could correctly and clearly describe the purpose of the genetic study (68.5%), but some
(17.6%) were not able to correctly describe the purpose of the study after the informed
consent process was completed. Female participants rated the influence of genetics on
smoking behavior higher than male participants, but there were no differences between the
other demographic or smoking status variables. Current smokers reported that they would
smoke less if they were to find out that they had a genetic risk for tobacco dependence
compared to no risk; t(536)=-13.92, p=0.000. In addition, current smokers also reported that
they would be more interested in quitting if they were to learn that they had a genetic risk
factor compared to no risk; t(534)=--9.86, p=0.000. The implications for participation in
genetic studies and understanding young adult beliefs about genetic testing for complex
behavior like cigarette smoking will be discussed.
S2314 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Kathleen Diviak, PhD, University of Illinois at Chicago, Chicago, IL,
60608; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2315

Saturday

April 1, 2017

8:30 AM-9:45 AM
S2316 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 70 8:30 AM-9:45 AM

SEDENTARY BEHAVIOR IN THE WORKPLACE: UNDERSTANDING AND INTERVENING

Mark A. Pereira, PhD1, Meynard Toledo, B.S., M.S.2, Sarah Mullane, Ph.D.2, Charlotte
Brakenridge, BPsySc3, Nirjhar Dutta, M.S., D.O.1

1
University of Minnesota, Minneapolis, MN; 2Arizona State University, Phoenix, AZ; 3The
University of Queensland, Brisbane, Queensland, Australia

The health benefits of moderate-vigorous physical activity are well-established across the
lifespan. However, guidelines also state the need to minimize the amount of time spent
being sedentary. Adults spend an average of 7.5h/day sedentary, yet recent studies suggest
reallocating just 30 min/day from sedentary to light activity may significantly improve health
risk factors. In recent years a paradigm shift has emerged in the area of physical activity and
health stemming from: 1) emergence of new technologies to accurately assess sedentary time
and very light intensity physical activity, 2) accumulation of scientific evidence supporting
sedentary time (adverse) and light activity (protective) having important impacts on chronic
disease risk independent of moderate and vigorous activity, and 3) emergence of new,
innovative ways to modify posture and behavior through capitalizing on various layers of the
socioecological framework. This symposium focuses on the worksite of adult populations in
the U.S. and Australia. The worksite is a logical intervention target for sedentary behavior, as
adult workers spend 70-80% of their time at work sitting. We will present observational and
experimental studies aimed at understanding the correlates of sedentary time in the
workforce, and evaluating approaches to intervening on sedentary time using novel
experimental strategies. Unique aspects of this collection of scientific endeavors towards
understanding and intervening on sedentary behavior include 1) modern technologies in
objective assessment of sedentary time, posture (sitting v. standing), and light activity, 2)
ecological momentary assessments in 271 adults with 673 EMA data points, 3) parsing
correlates of sedentary behavior into environmental and individual-level factors in 279 adults,
4) efficacy of wearable tracking devices in a worksite intervention of 153 adults in a cluster-
randomized trial, and 5) examination of long-term sustained changes in follow-up to a within-
person cross-over randomized trial of sit-stand workstations. These collective research
studies significantly add to the accumulating body of rigorous studies to date on the efficacy
of workplace interventions to increase light activity, reduce sitting, and improve health.
Findings will inform the potential health benefits of sit-stand workstations and other
approaches that are becoming increasingly popular, but for which rigorous science is needed.
The findings also stand to be generalizable to much of the sedentary workforce, and will help
lead towards more informed health practitioner and employer preventive plans and decisions.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2317

Symposium 70A

ECOLOGICAL MOMENTARY CORRELATES OF WORKPLACE SEDENTARY BEHAVIOR

Mr. Meynard Toledo, B.S., M.S.

PURPOSE: To examine the bi-directional relationship between momentary perceptions and

affective states and objectively-measured workplace SB. METHODS: Participants were
recruited from 13 worksites in the greater Minneapolis and Phoenix regions. Participants
completed an ecological momentary assessment (EMA) survey (delivered via smartphone or
e-mail) on three random days during work hours on a 7-day assessment period. Participants
also wore an activPAL3c accelerometer. The EMA survey assessed momentary perceptions
and affective states (i.e., feeling hungry, tired, relaxed, and happy). SB outcomes were sitting
time, sit-stand transitions, long sitting bouts (sitting bouts that are >30 mins), and light
physical activity (LPA) (standing or stepping< 3.0 metabolic equivalents), one hour preceding
and proceeding each EMA response. Multilevel models were used to examine bi-directional
associations of between- and within-person SB and physical feeling and affective states. All
models were adjusted for age, gender, BMI, and job type. RESULTS: A total of 673 EMA
responses were collected from 271 participants (73% female; age = 42.7 11.3 yrs; BMI= 28.5
6.8 kg/m2). At the between-person level, more happiness was associated with less time in
long sitting bouts during the next hour (b[SE]= -3.3 [1.58], =0.04). No other between-person
relationships were significant. At the within-subject level, participants reported more hunger
following periods of higher than usual LPA (b[SE]= 0.06[0.02], b[SE]= -0.03[0.02], b[SE]= -
0.03[0.02], b[SE]= 0.04 [0.02], =0.02). CONCLUSION: Sedentary behavior was bi-directionally
associated with subsequent physical feeling states; conversely, feelings of happiness were
associated with subsequent SB. Within-person relationships may have been attenuated due to
a limited number of observations per person. Nevertheless, these results can inform future
studies aimed at reducing workplace sedentary behavior through contextually-aware
interventions.

CORRESPONDING AUTHORS: Mullane Sarah, Ph.D., Arizona State University;

[email protected]; Sarah Rydell, MPH, University of Minnesota; [email protected]; Miranda
Larouche, Arizona State University; [email protected]; Matthew Buman, PhD, Arizona State
University; [email protected]; Mark Pereira, Ph.D., University of Minnnesota; [email protected]

Symposium 70B

ENVIRONMENTAL AND INDIVIDUAL LEVEL CORRELATES OF SEDENTARY BEHAVIOR IN THE

WORKPLACE

Dr. Sarah Mullane, Ph.D.

S2318 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

PURPOSE: To identify environmental and individual level correlates of objectively measured

sedentary behavior in the workplace. METHODS: Participants were recruited from thirteen
worksites in the greater Minneapolis and Phoenix regions. Participants wore an activPAL3c
accelerometer continuously for seven consecutive days. Work time was identified and
extracted using daily logs. Workplace sedentary behavior outcomes were sit time, stand time,
light-intensity physical activity (LPA, stepping < 3.0 metabolic equivalents), number of sit-
stand transitions per hour, and time in long bouts of sitting (30 min). Outcomes were
standardized to an 8h work day. An electronic survey was completed to assess environmental
factors (i.e., desk-based vs. centralized amenities, private vs. public office, floor level) and
individual factors (i.e., perceptions of work engagement [UWES], work satisfaction and
performance). Mixed model analyses (worksite-level clustering) were performed to examine
relationships between sedentary behavior and environmental and individual factors. All
models were adjusted for work sector (academic, state, industry), job type, age, gender and
BMI. RESULTS: Participants (N = 297; 70% female; age: 42.3 11.3 yrs) wore the activPAL3c
for 90.2 15.5% of the workday. Presence of a centralized recycling area was associated with
less sitting time (b[SE] = -54.2[17.2], P = 0.002), less prolonged sitting (b[SE] = -64.4[21.5], P =
0.034) and more standing time (b[SE] = 48.7[16.6], P = 0.004), sit-stand transitions (b[SE] =
2.0[0.9], P = 0.02) and LPA (b[SE] = 2.2[0.9], P = 0.01). Conversely, perceived job performance
was associated with more sitting time (b[SE] = 11.7[5.8], P = 0.047) and less standing time
(b[SE] = -13.5[5.6], P = 0.02). Private office use was marginally associated with more
prolonged sitting (b[SE] = 31.0[17.4], P = 0.08). CONCLUSIONS: Environmental factors such as
centralized amenities and office type may influence workplace sedentary behavior.
Surprisingly, high perceived job performance may be a barrier to workplace sedentary
behavior reduction strategies. These findings may inform future longitudinal examinations
and interventions targeting reductions in workplace sedentary behavior.

CORRESPONDING AUTHORS: Meynard Toledo, M.S., Arizona State University;

[email protected]; Sarah Rydell, MPH, University of Minnesota; [email protected]; Linda
Feltes, M.S., Sate of Minnesota; [email protected]; Brenna Vuong, MPH, Fairview
Health Services; [email protected]; Noe Crespo, Ph.D., San Diego State University;
[email protected]; Mark Pereira, Ph.D., University of Minnesota; [email protected];
Matthew Buman, Ph.D., Arizona State University; [email protected]

Symposium 70C

EVALUATION OF A PILOT 12-MONTH ORGANISATIONAL INTERVENTION TO REDUCE SITTING

TIME WITH OR WITHOUT AN ACTIVITY TRACKER.

Ms. Charlotte Brakenridge, BPsySc

ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2319

Introduction: Effective, worksite-driven strategies are needed for sustainable ways to reduce
sitting time. This study evaluated the efficacy of reducing sitting time of minimally intensive
organisational-support strategies alone or combined with a novel activity tracker. Methods: A
pilot 12-month cluster-randomised trial was conducted in 153 Australian office workers (18
teams) in 2014-2015. Nine teams received an organisational-support intervention (e.g.,
emails, manager support; n=87), and nine teams (n=66) also received a waist-worn LUMOback
activity tracker. Data were collected (baseline, 3 months, 12 months) by activPAL3 monitors
(work and overall sitting), other objective assessments (anthropometrics, activity tracker), and
questionnaire. Data were analysed by linear mixed models (n=102-112). Results: After 12
months, both groups significantly (p<0.05) reduced sitting time at work and overall (by
32.112.8 to 40.510.4 min/day [meanSE]) with no significant difference between
interventions, adjusting for confounders. Baseline sitting, back problems, higher job
performance and manager support significantly predicted greater sitting reductions at work.
Work satisfaction and uptake of other activity apps or trackers significantly predicted greater
overall sitting reductions. Longer LUMOback usage tended towards reductions in work and
overall sitting (p<0.10). Conclusions: The findings demonstrate that organisational-support
strategies may effectively reduce office workers sitting time over 12 months. Management
support and addressing performance issues, for example, may facilitate sitting reductions. The
LUMOback may require longer usage to be effective; uptake of other apps and trackers were
associated with improvements and merit investigation.

CORRESPONDING AUTHORS: Charlotte Brakenridge, BPsySc, University of Queensland;

[email protected]; Elizabeth Winkler, University of Queensland;
[email protected]; Brianna Fjeldsoe, Ph.D., University of Queensland;
[email protected]; Genevieve Healy, Ph.D., University of Queensland;
[email protected]

Symposium 70D

ONE YEAR FOLLOW-UP OF A RANDOMIZED TRIAL ON THE EFFICACY OF SIT-STAND

WORKSTATIONS TO DECREASE SEDENTARY TIME AT WORK

Dr. Nirjhar Dutta, M.S., D.O.

Background: We previously reported on a four week within-person randomized cross-over

trial using a sit-stand desk (SSD) intervention that resulted in decreased sedentary time at
work by 3.2 hour per week. Purpose: The purpose of this study was to measure the long term
effect of the SSD intervention on sitting time and activity level 14 months after the original
trial was completed. Methods: A pre-post design was used where the control period from the
original study was regarded as pre and the measurements made in the follow-up study were
post . The intervention SSDs were left with the participants at the end of the original study
in April 2012. The follow-up study was conducted in the same participants over a two week
S2320 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

period in June 2013. The primary outcome for the study was objectively measured physical
activity with the Gruve accelerometer. Results: 15 out of the 23 eligible participants took
part in the follow-up study. Self-reported sitting time (OSPAQ survey) during work-hours was
decreased by 22% (95% CI: 15% to 29%; p<0.001), replaced almost entirely by standing.
Activity measured by accelerometer during work-hours was significantly higher at follow-up
compared to baseline (+24748 AU/hr; 95% CI: 7150 to 42347; p<0.05). There was also a trend
towards lower sedentary time during work-hours at follow-up compared to baseline (p=0.17).
As suggested in the original trial, perceived hunger assessed by ecological momentary
assessment was lower at follow-up compared to baseline. Conclusion: These findings suggest
that sit-stand workstations use may be sustained over the long-term and they may be an
effective tool not only decreasing sitting but also for increasing light physical activity at work.

CORRESPONDING AUTHORS: Nirjhar Dutta, M.S., D.O., University of Minnesota;

[email protected]; Mark Pereira, Ph.D., University of Minnesota; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2321

Symposium 71 8:30 AM-9:45 AM

ETHICAL CONSIDERATIONS FOR USING ONLINE STRATEGIES FOR RECRUITING AND

INFORMING PARTICIPANTS IN GENOMIC SEQUENCING STUDIES

Christine Rini, PhD1, Jennifer Hay, PhD2, Megan Lewis, PhD3, Gail E. Henderson, PhD4,
Kimberly Kaphingst, ScD5

1
University of North Carolina at Chapel Hill, Chapel Hill, NC; 2Memorial Sloan Kettering Cancer
Center, New York, NY; 3RTI International, Research Triangle Park, NC; 4UNC School of
Medicine at Chapel Hill, Chapel Hill, NC; 5University of Utah and Huntsman Cancer Institute,
Salt Lake City, UT

Genomic sequencing is rapidly becoming the technology of choice in many clinical fields for
screening and diagnostic testing. Using genomic sequencing for these purposes creates ethical
considerations for researchers due to potential disparities in uptake of this technology across
groups characterized by age, education, race/ethnicity, family composition or family history.
In this symposium we will explore the ethical and social issues raised by genomic sequencing
studies that use online strategies to recruit, educate or inform participants about genomic
risk. First, we will describe a randomized controlled trial examining uptake and outcomes
(screening, sun protection) of skin cancer genetic testing, which involved an on-line MC1R test
offer. Racial/ethnic disparities in test interest, as well as potential explanations for these
disparities in test interest, will be explored. Second, we will present ongoing work describing a
randomized trial in which genomic sequencing is offered to two family cohorts: those with a
newborn or those with a young child that has a previously diagnosed disorder. All families can
opt to learn about medically actionable childhood conditions and then they are randomized to
learn about other categories of information that vary by age of onset and medical
actionability. Early recruitment data suggest differences in interest between the two cohorts
of parents, and that family conflict may play a role in whether to participate. Third, we will
describe clinic-based recruitment of adults for genomic screening that targets rare, medically
actionable conditions. Education and consent materials are provided online. Those who join
take a survey about decision making and understanding. Preliminary recruitment data reveal
differential uptake by age, gender, and race/ethnicity. Individuals often make the decision to
join before reading the online consent materials, and two-thirds did not open additional items
on the consent summary page. The only predictor of knowledge of the study objectives was
income. Our discussant, an expert in the communication of genomic information, will discuss
common themes across presentations and health communication best practices for increasing
diversity and access for genomic sequencing studies that use online strategies. This
S2322 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium will highlight important issues that need to be considered as the field moves
toward the recruitment of large population-based studies.

Symposium 71A

ONLINE SKIN CANCER GENETIC TESTING IN DIVERSE PRIMARY CARE

Jennifer Hay, PhD

Translational research in genomics has disproportionately engaged highly-educated and

Caucasian populations, but must broaden reach to fully realize the public health potential.
Rates of melanoma are on the rise. Germline variants in the MC1R gene are common in the
general population and confer moderate melanoma risk in those with varied skin types. In a
randomized controlled trial examining uptake and outcomes (screening, sun protection) of
MC1R testing in Albuquerque primary care clinics, 1,339 patients (English or Spanish speakers)
have been approached; 25% have agreed to participate. Participants (N=325; 45% Hispanic,
49% non-Hispanic White, 74% female, mean age=56) were randomized 1:6 to a usual care
condition (NCI skin cancer brochure for diverse skin types) or an online MC1R test offer. Of
the 285 patients randomized to the test offer to date, 42% (n=119) elected to learn about
MC1R testing by logging onto the study website; most (91%) that log on request testing (via
saliva test kit). Those who log onto the website were more likely to be non-Hispanic, have
lower cancer fatalism, report lower family influence on health information seeking, and report
fewer misconceptions about skin cancer (psps>.05). In a logistic regression including all
significant predictors of logging on and participant education level and income, only ethnicity
was significant (pMC1R testing). Level of comprehension of MC1R risk feedback (average
versus higher risk) and three-month sun protection and screening outcomes (target N=885)
will ultimately be examined, as well.

We acknowledge the funding support of NIH R01 CA181241 to Jennifer Hay and Marianne
Berwick (MPIs)

CORRESPONDING AUTHOR: Jennifer Hay, Ph.D., Memorial Sloan Kettering Cancer Center;
[email protected]

Symposium 71B

PARENTAL DECISION MAKING ABOUT GENOMIC SEQUENCING FOR THEIR CHILDREN: ETHICAL
AND PRACTICAL CONSIDERATIONS
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2323

Megan Lewis, PhD

Genomic sequencing is becoming the screening technology of choice in many clinical fields.
Using genomic sequencing for screening creates ethical considerations for researchers. These
issues are compounded when parents make screening decisions for their child or newborn.
Negotiating whether to have a childs genome sequenced and what types of information to
learn can increase conflict between parents during decision making. The NC NEXUS study is
offering genomic sequencing to two cohorts of parents: those with seemingly healthy
newborns or with children previously diagnosed with a disorder. Online decision aids (DA) are
being used at different points in the study along with clinical consultation to support decision
making. We will report on enrollment data between the two cohorts and describe differences
between those who enroll and accept sequencing and those who decline on demographic
variables, reported barriers and psychosocial factors. To date, the response rate is high in
both cohorts (74% in the healthy newborn cohort; 50% in the previously diagnosed cohort). Of
parents who complete the DA, most choose to have sequencing for their child (56% in the
healthy newborn cohort; 83% in the previously diagnosed cohort) suggesting potential
differences in interest between the cohorts. Early reports on barriers to participation between
spouses include the need for both to consent. Dual consent is a regulatory requirement for
participation. In some cases, one parent wants to participate, but the other does not. In two
families, so far, custody battles have made participation impossible. Several families
discontinued participation because one parent completed part of the study procedures, but
the other did not, indicating passive withdrawal from the study. Another barrier to
participation is that all questionnaires and DAs are online. So far, 10 families report lack of
Internet access or not having a device that can display both the DA and questionnaires as
challenges for participation. We will discuss ethical issues raised by differences in participation
and lessons learned about how to recruit parents into studies focused on genomic sequencing
for their children.

CORRESPONDING AUTHORS: Megan Lewis, Ph.D., RTI International; [email protected]; Rita

Butterfield, Ph.D., University of North Carolina, Chapel Hill; [email protected]; Christine
Rini, Ph.D., University of North Carolina, Chapel Hill; [email protected]; Ryan Paquin, RTI
International; [email protected]; Myra Roche, MS, University of North Carolina, Chapel Hill;
[email protected]; Jonathan Berg, M.D., Ph.D., University of North Carolina, Chapel
Hill; [email protected]; Cynthia Powell, M.D., University of North Carolina, Chapel Hill;
[email protected]; Donald Bailey, Ph.D., RTI International; [email protected]

Symposium 71C

REPORT FROM GENESCREEN: ONLINE RECRUITMENT FOR PREVENTIVE GENOMIC SCREENING

AMONG HEALTHY ADULTS

Dr. Gail E. Henderson, PhD

S2324 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Programs that screen adults for rare, medically actionable conditions will likely use web-based
education and consent. Yet, there is little evidence for the effectiveness of e-consent to
convey study information. In GeneScreen, we investigate issues raised by such screening,
focusing on 11 medically actionable conditions. We piloted recruitment at a hospital-based
primary care practice (n=436) and a research biobank (n=650), attempting to recruit equal
numbers by gender, race/ethnicity, and age. Individuals were mailed a brochure and letter
directing them to the GeneScreen website. It provided a summary of 14 study features
relevant for informed consent, which could be expanded to display more information.
GeneScreen joinerswere surveyed about decision making and understanding, and a subset
were interviewed in more detail. Here we describe recruitment results, joiners decision
making processes and characteristics, and factors related to understanding the study. 1,086
individuals were approached, 275 (25.3%) joined. Among those approached, women and
whites were more likely to join; age was related to joining, but in opposite ways at the two
sites. 68% of joiners were female, 79% non-Hispanic white, 9.6% Hispanic; and average age
was 59 (range 24-95). Joiners had generally high literacy and SES, typical of early adopters of
new technology. Median Income was $75,000-99,999; two-thirds had at least a college
degree. 87.9% reported being very or extremely confident with medical information. Genetic
self-efficacy was high: mean on scale of 1-6 was 5.01. Compliance with doctors advice was
high: mean on scale of 1-5 was 4.74. The only significant difference between white and
African American joiners was age: 58.16 vs 47.27 (t=2.43, p= .018). Data on website behavior
revealed that of 14 consent summary items, the mean number opened was 1.67. Two-thirds
did not open any item; 20% opened 1-3 items. On a 6-item scale measuring understanding
study objectives, mean score was 4.51 (sd=1.19). In multivariate analysis, study knowledge
was not associated with expanding more consent summary items; nor was it related to study
site, education, health literacy, race (white vs. AA), or self-efficacy. Higher study knowledge
was only statistically significantly associated with higher income. Most found the decision to
join extremely or very easy. Telephone interviews (n=50) found that over two-thirds made the
decision to join after reading the printed recruitment materials, before going online. We
discuss implications of these findings in the context of e-consent for genomic screening.

CORRESPONDING AUTHORS: Gail Henderson, Ph.D., University of North Carolina, Chapel Hill;
[email protected]; R. Jean Cadigan, Ph.D., University of North Carolina, Chapel
Hill; [email protected]; Rita Butterfield, Ph.D., University of North Carolina, Chapel Hill;
[email protected]; Kriste Kuczynski, B.A., University of North Carolina, Chapel Hill;
[email protected]; Chrstine Rini, Ph.D., University of North Carolina, Chapel Hill;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2325

Symposium 72 8:30 AM-9:45 AM

NEW FINDINGS ON UNDERSTANDING AND TREATING BINGE EATING IN OBESITY

Niloofar Afari, PhD1, Kathryn Godfrey, MS2, Gina Evans-Hudnall, PhD3, Jessica Gundy Cuneo,
Ph.D. 4, Shira Maguen, PhD5

1
University of CA, San Diego & VA San Diego Healthcare System, La Jolla, CA; 2SDSU/UCSD
Joint Doctoral Program in Clinical Psychology, La Jolla, CA; 3Michael E. Debakey VA Medical
Center, Humble, TX; 4University of California, San Diego, Sonoma, CA; 5San Francisco VA
Medical Center and UCSF, San Francisco, CA

Two-thirds of Americans are overweight or obese and the prevalence is even higher among
veterans. Binge eating disorder is characterized by persistent and frequent episodes of
consuming large quantities of food in addition to other symptoms. Regardless of meeting
diagnostic criteria, binge eating is a behavioral problem that can interfere with attempts to
lose weight and contributes to obesity. Additionally, binge eating alone or in the context of
obesity is linked to substantial morbidity and undesirable physical and mental health
consequences. Therefore, it is important to better understand the physiological and
behavioral markers of binge eating as well as to develop and test interventions that address
the underlying mechanisms that may maintain binge eating in obesity and the associated
health-related functioning. The goal of this symposium is to highlight several new findings that
may shed light on the processes involved in the maintenance of binge eating and its
amelioration. The first presentation will describe findings from a novel study of the stress and
binge eating link in obesity by examining associations between lab-based heart rate variability
(HRV) and binge eating behaviors. Presenter 2 will explore the relationships between binge
eating and patient activation and social support in a sample of participants recruited from the
VAs MOVE! Program. Presenter 3 will discuss findings from mediation analyses to examine
the links between psychological flexibility, binge eating, and quality of life in obese or
overweight veterans receiving psychological treatment for binge eating. Finally, the discussant
will critically discuss the presented findings, implications for clinical care, and potential future
research directions.

Symposium 72A

BINGE EATING AND HEART RATE VARIABILITY: IDENTIFYING PHYSIOLOGICAL MARKERS OF

RETROSPECTIVE AND NATURALISTIC EATING BEHAVIOR
S2326 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Kathryn Godfrey, MS

Advancing prevention and treatment of obesity requires developing interdisciplinary

paradigms spanning physiological, psychological, and behavioral bases and designs that
explain eating behaviors within real world settings. This study examined the stress and binge
eating link by modeling associations between lab-based heart rate variability (HRV) and binge
eating behaviors measured through retrospective diagnostic interview and ecological
momentary assessment (EMA). A community-recruited sample of 32 male and female
participants with obesity completed a lab visit to measure HRV and assess binge eating in the
past month. HRV was measured through a lab protocol containing baseline and stressed
conditions. A subsample (n=16) of participants completed a 7 day at-home EMA protocol to
assess binge eating using a smartphone. Multiple linear regressions modeled the relationships
between HRV variables at baseline and when stressed with the frequency of retrospective
binge eating and during the at-home EMA protocol. At baseline, participants with higher
SDNN reported fewer retrospective episodes of overeating (p = 0.03). Similarly, participants
with higher sympathetic activation measured by LFn at baseline had more retrospective
episodes of loss of control over eating (p = 0.04). During the at-home protocol, overeating,
loss of control over eating, consumption of non-nutritious, high calorie foods and breaking
dietary rules were non-statistically significantly associated with time and frequency-domain
HRV variables and heart rate with small to medium effect sizes (semi-partial r2 = 0.10 to 0.35),
which can be detected in larger samples. Lower autonomic flexibility at baseline is related to
binge eating severity in individuals with obesity. Further elucidating the link between HRV and
eating behaviors and examining the clinical potential of integrating physiological,
psychological, and behavioral assessment will inform the theoretical models for weight
management and the development of individualized, real-time intervention.

CORRESPONDING AUTHOR: Niloofar Afari, PhD, University of CA, San Diego; [email protected]

Symposium 72B

EXAMINING RELATIONSHIPS BETWEEN PATIENT ACTIVATION, SOCIAL SUPPORT, & MENTAL

ILLNESS AMONG VETERANS WITH BINGE EATING IN MOVE!

Dr. Gina Evans-Hudnall, PhD

Nearly 78% of Veterans are estimated to be overweight or obese and those poor patient
activation (taking an inactive role in managing health conditions) and limited social support
have difficulties with weight loss. Binge eating (BE) is associated with obesity in Veterans.
Veterans with BE often have co-occurring depression, anxiety and PTSD, which creates
barriers to weight loss. The VA developed an evidenced based weight management program
(MOVE!) that targets change in obesity related health behaviors but does not target patient
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2327

activation or social support. The HELP program is a pilot study of a mental health treatment
effectiveness trial designed to be adjunctive to MOVE!. Participants were recruited during the
first MOVE! session. At baseline, participants completed the MOVE 11 and validated measures
of patient activation, depression, anxiety and PTSD symptom severity, and social support for
healthy eating. The MOVE 11 is an assessment of medical, psychological, and behavioral
contributors to obesity. BE was assessed by a positive response to a single item on the MOVE
11.Correlational analysis was used to assess the relationships between patient activation,
social support, and mental health variables with BE at baseline. Fifty-four participants met the
criteria for BE. BE was significantly associated with low baseline scores for patient activation
(p=.031) and social support for healthy eating (p=.030). BE was also significantly associated
with high scores for anxiety symptom severity (p=.039). Thus, BE among Veterans is
associated with heightened symptoms and anxiety, suboptimal personal activation for
effectively managing health conditions and perceived deficits in support for making healthy
dietary practices. Future investigation of the direct and mediating relationship between BE,
psychological symptoms, patient activation and social support is needed.

Symposium 72C

CHANGES IN PSYCHOLOGICAL FLEXIBILITY, BINGE EATING, AND QUALITY OF LIFE IN

VETERANS: EXAMINING MEDIATED TREATMENT EFFECTS

Dr. Jessica Gundy Cuneo, Ph.D.

Weight-related avoidance and inflexibility play a central role in problematic health behaviors,
such as binge eating. Binge eating in obese individuals is also associated with a greater degree
of morbidity and impairment in quality of life (QOL) than in obesity alone. Psychological
treatments targeting psychological flexibility (PF) improve openness and awareness of
thoughts and feelings related to weight and may improve binge eating behavior and, in turn,
increase QOL. The current study explored the binge eating mediation pathway linking weight-
related PF and QOL in obese/overweight veterans receiving psychological treatment for binge
eating. 90 veterans with an average BMI of 38.9 kg/m2 (SD = 7.6) who reported binge eating
after completing a standard 8-week weight management program (MOVE!) were randomized
to either Acceptance and Commitment Therapy (ACT) or active control. Measured included
the Binge Eating Scale, Acceptance and Action Questionnaire for Weight-Related Difficulties,
and Obesity Related Well-Being questionnaire. As both groups demonstrated comparable
changes in PF during treatment, groups were combined for mediation analyses. Change
scores on the measures were used from baseline to 6-month post treatment follow-up
(n=82). Higher levels of PF were positively related to improvements in binge eating (b =
.2150, SE = .055, p = .0002), and improved binge eating was positively related to increased
QOL (b = .8468, SE = .305, p = .007). A statistically significant indirect effect of increased PF via
decreased binge eating was found on improvements in QOL (b = .182, SE = .083, 95% CI = .052,
S2328 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

.090). Thus, the relationship between increased weight-related PF and increased QOL could be
a function of improved binge eating for Veterans receiving psychological treatment after
completing a weight management program. Future studies can further examine the efficacy of
weight-related psychological treatments that focus on increasing PF.

CORRESPONDING AUTHORS: Kathryn Godfrey, SDSU/UCSD JDP in Clinical Psychology;

[email protected]; Niloofar Afari, PhD, University of CA, San Diego; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2329

Symposium 73 8:30 AM-9:45 AM

A STRENGTH-BASED APPROACH TO UNDERSTANDING HEALTH DISPARITIES: EXAMINING RISK

AND RESILIENCE IN MOTHERS LIVING WITH SYNDEMICS

Idia B. Thurston, PhD1, Kathryn Howell, PhD1, Rebecca C. Kamody, M.S.2, Courtney Peasant,
PhD3

1
University of Memphis, Memphis, TN; 2The University of Memphis, Memphis, TN; 3RTI
International, Raleigh, NC

Physical and emotional adversities in mothers have rippling effects across the family system.
Consistent with syndemic theory, violence, HIV, and substance use are adversities that
frequently co-occur among racial/ethnic minorities and the urban poor. This health disparity
may unfairly create an excess burden of poor health and functioning among vulnerable
populations. To date, there is a gap in our understanding of how syndemics affects the
functioning of mothers and their children, as well as a dearth of information on factors that
protect these women from poor health outcomes. Thus, this symposium utilizes a strength-
based framework to examine risk and resilience in mothers living with syndemics. Such an
approach recognizes healthy adaptation in the midst of adversity, which is critically needed to
improve maternal and child health equity. We will begin this symposium with an overview of
syndemic theory and provide context for the examination of both risk and resilience factors
among mothers. The first presentation will assess how mothers historical and current
experiences of partner violence impact communication with their children about violence,
HIV, and substance use. The second presentation will evaluate how resilience serves as a
protective factor against depressive symptoms in mothers living with HIV. The third
presentation will explore the relationship between poverty and gender-based violence as
perpetrated by a partner among high-risk South African mothers using substances. The fourth
presentation will examine the protective role of support from family and friends on the
relationship between syndemics and posttraumatic stress among high-risk U.S. mothers. This
panel will highlight the impact of cumulative adversity among mothers exposed to violence,
HIV, and substance use. It will also underscore the many strengths these women utilize in the
midst of experiencing syndemics. Consistent with the 2017 SBM conference theme to expand
horizons in behavioral health, this innovative symposium will advance knowledge about the
social context of syndemics, offer novel ways to decrease maternal and child health
disparities, and inform future strength-based treatments.

Symposium 73A
S2330 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

THE MEDIATING ROLE OF RESILIENCE ON THE RELATIONSHIP BETWEEN SYNDEMICS AND

DEPRESSIVE SYMPTOMS AMONG MOTHERS LIVING WITH HIV

Dr. Idia B. Thurston, PhD

Background: According to syndemic theory, people living with HIV experience other
adversities, including substance use and violence, with risk elevated among the urban poor.
The consequences of syndemics are pronounced in mothers, given their central role as
caretakers in the family. While women experiencing syndemics endure poor mental health,
many show resilience (ability to bounce back from adversity). We examined the mediating
role of resilience on the relation between syndemics and depression in mothers living with
HIV.

Method: Participants were 55 mothers living with HIV, aged 25-62 (M=41.2, SD=9.0; 81%
Black). 85% of mothers had a yearly income of $20,000. Women were recruited via HIV/AIDS
service organizations in the U.S. MidSouth and reported time since HIV diagnosis. Syndemics
were assessed via report of illicit drug use and physical/sexual violence with an intimate
partner (both in the past 6 months). Women completed measures of depression, resilience,
and life stress. Women living with HIV (1 syndemic) were compared to those with HIV plus
substance use and/or partner violence (2 or 3 syndemics).

Results: Mediation analyses were conducted using SPSS PROCESS to generate 5000 bootstrap
samples with 95% CI. Controlling for lifetime stressors and time since HIV diagnosis, the
mediation model was significant, R2=.23, F(4, 50)=3.78, p=.009. Being in the syndemic group
had a direct effect on depression, B=6.19, SE=2.89, pB=-.039, SE=0.15, p=.01. When resilience
was included, the direct effect of syndemic group membership on depression was no longer
significant, B=3.50, SE=2.91, p=0.24, indicating full mediation. Thus, resilience had an indirect
effect on the association between syndemic group membership and depression, B=2.70,
SE=1.64, 95% CI=0.43, 6.91.

Conclusions: Findings suggest that while mothers who experience syndemics are at greater
risk for psychopathology, they are protected by the ability to bounce back following adversity,
even after accounting for other stressors. Interventions aimed at improving the wellbeing of
mothers living with HIV should move beyond assessing risk and instead consider positive
factors that protect women from poor health outcomes.

CORRESPONDING AUTHORS: Kathryn Howell, PhD, University of Memphis;

[email protected]; Rebecca Kamody, M.S., University of Memphis;
[email protected]; Courtney Maclin-Akinyemi, M.S., University of Memphis;
[email protected]; Mia Cotton, Friends for Life; [email protected]

Symposium 73B
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2331

SOCIAL SUPPORT AS A MEDIATOR OF THE RELATIONSHIP BETWEEN SYNDEMICS AND

POSTTRAUMATIC STRESS

Dr. Kathryn Howell, PhD

Background: Syndemics (co-occurrence of multiple adversities including intimate partner

violence (IPV), HIV, and substance abuse) in mothers has a rippling effect across the family
system. An understanding of factors that might protect against the negative impact of
syndemics, such as social support is lacking. This study explored the mediating role of social
support from friends and family on the relation between syndemics and posttraumatic stress.

Method: Participants were 178 mothers who experienced physical or emotional adversities
(i.e., living with HIV, experiencing IPV, or using substances). Mothers were aged 22-62
(M=35.2, SD=8.4; 71% Black) who were recruited from community organizations in the U.S.
MidSouth. Syndemics were examined based on report of being HIV positive, illicit drug use,
and physical/sexual IPV (both in the past 6 months). Mothers completed measures of
posttraumatic stress, social support, and life stress. Women with one syndemic were
compared to those with two or three.

Results: Mediation analyses were conducted in SPSS PROCESS to examine the indirect effects
of social support (family vs friends) on syndemics and posttraumatic stress. Controlling for
other stressors, the family support model was significant, R2=.04, F(2, 175)=3.27, p=.04.
Syndemic group membership had a direct effect on posttraumatic stress, B=.41, SE=.16, p=.01
and family support, B=-2.70, SE=1.09, p=.01. When family support was added as a mediator,
the effect of syndemics on posttraumatic stress was weakened, B=.35, SE=.16, p=.03,
indicating partial mediation. Family support had an indirect effect on the association between
syndemic group membership and posttraumatic stress, B=-.02, SE=.01, 95% CI=-0.436, -.0001.
The model examining social support from friends was not significant.

Discussion: Results show the protective role of social support on mitigating the negative
impact of syndemics on traumatic stress. The finding that social support from family has a
stronger protective role highlights the importance of examining social support from multiple
sources. Researchers and clinicians working with mothers should examine the unique role
family support plays, as this could be a target for resilience-focused interventions.

CORRESPONDING AUTHORS: Idia Thurston, PhD, University of Memphis;

[email protected]; Lauren Schaefer, M.A., University of Memphis;
[email protected]; Amanda Hasselle, B.A., University of Memphis;
[email protected]; Olliette Murry-Drubot, Family Safety Center of Memphis and Shelby
County; [email protected]

Symposium 73C
S2332 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

THE RELATIONSHIP BETWEEN INTIMATE PARTNER VIOLENCE AND MOTHER-CHILD

COMMUNICATION ABOUT SYNDEMICS.

Ms. Rebecca C. Kamody, M.S.

Background. Syndemic theory suggests a clustering of substance abuse, violence, and

HIV/AIDS (i.e., SAVA) among racial/ethnic minorities and the urban poor. Parent-child
communication about syndemics may protect against the intergenerational risks related to
substance use, violence, and HIV/AIDS. Among children, open communication may enhance
their resilience (i.e., ability to bounce back) to combat adversities in their high risk
environment. As such, we aimed to identify variables that may influence mothers willingness
to communicate with their children about SAVA.

Method. Participants were 174 mothers (70% Black, Mage = 34.7) of children ages 6-14 (51%
female, Mage = 10.3), recruited from community centers. Mothers completed measures of
parent-child communication about SAVA, history of violence perpetration, and the Revised
Conflict Tactics Scale. Path analysis was used to evaluate a hypothesized model, in which
mothers current IPV was related to willingness to communicate about SAVA, as mediated by
violence perpetration.

Results. An excellent model fit was found (Comparative Fit Index = .99; Root Mean Square
Error of Approximation = .02; Standardized Root Mean Square Residual = .02). IPV had a direct
path to violence perpetration (est.=.29, p < .01), which had a direct path to willingness to
communicate about SAVA (est. = -.60, p < .01). Further, Maternal IPV had both a direct (est. =
-.41, p < .01) and indirect (est. = -.17, p < .01) path to communication about SAVA via the
mediator of violence perpetration. Childs age was a moderator of the relationship between
violence perpetration and communication about SAVA (est.=.56, p

Discussion. Findings highlight the negative consequences of bidirectional violence, such that
mothers who are both victims and perpetrators of violence are less willing to communicate
with their children about SAVA. Given that parent-child communication is a protective factor
for positive health and well-being, this lack of communication may place children on a
problematic health trajectory. Accordingly, researchers should develop parenting
interventions that target the family system and promote open communication, which may
enhance resilience in children.

CORRESPONDING AUTHORS: Idia Thurston, PhD, University of Memphis;

[email protected]; Kathryn Howell, PhD, University of Memphis;
[email protected]; Laura Schwartz, M.S., University of Memphis;
[email protected]; Lauren Schaefer, M.A., University of Memphis;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2333

Symposium 73D

CONTROL IS KEY: POVERTY, VIOLENCE, AND ALCOHOL ABUSE AMONG HIGH-RISK SOUTH
AFRICAN MOTHERS.

Dr. Courtney Peasant, PhD

Background: Interconnected epidemics such as poverty, gender-based violence (GBV), and

substance abuse contribute to the disproportionately high rates of HIV among South African
women, particularly among mothers who must provide for their children. Th e current study
takes a resilience approach to explore how sexual control, a factor that is associated with
lower HIV risk, may moderate the relation between poverty, GBV specifically, intimate
partner violence (IPV), and alcohol abuse among most-at-risk mothers in Pretoria, South
Africa.

Method: Data for this study are derived from a subset of participants of a larger trial
examining the efficacy of a women-focused HIV prevention interventions. Participants were
416 Black/African mothers who were 15 years or older and reported use of at least two
substances on 13 of the past 90 days, had unprotected vaginal sex with a male partner in
the last six months, and had a boyfriend. Women completed computer-assisted interviews
that assessed the frequency of going to bed hungry in the past year (proxy for poverty);
physical, emotional, or sexual IPV by their boyfriend in the past 90 days; problems with
alcohol use in the past year; and sexual control (i.e., ability to refuse unwanted sex from their
boyfriend).

Results: Analyses indicated that sexual control moderated the associations between poverty
and physical IPV including being attacked with a weapon (CI = -.11, -.02), and being struck or
beaten (CI = -.09, -.009). Tests of simple slopes revealed that there was a positive association
between poverty and being attacked or struck/beaten among women with sexual control
scores of 14.42 and 14.20 (Range = 6-30) or less, respectively. However, the associations
between poverty and IPV were not significant for mothers with higher levels of sexual control.

Discussion: These findings indicate that empowering mothers to have sexual control with
their partners and educating couples about the importance of respecting sexual control may
lessen the impact of poverty on gender-based violence. Given that sexual control is amenable
to change, this construct could be a key target for future strength-based interventions.

CORRESPONDING AUTHORS: Jackie Ndirangu, PhD, RTI International Research Triangle Park;
[email protected]; Tracy Kline, PhD, RTI International Research Triangle Park; [email protected];
William Zule, DrPH, RTI International Research Triangle Park; [email protected]; Charlie van der
Horst, M.D., University of North Carolina Chapel Hill; [email protected]; Wendee
Wechsberg, PhD, RTI International Research Triangle Park; [email protected]
S2334 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 74 8:30 AM-9:45 AM

SYSTEMATIC REVIEWS IN CANCER: LEARNING FROM COCHRANE METHODS AND NEW

FINDINGS

Joanna Buscemi, Ph.D.1, Sherri Sheinfeld Gorin, PhD, FSBM2, Dorothy McLeod, N/A3, Heather
L. McGinty, PhD4, Kristen Wells, Ph.D., MPH5, Paul B. Jacobsen, PhD6

1
DePaul University, Chicago, IL; 2New York Physicians Against Cancer, http://nypac.center/,
New York, NY; 3Loyola University Chicago, Chicago, IL; 4The Ohio State University & James
Cancer Hospital and Solove Research Institute, Columbus, OH; 5San Diego State University,
San Diego, CA; 6National Cancer Institute, Bethesda, MD

There are more than 15 million citations in MEDLINE, with 10,000 to 20,000 new citations
added each week, many in cancer, and more than researchers, clinicians, and policy makers
can regularly review, parse, and apply to practice and policy. According to the Cochrane
Collaboration, a 20-year old organization that publishes the Cochrane Database of Systematic
Reviews, internationally-recognized as the highest standard in evidence-based healthcare
resources, systematic reviews attempt to identify, appraise, and synthesize all of the empirical
evidence that meet pre-specified eligibility criteria to reduce bias and answer a given research
question. Using Cochrane criteria produces reliable findings that can be used to inform
decision-making. Our symposium will focus on lessons learned from systematic reviews and
meta-analyses across the cancer continuum, using Cochrane Collaboration and non-Cochrane
approaches. Dr. McGinty will examine the processes and findings from two recent systematic
reviews of multiple cognitive outcomes following cancer treatment in two distinct
populations, men with androgen deprivation therapy (ADT) and individuals receiving
hematopoietic cell transplant (HCT). Ms. McLeod will describe a recent systematic review
exploring the effects of acculturation on weight among Latino children in the US. Drs. Kumar
and Wells, novice and experienced Cochrane reviewers, will discuss their experiences
conducting Cochrane systematic reviews in the fields of medicine and behavioral medicine
across the cancer continuum. They will highlight the differences between Cochrane and other
systematic reviews, as well as the supports and barriers to conducting them. The symposium
aims to increase the implementation of systematic reviews and meta-analyses, particularly
using Cochrane Collaboration processes.

Symposium 74A
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2335

A SYSTEMATIC REVIEW OF ACCULTURATION AND WEIGHT AMONG LATINO CHILDREN

Dorothy McLeod, MA

We propose to describe a recent systematic review, which examined the effects of

acculturation on weight among Latino children in the US. Using this review as a case example,
we also propose to highlight methods and lessons learned over the course of the process.

In the systematic review Becoming American, Becoming Obese? findings across studies on
the relations between acculturation and obesity in Latino children were explored. Previous
research has found that there is a positive relationship between US acculturation and obesity,
a known risk factor for cancer development, among Latino adults (Oza-Frank & Cunningham,
2010); however, less is known about how the process of acculturation may affect childrens
weight. The identification of potential factors associated with obesity in children and adults is
critical to inform future interventions that may reduce obesity and obesity-related cancers.
Using search terms referring to children, acculturation, and obesity to search PsychInfo,
ERIC, Cochrane, PubMed, and CINAHL, the authors located 381 records. The first two authors
independently assessed 59 of these articles in their full-text form, and 29 of these were
included in the qualitative synthesis. The authors concluded that there was heterogeneity
across studies of acculturation and weight, likely stemming from the variability in the
measurement and definition of acculturation across the studies. However, overall, relations
between US acculturation and weight were more likely to be positive among adolescents and
toddlers, whereas they were more likely to be negative among school-aged children.

Throughout conducting the systematic review, the authors recognized three important
lessons: (1) carefully consider the definition and measurement of your variables of interest,
(2) keep detailed records of every step of the process, and (3) examine findings through as
many lenses as possible in order to determine best next steps for the field. In addition to
these three lessons for preparing a review, the authors also recognized a lesson for authors of
empirical (non-review) manuscripts: define your methods and findings with enough detail that
they may be incorporated into a review. We propose to present these lessons, contextualized
by real-world examples from our own review, in order to inform and empower early career
researchers with the tools to conduct their own reviews.
S2336 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHORS: Joanna Buscemi, Ph.D., DePaul University;

[email protected]; Amy Bohnert, Ph.D., Loyola University Chicago; [email protected]

Symposium 74B

CONDUCTING SYSTEMATIC REVIEWS AND META-ANALYSES OF MULTIDIMENSIONAL

OUTCOMES WITH MULTIPLE COMPARISON GROUPS

Dr. Heather L. McGinty, PhD

Two recent systematic reviews of cognitive outcomes following cancer treatment in two
distinct populations are described and used to illustrate special considerations for conducting
complex systematic reviews and meta-analyses.

The first systematic review and meta-analysis examined cognitive outcomes for men treated
with androgen deprivation therapy (ADT) across 7 cognitive domains and 3 types of control
comparisons (cancer patient controls, non-cancer controls, and pre-ADT baselines). A total
157 abstracts were reviewed for inclusion by independent pairs of raters, with 14 studies with
unique samples (n = 417 ADT patients) included in the meta-analysis. Separate meta-analyses
were conducted for each identified cognitive domain. For the random effects model, ADT
patients performed significantly worse than controls on visuomotor tasks (g = -0.67, p = .008).

The second systematic review and meta-analysis evaluated cognitive functioning before and
after hematopoietic cell transplant (HCT) across 8 cognitive domains. 732 abstracts were
pulled and a total of 17 studies were included in a systematic review; of those, 11 studies had
sufficient data for meta-analysis (n = 404 HCT patients). Again, separate meta-analyses were
conducted for each domain and several possible study characteristics were evaluated
including moderators and meta-regression. No significant cognitive changes were observed
across studies (ps > .05).

Specific recommendations are offered for 1) defining multidimensional outcomes a priori, 2)

developing and honing literature search strategies, 3) designing and conducting detailed data
abstraction procedures, and 4) conducting analyses for complex outcomes. Guidance for
graduate students interested in conducting meta-analyses is also discussed.

CORRESPONDING AUTHOR: Heather McGinty, PhD, The Ohio State University / James Cancer
Hospital and Solove Research Institute; [email protected]

Symposium 74C

LESSONS LEARNED IN CONDUCTING COCHRANE SYSTEMATIC REVIEWS

ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2337

Dr. Kristen Wells, Ph.D., MPH

The goal of a systematic review is to identify, evaluate, synthesize, and summarize the
findings from all relevant studies on a specific question related to a specific question domain
(e.g., treatment). Systematic reviews provide decision makers with the best possible evidence
to make informed decisions. As a global independent network of researchers, professionals,
patients, carers, and people interested in health, Cochrane has a mission to promote
evidence-informed health decision-making by producing high-quality, relevant, accessible
systematic reviews and other synthesized research evidence. Historically, most Cochrane
systematic reviews have been conducted in the field of medicine, with relatively few
behavioral medicine systematic reviews having been completed. As more behavioral medicine
researchers become interested in conducting Cochrane systematic reviews, it is clear that
these researchers will need to become familiar with the steps required to complete a review.
We will share our experiences with conducting Cochrane systematic reviews in the fields of
medicine and behavioral medicine by incorporating the perspectives of both a novice
Cochrane review author and an experienced Cochrane author and reviewer. First, we will
summarize major differences between Cochrane and non-Cochrane systematic reviews. Then,
we will provide an overview of the supports and barriers to conducting a Cochrane review,
with a focus on the following topics: understanding and following Cochrane procedures,
adhering to Cochrane timelines, funding for the conduct of systematic reviews, determining a
specific research question, assembling a team with content and methodological expertise,
developing a systematic review protocol, collaborating with a librarian, training of data
abstractors, and evaluating publication venues. We will conclude with lessons learned from
our experiences conducting Cochrane systematic reviews on a wide range of behavioral
medicine and medical topics.

CORRESPONDING AUTHOR: Ambuj Kumar, M.D., M.P.H, University of South Florida;

[email protected]
S2338 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 75 8:30 AM-9:45 AM

EXPANDING BEHAVIORAL MEDICINES IMPACT ON THE PREVENTION AND TREATMENT OF

SUBSTANCE USE IN THE CONTEXT OF INFECTIOUS DISEASE

Amy Starosta, PhD1, Andrea L. Hobkirk, PhD2, Sarah M. Wilson, Ph.D.3, Sara N. Edmond, PhD4,
Courtney Peasant, PhD5

1
Denver VA Medical Center, Denver, CO; 2Duke University School of Medicine, Durham, NC;
3
Department of Veterans Affairs, Durham, NC; 4VA Connecticut Healthcare System, West
Haven, CT; 5RTI International, Research Triangle Park, NC

For those living with chronic infectious diseases in the United States, substance use
contributes to deteriorating health and poor treatment adherence. Substance use-related
health disparities exist despite significant advances in HIV and Hepatitis C pharmacological
treatment and prevention. More than ever before, infectious disease clinical practice and
research can be informed by behavioral medicine with regard to reducing health risk behavior
and increasing health promotion.

This symposium, conducted by behavioral health investigators from research institutions

across the country, focuses on the intersection of non-intravenous substance use and
infectious disease and will elaborate on empirical research and clinical models from a variety
of perspectives. Sessions will discuss the prevention and treatment of substance use and
infectious disease risk reduction in the context of clinical care, address health behaviors in
complex social contexts that include stigma and addiction, and highlight the translational
potential of the research findings.

Each session will review a non-intravenous substance of abuse that impacts infectious disease
transmission and/or clinical care. The use of alcohol and crack cocaine continue to drive HIV
transmission risk in the US through risky sex behaviors. For those living with infectious
diseases like HIV, smoking now accounts for more morbidity and mortality than HIV itself. In
addition, prescription opioid abuse is a rapidly growing epidemic that affects medical
populations at disproportionately high rates.

The symposium will begin with a discussion of research findings from two studies assessing
neurobehavioral and individual factors that contribute to sex risk behavior among cocaine
users in the community and college-aged women who use alcohol. The implications for
incorporating these substance use factors into prevention interventions in primary care
settings will be discussed. The second half of the symposium will focus on substance use
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2339

among those with infectious diseases. This will include findings from the first nationally
representative study to examine the prevalence of smoking among veterans living with HIV
and problematic opiate use among OIF/OEF veterans. A discussion of the implications for
behavioral health treatment within multidisciplinary teams in the VA system will follow.

Our discussant will synthesize the findings by reviewing the over-arching impact of substance
use on sexual risk and infectious disease care. This review will highlight the importance of
interventions that address the social context of substance users. Given the expanding role of
behavioral medicine in the prevention and treatment of substance use and infectious disease,
we believe this symposium will be of interest to the attendees at the Society of Behavioral
Medicine 38th Annual Meeting and Scientific Sessions.

Symposium 75A

ALCOHOL AND CONDOMS: ITS COMPLICATED.

Dr. Amy Starosta, PhD

Although it has frequently been assumed that alcohol use negatively impacts condom use, the
research on this topic has been mixed. Some studies find no difference between condom use
following drinking, while others find a negative association between number of drinks and
condom use. While actual alcohol consumption is an important factor to consider, there may
also be attitudes and expectancies associated with drinking that may be important
mechanisms in the mixed association between drinking and risky sexual behavior. The current
study examines the relationship between condom attitudes and intentions with drinking
behaviors.

Participants (318 single college women Mage=18.84 years, SD=1.78) completed demographic,
sexual, and condom attitudes, condom and drinking intentions questionnaires as well as a
Timeline Followback Interview to assess their drinking and condom use for the previous three
months.

Drinking more times per week and more drinks during each drinking event were both
associated with more sexual partners (frequency of drinking r=.278, p<.0001; number of
drinks r=.281; p<.0001), and poorer attitudes toward condoms (frequency of drinking r= -.208,
p<.0001; number of drinks r= -.150; p<.0001). Drinking behavior, as measured by frequencies
of events, number of drinks, or binge drinking over the previous 90 days, was not associated
with condom use over the previous 90 days or during the most recent sex act. Intentions to
use condoms were negatively associated with the frequency of drinking (r=-.115, p=.041), but
not with number of drinks or binge drinking over the previous 90 days. Condom attitudes
were positively associated with actual condom use over the previous 90 days (r=.376,
p<.0001) and during the last act sex act (F=10.11, p<.0001).
S2340 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Overall, drinking behavior was not related condom use. However, increased drinking was
associated with poorer attitudes toward condoms, lower intentions to use condoms, and
more sexual partners. These findings further highlight the complex relationship between
alcohol use and risky sexual behavior. The implications of these findings on behavioral
interventions to reduce risky sexual behavior will be discussed.

Symposium 75B

NEUROBEHAVIORAL CHARACTERISTICS AND HIV SEX RISK BEHAVIOR: A COMPARISON OF

COCAINE USERS WITH AND WITHOUT HIV TO NON-DRUG USERS

Dr. Andrea L. Hobkirk, PhD

Background: Infectious disease is the second leading cause of death among high risk
substance users in primary care settings. While there are well-known neurobehavioral
characteristics that differentiate substance users from non-drug users (e.g., impairments in
neurocognitive function, altered decision-making, trait impulsivity), few behavioral HIV risk
reduction interventions are designed with these in mind.

Method: The current study sought to identify neurobehavioral characteristics that differ
between HIV+ cocaine users (n=19), HIV- cocaine users (n=22), and a control sample of HIV-
non-drug users (n=21), and how these neurobehavioral factors relate to sex risk behavior.
Participants completed a clinical interview, computerized survey and decision-making tasks,
and a neuropsychological battery.

Results: Cocaine users had significantly higher rates of sex risk behavior than non-drug users.
Specifically, HIV- cocaine users had the highest scores on the Risk Assessment Battery (RAB)
followed by HIV+ cocaine users and HIV- non-drug users (F[2,59]=7.12, p=.002).
Neurobehavioral function was also altered among cocaine users compared to non-drug users.
The HIV+ cocaine users had the highest number of errors on the Wisconsin Card Sorting Task
(WCST) (F[2,59]=4.13, p=.021) and the worst deficits in attention (F[2,59]=3.62, p=.033), while
the HIV- cocaine users had the highest trait impulsivity (F[2,59]=13.31, p=.001).

Controlling for group status, several neurobehavioral characteristics were associated with sex
risk. Higher deficits in memory and motor function were associated with unprotected
intercourse (r=.280, p= .029 and r=.473, p= .001) and total number of sex partners in the past
year (r=.296, p=.020 and r=.457, p=.001). Having multiple partners in the past 3 months was
associated with higher deficits in verbal fluency (F[1,59]=7.079, p=.010). Less impairment in
executive function was associated with higher RAB scores (r=-.295, p=.021) and sex with a
prostitute (F[1,59]=5.268, p=.025). Finally, higher sensation seeking was associated with
higher RAB scores (r=.298, p=.020), unprotected intercourse (F[1,59]=11.710, p=.001), and sex
while high or drunk in the past year (F[1,59]=5.020, p=.029).
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2341

Using multiple regression, the effect of group status on RAB scores was moderated by
sensation seeking (B=-0.11, p=.038) and the effect on total number of sex partners in the past
year was moderated by deficits in attention (B=-0.81, p=.025). Sex risk increased with higher
sensation seeking and worse deficits in attention for the HIV+ cocaine users, but not for the
other groups.

Implications: The findings highlight the important role of neurobehavioral function in the risk
for contracting and transmitting infectious diseases like HIV among cocaine users. The
implication of these findings for behavioral interventions in primary care settings will be
discussed.

CORRESPONDING AUTHORS: Sheri Towe, PhD, Duke University School of Medicine;

[email protected]; Christina Meade, PhD, Duke University School of Medicine;
[email protected]

Symposium 75C

VETERANS LIVING WITH HIV: A PARTICULARLY HIGH-RISK GROUP FOR CIGARETTE SMOKING

Dr. Sarah M. Wilson, Ph.D.

The Department of Veterans Affairs (VA) is the largest provider of medical care to persons
living with human immunodeficiency virus (HIV) in the United States. Although HIV status and
veteran status are known risk factors for smoking, there is no population-level evidence of
smoking prevalence among veterans with HIV. Data from the National Survey on Drug Use and
Health (NSDUH) were examined to characterize smoking as a function of veteran and HIV
status. Results indicated that 51% of veterans living with HIV were daily smokers, as compared
to 25% of veterans without HIV and 31% of non-veterans living with HIV. Additionally,
veterans living with HIV were more likely than HIV-negative veterans to endorse past-month
smoking. This study provides preliminary evidence that veterans living with HIV may be a
particularly high-risk group for daily tobacco smoking, with consequent poor health. Findings
are discussed in the context of multiple ecological levels of health risk and behavior change
for this at-risk population, including public policy, institutional policies, interpersonal models,
and intrapersonal behavior change.

CORRESPONDING AUTHORS: Lauren Pacek, Ph.D., Duke University Medical Center;

[email protected]; Paul Dennis, Ph.D., Duke University Medical Center;
[email protected]; Lori Bastian, M.D., M.P.H., PRIME Center of Innovation (COIN);
[email protected]; Jean Beckham, Ph.D., Durham VA Medical Center;
[email protected]; Patrick Calhoun, Ph.D., Duke University Medical Center / Durham VA
Medical Center; [email protected]
S2342 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 75D

CHRONIC PAIN MANAGEMENT IN THE CONTEXT OF OPIOID USE AND INFECTIOUS DISEASES

Sara N. Edmond, PhD

In the context of the national opioid crisis, both prescribed and unprescribed opioid use is
fueling opioid misuse and adverse events including overdose and death. Chronic pain, the
most common indication for opioid therapy, is common among patients with infectious
diseases such as Hepatitis C (HCV), and data suggest patients with chronic pain frequently use
alcohol and illicit substances to manage pain. Among veterans of recent conflicts with HCV
(n=2942), 51% endorse moderate to severe pain. Given high rates of chronic pain, unhealthy
alcohol use (46%), and tobacco use (58%) in this at-risk population, innovative models of care
are needed for veterans with HCV. The Opioid Reassessment Clinic (ORC) is a multidisciplinary
care team staffed by an internist, addiction psychiatrist, APRN, and health psychologist to
work with patients with problems with safety, efficacy, or misuse of opioids. Within the ORC,
patients receive care from a multidisciplinary team with attention to safer pain management
including non-opioid pain medications and non-pharmacological pain management
approaches. Although the ORC is not explicitly designed for patients with infectious diseases,
ORC patients are often medically complex. Of 87 consecutively referred patients over a two-
year period, 33% had a diagnosis of HCV and 1% had HIV. At time of referral, average
morphine equivalent daily dose (MEDD) was 121.4; average pain score was 6.8/10. Over half
(57%) had an aberrant urine toxicology screen prior to their ORC referral. The majority had a
history of substance use disorder (84%); 54% had opioid use disorder at initial assessment.
Two-thirds met criteria for alcohol use disorder and 47% met criteria for cocaine use disorder,
yet only 36% were seen for addiction treatment in the prior year. Mental health co-
morbidities were common with diagnosis rates of 56% for depression, 39% for PTSD, 24% for
an anxiety disorder, and 19% for bipolar disorder. Integrated clinics such as the ORC may be
useful in managing complex patients such as those with HCV and co-morbid pain, substance
use, and mental health concerns.

CORRESPONDING AUTHORS: Sarah Wilson, PhD, Durham VA Medical Center;

[email protected]; John Sellinger, PhD, VA Connecticut Healthcare System;
[email protected]; Ellen Edens, MD, VA Connecticut Healthcare System;
[email protected]; Kristin Mattocks, PhD, VA Central Western Massachusetts;
[email protected]; Cynthia Brandt, MD, VA Connecticut Healthcare System;
[email protected]; Sally Haskell, MD, VA Connecticut Healthcare System;
[email protected]; Lori Bastian, MD, VA Connecticut Healthcare System;
[email protected]; William Becker, MD, VA Connecticut Healthcare System;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2343

Symposium 76 8:30 AM-9:45 AM

EXERCISE AS A NON-PHARMACOLOGICAL APPROACH TO THE MANAGEMENT OF CHRONIC

PAIN

E. Amy Janke, PhD1, Erin Dannecker, PhD, ATC2, Francis Keefe, PhD.3, Linda C. Li, PT, PhD4

1
University of the Sciences, Philadelphia, PA; 2University of Missouri, Columbia, MO; 3Duke
University, Durham, NC; 4University of British Columbia, Richmond, BC, Canada

Exercise is Medicine (EIM) is a global initiative started by the American Medical Association
and the American College of Sports Medicine (ACSM) to make physical activity (PA) a first-line
behavioral medicine strategy to prevent and treat many forms of disease and disability.
Approximately 1/3 of all Americans experience persistent pain, and such pain is associated
with significant distress, disability, and economic cost. An estimated 41% of adults with
chronic, non-cancer pain report their pain symptoms are not controlled. Furthermore, the
growing opioid addiction crisis demands exploration and support of non-pharmacological
approaches to pain management. The aims of this jointly sponsored symposium by ACSM and
the Society of Behavioral Medicine are to discuss exercise as a non-pharmacological approach
to aid in the management of chronic pain. Specifically, this symposium will review: 1) the
relationship between chronic pain and physical activity, and specifically the role of exercise in
the treatment of chronic pain; 2) clinic-based brief intervention approaches delivered
alongside allied health professionals to increase activity in individuals with pain; and 3)
community-based approaches to safely increase exercise in individuals with pain. The first
panelist will review current prevalence and trends in pain and physical activity, shared risk
factors for chronic pain and insufficient physical activity, and provide an overview of the
relationship between pain and exercise within the larger context of EIM. The second panelist
will discuss brief intervention approaches with physical therapists and other health
professionals to increase exercise in individuals with osteoarthritis and discuss some of the
challenges faced when implementing these approaches. The third panelist will discuss
community based approaches for remotely counseling individuals with arthritis to increase
daily activity and exercise, linking these approaches to the EIM model of engaging patients in
community resources. Finally, the discussant will comment on the presented approaches for
increasing physical activity in individuals with pain and discuss current opportunities and
challenges implementing exercise as a non-pharmacological approach to pain management.

Symposium 76A
S2344 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

THE RELATIONSHIP BETWEEN EXERCISE AND PAIN

Erin Dannecker, PhD, ATC

Improving the management of chronic pain and increasing physical activity are both current
public health priorities. The purpose of this presentation is to review the prevalence of
chronic pain and physical activity, shared risk factors for chronic pain and insufficient physical
activity, and the complex relationship between pain and exercise. Based on best estimates,
chronic pain conditions are highly prevalent and represent an increasing burden on the U.S.
healthcare system as the population ages. Current pharmacologic practices to address these
conditions have proven inadequate while inadvertently contributing to an epidemic of opioid
abuse and deaths from overdose. In comparison, exercise can be an effective pain-
management approach, which is why recent guidelines from the Centers for Disease Control
and Prevention prioritize non-pharmacological treatments before initiating opioid therapy in
some patients. Actually, many clinical practice guidelines for various painful diagnoses have
recommended exercise. Systematic reviews support these guidelines, but the relationship
between exercise and pain is complex. For example, the exercise parameters and/or the
condition of the exerciser can affect the magnitude and direction of change in pain. Also, the
mechanisms of exercises effects on pain vary depending upon the exercise parameters.
Moreover, pain is a barrier to exercise behavior and, in the U.S., physical activity levels are
inadequate. Inadequate physical activity is a leading cause of death worldwide. These issues
and the shared risk factors for chronic pain and inadequate physical activity must be
considered by efforts to use exercise as an intervention for chronic pain.

Symposium 76B

PAIN COPING SKILLS TRAINING & PHYSICAL EXERCISE TRAINING FOR PATIENTS WITH
OSTEOARTHRITIC PAIN: OPPORTUNITIES AND CHALLENGES

Dr. Francis Keefe, PhD.

Many patients with osteoarthritis (OA) pain report that difficulties coping with pain are
important factor limiting their ability to follow recommendations about remaining active and
engaging in physical exercise. This presentation highlights a program of research that has
examined the separate and combined effects of physical exercise training and training in pain
coping skills. Three randomized studies (collaborator: Kim Bennell, U. of Melbourne) are
presented. In the first 201 OA knee pain patients were randomized to physical exercise alone,
pain coping skills training (PCST), or a combined physical exercise training/PCST protocol.
Though no significant between-group differences in pain were found, the combined physical
exercise training/PCST condition had greater improvements in function compared to each
intervention alone. One limitation is that the PCST protocol involved ongoing training and
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2345

supervision of physical therapists. Two more recent studies tested an internet-based

interactive PCST program that did not require therapist training/supervision. In a study of 148
individuals with OA knee pain we found that a protocol that combined an online interactive
PCST program with a physiotherapist home exercise program delivered via videoconferencing
was effective in reducing walking pain and improving physical function. A third study of 144
OA hip pain patients found the online PCST program produced immediate improvements in
pain coping and physical function but did not confer any additional benefits to a subsequent
exercise program, though it did produce sustained long-term improvements in pain coping.
The presentation concludes with a discussion of lessons learned and challenges and
opportunities in this emerging research area.

Symposium 76C

CAN PHYSICAL ACTIVITY MONITORING TOOLS SUPPORT PEOPLE WITH CHRONIC PAIN TO BE
PHYSICALLY ACTIVE?

Dr. Linda C. Li, PT, PhD

The evidence supporting a physically active lifestyle among people with chronic diseases is
compelling. Many patients, however, do not meet the minimum recommended amount of
physical activity, and evidence suggests individuals with persistent pain may be particularly
vulnerable to inactivity. For example, in people with arthritis, pain and fatigue are the major
barriers. It is known that people with arthritis were less active during their leisure time
compared to the general population. A recent study also found 42% of rheumatoid arthritis
patients accumulated 0 minute of bouted moderate/vigorous physical activity, performed in
bouts of 10 minutes or more, in a week. Commercially available self-monitoring tools, such as
pedometers and accelerometers, are popular to support an active lifestyle. The purpose of
this presentation is to examine opportunities and challenges of using self-monitoring tools to
promote a physically active lifestyle in patients with chronic pain. Using arthritis as an
example, the feasibility of a remote counselling intervention by a physical therapist with the
use of a popular wearable device will be presented. Findings of a subsequent ongoing
randomized controlled trial in patients with knee osteoarthritis will be discussed. The
presentation will end with findings from a qualitative study in three Canadian provinces on
the views and experiences of patients and health professionals on the use of physical activity
monitoring technologies in arthritis care. A key finding was the discrepancy in how the two
groups perceived the ability of monitoring tools to motivate users, the sustainability of these
tools, and how they were changing patient-health professional interactions. Potential
opportunities in the use of self-monitoring tools to support the Exercise is Medicine (EIM)
initiative in the community will be discussed.
S2346 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 77 8:30 AM-9:45 AM

PREVALENCE AND PREDICTORS OF DEPRESSION AND STRESS IN DIVERSE SAMPLES OF

PREGNANT WOMEN

Karen L. Weis, PhD, RNC-OB, FAAN1, Patricia Kinser, PhD, WHNP-BC, RN, FNAP2, Jeni
Matthews, MS3, Jenn Leiferman, PhD4

1
University of the Incarnate Word, San Antonio, TX; 2Virginia Commonwealth University
School of Nursing, Richmond, VA; 3Arizona State University, Phoenix, AZ; 4colorado school of
public health, Aurora, CO

Karen L. Weis, PhD, RNC1, Patricia Anne Kinser, PhD, WHNP, RN2, Jennifer Matthews, MS3 and
Jenn Leiferman4

1
University of the Incarnate Word, San Antonio, TX; 2Virginia Commonwealth University,
Richmond, VA; 3Arizona State University, Phoenix, AZ; 4Colorado School of Public Health
University of Colorado, Denver, CO

Prenatal maternal stress and depression are associated with pregnancy complications,
negative birth outcomes, poor maternal-infant attachment and infant neurodevelopmental
disorders. Research is just beginning to uncover the etiology of fetal programming to adult
cardiovascular, metabolic and neuroendocrine disorders. Prenatal maternal stress and
depression increase the likelihood of tobacco and substance use, poor dietary habits and late
or inadequate prenatal care. All of these risk factors are modifiable but the development of
well-designed effective interventions requires a clear understanding of population
characteristics and the influence on the outcome variables.

This symposium will address predictors and correlates of prenatal maternal stress and
depression in three diverse populations. The first presentation will focus on demographic
predictors of prenatal maternal anxiety, depression, self-esteem and resilience within a
military sample. The second presentation will concentrate on pre-conception and early
prenatal predictors of depression within an urban sample of black and white women. The final
presenter will compare various stressors and coping mechanisms reported in a sample of
pregnant women with low to severe stress levels (Perceived Stress Scale) - . Lastly, the Chair
will comment on how these identified predictors and correlates of prenatal depression/stress
differ across populations and how these findings may inform future intervention approaches.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2347

Symposium 77A

PREDICTORS OF INCREASED PRENATAL MATERNAL ANXIETY, SELF-ESTEEM AND RESILIENCE IN

A SAMPLE OF MILITARY WOMEN

Dr. Karen L. Weis, PhD, RNC-OB, FAAN

Background: A relationship exists between prenatal maternal anxiety and depression to

pregnancy complications, poor birth outcomes and infant/childhood cognitive delays.
Traditionally, high-risk groups of young, unmarried, undereducated primigravidas have been
the focus of prenatal assessment and interventions. Less understood is anxiety and
depression in differing samples. Purpose/Methods: Two hundred and forty-six military
women were consented and randomized to either a prenatal support program or regular
prenatal care starting in the first trimester. All participants completed the PSEQ, EPDS, RSE
and BRIEF in each trimester. Women in the intervention attended 8, 1-hr mentored sessions
aimed at decreasing prenatal pregnancy-specific anxiety and depression. The efficacy of the
intervention across pregnancy was examined for prenatal anxiety, self-esteem, depression
and resilience using linear mixed models with autoregressive correlation. Demographic
covariates were: age, employment, parity, marital status, education, deployment history,
military branch, race, and active duty status. Results: Compared to multigravidas,
primigravidas had lower anxiety for accepting pregnancy and concerns related to partner
relationships, but a five-fold increase in anxiety related to preparation for labor and higher
fears of helplessness in labor. Marital status afforded a five-fold decrease in anxiety for
accepting pregnancy, a two-fold decrease in fears of well-being but non-married women had
resilience scores two times higher. Those having a high school degree had two times greater
anxiety for accepting pregnancy than those with some or a college degree, unrelated to age.
College graduates had greater fears related to well-being. Age was only predictive of
increased anxiety for relationship with mother. Deployment significantly effected maternal
identification and self-esteem was two-times greater for women of color. Conclusion: This
military sample experienced lower overall anxiety and depression scores compared to
previous studies in military and civilian samples. The results reflect a different picture than
generally anticipated in that women with higher education and parity had increased anxiety
for certain dimensions and age was not a significant factor. The military environment
promotes opportunities to gain increased self-esteem and resilience regardless of sex or race,
and the study bears this out. The impact of the fathers absence on a womens identification
as a mother is extremely important, particularly within a military population. The findings
provide relevant information indicating the unique differences that exist for women regarding
pregnancy anxiety and the need for appropriate assessments and interventions for the
particular population.

CORRESPONDING AUTHORS: Regina Lederman, Phd, RN, FAAN, University of Texas Medical
Branch; [email protected]; Katherine Walker, MSN, RN, 59th Medical Wing;
S2348 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

[email protected]; Wenyaw Chan, PhD, University of Texas - Houston;

[email protected]

Symposium 77B

DEPRESSIVE SYMPTOM PREVALENCE AND PREDICTORS IN THE FIRST HALF OF PREGNANCY

Dr. Patricia Kinser, PhD, WHNP-BC, RN, FNAP

Background/ Purpose/Methods: Despite recent guidance about screening for depression by

the US Preventive Services Task Force, depressive symptoms continue to be under-identified
and under-treated in pregnant women. The purpose of this IRB-approved study is to examine
the predictors of depression in white and black women in early second trimester of pregnancy
as part of a larger study designed to explore how environmental and social exposures
contribute to inter-individual differences in the timing of birth. A sample of 230 black and
white women receiving prenatal care in obstetric clinics in Richmond, VA were recruited for
the study from 2013 to 2016.

Results: Univariate logistic regression indicated that multiple variables were associated with
CIDI-SF depression, such as demographic (race, relationship status, employment status and
education), preconception health (having visited a healthcare worker for depression/ anxiety
in the past 12 months, lifetime history of depression), coping strategies, the prenatal social
environment, and other symptoms (perceived stress, pregnancy-related anxiety). These
variables were then entered into a multivariable logistic regression model with a backwards
elimination procedure to arrive at two parsimonious models: the first included a single
covariate of a lifetime history of a depression episode (AUC=0.87; 95% CI 0.82-0.90; sensitivity
97%; specificity 76%) and the second eliminated lifetime history and included race,
employment status, having visited a healthcare worker for depression/anxiety in the past 12
months, and the Prenatal Social Environment Inventory scale (AUC=0.74; 95% CI 0.64-0.82;
sensitivity 49%; specificity 88%).

Conclusions: This study contributes to the body of knowledge predictors of depressive

symptoms with onset during in preconception and early pregnancy. Future research is
warranted to explore differences in clinical presentation and underlying biological patterns
because, in clinical settings, the use of predictive variables in combination with identifiable
biomarkers related to depressive symptoms may allow for earlier diagnosis and/or
differentiate appropriate treatment modalities.

CORRESPONDING AUTHORS: Leroy Thacker, PhD, VCU School of Nursing; [email protected];

Lapato Dana, PhDc, Virginia Commonwealth University; [email protected]; Sara Wagner, RN,
Virginia Commonwealth University; [email protected]; Roberson-Nay Roxann, PhD,
Virginia Commonwealth University; [email protected]; Ananda Amstadter,
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2349

PhD, Virginia Commonwealth University; [email protected]; Timothy York,

PhD, Virginia Commonwealth University; [email protected]

Symposium 77C

DIFFERENCES IN STRESSORS AND COPING STRATEGIES OF PREGNANT WOMEN WITH LOW-

MODERATE AND HIGH-SEVERE PERCEIVED STRESS SCORES

Jeni Matthews, MS

Background: Up to 78% of women experience elevated levels of stress during pregnancy.

Elevated levels of stress have been linked to negative birth outcomes such as preterm birth (
Methods: We conducted a national survey in women 8 weeks pregnant, 18 years old, and
residing in the US (N=971). Descriptive statistics, frequencies, and percentages were
calculated. The Perceived Stress Scale (PSS) was used to determine levels of stress. Data was
split into quartiles to determine PSS cut points within the 25th, 50th, and 75th percentiles.

Results: Fifty-one percent of women scored within the low-moderate range (0-15) and 49%
scored within the high-severe range (16). Women with low-moderate levels of stress
reported financial burden (N=190, 38%), work (N=164, 33%), and time (N=152, 30%) as the
top three stressors. Women with high-severe levels of stress reported financial burden
(N=271, 58%), work (N=187, 40%), and pregnancy (N=170, 36%) as the top three stressors.
Women with low-moderate levels of stress reported reading (N=232, 46%), exercising (N=179,
36%), and eating (N=127, 25%) as the most frequently used coping strategies. Women with
high-severe levels of stress reported reading (N=145, 31%), eating (N=135, 29%), and nothing
(N=126, 27%) as the most frequently used coping strategies.

Conclusion: Our data suggest that pregnant women with high-severe levels of stress
frequently do nothing to manage their stress and women in both groups eat as a response to
stress. This is important, as researchers may want to consider focusing on limiting these types
of negative coping strategies when designing interventions to decrease stress during
pregnancy. Future studies may want to consider providing specific resources for healthy
coping strategies in women with high-severe stress (e.g., exercise, meditation).

CORRESPONDING AUTHORS: Jeni Matthews, MS, Arizona State University;

[email protected]; Jennifer Huberty, PhD, Arizona State University; [email protected];
Jenn Leiferman, PhD, University of Colorado Denver; [email protected]; Linda
Larkey, PhD, Arizona State University; [email protected]
S2350 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 78 8:30 AM-9:45 AM

OPTIMIZATION OF ONLINE SUBSTANCE USE INTERVENTIONS TARGETING COLLEGE STUDENTS

Kari C. Kugler, PhD, MPH1, Emily Grekin, PhD2, Kari C. Kugler, PhD, MPH1, David Wyrick, PhD3,
Gary G. Bennett, PhD4

1
The Pennsylvania State University, University Park, PA; 2Wayne State University, Detroit, MI;
3
University of NC - Greensboro, Greensboro, NC; 4Duke University, Durham, NC

The multiphase optimization strategy (MOST) is an engineering-inspired framework for

building more efficacious, effective, cost-effective, and scalable behavioral interventions. This
phased approach (preparation, optimization, and evaluation phases) often uses factorial
experimentation to identify which components to include in an optimized intervention. Online
interventions are useful platforms for taking advantage the efficiency of a factorial
experiment. The talks in this symposium showcase three studies at different stages of MOST,
each interested in optimizing an online intervention targeting substance use among college
students.

The first talk presents the results from a pilot study examining elements of a brief computer-
delivered intervention targeting heavy alcohol use. This talk highlights the importance of pilot
testing intervention components prior to experimentation during the optimization phase. The
second talk presents the results from the first (of two) screening experiments during the
optimization phase. This talk demonstrates how the conceptual model of alcohol use and
sexual risk behaviors was used to identify the intervention components for the factorial
experiment. The third talk presents the results evaluating the effectiveness of an optimized
substance use intervention during the evaluation phase. This study shows how repeated
experimentation during the optimization phase engineered a potent intervention.

Collectively, these talks demonstrate how the phases of MOST are currently being used to
optimize online behavioral interventions among college students. Following the principles and
phases of MOST increases the likelihood that a behavioral intervention will achieve its desired
outcome. The discussant for this symposium has extensive expertise in technology-delivered
behavioral interventions and will review the benefits of using MOST to improve the public
health impact of behavioral interventions.

Symposium 78A
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2351

OPTIMIZING E-INTERVENTIONS FOR ALCOHOL USE

Dr. Emily Grekin, PhD

The flexibility and modularity of computer-delivered brief interventions (CDBIs) makes them
an ideal platform for (a) isolating the active ingredients that are associated with positive
outcomes; and thereby (b) continually optimizing CDBIs to achieve cumulative increases in
efficacy. The current study seeks to systematically manipulate four CDBI elements presence
versus absence of (1) empathy/positive regard, (2) voice, (3) animated narrator, and (4) use of
techniques adapted from Motivational Interviewing (MI) to determine which combination of
elements is most associated with reductions in drinking. The current project will recruit 352
university students meeting NIAAA criteria for heavy drinking and randomly assign them to 1
of 16 different conditions representing all combinations of the four intervention elements
noted above (2 X 2 X 2 X 2). Outcomes include state motivation immediately following
baseline participation and alcohol use at 1- and 3-months post-baseline. As a critical first
step, we pilot tested content for three specific cells (high empathy only, high positive regard
only, and education-only control) to evaluate discriminability. Pilot participants viewed the
content for these cells in counterbalanced order and ranked them for preferability,
understanding, and supportiveness. A total of 63.6% of participants preferred the high
empathy content, compared to 18.2% for both the high positive regard and education-only
CDBIs (z = 3.01, p < .01). Participants also found the high empathy CDBI to be the most
supportive (z = 3.01, p < .01), understanding (z = 3.47, p < .01), and affirming (z = 2.56, p =
.01). Interestingly, participant ratings for the control and positive regard conditions did not
differ. These results suggest that either more work is needed on the positive regard content,
or that expressions of positive regard have no impact on participant preferences or
perceptions. These results also support the Nass et al. Media Equation Theory, which suggests
that interactions with even minimally lifelike technology leads to social responses that are
indistinguishable from those with human beings.

CORRESPONDING AUTHORS: Stephen Ondersma, PhD, Wayne State University;

[email protected]; Lucy McGoron, PhD, Wayne State University;
[email protected]; Jessica Beatty, PhD, Wayne State University;
[email protected]

Symposium 78B

OPTIMIZATION OF AN ONLINE STI PREVENTIVE INTERVENTION TARGETING COLLEGE

STUDENTS

Dr. Kari C. Kugler, PhD, MPH

S2352 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Drinking alcohol is often implicated in the sexual behaviors that increase college students STI
risk. Most existing interventions target alcohol and sex separately, rather than focus on the
alcohol-sex intersection. The current study will present results from an ongoing factorial
experiment of itMatters, an online intervention focused at the alcohol-sex intersection
(completion fall 2016). The experiment is designed to identify which intervention components
(i.e., descriptive norms, injunctive norms, expectancies, perceived benefits of protective
behavioral strategies, and self-efficacy to use strategies) are effective at changing the
corresponding primary outcomes. All first-year students from four universities were recruited
and randomized to one of 32 experimental conditions that included a combination of the
aforementioned components. The 3439 students who completed the baseline survey were
mostly female (61%), White (52%), heterosexual (94%), and lived on campus (74%). Baseline
results provide behavioral context for the intervention. Most students reported past alcohol
use (63%) and sex (oral, anal, and/or vaginal; 58%). Among those who had had sex, 9%
reported using alcohol (33% reported being drunk) and 38% did not use a condom at their
most recent vaginal sex. Among the 35% of students who reported a hookup, 20% used
alcohol (42% which reported being drunk) and 31% did not use a condom for vaginal sex.
There were no statistically significant baseline differences in outcomes based on
randomization levels. The results from this experiment will be used to guide decisions about
which intervention components to revise and reexamine in a subsequent factorial
experiment. Complete results from the first screening experiment, the rationale for
conducting two screening experiments to optimize the STI preventive intervention, and
overall study design and goals will be discussed.

CORRESPONDING AUTHORS: David Wyrick, PhD, University of NC - Greensboro;

[email protected]; Amanda Tanner, PhD, University of NC - Greensboro;
[email protected]; Jeff Milroy, PhD, University of NC - Greensboro; [email protected];
Brittany Chambers, MPH, University of NC - Greensboro; [email protected]; Alice Ma,
MPH, University of NC - Greensboro; [email protected]; Linda Collins, PhD, Penn State
University; [email protected]

Symposium 78C

ENGINEERING AN EFFECTIVE SUBSTANCE USE PREVENTION PROGRAM FOR COLLEGE

STUDENT-ATHLETES

David Wyrick, PhD

College student-athletes are at increased risk of substance use compared to non-athletes. We

developed myPlaybook, an online substance use intervention tailored to college student-
athletes using the Multiphase Optimization Strategy (MOST), to optimize its impact. As part of
optimization phase, we tested whether a set of intervention components (i.e., lessons) met
specific optimization criteria in a series of three component screening experiments. The
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2353

results from 3 factorial experiments (ranging from 8 to 16 conditions) identified the optimized
myPlaybook intervention, which was then evaluated using a randomized control trial to test
effectiveness. The optimized intervention package consisted of 4 components: (1) knowledge
lesson, (2) normative perceptions lesson, (3) expectations lesson, and (4) harm prevention
lesson. Schools (n=73) were randomly assigned to receive the packaged intervention or just
the introductory lesson. First-year and transfer student athletes (n=5,860) completed a
baseline survey, a 30-day follow-up survey, and a 4-month follow-up survey. There was a
significant and sustained effect on descriptive and injuctive norms for drug use, and positive
expectancies for alcohol use; however, there was no significant effect on harm prevention
behaviors or intentions. In terms of drug use behavior, there was a significant effect on
frequency of alcohol use and marijuana frequency at 30-day follow-up, but no effect for binge
drinking, heavy drinking, or tobacco use frequency. Finally, there was a significant effect on
non-alcohol and other drug related consequences (e.g., missing class) for in-season (but not
off-season) student-athletes at 30 day follow-up. These results suggest that the optimized
myPlaybook intervention was successful in changing short-term behaviors. Detail about the
process of using MOST and next steps will be discussed.

CORRESPONDING AUTHORS: Jeff Milroy, PhD, University of NC - Greensboro;

[email protected]; Kelly Rulison, PhD, University of NC - Greensboro; [email protected];
Linda Collins, PhD, Penn State University; [email protected]
S2354 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 79 8:30 AM-9:45 AM

INTIMATE PARTNER VIOLENCE AND TRAUMA-INFORMED CARE: PRIORITY ISSUES FOR

BEHAVIORAL MEDICINE

Emily F. Rothman, ScD1, Sheela Raja, PhD2, Ivy Ho, PhD3, Rose Constantino, PhD, JD, RN,
FAAN, FACFE, Fulbright Scholar4

1
Boston University School of Public Health, Boston, MA; 2University of Illinois at Chicago,
Chicago, IL; 3University of Massachusetts Lowell, Lowell, MA; 4University of Pittsburgh School
of Nursing, Pittsburgh, PA

Increasingly behavioral medicine experts are recognizing the importance of addressing

patients' history of trauma and violence victimization in order to optimize all kinds of health-
promoting interventions and to improve individuals' overall well-being. However, precisely
how individual physicians and other healthcare providers, or healthcare organizations, should
tackle the complicated problem of "addressing trauma" is still emerging. "Trauma-informed
care" is now a somewhat commonly-used phrase (or buzzword), but not everyone
understands what is meant by "trauma-informed care" in the same way. Further, we lack a
broad and deep body of evidence about the most effective ways to approach certain victims
of trauma, such as survivors of intimate partner violence. In this session we bring together
three nationally-recognized experts in addressing trauma in the healthcare setting in order to
accomplish three aims. First, we will provide attendees with a neurophysiological explanation
for why and how trauma can influence somatic health in the short- and long-term. Second,
we will provide attendees with evidence that intimate partner violence, specifically, is
associated with physical health outcomes in Southest Asian American women, and finally we
will present data that explores the ways in which caring for survivors of interpersonal violence
can cause compassion fatigue in medical care providers and discuss how compassion fatigue
can be prevented or treated. Understanding and addressing experiences of interpersonal
violence and traumatic experiences so that patients have positive healthcare and health-
promoting experiences is a critically important priority for behavioral medicine, and this
session will provide an ideal forum for anyone with an interest in these topics to engage with
other thought-leaders and experts in the field.

Symposium 79A

TRAUMA INFORMED CARE FOR INTERPERSONAL VIOLENCE SURVIVORS: UNDERSTANDING

NEUROPHYSIOLOGY AND BEHAVIORAL COPING
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2355

Sheela Raja, PhD

The World Health Organization defines interpersonal violence (IPV) as physical, sexual, and
emotional abuse, as well as controlling behaviors that take place in an intimate relationship. A
sizable portion of the US population have experienced IPV, and many children witness IPV on
a daily basis. This session will provide an overview of the health effects of IPV, with a
particular emphasis on the neurophysiological changes resulting from chronic IPV exposure.
Children who witness IPV on a regular basis may be particularly vulnerable to toxic stress,
which results in long-term neurophysiological disruption. The Hypothalamic-Pituitary-Adrenal
(HPA) axis may become chronically over-activated in some individuals. We will focus on how
these physiological changes may encourage maladaptive coping methods later in life,
including overeating, smoking, drinking, and high risk sexual behavior. In addition, trauma
history is likely to influence healthcare utilization patterns, including underutilization of
preventive care and more emergency and sick visits, resulting in a higher disease burden. The
session will explore how trauma-informed approaches can help survivors break
intergenerational patterns of violence. These approaches include psychological interventions
that alleviate neurophysiological disruption and community-based supports that can decrease
the prevalence of IPV.

Symposium 79B

INTIMATE PARTNER VIOLENCE AND PHYSICAL HEALTH OUTCOMES AMONG SOUTHEAST

ASIAN AMERICAN WOMEN

Dr. Ivy Ho, PhD

Although intimate partner violence (IPV) is prevalent among Southeast Asian (SEA) American
women, little is known about the associations between the experience of IPV and negative
health outcomes in this population. Resnick, Acierno and Kilpatrick (1997) proposed a model
explaining the development of health problems following violent assault, which includes three
direct outcomes (acute physical injury, increased stress, and increased risk of mental health
problems) that in turn precipitate a second layer of mediating factors (chronic physical injury,
impaired immune system functioning, increased health risk behavior, and ineffective health
care utilization). The present paper assesses the applicability of Resnick et al.s model to SEA
American women who have experienced IPV by reviewing cultural, historical, and social
factors in this population. Our review indicates that the applicability of Resnick et als (1997)
model to SEA American women is mixed. Some components of the model such as factors
that pertain to physical injury, both acute and chronic fit well with this population. Other
components such as factors pertaining to increased stress, increased risk for mental health
problems, and increased health risk behavior require a more nuanced and complex
S2356 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

perspective. This presentation will conclude with a discussion on the research and clinical
implications of with regard to IPV among SEA American women.

CORRESPONDING AUTHORS: Khanh Dinh, PhD, University of Massachusetts Lowell;

[email protected]; Sable Smith, University of Massachusetts Lowell;
[email protected]

Symposium 79C

EXPLORING TRAUMA-INFORMED CARE AND COMPASSION FATIGUE IN HEALTHCARE

PROVIDERS

Dr. Rose Constantino, PhD, JD, RN, FAAN, FACFE, Fulbright Scholar

Purpose: The purpose of this presentation is to inform the attendees of the signs, symptoms,
and consequences of Compassion Fatigue (CF), the various ways it is displayed and the healing
effects of trauma-informed care. The signs and symptoms of CF are not readily seen by the
untrained eye especially by the sufferer; it is an insidious and gradual lessening of compassion
over time to the observant, trauma-informed, and compassionate co-worker, supervisor or
partner.

Literature Review: We present applicable evidence based literature, examine protective

factors and trauma-informed care in both the healthcare provider and the CF sufferer. For
example, Berg and others (2016) studied CF in a trauma team and found that team members
were not as managing work stresses as well as they perceived themselves to be. Raja and
others (2016) developed a program on educating healthcare providers on trauma-informed
care: the trauma-informed medicine eCases (TIMe) where every part of the institution (from
staff, administrators, supervisors, to healthcare providers) understand the consequences of
traumatic events, sensitively interact trauma supervisors, avoid re-traumatization, and engage
in trauma screening and prevention.

Case Study: An extreme case of tension and preoccupation with the suffering of those being
helped to the degree that it traumatizes the helper. Gentry suggests that we enter into the
healthcare giving field already with CF. CF differs from burnout as burnout is exhaustion from
the chronic tedium of the job (Tabor 2011). The healthcare provider in the case study will
perform her duties and provide compassionate care to the patient devoid of CF. We will have
an interactive session with the audience what if the healthcare provider in this case was
suffering from CF.

Implications: Whether we like it or not, believe or not, realize it or not, healthcare providers
are in the ministry of compassionate and trauma-informed care because we "help" and "care"
for the sick and traumatized; dedicated to caring for the displaced, traumatized and relief of
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2357

human suffering. Trauma-informed care prods us to suffer with the others, to enter the dark
places where it hurts and be able to leave those dark, bloody and chaotic places after we put
them in some order.

Conclusions: No one is immune to job-related trauma and stress. We need to care for each
other. We are indeed our brothers'/sisters' keeper. Being a trauma-informed healthcare
professional is not expensive nor does it remove anything from us in fact it adds something to
the vast knowledge, information and evidence-based practice we already possess.

CORRESPONDING AUTHORS: Corrie Berk, RN, DNP, UPMC; [email protected]; Beth Reefer,
RN, BSN, UPMC; [email protected]; Linden Wu, RN, BS, University of Pittsburgh School of
Nursing; [email protected]
S2358 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 80 8:30 AM-9:45 AM

PERSONALITY PREDICTS HEALTH COGNITIONS, EXPERIENCES, AND BEHAVIORS: FINDINGS

FROM DIVERSE PERSPECTIVES

Jennifer Howell, PhD1, Jennifer M. Taber, PhD2, Jenna R. Cummings, M.A.3, Kate Sweeny,
PhD4, Jennifer Howell, PhD1

1
Ohio University, Athens, OH; 2Kent State University, Kent, OH; 3University of California, Los
Angeles, Los Angeles, CA; 4University of California, Riverside, Riverside, CA

The field of behavioral medicine has long recognized the influence of demographic
characteristics on health cognitions and behaviors. Recently, emerging technologies have
made individually-tailored health communications and interventions possible. However, many
of these endeavors have focused exclusively on demographic differences and ignored the role
of personality. In this symposium, we attempt to extend knowledge about how individual
differences can affect health cognitions and behaviors to include the domain of personality. In
four talks, we offer evidence that personality can affect health cognitions, experiences, and
behaviors across a diverse set of domains. First, Taber will discuss research suggesting that
personality predicts perceptions of risk in a sample of adults undergoing genomic sequencing.
Taber will highlight how personality factors like optimism, spontaneous self-affirmation, and
big-five personality traits predict the discrepancy between peoples understanding of their
absolute risk and their feelings about their risk. Next, Sweeny will discuss how personality
affects physical health and psychological well-being as people wait for uncertain news.
Specifically, she examines the health of law school graduates during four months as they
waited to learn whether they passed the bar exam. Sweenys results demonstrate that
personality traits associated with handling stressors (e.g., spontaneous self-affirmation, trait
mindfulness) are associated with better health trajectories during the waiting period. Third,
Cummings will discuss the role of personality and demographics in determining the social
aspects of drinking and eating. Specifically, her data suggest that bonding over food and drink
is moderated by both gender and the trait emotional eating. Finally, Howell will discuss
evidence that personality traits predict the likelihood that people will avoid health
information. Data from several studies suggests that traits that indicate dispositional
openness to information (e.g., curiosity, openness to experience) and emotional coping
resources (e.g., low neuroticism, high spontaneous self-affirmation, and high optimism)
predict less avoidance of health information and feedback. Together, these four talks provide
evidence of the nuanced and important role of personality in health cognitions, experiences,
and behavior.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2359

Symposium 80A

THINKING ONE WAY ABOUT RISK BUT FEELING ANOTHER: PERSONALITY CORRELATES OF
HAVING MISMATCHED RISK PERCEPTIONS

Dr. Jennifer M. Taber, PhD

People think about personal disease risk in diverse and nuanced ways, and conceptualizations
of risk may depend on personality factors, including individual differences in psychological
resources (i.e., characteristics that help individuals to bolster against emotional, social, or
physical threats). Characteristics of people with discrepancies between what they think and
feel about their risk are relatively unexplored. We examined the prevalence and correlates
(personality/sociodemographic) of mismatches between deliberative (i.e., cognitions) and
experiential (i.e., intuitions) risk perceptions among adults enrolled in an NIH genome
sequencing trial (n=536). Participants completed a survey of their risk perceptions before
learning sequencing results, reporting the extent to which they believed their results would
confer personal risk for a common disease (deliberative) and the extent to which they felt
their genes put them at risk for a common disease (experiential). Participants were
categorized into four groups based on responses to these questions. Excepting those who
selected the midpoint of unsure for both risk items (9.3%) and were removed from analyses,
responses of unsure were categorized as estimates of low risk. Nearly one-quarter (24.3%)
perceived high deliberative but low experiential risk. Fewer (14.7%) reported low deliberative
but high experiential risk. Slightly over half did not have risk mismatches: 28.2% perceived
high deliberative and experiential risk and 23.5% perceived low deliberative and experiential
risk. Compared to all other participants, those in the group whose experiential beliefs were
greater than their deliberative beliefs reported lower optimism (F(1,475)=11.80, p=.001),
greater worry (F(1,478)=16.54, pF(1,450)=5.37, p=.021), somewhat lower extraversion
(F(1,450)=3.58, p=.059), and greater education (F(1,468)=8.06, p=.005). Those whose
deliberative risk beliefs were higher than their experiential risk beliefs were older
(F(1,482)=9.52, p=.002). There were no differences in education, gender, race, resilience,
spontaneous self-affirmation, conscientiousness, openness to experience, or agreeableness
across groups. These data suggest that people with lower levels of trait-level psychological
resources but higher education might feel that they are at risk for disease but acknowledge
low risk or be unsure of their actual objective risk. This is consequential, as deliberative
measures (most commonly used in health behavior research) may not adequately capture
how such individuals engage with risk. Future research may examine whether these
individuals distribute resources (e.g., protective actions) to threats that feel risky rather
than targeting resources to threats that are perceived as objectively risky.

CORRESPONDING AUTHORS: William Klein, PhD, National Cancer Institute;

[email protected]; Rebecca Ferrer, PhD, National Cancer Institute; [email protected];
Emily Grenen, National Cancer Institute; [email protected]
S2360 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Symposium 80B

TRAIT AND DEMOGRAPHIC PREDICTORS OF BONDING OVER ALCOHOL AND UNHEALTHY FOOD
CONSUMPTION

Ms. Jenna R. Cummings, M.A.

Excessive consumption of alcohol and unhealthy food is a modifiable risk factor for multiple
chronic illnesses. However, positive social experiences that coincide with eating and drinking,
such as improved social connection, may reinforce consumption of alcohol and unhealthy
food. This may be especially pertinent for individuals with certain traits and demographic
profiles. The objective of the current study was to experimentally examine how individual
differences in sex, emotional eating, and endorsement of alcohol expectancies may moderate
social experiences that reinforce eating and drinking. Participants (N = 200) completed
demographic and personality questionnaires online. Then, participants were randomly
assigned to consume milkshakes, placebo beer, or water (control) with a stranger in a
laboratory space similar to real-world settings where people eat and drink with others.
Research assistants coded how often participants talked, smiled, and laughed. Participants
self-reported how close they felt with, and chose how close to sit to, the stranger. Results
indicated several effects moderated by individual differences: (1) Men, but not women,
laughed (p = .032) and smiled more (p = .082), and reported feeling closer with the stranger if
they consumed placebo beer but not milkshakes or water (p < .001). (2) Individuals scoring
higher in trait emotional eating laughed (p = .007), smiled (p = .007), and talked more (p =
.063), and reported feeling closer with the stranger (p = .062), if they consumed milkshakes
but not placebo beer or water. (3) Individuals with stronger expectancies that alcohol reduces
tension or promotes social assertiveness reported feeling closer with the stranger if they
consumed placebo beer but not milkshakes or water (p = .037 and p = .032, respectively).
Challenging positive social experiences that coincide with eating and drinking may prevent
excess consumption of alcohol and unhealthy food, particularly if targeted at individuals with
certain traits and demographic profiles: men, individuals that emotionally eat, and individuals
with strong alcohol expectancies.

CORRESPONDING AUTHOR: A. Janet Tomiyama, PhD, University of California, Los Angeles;

[email protected]

Symposium 80C

PERSONALITY AS A HEALTH BUFFER DURING STRESSFUL WAITING PERIODS

Kate Sweeny, PhD

ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2361

Research suggests that waiting for uncertain news can be both emotionally and physically
taxing. As the news approaches, people typically experience increasing levels of anxiety,
rumination, sleep disruption, and poor health. In the present study we examined whether a
variety of personality traits moderate these trends in psychological and physical well-being.
Specifically, we examined whether dispositional optimism, trait mindfulness, tolerance of
uncertainty, and the tendency to spontaneously self-affirm (i.e., to think about ones values
and strengths) might moderate these trends in well-being. To do so we investigated the
subjective health, sleep disruption, well-being, and personality profiles of a sample of law
school graduates awaiting their bar exam results. The results showed that those high in
spontaneous self-affirmation, trait mindfulness, dispositional optimism, and tolerance of
uncertainty generally reported better physical and psychological well-being during the waiting
period. Moreover, participants high in these traits also showed different longitudinal
trajectories in anxiety, rumination, subjective health, and sleep disruption. Specifically,
participants high in these traits tended to show less deterioration in their well-being as the
news approached and flatter trajectories generally, suggesting that the waiting period was
not as unhealthy for them. Taken together, these results suggest that at least personality is
associated with greater well-being during a stressful waiting period and that personality may
buffer people from the escalations in anxiety, rumination, sleep disruption, and poor health
associated with forthcoming news.

CORRESPONDING AUTHOR: Jennifer Howell, PhD, Ohio University; [email protected]

Symposium 80D

THE PERSONALITY PROFILE OF HEALTH INFORMATION AVOIDERS

Jennifer Howell, PhD

Although knowing health information can be critical for disease prevention, people
sometimes intentionally avoid learning information about their health. Indeed, research
suggests that up to 39% of Americans would avoid learning their cancer risk (Emanuel et al.,
2015). In this talk, we examine who avoids information about their health. That is, using
evidence from four distinct studies, we examine what personality traits predict information
avoidance. Broadly, these studies reveal two categories of dispositional predictors of
avoidance: orientation toward information and coping resources. First, we offer evidence that
people who are dispositionally open to information display less health information avoidance.
Using evidence from a longitudinal study and a nationally representative sample, we
demonstrate that personality traits associated with an open orientation toward the world
(e.g., dispositional curiosity, openness to experience) predict less information avoidance.
Next, we discuss two studies that demonstrate that personality traits associated with better
internal coping resources (e.g., optimism, self-esteem, emotional stability) are associated with
less health information avoidance. Throughout the talk, we will also highlight personality
S2362 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

predictors that do not predict health information avoidance (e.g., extraversion,

agreeableness, and intolerance of uncertainty). Discussion centers around identifying people
whose personality profiles indicate they are particularly prone to health information
avoidance, and on possible tailored interventions to reduce problematic health information
avoidance.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2363

Symposium 81 8:30 AM-9:45 AM

NON-PHARMACOLOGICAL APPROACHES TO MANAGEMENT OF CHRONIC MUSCULOSKELETAL

PAIN AMONG VETERANS

Robert D. Kerns, PhD1, Eve E. Reider, Ph.D.2, Erin E. Krebs, MD, MPH3, Rollin Mac. Gallagher,
MD, MPH4, Ranjana Banerjea, PhD5

1
Yale University, West Haven, CT; 2National Center for Complementary and Integrative Health,
National Institutes of Health, Bethesda, MD; 3Minneapolis VA Health Care System,
Minneapolis, MN; 4University of Pennsylvania, Gladwyne, PA; 5Veterans Health
Administration, Princeton, NJ

Chronic pain and prescription opioid abuse and harms have been recognized by
the Department of Health and Human Services (HHS) and the Centers for Disease Control and
Prevention (CDC) as public health crises. Both HHS and CDC have encouraged greater access
to evidence-based non-pharmacological approaches to chronic pain management as
alternatives to long term opioid therapy, The Department of Veterans Affairs (VA) in
collaboration with the Department of Defense (DOD) and the National Institutes of Health
(NIH) have established a partnership that includes funding research to develop and test as
well as adapt efficacious interventions for military and veteran populations who have
problems with pain and comorbid conditions. Objectives are to promote equitable and timely
access to these approaches that have existing evidence to support their effectiveness and to
promote research to address gaps in scientific knowledge. Dr. Robert Kerns, health services
investigator and advocate for improved pain care for veterans, will introduce the symposium
and provide a brief background as context for three presentations and discussion. Dr. Eve
Reider, a Health Scientist Administrator at the National Center for Complementary and
Integrative Health, National Institutes of Health (NIH), will describe a joint 2014 NIH-VA
funding initiative and provide a brief overview of early findings from 13 funded projects
addressing this topic. Dr. Erin Krebs, a VA health services investigator will review the
recommendations from a VA sponsored State-Of-The-Art Conference on this topic held
in November 2016. Recommendations related to four categories of non-pharmacological
approaches will be presented: Behavioral/psychological/ mind-body, exercise/movement,
manual therapies, and model of care. Dr. Rollin Gallagher, VA National Director for Pain
Management, will describe ongoing practice and policy initiatives designed to promote
evidence-based, integrated, multimodal and interdisciplinary chronic pain management for
military service members and veterans. VA's Opioid Safety Initiative, Joint Pain Education
Project, and Acupuncture Training Across Clinical Settings initiatives will be highlighted. Dr.
Ranjana Banerjea, Scientific Program Manager for VA's Health Services Research and
S2364 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Development Service, will discuss the implications of these initiatives and describe
opportunities for continued research in this area.

Symposium 81A

A JOINT NIH-VA COLLABORATION ON NON-PHARMACOLOGICAL APPROACHES TO PAIN

MANAGEMENT N U.S. MILITARY AND VETERAN POPULATIONS

Dr. Eve E. Reider, Ph.D.

Since 2001, more than 2.5 million U.S. troops have been deployed for Operation New Dawn,
Operations Enduring Freedom (OEF) in Afghanistan, and Iraqi Freedom (OIF) in Iraq. Many
service members returning from these operations experienced pain and related comorbidities
(e.g., mental health, substance abuse, sleep, etc.). Studies report approximately 45-50% of
soldiers and veterans experience pain on a regular basis and there is significant overlap
between chronic pain and co-morbid conditions. In addition, there is an incomplete evidence
base for effective pain management. Opioid medications are often prescribed for the
treatment of chronic pain, but chronic use is associated with the potential for misuse, abuse,
and dependence and sometimes fails to adequately control pain. As a result, there is a need
for non-pharmacological approaches, which may complement pharmacological strategies, for
pain management and to reduce the needs and hazards of excessive reliance on opioids.

The National Institutes of Health (NIH), Department of Defense (DoD), and Veterans
Administration (VA) have been working individually and collaboratively to develop and
improve pain management approaches for military personnel, veterans and their families by
adopting changes in clinical protocols and implementing research initiatives. In 2014 a joint
NIH-VA funding initiative took place that included the National Center for Complementary and
Integrative Health (NCCIH), the National Institute on Drug Abuse (NIDA) and the VA Health
Services Research and Development (HSR&D) Division. Thirteen research projects totaling
approximately $21.7 million over five years were funded to enhance non-pharmacological
approaches to managing pain and co-morbid conditions in U.S. military personnel, veterans
and their Families. Projects funded included pilot/feasibility, efficacy/effectiveness, and
health care services studies. This presentation will provide an overview of the studies funded
from this initiative as well some early findings that are available from this initiative.

Symposium 81B

VA STATE OF THE ART (SOTA) CONFERENCE ON NON-PHARMACOLOGICAL APPROACHES TO

CHRONIC MUSCULOSKELETAL PAIN MANAGEMENT
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2365

Dr. Erin E. Krebs, MD, MPH

Chronic pain is an important public health problem that has long been a major cause of
disability in the US. In recent years, an emerging epidemic of opioid-related harmstriggered
by decades of increased opioid analgesic prescribinghas added substantially to the public
health burden of chronic pain. To address these dual challenges, a transformation in the
approach to chronic pain care is needed. The National Pain Strategy envisions this
transformed approach as involving integrated, multimodal care that meets patients
biopsychosocial needs. Achieving this new standard of pain care will require greater access to
evidence-based non-pharmacological therapies.

At an October 2015 White House summit on the opioid epidemic, the Department of Veterans
Affairs (VA) committed to lead a research initiative evaluating non-opioid alternative
approaches to pain management. Toward this end, the VA Health Services Research and
Development service is organizing a State of the Art (SOTA) conference focused on non-
pharmacological approaches to chronic musculoskeletal pain management, including
complementary, conventional, and integrative therapies. The SOTA conference will be held in
November 2016 and will generate both consensus recommendations and peer-reviewed
publications. The three main goals are as follow: 1) to synthesize existing evidence and
evidence gaps related to non-pharmacological approaches, including psychological/behavioral
therapies, exercise/movement therapies, manual therapies, and models for delivery of multi-
modal pain care; 2) to identify promising non-pharmacological approaches that may be ready
for wider dissemination and implementation in VA and non-VA settings; and 3) to identify a
research agenda that can lead to increasing use of evidence-based non-pharmacological
approaches for the benefit of chronic musculoskeletal pain management. This presentation
will include a discussion of SOTA processes and products, including recommendations for
implementation and research priorities related to non-pharmacological pain care approaches.

Symposium 81C

IMPLEMENTING NON-PHARMACOLOGIC PAIN CARE IN THE STEPPED CARE MODEL IN THE

DEPARTMENTS OF VETERAN AFFAIRS AND DEFENSE

Dr. Rollin Mac. Gallagher, MD, MPH

The large number of older Veterans (>50%) with disabling chonic pain conditions and pain-
related co-morbidities and the even higher rates in Veterans from recent wars (>60%), has
prompted the Department of Veteran Affairs, working closely with the Department of
Defense, to engage in an almost decade-long multi-pronged population-based program to
implement a multi-modal Stepped Care Model of pain management across our health
systems. The urgency of this transformational effort is heightened by data demonstrating the
S2366 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

dangers of over-reliance on opioid analgesics for pain care, and has encouraged the
development of team-based, multi-modal care with a specific focus on engaging the Veteran
in a developing self-management as a foundation of effective Stepped Care. This presentation
will outline five core domains in our strategic approach to implementing system-
wide access to evidence-based stepped integrative, physical, and psychological therapies for
pain and measuring the quality of that care. First, system-wide training programs for clinical
teams and specific professionals on those teams, including cased-based virtual SCAN ECHO
training, the Mini-residency, the Joint Pain Education Project(JPEP) for training primary care
teams and providers, the Tiered Acupuncture Training Across Clinical Settings (ATACS)
program, the Pain Psychology Training program, and the Nurse Resource Program. Second,
comprehensive patient/family education programs to empower Veterans in learning pain self-
management, including pain schools and education websites with interactive learning
programs and videos . Third communication tools and resources to support the pain
management strategy in both training and clinical care, including those that directly address
the need for safe opioid prescribing such as the academic detailing approach(measure,
monitor, feedback, train) in the Opioid Safety Initiative to changing practice, the use of
electronic medical tools such as OTRR and STORM to support safe prescribing, and the use of
process improvement tools such as the Patient Aligned Clinical Team (PACT) Pain Roadmap.
Fourth, strengthened communication amongst pain leadership in VHA's Central Office
(VHACO), Veterans Integrated Service Networks (VISN), and Veterans Affairs Medical Centers
(VAMC) to assure dissemination of strong practices and implementation of strategies to
standardize and measure care quality. Fifth, implementation of standardize
pain care between the Department of Defense (DoD) and VHA health systems and facilitation
of effective transitions of care from DoD to VHA. The presenter will discuss challenges in
implementation and in measuring outcomes to foster change.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2367

Symposium 82 8:30 AM-9:45 AM

THE ROLE OF FATHERS IN SUPPORTING CHILD HEALTH BEHAVIORS

Deirdre Dlugonski, PhD1, Kirsten Davison, PhD2, Elena Jansen, PhD3, Emily L. Mailey, PhD4

1
East Carolina University, Greenville, NC; 2Harvard TH Chan School of Public Health, Boston,
MA; 3Centre for Children's Health Research - Queensland University of Technology, South
Brisbane, Queensland, Australia; 4Kansas State University, Manhattan, KS

This symposium addresses the role of fathers in supporting physical activity, healthy eating,
and obesity prevention among children. It is well-established that parents play an important
role in the uptake of health behaviors among children. Yet, the majority of research studies
focus primarily on the behaviors and beliefs of mothers, rather than fathers, for
understanding and promoting child health behaviors. Thus, the overall goals of this
symposium are to: quantify the inclusion of fathers compared to mothers in childhood obesity
prevention interventions; provide evidence for the shared behavioral patterns of fathers and
young children; and describe opportunities and challenges related to engaging fathers in
health behavior research. The first presenter will describe results of a systematic literature
review and quantitative content analysis of the representation of fathers in family based
interventions targeting childhood obesity prevention. The second presenter will describe the
physical activity patterns of father-young child dyads. The final presenter will share data from
multiple studies on factors associated with recruitment and engagement of fathers within
childhood nutrition research. After this symposium, attendees will have a better
understanding of the role of fathers in the promotion of child health behaviors.

Symposium 82A

PHYSICAL ACTIVITY PATTERNS AMONG FATHER-YOUNG CHILD DYADS

Dr. Deirdre Dlugonski, PhD

Approximately 50% of preschool aged children are not meeting physical activity guidelines.
Parents can serve as role models for child health behaviors, but evidence for the relationship
between parent and child physical activity is mixed. Role modeling is most effective when the
parent and child engage in the behavior together (shared physical activity). As such, shared
physical activity might be more strongly associated with child physical activity than parental
physical activity. No previous studies have examined shared physical activity among parent-
S2368 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

young child dyads using an objective measure of physical activity and most studies have
focused on mothers rather than fathers.

The aims of this study were to: describe patterns of shared physical activity among father-
child dyads and examine the associations among shared physical activity, father physical
activity, and child physical activity. Participants (N = 52; 26 father-child dyads) in this cross-
sectional study made two visits to a research lab. Visit 1 included a demographic survey,
height and weight measurement, and instructions for wearing a Bluetooth enabled
accelerometer. Fathers and children were asked to wear the accelerometer on their right hip
during waking hours for 8 days. The accelerometers produced individual physical activity data
and a proximity file with a minute-by-minute analysis of whether or not the parent and child
were in proximity. Individual data files were overlaid with proximity data and processed to
yield shared physical activity. During visit 2, fathers returned the accelerometers.

Fathers and children engaged in an average of 40.5 24.3 and 56.7 21.2 minutes of daily
moderate to vigorous physical activity (MVPA), respectively. Father-child dyads spent 107.1
51.8 minutes in daily shared time. Shared minutes were primarily spent in shared sedentary
activities (36%) and shared light activities (25%). Less than 1% of shared minutes were spent
in shared MVPA. The relationships between child physical activity with shared and father
physical activity were not statistically significant. The magnitude of the relationship between
shared physical activity and child MVPA (r = .28, p = .19) was stronger than the relationship
between father MVPA and child MVPA (r = -.16, p = .47). This suggests that child physical
activity might be more strongly associated with father-child shared physical activity
participation rather than a father's physical activity level.

These data provide an initial description of shared patterns of physical activity between
father-young child dyads using a novel application of accelerometers that will inform future
efforts to understand and promote physical activity within family contexts..

CORRESPONDING AUTHORS: Katrina DuBose, PhD, East Carolina University;

[email protected]; Patrick Rider, MA, East Carolina University; [email protected]

Symposium 82B

THE FORGOTTEN PARENT: FATHERS MAKE UP ONLY 6% OF PARENT PARTICIPANTS IN FAMILY-

BASED CHILDHOOD OBESITY INTERVENTIONS SINCE 2008

Kirsten Davison, PhD

It has been suggested that child health research disproportionately focuses on mothers which
may hinder the design of family-based interventions. In this study, we used rigorous,
systematic methods to quantify the inclusion of fathers, relative to mothers, in family-based
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2369

childhood obesity interventions published since 2008. We conducted a systematic review and
quantitative content analysis of the literature using methods consistent with PRISMA
(Preferred Reporting Items for Systems Reviews and Meta-Analyses) guidelines. Eligible
articles included original studies reporting the results of family-based interventions for
childhood obesity control and prevention published in English between January 1st, 2008 and
July 31st, 2015. Family-based was defined as active and repeated involvement by parents or
legal caregivers. Articles were excluded where parent/caregiver involvement was passive (eg.
educational brochure sent home to parents) and only children were included as research
participants. Two trained coders used a standardized coding scheme to record parent (e.g.,
number of fathers and mothers, parent race/ethnicity) and intervention (eg. setting,
intervention length, attrition rate, child age) characteristics for all eligible interventions.
Differences in parent and intervention characteristics for studies that did and did not include
fathers were tested using chi-square analysis. Seventy nine eligible interventions were
identified. Twenty nine (37%) interventions included only mothers, 27 (34%) included mothers
and fathers, and 1 included only fathers (1%); parent gender was not reported for 22 (28%)
interventions. Of the interventions that included fathers (N=28), the majority (N=14, 52%)
included 10 or fewer fathers. The estimated number of participating fathers across all
interventions was 880; the comparable figure for mothers was 12,980. There is no evidence
that interventions that included fathers varied systematically in intervention design or
participant demographics (beyond gender) compared to interventions that did not include
fathers. In summary, only 6% of all parent participants in family-based childhood obesity
interventions published since 2008 were fathers. This study is a call to action for increased
inclusion of fathers in childhood obesity interventions.

CORRESPONDING AUTHORS: Nicole Kitos, MA, MPH, Harvard TH Chan School of Public Health;
[email protected]; Tayla Ash, MPH, Harvard TH Chan School of Public Health;
[email protected]; Alen Agaronov, MS, RDN, Harvard TH Chan School of Public Health;
[email protected]; Alyssa Aftosmes-Tobio, MPH, Harvard TH Chan School of
Public Health; [email protected]; Fana Hagos, San Diego Mesa College;
[email protected]

Symposium 82C

ENGAGING FATHERS IN CHILD NUTRITION RESEARCH

Dr. Elena Jansen, PhD

Parents feeding practices influence development of childrens eating behaviour and growth
trajectory. The nature and impact of feeding practices may differ between mothers and
fathers. Studies examining the role of fathers in child feeding are relatively scarce, however,
and recruitment to and engagement of fathers in child feeding research is a challenge.
S2370 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

This presentation draws on evidence from three Australian early childhood nutrition research
projects related to identifying factors associated with recruitment and engagement of fathers.
First, 436 fathers of 2-5 year old children were recruited via a university email list and two
community-based family research cohorts. Fathers reported their confidence and knowledge
of healthy eating, willingness to participate in healthy eating programs, with focus on type
and mode of delivery. Second, 28 fathers of 0-12 year old children participated in focus groups
conducted at their workplace and, amongst other themes, discussed their view on using their
workplace as point of contact for nutrition interventions. Third, 195 mother-father pairs of 2-5
year old children were recruited in a socioeconomically disadvantaged community. A variety
of engagement methods and choice of incentives were examined to determine uptake and
identify the most successful strategies for engagement and recruitment of hard-to-reach
families and pairs in particular.

Analyses examined five key issues in engagement of fathers in nutrition research: 1) fathers'
interest in research and intervention participation, 2) fathers' preferences for recruitment
strategies, 3) feasibility and acceptability for recruitment at the workplace, 4) role of
incentives, and 5) recruitment of mother-father pairs.

Our results indicate that recruitment and engagement of fathers in child nutrition research
may require modification to methodologies applied in mother only samples. Experience
from these three research projects will inform father-focused recruitment strategies and
intervention delivery which will help delineate and foster the father-specific impact on child
health.

CORRESPONDING AUTHORS: Holly Harris, Centre for Childrens Health Research, School of
Exercise & Nutrition Sciences, Queensland University of Technology; [email protected];
Lynne Daniels, Centre for Childrens Health Research, School of Exercise & Nutrition Sciences,
Queensland University of Technology; [email protected]; Karen Thrope, Centre for
Childrens Health Research, School of Psychology, Queensland University of Technology;
[email protected]; Anthony Rossi, School of Exercise & Nutrition Sciences, Queensland
University of Technology; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2371

Symposium 83 8:30 AM-9:45 AM

YOUTH IN FOSTER CARE: UNDERSTANDING AND INTERVENING UPON EARLY LIFE ADVERSITY

Rebeccah L. Sokol, B.S. 1, Kym Ahrens, MD/MPH2, Jennifer Blakeslee, PhD, MSW3, Susan
Spieker, PhD4, William D. Tynan, Jr., PhD, ABPP5

1
Gillings School of Public Health, UNC Chapel Hill, Carrboro, NC; 2Seattle Children's
Hospital/University of Washington, Seattle, WA; 3Portland State University, Portland, OR;
4
University of Washington, Seattle, WA; 5American Psychological Association, Washington, DC

In the United States, over a half-million youth are in foster care. A myriad of circumstances
cause youth to enter state care, but these situations broadly entail exposure to maltreatment
and family instability. Such early life exposures are risk factors for negative outcomes across
multiple domains, including poor mental and physical health. Experiences of abuse, neglect,
or household disruption increase the risks of future alcoholism, drug abuse, depression,
sexually transmitted diseases, severe obesity, and other negative health consequences by as
much as twelve-foldyet youth in foster care remain underserved by health promotion
programs.

This symposium provides an overview of the relationships between adverse experiences and
subsequent health for youth in or transitioning out of foster care, and it also details specific
intervention strategies that have exhibited success in improving health in this population. The
format of the symposium follows the early life course, starting with infancy and concluding
with young adults transitioning out of care. Dr. Susan Spieker begins with describing an
intervention program (and its associated randomized controlled trials) that teaches caregivers
how to be responsive to the unique needs of toddlers with child welfare involvement in order
to alleviate the burden that early maltreatment has on physical and mental health problems.
This transitions to Dr. Jennifer Blakeslees presentation, where she details various intervention
models that are aimed at improving psychosocial and socio-emotional development for youth
both in and transitioning out of foster care. The symposium concludes with Dr. Kym Ahrens
providing an evaluation of the health risks that youth in foster care face, with an emphasis on
reproductive health. Dr. Ahrens closes with discussing specific strategies that could be
employed by practitioners and policymakers to disrupt these risks.

Symposium 83A
S2372 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

UNDERSTANDING AND REDUCING PHYSICAL AND REPRODUCTIVE HEALTH RISKS AMONG

TRANSITION AGE YOUTH IN FOSTER CARE

Kym Ahrens, MD/MPH

Foster youth are an understudied population with high rates of chronic/untreated health
conditions both upon entrance into and while in foster care. Consequently, in 2009 the
Institute of Medicine highlighted them as a priority population for federally funded research.

There are many reasons to expect that former foster youth may be vulnerable to similarly
increased rates of chronic health problems once they transition to adulthood. They are
frequently exposed to poverty as well as many other adverse childhood experiences including
abuse, neglect, domestic violence, parental substance use, and many undergo frequent
placement and school changes while in foster care. Several theories exist regarding the effects
of chronic and early exposures to adversity, all of which suggest that the more stressors to
which one is exposed, the higher the likelihood of a chronic physical or mental health
condition later in life. However, few studies have focused on understanding health risks in
young adulthood for this group. One exception to this is reproductive health. We know that,
compared with youth in the general population, youth who have been in foster care have 2-4
times the risk of teen pregnancy as well as 2-14 times the risk of several sexually transmitted
infections (STIs).

In my presentation, I will review findings from several studies in which my collaborators and I
and other investigators have evaluated physical and reproductive health risks during the
transition to adulthood for youth who emancipate from foster care. I will also review research
my collaborators and I have conducted to better understand the cumulative effects of
historical and current risk and protective factors influencing these outcomes. My
presentation will include data on a variety of health conditions, with an emphasis on
reproductive health outcomes. Finally, I will discuss potential strategies relevant to
practitioners and policymakers, which could be used to disrupt negative health trajectories
and mitigate health risks for youth in foster care.

Symposium 83B

DEVELOPING INTERVENTION MODELS TO BETTER ADDRESS FOSTER YOUTH WELL-BEING

NEEDS

Jennifer Blakeslee, PhD, MSW

Research shows that many young people who age out of the child welfare system experience
poor outcomes during the transition from foster care, and multiple studies identify a large
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2373

subgroup characterized by limited service engagement and/or minimal social support who
experience higher rates of mental health challenges, delinquency, and substance abuse.
Accordingly, there is growing interest in developing services that better address aspects of
foster youth well-being in a system where policy and funding are geared towards child
protection. In general, existing approaches do not adequately attend to socio-emotional
development for youth in care which likely contributes to poor outcomes in early adulthood.

I will discuss intervention models that demonstrate effectiveness in improving foster youth
wellbeing by focusing on psychosocial skills and socio-emotional development. One approach
uses weekly curriculum-based coaching to increase youth self-determination, and this model
has been adapted and tested with foster youth in high school, foster youth experiencing a
disability (physical, emotional/behavioral, or intellectual), and undergraduate foster youth
experiencing mental health challenges. This model improves transition readiness, increases
college enrollment and retention, and reduces criminal justice involvement in young
adulthood, all of which are broadly associated with better-functioning subgroups of foster
youth.

Additionally, I am currently developing an intervention model to enhance foster youth

support networks. This model uses practice strategies identified as acceptable and feasible by
youth and service providers to increase service engagement and mobilize informal networks
and natural supports. This approach is grounded in systematic support network assessment
methods that are linked to practice strategies intended to counter the impact of ongoing
social network disruption related to child welfare system involvement. I will discuss how these
two intervention approachesrelationship-based curriculum delivery and strategic network
enhancementmay be combined to improve youth psychosocial and socio-emotional
development in preparation for the transition from foster care.

Symposium 83C

PROMOTING FIRST RELATIONSHIPS: INTERVENING WITH CAREGIVERS OF INFANTS AND

TODDLERS IN FOSTER CARE OR AT RISK OF REMOVAL

Susan Spieker, PhD

Infants and toddlers are a growing population in child welfare and foster care. The obvious
vulnerability of infants and toddlers, and the fact that critical aspects of brain development
occur in the contexts of attachment relationships in the first 3 years of life, necessitate a
preventative approach that supports caregiving relationships. This is the primary aim of the
Promoting First Relationships (PFR) program. PFR is a manualized, relationship-based,
strengths-based, 10-week home visiting program with video feedback techniques.
S2374 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

PFR helps caregivers be sensitive and responsive to the unique needs and challenges of caring
for toddlers and young children with child welfare involvement. Infants in foster care must
develop a new attachment relationship with each placement. Children grieve when
attachment relationships are disrupted, even when they receive excellent care in their new
home. Infants younger than 12 months tend to seek comfort and closeness from new
attachment figures within a week of placement, but toddlers do not readily seek out new
attachment figures, and may be inconsolable or push them away. The process of attachment
formation in toddlers and older children is complicated by these difficult child behaviors,
which may be further exacerbated if there are serial placements resulting in multiple
caregiver changes. These events can compound the effects of early maltreatment and lead
to physical and mental health problems. However, caregiver sensitivity and understanding of
the meaning of the childs behavior is associated with child attachment security, even for
older toddlers and preschoolers with a child welfare history.

This presentation has two parts. The first part includes descriptions of the active elements of
PFR, methods that have been effective for training and ensuring fidelity by community service
providers, and current adaptations of PFR for primary care providers during the first 3 years of
life. The second part presents an overview of results from two randomized control trials of
PFR involving about 450 toddlers in child welfare and their caregivers. The studies address the
need for parenting services that are suitable for community delivery systems, have a training
mechanism accessible to service providers, are brief enough to keep families engaged, and are
effective.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2375

Symposium 84 8:30 AM-9:45 AM

PASSIVE SENSING OF EATING BEHAVIORS: CURRENT STATE OF THE ART AND IMPLICATIONS
FOR BEHAVIORAL MEDICINE RESEARCH

Nabil Alshurafa, PhD1, Angela Fidler Pfammatter, PhD2, Edward Sazonov, PhD3, Bonnie Spring,
PhD1

1
Northwestern University, Chicago, IL; 2Northwestern University Feinberg School of Medicine,
Chicago, IL; 3The University of Alabama, Tuscaloosa, AL

Eating behaviors contribute to a variety of health outcomes such as cardiovascular disease,

diabetes, obesity, and cancer. To date, the vast majority of behavioral medicine research has
relied on measurements of eating that are inherently flawed. Daily self-monitoring is often
burdensome, incomplete, and inaccurate. 24 hour recalls, considered the gold standard in
dietary intake assessment, is burdensome, often limited in accuracy, and only captures a
snapshot in time of what was consumed. Further, using todays technology, behavioral
scientists are increasingly able to intervene on individuals just-in-time, when the person most
needs or would most benefit from support. However, with both popular methods of dietary
intake, the measures are retrospective and not assessed while eating is occurring, limiting the
potential of using just in time interventions. Other behaviors such as physical activity have
benefitted from passive sensing technologies to overcome this barrier. Passive sensing,
detection, and prediction of eating has the potential to decrease biased reports of eating,
gather real time information about eating, and predict problematic eating behaviors. In this
symposium, we seek to review the state of the science of passive sensing of eating in the first
presentation. We will then move on to discuss the guts of how detection and prediction of
eating works along with privacy, stigma, and user-comfort of wearable sensors. In our third
presentation, we will illustrate a package of real time sensing and the potential for
intervention. Finally, we will review the implications of moving this aspect of science forward
as it pertains to prevention and treatment of problematic eating.

Symposium 84A

FROM LAB TO FIELD: EATING DETECTION MACHINE LEARNING MODELS, PRIVACY, STIGMA
AND USER-COMFORT OF WEARABLES
S2376 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Prof. Nabil Alshurafa, PhD

Wearable sensors combined with computer science machine learning methods have given rise
to passive sensing, making it feasible to validly infer in real time physical features of
psychological states, such as stress, and appetitive behaviors, such as overeating. The
advantage of these real time passive sensing capabilities is their ability to use machine
learning models to measure objectively (minimizing self-report bias), precisely (with high
sensitivity and specificity), passively (without burden or disruption), and dynamically
(detecting rapidly changing states). However, to succeed in passively sensing, people must be
willing to wear them, and sensing algorithms must be able to sense eating in the real world.
Despite the continuous growth in the wearable market, adherence continues to be impacted
by concerns of user privacy, social stigma associated with wearing eating-detection based
sensors, and user-comfort. Understanding these concerns will help the wearable community
and researchers ultimately improve our ability to design wearables that mitigate these effects,
increasing adherence to passive sensing and ultimately paving the way towards personalized
behavioral medicine. This talk will present the process of building machine learning models to
detect eating both in-lab and in the field. A study will also be presented aimed at
understanding the factors that influence adherence to wearing sensors used to detect eating-
habits including a wrist-worn, neck-worn sensor and a camera to validate ground truth in
different environments and situations.

CORRESPONDING AUTHOR: Rawan Alharbi, Northwestern University;

[email protected]

Symposium 84B

LEVERAGING PASSIVE SENSING FOR JUST IN TIME ADAPTIVE INTERVENTIONS ON EATING

Dr. Angela Fidler Pfammatter, PhD

Current technologies are moving closer to realizing passive sensing of eating. Leveraging these
technologies will be a new frontier for behavioral medicine and could be used to deploy just in
time adaptive interventions (JITAIs). However, computer and behavioral scientists must work
in tandem to actualize the potential. Specifically, types of eating and temporality must be
specified when sensing, particularly in real time, as part of an intervention decision rule. An
ongoing trial, Sense2Stop, will be presented as a case study in developing a JITAI using passive
sensing to intervene on smoking behaviors given sensing of stress. Elements of the project
such as the sensing capabilities, algorithm development, and decision rules for intervention
will be presented. Lessons learned will be highlighted and a proposal for moving forward and
generalizing to other sensing technologies elucidated. Paramount to this work, a strategy for
working alongside computer scientists will be discussed. Primary questions around eating
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2377

sensing including discovery of context and factors that contribute to typical and problematic
eating behaviors must be explored prior to intervention development. Once these factors and
contexts are identified, they may be used within decision rules to deploy JITAIs. Finally,
examples of microrandomized trials to optimize JITAIs in eating behaviors will be presented.

CORRESPONDING AUTHORS: Kevin Moran, Northwestern University;

[email protected]; Bonnie Spring, PhD, Northwestern University;
[email protected]

Symposium 84C

WEARABLE SENSORS FOR DETECTION AND CHARACTERIZATION OF FOOD INTAKE

Dr. Edward Sazonov, PhD

Food intake provides energy and nutrients to sustain human life. Studying the ingestive
behavior of individuals is of particular interest for understanding and treatment of medical
conditions strongly associated with food intake, such as obesity and eating disorders.
Traditionally, ingestive behavior in humans has been assessed through self-monitoring of food
intake. However, this approach is time-consuming and suffers from the observation and
misreporting effects. Wearable sensors present a compelling alternative to overcome the
limitations of self-reporting methods. These sensors can potentially provide more objective
measurements of food intake by monitoring behavioral and physiological phenomena related
to one or more stages of the food consumption process: hand-to-mouth gestures, bites,
chewing or swallowing. Specialized signal processing and pattern recognition methodologies
applied to the sensor signals are used to automatically detect and characterize each intake
episode. Timing, duration of the meals, the mass and volume, caloric and nutritional content
of ingested food, and the rate of ingestion could potentially be estimated from sensor data.
Furthermore, the sensor-derived measurements of ingestive behavior may inform tailored,
real-time interventions aimed at modification of food intake. This talk presents an overview of
the wearable sensors and accompanying methodologies that have been proposed for
monitoring and modifying ingestive behavior in humans.
S2378 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Saturday

April 1, 2017

10:00 AM-11:00 AM
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2379

10:00 AM-11:00 AM

HEALTH LITERACY AND DISEASE SELF-MANAGEMENT IN OLDER COPD PATIENTS: WHEN

SOCIAL SUPPORT IS NOT SUFFICIENT

Kimberly Muellers, MPH, Li Chen, MPH, Juan Wisnivesky, MD, DrPH

Icahn School of Medicine at Mount Sinai, New York, NY

Background

Although the importance of social support to health outcomes is well established, its role as a
moderator of the effect of health literacy (HL) on medication adherence in patients with
chronic disease has yet to be explored. An investigation of this pathway would further
understanding of the mechanisms through which social support modifies health behaviors
among patients with chronic illness. We predicted that the negative effect of inadequate HL
on self-management in a population with Chronic Obstructive Pulmonary Disease (COPD)
would be mitigated by caregiver support. Moreover, we expected that caregivers with
adequate HL would have greater impact on patient medication adherence than those whose
caregiver had inadequate HL.

Methods

We enrolled 400 community-dwelling individuals who were 55 years of age, had a diagnosis
of moderate to severe COPD made by a health care provider, and spoke English or Spanish
over a 2-year period. We excluded patients with asthma or other chronic respiratory illness,
dementia, or other neurological and psychological conditions. Participant HL was assessed
through the Short Test of Health Literacy in Adults (S-TOFHLA), and caregiver presence was
based on patient self-report. STOFHLA scores (scale 0-36) were dichotomized as adequate
(23) or inadequate/marginal ( < 23). Adherence to COPD medications was measured with the
Medication Adherence Reporting Scale (MARS) and categorized as low ( < 4.5) vs. adequate
(4.5). We used logistic regression to test the relationship between HL and adherence; the
model included an interaction term between HL and caregiver presence to test for
moderation. We also examined a subset of patients with caregivers (N=97 dyads).
S2380 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Results

Our sample included 342 patients, 47% of whom reported having an informal or formal
caregiver. Of the patients, 61% had inadequate HL, and 17% of caregivers had inadequate HL.
Patients with inadequate HL had significantly higher odds of poor medication adherence
(OR=2.48, CI: 1.47-4.19). Contrary to our hypothesis, the presence of an informal caregiver did
not attenuate the negative impact of low HL (p>.05). Among patients with caregivers,
inadequate HL was again associated with higher odds of poor medication adherence (OR:3.5,
95%CI: 1.4-8.8). This relationship remained significant with the inclusion of caregiver HL,
however the interaction term of caregiver HL did not significantly contribute to medication
adherence.

Conclusions

Low HL among patients with moderate with severe COPD is associated with lower medication
adherence. However, social support from caregivers does not appear to attenuate the
negative impact of low HL on patients self-management, regardless of the caregivers HL
level. These findings suggest that interventions to improve adherence to COPD medications
are specially needed for low literacy patients, even those with adequate caregiver support.

CORRESPONDING AUTHOR: Kimberly Muellers, MPH, Icahn School of Medicine at Mount

Sinai, New York, NY, 10029; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2381

10:00 AM-11:00 AM

DOES SELF-WEIGHING AFFECT MOOD? FOR WHOM?

Carly R. Pacanowski, Phd, RD1, Jonathan Miller, MURP2, Jennifer A. Linde, PhD2, Dianne
Neumark-Sztainer, PhD MPH RD2, Marla Eisenberg, Sc.D. M.P.H.2

1
University of Delaware, Newark, DE; 2University of Minnesota, Minneapolis, MN

Abstract
Frequent self-weighing may prevent weight gain in young adults, but concerning psychological
correlates of self-weighing have been documented. Self-weighing may impact mood, which
may mediate relationships with adverse psychological outcomes (e.g. self-esteem, depressive
symptoms). Individuals more likely to be concerned with shape and weight (e.g. females;
those who are overweight or obese; those having experienced an eating disorder (ED)) may be
at heightened risk for self-weighing impacting their mood. This study assesses whether the
frequency of self-weighing is related to mood and if this relationship differs by gender and
lifetime presence/absence of an ED.

Methods
Young adult (ages 25-36) respondents to Project EAT-IV (n = 1830) reported self-weighing
frequency and whether viewing their weight affects their mood. Mean weekly self-weighing
frequencies and proportions of participants reporting an effect on mood were compared
across gender, overweight status, and presence/absence of lifetime ED diagnosis. The
association of self-weighing frequency with likelihood of effect on mood was assessed using
linear regression adjusting for gender, report of eating disorder, BMI, age, race, education and
income, and including interactions of weighing frequency with gender and ED.

Results
Respondents reporting ED reported higher self-weighing frequency (3.1 times/week) than
respondents who had recovered from an ED (2.0 times/week) or never had an ED (1.5
times/week, p = 0.03). Female and overweight respondents and respondents who reported
ED were more likely to report that self-weighing affected their mood compared to male, non-
overweight participants, and those who never had an ED (all p < 0.001). Frequency of self-
weighing was associated with higher likelihood of reporting that self-weighing affects mood (p
< 0.0001).

Conclusions
These data suggest that female participants, those who are overweight, and those with
S2382 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

history of an ED may be more emotionally impacted by self-weighing. The probability of self-

weighing affecting mood significantly increased with each additional self-weighing occurrence
per week. Given the rates of overweight and obesity in young adults, and the promise of self-
weighing to facilitate weight control, but potential effects on mental health, it is imperative to
consider individual characteristics and exercise caution when recommending self-weighing as
a weight control strategy.

CORRESPONDING AUTHOR: Carly R. Pacanowski, Phd, RD, University of Delaware, Newark, DE,
19716; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2383

D001 10:00 AM-11:00 AM

EXAMINING FAITH AS A MEDIATOR OF MOTIVATION FOR DIET AND PHYSICAL ACTIVITY

AMONG AFRICAN AMERICAN WOMEN

Leilani Dodgen, MPH1, Sumihiro Suzuki, PhD2, Heather Kitzman-Ulrich, PhD3

1
University of North Texas Health Science Center, Fort Worth, TX; 2UNT Health Science
Center, Fort Worth, TX; 3Baylor Scott & White Health, Dallas, TX

African American (AA) women have the highest prevalence of overweight and obesity in the
U.S. and suffer disproportionately from chronic diseases. These conditions are primarily
preventable through behavioral lifestyle changes, however interventions with AA women
have been less effective than in other populations. Due to the importance of faith among AA
women, a popular way to address obesity has been through faith-adapted or faith-based
interventions. However, little is known about how these adaptations influence outcomes. In
this study, possible mediating effects of motivation on the relationship between religiosity
(measured by spiritual health locus of control (SHLOC)) and weight loss behaviors were
examined.

Baseline data from AA female participants in the NIH-funded Better Me Within program (N=
158) were collected through surveys from January 2014 to March 2015. Mean age was 48.9
years (SD 11.7) and participants were overweight or obese with mean body mass index (BMI)
of 38.2 (SD 9.7), and waist circumference of 43.0 inches (SD 5.9). Two linear regression
models were run to examine motivation for diet and motivation for physical activity. Since
motivation for diet and physical activity were closely correlated (r = 0.67, p < 0.0001) both
variables were included in each model and stratified to examine participants with high
motivation versus low motivation. In the diet model, among women with low motivation for
physical activity, active SHLOC (e.g., belief that God is a partner in health empowering a
person to take care of him/herself) was the primary predictor of motivation for diet (p < .05);
while in women with high motivation for physical activity, self-efficacy predicted motivation
for diet (p < 0.001). In the physical activity model, among women with both high and low
motivation for diet, self-efficacy for exercise, not SHLOC, was the primary predictor of
motivation for physical activity (p = 0.01).

By incorporating scripture and other faith elements that support an active SHLOC, faith may
serve as a change agent to increase motivation to eat healthy. Faith-based programs may
need to focus faith elements on the connections between faith, health and food in addition to
self-efficacy to enhance motivation for diet. Research focused on defining and measuring
S2384 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

specific constructs within faith-based programs is needed to understand their impact on

obesity-related outcomes.

CORRESPONDING AUTHOR: Leilani Dodgen, MPH, University of North Texas Health Science
Center, Fort Worth, TX, 76107; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2385

D002 10:00 AM-11:00 AM

EXERCISE INTENTIONS MODERATE THE EFFECTS OF IMPLEMENTATION INTENTION

INTERVENTIONS FOR PHYSICAL ACTIVITY

Scout M. Kelly, M.A., John A. Updegraff, PhD

Kent State University, Kent, OH

Background: Implementation intentions (IIs) are plans made in the form of if-then
contingencies and have been shown to be effective in promoting physical activity (Gollwitzer
& Sheeran, 2006). However, IIs are not equally effective for all people. Recent research
indicates that exercise intentions and plan quality may moderate II effectiveness. The present
study examined whether pairing II interventions with a novel online plan quality training
module promotes greater physical activity than a traditional II intervention or control activity,
and whether the intervention effects are moderated by baseline exercise intentions. Method:
166 undergraduates (69.5% female, 69.3% non-Hispanic white, mean age = 19.9) completed
baseline questionnaires and were randomly assigned to complete a traditional II intervention,
a training-enhanced II intervention, or an information-only control activity at the end of the
lab session. Participants self-reported physical activity one- and two- weeks later. Results:
Controlling for baseline physical activity, intentions, and time, a bootstrapped generalized
estimating equation revealed no main effect of either intervention on subsequent physical
activity compared to the control condition (ps > .10). However, there was a significant
interaction between exercise intentions and conditions (II: b = .34, p = .006; II+T: b = .42, p =
.001), such that both II interventions led to significantly greater physical activity than the
control condition among participants with high intentions, but led to significantly less physical
activity than the control condition among participants with low intentions. There were no
significant differences in the simple slopes of intentions and physical activity between the two
intervention conditions. Conclusions: Both II interventions were effective for individuals with
high exercise intentions, but backfired for those with low intentions. Additionally, the
enhanced II intervention had no effect beyond the traditional II intervention on self-reported
physical activity. Potential explanatory factors, such as changes in plan quality, and memory
for plans will be discussed. While implementation intentions offer great promise as time-and
cost-effective interventions for physical activity, providers should account for existing exercise
intention levels of II intervention recipients.

CORRESPONDING AUTHOR: Scout M. Kelly, M.A., Kent State University, Kent, OH, 44242;
[email protected]
S2386 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D003 10:00 AM-11:00 AM

FACTORS ASSOCIATED WITH DIETARY DECISION MAKING IN PATIENTS WITH TYPE 2 DIABETES
AND CHRONIC KIDNEY DISEASE IN A BEHAVIORAL TRIAL

Lu Hu, PhD1, Huilin Li, PhD2, Kathleen Woolf, PhD, RD3, David St-Jules, RD, PhD1, Ram
Jagannathan, PhD4, David S. Goldfarb, MD5, Stuart Katz, MD5, Aditya Mattoo, MD6, Stephen
Williams, MD7, Lisa Ganguzza, MS, RD, CDN8, Mary Lou Pompeii, RDN, CSR, CDN, CDE4, Alex
Sierra, MS9, Zhi Li, MPH7, Mary Ann Sevick, ScD, RN10

1
Center for Healthful Behavior Change, NYU School of Medicine, New York, NY; 2New York
University, New York, NY; 3NYU Steinhardt, New York, NY; 4Center for Healthful Behavior
Change Department of Population Health New York University School of Medicine, New York,
NY; 5NYU Langone Medical Center, New York, NY; 6NYU Langone Medical Center, NEW YORK,
NY; 7New York University Langone Medical Center, new york, NY; 8NYU School of Medicine,
New York, NY; 9Center for Healthful Behavior Change Department of Population Health New
York University School of Medicine, NEW YORK, NY; 10Center for Healthful Behavior Change
Department of Population Health New York University School of Medicine, new york, NY

Purpose

Patients with type 2 diabetes (T2D) often are encouraged to make dietary changes. Yet, little
is known about the extent to which they make autonomous dietary decisions or associated
factors. In this report, we 1) describe participants involvement in dietary decisions regarding
food purchase and preparation, and 2) examine associations of dietary decision making
autonomy with sociodemographics and self-efficacy.

Methods

Participants were asked to describe their involvement in: 1) choosing or buying food brought
to their homes, and 2) making their meals and snacks. Scores for each question ranged from
0% (none) to 100% (all of the time), and were averaged to yield an autonomy score. The
Weight Efficacy Lifestyle survey was used to assess self-efficacy to lose weight. Linear
regression modeling was used to examine each factor separately and only those significant in
univariate models were entered in the multivariate model to examine joint effect.

Results
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2387

The sample (N=76) was mostly white (80.3%), male (60.5%), currently or living as married
(56.6%) with a mean age of 66.1 (SD=8.0) and BMI of 32.5 (SD=4.2) kg/m2. Participants
reported a mean autonomy score of 77.2% (SD=27.8%). In the univariate models, there were
no differences due to age, race, education, income, number of comorbidities, employment
status, and self-efficacy, except gender, marital status and BMI. When entering these 3 factors
into a multivariate model, only marital status (p < .001) and BMI (p=.048) remained
significant. More specifically, decision-making autonomy decreased by 1.4% for each unit
decrease in BMI and those who were or living as married scored 22.1% lower than those who
were not.

Conclusions

Participants who had a lower BMI, and were or living as married, had less dietary decision-
making autonomy. Engaging those with low dietary decision-making autonomy may be
important for lifestyle interventions involving dietary behavior change, but requires further
study.

CORRESPONDING AUTHOR: Lu Hu, PhD, Center for Healthful Behavior Change, NYU School of
Medicine, New York, NY, 10016; [email protected]
S2388 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D004 10:00 AM-11:00 AM

HOW ASSESSMENT MODALITIES AFFECT THE ENDORSEMENT RATE OF SKIN SELF-

EXAMINATION FOR EARLY SIGNS OF MELANOMA

Annett Krner, PhD, Adina Coriou, MA, Chelsea Moran, BSc, Beatrice Wang, MD CM, FRCP,
CSPQ, FAAD

McGill University, Montreal, PQ, Canada

Background: Early detection and melanoma removal are the most effective methods to
decrease melanoma-related mortality and the majority of melanomas are self-detected.
Regular skin self-examination (SSE) for early signs of subsequent melanomas is recommended
worldwide by clinical guidelines as essential part of melanoma follow-up care. A systematic
review in preparation (Coroiu et al., also submitted for presentation at SBM meeting)
identified 14 randomized controlled trials testing psycho-educational interventions to improve
SSE practice in individuals at increased risk for melanoma. It was found that the
operationalization of SSE varies widely across studies. As a follow up, the current study
examines how different assessment modalities may result in different endorsement rates of
SSE behaviors among melanoma survivors. Further, our study examines how predictors of SSE
behavior may change depending on the operationalization of SSE.

Methods: Our longitudinal study enrolled 242 melanoma survivors (Mage 59 years, 49%
female). SSE was assessed with a 1-item measure commonly used in melanoma research and
by inquiring about the frequency of SSE for 5 body areas (head, front upper and lower body,
back upper and lower body). SSE behavior data were collected at baseline and at 3 and 12
months after a dermatological education about SSE. Analyses included descriptive statistics
for the different SSE assessment modalities and linear regressions using medical and socio-
demographic data, as well as psychosocial variables (assessed at 3 months post intervention)
to predict SSE (assessed at 12 months post intervention)

Results: When we asked How often did you carefully and purposefully examine your own
skin for suspicious changes?, 3% of melanoma survivors indicated that they had not self-
examined their skin over the past 3 months whereas 97 % reported checking their skin at least
once. When we inquired about the frequency of SSE for each of the 5 body parts: monthly
whole-body SSE was endorsed by 46% and 80% reported at least 1 whole-body SSE over the
past 3 months.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2389

In regression analyses, psychosocial variables such as tumor fear, social support seeking,
physician support, SSE attitude and SSE self-efficacy explained between 12% and 31% of the
SSE outcome depending on which operationalization of SSE practice was used; while
melanoma stage and time since diagnosis explained between 1% and 6% of the variance.

Conclusions: Our study illustrates that findings differ depending on the way SSE behavioral
practice is operationalized. This stresses the need for the development and validation of a
measure that could be used across studies in order to allow for the synthesis of findings of
different research teams. We hope the current study will stimulate a lively discussion about
the most promising ways to operationalize SSE and to test the psychometric properties of this
measure.

CORRESPONDING AUTHOR: Annett Krner, PhD, McGill University, Montreal, PQ, H3A 1Y2;
[email protected]
S2390 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D005 10:00 AM-11:00 AM

SIMULATION AND COMPARISON OF SUN PROTECTION INTERVENTIONS UTILIZING THE

THEORY OF PLANNED BEHAVIOR: AN AGENT-BASED APPROACH

W. Kyle Hamilton, BA, Linda D. Cameron, Ph. D.

University of California, Merced, Merced, CA

One of the most common forms of cancer is melanoma. Cancer treatment costs directly
attributable to melanoma are about $3.3 billion per year in the U.S. Sun protection behaviors
such as sunscreen use and minimizing sun exposure are advertised as ways to prevent skin
cancers. Yet despite the costs and dangers of sun exposure, protective sun behaviors such as
sun screen use remains low. Influencing people to adopt these behaviors remains a
challenging public health issue. According to the Theory of Planned Behavior, intentions and
behaviors are shaped by attitudes, subjective norms, and perceived behavioral control. These
psychological factors could therefore be key targets of health communications aimed at
promoting sun protection behaviors. This study tests an agent-based model that compares
sun protection interventions aimed at young adults. This model is based on previous
interventions which have targeted all three facets of the Theory of Planned Behavior as well
as recent evidence that norms are the strongest predictors of intentions to engage in sun
protective behavior.

Using simulations of this agent-based model, we observed the transmission and spread of a
behavioral intervention within a social network that included users who have not previously
adopted the intervention. We used effect sizes from a recent meta-analysis on sun protection
interventions on reported subjective norms for our models initial conditions. For our network
size, we designed the model to have 375 nodes (users), which is the average number of
Facebook friends for a college undergraduate. By identifying the diffusion of norms and
intentions to engage in sun protection behaviors within the simulated social network, we
identified the factors which best influence change or maintain behavioral adherence on a
larger scale and thus represent key targets for interventions.

CORRESPONDING AUTHOR: W. Kyle Hamilton, BA, University of California, Merced, Merced,

CA, 95340; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2391

D006 10:00 AM-11:00 AM

SINGLE AND MULTIPLE HEALTH BEHAVIOR CHANGE INTERVENTIONS IN CANCER SURVIVORS:

A SYSTEMATIC REVIEW

Ellen M. Kinner, MA, Anne I. Roche, BS, Alan J. Christensen, PhD

University of Iowa, Iowa City, IA

Single and Multiple Health Behavior Change Interventions in Cancer Survivors: A Systematic
Review

Health behavior change has become an important addition to cancer survivorship care.
Favorable changes in physical activity, diet, and weight management behaviors promote long-
term health and improve quality of life while simultaneously reducing the risk of additional
health and treatment-related problems in cancer survivors. Although there is a large and
growing literature on health behavior change interventions in cancer survivors, it is currently
not known if these interventions differ in effectiveness when addressing more than one
behavior or what factors may influence such changes when more than one behavior is
targeted. This review examined the effects of single health behavior change interventions
(SHBC) and multiple health behavior change (MHBC) interventions in cancer survivors that
targeted physical activity (PA), diet quality (DQ), and weight management (WM). The specific
aims of this review are to 1) assess and compare the effectiveness of SHBC and MHBC
behavior change interventions; and 2) identify and compare the sample and intervention
characteristics that may impact observed effects. A systematic search strategy yielded 31
randomized controlled trials that met inclusion criteria with nine identified as SHBC and 22
identified as MHBC. Generally, MHBC interventions in comparison to SHBC appear to be more
effective in promoting behavior change in the physical activity and diet quality trials, while no
discernable differences were seen in weight management trials. Due to the substantial
variability observed in sample and intervention characteristics within the MHBC and SHBC
domains, few notable comparisons between the domains emerged. Overall, randomized
controlled trials of health behavior change interventions in cancer survivors have shown
interventions to be effective in facilitating behavior change in comparison to control groups. ,
Additionally, MHBC interventions appear to be superior to SBHC interventions when targeting
physical activity and dietary quality behaviors. This difference in effectiveness was not
observed when examining weight management behaviors. Given the amount of variability
seen across these studies, a great deal of work is still needed to clarify what aspects of these
S2392 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

interventions are most effective in facilitating and maintaining health behavior change in
cancer survivors.

CORRESPONDING AUTHOR: Ellen M. Kinner, MA, University of Iowa, Iowa City, IA, 52242;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2393

D007 10:00 AM-11:00 AM

STAGES OF CHANGE AND FRUIT AND VEGETABLE CONSUMPTION IN MINORITY POPULATIONS

Elizabeth M. Orsega-Smith, PhD, Shannon Robson, PhD, Katie Greenawalt, BS

University of Delaware, Newark, DE

Data from 2007-2010 show 75.5% of adults were below recommendations for fruit and 87.3%
were below for vegetables. Diet-related disparities exist among minority populations with
Black and Hispanic populations often consuming less fruits and vegetables compared to
Caucasian populations in the United States. Due to the disparities in health behaviors and
outcomes between populations, nutrition education programs such as those offered through
Cooperative Extension Services often target the most vulnerable communities. We
investigated the stage of change along with other factors that influence fruit and vegetable
consumption in these diverse populations.

Methods:

A total of 126 adults (18 years-old) from two communities in Delaware were enrolled in a
seven-month community-based participatory research study through the Universitys
Cooperative Extension Agency, which formed the basis of the educational programs that were
part of the challenge. Participants who agreed to participate in the research study completed
a comprehensive baseline questionnaire that assessed demographics, health-related status,
fruit and vegetable staging, fruit and vegetable self-efficacy, and self-reported intake of fruits
and vegetables. Bivariate correlations were conducted to determine what demographic and
health variables were associated with fruit and vegetable intake and were included as
covariates in subsequent analyses. A block linear regression examined the relationship
between fruit and vegetable staging, fruit and vegetable self-efficacy and fruit and vegetable
intake.

Results:

Of the 126 participants, 64 identified as Hispanic and 62 as African American, corresponding

to the community of residence. The mean age for was 46.7 15.3 years and 77.9% of
participants were female. Most participants reported being in the contemplation (39.2%) or
preparation stage (48.5%).Fruit and vegetable staging was significantly associated with fruit
and vegetable consumption (F (7,69) = 7.49). The total variance explained by the model was
37.1%, p < 0.001.

Conclusion:
S2394 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

These findings suggest stages of change is strongly related to fruit and vegetable consumption
in these diverse minority populations. Utilizing this information may be advantageous when
developing programs focused on increasing the behavior of fruit and vegetable intake of
ethnic minorities by developing programs to move participants along the stages of change
continuum toward action.

CORRESPONDING AUTHOR: Elizabeth M. Orsega-Smith, PhD, University of Delaware, Newark,

DE, 19716; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2395

D008 10:00 AM-11:00 AM

THE EFFECT OF A PEER MODELING WORKPLACE PHYSICAL ACTIVITY INTERVENTION FOR

WOMEN

Sheri Rowland, MSN, APRN-BC1, Bernice Yates, PhD, RN2, Marlene Cohen, PhD, RN, FAAN2,
Kris Berg, PhD3, Carol Pullen, EdD, RN2, Kevin A. Kupzyk, PhD2

1
University of Nebraska Medical Center, Gretna, NE; 2University of Nebraska Medical Center,
Omaha, NE; 3University of Nebraska Omaha, Omaha, NE

Purpose
People compare themselves to others for self-evaluation, practical information, and
motivation for health behavior change. Peers models have been used effectively in health
behavior interventions to recruit, educate, and motivate but not specifically to provide
vicarious experience for healthy physical activity (PA). The purpose of this study was to
evaluate the effect of a 12-week peer-modeling intervention on self-efficacy, motivation, and
social comparison for PA.

Theoretical Framework
Social cognitive theory and social comparison theory support the intervention. Peer-models
in this study provided vicarious experience and were a focus of comparison for an active
lifestyle.

Methods
Inactive women 19 to 65 years old were recruited from a Midwestern health system.
Participants were randomized to either the attention control group (ACG) (n = 24) or the
intervention group (IG) (n = 26). Both groups met every other week for six sessions for a lunch
and learn presentation. The ACG received general health information. The IG received specific
information on PA and observed a different active peer each session who presented their
personal story for being active. Both groups completed pre- and post-intervention measures
of: self-efficacy and motivation for PA, social comparison orientation (abilities and opinions),
and motives for making social comparisons (distancing, similarity, enhancement, modeling,
future self). Hierarchical linear modeling was used for data analysis.

Results
Forty-three women (ACG n = 22; IG n = 21) completed both pre and post testing. There were
no significant group X time interaction effects. There were trends favoring the IG in a majority
of social comparison measures. The IG increased comparison on abilities (pre/post M = 15.1 to
15.5), opinions (pre/post M = 18.3 to 19.1), modeling (pre/post M = 8.6 to 9.9), and future self
S2396 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

(pre/post M = 18.5 to 20.2) The ACG decreased comparison on abilities (pre/post M = 15.6 to
15.0), opinions (pre/post M = 20.0 to 18.8), modeling (pre/post M = 8.46 to 8.41), and future
self (pre/post M = 18.7 to 18.4).

Conclusions
Active peer-models may be a viable source of vicarious experience to increase PA through
social comparison processes. A fully powered study is needed to evaluate intervention effects
on self-efficacy and motivation for PA.

CORRESPONDING AUTHOR: Sheri Rowland, MSN, APRN-BC, University of Nebraska Medical

Center, Gretna, NE, 68028; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2397

D009 10:00 AM-11:00 AM

THE EFFECT OF COGNITIVE-BEHAVIORAL WEIGHT LOSS INTERVENTIONS ON PSYCHOLOGICAL

OUTCOMES: A META-ANALYSIS

Ariane Jacob, Ph.D student1, Simon L. Bacon, PhD2, Gregory Moullec, PhD3, Catherine Laurin,
PhD4, Tovah Cowan, BSc.5, Christina Kazazian, BA2, Candace Raddatz, BSc6, Kim L. Lavoie,
PhD7

1
Universit du Qubec Montral, Montral, PQ, Canada; 2Concordia University, Montreal,
PQ, Canada; 3Public Health School, University of Montreal, Montreal, PQ, Canada; 4hpital du
sacr-Coeur de Montral, Montreal, PQ, Canada; 5Concordia University, Beaconsfield, PQ,
Canada; 6Concordia University, Haley Station, ON, Canada; 7University du Quebec a Montreal,
Montreal, PQ, Canada

Objectives: To examine the effects of cognitive behavioral therapy weight loss (CBTWL)
interventions on psychological outcomes (eating behaviors [cognitive restraint,
emotional/binge eating], and depressive/anxiety symptoms) and on weight reduction in
adults with overweight or obesity.

Methods: Selection criteria: 1) randomised controlled clinical trials of CBTWL intervention

versus a comparison intervention, 2) study that included weight loss and psychological
outcomes, 3) patients who were overweight/obese.

Results: Twelve studies (6805 participants) were included. The average weight loss difference
between each arms was -1.70 kg (95%CI, -2.52 to -0.86, p

Conclusions: Beyond weight reduction, literature provides evidence suggesting that CBTWL is
an efficacious therapy for increasing cognitive restraint and reducing emotional eating.
However, CBTWL does not seem to be superior to other interventions for decreasing
depressive symptoms. Future studies should focus on psychological outcomes to better
understand the mechanistic role of these factors.

CORRESPONDING AUTHOR: Ariane Jacob, Ph.D student, Universit du Qubec Montral,

Montral, PQ, H2G 2J8; [email protected]
S2398 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D010 10:00 AM-11:00 AM

THE EFFECT OF IN-SESSION CHANGE LANGUAGE ON ALCOHOL OUTCOMES AMONG

EMERGENCY DEPARTMENT PATIENTS.

Reyna Puentes, BS1, Jon Houck, PhD2, Miriam Alvarez, MA1, Sandra Oviedo-Ramirez, BA1,
Craig Field, PhD, MPH1

1
Latino Alcohol and Health Disparities Research (LAHDR) Center, Department of Psychology,
The University of Texas at El Paso, El Paso, TX; 2Center on Alcoholism, Substance Abuse, and
Addictions, University of New Mexico, Albuquerque, NM

Motivational Interviewing (MI) is a clinical approach that focuses on enhancing a patients

readiness to change by exploring ambivalence while evoking and strengthening the patients
own verbalized reasons and motivations for change. The technical component of MI is based
on an implicit causal chain hypothesis, in which increased patient in-session verbalized
arguments for change (i.e. change talk) and decreased language in favor of maintaining the
status quo (i.e. sustain talk) predict behavior change. Researchers have found that frequency
of patient in-session change and sustain talk independently predict drinking outcomes. Efforts
to enhance the efficacy of brief interventions in reducing risky drinking among patients has
led to the adoption of the theoretical approach of MI. The present study tests this causal
chain hypothesis by investigating the relationship between in-session client language and
drinking reported at 3-month follow-up.

Emergency department patients were screened for heavy drinking and were randomly
assigned to one of three treatment conditions: brief advice, brief motivational intervention,
and BMI+Booster. Participants (n = 162, Mage = 36.6, SD = 12.4) were 75% male and 45% non-
Hispanic White. Interview audio recordings were coded using the Motivational Interviewing
Skill Code 2.5 and analyzed using the CASAA Application for Coding Treatment Interactions
software to assess clinician and client in-session behaviors. Patient maximum number of
drinks per week was regressed on sustain talk and change talk, covarying for baseline
drinking. Patient sustain talk was significantly related to maximum drinks per week at the 3-
month follow-up, accounting for approximately 3.1% of the variance in the outcome (b = 4.24,
SE = 1.79, t = 2.38, p = .019). Change talk was not significantly related to drinks reported at 3-
months, however the relationship was in the expected direction.

Our results provide support for the causal chain mechanism proposed by MI theory, given the
significant effect of patient in-session sustain talk on drinking outcomes at 3-months. More
specifically, our findings suggest higher levels of sustain talk (i.e. patient language against
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2399

change, in favor of maintaining the status quo) is related to higher maximum number of drinks
per week at 3-month follow-up, highlighting the influence sustain talk can have on predicting
behavior change. Future studies should pay particular attention to how to properly address in-
session sustain talk and could investigate how to enhance the effectiveness of change talk on
future behavior change.

CORRESPONDING AUTHOR: Reyna Puentes, BS, Latino Alcohol and Health Disparities Research
(LAHDR) Center, Department of Psychology, The University of Texas at El Paso, El Paso, TX,
79912; [email protected]
S2400 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D011 10:00 AM-11:00 AM

TRANSLATING AN INTENSIVE BEHAVIORAL INTERVENTION FOR DELIVERY BY CHWS

April Idalski Carcone, PhD1, Deborah Ellis, PhD1, Gloria Palmisano, na2

1
Wayne State University, Detroit, MI; 2CHASS Center, Inc., Detroit, MI

We previously demonstrated the efficacy of Multisystemic Therapy (MST), an intensive,

home-basedpsychotherapy, to improve diabetes management and metabolic control and
reduce diabetes-related hospital admissions among urban, minority youth with poorly
controlled type 1 diabetes (T1D). The goal of this study was to test an adapted version of MST
Fit Families (FF) for delivery by community health workers (CHWs).FF was developed and
pilot tested in collaboration with a federally qualified health center in Detroit that provides
primary health care and support services to underserved minority families. Forty-eight youth
with poorly controlled T1D (HbA1c9 %) and their primary caregivers were randomized to
either FF plus treatment as usual (TAU) (N=24) or TAU alone (N=24). Adolescents were 14.2
years old (SD=2.4), 37% male, and 78% African American. Mean HbA1c was 11.5% (SD=1.9%)
at study entry. Caregivers were 88% biological mothers, 47% were single parents, and the
median income was $20-29K (IQR= $10-19K, $50-59K). Families randomized to FF received 6
months (M=12.3, SD=9.5) of home-based cognitive-behavioral treatment delivered by a CHW,
and included disease education and skills training, but also parenting skills, family
communication skills, and skills for communicating effectively with medical care providers and
school personnel regarding diabetes care. Treatment began with a functional assessment to
assess areas of strengths and needs and an individualized treatment plan was developed to
address the specific barriers to illness management for each family. Data were collected at
baseline and six-month follow-up (treatment completion). FF adolescents demonstrated
improvements in HbA1c (-0.78%) and frequency of blood glucose monitoring (+0.27
tests/day). TAU adolescents had worsening HbA1c (+0.13) and decreased frequency of blood
glucose monitoring (-0.45) [tHbA1c(41)=1.48, p=.073 and tBGM(42)=-1.80, p=.040]. In per
protocol analysis, treatment completers demonstrated an even greater improvement in their
HbA1c (-1.12%) versus TAU (+0.15) [t(37)=2.02, p=.026]. BGM results were similar (MFF=+0.24
versus MTAU=-0.54, t(38)=-1.81, p=.039. These outcomes are comparable to those observed in
our previous trials where MST was delivered by masters-level mental health professionals.
Results suggest that an evidence-based intervention to improve illness management in youth
with poorly controlled diabetes can be delivered by CHWs. Delivery by paraprofessionals
could increase real-world utility of the intervention.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2401

CORRESPONDING AUTHOR: April Idalski Carcone, PhD, Wayne State University, Detroit, MI,
48202; [email protected]
S2402 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D012 10:00 AM-11:00 AM

USE OF STATE SPACE GRIDS TO CHARACTERIZE DYADIC INTERACTIONS IN SMOKING

CESSATION COUNSELLING SESSIONS: A FEASIBILITY STUDY

Heather Gainforth, PhD1, Fabiana Lorencatto, PhD2, Karl Erickson, PhD3, Kristy L. Baxter,
Human Kinetics4, Kailey Owens, Bachelor of Human Kinetics5, Susan Michie, BA, MPhil, DPhil6,
Robert West, BSc PhD7

1
University of British Columbia, Kelowna, BC, Canada; 2City University London, London, N/A,
United Kingdom; 3Michigan State University, East Lansing, MI; 4University of British Columbia
Okanagan, Kelowna, BC, Canada; 5UBC Okanagan, Edmonton, AB, Canada; 6University College
London, London, N/A, United Kingdom; 7University College London, London, England, United
Kingdom

Background: The content of behavior change counselling sessions is typically characterized in

terms of the presence (or absence) behavior change techniques (BCTs) used by practitioners.
However, understanding how BCTs operate in practice requires a method for characterizing
the reciprocal, dynamic and real-time nature of interactions between practitioners and
clients. State-space grids (SSGs) are an observational, dynamic systems methodology used to
map the trajectory of dyadic interactions in real time. By mapping the flow of events in terms
of practitioner and client actions, SSGs seem potentially well suited to characterize behavior
change counselling sessions. Smoking cessation counselling sessions can be reliably and
usefully characterized in terms of practitioner-delivered and client-received BCTs and so this
domain provides a useful starting point for assessing the feasibility of using the SSG
methodology. Aim: To examine the feasibility of using the SSG methodology for
characterizing practitioners delivery of and clients response to BCTs in smoking cessation
behavioral counselling sessions. Methods: Six smoking cessation counselling sessions were
video recorded, transcribed and divided into single speech acts by either the practitioner or
the client (n = 3404 speech acts). All speech acts were coded for both content and duration
using published frameworks for specifying practitioner-delivered and client-received BCTs in
stop cessation interactions. This coding scheme was used independently by two researchers
and inter-rater reliability was assessed. Indices of practitioner-client interaction dynamics
were derived: 1) reciprocity (i.e. attractor states, content congruence, conditional pairing) and
2) temporal patterning (i.e. variability, inter-grid distance, lagged phase plots, sessional
macro-patterning). The ability of the indices to differentiate between sessions involving
different practitioners and clients was assessed. Results: Inter-rater reliability was moderate
to high with 72% initial agreement which was resolved through discussion. Indices of
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2403

reciprocity and temporal patterning were able to differentiate between sessions involving
different practitioners and clients. Discussion: State space grids provide a useful method for
characterizing the dynamics of practitioner-delivered and client-received BCTs in smoking
cessation counselling sessions.

CORRESPONDING AUTHOR: Heather Gainforth, PhD, University of British Columbia, Kelowna,

BC, V1V1V7; [email protected]
S2404 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D013 10:00 AM-11:00 AM

USING BETWEEN-SESSION TELEPHONE CALLS TO OVERCOME BARRIERS TO PHYSICAL ACTIVITY

IN A GROUP-BASED INTERVENTION

JoEllen Wilbur, PhD, RN, FAAN1, Michael Schoeny, PhD1, Susan Buchholz, PhD, APN2, Arlene
Miller, PhD, RN, FAAN2

1
Rush University, Chicago, IL; 2College of Nursing Rush University, Chicago, IL

Introduction: The efficacy of interventions to increase physical activity (PA) may vary as a
function of participants barriers to PA. Prior research has been limited and considered
primarily demographic moderators of intervention efficacy. The aim of this presentation is to
determine whether individual barriers (demographic, physical health, psychological health,
neighborhood factors, perceived barriers to PA, social support for PA) moderate treatment
effects on increases in PA. Methods: Three treatment conditions tested the relative efficacy
of a group-based PA intervention alone or supplemented by either personal or automated
phone calls made between group meetings. From 2010 to 2012, 284 African American women
(ages 40-65) living in the Chicago, IL, area were randomized to one of the three treatment
conditions. Data collection occurred at baseline as well as 24 and 48 weeks after baseline.
Moderation of intervention effects by barriers to PA were tested across four outcome
measures (self-reported moderate-vigorous PA, self-reported walking, accelerometer steps,
and aerobic fitness) using multilevel mixed-effects analyses. Results:Significant condition by
barrier interaction effects for the accelerometer steps outcome were found for economic
hardships, F(2, 171)=3.37, p=.037; depressive symptoms, F(2, 173)=4.96, p=.008;
neighborhood crime rate, F(2, 172)=2.46, p=.088; and perceived barriers to PA, F(2, 171)=2.65,
p=.073. For aerobic fitness, intervention effects were moderated by economic hardships, F(2,
218)=4.00, p=.020; and perceived pain, F(2, 217)=3.50, p=.032. Increases in the outcome
variables were greater for the conditions in which group sessions were supplemented with
personal and/or automated calls. Conclusion: Among participants with greater barriers to PA,
supplementing the intervention group meetings with between-session personal and/or
automated phone calls may be an effective way to strengthen intervention effects. Future
research should examine the use of between-session phone calls as treatment supplements
for non-responders in the context of adaptive interventions.

CORRESPONDING AUTHOR: JoEllen Wilbur, PhD, RN, FAAN, Rush University, Chicago, IL,
60126; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2405

D014 10:00 AM-11:00 AM

USING MACHINE LEARNING TO REFINE BEHAVIOR CHANGE THEORY: THE CASE OF

PARTICIPANT ENGAGEMENT IN MHEALTH INTERVENTION

Angela Fidler Pfammatter, PhD1, Nabil Alshurafa, PhD1, Bonnie Spring, PhD2

1
Northwestern University Feinberg School of Medicine, Chicago, IL; 2Northwestern University,
Chicago, IL

Current technology allows for logging of specific actions within an app to track engagement,
thus an opportunity to use strategies like machine learning emerges. Leveraging this depth of
information enables refinement of our behavioral theories and constructs as it relates to
mHealth interventions. Participant engagement is one construct that is assumed to lead to
change in behavior during mHealth interventions. Engagement refers to a myriad of concepts
within behavioral medicine research and generally refers to any action done to achieve the
best health outcome from available services. Though many researchers propose engagement
as an outcome or mediator, there is not a standardized operational definition for engagement
when evaluating interventions. Despite mHealth apps requiring various types of engagement,
little to no work has been done to conceptualize these types or their differential contribution
to outcomes. At worst, trials do not report how often or to what extent an app is used when
evaluating the effect on health outcomes. At best, few report how many days a participant
opens the app, wears a device, or logs into an mHealth system. Machine learning was used in
an ongoing mHealth multiple behavior change trial in college students over 6 months to
identify actions that comprehensively define engagement such that specific actions or profiles
may be linked to better health outcomes. Aligned with intended use of mHealth apps,
engagement was divided into two primary categories: reviewing feedback and self-
monitoring. We applied a head-tail classification to classify sub-types of engagement. Five
sub-types of feedback engagement emerged: glances (< 7 seconds), check (7 17 seconds),
brief review (17-38 seconds), detailed review (38-68 seconds), and deep review (> 68
seconds). For self-monitoring, three subtypes emerged: brief interact (3 - 29 sec), detailed
interact (29-51 sec), and deep interact (> 52 seconds). This new operationalization enables
identification of profiles that can be used to label participants and to predict outcomes.
Future work will determine what type or combination of types of engagement predict
important mechanisms and clinical outcomes in health behavior change research.

CORRESPONDING AUTHOR: Angela Fidler Pfammatter, PhD, Northwestern University Feinberg

School of Medicine, Chicago, IL, 60611; [email protected]
S2406 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D015 10:00 AM-11:00 AM

USING PSYCHOPHYSIOLOGICAL METHODS TO UNDERSTAND THE EFFECTS OF DIFFERENT

POSITIVE FACIAL EXPRESSION MANIPULATIONS

Diana Dominguez Ortega, Undergraduate1, Tina Singh, Undergraduate 2, Shuo Zhang,

Undergraduate3, Susan E. Koh, Bachelor of Arts in Psychology4, Aline Carcano, Bachelor of
Arts in Psychology & Social Behavior5, Liana Gheorma, BA6, Marie P. Cross, M.A.1, Sarah D.
Pressman, PhD7

1
University of California, Irvine, Irvine, CA; 2University of California, Irvine, Newport Beach,
CA; 3University of California Irvine, Newport Beach, CA; 4University of California, Irvine,
Foothill Ranch, CA; 5University of California, Irvine, Los Angeles, CA; 6University of North
Carolina, Chapel Hill, Chapel Hill, NC; 7University of California- Irvine, Irvine, CA

Differences in positive facial expressions have long been implicated in psychological disorders,
physiological activity, and even health outcomes. Despite this, little work has objectively
validated experimental methods of positive expression manipulation. The current study
tested the differences between three methods of inducing smiles: a newly developed
commercial device (the Smile Stick), chopsticks held in the mouth (used in past studies), and
natural smiling. Participants (N = 32, data collection underway) performed three different
tasks while electromyography (EMG) recorded orbicularis oculi (eye) and zygomaticus major
(cheek) muscle activation in the face while participants demonstrated their assigned
conditions. An analysis of variance revealed that natural smiling was associated with higher
zygomaticus major muscle activation compared to the other conditions, F(1.302,
36.46)=10.629, p=.001. Paired contrasts of smile stick versus natural or chopsticks versus
natural also revealed significantly higher zygomaticus activity in the natural condition
(F(1,28)=10.960, p=.003 and F(1.28)=12.356, p=.002, respectively). There was no significant
difference in orbicularis oculi muscle activation between groups, F(2,56)=1.689, p=.194.
Participants self-reported feeling less tired after naturally smiling or holding chopsticks as
compared to the smile stick, F(1.587,44.441)=7.206, p=.004. Of note, while all conditions
activated both muscle groups above baseline, cheek muscles were 52.8% more activated from
natural smiles. When using the Smile Stick, there was a 1,478.2% increase in the zygomaticus
major (cheek) muscle activity and a 277.2% increase in the orbicularis oculi (eye) muscle
activity compared to baseline. Thus, while future studies can continue to use oral device
methodologies with confidence, for the largest zygomaticus major effects in future health and
psychology research, new methods should be developed where natural smiles could be
encouraged covertly via other non oral device methods.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2407

CORRESPONDING AUTHOR: Diana Dominguez Ortega, Undergraduate, University of California,

Irvine, Irvine, CA, 92612; [email protected]
S2408 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D016 10:00 AM-11:00 AM

VACCINE HESITANCY AMONG MILLENNIALS: CONCERNS, BELIEFS AND ATTITUDES

Julie E. Volkman, PhD, Kirsten L. Hokeness, PhD, Alyce Viens, MA, Alexandra Dickie, N/A

Bryant University, Smithfield, RI

Background: Vaccines have proven to be the most effective mechanism available to prevent
disease; yet an increase in vaccine hesitancy has led to a subsequent increase in vaccine
refusal. It is suggested that the largest population of vaccine-hesitant individuals are young
adults ranging from 18-29 years of age.

Objective: To assess vaccine hesitancy in the millennial generation using the WHO Vaccine
Hesitancy conceptual items (Larsen et al., 2015) and understand if uncertainties exist related
to vaccines among millennials.

Method: Participants (N=180) from two northeastern universities answered the WHO Vaccine
Hesitancy items and additional questions related to uncertainty about vaccines. Questions
assessed contextual, individual and group influences, and vaccine specific issues related to
vaccine hesitancy along a Likert scale (1=Strongly Disagree to 5=Strongly Agree). Frequencies
were assessed using SPSS 21.0.

Results: Most participants were female (n=114; 63.3%), White/Caucasian (n=123; 68.3%) and
business majors (n=104; 57.8%). Respondents agreed on mandatory vaccines (M=3.34;
SD=.85) and agreed vaccines are a socially accepted norm (M= 3.71; SD=.62), but expressed
uncertainty and a desire to know more about vaccines (M=3.46; SD=.77). They believed
vaccines are safe (M=3.47; SD=.72), yet reported concern with getting new vaccines (M=2.93;
SD=.47) and the motives of the pharmaceutical industry in designing vaccines (M=2.61;
SD=1.05). For vaccine related information, participants trusted healthcare providers (M=4.10;
SD=.85) and parents (M=3.83; SD=.94), and less trust from the government (M=3.00;
SD=1.01), friends (M=2.92; SD=1.03), news media (M=2.71; SD=1.02) and social media
(M=2.20; SD=1.01).
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2409

Conclusion: Findings suggest young adults believe vaccines should be mandatory and are safe,
but have concerns with vaccine processes. Healthcare providers and parents are the best
sources of information for decision-making among this age group. Interventions to decrease
vaccine refusal may want to encourage more communication between healthcare providers,
parents and young adults about vaccines and their processes. Future research can begin to
examine uncertainties related to specific vaccines and the vaccine process.

CORRESPONDING AUTHOR: Julie E. Volkman, PhD, Bryant University, Smithfield, RI, 02917;
[email protected]
S2410 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D017 10:00 AM-11:00 AM

VALIDATING STAGES OF CHANGE FOR OBESOGENIC BEHAVIORS ACROSS ADOLESCENT

MINORITIES

Sasha A. Fleary, PhD1, Michelle Tagorda, MPH2, Sophia Kim, Phd3, Michelle Thompson, N/A3,
Claudio R. Nigg, PhD, FSBM3

1
Tufts University, Medford, MA; 2University of Hawaii at Manoa, Honolulu, HI; 3University of
Hawaii, Honolulu, HI

Background: Filipino adolescents are underrepresented in obesity research although Filipinos

are the second largest Asian American and Pacific Islander subpopulation in the US. Given that
increasing obesity rates, and inadequate physical activity (PA) and fruit and vegetables (FV)
consumption are characteristic of US adolescents with minority adolescents being more likely
to experience obesity-related chronic illnesses into adulthood, an understanding of how well
the theories of behavior change apply to Filipino adolescents is critical to addressing
obesogenic behaviors in this group. This study aimed to validate the Transtheoretical Model
of behavior change (TTM) for PA and FV among a majority Filipino adolescent population. The
convergent validity hypothesis was that adolescents grouped in the precontemplation,
contemplation and preparation stages would report less PA and FV compared to those in
action and maintenance stages.

Methods: Adolescents in grades 9-11 (N=164, 82% female, 70% Filipino) who were part of the
Health Action and Research Training project completed measures of PA (minutes/week) and
FV (times/day) in the past seven days, and PA and FV stages of change. To assess the validity
of the PA and FV stages of change with the respective behaviors, one-way ANOVAs and Tukey
HSD post-hoc tests were computed.

Results: There was a significant effect for fruits (action [M=1.53, SD=1.26] > contemplation
[M=.74, SD=.76], preparation [M=.79, SD=.90]) and vegetables (maintenance [M=1.39,
SD=1.24], action [M=1.30, SD=1.13] > contemplation [M=.45, SD=.31]) intake across the FV
stages of change. There was a significant effect of strenuous PA (contemplation [M=25.8,
SD=30.4], preparation [M=45.2, SD=52.4] < action [M=141.7, SD=111.4] < maintenance
[M=214.8, SD=114.4]) and moderate PA (contemplation [M=63.2, SD=74.9] < action [M=134,
SD=112.3], maintenance [M=139.6, SD=98.3]) across the PA stages of change.

Discussion: This study provides validity evidence for the TTM stages of change for FV and PA
among minority (predominantly Filipino) adolescents. This validation, in turn, extends the
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2411

generalizability of the stages of change construct across health behaviors, replicating other
adolescent studies and in other age groups.

CORRESPONDING AUTHOR: Sasha A. Fleary, PhD, Tufts University, Medford, MA, 02155;
[email protected]
S2412 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D018 10:00 AM-11:00 AM

WELLNESS SELF-COACHING AS A BOOSTER INTERVENTION FOR THE STANFORD SMALL GROUP

SELF-MANAGEMENT PROGRAMS

Thelma J. Mielenz, PT, PhD, OCS, Laura Durbin, MPH, MPH, Haomiao Jia, PhD

Columbia University Mailman School of Public Health, New York, NY

The enormous burden of multiple chronic diseases is clear in older adults, especially among
low-income and underserved minority populations. We evaluated the Stanford Small Group
Self-Management Programs (SSGSMP) plus wellness coaching as a booster intervention. We
conducted a pragmatic randomized controlled trial (RCT) with complete block design using
two intervention arms: 1) SSGSMP and 2) SSGSMP plus a wellness self-coaching program
initiated as a booster after SSGSMP completion. In addition to completion of a SSGSMP
course, inclusion criteria for the 125 participants were: 1) 55 or greater years old, 2) resident
of one of the five South Bronx New York City Housing Authority communities, 3) self-report of
two or more chronic diseases, 4) cognitively competent, 5) ambulatory (independently or
walker/canes), and 6) English- or Spanish-speaking. Participants (n=125) were enrolled from.
Certified wellness coaches led three groups (two in Spanish) for 24 one-hour weekly wellness
self-coaching phone sessions. Physical activity was the primary outcome, measured by the
Community Health Activities Model Program for Seniors. Secondary outcomes included the
following PROMIS v1.0 measures health domains: depression, fatigue, pain, physical function,
and sleep disturbance. Participant surveys were completed at baseline, 3-months, and 6-
months time points. We evaluated the outcomes of 121 participants in the RCT using
appropriate general linear models to test directly the joint effects of SSGSMP and the wellness
self-coaching as a booster. Data collectors and analyzers were blinded to group allocation,
and intent-to-treat analyses were conducted. Participants in the wellness coaching
intervention did not differ significantly from control participants on an improvement in
physical activity but did differ on self-reported physical functioning (1.3 units higher, p=0.03
on the Physical Function SF 20a). Wellness coaching post-SSGSMP was a booster to physical
function, a plausible upstream outcome for physical activity.

CORRESPONDING AUTHOR: Thelma J. Mielenz, PT, PhD, OCS, Columbia University Mailman
School of Public Health, New York, NY, 10032; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2413

D019 10:00 AM-11:00 AM

WHO BENEFITS FROM A DIETARY ONLINE INTERVENTION? EVIDENCE FROM ITALY, SPAIN,
AND GREECE

Ralf Schwarzer, Dr.1, Lena Fleig, PhD2, Lisa Warner, PhD3, Maryam Gholami, PhD4

1
Freie Universitt, Berlin, Berlin, Germany; 2Freie Universitt Berlin, Berlin, Berlin, Germany;
3
Freie Universit Berlin, Berlin, Berlin, Germany; 4University of California San diego, Berlin,
Berlin, Germany

Objective: The traditional Mediterranean diet includes high consumption of fruits, vegetables,
olive oil, legumes, cereals, nuts, moderate to high intake of fish and dairy products, and low
consumption of meat products. Intervention effects to improve adoption of this diet may vary
in terms of individuals motivational or volitional prerequisites. In the context of a three-
country research collaboration, intervention effects on these psychological constructs for
increasing adoption of the Mediterranean diet were examined.

Design: An intervention was conducted to improve Mediterranean diet consumption with a

two-month follow-up. Linear multiple level models examined which psychological constructs
(outcome expectancies, planning, action control, stage of change) were associated with
changes in diet scores. Setting: Web-based intervention in Italy, Spain, and Greece. Subjects:
N=454 adults; age range=18 to 65 years; Mage=42.2, SDage=10.4; n=112 at follow-up.
Results: Analyses yielded an overall increase in the Mediterranean diet scores. Moreover,
there were interactions between time and all four psychological constructs on these changes.
Participants with lower levels of baseline outcome expectancies, planning, action control, and
stage of change were found to show steeper slopes, thus greater behavioral adoption, than
those who started out with higher levels.

Conclusion: The intervention produced overall improvements in Mediterranean diet

consumption, with outcome expectancies, planning, action control, and stage of change
operating as moderators, indicating that those with lower motivational or volitional
prerequisites gained more from the online intervention. Individual differences in participants
readiness for change need to be taken into account to gauge who would benefit most from
the given treatment.

CORRESPONDING AUTHOR: Ralf Schwarzer, Dr., Freie Universitt, Berlin, Berlin, 14195;
[email protected]
S2414 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D020 10:00 AM-11:00 AM

A TOOL, NOT A CRUTCH': PATIENTS WITH BREAST CANCERS PERCEPTIONS OF THE IBM
WATSON FOR ONCOLOGY CLINICAL DECISION SUPPORT SYSTEM

Jada G. Hamilton, Ph.D., M.P.H., Elyse Shuk, M.A., Joy S. Westerman, B.A., Margaux C. Genoff,
B.S., Rachel S. Werk, B.S., Val Pocus, B.S., Jennifer L. Hay, Ph.D., Chasity Walters, Ph.D., R.N.,
Elena Elkin, Ph.D., Ayca Gucalp, M.D., Andrew D. Seidman, M.D., Marjorie G. Zauderer, M.D.,
Andrew S. Epstein, M.D., Mark G. Kris, M.D.

Memorial Sloan Kettering Cancer Center, New York, NY

Background: IBM Watson for Oncology (WFO) is a cognitive, cloud-based clinical decision
support system designed to inform oncologists choices about personalized, evidence-based
treatments and clinical trials for their patients. Although substantial technical and clinical
expertise has been devoted to the development of WFO, patients perspectives of this
emerging technology have yet to be examined. Clarifying and addressing patients preferences
about WFO could increase patient satisfaction and contribute to greater physician adoption of
this technology.

Methods: To understand patients perspectives about the acceptability of WFO, we are

conducting nine focus groups with patients who received therapies at our institution. Here,
we report preliminary findings from the first three focus groups conducted with patients with
breast cancers who received either neoadjuvant/adjuvant chemotherapy (n=6; in-person
group), chemotherapy for metastatic disease (n=3; in-person group), or chemotherapy
through a clinical trial (n=4; virtual group). Focus groups were audio-recorded, transcribed,
and analyzed through inductive thematic text analysis to identify common thematic findings.

Results: Across all groups, participants felt comfortable with the prospect of their oncologists
using WFO, as long as it was used as a supplementary tool to inform their treatment decision-
making. Participants strongly believed that their oncologists were the best providers of
personalized care, and questioned whether a computer could replicate this highly valued facet
of patient care. Participants raised concerns about potential physician over-reliance on the
decision support tool, which could undermine their trust in the physician. Participants
perceived the ability to expand the body of up-to-date medical knowledge available to
physicians at the point of care as a primary advantage of WFO; however, concerns about data
quality and accuracy were common. Participants also expressed information needs regarding
issues such as cost and confidentiality that, if unaddressed, would undermine their
acceptance of this decision support tool.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2415

Conclusion: Participants generally reacted favorably to WFO, but with serious caveats
regarding the quality of content and clinical implementation of this technology. Results
suggest a need for transparency to ensure that patients are comfortable with their physicians
integrating this tool into the treatment decision-making process.

CORRESPONDING AUTHOR: Jada G. Hamilton, Ph.D., M.P.H., Memorial Sloan Kettering Cancer
Center, New York, NY, 10022; [email protected]
S2416 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D021 10:00 AM-11:00 AM

CORRELATES OF TANNING DEPENDENCE AMONGST SEXUAL MINORITY MALES

Kelsey A. Nogg, B.A.1, Benjamin M. Rooney, B.A.1, Kalina M. Lamb, B.A.2, Aaron J. Blashill,
Ph.D. 1

1
San Diego State University, San Diego, CA; 2San Diego State University, San Deigo, CA

Background: Sexual minority males are an at-risk population for developing skin cancer.
Disparities in skin cancer diagnoses among sexual minority males are likely the result of
increased skin cancer risk behaviors. Indoor and outdoor tanning have been linked to
increased likelihood of developing skin cancer, and sexual minority males tan at rates
comparable to or greater than heterosexual females. One construct that may underlie
frequent tanning is tanning dependence: compulsive tanning behaviors associated with
tolerance, withdrawal, and craving of UV exposure. To date, preliminary research has shown
tanning dependence to be associated with increased skin cancer risk behaviors; however, no
known studies have examined tanning dependence among sexual minority males. Thus, the
current study aimed to explore the correlates of tanning dependence among a sample of
young, sexual minority males.

Methods: This study employed Facebook to recruit 238 sexual minority males ages 15 to 35
years (M = 24.56, SD = 5.44) living in San Diego County. Participants completed an online self-
report questionnaire, which included items assessing skin cancer risk behaviors (e.g., indoor
and outdoor tanning behavior, use of sunscreen). Tanning dependence was measured via the
Behavioral Addiction Indoor Tanning Screener (BAITS).

Results: Elevated tanning dependence was significantly and positively associated with
intention to indoor tan within the next three months, (r = .31, p < .001), frequency of indoor
tanning over the past three months (r = .44, p < .001), and frequency of outdoor tanning over
the past three months (r = .20, p = .002). Elevated tanning dependence was also significantly
and negatively associated with frequency of sunscreen use (r = -.15, p = .026).

Conclusion: The current study is the first known to explore tanning dependence among sexual
minority malesa population at risk for developing skin cancer. Elevated tanning dependence
was associated with several established skin cancer risk behaviors, suggesting that tanning
dependence may play a role in skin cancer disparities seen among sexual minority males.
Thus, future skin cancer prevention efforts aimed at this population may benefit from
exploring strategies to reduce tanning dependence in order to prevent the development of
skin cancer.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2417

CORRESPONDING AUTHOR: Kelsey A. Nogg, B.A., San Diego State University, San Diego, CA,
92104; [email protected]
S2418 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D022 10:00 AM-11:00 AM

DISGUST AS A PREDICTOR OF COLONOSCOPY SCREENING INTENTIONS

Maria M. Keller, BS

University at Buffalo, SUNY, Buffalo, NY

BACKGROUND: Colorectal cancer (CRC) is responsible for nearly 50,000 deaths each year and
though screening is an effective way to reduce deaths, screen compliance is only 50-60%
(American Cancer Society, 2016). An underlying factor in failure to screen or to comply with
screening preparation procedures may be disgust elicitation. Disgust causes the urge to
withdraw from potential contaminants. Fecal matter and violations of the bodily envelope are
two known elicitors of disgust (Curtis & Biran, 2001). Both are central features of colonoscopy
screening, making it plausible that disgust reactions result in failure to screen. However, this
has not been well tested and the affective mechanisms through which disgust impacts
screening are not known. METHODS: We conducted an online survey of US adults (N = 298)
using Amazon mTurk. We examined affective associations with colonoscopy, including disgust
associations, current feelings of disgust, dispositional disgust sensitivity, and three
components of intentions to be screened (intent to screen, intent to complete preparation
regimen, and intent to discuss screening with provider). RESULTS: Both overall negative
affective associations and specific associations with disgust significantly predicted intent to
screen, intent to complete preparation regimen, and intent to discuss screening with provider
( = -.15, p < .001, = -.10, p < .05, and = -.12, p < .01, respectively). By contrast, neither
current disgust nor disgust sensitivity were significant predictors, nor did either moderate the
relation of affective associations to screening. DISCUSSION: Affective associations of disgust
(that is, associated feelings of disgust with the colonoscopy screening procedure) was a
significant predictor of intentions to be screenedthe more disgust associated with
screening, the lower the screening intentions. This effect of disgust on screening seems to be
through the behavioral affective associations pathway (Kiviniemi, Voss-Humke, and Seifert,
2007)neither experienced disgust when thinking about screening nor dispositional
sensitivity to disgust were related to screening intentions. Interventions to encourage
screening colonoscopy should address the impact of disgust on noncompliance.

CORRESPONDING AUTHOR: Maria M. Keller, BS, University at Buffalo, SUNY, Buffalo, NY,
14261; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2419

D023 10:00 AM-11:00 AM

INDOOR TANNING PREDICTS SOME SKIN CANCER COGNITIONS AMONG FEMALE COLLEGE
STUDENTS

Jennifer Bowers, B.A.1, Anne Moyer, Ph.D.2

1
Stony Brook University, Port Jefferson, NY; 2Stony Brook University, Stony Brook, NY

Introduction

Skin cancer, a behaviorally influenced disease, which increasingly and adversely affects college
age individuals, is the most common cancer in the United States. One of the most potent risk
factors for developing skin cancer is the use of indoor tanning. This study aimed to understand
if self-reported indoor tanning predicts cognitions about skin cancer, specifically, risk
perceptions, attitudes, and behavioral intentions.

Method

Participants were 158 female students attending a Northeastern US university who indicated
that they were or were not indoor tanners.

In addition to demographic questions, measures were administered to assess previous

tanning behavior, attitudes about skin cancer, behavioral intentions, and risk perceptions.

Results

Ages ranged from 18 to 47, with a mean of 20.88 (SD 4.107). Sixty percent were White, 22.8%
Asian, and 4.4% Black. Forty-one percent self-reported tanning indoors at least once.

Three two-stage hierarchical multiple regressions were conducted with behavioral Intentions,
risk perceptions, and attitudes as dependent variables. Age, ethnicity, and skin type were
entered at stage one of each regression to control for these demographic factors. Tanner
status was entered at stage two.

For behavioral Intentions, the demographic variables contributed significantly to the

regression model, F (4,150) = 4.85, p < .001, and accounted for 11.5% of the variance.
Introducing the indoor tanning variable explained an additional 12.7% of the variance and this
change in R was significant, F (1,149) = 24.97, p < .001.
S2420 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

For risk perceptions, the demographic variables contributed significantly to the regression
model, F (4,152) = 7.25, p < .001 and accounted for 16.0% of the variance. Introducing the
indoor tanning variable explained an additional 3.0% of variation and this change in R was
significant, F (1,151) = 5.56, p < .03.

Multiple regression results were not significant for the Attitudes variable.

Discussion

These results demonstrate that indoor tanning may contribute to various cognitions about
skin cancer, which may negatively impact future health behavior. Indoor tanners were more
likely to view themselves as at-risk for skin cancer, but were less likely to intend to participate
in skin protective behaviors. In order to facilitate behavior change, future studies and
interventions should consider these differences to create tailored messages for those who
have used indoor tanning.

CORRESPONDING AUTHOR: Jennifer Bowers, B.A., Stony Brook University, Port Jefferson, NY,
11777; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2421

D024 10:00 AM-11:00 AM

LIMITED EFFECTS OF USING MASS MEDIA TO COMMUNICATE GENOMIC RISK INFORMATION

TO DEMOGRAPHICALLY DIVERSE ADULT SMOKERS

Erika A. Waters, PhD, MPH

Washington University in St. Louis, St. Louis, MO

Background: Past research suggests that using the mass media to communicate information
about the genetic basis for nicotine addiction may discourage smoking cessation. Yet, most
prior research has involved primarily white, highly educated samples.

Objective: To examine how alerting demographically diverse adult smokers about the
discovery of a genetic variant that confers a high risk of severe nicotine dependence and lung
cancer affects key smoking-related health cognitions and quitting behavior.

Methods: Adult smokers (n=343) were recruited from community locations (e.g.,
laundromats, bingo halls). Participants read either a real news article describing a pharmacys
decision to stop selling tobacco products or a real article describing the discovery of a gene for
nicotine addiction and lung cancer. Next, they completed measures of the following
outcomes: message acceptance, quit intentions, lung cancer-related cognitive and affective
perceived risk, worry, and perceived severity, and quitting self-efficacy and response efficacy.
Quit behaviors were assessed at 90-day follow-up (n=212). Data were analyzed using
ANCOVAs and logistic regressions, as appropriate. Race, sex, health literacy, numeracy,
perceived genetic knowledge, and nicotine dependence were covariates. The covariates were
explored as potential moderators.

Results: 51.0% of participants were racial minorities, 51.0% had no college experience, 63.3%
were women, and 42.0% were 50+ years old. The genetic article elicited more personal
reflection than the control article (p=.03, Meandiff=0.26), but there were no other main effects
on the remaining Time 1 outcomes (ps>.05). However, race moderated the effect of article
type on self-efficacy (p=.03) and quit intentions (p=.002). White smokers reported lower self-
efficacy (p=.002, Meandiff=0.53) and intentions (p=.01, Meandiff=0.61) in the genetic condition
but not in the control condition (ps>.05). This effect was not present among minority
smokers. Type of article did not affect quit behaviors at follow up (ps>.05).

Conclusion: Reading about the discovery of a genetic basis for nicotine addiction may produce
unfavorable changes in some health cognitions in white smokers, but not in minorities.
S2422 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Furthermore, there are no effects on actual quitting behavior. Mass media-based genomic
communication likely neither helps nor hinders smoking cessation campaigns.

CORRESPONDING AUTHOR: Erika A. Waters, PhD, MPH, Washington University in St. Louis, St.
Louis, MO, 63110; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2423

D025 10:00 AM-11:00 AM

MOBILE DISSEMINATION OF AN EFFICACIOUS DYSPHAGIA PREVENTION PROGRAM FOR HEAD

AND NECK CANCER PATIENTS DURING RADIATION

Shinn Shinn, Ph.D.

University of Texas M. D. Anderson Cancer Center, Houston, TX

Introduction: While cancer of the throat is highly curable, 39% of survivors experience serious
permanent swallowing problems. To prevent dysphagia, targeted swallowing exercises
performed during radiation have been shown to be effective in preventing radiation-induced
fibrosis of the swallowing musculature. Thirty-nine pharyngeal/ laryngeal cancer patients at a
community clinic in North Texas have been enrolled thus far on a mobile application pilot
study reinforcing adherence to swallowing exercises. The mobile app is based on a successful
10-session, in-person intervention program called PREPARE, which resulted in significantly
increased adherence to swallowing exercises in nonrecurrent pharyngeal/ laryngeal cancer
patients during radiation compared to randomized controls. Methods: All pilot participants
were provided with a pilot version of the mobile app, which includes video of exercises,
recipes, patient stories, didactic information and hyperlinks to over-the-counter products for
radiation-induced mucositis, dry mouth, skin burn and opioid-related constipation. Practical
skills to promote communication with caregivers and physicians during this stressful time are
also included. In-person speech pathology services are provided pre-radiation and 3 months
post-radiation. Participants were asked for feedback regarding the mobile system content
and delivery platforms both during radiation and 1 month post-intervention. Results:
Feedback from participants has been somewhat difficult to obtain remotely, but post-
radiation survey results indicate that patients are satisfied with the breadth and reading level
of the program content and valued the videos and patient stories most highly. Average time
spent per week was 40 minutes, with more time being spent in the first few weeks of
radiation and decreasing in the latter weeks as side effects worsened. Most participants
preferred to view the mobile app on their desktop computers rather than their phones
(average age of the sample was 57 years). While participants endorsed the importance of
having caregiver components in the mobile app, very few caregivers used the program. The
mobile application was delivered via secured email link but the full-scale study will be
delivered via a secure interactive website. Conclusion: This program has the potential to
significantly decrease the likelihood of long-term dysphagia in head and neck cancer survivors
being treated in rural, community-based clinics.

CORRESPONDING AUTHOR: Shinn Shinn, Ph.D. , University of Texas M. D. Anderson Cancer

Center, Houston, TX, 77030-3721; [email protected]
S2424 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D026 10:00 AM-11:00 AM

RELATIONSHIPS BETWEEN SALIVARY DIURNAL CORTISOL AND CANCER SURVIVORSHIP

OUTCOMES: A SYSTEMATIC REVIEW

Jennifer Hulett, PhD, APRN, FNP-BC1, Kristen Fessele, PhD, RN2, Margaret F. Clayton, PhD1,
Linda Eaton, PhD, RN, AOCN3

1
University of Utah, Salt Lake City, UT; 2University of Utah, Salt Lake City, UT, Bridgewater, NJ;
3
University of Utah, Mill Creek, WA

Background
Relationships between altered salivary diurnal cortisol (SDC) rhythms and clinical outcomes
are unclear, possibly due to variances in cortisol sampling. This systematic review examined
studies that assessed for SDC using a minimum of 2 collection times in 24 hours, including
awakening, in order to examine relationships between SDC and outcomes within cancer
patient populations.

Methods
Electronic database searching was conducted for the past 10 years using keywords: salivary
cortisol, cancer, and SDC assessment with combinations of subcategories for studies of adults
diagnosed with cancer in which salivary cortisol was an outcome measure. Main exclusion
criteria included endocrine-related disorders, exogenous glucocorticoids, and studies that did
not assess for SDC.

Results
The initial search yielded 13,913 articles, reduced to 165 after screening for inclusion and
exclusion criteria. An additional 135 studies did not meet SDC sampling criteria of interest,
resulting in a final sample of 30 articles. The majority of studies reported outcomes among
breast cancer survivors (n=20). Cortisol sampling across studies ranged from 1 to 4 days, with
a majority reporting 4 to 5 collection times per day (n=21) for 2 days (n=12). Few studies
(23%) reported objective monitoring of collection times (e.g., electronic bottle caps). Most
common SDC parameters reported were diurnal slope (n=19) and cortisol awakening response
(CAR) (n=12). Flattened cortisol slopes reported by 8 studies were associated with poorer
psychosocial outcomes and poorer cancer prognosis; however, 5 studies reported no change
in cortisol slope associated with yoga, sleep, stress, or psychosocial outcomes. CAR findings
were mixed, with higher CAR amplitude associated with tumor progression in 2 studies, while
a sharper dynamic rise in CAR was associated with less fatigue. A blunted CAR was associated
with poorer sleep, fatigue, and depression.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2425

Discussion
Despite our inclusion criteria focusing on recommended SDC sampling, directional
relationships between cancer outcomes and SDC across studies remain varied. Cortisol
variability within and between persons may account for differences across study findings.
Further, 2 to 3 days of SDC assessment may not consistently portray or detect measureable
changes in SDC. Further study is needed to inform rigor of SDC sampling with respect to
number of collection times per day and number of consecutive days. Finally, electronic device
monitoring of collection times would increase rigor and reproducibility of SDC sampling in
cancer research.

Learning Objectives
After attending this talk, participants should be able to:
1. Understand how the rigor of salivary diurnal cortisol sampling may affect the conclusions
regarding clinical outcomes.
2. Know the common SDC reporting parameters and potential relationships between these
parameters and survivorship outcomes.

CORRESPONDING AUTHOR: Jennifer Hulett, PhD, APRN, FNP-BC, University of Utah, Salt Lake
City, UT, 84112; [email protected]
S2426 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D027 10:00 AM-11:00 AM

SELF-COMPASSION, INSECURE ATTACHMENT, AND RELATIONSHIP FORMATION DIFFICULTIES

IN WOMEN POST-BREAST CANCER

Laura-Kate Shaw, BA Psych Hons1, Kerry Sherman, PhD1, Julie Fitness, PhD1, Elisabeth Elder,
PhD2, Breast Cancer Network Australia ., NA3

1
Macquarie University, Sydney, New South Wales, Australia; 2Westmead Breast Cancer
Institute, Sydney, New South Wales, Australia; 3BCNA, Camberwell, Victoria, Australia

Introduction: Many un-partnered women experience difficulty forming romantic relationships

after breast cancer, characterized by low interpersonal competence and dating anxiety. An
individuals insecure attachment style has been linked with these difficulties. Emerging
evidence suggests that self-compassion, a kind manner of treating oneself during suffering,
may underlie the relationship between attachment style and difficulties in forming romantic
relationships.

Aim: This study investigated whether self-compassion underlies the association between
insecure attachment style and relationship formation difficulties in women post-breast
cancer.

Methods: Women (N=156) diagnosed with breast cancer, who were either currently un-
partnered and desired a romantic relationship or had commenced a relationship post-
diagnosis, completed an online survey. Measures included attachment style (ECR-M36),
interpersonal competence (Interpersonal Competence Questionnaire), dating-related anxiety
(Dating Anxiety Scale), and self-compassion (Self-compassion Scale).

Results: Bootstrap mediation analyses revealed higher attachment anxiety and attachment
avoidance indirectly predicted lower interpersonal competence through lower self-
compassion. Higher attachment anxiety and attachment avoidance were also found to
indirectly predict greater dating anxiety through lower self-compassion.

Conclusions: Women with insecure attachment styles are at risk of experiencing difficulties in
forming new romantic relationships post-breast cancer. As attachment style can be somewhat
difficult to change, self-compassion may be a more appropriate target for interventions to
help facilitate romantic relationship formation in insecurely-attached breast cancer survivors.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2427

CORRESPONDING AUTHOR: Laura-Kate Shaw, BA Psych Hons, Macquarie University, Sydney,

New South Wales, 2109; [email protected]
S2428 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D028 10:00 AM-11:00 AM

SLEEP DURATION AND QUALITY OF LIFE IN KIDNEY CANCER SURVIVORS: INTERACTIONS WITH
SEDENTARY BEHAVIOR

Jaime Wong, B. A.1, Kerry S. Courneya, PhD2, Linda Trinh, PhD1

1
University of Illinois at Urbana Champaign, Urbana, IL; 2University of Alberta, Edmonton, AB,
Canada

Background: Sleep duration and sedentary behavior (SED) have been individually linked to
lower quality of life (QoL). As cancer survivors are more likely to report higher SED than
healthy individuals, they are at higher risk for lower QoL, negative mental health effects and
lower physical functioning, independent from the benefits of physical activity (PA).
Furthermore, cancer survivors have been shown to have shorter sleep duration that may
contribute to depression, fatigue and lower QoL. An understanding of how sleep duration
affects the relationship between SED and QoL in KCS is needed.

Purpose: To examine the associations between sleep duration and QoL in KCS, and the
interactive associations between SED and sleep duration on QoL for a work day (WD) and a
non-work day (NWD).

Methods: All 1985 KCS diagnosed between 1996 and 2010 were identified through a Canadian
Provincial Registry and mailed a survey including the Godin Leisure Time Exercise
Questionnaire, the modified domain-specific sitting time questionnaire (including sleep
duration), and several Functional Assessment of Cancer Therapy scales. Self-reported
demographic and medical variables were collected.

Results:Completed surveys were received from 481 KCS with Mage = 64.010.4 years, 64%
male, and 84% having localized kidney cancer. In the main hierarchical analysis, the
interaction effect of sleep duration by SED was significantly associated with physical well-
being (PWB) (p=.04), functional well-being (FWB) (p=.01), emotional well-being (EWB) (p=.05),
general QoL (p=.009) and fatigue, (p=.05) on a WD only. Follow-up simple slope analyses
demonstrated that among KCS with higher levels of sleep, higher levels of SED significantly
predicted lower levels of PWB (B=-.002; p=.09), lower levels of FWB (B=-.003; p=.08), and
higher levels of fatigue (B=-.009; p=.06). However, among KCS with lower levels of sleep,
higher levels of SED significantly predicted higher levels of general QoL (B=.006; p=.04) and
higher levels of EWB (B=.003; p=.00). There were no associations between SED and sleep
duration on QoL for a NWD.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2429

Conclusions: Sleep duration appears to play a role in moderating the relationship between
SED and QoL on a WD only. Higher levels of sleep duration coupled with higher levels of SED
may be detrimental to some aspects to the functional aspects of QoL (i.e. PWB, FWB, fatigue),
However, lower levels of sleep duration appear to positively influence general QoL and EWB
in the presence of higher SED. Having optimal levels of sleep and SED are likely to result in
better symptom management.

CORRESPONDING AUTHOR: Jaime Wong, B. A., University of Illinois at Urbana Champaign,

Urbana, IL, 61801; [email protected]
S2430 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D029 10:00 AM-11:00 AM

SOCIAL CONSTRAINT, SOCIAL SUPPORT, AND PERCEIVED BURDEN AMONG CHINESE CANCER
SURVIVORS: A MEDIATION MODEL

Kelly Yu-Hsin Liao, Ph.D.1, Qian Lu, Ph.D.2, Jennie Ho, Bachlor of Art3, Celia C.Y. Wong, MPhil2

1
Cleveland State University, Cleveland, OH; 2University of Houston, Houston, TX; 3University
of Houston, Sugar Land, TX

Although cancer has now become the leading cause of death among urban and rural residents
(Wang, Wei, Liu, Li, & Wang, 2012), very limited research has focused on Chinese cancer
survivors. The current study aimed to increase our understanding of this population by
examining a mediation model to determine what factors predict perceived burden, which in
turn contributes to fatigue and physical symptoms. Chwaliszs (1996) Perceived Stress Model
of Burden describes that coping and social support are both associated with perceived
burden, which can influence metal and health outcomes. Based on this model, the current
study hypothesized that social constraint or difficulty cancer survivors experience in talking
with others about cancer-related issues, would predict increased perceived burden, in turn,
influencing cancer survivors health outcomes. It was also hypothesized that social support
would be associated with decreased perceived burden, which would in turn predict health
outcomes (i.e., fatigue and physical symptoms). Data were collected from 136 Chinese cancer
survivors from cancer associations in the United States. The hypothesized model was tested
using path analysis in Mplus Version 6.1 program (Muthn & Muthn, 2010). The results
supported the hypotheses that social constraint and social support were associated with
perceived burden, which in turn predicted increased fatigue and greater number of physical
symptoms. Thus, perceived burden was the mediator or the mechanism through which social
constraint and social support were associated with health outcomes among Chinese cancer
survivors. Detailed findings and research/clinical implications will be presented.

CORRESPONDING AUTHOR: Kelly Yu-Hsin Liao, Ph.D., Cleveland State University, Cleveland,
OH, 44115; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2431

D030 10:00 AM-11:00 AM

SOCIAL CONSTRAINTS AND CARINGGUIDANCE PROGRAM EFFECTS ON PSYCHOLOGICAL

ADJUSTMENT IN MONTHS AFTER BREAST CANCER DIAGNOSIS

Robin M. Lally, PhD, MS, BA, RN, AOCN1, Kevin A. Kupzyk, PhD1, Steven M. Gallo, MS2, Karen
Meneses, PhD, RN, FAAN3

1
University of Nebraska Medical Center, Omaha, NE; 2University at Buffalo, Buffalo, NY;
3
University of Alabama at Birmingham, Birmingham, AL

Social constraints are negative human interactions that prevent cognitive processing of a life-
altering event. Avoidance, minimizing and preventing a person with cancer from expressing
concerns, exemplify social constraints. Breast cancer diagnosis alters ones life-view and
requires cognitive processing to reduce distress and depressive-symptoms toward achieving
psychological adjustment. CaringGuidance After Breast Cancer Diagnosis is a new, Internet-
based, self-guided, psychoeducational program developed to support psychological
adjustment in the initial months after diagnosis. Learning modules include content on
diagnosis disclosure, strategies for coping with social constraints and spouse/partner
relationship changes.

Method: 100 women diagnosed within 0 - 3 months with their first, stage 0 2 breast
cancer, randomized to self-guided use of the CaringGuidance Internet-program or usual care
for 3 months completed psychosocial measures including the Distress Thermometer (DT), CES-
D, Impact of Event Scale (IES), and Social Constraints Scale (SCS) (once for spouse/partner
(S/P) behavior and again for family/friend (F/F) behavior) at baseline and monthly. A custom
data analysis system monitored program use.

Results: Subjects mean age was 54.2 years (SD= 9.9). The majority was married; college
educated, employed, with no prior mental health diagnosis. Groups did not differ significantly
at baseline on demographic or psychosocial variables. The Intervention Group used the
program a median 4.4 hours. The SCS was highly reliable at each time point for S/P -focused
and F/F-focused responses (Cronbachs alpha >.90). Mean SCS scores ranged from 1.4 1.7.
Higher scores (significant only at month 1) were noted for F/F-focused scale responses versus
S/P. S/P and F/F-focused responses were highly correlated (r=.57 - .71). Significant positive
correlations were also found between SCS responses and DT, CES-D and IES scores. A non-
significant decrease in scores for both S/P and F/F-focused SCS responses occurred over the 3
months. However, a significant time by group interaction indicated that the Intervention
group significantly decreased in their reports of social constraints over time.
S2432 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusion: Greater social constraints correlated as predicted with greater reported distress
and depressive-symptoms among women newly diagnosed with breast cancer. Access to
CaringGuidance resulted in significantly fewer reports of constraints compared to usual
care.

CORRESPONDING AUTHOR: Robin M. Lally, PhD, MS, BA, RN, AOCN, University of Nebraska
Medical Center, Omaha, NE, 68198; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2433

D031 10:00 AM-11:00 AM

STUCK IN THE SPIN CYCLE: AVOIDANCE AND INTRUSIONS FOLLOWING BREAST CANCER
DIAGNOSIS

Margaret Bauer, M.A., C.Phil.1, Joshua F. Wiley, PhD2, Karen Weihs..., MD3, Annette L.
Stanton, Ph.D.4

1
University of California, Los Angeles, Los Angeles, CA; 2Monash Institute for Cognitive and
Clinical Neurosciences and School of Psychological Sciences at Monash University, Melbourne,
Victoria, Australia; 3University of Arizona, cancer center, Tucsn, AZ; 4UCLA, Los Angeles, CA

Cognitive processing theories of adjustment to chronic stress suggest that individuals

alternate between periods of intrusive thoughts and feelings and periods of avoidance.
Theoretically, phases of intrusions are characterized by stressor-related ideas and emotions
that are brought to mind for eventual integration. These phases of intrusions alternate with
periods of avoidance, which prevent individuals from becoming overwhelmed by intrusions.
However, most relevant studies assess avoidance as a predictor of intrusions or vice versa
rather than examining the theorized bidirectional process. The current study presents a
longitudinal empirical investigation of the reciprocal relationship between intrusions and
avoidance coping through the six to ten months following breast cancer diagnosis. Breast
cancer patients (N = 460) completed measures of cancer-related intrusions (Impact of Event
Scale) and avoidance (COPE Avoidance-oriented Coping composite) at study entry (i.e., zero -
four months after diagnosis), three-, and six-month follow-ups.

Cross-lagged panel analyses tested a series of nested models that contrasted a hypothesized
cross-lagged model demonstrating mutual influence between predictors with unidirectional
models (intrusions predicting avoidance, avoidance predicting intrusions). Wald tests revealed
that a model with bidirectional cross-lagged paths between avoidance and intrusions fit the
data better than a model with only avoidance (X2 =8.55 (2), p = .02) or intrusions (X2 =11.87
(2), p =.003) as a predictor. In the best-fitting model, intrusions at study entry predicted
greater avoidance three months later (b = .06, p < .001), and avoidance coping at study entry
predicted more intrusions three months later (b = .34, p = .01), controlling for the stability of
intrusions and avoidance as well as time since diagnosis. There was no significant relationship
between avoidance and intrusions at the three-month to the six-month assessment (ps > .05).
S2434 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

These findings provide empirical support for theories of cognitive processing following
stressful events in revealing a bidirectional relationship between avoidance and intrusions in
approximately the first six months after breast cancer diagnosis, which appears to diminish in
the next three months. Whether this process can be accelerated via psychosocial intervention
is worthy of investigation.

CORRESPONDING AUTHOR: Margaret Bauer, M.A., C.Phil., University of California, Los

Angeles, Los Angeles, CA, 90095; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2435

D032 10:00 AM-11:00 AM

SUN SAFETY, SKIN CANCER, AND SOCIODEMOGRAPHIC CHARACTERISTICS AMONG

APPALACHIAN ADULTS COMPARED TO NON-APPALACHIAN ADULTS

Minal Patel, PhD, MPH1, Katrina J. Serrano, PhD2, Elise L. Rice, PhD3, Chan Thai, PhD, MPH4,
Kelly D. Blake, ScD5, Robin Vanderpool, BS, MPH, DrPH6

1
National Cancer Institute, Rockville, MD; 2National Cancer Institute/Behavioral Research
Program, Bethesda, MD; 3National Cancer Institute, Bethesda, MD; 4Santa Clara University,
Santa Clara, CA; 5Behavioral Research Program, Division of Cancer Control and Population
Sciences, National Cancer Institute, National Institutes of Health, Bethesda, MD; 6University of
Kentucky College of Public Health, Lexington, KY

Background: Individuals living in the US Appalachian region experience greater cancer

disparities compared to their non-Appalachian counterparts. Although melanoma incidence is
comparable among Appalachians and non-Appalachians, melanoma mortality rates are 23%
higher in Appalachia compared to the national average. Protective behaviors related to
limiting UV exposure, such as sunscreen use and tanning bed avoidance, can help reduce skin
cancer mortality among Appalachians. Previous research examined sunscreen use among
Appalachian Ohio residents, but little is known about the entire 13 state region.

Objective: This study examines skin cancer and associated behaviors among Appalachian
adults vs. non-Appalachian adults, along with related sociodemographic characteristics in a
nationally representative US sample.

Methods: Data from four waves of the Health Information National Trends Survey, a
nationally representative population-based survey collected from 2011-2014, were combined
(N = 14,451). Descriptive analyses examined sunscreen use, tanning bed use, and history of
melanoma and non-melanoma skin cancers among Appalachian respondents vs. non-
Appalachians. Weighted multivariable analyses were conducted to examine sunscreen use,
controlling for sociodemographic characteristics.
S2436 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Results: Approximately 8% of the sample resided in Appalachia (n=1,015). Melanoma rates

among Appalachians were low (0.5%). Non-melanoma skin cancer (2.6%) also did not differ
statistically between Appalachian and non-Appalachian adults. Only 1.7% of Appalachian
respondents reported using sunscreen often or always when going outside for more than 1
hour on a warm, sunny day, compared to 25.2% among non-Appalachians. Only 0.5% of
Appalachian adults reported tanning bed use and residents appeared to be no more or less
likely to use a tanning bed than non-Appalachians in multivariable models. In a multivariable
logistic regression, individuals residing in Appalachia reported a lower odds of sunscreen use
often or always compared to non-Appalachian residents (OR=0.73, p < 0.05), controlling for
sociodemographics and general health status. Women were more likely than men to report
sunscreen use often or always (OR=2.18, p < 0.05), as were individuals with a college degree
or higher (OR=1.99, p < 0.05), compared with individuals with less than a high school degree.

Conclusions: Consistent with national surveillance data, Appalachian respondents were no

more or less likely to report being diagnosed with melanoma or non-melanoma skin cancer,
but were less likely to report sunscreen use than non-Appalachian residents. No differences in
tanning bed use were seen, indicating that interventions should focus on increasing regular
sunscreen use among Appalachian residents. Communication efforts in both occupational and
recreational settings are needed to increase sun safety behaviors.

CORRESPONDING AUTHOR: Minal Patel, PhD, MPH, National Cancer Institute, Rockville, MD,
20850; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2437

D033 10:00 AM-11:00 AM

SURVIVORSHIP CARE PLANNING AND ADHERENCE TO GUIDELINE-RECOMMENDED CANCER

SCREENING

L. Aubree Shay, PhD, MSSW1, Shayda I. Dioun, MPH Student2

1
UT School of Public Health, San Antonio Regional Campus, San Antonio, TX; 2University of
Texas School of Public Health, San Antonio, TX

Introduction: With improvements to cancer screening and treatment, the number of cancer
survivors living in the United States is increasing. Cancer survivors are at high risk of chronic
health conditions and additional cancer diagnoses. To aid in long-term follow-up care, the
National Cancer Institute recommends that all cancer survivors receive a written treatment
summary and plan for follow-up care, jointly called a survivorship care plan. However,
survivorship care planning is underutilized and there little evidence yet linking its use to
positive patient outcomes.

Purpose: Thus, for this study we used data from the 2014 Behavioral Risk Factor Surveillance
System (BRFSS) survey to look at the association between survivorship care planning and
adherence to guideline-recommended cancer screening.

Methods: We used 3 multivariable logistic regression models to test whether receipt of a

written treatment summary and receipt of instructions for follow-up care were associated
with being up-to-date on mammography, cervical cancer screening, and colorectal cancer
screening while controlling for patient demographic and cancer history related variables.

Results: In total, 4264 off-treatment cancer survivors completed the 2014 BRFSS cancer
survivorship module and were included in our analysis. Most survivors were female (63%),
white (93%) and 65 years old or older at the time of the survey (63%). Breast cancer was the
most common cancer diagnosis (22%) followed by melanoma (18%), and prostate cancer
(11%). In total, 40% of survivors reported receiving a treatment summary and 71% reported
receiving instructions for follow-up care. Additionally, 79% of eligible survivors were up-to-
date on mammography, 83% on cervical cancer screening, and 78% on colorectal cancer
screening. In our adjusted models we found that receipt of a plan for follow-up care was
significantly associated with adherence to guideline-recommended mammography (OR=2.2,
95% CI (1.6-2.9), cervical cancer screening (OR=1.9 (1.1-3.1)) and colorectal cancer screening
(OR=1.7 (1.4-2.2). Receipt of a written treatment summary was not associated any of the
cancer screening behaviors.
S2438 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusion: Timely cancer screening can be life-saving and is particularly important among
cancers survivors who have higher odds of subsequent cancer diagnoses. Our study shows
initial evidence for the impact of survivorship care planning on adherence to guideline-
recommended cancer screening.

CORRESPONDING AUTHOR: L. Aubree Shay, PhD, MSSW, UT School of Public Health, San
Antonio Regional Campus, San Antonio, TX, 78229; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2439

D034 10:00 AM-11:00 AM

SYMPTOM CLUSTERS IN METASTATIC BREAST CANCER PATIENTS

Danielle Tometich, M.S.1, Catherine E. Mosher, Ph.D.1, Kevin L. Rand, Ph.D.2, Adam T. Hirsh,
PhD1

1
Indiana University-Purdue University Indianapolis, Indianapolis, IN; 2IUPUI, Indianapolis, IN

Co-occurring symptoms, or symptom clusters, have a negative, compounding effect on

functional status and quality of life in cancer patients. Symptom clusters are also significant
because they point to common mechanismsboth biological and psychologicalthat may
underlie various symptoms. Psychoneurological symptom clusters consisting of mood
disturbances, cognitive dysfunction, fatigue, sleep disturbance, and pain have been found in
patients with early-stage breast cancer and other cancer types. Symptom cluster research in
breast cancer patients has largely focused on patients with early-stage disease or survivors in
remission, and little is known about symptom clusters in metastatic breast cancer patients
despite their increasing longevity and high symptom burden. In order to address this gap, this
study aimed to identify physical and psychological symptom clusters in metastatic breast
cancer patients. Eighty women with metastatic breast cancer (91% Caucasian, average age =
55 years, SD = 11 years) were recruited from the Indiana University Simon Cancer Center to
participate in this cross-sectional telephone interview study. The interview included valid
measures of the severity of 10 common symptoms (i.e., pain, fatigue, anxiety, depressive
symptoms, neuropathy, edema of arms or legs, nausea, hot flashes, sleep disturbance, and
cognitive concerns). An exploratory cluster analysis was performed on measures of all 10
symptoms. Anxiety, depressive symptoms, fatigue, sleep disturbance, and cognitive concerns
were found to cluster. A separate cluster consisted of pain, neuropathy, and nausea. Hot
flashes and swelling did not cluster. The first cluster is consistent with a psychoneurological
symptom cluster, which has been found in patients with non-metastatic breast cancer and
other cancer types. The co-occurrence of pain and neuropathy may reflect neuropathic pain,
which, along with nausea, is a common side effect of chemotherapy. Further research is
needed to replicate these clusters and elucidate common mechanisms underlying co-
occurring symptoms to inform intervention development.

CORRESPONDING AUTHOR: Danielle Tometich, M.S., Indiana University-Purdue University

Indianapolis, Indianapolis, IN, 46202; [email protected]
S2440 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D035 10:00 AM-11:00 AM

THE AUDACITY OF ENGAGEMENT: HEARING DIRECTLY FROM YOUNG ADULTS WITH CANCER
ON THE STATE OF CANCER SURVIVORSHIP RESEARCH

Catherine Benedict, PhD1, Dave Fuehrer, N/A2, David Victorson, PhD3, Brad Love, PhD4

1
Hofstra Northwell School of Medicine, Manhassat, NY; 2Stupid Cancer, New York, NY;
3
Robert H. Lurie Comprehensive Cancer Center, Northwestern University, Chicago, IL;
4
University of Texas at Austin, Austin, TX

Background: Adolescent and young adult (AYA) cancer survivors are an understudied and
underserved patient population. This study aimed to better understand AYAs perceptions of
cancer survivorship research in order to identify barriers and inform research approach and
methodology to ultimately improve evidence-based supportive care.

Methods: Individual interviews and small focus groups were conducted with AYA cancer
survivors (N=17) attending an oncology conference and social networking event. Analyses
were guided by grounded theory using an inductive data-driven approach to thematic content
analysis.

Results: Participants averaged 33 years old (SD=5.5), were 10 years post-diagnosis (SD=2.3),
and primarily female (79%); most common diagnoses were leukemia/lymphoma (37%) and
sarcoma (11%). Major themes included: perceptions of cancer research (2 subthemes) and
recommendations for improving research methodology and approach (3 subthemes). All
participants reported uncertainty as to what cancer survivorship research encompassed, but
drew from their clinical experiences and prior participation in clinical trials to inform their
opinions. Perceptions of research were both positive and negative. Positive feelings included
optimism about the potential to improve the cancer survivorship experience and wanting to
give back to the AYA community through research participation. Negative perceptions
included seeing research as inaccessible, overwhelming, and frustrating; participation as
inconvenient and burdensome; and viewing researchers with skepticism and mistrust.
Recommendations centered on: 1) making study materials age-appropriate and patient-
focused; 2) increasing AYAs control over participation requirements and the flexibility and
convenience of procedures; and 3) creating a back-and-forth dialogue between researchers
and the AYA community to improve study-specific communication (dissemination of findings),
stay up-to-date on medical information specific to ones disease, and to have a voice in
determining research priorities.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2441

Conclusion: Specific strategies are needed to address AYAs negative perceptions and
perceived barriers to research participation. Future work should determine how to best
improve communication with the AYA community including dissemination of research
findings, providing opportunities to participate in age-appropriate research advisory roles, and
building trust.

CORRESPONDING AUTHOR: Catherine Benedict, PhD, Hofstra Northwell School of Medicine,

Manhassat, NY, 11030; [email protected]
S2442 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D036 10:00 AM-11:00 AM

THE BEHAVIORAL MEASUREMENT AND INTERVENTION SHARED RESOURCE (BMISR)

SUPPORTS RESEARCHERS CONDUCTING BEHAVIORAL STUDIES

Loescher J. Lois, PhD, FAAN1, Angela Yung, RDN1, Cynthia Thomson, PhD, RD2

1
The University of Arizona, Tucson, AZ; 2University of Arizona, Tucson, AZ

Background: The University of Arizona Cancer Center (UACC) BMISR provides support and
services to UACC members and other University and national researchers studying human
lifestyle behavior (diet, physical activity, sun safety, sexual practices, tobacco use) and quality
of life (sleep, symptom management) related to cancer prevention and control. The BMISR
supports the design and implementation of highly innovative behavioral measurement and
interventions research within diverse populations. Specifically, the BMISR provides
consultation for theory and research design, offers a repository of data collection tools and
measurement equipment, facilitates development and use of ehealth and mHealth
measurement and intervention strategies, and provides training for data collection,
measurement, and multimodal lifestyle behavior coaching.

Purpose: To describe the structure of the BMISR and its effectiveness as a mechanism to
promote behavioral health research in multiple populations, organizations and settings.

Methods: We evaluated BMISR recruitment techniques, usage from 2014 to 2016 by UACC
members and nonmembers with and without peer-reviewed support, types of activities
requested by researchers, implementation in diverse populations, and cost.

Results: The BMISR has successfully recruited users using strategies of local and national
person-person networking, formal presentations to UACC and University faculty, and
publication in newsletters. Innovative partnerships with a smokers quitline program and
mHealth development program has increased usage. The BMISR provided services to over 35
researchers nationally, including those in 9 comprehensive cancer centers. Of those
researchers, 35% were UACC members with peer-reviewed support and 65% were
nonmembers. Researchers requested support for questionnaire development (n=8
researchers) dietary measurement (n=21), health coaching (n=2), mHealth studies (n=4).
These researchers were conducting studies in populations of adolescents, American Indians,
Hispanics, healthcare providers, and cancer survivors. Fees for BMISR services such as survey
analysis, dietary recalls, coaching calls, and accelerometer use were at least 50% less than
charges from outside vendors.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2443

Conclusions: The BMISR is an effective and economical means of supporting researchers who
conduct behavioral research in diverse populations and settings.

CORRESPONDING AUTHOR: Loescher J. Lois, PhD, FAAN, The University of Arizona, Tucson, AZ,
85721-0203; [email protected]
S2444 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D037 10:00 AM-11:00 AM

THE BOTTOM LINE: WHY MENTAL HEALTH PROVIDERS NEED TO PARTICIPATE IN THE CARE OF
NON-ELDERLY ADULTS WITH CANCER

Diana D. Jeffery, PhD1, C. Allison Russo, PhD2, Laura Hopkins, M.P.A.2, Harry B. Burke, MD,
PhD3

1
Department of Defense, Defense Health Agency, Falls Church, VA; 2Kennell and Associates,
Inc., Falls Church, VA; 3Department of Defense, USUHS, Bethesda, MD

Background: This study examines the relationship of mental health comorbidities and cost in
non-elderly cancer patients.
Methods: A cross-sectional analysis of Military Health System data was conducted on all
beneficiaries (9.5 million) to select patients ages 18-64 with a primary diagnosis of cancer for
fiscal years 2007 2014. General linear models were used to examine the relationship
between mental health comorbidities and annual costs, the number of ambulatory visits, the
number of hospitalizations, and the number of admission days. Predictors included
demographic characteristics, type and location of enrollment, types of treatment modalities,
number of chronic disease comorbidities, presence and type of mental health comorbidities,
and tobacco use.
Results: On average, 92,318 patients had a primary cancer diagnosis each year of observation.
The Department of Defense paid an average of $20,357 (adjusted to U.S. 2014 dollars) per
patient per fiscal year. The mean number of annual visits per patient was 26.44; the mean
number of annual hospitalizations was 1.62; and the mean number of admission days was
16.13. The majority of patients were ages 45 - 64, female, and military retired or a family
member of a retiree. After controlling for other predictors, concurrent diagnosis of
depression, anxiety, or adjustment disorder (DAA) was significantly related to annual cost (p <
0.0001), with only the receipt of chemotherapy being a stronger predictor. A concurrent
diagnosis of DAA, or of a serious, persistent mental illness, were significant predictors of
annual number of ambulatory visits (p < 0.0001).DAA comorbidity was highly related to
annual number of hospital readmissions (p < 0.0001), as well as annual number of admission
days (p < 0.0001). A comorbidity of alcohol and drug use disorder was statistically significant
(p < 0.0001) in all regression models, though not as strong a predictor as DAA. Additionally,
tobacco use was significantly related to annual cost (p < 0.0001), ambulatory visits (p <
0.0001), and hospital readmissions (p < 0.0001), but not to the annual number of admission
days.
Conclusions: This study found that cancer patients had high levels of depression, anxiety, and
adjustment disorder and that these patients generated increased annual costs, including costs
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2445

related to ambulatory visits, hospital readmissions, and admission days. These findings
suggest that the increased involvement of mental and behavioral health professionals may be
beneficial in improving patient care and reducing costs.

CORRESPONDING AUTHOR: Diana D. Jeffery, PhD, Department of Defense, Defense Health

Agency, Falls Church, VA, 22042; [email protected]
S2446 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D038 10:00 AM-11:00 AM

THE BREAST CANCER ONLINE REHABILITATION PROGRAM FEASIBILITY OF A MUNICIPALITY-

BASED UPPER-BODY REHABILITATION PROGRAM

Bolette Skjdt. Rafn, PT. MSc, PhD student1, Carina Nees, PT2, Jette Vibe-Petersen, MD3, Julie
Midtgaard, PhD4, Pat Rafn, PhD5, Kristin Campbell, PhD6

1
University of British Columbia, Vancouver, BC, Canada; 2Copenhagen Center for Cancer and
Health, Copenhagen, Hovedstaden, Denmark; 3Copenhagen Centre for Cancer and Health,
Copenhagen, Hovedstaden, Denmark; 4Section of Social Medicine, Department for Public
Health, University of Copenhagen and The University Hospitals Centre for Health Research,
Copenhagen University Hospital, Copenhagen, Hovedstaden, Denmark; 5St. Pauls Hospital
and University of British Columbia, Vancouver, BC, Canada; 6University of British Columbia,
Department of Physical Therapy, Vancouver, BC, Canada

Background Following treatment for breast cancer, a considerable amount of women

experience persistent upper-body issues. Innovative interventions for early post-surgery
rehabilitation are needed. An online resource offers the opportunity to deliver homogenous,
high-quality breast cancer rehabilitation that may reduce the prevalence or severity of
persistent upper-body issues. We developed the Breast Cancer Online Rehabilitation
(BRECOR) program through an iterative process (usability testing, discussion panels, and focus
group interviews with women with breast cancer and physiotherapists). The program consists
of a clinical assessment tool for physiotherapists to prescribe individualized exercises, and a
website and education pamphlet to support the home-based rehabilitation prescriptions.

Objective To test the feasibility of integrating the BRECOR program into clinical practice to
support upper-body rehabilitation for women after breast cancer surgery.

Design and setting Single-group, feasibility study in 11 municipalities in the Capital Region of
Denmark.

Subjects Women age 18-80 years with 1) a surgery for breast cancer within the past 8 weeks,
2) access and ability to use internet, and 3) ability to read and understand Danish.

Methods All participants underwent an in-person assessment with a physiotherapist using the
BRECOR clinical assessment tool, which then formed the therapists decision-making for the
prescription of specific exercises for a 12-week home-based upper-body rehabilitation
program tailored to each participants needs. Each participant received in-person instruction
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2447

by the physiotherapist for the specific exercises at that visit. To support this program,
participants were asked to use the BRECOR education pamphlet and the BRECOR website to
complete 5 to 7 exercise sessions weekly. At 12-weeks, the in-person assessment with a
physiotherapist was repeated and used as a measure of retention. Primary outcomes were
recruitment and retention rate, adherence, participant satisfaction, capacity/resources and
secondly, self-reported shoulder function (QuickDASH).

Results From 49 women, who were eligible, 39 (79.6%) participated and 29 (74.6%)
completed the 12-week follow-up assessment. 72% (21 of 29) adhered to the BRECOR
program. 90% (26 of 29) reported to have benefitted much or very much from the
BRECOR program. Capacity (total time spent by the physiotherapist in delivering the program)
was 1.5 hour 27 min at baseline and 5413 min at follow-up. Self-reported shoulder function
improved from baseline (mean 18.6012.5) to 12-weeks (mean 13.9716.1) but was not
significant (p=0.08).

Conclusion The BRECOR program can reach and retain the majority of women in the early
phase after surgery for breast cancer. Participants are satisfied with and adhered to the
program.

CORRESPONDING AUTHOR: Bolette Skjdt. Rafn, PT. MSc, PhD student, University of British
Columbia, Vancouver, BC, V6K 3J7; [email protected]
S2448 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D039 10:00 AM-11:00 AM

THE DEVELOPMENT, IMPLEMENTATION, AND EVALUATION OF A COMMUNICATION SKILLS

TRAINING PROGRAM FOR ONCOLOGY NURSES

Smita Banerjee, Ph.D.1, Ruth Manna, MPH2, Nessa M. Coyle, PhD, RN1, Stacey Penn, BSN, RN,
OCN3, Tess Gallegos, MPH4, Talia Zaider, PhD1, Carol Krueger, LCSW-R5, Philip Bialer, MD1,
Carma Bylund, PhD6, Patricia A. Parker, PhD1

1
Memorial Sloan Kettering Cancer Center, New York, NY; 2Memorial Sloan Kettering Cancer
Center, new york, NY; 3Memorial Sloan Kettering Cancer Center, Riverdale, NY; 4Memorial
Sloan Kettering Cancer Center, Montclair, NJ; 5Memorial Sloan-Kettering Cancer Center, New
York, NY; 6Hamad Medical Corporation, Houston, TX

Many nurses express difficulty in communicating with their patients, especially in oncology
settings where there are numerous challenges and high stakes decisions during the course of
diagnosis and treatment. Providing specific training in communication skills is one way to
enhance the communication between nurses and their patients. We developed and
implemented a communication skills training program for nurses, consisting of three teaching
modules: Responding Empathically to Patients, Discussing Death, Dying, and End-of-Life Goals
of Care, and Responding to Challenging Interactions with Families. Training included didactic
and experiential small group role plays. This paper presents results on program evaluation,
self efficacy and behavioral demonstration of learned communication skills. Three hundred
and forty two in-patient oncology nurses participated in a one-day communication skills
training program, and completed course evaluations, self-reports, and pre- and post-
Standardized Patient Assessments. Overall nurse participants rated the training favorably.
Specifically, more than 90% of nurse participants indicated that they agreed or strongly
agreed with 5 of the 6 evaluation items (with one item receiving endorsement by more than
80% but less than 90% of participants). As well, a majority of nurse participants (>80%) rated
each individual module component as aiding in learning (as indicated by ratings of somewhat
aided my learning to aided my learning a lot). Participants reported significant gains in self
efficacy in their ability to communicate with patients in various contexts. Overall, nurses self-
efficacy significantly improved [t(1016) = 31.17, p < .001] from pre- (M = 3.31, SD = .88) to
post-training (M = 4.05, SD = .65). In addition, there was a significant increase in overall skill
use from pre- to post-training. The biggest gain was observed in empathic skills. In particular,
4 out of 5 empathic skills (encourage expression of feelings, acknowledge, normalize, and
praise patient efforts) significantly increased from pre- to post-training. Additionally,
significant increase was observed in 1 out of 5 questioning skills (clarify) and 1 out of 4
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2449

information organization skills (summarize). This work demonstrates that implementation of a

nurse communication skills training program at a major cancer center is feasible, acceptable,
and has a significant impact on participants' self-efficacy and uptake of communication skills.
A key practical implication of this work is that communication skills training for oncology
nurses (as well as for other health providers) is an effective way of improving communication
between nurse and patient/families, with putative improvements in patient-related
outcomes. Future research should examine the impact of communication skills training on
relevant patient outcomes.

CORRESPONDING AUTHOR: Smita Banerjee, Ph.D., Memorial Sloan Kettering Cancer Center,
New York, NY, 10022; [email protected]
S2450 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D040 10:00 AM-11:00 AM

THE EFFECT OF ACCEPTANCE AND COMMITMENT THERAPY ON COGNITIVE FUSION, DISTRESS,

AND MOOD DISORDERS: A PILOT STUDY

Symone McKinnon, Bachelors of Arts in Psychology1, Maya Santoro D'Eon, Ph.D.2, April May,
Masters of Arts in Psychology3, Terry A. Cronan, Ph.D.4

1
San Diego State University, La Mesa, CA; 2University of California, San Diego, San Diego, CA;
3
University of California San Diego/San Diego State University Joint Doctoral Program, San
Diego, CA; 4San Diego State University, San Diego, CA

Acceptance and Commitment Therapy (ACT) is designed to promote psychological flexibility

through various cognitive and behavioral processes, including cognitive defusionthe act of
detaching from ones thoughts and emotions. Defusion promotes greater flexibility in coping
with difficult thoughts and emotions, which may help newly diagnosed breast cancer patients
who experience psychological distress and mood disorders. A pilot study was conducted to
examine whether ACT would reduce cognitive fusion, psychological distress, depression, and
anxiety in a sample of 22 women diagnosed with breast cancer (stages 0 to 3) within the past
year, who reported experiencing significant levels of distress. Participants were randomly
assigned to an 8-week ACT intervention group or an attention-control group and were
administered the Brief Symptom Inventory-18, the Cognitive Fusion Questionnaire-13, and the
Hospital Anxiety and Depression Scale pre- and post-intervention. The small sample size
limited power to detect interactions between the intervention groups and time points; thus,
effect sizes were calculated to compare the two groups on change for each of the variables of
interest. The results showed a reduction in cognitive fusion in both groups; however, the
effect size was moderate for the ACT group (d = 0.614) and small for the attention-control
group (d = 0.246). The attention-control group reported a greater reduction in psychological
distress (d = 1.615) than the ACT group (d = 1.475), with both groups having large effect sizes.
A large effect was detected in the ACT group for reduction of depressive symptoms (d =
1.740), with only a moderate effect size for those in the attention-control group (d = 0.773).
Lastly, moderate effect sizes were obtained for both groups in the reduction of anxiety
symptoms (ACT: d = 0.765; Attention-control: d = 0.644). A large randomized controlled trial is
warranted to determine whether the promising effects of ACT found in this pilot study will be
sustained among newly diagnosed breast cancer patients.

CORRESPONDING AUTHOR: Symone McKinnon, Bachelors of Arts in Psychology, San Diego

State University, La Mesa, CA, 91942; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2451

D041 10:00 AM-11:00 AM

THE IMPACT OF CANCER ON THE SEXUAL FUNCTION OF ADOLESCENTS AND YOUNG ADULTS:
A SYSTEMATIC REVIEW

Amelia Stanton, BA, Ariel Handy, BA, Cindy Meston, PhD

University of Texas at Austin, Austin, TX

Background: Adolescence and young adulthood (ages 15-39) is a critical period for sexual
development, and a diagnosis of cancer during this time may have negative implications for
sexual health (Nathan, Hayes-Lattin, Sisler, & Hudson, 2011). To date, no review has examined
the effect of cancer on the sexual function of either men or women within this age range.

Objective: To identify, with supporting evidence, the effect cancer and its various treatments
on the sexual function of adolescent and young adult men and women.

Data sources: A systematic review of articles published in English using PsycINFO, PubMed,
and CINAHL was conducted. Additional studies were identified through reference lists. A total
of 1,465 articles were retrieved after duplicates were removed. Search terms included:
cancer, adolescents, young adults, sexual function, sexual satisfaction, sexual desire, sexual
arousal, sexual pain, orgasm.

Study selection: Studies were eligible for inclusion if they a) were published in peer-reviewed
journals; b) presented original findings; c) used validated measures; and d) included men
and/or women who were diagnosed with cancer between the ages of 15 and 39.

Data extraction: Quality assessments and data extraction were performed independently by
the first two authors using The Strengthening the Reporting of Observational Studies in
Epidemiology Statement (STROBE; von Elm et al., 2008).

Results: In the 14 final studies, there was a total of 2,265 participants (849 females), with
sample sizes ranging from 9 to 793. The results showed that a diagnosis of cancer during the
adolescent and young adult period affects the following domains of sexual function: desire,
arousal, lubrication/erection, orgasm/ejaculation, pain, enjoyment/satisfaction, and
frequency of sexual activity. For both women and men, a diagnosis of cancer during this time
significantly impacted sexual satisfaction. In women specifically, the domain of sexual function
that was most impacted was desire, whereas for men erectile function was most
compromised.
S2452 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusion: Adolescents and young adults have been identified as particularly vulnerable to
the negative impact of cancer and its treatments. Increased awareness of the potential
consequences of cancer on sexual function in this population is essential to developing
targeted interventions.

CORRESPONDING AUTHOR: Amelia Stanton, BA, University of Texas at Austin, Austin, TX,
78712; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2453

D042 10:00 AM-11:00 AM

THE IMPACT OF MATCHING TO PSYCHOTHERAPY TREATMENT PREFERENCE ON THERAPEUTIC

ALLIANCE IN PATIENTS WITH ADVANCED CANCER

Allison Marziliano, MA1, Allison Applebaum, PhD2, Anne Moyer, Ph.D.3, Barry Rosenfeld,
PhD4, William Breitbart, MD2, Hayley Pessin

1
Memorial Sloan-Kettering Cancer Center, New York, NY; 2Memorial Sloan Kettering Cancer
Center, New York, NY; 3Stony Brook University, Stony Brook, NY; 4Fordham University, Bronx,
NY

Background: A common challenge in randomized controlled trials (RCTs) is that participants

may inherently prefer one treatment over another, which may impact attrition, outcomes,
and therapeutic alliance. The purpose of these analyses is to examine whether matching
patients with advanced cancer to their preferred psychotherapy treatment will positively
impact their alliance with their therapist during treatment.

Method: Data were drawn from an RCT comparing the efficacy of Individual Meaning-
Centered Psychotherapy (IMCP), Individual Supportive Psychotherapy (ISP), and Enhanced
Usual Care (EUC) in a sample of patients with advanced cancer. Patients preference for
treatment was assessed with a categorical question, if you were able to choose, which arm
would you prefer?. Therapeutic alliance was measured before the fourth session by a
member of the study team other than the therapist using the Working Alliance Inventory-
Short Form.

Results: For the purposes of these analyses, only participants (N = 168) who prefer and are
randomized to one of the two psychotherapy treatments (either IMCP or ISP) will be included.
Eighty-six (51%) participants were matched to their preferred type of psychotherapy
treatment. Of those 86, 49 participants preferred and received IMCP, and 37 participants
preferred and received ISP. Eighty-two (49%) participants were categorized as having been
mismatched to their preferred type of psychotherapy treatment. Of those 82 participants, 46
preferred IMCP but received ISP, and 36 preferred ISP but received IMCP. Results
demonstrated that patients who matched to their preferred psychotherapy treatment had a
significantly stronger therapeutic alliance than those who did not, (t = -3.10, p =.003). Further
breaking down the mismatched group, those who preferred IMCP and received ISP, compared
to other mismatched patients, demonstrated a significantly weaker therapeutic alliance, (t =
2.93, p = .005). By contrast, there was no significant difference in therapeutic alliance
between those who preferred ISP but received IMCP and other mismatched patients, (t =
S2454 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

0.76, p = .451).

Conclusions: Matching to treatment preference appears to impact the therapeutic alliance

between the therapist and patient, as reported by the patient, in a RCT comparing
psychotherapeutic treatments for patients with advanced cancer. Mismatch seems to
differentially impact therapeutic alliance, and might be less relevant for people who prefer ISP
than those who prefer IMCP.

CORRESPONDING AUTHOR: Allison Marziliano, MA, Memorial Sloan-Kettering Cancer Center,

New York, NY, 10022; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2455

D043 10:00 AM-11:00 AM

THE PSYCHO-PHYSIOLOGICAL IMPACTS OF SELF-COMPASSION TRAINING FOR CANCER

SURVIVORS: A RANDOMIZED-CONTROLLED PILOT STUDY

Adrian H.Y. Wan, MSW1, Tiffany Hon, BSc(Psy)2, Rainbow T. H. Ho, PhD1

1
Centre on Behavioral Health, Hong Kong, N/A, Hong Kong; 2The University of Hong Kong,
Hong Kong, N/A, Hong Kong

Background: Being self-compassionate is to treat yourself with kindness at times of adversity.

Although self-compassion training was found to be associated with domains of positive
mental health outcomes among non-clinical population, few studies examined the
effectiveness of compassion training among people with chronic medical conditions. Utilizing
both subjective self-report and the biomarkers Heart Rate Variability, this pilot study explored
the benefits of mindfulness-based self-compassion training among cancer patients.

Methods: This pioneer study adopted a 2-arm randomized-controlled study design. A total of
37 Chinese cancer survivors were surveyed once at baseline and at post-training. They were
surveyed on a self-report inventory of self-compassion, symptoms of anxiety and depression,
positive and negative affect, and non-reactivity; they were also invited to participate in a
stress task to capture clinical data heart rate variability.

Results: On subjective measures, co-varying age,participants of the Self-Compassion Training

showed significant Time x Group effects on self-compassion, positive affect, non-reactivity,
and depressive symptoms. However, no significant Time x Group effect was reported for
negative affect. In terms of mood regulation, participants also showed significant
improvements in heart rate variability as reflected on the measures of RMSSD, HF, and LF
during a stress test at post-intervention.

Conclusions: Results of the pilot study revealed the potential effectiveness of cultivating self-
compassion among cancer survivors through mindfulness practice. Participants showed
improvements in self-compassion, non-reactivity, positive affect, depressive symptoms after
the program, and they reported better emotional resilience during a stress task. Further study
with a discrete sample of larger size is recommended to yield further evidence on
effectiveness of self-compassion training.

CORRESPONDING AUTHOR: Adrian H.Y. Wan, MSW, Centre on Behavioral Health, Hong Kong,
N/A, 852; [email protected]
S2456 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D044 10:00 AM-11:00 AM

THE ROLE OF SELF-EFFICACY IN INTENTIONS TO RECEIVE HPV VACCINATION AMONG FEMALE

COLLEGE STUDENTS

Shannon M. Christy, PhD1, Joseph Winger, M.S.2, Catherine E. Mosher, Ph.D.3

1
Moffitt Cancer Center, Tampa, FL; 2Indiana University-Purdue University Indianapolis,
Durham, NC; 3Indiana University-Purdue University Indianapolis, Indianapolis, IN

College women are at high risk for contracting human papillomavirus (HPV), and yet, up to
70% of young women do not complete the three-shot HPV vaccination series. Drawing upon
the Health Belief Model (HBM), prior studies have identified cognitions (e.g., perceived risk of
disease, perceived barriers) associated with young womens intentions to receive the HPV
vaccine. Self-efficacy, a common target for intervention, may be a factor underlying the
relationship between HBM variables and young womens vaccine intentions. The current
study tested the hypothesis that self-efficacy for HPV vaccination would mediate relationships
between five HBM variables (i.e., perceived risk, perceived benefits, perceived barriers,
perceived severity, cues to action) and intentions to receive the vaccine among unvaccinated
college women. Data were collected via an anonymous, web-based survey from 115
unvaccinated women aged 18-26 who were attending a Midwestern university. Five separate
mediation analyses were conducted to examine whether self-efficacy mediated relationships
between the five HBM variables and intentions while controlling for the other four
unexamined HBM variables and prior knowledge of the HPV vaccine. As hypothesized, there
were significant indirect effects of perceived risk on intentions through self-efficacy (indirect
effect=.16, SE=.09, 95% CI=.02-.37) and perceived barriers on intentions through self-efficacy
(indirect effect=-.16, SE=.06, 95% CI=-.31--.06). Specifically, perceived risk of developing an
HPV-related condition was positively associated with self-efficacy for HPV vaccination which,
in turn, was positively associated with vaccine intentions. In addition, perceived barriers to
HPV vaccination were negatively associated with self-efficacy for vaccination which, in turn,
was positively associated with vaccine intentions. Self-efficacy did not mediate the effects of
the other HBM variables on intentions. Future intervention research might examine whether
increasing self-efficacy for HPV vaccination by providing risk information and reducing
perceived barriers promotes vaccination among college women.

CORRESPONDING AUTHOR: Shannon M. Christy, PhD, Moffitt Cancer Center, Tampa, FL,
33612; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2457

D045 10:00 AM-11:00 AM

TRUENTH LIFESTYLE MANAGEMENT: FEASIBILITY AND BENEFITS OF PHYSICAL ACTIVITY AND

YOGA PROGRAMS FOR PROSTATE CANCER SURVIVORS

Nicole Culos-Reed, PhD1, Ashley Zahavich, MSc, RN2, Kathryn Wytsma-Fisher, MKin, CSEP
CEP3, Michael Dew, MSc, CSEP CEP4, Bronte Viznaugh, N/A4, Christine MacSween, BA5, Daniel
Santa Mina, PhD6

1
University of Calgary, Calgary, AB, Canada; 2Kinesic Sport Lab, Halifax, NS, Canada; 3Faculty of
Kinesiology, calgary, AB, Canada; 4Faculty of Kinesiology, University of Calgary, calgary, AB,
Canada; 5Faculty of Kinesiology, University of Calgary, Calgary, AB, Canada; 6Faculty of
Kinesiology and Physical Education, University of Toronto, Toronto, ON, Canada

TrueNTH Lifestyle Management (LM) aims to improve the quality of life for prostate cancer
survivors through community-based physical activity and yoga programs. The LM physical
activity program is a 12-week, evidence-based program with weekly group fitness and yoga
classes, as well as education on key behaviour change skills that will help facilitate physical
activity adoption and adherence. Instructors with specialized training followed a manual of
operations that included (a) class plans that could be modified to the unique needs of
participants based on baseline assessments, and (b) education topics to be delivered on a bi-
weekly basis during the weekly classes. Assessments at baseline and 12-weeks included
quality of life (Functional Assessment of Cancer Therapy-Prostate and PORPUS), weekly
activity levels (Godin Leisure Time Exercise Questionnaire), body composition measures, and
fitness assessments (functional aerobic capacity, flexibility, balance, muscular strength and
endurance). Participants (n=69) mean age was 66 years, with 18% on active surveillance, 15%
on active treatment, and 68% completed treatment. Treatments included surgery (60%),
radiation (24%), ADT (26%), and chemotherapy (3%). Significant pre to post intervention
improvements were seen in quality of life (FACT-P, p=.046; PORPUS, p=.03), total
moderate/strenuous exercise (GLTEQ, p=.018), body composition (waist-to-hip ratio, p=.012),
functional aerobic capacity (6-minute walk test, p=.003), flexibility (sit and reach, p=.002),
dynamic balance (8-foot timed-up-and-go, p=.000), and muscular endurance (push-ups,
p=.042; 30-second sit-to-stand, p=.01). Previously sedentary participants also reached ACSM
physical activity guidelines at 12-weeks (GLTEQ, p=.012). Ongoing program implementation in
additional sites (n=8) includes quality improvement cycles to update programming based on
feedback from both participants and instructors, as well as on current evidence. The in-class
programming is supplemented by a dedicated website that provides program participants as
well as those men who may not have access to the programs (i.e., in rural locations) with the
(a) educational resources for physical activity behaviour change as well as key nutrition and
S2458 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

stress-reduction topics, and (b) connection with experts online to aid in physical activity
screening and prescription.

CORRESPONDING AUTHOR: Nicole Culos-Reed, PhD, University of Calgary, Calgary, AB, T2N
1N4; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2459

D046 10:00 AM-11:00 AM

'ULTIMATELY, MOM HAS THE CALL:' VIEWING CANCER PATIENTS' CLINICAL TRIAL DECISION
MAKING THROUGH THE LENS OF RELATIONAL AUTONOMY

Gladys Asiedu, PhD1, Jennifer L. Ridgeway, MPP2, Aminah Jatoi, MD3, Katherine Carroll, PhD4,
Carmen Radecki Breitkopf, PhD1

1
Mayo Clinic, Rochester, MN; 2Robert D. and Patricia E. Kern Center for the Science of Health
Care Delivery, Rochester, MN; 3Mayo Clinic Rochester, Rochester, MN; 4The Australian
National University, Wandin North, Victoria, Australia

Rational decision models and shared decision making processes offer principles regarding how
cancer patients should approach clinical trial (CT) decisions, but actual social circumstances
may influence patients self-determination, resulting in departures from these decision
processes. This study aimed to understand the social and familial contexts that may impact CT
decisions. In-depth interviews were conducted with 33 ovarian cancer patients who had been
offered a CT and 39 of their family members (FM). Data were analyzed thematically using
constructs derived from the theory of relational autonomy (RA), which argues that social and
structural contexts can support or undermine self-determination. Findings suggest that
knowledge and self-reliance are important to patients as they actively seek the opinion of FM
regarding participation and yet create boundaries around who to include, how, and when.
Some patients exhibited strong self-agency by informing FM of their CT participation only
after the decision had been made. While RA is focused on the self-determination of patients,
this study also offered unique insights into how FM view patients autonomous decision
making. FM largely confirmed that the decision ultimately rests with the patient (ultimately,
mom has the call), as the patient is the one who will bear the risks and burdens of trial
participation. Although most FM appear to be passive in this process there was evidence of
deep relational engagement as FM sometimes sought clarification on the patients position
and said they would influence a shift in perspective or intervene if they felt patients did not
make the right decision. Similar to patients, FM were cautious and created boundaries around
what to say and how to present their views to patients. Other influential social and contextual
factors present in the data included altruism, provider recommendations, past CT experience,
structural and logistic factors. Our study provides a comprehensive understanding of how
social circumstances, significant relationships, and contextual factors may influence cancer
patients autonomy when making CT decisions. It also provides novel insight into how FM
view their role in supporting patient autonomy. CT presentations emphasize patient
autonomy but also should call attention to the relational aspects that influence decision
S2460 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

making. Tools are needed to guide cancer patients and their FM through complex social
factors during CT decision making.

CORRESPONDING AUTHOR: Gladys Asiedu, PhD, Mayo Clinic, Rochester, MN, 55902;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2461

D047 10:00 AM-11:00 AM

UNDERSTANDING BARRIERS AND FACILITATORS TO COLORECTAL CANCER SCREENING

AMONG LATINOS IN A BORDER REGION:FOCUS GROUP FINDINGS

Jessica Haughton, MPH, MA1, Jesse Nodora, DR. PH2, Carolina Lopez, MPH1, Kristen Wells,
Ph.D., MPH3, Rebeca Espinoza Giacinto, MA/MPH, PhD4, Balambal Bharti, Staff Res Assoc2,
Samir Gupta, MD2, Christian Ramers, MD5, Elva M. Arredondo, PhD3

1
San Diego State University/UC San Diego Moores Cancer Center Partnership Institute for
Behavioral and Community Health, San Diego, CA; 2University of California, San Diego, La Jolla,
CA; 3San Diego State University, San Diego, CA; 4San Diego State University/UC San Diego,
Auburn, CA; 5Family Health Centers of San Diego, San Diego, CA

Title: Understanding barriers and facilitators to colorectal cancer screening among Latinos
living in a border region: Focus group findings

Background. Latinos with colorectal cancer (CRC) are diagnosed at late stages and experience
disproportionately higher rates of related mortality. CRC screening has been found to be an
effective tool for management and prevention of this type of cancer, yet screening continues
to be underutilized by Latinos.

Methods. As part of the formative research for developing an intervention to increase

colorectal cancer screening in Latino men and women aged 50 and older, five focus groups
were conducted with Latino males (n=8) and females (n=31); n=11 were not up-to-date with
CRC screening. These focus groups aimed to further understand knowledge and attitudes
about colorectal cancer, perceived barriers to participate in CRC screening as well as to inform
the development of a related community-health care intervention. All focus groups were
conducted in Spanish, audio recorded and transcribed verbatim. Transcripts were analyzed
using inductive content analysis and thematic coding.

Results. Primary themes included: 1) poor access to care and low knowledge of resources for
CRC as barriers to cancer screening; 2) among uninsured participants, fear of incurring
economic costs/debt as a barrier to engaging in preventive health services; 3) significant
influence of primary care provider recommendations to be screened. Overall, participants
equated cancer and cancer treatment with negative outcomes (i.e. death) and were poorly
informed about CRC and types of screening.
S2462 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions. Study findings suggest that efforts to increase knowledge of CRC and screening
are needed among Latinos. Further, low or no-cost CRC screening resources as well as
culturally relevant health education is needed to increase CRC participation by individuals who
are economically disadvantaged and underserved. As seen in prior work, provider
recommendations may have a significant impact on CRC screening. Findings from this
qualitative research will be used to further inform the development of an intervention in a
culturally appropriate CRC prevention study.

Keywords. Colorectal cancer; Screening; Qualitative research; Latinos

CORRESPONDING AUTHOR: Jessica Haughton, MPH, MA, San Diego State University/UC San
Diego Moores Cancer Center Partnership Institute for Behavioral and Community Health, San
Diego, CA, 92123; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2463

D048 10:00 AM-11:00 AM

UNDERSTANDING BARRIERS TO CERVICAL CANCER SCREENING USING PHOTOVOICE AMONG

RURAL INDIAN WOMEN

Prajakta Adsul, MBBS, MPH, PhD1, Purnima Madhivanan, MBBS, M.P.H., Ph.D2

1
National Cancer Institute, Bethesda, MD; 2Florida International University, Miami, FL

Objectives:

Cervical cancer is the second most common cancer diagnosed among women in India and
current estimated indicate that coverage of cervical cancer screening in developing countries
is on average 19% compared to 63% in developed countries. The goal of the present study
was to explore perceptions and beliefs that can increase cervical cancer screening uptake
among women using a community-based participatory approach called photovoice.

Methods

A total of 14 women between 30-60 years, residing in rural villages around Mysore,
Karnataka, India participated in the project. Each participant was provided with a digital
camera and asked to photo document their everyday realities which reflected the perceived
barriers towards cervical cancer screening. During the 3 weeks of photo collection,
participants met individually with the researchers to discuss the meaning of the clicked
photographs. Group discussions were conducted to discuss the action steps required to
address these barriers.

Results

Seven themes emerged from the data: lack of knowledge regarding cervical cancer, methods
and the concept of screening for cancer; socioeconomic factors such as poverty and childcare
that affect participation in cancer screening initiatives; lack of family support from husbands
and mothers-in-law for participating in health care related activities including cancer
screening and further treatment; personal religious beliefs that are not aligned with health
care seeking behaviors in the present health systems for participating in cancer screening;
misconceptions about screening test methodology and the role of physicians and health care
workers in the cancer care continuum; fear of a positive screening test, embarrassment
associated with a pelvic exam, and stigma surrounding a cancer diagnosis that is prevalent in
the community; and lack of appropriate health care services including lack of privacy for
physical examination in primary care clinics and access to a female doctor. Action steps were
S2464 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

identified for each of these themes in group discussions.

Conclusions

Study findings helped identify elements of the social and cultural context of rural communities
thereby providing a rich understanding of the barriers of cervical cancer screening which can
be integrated into pre-intervention capacity development in the future.

CORRESPONDING AUTHOR: Prajakta Adsul, MBBS, MPH, PhD, National Cancer Institute,
Bethesda, MD, 20892-9712; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2465

D049 10:00 AM-11:00 AM

UNDERSTANDING MOTIVATION TO CHANGE THROUGH THE LENS OF A MOTIVATIONAL

INTERVIEWING FRAMEWORK

Heatherlun Uphold, PHD1, Felicity Harper, PhD1, Terrance L. Albrecht, PhD2, Louis A. Penner,
PhD2

1
Karmanos Cancer Institute/Wayne State University, Detroit, MI; 2Wayne State
University/Karmanos Cancer Institute, Detroit, MI

Introduction

Previous research indicates that for parents of pediatric cancer patients, their level of self-
efficacy influences their childs level of distress and cooperation during cancer-related
treatment procedures. Understanding factors that motivate parents to engage in Motivational
Interviewing (MI) interventions to increase their self-efficacy is essential to improving parent
and child coping and emotional well-being. The purpose of this study was to develop a
typology, based on MI principles of importance and self-confidence, to characterize parents
motivation to change.

Methods

Participants were 19 parents of pediatric cancer patients receiving treatment in a Midwestern

urban childrens hospital. Data were compromised of audio-recorded treatment sessions
between parents and a trained MI interventionist. Using qualitative methods, audio-
recordings were analyzed to develop an exhaustive list of parent cognitions and behaviors
that reflected reasons or motivation to change (or not change).

Results

A typology was developed consisting of four categories based on the MI dimensions of

confidence and importance. Parents who were low importance/low confidence were
characterized as unaware or cynical about the need for change and believing it was not their
job to calm their child. Parents in the low importance/high confidence category were skeptical
S2466 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

of the need for change, guarded in admitting weakness, and not open to suggestions for
helping their child. Parents in the high importance/low confidence category were
characterized as wanting to help their child but lacking confidence in their ability to help.
Finally, parents in the high importance/high confidence category were characterized as
optimistic about their ability to help and always having a Plan B.

Conclusions/Implications

This typology provides a framework for assessing parental levels of importance and self-
confidence with respect to increasing their caregiving self-efficacy. This framework furthers
our ability to identify parents motivation to change and ultimately will guide future MI
interventions designed to improve coping and emotional well-being.

CORRESPONDING AUTHOR: Heatherlun Uphold, PHD, Karmanos Cancer Institute/Wayne State

University, Detroit, MI, 48201; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2467

D050 10:00 AM-11:00 AM

USE OF EHEALTH/MHEALTH TECHNOLOGIES FOR SELF-MANAGEMENT AMONG CANCER

SURVIVORS

Yun Jiang, PhD, MS, RN, Brady T. West, PhD, Marcelline Harris, PhD, RN, Debra L. Barton, PhD,
AOCN, FAAN

University of Michigan, Ann Arbor, MI

Background and Purpose: Cancer survivors use of eHealth/mHealth technologies for self-
management needs is not well understood. The purpose of this study was to examine
survivors use of the Internet and mHealth apps for four self-management processes,
including resource utilization, partnerships with healthcare providers, treatment decision-
making, and taking action, and to explore correlates of technology use for each process.

Methods: Secondary analysis utilized data from the Health Information National Trends
Survey 4 (HINTS 4) Cycle 4, a national health information collection program administered by
the National Cancer Institute. Variables examined were chosen based on the Individual and
Family Self-Management theory. All analyses accounted for the complex sampling features,
including weights and stratification. Descriptive statistics were used to characterize the cancer
survivor subpopulation in terms of self-management behaviors, and bivariate associations
between the correlates and indicators of engaging in each process were tested.

Results: Just under 1/3 of cancer survivors had used technologies for seeking online cancer
information (32%), accessing the patient portal (29%), and exchanging medical information
electronically with healthcare professionals (30%). Forty one percent of cancer survivors did
not have a smartphone or tablet, and 36% of those who had a smartphone or tablet did not
have mHealth apps. Among cancer survivors having mHealth apps, 56.3% used the apps for
achieving health-related goals, and 32.9% used the apps to help treatment decision making.
Cancer survivors who were younger, married, employed, and had higher income had higher
odds of using technologies for self-management. In addition, new diagnosis, receiving
treatment, and the belief that everything causes cancer were correlates of online information
seeking or exchanging information with healthcare professionals. BMI and rural residence
were correlates of the use of mHealth apps.

Conclusion: Less than one-third of cancer survivors used eHealth/mHealth technologies for
self-management. Their use may be limited by the availability of technologies and health-
related applications, and was associated with their socio-demographic characteristics, as well
as certain cancer-specific factors and beliefs. Findings from the study may aid the
S2468 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

development of appropriate technology-supported self-management programs for cancer

survivors.

CORRESPONDING AUTHOR: Yun Jiang, PhD, MS, RN, University of Michigan, Ann Arbor, MI,
48109; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2469

D051 10:00 AM-11:00 AM

USING THE COMMON SENSE MODEL TO UNDERSTAND BRCA PREVIVORS

CONCEPTUALIZATION OF THEIR INCREASED RISK FOR CANCER

Courtney L. Scherr, PhD1, Marleah Dean, PhD2, Amy A. Ross, Ph.D. Student3, Charlotte G.
Marshall-Fricker, N/A3, Renee Jacoby, N/A1, Meredith L. Clements, BA, MA2

1
Northwestern University, Chicago, IL; 2University of South Florida, Tampa, FL; 3Northwestern
University, Evanston, IL

Background: Women who have not had cancer, but test positive for a BRCA mutation
(previvors) are at significantly increased risk for developing breast and ovarian cancer. Little is
known about how BRCA previvors conceptualize their increased cancer risk, and
subsequently, make risk management decisions. The Common Sense Model (CSM) provides a
cognitive framework for exploring individuals risk perceptions and management decisions.
Using this framework, we explored how BRCA previvors conceptualize and manage their
cancer risk.

Methods: Female BRCA previvors over 18 years were recruited through social network sites to
participate in a semi-structured telephone interview. Demographic information was collected
during the interview. Interviews were audio recorded and transcribed verbatim. Two coders
conducted content analysis based on an a priori code list including four CSM constructs:
cause, timeline, control, consequence, and emotional response. Inter-coder reliability was
established using Krippendorfs alpha.

Results: Twenty-five BRCA previvors participated. Coders were reliable at = 0.73. Most
previvors obtained genetic testing due to a family history of cancer and/or BRCA mutation.
They described their increased risk as something they always had, but became aware of
through testing. Many identified a certain age at which they were most concerned about
developing cancer, an age that often aligned with the age of the first cancer diagnosis in their
family. This age became a self-imposed deadline for prophylactic surgery decisions,
Knowledge of their BRCA mutation impacted family planning, and in some cases, personal
relationships. Emotions were higher immediately after receiving genetic test results, but
lessened over time, especially after completing prophylactic surgeries.

Conclusion: BRCA previvors conceptualized cancer risk as something they have always had,
and managed their risk by engaging in surveillance prior to undergoing surgery at a self-
imposed age based on family history. Prophylactic surgeries provided a return to normalcy
S2470 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

and eliminated negative outcomes. Understanding how BRCA previvors conceptualize their
risk can inform intervention development.

CORRESPONDING AUTHOR: Courtney L. Scherr, PhD, Northwestern University, Chicago, IL,

60611; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2471

D052 10:00 AM-11:00 AM

WHAT GOOD IS SELF-COMPASSION? BRIDGING MINDFULNESS PRACTICE AND POSITIVE

PSYCHOLOGICAL OUTCOMES AMONG CANCER SURVIVORS

Adrian H.Y. Wan, MSW1, Tiffany Hon, BSc(Psy)2, Rainbow T. H. Ho, PhD1

1
Centre on Behavioral Health, Hong Kong, N/A, Hong Kong; 2The University of Hong Kong,
Hong Kong, N/A, Hong Kong

Background: The practice of mindfulness teaches one to respond to stress in non-reactive

manner. Ample of evidence suggested the benefit of mindfulness in improving psychosocial
well-being of cancer patients. However, little is known about how such transformations took
place. With reference to the Buddhist root of mindfulness practice, this present study
explored the role of self-compassion in mindfulness training for cancer patients.

Methods: 47 Chinese adult cancer survivors attending an 8-week mindfulness-based self-

compassion training program were surveyed on a self-report questionnaire on their level of
reactivity, self-compassion, positive and negative affect, as well as anxiety and depression
symptoms.

Findings: Participants reported significant improvement in non-reactivity, self-compassion, as

well as symptoms of anxiety and depression after attending the 8-week program. Regression
analyses showed that non-reactivity was positive associated with self-compassion and positive
affect; but association was not found either between non-reactivity and negative affect or
anxiety symptoms. Further analyses showed that self-compassion completely mediated the
relationship between non-reactivity and positive affect as well as depression.

Conclusions: Mindfulness practice emphasizes the non-reactive way in responding to daily

life stress; the present study suggested that the association between non-reactivity and
positive coping outcomes were mediated by the presence of self-compassion. Cancer
survivors who are less reactive reported more positive affect and fewer depressive symptoms
by means of becoming more self-compassionate. The cultivation of self-compassion could be
one of the ways to cancer survivors to alleviate mood and nurture positive affect for cancer
survivors.

CORRESPONDING AUTHOR: Adrian H.Y. Wan, MSW, Centre on Behavioral Health, Hong Kong,
N/A, 852; [email protected]
S2472 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D053 10:00 AM-11:00 AM

WHEN CANCER STIGMA GETS UNDER THE SKIN: ASSOCIATIONS OF CANCER STIGMA WITH
COGNITIONS AND QUALITY OF LIFE

Celia C.Y. Wong, MPhil, Qian Lu, Ph.D.

University of Houston, Houston, TX

Objective:

The pink ribbon culture presents breast cancer as a disease of innocence, which results in
much less public stigma against women with breast cancer. While surviving breast cancer is
less stigmatized and even honored in the American culture, stigma remains a significant
stressor among Asian breast cancer survivors. Therefore, the present study aimed to examine
the association between internalized stigma and quality of life among Chinese American
breast cancer survivors, and how impeded cognitive processing (i.e., reduced posttraumatic
growth and more intrusive thoughts) may mediate this association

Methods:

One hundred and thirty-five Chinese breast cancer survivors aged between 33 and 82 (M =
54.55, SD = 9.37) were recruited from Chinese community organizations in Southern
California. Interested and eligible individuals were invited to complete a questionnaire
package assessing their levels of internalized stigma, posttraumatic growth, intrusive
thoughts, and quality of life.

Results:

Results of correlation analysis showed internalized stigma was significantly associated with
increased intrusive thoughts (r = .67) but reduced posttraumatic growth (r = -.28) and quality
of life (r = -.48). Results of path analysis indicated the data fit well to the proposed mediation
model, 2(1) = 0.45, p = .50, CFI = 1.00, TLI = 1.02, RMSEA < .001, SRMR = .01. The indirect
effects of internalized stigma on quality life through posttraumatic growth (standardized
indirect effect = -0.09, p < .05, 95% CI: -.72 to -.07) and intrusive thoughts (standardized
indirect effect = -0.39, p < .01, 95% CI: -1.79 to -.87) were significant, whereas the direct effect
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2473

of internalized stigma on quality of life was not significant (standardized indirect effect = 0.00,
p = .93, 95% CI: -.66 to .63).

Discussion:

Findings suggested internalized stigma was associated with poorer quality of life among
Chinese American breast cancer survivors, and impeded cognitive processing may be the
underlying mechanism. To attenuate the negative associates of internalized stigma and
promote quality of life among Asian breast cancer survivors, cognition-focused interventions
should be developed to foster posttraumatic growth and reduce intrusive thoughts.

CORRESPONDING AUTHOR: Celia C.Y. Wong, MPhil, University of Houston, Houston, TX,
77071; [email protected]
S2474 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D054 10:00 AM-11:00 AM

YOUNG BREAST CANCER SURVIVORS' SPECTRUM OF REPRODUCTIVE AND SEXUAL HEALTH

NEEDS FROM DIAGNOSIS TO SURVIVORSHIP

Kristin Z. Black, PhD, MPH1, Eugenia Eng, DrPH1, Jennifer C. Schaal, MD2, La-Shell Johnson,
MA3, Hazel B. Nichols, PhD1, Diane L. Rowley, MD, MPH1

1
University of North Carolina at Chapel Hill, Gillings School of Global Public Health, Chapel Hill,
NC; 2The Partnership Project, Greensboro, NC; 3UNC Center for Health Promotion and Disease
Prevention, Chapel Hill, NC

Background: The long-term reproductive health impact of cancer treatments is a legitimate

concern for premenopausal women with a breast cancer (BC) history, particularly since it is
recommended that women delay childbearing for 2-3 years after diagnosis. Whether or not
they desire to conceive after treatment completion, all young women with a BC history may
experience treatment-induced reproductive/sexual health issues that potentially affect their
quality of life in survivorship. This study aimed to determine the unmet reproductive
healthcare needs of premenopausal BC survivors. Methods: We recruited women with a BC
history, ages 18-45 at diagnosis, and living in North Carolina to participate in an in-person 2-
part semi-structured interview that inquired about reproductive health history, BC
experience, and reproductive/sexual health concerns around the time of diagnosis. To analyze
the data, we used ResearchTalks Sort and Sift, Think and Shift Method, a multidimensional
qualitative analysis approach. This approach along with member checking with a community
advisory committee and interview participants allowed us to identify the most salient themes
from the data. Findings: Seventeen women with a BC history completed the interview. The
mean age at first BC diagnosis was 38.6 years. Of the 17 women, 12 delivered all of their
children prior to diagnosis, 1 delivered her children after her diagnosis, 2 delivered children
before and after their diagnoses, and 2 did not have any children. Ten of the women were
white and 7 were women of color (e.g., black, Latina, and Asian). The 5 overarching themes
were that women with a BC history: 1) received limited reproductive health information; 2)
desired realistic expectations of conceiving post-cancer; 3) made lifestyle choices based on
family history of BC; 4) struggled with adjusting to their altered physical appearance; and 5)
had menopausal symptoms that led to sexual health and quality of life issues. Despite many
commonalities in these womens experiences, there were subtle differences between women
who did and did not have a child after their BC diagnosis, as well as between women of color
and white women. Conclusions: Women with a BC history are in need of and desire more
education and resources to address their reproductive and sexual health concerns. Young BC
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2475

survivors would greatly benefit from receiving reproductive health counseling to address their
concerns during the BC treatment phase and beyond.

CORRESPONDING AUTHOR: Kristin Z. Black, PhD, MPH, University of North Carolina at Chapel
Hill, Gillings School of Global Public Health, Chapel Hill, NC, 27514; [email protected]
S2476 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D055 10:00 AM-11:00 AM

METABOLIC SYNDROME, INFLAMMATION, AND COGNITIVE FUNCTIONING IN A COMMUNITY

SAMPLE OF AFRICAN AMERICANS

Kanesha Simmons, B.A. 1, Olga Herren, M.S. 2, Denee Mwendwa, Ph.D.2, Clive Callendar,
M.D.3, Alfonso Campbell, Ph.D.2

1
Howard University, Acccoke, MD; 2Howard University, Washington, D.C., DC; 3Howard
University, Washington, D.C>, DC

Research has shown that Metabolic Syndrome (MetS) has adverse effects on cognitive
functioning. Studies have illustrated that pro-inflammatory cytokines such as interleukin-1a
(IL-1a), interleukin-6 (IL-6), and C-Reactive protein (CRP) mediate this relationship. African
Americans are disproportionately affected by MetS and at greater risk for earlier onset and
accelerated decline in cognitive functioning. Few studies have examined this relationship in
African Americans. The current study investigated the relationship between MetS and
cognitive functioning in an African-American community sample. The study also determined if
the association between MetS and cognitive functioning is mediated by IL-1a, IL-6, and CRP
and whether it is moderated by gender and socioeconomic status (SES). The study consisted
of 214 African Americans in the Washington, D.C. metropolitan area. Participants underwent a
medical examination conducted at the General Clinical Research Center at Howard University
Hospital. Blood serum was collected using venipuncture procedure to determine cholesterol,
triglyceride, IL-1a, IL-6, and CRP levels. Weight, height, and blood pressure were also taken.
Participants completed a full health history, as well as a battery of psychosocial and
neurocognitive measures, which included the Stroop, Trail-making A, and Trail-making B.
Results showed no significant main effects of MetS on cognitive functioning. However, the
relationship between MetS and cognitive functioning was mediated by Il-6 and CRP. This
relationship was not moderated by gender or SES. Results emphasize the important role of
inflammation as a mechanism by which MetS impacts cognitive functioning. Future research
should consider the role of inflammation when determining risk for cognitive impairment in
individuals with MetS. The central role of inflammatory processes associated with other
chronic conditions and their impact on cognitive decline should be further investigated.

CORRESPONDING AUTHOR: Kanesha Simmons, B.A. , Howard University, Acccoke, MD, 20607;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2477

D056 10:00 AM-11:00 AM

MILD COGNITIVE DEFICITS IN MEN AND WOMEN WITH AND WITHOUT CORONARY ARTERY
DISEASE: PERSONALITY MATTERS

Bianca D'Antono, PhD

Montreal Heart Institute/Universit de Montral, Montreal, PQ, Canada

BACKGROUND: Mild cognitive deficits and dementias are more prevalent with aging. Pro-
inflammatory and vascular processes likely play a significant role in their pathophysiology.
However, personality traits and affective states that drive much of how individuals interact
with their world, and that have been shown to increase the risk of coronary artery disease
(CAD) morbidity and mortality, may also be of importance. To examine the presence and
correlates of mild cognitive deficits in men and women with and without CAD and no known
cognitive disorder.
METHODS: Five hundred and eighteen men and 212 women with and without CAD (N = 478
and 252, respectively; Age=65.7 6.7 years), and no known cognitive disorder, were
administered the Montreal Cognitive Assessment (MoCA) as well as various psychological and
health questionnaires. Blood was drawn in fasting state and anthropomorphic and BP data
obtained to ascertain metabolic syndrome (MS), as well as CRP and Il-6 levels. Presence of
mild cognitive deficits was determined on the basis of MoCA scores RESULTS: Patients with
CAD showed significantly more mild cognitive deficits compared with similarly aged
individuals with no CAD (41.4% vs 26.6%; p < 0.001). Differences were maintained after
controlling for age, lifestyle, personality, depressive state, and Il-6 (p < 0.007). A forward
logistic regression to predict presence of cognitive deficits led to the following stepwise
inclusions: age (OR=1.071), Il-6 (OR=4.165), hostility (OR=1.044), CAD status (OR=1.634), social
desirability (OR=1.082), and alcohol frequency (OR=0.907) (all p < 0.001, except for alcohol;
Nagelkerke R2=0.132, p < 0.007).
CONCLUSION: Prevalence of mild cognitive deficits among our participants without known
cognitive disorders was high. We found a pro-inflammatory state, vascular impairment, as
well as higher hostility and social desirability traits to predict concurrent cognitive deficits.
Whether interventions targeting these physiological and psychological dimensions could delay
conversion to dementia requires further study.

CORRESPONDING AUTHOR: Bianca D'Antono, PhD, Montreal Heart Institute/Universit de

Montral, Montreal, PQ, H1T 1C8; [email protected]
S2478 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D057 10:00 AM-11:00 AM

OPTIMISM INCREASES THE ODDS THAT ADULTS WITH CARDIOVASCULAR DISEASE ENGAGE IN
EXERCISE

Jacob Julian, B.A.1, Rachel Monane, B.A.1, Laura Meli, M.S.1, Carmela Alcantara, PhD2, Keith
M. Diaz, PhD1

1
Center for Behavioral Cardiovascular Health, Columbia University Medical Center, New York,
NY; 2Columbia University, New York, NY

Background: Exercise following a myocardial infarction (MI) is a critical part of care for cardiac
patients. However, few cardiac patients regularly exercise. One key factor associated with
adherence to recommended cardiovascular health regiments in the general population is
optimism. To examine a potential factor that influences exercise participation among cardiac
patients, we tested the association of dispositional optimism with exercise in a representative
U.S. sample of adults with a history of MI. Methods: Data were collected from a national
sample of 1485 adults with a history of MI (64% male; mean age: 68 11 years). Participants
completed an online survey that queried health beliefs and behaviors. Physical activity was
self-reported by the item, In the past week, on how many days have you done a total of 30
minutes or more of physical activity, which was enough to raise your breathing rate? A
regular exerciser was defined as those who reported >1 day of physical activity of 30 minutes
or more. A revised version of the Life Orientation Test was used to assess optimism.
Participants were stratified into low, moderate, and high optimism groups in accordance with
previous publications. Results: Among the 1485 participants, 178 (11.9%), 736 (49.6%), and
571 (38.5 %) participants reported low, moderate, and high levels optimism, respectively. In a
multivariable adjusted model that included age, gender, race, education, number of heart
attacks, years since last MI, and depressive symptoms, the odds of being a regular exerciser
were significantly higher with greater levels of optimism. Compared to participants with low
optimism, participants with moderate and high optimism were 1.94 (95% Confidence Interval
(CI): 1.35-2.79) and 2.26 (95% CI: 1.51.-3.40) times more likely to be regular exercisers,
respectively (P-trend < 0.001). Conclusion: In this nationally representative sample of adults
with a history of MI, our findings show that optimism is associated with a greater likelihood of
exercising. Future studies should examine whether psychological interventions to improve
positive affect improve adherence to health behaviors among adults with cardiovascular
disease.

CORRESPONDING AUTHOR: Jacob Julian, B.A., Center for Behavioral Cardiovascular Health,
Columbia University Medical Center, New York, NY, 10032; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2479

D058 10:00 AM-11:00 AM

PHYSICAL ACTIVITY BUFFERS DETRIMENTAL INFLUENCE OF STRESS ON DEPRESSION AND

ANXIETY AMONG PATIENTS WITH CORONARY HEART DISEASE

Biing-Jiun Shen, PhD1, Ester Pei Xuan Lee, N/A2, Hung Yong Tay, MS3, Cindy KS Lim, N/A3

1
Nanyang Technological University, Saratoga, CA; 2Nanyang Technological University,
Singapore, N/A, Singapore; 3Singapore Heart Foundation, Singapore, N/A, Singapore

Background: For patients with coronary heart disease (CHD), promotion of physical activity is
an essential task in behavioral intervention to encourage comprehensive healthy lifestyle
changes. Research has shown that stress, depression, and anxiety are prevalent and
detrimental to CHD patients. This study investigated whether physical activity would buffer
the negative effect of stress on depressive and anxiety symptoms among CHD patients.
Methodology: Participants were 161 CHD patients (84% men), with a mean age of 63.7
(SD=9.1), who participated in a community-based cardiac rehabilitation program. Physical
activity was assessed with the International Physical Activity Questionnaire, stress by
Perceived Stress Scale, depression by the Center for Epidemiologic Studies Depression Scale,
and anxiety by the Hospital Anxiety and Depression Scale. Hierarchical multiple regression
was conducted to examine (1) whether higher stress and lower physical activity were
associated with elevated depression and anxiety symptoms, and (2) whether physical activity
interacted with stress to buffer its negative impact on depression and anxiety. All models
were adjusted for age, gender, and ethnicity as standard covariates.
Results: In the model to predict depressive symptom severity, both higher stress ( = .55, p <
.001) and lower physical activity ( = -.15, p = .03) were significantly associated with higher
depressive symptoms, and there was also a significant stress by physical activity interaction (
= -.20, p = .004), indicating that physical activity buffered the negative impact of stress on
depression. The simple slope analysis showed that for patients with higher stress levels (1 SD
above the mean), higher physical activity was strongly associated with lower depressive
symptoms ( = -.36, p = .002), whereas for those with lower stress (1 SD below mean),
physical activity ( =.07, p = .47) was not associated with depression. In predicting anxiety,
higher stress ( = .44, p < .001) and lower physical activity ( = -.16, p = .03) were both
significantly associated with more elevated anxiety, and there was a significant stress by
physical activity interaction ( = -.32, p < .001) indicating a buffering effect. The simple slope
analysis demonstrated that for patients with higher stress, higher physical activity ( = -.51, p
< .001) was strongly associated with lower anxiety; in contrast, for those with lower stress,
physical activity was related to higher anxiety ( =.20, p = .03).
Conclusion: The findings suggest that physical activity appears to be an effective stress buffer
S2480 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

among CHD patients. Promoting physical activity may be associated with lower depression
and anxiety, especially for patients with higher stress levels.

CORRESPONDING AUTHOR: Biing-Jiun Shen, PhD, Nanyang Technological University, Saratoga,

CA, 95070; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2481

D059 10:00 AM-11:00 AM

POST-TRAUMATIC STRESS DISORDER SYMPTOMS AND AVERSIVE COGNITIONS ABOUT

PHYSICAL ACTIVITY IN ACUTE CORONARY SYNDROME PATIENTS

Rachel Monane, B.A.1, Brendan Swan, B.S.1, Othanya Garcia, B.A.1, Ian Kronish, MD, MPH2,
Keith M. Diaz, PhD1

1
Center for Behavioral Cardiovascular Health, Columbia University Medical Center, New York,
NY; 2Columbia University Medical Center, Center for Behavioral Cardiovascular Health, New
York, NY

Background: ~1 out of 8 patients who survive an acute coronary syndrome (ACS) event
develop post-traumatic stress disorder (PTSD). Patients with ACS-induced PTSD have twice the
risk of recurrent cardiac events; however the contributing mechanisms have not been
elucidated. The purpose of this study was to test our hypothesis that patients with ACS-
induced PTSD symptoms avoid physical activity due to fear of trauma-related bodily
sensations (e.g. increased heart rate, shortness of breath). Methods: We studied 133 patients
who presented to the emergency room of a university hospital with ACS symptoms. PTSD
symptoms were evaluated 1-month post-hospitalization using the 17 item PTSD checklist
(PCL-5), tailored to those who incurred a cardiac-related event. Aversive cognitions about
physical activity were also measured with the following items: 1) It scares me when my heart
beats rapidly during physical activity; 2) When I notice my heart beating rapidly during
physical activity, it reminds me of my heart event; 3) When I become short of breath during
physical activity, it reminds me Im at risk of having another heart event; 4) I sometimes
avoid physical activity because Im afraid it will cause me to have a heart event. Results: In
multivariable adjusted analyses, a 5 point increase in PTSD symptoms was associated with a
1.55 (95% CI 1.23-2.04; p < 0.001) odds of feeling scared when the heart beats rapidly during
physical activity, a 2.21 (95% CI 1.58-3.57; p < 0.001) odds of a rapid heartbeat during physical
activity triggering reminders of the cardiac event, a 1.77 (95% CI 1.39-2.40; p < 0.001) odds of
shortness of breath during physical activity triggering worry about future risk of another
cardiac event, and a 1.30 (95% CI 1.10-1.58; p=0.004) odds of avoiding physical activity due to
fear that it could trigger another cardiac event. Conclusion: ACS-induced PTSD is associated
with aversive cognitions about physical activity, a critical secondary prevention behavior.
Interventions that promote positive associations with physical activity may have potential to
reduce the incidence of recurrent events in the high-risk group of ACS-survivors who develop
PTSD symptoms.
S2482 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Rachel Monane, B.A., Center for Behavioral Cardiovascular

Health, Columbia University Medical Center, New York, NY, 10032;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2483

D060 10:00 AM-11:00 AM

PSYCHOLOGICAL DISTRESS IN FEMALE PATIENTS WITH SPONTANEOUS CORONARY ARTERY

DISSECTION

Kelsey Clemson, M.S.1, Katharine Sears Edwards, Ph.D.2, Jennifer Tremmel, M.D.2

1
PGSP-Stanford Psy.D. Consortium, Mountain View, CA; 2Stanford University School of
Medicine, Stanford, CA

Symptoms of psychological distress often occur in patients after myocardial infarction (MI)
and may be more common when MI is caused by Spontaneous Coronary Artery Dissection
(SCAD) (Starovoytov et al., 2016). SCAD is an atypical vascular condition that occurs in patients
who are often young (mean age = 42), female (>80%), and do not have traditional cardiac risk
factors (Hayes, 2014; Giacoppo et al., 2014). Women with SCAD may be particularly
susceptible to emotional distress due to the suddenness and severity of the condition, its lack
of warning signs, and its relative untreatability. Three studies have described high rates of
anxiety and depression in SCAD patients (Liang et al., 2013; Silber et al., 2015; Chou et al.,
2016). As the condition is rare, most published studies include a small n. The goal of this study
was to further characterize female SCAD patients psychosocial distress. Female participants,
identified by Stanford physicians as having had a SCAD event, were invited to participate in a
psychosocial support group. Data was collected in two waves, first in October 2015 for group
participants and second in April 2016 for women who had declined to participate in the group.
Fourteen women (mean age = 51) returned self-report measures of physical and mental
health and quality of life. Demographic information showed that, of 11 women reporting a
history of mental health treatment, 8 indicated their treatment was related to having SCAD.
Scores on self-report questionnaires suggest that most participants were experiencing
clinically significant levels of anxiety and depression or PTSD warranting clinical concern or
diagnosis. Nearly all participants reported high stress levels and many suffered from clinical
levels of insomnia. Descriptive statistics revealed ongoing physical impairment, mostly high
levels of social support, and beliefs that Doctors and Chance had the strongest influence on
their condition. Findings are consistent with earlier studies documenting high rates of anxiety
and depression; this study extends prior research by including other indicators of
psychological distress and as the first known study to investigate PTSD symptoms in SCAD
patients.

CORRESPONDING AUTHOR: Kelsey Clemson, M.S., PGSP-Stanford Psy.D. Consortium,

Mountain View, CA, 94043; [email protected]
S2484 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D061 10:00 AM-11:00 AM

PSYCHOMETRIC PROPERTIES OF THE CONSULTATION AND RELATIONAL EMPATHY (CARE)

MEASURE IN A CARDIOLOGY SETTING

Robert Scales, Ph.D.1, Stewart Mercer, M.D.2, Douglas Murphy, M.D.3, Cengiz Akalan, Ph.D.4,
Farouk Mookadam, M.D.1, Elise Storey, B.S.5, Loreley Hall, B.S.1, Wilansky Susan, M.D.1

1
Mayo Clinic-Arizona, Scottsdale, AZ; 2University of Glasgow, Glasgow, Scotland, United
Kingdom; 3University of Dundee, Dundee, Scotland, United Kingdom; 4Ankara University,
Ankara, Ankara, Turkey; 5Mayo Clinic-Arizona, Sottsdale, AZ

The Consultation and Relational Empathy (CARE) Measure is a 10-item questionnaire that
rates the patient experience and the interpersonal quality of a medical visit. The measure has
been validated in primary care and some sub-speciality disciplines, but it has not been
assessed in cardiology. This was a preliminary study of the utility of the CARE Measure to
assess physicians in cardiology fellowship training. Adult patients completed the CARE
Measure after a medical visit. Face validity was estimated by the number of not applicable
responses. Statistical analysis included principal component analysis (construct validity),
Cronbachs alpha (internal consistency), and Generalisability theory (inter-rater
reliability).Twenty-two cardiology trainees were assessed after 1372 total visits. The level of
not applicable responses was low, ranging from 0.2% to 5.3% per item (average 1.4% across
all 10 items). Missing values were rare (0.2% overall). The measure showed high internal
consistency (Cronbachs alpha coefficient=0.97) and the removal of any of the items reduced
reliability. Factor analysis revealed a single factor with high factor loading for each item. The
mean CARE Measure score at the patient level (n=1372) was 48.3 (SD=4.00), with mean scores
ranging from 23-50. The skew was -2.721 and the kurtosis 7.743 with 48% of patients
reporting a maximum CARE score of 50. The mean CARE Measure score at the physician level
(n=22) was 48.2 (SD=0.85), with an essentially normal distribution (skew -0.314, kurtosis -
0.594). Although mean CARE scores did vary significantly between physicians, the variation
was limited (range 46.5-49.5, p < 0.005) and thus the ability of the measure to discriminate
between individuals in this sample was correspondigly low (G=0.36-0.48). Higher reliability,
suitable for formative feedback (G=>0.5) would require >60 completed questionnaires per
physician. Preliminary findings suggest that the CARE Measure has face and construct validity
with internal reliability in cardiology fellowship training and it may provide a practical
measure of interpersonal communication. Further investigation is needed with a larger
number of physicians and patients in different cardiology settings to determine if the CARE
Measure can differentiate between individuals within this sub-speciality.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2485

CORRESPONDING AUTHOR: Robert Scales, Ph.D., Mayo Clinic-Arizona, Scottsdale, AZ, 85259;
[email protected]
S2486 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D062 10:00 AM-11:00 AM

RACIAL/ETHNIC DIFFERENCES IN PROGRESSION OF SUBCLINICAL ATHEROSCLEROSIS IN THE

NORTH TEXAS HEART STUDY

John M. Ruiz, Ph.D.1, James Garcia, M.S.2, Melissa Flores, MS3, Chul Ahn, PhD4, Matthew
Allison, MD, MPH, FAHA5, Timothy Smith, PhD6, Joshua Smyth, Ph.D7, Daniel J. Taylor, Ph.D.8,
Bert N. Uchino, PhD9

1
University of Arizona, Tucson, AZ; 2Southern Arizona VA Health Care System, Tucson, AZ;
3
The University of Arizona, Tucson, AZ; 4UT Southwestern Medical Center, Dallas, TX;
5
University of California San Diego, La Jolla, CA; 6University of Utah, Park City, UT; 7The
Pennsylvania State University, University Park, PA; 8University of North Texas, Denton, TX;
9
University of Utah, Salt Lake City, UT

Compared to Whites, Blacks/African Americans experience a disproportionate burden of

cardiovascular disease including higher prevalence, incidence, morbidity, and mortality. In
contrast, Hispanics experience similar or relatively better outcomes compared to non-Hispanic
(NH) Whites despite a significantly worse risk profile. One hypothesis is that atherosclerotic
disease progresses more slowly among Hispanics contributing to lower early mortality. The
aim of this study was to examine racial/ethnic differences in progression of carotid intima-
media-thickness (cIMT) as a marker of preclinical atherosclerosis.

Participants were a community sample of 300 healthy adults (150 men, 150 women) ages 21
to 70 years (M=42.44 yrs), stratified by age within gender and race/ethnicity. The sample
included 60% NH Whites, 15% NH Blacks, and 19% Hispanic/Latino. Participants completed a
physical and underwent a bilateral ultrasound imaging of the extracranial carotid vasculature
as part of a larger study. A returning sample of 239 participants (80% retention) repeated all
procedures at a 2-year follow-up. Procedures were used to optimize test-retest over time.
Images were scored by two blinded raters using Vascular Research Tools (V 5.0). Consistent
with field standards, mean of the maximum scores was derived for the common carotid artery
(CCA), bifurcation (BIF), and internal carotid artery (ICA) and difference scores (time 2 time
1) were calculated.

Analyses of Covariance adjusting for demographics (gender, age), socioeconomic status

(household income, education), BMI, and time 1 values of the matched cIMT variable revealed
main effects for both the CCA and BIF, all Fs(2,145) > 3.515, p < .05, np2 > .045. Progression
was greater for NH Blacks than NH Whites with no difference between NH Whites and
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2487

Hispanics. A marginal interaction between race/ethnicity and gender was observed for ICA
with the effect driven by differences among men, F(2,46) = 5.237, p = .009, np2 = .185.
Progression was significantly greater for NH Blacks compared to Hispanics.

These findings document important racial/ethnic differences in atherosclerotic progression

and suggest that the rate of progression among Hispanics is similar to that of NH Whites
despite a greater risk profile.

CORRSPONDENCE: John M. Ruiz, Ph.D., Department of Psychology, University of Arizona.

Email: [email protected]

CORRESPONDING AUTHOR: John M. Ruiz, Ph.D., University of Arizona, Tucson, AZ, 85721;
[email protected]
S2488 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D063 10:00 AM-11:00 AM

RACIAL/ETHNIC DIFFERENCES IN SUBCLINICAL ATHEROSCLEROSIS IN THE NORTH TEXAS HEART

STUDY

John M. Ruiz, Ph.D.1, Chul Ahn, PhD2, Matthew Allison, M.D., MPH3, Timothy Smith, PhD4,
Joshua Smyth, Ph.D5, Daniel J. Taylor, Ph.D.6, Bert N. Uchino, PhD7, James Garcia, M.S.8

1
University of Arizona, Tucson, AZ; 2UT Southwestern Medical Center, Dallas, TX; 3University
of California at San Diego, San Diego, CA; 4University of Utah, Park City, UT; 5The Pennsylvania
State University, University Park, PA; 6University of North Texas, Denton, TX; 7University of
Utah, Salt Lake City, UT; 8Southern Arizona VA Health Care System, Tucson, AZ

Evidence documents significant racial/ethnic differences in heart disease prevalence,

incidence, morbidity, and mortality (Mozzafarian et al., 2015). Data also document black-
white disparities with more recent data characterizing Hispanics as having rates similar to or
better than non-Hispanic (NH) Whites; a paradox given their elevated risk factor profile.
Diagnosis and outcomes can be influenced by access to care. Less known is whether
racial/ethnic groups differ in the undiagnosed or objective measures of subclinical
atherosclerosis. Therefore, the aim of this study is to examine differences in carotid intima-
media-thickness (cIMT) as an objective marker of preclinical disease.

Participants were a community sample of 300 healthy adults (150 men, 150 women) ages 21
to 70 years (M=42.44 yrs), stratified by age within gender and race/ethnicity. The sample
included 60% NH Whites, 15% NH Blacks, and 19% Hispanic/Latino. Participants completed a
physical and underwent a bilateral ultrasound imaging of the extracranial carotid vasculature
as part of a larger study. Images were scored by two raters using Vascular Research Tools (V
5.0). Consistent with field standards, mean of the maximum scores was derived for the
common carotid artery (CCA), bifurcation (BIF), and internal carotid artery (ICA).

Racial/ethnic differences in each segment of cIMT were cross-sectionally evaluated with a

series of Analyses of Covariance adjusting for gender, age, and BMI. No differences were
observed for either the CCA or BIF. However, a main effect of Race/Ethnicity was found at the
ICA, F(2,223) = 6.261, p = .013, np2 = .03. Follow-up revealed that Hispanics had greater cIMT
at the ICA than NH Whites (.777 vs. .719), F(1, 185) = 4.833, p = .029, np2 = .025. Importantly,
inclusion of household income rendered the main effect non-significant, F(2,213) = 0.357, p =
.70, np2 = ns.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2489

Contrary to expectations, Hispanics in this study were found to have somewhat greater
subclinical atherosclerosis in one segment of the carotid vasculature. However, these
differences were accounted for by disparities in SES indicators.

CORRESPONDING AUTHOR: John M. Ruiz, Ph.D., University of Arizona, Tucson, AZ, 85721;
[email protected]
S2490 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D064 10:00 AM-11:00 AM

RELATIONSHIP BETWEEN MINDFULNESS AND HEART RATE VARIABILITY IN PATIENTS WITH

CHF

Brandon Khuu, BS, Paul Mills, Ph.D, Meredith A. Pung, Ph.D, Kathy Wilson, MS, Brook Henry,
Ph.D., Srikanth Kalluri, BS, Laura Redwine, Ph.D

University of California, San Diego, La Jolla, CA

Relationship between mindfulness and heart rate variability in patients with CHF

Chronic heart failure (CHF) is a growing concern, with over 5.1 million cases in the United
States alone and a 5-year mortality rate. Reduced heart rate variability (HRV) is associated
with increased morbidity and mortality. Research shows that psychological factors can impact
HRV, which presents a possible mechanism influencing cardiac health. While mindfulness is
associated with increased HRV in healthy individuals, it is unknown whether this relationship
occurs in patients with serious heart conditions. Methods: Forty-three stage B and C heart
failure patients (mean age = 65.69 +/- 8.23 and left ventricular ejection fraction (LVEF%) =
60.03% +/- 11.48%) from two separate studies were examined for HRV and components of
mindfulness. Components of mindfulness were accessed through the Five Facets Mindfulness
Questionnaire (FFMQ) in both groups. The resting HRV was collected using the Equivital EQ-
01 heart rate sensor, while the participants sat quietly in an upright position for 10 minutes.
Results: Regression analyses adjusted for LVEF% revealed a positive association between HRV
and combined awareness and nonjudging facets of mindfulness (change R2 = .125, t = 2.252, p
= .034). Also, combined mindfulness facets, awareness and nonreactivity showed a positive
association with HRV (change R2 = .101, t = 2.204, p = .035). However, total mindfulness
scores were not related to HRV. Conclusions: These results indicate that there are aspects of
mindfulness that may promote increased HRV in CHF patients. Through the understanding of
the various interactions of the facets of mindfulness, this can help lead to the development of
interventions that may improve physiological responses and potentially improve CHF
prognosis.

NIH/NHLBI -R01HL07355; R01HL096784

CORRESPONDING AUTHOR: Brandon Khuu, BS, University of California, San Diego, La Jolla, CA,
92093; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2491

D065 10:00 AM-11:00 AM

RISK FACTORS FOR HYPERTENSION IN EARLY TO MID-LIFE KENYANS: THE KENYA

DEMOGRAPHIC AND HEALTH SURVEY

Muhammad Abdul Baker Chowdhury, MPH, MPS, M.Sc1, Rahma S. Mkuu, MPH, CPH2, Katrina
Epnere, MPHc, BS3, Nafissatou Cisse-Egbuonye, PhD4, Idethia Harvey, DrPH5, Md Aminul
Haque, PhD6, Gerald Yonga, MBChB, MMed, MD, FRCP (Edin), MBA(healthcare), FESC, FACC7

1
Department of Data Science, The University of Mississippi Medical Center, Jackson MS
39216, Dhaka, Dhaka, Bangladesh; 2Texas A&M University, College Station, TX; 3Florida
International University/Robert Stempel College of Public Health and Social Work, Miami, FL;
4
Black Hawk County Health Department, Waterloo, IA; 5Texas A&M University , Department
of Health and Kinesiology, Transdisciplinary Center for Health Equity Research, College
Station, TX; 6University of Dhaka, Dhaka, Dhaka, Bangladesh; 7Aga Khan University, Nairobi,
Nairobi Area, Kenya

Introduction: Hypertension is the leading cause of adult mortality in developing nations.

However, data on the risk factors of hypertension, especially in Kenya is lacking. This study
identified the risk factors associated with hypertension in Kenya.

Methods: Cross-sectional data from the 2014 Kenya Demographic and Health Survey (KDHS)
was used to assess the risk factors among 12,578 men (age 15-54 years) and 30,431women
(age 15-49 years). Multivariable logistic regression was performed to assess the association
between risk factors and self-reported hypertension.

Results: We found a wide variation of risk factors of hypertension among males and females.
Only age, diabetes, ethnicity, and region were common risk factors for hypertension in both
gender. It was found that older age (OR: 1.80, 95% CI: 1.24, 2.61), having higher education
(OR: 1.55, 95% CI: 1.23, 2.14), having diabetes (OR: 26.45, CI: 16.79, 41.67), and drinking
alcohol (OR: 1.31, 95% CI: 1.05, 1.63) were significantly associated with the increased risk of
hypertension for males. Moreover, living in urban areas (OR: 1.24, 95% CI: 1.06, 1.45) and
being obese (OR: 2.44, 95% CI: 2.04, 2.93) were factors significantly associated with the higher
risk of hypertension for females.

Conclusions: Regardless of gender, individuals over the age of 35 years were more likely to
self-report hypertension compared to the individuals younger than 35 years. However, there
were significant variations between gender and the risk factors associated hypertension.
S2492 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Focus on gender-specific health education programs and interventions are critical in the early
diagnosis of this debilitating disease in Kenya.

CORRESPONDING AUTHOR: Muhammad Abdul Baker Chowdhury, MPH, MPS, M.Sc,

Department of Data Science, The University of Mississippi Medical Center, Jackson MS 39216,
Dhaka, Dhaka, 1000; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2493

D066 10:00 AM-11:00 AM

TREATMENT OF METABOLIC ACIDOSIS IN CHRONIC KIDNEY DISEASE WITH FRUITS AND

VEGETABLES IMPROVES CARDIOVASCULAR RISK

Nimrit Goraya, MD1, Jan Simoni, MD2, Heather Kitzman-Ulrich, PhD3, Donald E. Wesson, MD
MBA3

1
Baylor Scott & White Health, Temple, TX; 2Department of Internal Medicine, Lubbock, TX;
3
Baylor Scott & White Health, Dallas, TX

Background: Individuals with chronic kidney disease (CKD) have dramatically increased risk for
cardiovascular disease (CVD), a major contributor to their premature mortality. Metabolic
acidosis is a common co-morbid condition in CKD for which the currently recommended
treatment is oral Na+-based alkali therapy, commonly NaHCO3. Base-producing fruits and
vegetables (F+V) also improve metabolic acidosis in CKD and might have additional benefits.
We tested the hypothesis that treatment of metabolic acidosis in CKD with F+V rather than
NaHCO3 yields a better cardiovascular risk profile. Methods: Individuals (N=108) with CKD
(stage 3 eGFR 30-59 ml/min) and metabolic acidosis and plasma total CO2 (PTCO2) between
22-24 mM were randomized to either: F+V (n=36) to reduce dietary potential renal acid load
(PRAL) by 50%; Oral NaHCO3 (n=36) to reduce PRAL by 50%; or Usual Care (n=36), and
followed for five years. Outcome variables were collected annually including PTCO2, cystatin
C calculated eGFR, body mass index (BMI), systolic blood pressure (SBP), and low-density
lipoprotein cholesterol (LDL). Results: Baseline values were not different among groups. Five-
year values for PTCO2 were higher in both F+V (23.80.4 mM) and Oral NaHCO3 (23.90.4
mM) than Usual Care (21.90.4 mM, p < 0.01) without significant differences between Oral
NaHCO3 and F+V. Similarly, 5-year values for eGFR were higher in F+V (29.36.1 ml/min/1.73
m2) and Oral NaHCO3 (27.46.1 ml/min/1.73 m2) than Usual Care (20.56.1 mM, p < 0.01),
with no differences between F&V and Oral NaHCO3. By contrast, F+V had lower 5-year values
than Oral NaHCO3 for SBP (1255 vs. 1355 mm Hg, p < 0.01), BMI (26.61.7 vs. 28.41.9, p <
0.01), and LDL (1169 vs. 13610 mg/dl, p < 0.01). Conclusions: Treating CKD individuals with
either F+V or NaHCO3 improved metabolic acidosis and preserved kidney function. By
contrast, those treated with F+V rather than the currently recommended NaHCO3 had lower
values for SBP, BMI, and LDL, indicating an improved CVD risk profile. Integrating F+V into CKD
treatment has the potential to improve overall cardiovascular risk and population health.
S2494 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Nimrit Goraya, MD, Baylor Scott & White Health, Temple, TX,
76508; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2495

D067 10:00 AM-11:00 AM

I JUST WANT TO DO EVERYTHING RIGHT: PRIMIPAROUS WOMENS ACCOUNTS OF

BREASTFEEDING FROM BIRTH TO EIGHT WEEKS VIA AN APP DIARY

Jill Demirci, PhD, RN, IBCLC, Erin Caplan, student, Nora Murray, student, Susan Cohen, PhD

University of Pittsburgh, Pittsburgh, PA

Background: First-time mothers are at risk for breastfeeding problems and early, unintended
breastfeeding cessation for reasons yet to be fully explored.

Purpose: To describe the lived, early breastfeeding experience of primiparous women.

Methods: Sixty-one healthy, primiparous women intending to exclusively breastfeed

downloaded a commercial infant feeding app during their postpartum hospitalization. Women
were instructed to free-text breastfeeding thoughts and experiences daily through eight
weeks postpartum using the apps diary feature; diary data were emailed to investigators
weekly. Diaries were independently coded by three authors line-by-line for content, implicit
meanings, and narrative structure. Coding decisions and emergent themes were discussed in
regular team meetings.

Results: Thirty-five participants completed at least one diary entry and were included in the
analysis. Participants were primarily white, married, and college-educated. The overarching
theme that encompassed the primiparous breastfeeding experience was seeking validation
and sustainability. Particularly in the first weeks post-birth, women were overwhelmed,
frustrated and anxious regarding infant satiety at-breast and the relentless nature of frequent
feeds. Adequate sleep and opportunities for self-care were tipping points for the escalation
and de-escalation of problemsas mothers despaired or expressed hope that they could
withstand breastfeeding under evolving circumstances. Mothers yearned for, and actively
sought work-arounds to achieve predictability in feeding routines. Perceived breastfeeding
success was inextricably tied to maternal self-worth and identity, and almost entirely
adjudicated (and re-adjudicated) through point-in-time weight checks and pediatrician
feedback on adequacy of breast milk volume, rather than mothers own day-to-day
observations and enjoyment of the breastfeeding experience. When success was not
immediate, or when setbacks occurred, self-trust plummeted; some sought absolution from
perceived personal culpability. In later weeks, as problems resolved and the intensity of
breastfeeding abated, women voiced a tentative confidence, future orientation, and resolve
to establish a sustainable feeding schedule.
S2496 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions: The primiparous breastfeeding experience is fraught with internally-imposed,

and externally reinforced, negative pressure to produce rather than connect, to succeed
rather than develop, and to control rather than accommodate.

CORRESPONDING AUTHOR: Jill Demirci, PhD, RN, IBCLC, University of Pittsburgh, Pittsburgh,
PA, 15241; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2497

D068 10:00 AM-11:00 AM

ADAPTIVE SIGNIFICANCE OF PERSONAL PRONOUN USE IN FAMILIES OF ADOLESCENT

SUBSTANCE ABUSERS

Florencia Lebensohn-Chialvo, Ph.D.

University of San Diego, San Diego, CA

A growing body of research suggests that patterns of personal pronoun use in couples
particularly we-talk (first person plural pronouns) and you-talk (second person pronouns)
are potentially meaningful markers of adaptive and maladaptive functioning, respectively.
These associations have not only been demonstrated in terms of relationship quality, but
couples pronoun use has also shown to correlate with various indicators of physical health.
Very little, however, is known about the relational implications of personal pronoun use in
larger social units like families, where relevant interaction patterns are often triadic and
multigenerational. The present study employed data from a multi-site study of family therapy
for adolescent substance drug abuse to (a) describe patterns of pronoun use in families during
conversations that had a collaborative and a conflictual valence; and (b) explore associations
between pronoun patterns and indicators of adaptive adolescent and family functioning. As
hypothesized, automated text analysis of transcripts from 74 families revealed more we-talk
in the cooperative task, more you-talk and I-talk in the conflict task, and significant variations
in pronoun frequency by family role. Additional coding, guided by structural family systems
theory, took into account the source and referent of each pronoun utterance (e.g., parent-
parent we-talk, cross-generation you-talk). As predicted, structural pronoun variables showed
stronger associations with adolescent and family functioning than global (raw count) pronoun
variables. Contrary to expectation, you-talk was a stronger predictor of adaptive functioning
than we-talk, and associations between pronoun patterns and indicators of functioning were
stronger for the conflict task than the cooperation task. The results suggest that relational
meanings of pronouns are substantially more complex in triadic intergenerational family
interactions than in dyadic romantic relationships. Discussion of these results includes study
limitations and possible directions for further research.

CORRESPONDING AUTHOR: Florencia Lebensohn-Chialvo, Ph.D., University of San Diego, San

Diego, CA, 92117; [email protected]
S2498 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D069 10:00 AM-11:00 AM

PROCESS EVALUATION OF A PROMOTORA-LED, HOME-BASED OBESITY PREVENTION

PROGRAM TARGETING LATINO PRESCHOOL CHILDREN

Sharon E. Taverno Ross, PhD1, Ivonne Smith-Tapia, MSW1, Ruth P. Saunders, PhD2, Patricia I.
Documet, DrPH1, Russell R. Pate, PhD2

1
University of Pittsburgh, Pittsburgh, PA; 2University of South Carolina, Columbia, SC

BACKGROUND. Latino preschool children have higher rates of obesity than children from
other racial/ethnic backgrounds. Few effective, culturally-tailored obesity prevention
interventions exist that target preschool children. The purpose of this study was to test the
feasibility and acceptability of a promotora-mediated, home-based intervention to prevent
obesity in Latino preschool children. ANDALE Pittsburgh included 10, 90 minute weekly
interactive and tailored sessions that targeted changes in the home social and physical
environment to build parental self-efficacy.

METHODS. Guided by a logic model, we assessed fidelity, dose, reach, recruitment, and
contextual factors using multiple data sources and quantitative and qualitative methods.
Feasibility of recruitment and intervention protocols were assessed using qualitative data
from promotoras and parents, and Project Coordinator logs. Intervention completeness and
fidelity were assessed through descriptive analyses of multiple data sources.

RESULTS. Six promotoras recruited participants primarily through their own social networks
and delivered the intervention to 50 families (child-mother dyads) that were majority
Mexican-origin, low-income, dual-parent households. Promotoras delivered 75% of sessions
with fidelity (i.e., based on core elements identified pre-intervention, including physical
activity breaks, recipe preparation, goal setting, supportive home environment, barrier
counseling) and 96% of participants completed the whole 10-week intervention. Participants
expressed that session topics were very or extremely useful (95%) and they intended to
continue with changes related to physical activity (74.9%) and healthy eating (70.7%) after the
intervention. Qualitative analysis revealed that shared cultural characteristics between
promotora and participants and the flexible and adaptive intervention approach facilitated
implementation. Barriers to implementation included lack of support from family members
(e.g., husbands) and multiple families living in one home.

CONCLUSION. Initial evidence suggests that ANDALE Pittsburgh is a highly feasible and
acceptable intervention in this sample of Latino families with preschoolers living in an
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2499

Emerging Latino Community. Additional research is needed to determine the effectiveness of

the intervention on outcomes including child weight status, physical activity, and nutrition.

CORRESPONDING AUTHOR: Sharon E. Taverno Ross, PhD, University of Pittsburgh, Pittsburgh,

PA, 15261; [email protected]
S2500 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D070 10:00 AM-11:00 AM

PROMOTING RESILIENCE IN PARENTS OF YOUTH WITH NEUROFIBROMATOSIS: USING

QUALITATIVE DATA TO INFORM INTERVENTION DEVELOPMENT

Catherine Pierre-Louis, BA1, Carmen Rugamas, N/A1, Eric Riklin, BS2, Jessica N. Coleman, BA3,
Ana-Maria Vranceanu, PhD4

1
Massachusetts General Hospital, Boston, MA; 2Fordham University, Boston, MA;
3
Massachusetts General Hospital, Cambridge, MA; 4Massachusetts General Hospital/Harvard
Medical School, Boston, MA

Introduction: Neurofibromatosis (NF1 and 2) are multisystemic, complex and rare diseases
with many still unrecognized features in pediatric patients. Parents of children with NF are at
an increased risk of emotional and physical health problems due to the demands associated
with carrying for a child with NF, including guilt associated with passing on or potential causes
for the NF, dealing with multiple medical, appointments, uncertainty of disease management
and progression, cognitive, social and emotional difficulties experienced by the child, and
trouble around self prioritization. Anecdotal information and clinical experience has shown
that often children sense their parents level of stress and worry, and shy away from sharing
potential concerns, as to not add to their parents distress; ultimately, this leads to increased
stress for children. However, no psychosocial interventions targeting parental stress are
available for this population.

Methods: We aimed to determine parents perceptions of stressors associated with parenting

a child with NF in order to inform the development of a resiliency intervention. We conducted
three live video semi structured focus groups with parents of youth with NF (N=40), which
were subsequently transcribed and coded using grounded theory.

Results: Parents reported heightened stress associated with the childs educational, medical
and social needs, as well as concerns about their childs physical and mental health. They also
reported stress associated with managing finances, multiple medical appointments, role
challenges (i.e. being a parent or partner), and managing uncertainty/unpredictability around
their childs NF diagnosis. These stressors reportedly affected employment status (i.e. work
scale backs), relationships (i.e. social, familial, with partner, other children), and the self (i.e.
negative effects on parents physical and mental health). Few parents reported engaging in
adaptive coping such as mindfulness, exercise, or use of social support.

All participants expressed interest in a mind body program aimed at improving resiliency by
teaching coping skills (e.g., mindfulness, adaptive thinking, positive psychology skills) and
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2501

providing support. Additional topics initiated by parents included facilitating healthy social
interactions for the child, resource sharing, managing financial stressors, and parental role
challenges. Barriers to participation included travel and time of intervention sessions, and all
participants were in favor of the intervention being delivered via a web-based platform (i.e.
Skype).

Conclusion: Results show parents enthusiasm for a resiliency intervention targeting stress
associated with parenting a child with NF, and provide valuable information for the content of
the intervention and its delivery modality.

CORRESPONDING AUTHOR: Catherine Pierre-Louis, BA, Massachusetts General Hospital,

Boston, MA, 02114; [email protected]
S2502 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D071 10:00 AM-11:00 AM

REDUCTION OF PRETERM BIRTH RATE AMONG MEDICAID ELIGIBLE POPULATION THROUGH

AN OUTREACH PROGRAM USING COMMUNITY HEALTH WORKERS

Yondell Masten, PhD, WHNP-BC, RNC-OB, Huaxin Song, PhD

Texas Tech University Health Sciences Center, Lubbock, TX

Introduction
The Centers for Medicaid and Medicare (CMS) funded the Strong Start for Mothers and
Infants grant in 27 centers across the United States (2013-2017). The focus of the grant was
on enhancing prenatal care for at-risk clients to reduce the preterm birth rate and decrease
the number of intensive care days required for preterm neonatal care. As one of the funded
centers, the Maternity Care Home model using certified Community Health Workers (CHWs)
was implemented to provide enhanced prenatal care.
Methods
The Medicaid eligible pregnant clients were referred for care navigation by CHWs from Texas
Tech University Health Sciences Center following the initial nurse intake visit. The CHWs
followed the clients until 6 weeks after birth. The CHWs provided education and services
directed at improving the impact of the social determinants of health on pregnancy and
enhancing routine prenatal care with additional support, services, and home visits. The
primary outcomes include number of prenatal visits, birth weight, gestational age, and NICU
length of stay and secondary outcomes include satisfaction with services, rate of birth control
usage after delivery, and breast feeding rate tracked using both electronic medical record and
surveys. Chi-Square tests, ANOVA and mixed model were conducted using SAS 9.4.
Results
Among the total of 605 clients who delivered between November 2013 to September 2016,
412 clients received at least one CHW visit in the third trimester and were considered as
Active clients. The 193 clients who did not receive a CHW visit during the third trimester due
to the shortage of CHWs or lost client contact were considered as Inactive clients. The
preterm birth rate for the Active group compared with the Inactive group is 7.77% vs. 17.62%
(OR=0.39 [0.23, 0.66], p=0.0004). The average gestational age at delivery among the preterm
birth deliveries for Active group compared with Inactive group is 35.101.85 vs. 31.894.81 (p
< 0.001). Compared with Inactive group, the Active group also experienced lower NICU
admission rate (18.75% vs. 38.24%, p=0.08) and shorter NICU length of stay (26.1710.80 days
vs. 66.5835.05 days, p < 0.05).
Conclusion
This Strong Start Maternity Care Home project provided evidence that enhancing prenatal
care through CHW-provided services can significantly reduce the preterm birth risk among
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2503

Medicaid eligible clients. Additionally, such outcomes may reduce the cost for NICU length of
stay for the preterm new born babies.

CORRESPONDING AUTHOR: Yondell Masten, PhD, WHNP-BC, RNC-OB, Texas Tech University
Health Sciences Center, Lubbock, TX, 79430-6264; [email protected]
S2504 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D072 10:00 AM-11:00 AM

SOCIAL RELATIONSHIP DYNAMICS AMONG ADOLESCENTS WITH AND WITHOUT CHRONIC

MEDICAL CONDITIONS

Lauren E. Wisk, PhD, Elissa R. Weitzman, ScD, MSc

Boston Children's Hospital and Harvard Medical School, Boston, MA

Purpose: Social relationships have been shown to exert short- and long-term effects on
health. As youth with chronic conditions (YCC) may struggle with behavior and adjustment
problems or activity limitations that inhibit their social engagement, their social relationships
may suffer, contributing to poor health. We sought to quantify disparities in various domains
of peer relationships for YCC compared to their healthy peers. Methods: Data are from 2721
adolescents (ages 8-19 years) interviewed for the 2002 and/or 2007 Child Development
Supplement of the Panel Study of Income Dynamics. We utilized multivariate-adjusted
generalized linear mixed models to assess the relationship between having a chronic
condition and each of seven scales measuring various domains of social relationships. Models
included natural cubic splines (to model non-linear effects of age) and evaluated interactions
between splines for age and chronic conditions to determine if age-trends differed for YCC
compared to their healthy peers. Results: Compared to their healthy peers, YCC (58% of
sample) reported experiencing significantly more bullying (1.46 vs 1.34, p < 0.01), less
closeness to friends (2.59 vs 2.66, p=0.01), greater negative peer influence (1.55 vs 1.46, p <
0.01), and lower best-friend relationship quality (2.50 vs. 2.59, p < 0.01). Adjusted analyses
revealed non-linear age-trends in closeness to friends, negative peer influence, and best-
friend relationship quality; however, bullying decreased linearly with age. YCC reported higher
levels of bullying (=0.15, p < 0.01) and negative peer influence (=0.06, p=0.03) across all
ages. Disparities in reported closeness to friends appeared in middle adolescence for YCC
(e.g., predicted scores: 2.58 vs 2.61 at age 13; 2.71 vs 2.86 at age 16). Disparities in
relationship quality diminished in middle adolescence for YCC (e.g., 2.46 vs 2.56 at age 13;
2.52 vs 2.54 at age 16). Conclusion: Compared to their healthy peers, YCC report experiencing
poorer social relationships and more bullying. As poor social relationships contribute to worse
future health, while reports of strong social relationships contribute to better future health, it
may be important for providers to assess social well-being as part of overall health, both to
better understand disease burden and to improve health outcomes. Yet, additional work is
needed to benchmark these measures against clinical metrics to guide clinician evaluation and
interpretation of social health.

CORRESPONDING AUTHOR: Lauren E. Wisk, PhD, Boston Children's Hospital and Harvard
Medical School, Boston, MA, 02115; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2505

D073 10:00 AM-11:00 AM

SOCIO-ECOLOGICAL FACTORS THAT INFLUENCE PERSISTENCE IN BREASTFEEDING AMONG

APPALACHIAN WOMEN: A MIXED METHOD STUDY

Ebtihag O. Alenzi, BPharm, PhD Candidate1, Elvonna Atkins, MSc, BSc2, Brandon Dolly,
PharmD3, Paul Cornelius, MBA4, Kimberly M. Kelly, PhD, MS1

1
West Virginia University, Morgantown, WV; 2Life Treatment Centers, South Bend, IN; 3West
Virginia University, Clay, WV; 4West Virginia University, Star City, WV

Background: The American Academy of Pediatrics recommends breastfeeding children for at

least 1 year, with exclusive breastfeeding for the first 6 months of life. However, the state of
West Virginia is among the lowest for persistence in breastfeeding compared to national
rates. Objective: Guided by Sokols Social Ecological Model, this study examined the
association between persistence in breastfeeding (6-month breastfeeding and extended
breastfeeding) and individual, interpersonal, community and public policy level factors among
Appalachian mothers. Methods: A mixed-method study with convergent parallel design of
women who lived and/or worked in WV who had at least one child aged < 36 months was
conducted. For quantitative data, unadjusted analyses were conducted using Chi-square and
Fishers exact tests, and adjusted analyses were conducted using multivariable and
multinomial logistic regressions. For qualitative data, kappa coefficient coding was conducted
by two researchers. Results: Out of 131 women, 30.5% had breastfed for < 6 months, 39.7%
for 6-12 months, 21.4% for 12-24 months, and 8.4% for 24 months. In the adjusted analysis,
the likelihood of breastfeeding 6 months was associated with not having full-time
employment (AOR=0.3, 95% CI; 0.1-0.9) and high perceived social support (AOR= 28.1, 95% CI;
2.1-381.2) as compared to full-time work and low perceived social support, respectively.
Extended breastfeeding (24 months) was negatively associated with higher education (AOR=
0.06, 95% CI; 0.01-0.1), employment in a full-time work (AOR=0.16, 95% CI; 0.03, 1.0), and
being a first-time mother (AOR=0.09, 95% CI; 0.01-0.8). With regard to qualitative data,
insufficient milk supply, supplement with formula, work, and health issues for child were the
most frequent obstacles reported by mothers who breastfed their children < 6 months.
Natural nutrition, bonding, economic, and health benefits for children were the most frequent
incentives reported by mothers who breastfed their children 24
months. Conclusion: Individual and interpersonal level factors from the Social Ecological
Model were key to persistence in breastfeeding among Appalachian mothers. In light of lost
generations of breastfeeding mothers, social support is critical during the first 6 months to
overcome issues that mothers could experience, such as insufficient milk supply, work, and
S2506 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

health issues, and our study informs future interventions to improve the breastfeeding
experience.

CORRESPONDING AUTHOR: Ebtihag O. Alenzi, BPharm, PhD Candidate, West Virginia

University, Morgantown, WV, 26506; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2507

D074 10:00 AM-11:00 AM

MERITORIOUS AWARD WINNER

THE ACROSS-TIME RELATIONS BETWEEN MOTHERS' DEPRESSIVE SYMPTOMS, MOTHERS'

PROVISION OF SUPPORT, AND CHILD VERBAL COGNITION

Victoria Wu, B.A.1, Patricia East, Ph.D.2, Erin Delker, M.P.H.3, Sheila Gahagan, M.D., M.P.H4

1
UC San Diego School of Medicine, San Diego, CA; 2University of California, San Diego, La Jolla,
CA; 3University of California - San Diego / San Diego State University, San Diego, CA; 4UC San
Diego School of Medicine, La Jolla, CA

The trajectory of childrens cognitive development is shaped by the interaction of supportive

engagement, responsiveness, and provision of learning materials by mothers. Maternal
postnatal depression is known to dampen mothers emotional sensitivity and verbal
stimulation to her child and has been linked with lower child cognitive outcomes. However,
little is known about this interplay on a longitudinal scale. The current study sought to
examine how mothers depressive symptoms predict their subsequent ability to provide
emotional-verbal and learning material support to their child, and how this influences their
childs verbal cognitive outcomes from infancy to adolescence.

A sample of 880 Chilean infants and their mothers was studied across a 16-year period as part
of an iron deficiency anemia preventive trial and follow-up study. Mothers depressive
symptoms, their emotional-verbal responsivity and learning material support to their child,
and child verbal cognition were measured when children were 1 year, 5 years, 10 years, and
16 years. Child gender, family socioeconomic status, family stress, and maternal educational
level were controlled at all time points.

Using cross-lagged structural equation modeling, we found that mothers depressive

symptoms were both directly and indirectly related to their childs verbal cognition scores
across the 16-year period. Indirect effects indicated that higher levels of maternal depressive
symptoms at infancy contributed to lower levels of maternal support at age 5 - both
emotional-verbal responsiveness and learning material support, which in turn was related to
lower child cognitive scores at age 10. Similarly, higher maternal depressive symptoms at age
5 contributed to lower maternal support at age 10 (both emotional-verbal responsiveness and
learning material support), which in turn was related to lower child cognitive scores at age 16.

Findings suggest that the relation between mothers depressive symptoms and child verbal
S2508 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

cognition is mediated by maternal emotional-verbal and material support throughout a childs

lifetime. Ultimately, this may provide clearer entryway into the developmental pathways of
child cognition as well as new insight into age-appropriate mother-child intervention
approaches for clinical practice.

CORRESPONDING AUTHOR: Victoria Wu, B.A., UC San Diego School of Medicine, San Diego,
CA, 92122; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2509

D075 10:00 AM-11:00 AM

THE LIVED EXPERIENCES OF MALE PARTNERS OF WOMEN WHO HAVE BEEN PREVIOUSLY
DIAGNOSED WITH POSTPARTUM DEPRESSION

Judy Kaye Smith, PhD, MSN, RN

Lamar University, Beaumont, TX

Postpartum depression (PPD) is a real complication in the postpartum period that affects 50
to 80% of all women giving birth but is not a condition that solely affects women (Beck, 2006).
This qualitative descriptive phenomenological study based upon Husserls (1960, 1970)
philosophical underpinnings was designed to gain a broader perspective about the
phenomenon of postpartum depression (PPD) and its impact on the family structure through
the lived experiences of male partners of women previously diagnosed with the disorder. A
sample of seven men recruited through a community hospital participated in face-to-face
audio taped interviews that were later transcribed verbatim. The transcripts were rigorously,
critically, systematically analyzed and compared to identify common thematic patterns within
and between the fathers individual experiences using a two group analysis and Spiegelbergs
(1965, 1975) six step process. The men experienced overarching feelings of being vulnerable
when their partners behavior began to change in such a way that they did not recognize the
person their partner had become after the birth of their baby. They began to rationalize the
cause for the changes, with feelings ranging from annoyance to wanting a divorce. But as
things changed within their family, fathers felt the overwhelming need to try to make things
better for their families. The second major theme was one of being helpless to know what to
do or say; but whatever they did was not right or good enough which they attributed to their
lack of knowledge about postpartum depression. Given time, the third overall theme of
coping emerged in which they were able to identify methods of dealing with the changes that
occurred in their lives when their partner was diagnosed with postpartum depression.
Suggestions included the need for more one to one education with parents, Also fathers need
to be patient and more attuned to the needs of their partner. The foremost clinical
implication from the study is the need for healthcare providers to develop better educational
methods to relate information about postpartum depression to childbearing couples.

CORRESPONDING AUTHOR: Judy Kaye Smith, PhD, MSN, RN, Lamar University, Beaumont, TX,
77706; [email protected]
S2510 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D076 10:00 AM-11:00 AM

THE ROLE OF FAMILY IN CURRENT LIFESTYLE CHOICES: A QUALITATIVE SECONDARY DATA

ANALYSIS OF INTERRACIAL AND SAME-RACE COUPLES.

Ana Fonseca, M.A. , Emily Butler, Ph.D.

University of Arizona, Tucson, AZ

Loved ones, including family of origin, have a large impact on our long-term eating behavior
(Bandura, 1999; Savoca & Miller, 2001). In addition, recent studies have established a
recursive link in health patterns (e.g., exercise) between romantic partners; however, how
family of origin and romantic partners interact together to impact current individual health
related behaviors including eating and exercise has received less attention in the empirical
literature. More importantly, how these influences play out between same-versus-interracial
couples may differ, given that racial disparities exist among health outcomes (e.g., higher
probability of chronic diseases, life expectancy, etc.). Notably, it is important to focus on
Hispanic and Non-Hispanic White individuals due to the higher prevalence of obesity found
among Hispanics and general unhealthy patterns in ethnic minorities (Morales et al., 2007;
Ogden & Elder, 1998). Therefore, the goal of this study is to extent results from previous
research by qualitatively exploring how the families of origin for both partners in romantic
relationships (same versus interracial) interact together to impact current health related
behaviors, with an emphasis on the viewpoints that were held in the families of origin about
eating and exercise. Qualitative data analysis and interpretations were based on 14 romantic
couples 7 same-race couples (Non-Hispanic White partners) and 7 interracial couples (Non-
Hispanic White partner and Hispanic partner). A grounded theory approach was taken to
analyze four interview questions to fit the unique needs of the study (Patton, 2002; Strauss &
Corbin, 1998). The data was analyzed within two coding cycles using five different coding
methods. A total of 4 central themes emerged from the data: family of origin eating patterns;
family influence in the adult diet; family of origin physical activity approach; and family
influence in physical activity. Individuals described families as being aware and unaware of
families making healthy and unhealthy food choices during upbringing. Additionally,
individuals described families as making high effort and no effort to be physically active while
growing up. The different perceptions of family of origin in food and exercise were narratively
related to certain eating habits and physical activity behaviors in romantic couples. Different
perspectives from family of origin were found in food and exercise across same-race and
interracial couples. Our results advise that prevention and treatment programs should be
targeting family systems rather than individuals to captivate permanent healthy lifestyle
changes.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2511

CORRESPONDING AUTHOR: Ana Fonseca, M.A. , University of Arizona, Tucson, AZ, 85719;
[email protected]
S2512 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D077 10:00 AM-11:00 AM

RELIGIOSITY AND DIET IN A COLLEGE AND COMMUNITY SETTING

Natalie A. Cyphers, PhD, RN1, Andrea D. Clements, PhD2

1
DeSales University, Center Valley, PA; 2East Tennessee State University, Johnson City, TN

Some researchers have indicated more religious individuals have healthier diets; however,
research is inconsistent. Diet is known to predict the incidence of a variety of diseases,
therefore identifying predictors of diet may inform methods to improve individuals overall
health. A sample of 903 college students and community members (age 18 to 60) participated
in an online survey during spring 2013 or fall 2014. Measures included a demographic
questionnaire, the Perceived Stress Scale (PSS), the Religious Surrender and Attendance
Satisfaction Scale (RSASS), an Intrinsic Religiosity (IR) question, and Starting the Conversation
(a brief dietary questionnaire). Our aim was to determine whether religiosity variables
predicted diet, while controlling for age, gender, and stress. Hierarchical multiple linear
regression indicated 3.5% of diet was predicted by covariates in block one (R2 = .035, F change
(3, 864) = 10.368, p < .001), with the largest proportion predicted by stress ( = .175, p < .001).
After entry of IR, religious commitment (RC) and satisfaction with RC in block two, the total
variance explained by the model was 4.5%, with one percent of the variance being explained
by RC and satisfaction with RC (R2 = .045, R2 change = .011, F change (3, 861) = 3.172, p =
.024). Healthier diet was predicted by lower levels of religious commitment ( = .088, p < .05)
and by being satisfied with religious commitment ( = -.077, p < .05). In the study region,
breaking bread is a central activity within religious communities and may partially explain
the negative association between RC and healthy diet. Future studies should investigate
whether RC may buffer effects of poor diet on health, as the religion/health association is
robust across studies.

CORRESPONDING AUTHOR: Natalie A. Cyphers, PhD, RN, DeSales University, Center Valley,
PA, 18034; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2513

D078 10:00 AM-11:00 AM

OLDER ADULTS AGE AND SEX PREFERENCES FOR VARIOUS PROFESSIONAL PROVIDERS

Ilana J. Engel, B.A., Psychological & Brain Sciences, Washington University in St. Louis1, Brian
D. Carpenter, Ph.D., Clinical Psychology, Case Western Reserve University2

1
Current: University of Chicago, Past: Washington University in St. Louis, Birmingham, AL;
2
Washington University in St. Louis, St. Louis, MO

Purpose: Previous research has found that people have age and sex preferences when
choosing physicians and mental health counselors, but it is not clear whether older adults in
particular have similar preferences when choosing these and other service providers. The
purpose of this study was to explore older adults age and sex preferences for various kinds of
professionals to understand factors that may influence older adults help-seeking behaviors,
which may have important implications for their health outcomes. Methods: Fifty Caucasian,
community-dwelling older adults ages 64 and over completed an online survey. The sample
was 48% female, and participants ranged in age from 64 to 89 years (M = 72.8). The survey
included photos of six individuals representing three age categories (young, middle aged,
older adults) and two sexes (male, female). For each photograph, participants rated their
comfort with and perceived competence of the individual, if their services were sought in five
different professional settings (financial, legal, medical, mental health, and religious). Ratings
were given on a 6-point scale and photographs were presented in a randomized order. In
total, each participant provided 60 ratings (comfort and competence, for each photograph,
for each setting). Responses were analyzed using a within-subjects three-way repeated
measures ANOVA. Results: Older adults preferred older professionals across all five settings.
Female participants preferred female professionals across all settings, and a significant
interaction between sex and profession revealed that in mental health settings, both male
and female participants preferred female therapists. Professionals in the religious setting
were rated lower on comfort and competence than professionals in other settings.
Conclusions: Findings suggest that older adults have age and sex preferences for a broad
range of professionals. It may be beneficial for service providers to inquire about preferences
and accommodate them when possible. Future work is needed to inform whether these
preferences influence actual help-seeking behaviors.

CORRESPONDING AUTHOR: Ilana J. Engel, B.A., Psychological & Brain Sciences, Washington
University in St. Louis, Current: University of Chicago, Past: Washington University in St. Louis,
Birmingham, AL, 35223; [email protected]
S2514 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D079 10:00 AM-11:00 AM

PROSTATE CANCER PATIENTS UNDERSTANDING OF THE GLEASON SCORING SYSTEM:

IMPLICATIONS FOR CLINICAL DECISION MAKING

Erin K. Tagai, PhD, MPH1, Suzanne M. Miller, PhD2, Alexander Kutikov, MD1, Tahseen Al-
Saleem, MD1, Michael A. Diefenbach, Ph.D.3, Sara Fleszar, N/A3, David Y.T. Chen, MD1, Ronak
A. Gor, DO4

1
Fox Chase Cancer Center, Philadelphia, PA; 2Fox Chase Cancer Center/Temple University
Health System, Philadelphia, PA; 3Northwell Health, Manhasset, NY; 4Fox Chase Cancer
Center, Minneapolis, MN

The Gleason Scoring System is critically important for patients in making decisions about their
prostate cancer treatment. A patients Gleason score is designed in part to facilitate risk
communication and to help achieve informed treatment decision making. However, the
system is complex and difficult to communicate to patients, often undermining well-informed
and high-quality decision making. Prostate cancer patients (N=50) that had completed
treatment were recruited to complete a survey during follow-up appointments. Measures
assessed included demographics, Gleason Scoring System knowledge, perceptions of their
Gleason scores impact on their decision making, and anxiety. Biopsy date, cancer stage,
Gleason score at diagnosis, PSA score at diagnosis, and treatment type and completion date
were obtained from patients medical records. Patients had low overall knowledge of the
Gleason Scoring System (66% scored below 50% on a Gleason knowledge scale). However, a
majority of patients (58%) felt their physicians explanation of their Gleason score made sense
to them. This was positively associated with health literacy, Gleason Scoring System
knowledge, and Gleason score at diagnosis (p < .05). Most patients (64%) stated their
understanding of their Gleason score significantly impacted their treatment decision. This was
positively associated with patients age, health literacy, Gleason score knowledge, and cancer
staging (p < .05). PSA anxiety was positively associated with education and time since
treatment completion (p < .05) and a striking 80% of participants reported a fear of
recurrence. While patients have low knowledge of the Gleason Scoring System, most
understand their score is an important part of decision making. However, factors such as low
health literacy and stage of cancer diagnosis are associated with poorer patient understanding
and use of their Gleason score in their decision making. Future research is needed to better
understand how physicians can better communicate patients Gleason score and its impact on
decision making.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2515

CORRESPONDING AUTHOR: Erin K. Tagai, PhD, MPH, Fox Chase Cancer Center, Philadelphia,
PA, 19111; [email protected]
S2516 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D080 10:00 AM-11:00 AM

THE IMPACT OF FINANCIAL INCENTIVES ON ORGAN DONATION INTENTIONS AMONG AFRICAN

AMERICANS

Nakeva Redmond, MPH1, Kimberly J. Arriola, PhD/MPH2, Dana HZ. Williamson, MPH3, Shauna
Flemming, MPH, MDiv4, Laura Harker, MPH5, Jennie P. Perryman, RN, PhD6, Nancy J.
Thompson, Ph.D., M.P.H.7, Rachel E. Patzer, PhD, MPH8

1
Emory University, Atlanta, GA; 2RSPH of Emory University, Atlanta, GA; 3Rollins School of
Public Health, Emory University, Atlanta, GA; 4Emory University, Rollins School of Public
Health, Atlanta, GA; 5Georgia Budget and Policy Institute, Atlanta, GA; 6Emory Transplant
Center/Emory Healthcare/Emory University, Atlanta, GA; 7Department of Behavioral Sciences
and Health education/Rollins School of Public Health/Emory University, Atlanta, GA; 8Emory
University School of Medicine, Atlanta, GA

The Impact of Financial Incentives on Organ Donation Intentions among African Americans

Transplantation is the therapy of choice for many patients experiencing end-stage organ
failure, but a lack of available organs inhibits the number of life-saving transplants that occur.
Offering financial incentives in exchange for a commitment to organ donation has been
explored as a means to increase the donor pool, however, such an act is controversial.
Additionally, while the law of behavior asserts that incentives heighten behavioral effort1,
prior research shows mixed results regarding the extent to which incentives can influence
prosocial behaviors2, such as organ donation. Understanding how financial incentives may
influence donation decisions among African Americans (AA) is of particular importance, as AAs
are overrepresented among those in need of organs yet donate at less than comparable rates,
and historically have had a stronger distrust in the healthcare system than Whites 3. Little is
known about the extent to which donation decision making is influenced by financial
incentives among AAs.

In January 2016, 1,339 AA adults residing in Georgia completed a cross-sectional survey

measuring donation-related attitudes and knowledge, and views of financial incentives.
Approximately 56% of participants indicated that financial incentives would not influence
their decision to donate their organs. Among those who had no intention to donate,
approximately half indicated that financial incentives would have no effect on their decision,
27% said that financial incentives would make them less likely, and 24% more likely to donate.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2517

Greater knowledge of organ donation was associated with greater likelihood of changing a
decision to donate from no to yes, if given a financial incentive, OR=1.09 (95% CI=1.01,
1.17).

Financial incentives were associated with increasing organ donation intentions among AAs.
Efforts exploring the influence of financial incentives on donation decisions should continue to
examine the potentially negative effect of financial incentives, particularly among
communities with higher levels of distrust of healthcare.

References

1. Gneezy U, Meier S, Rey-Biel P. When and Why Incentives (Don't) Work to Modify Behavior.
Journal of Economic Perspectives. 2011;25(4):191-210.

2. Kamenica E. Behavioral Economics and Psychology of Incentives. Annual Review of

Economics. 2012;4:13.11-13.26.

3. Boulware LE, Cooper LA, Ratner LE, LaVeist TA, Powe NR. Race and trust in the health care
system. Public Health Rep. 2003;118(4):358-365.

CORRESPONDING AUTHOR: Nakeva Redmond, MPH, Emory University, Atlanta, GA, 30322;
[email protected]
S2518 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D081 10:00 AM-11:00 AM

TO TAKE OR NOT TAKE PREP: DIFFERENTIATING PERCEPTIONS OF APPROPRIATENESS AMONG

MEN WHO OBJECTIVELY MEET CRITERIA TO BEGIN PREP

Thomas H.F. Whitfield, MA1, Brett M. Millar, MA2, H. Jonathon Rendina, PhD, MPH3, Christian
Grov, PhD, MPH4, Jeffrey T. Parsons, PhD5

1
The CUNY Graduate Center / The Center for HIV Educational Studies and Training, New York,
NY; 2Graduate Center, City University of New York, New York, NY; 3Hunter College of the City
University of New York (CUNY), New York, NY; 4CUNY Graduate School of Public Health and
Health Policy, New York, NY; 5Hunter College and The Center for HIV Educational Studies and
Training, New York, NY

Background: Pre-exposure prophylaxis (PrEP) is the most effective biomedical HIV prevention
tool currently available, however implementation has been slow. Existing research has
identified that the largest barrier to uptake is not viewing oneself as a suitable candidate
for PrEP. The present study sought to investigate what factors contribute to viewing oneself as
being a PrEP candidate among GBM who all meet objective CDC qualifications for eligibility.

Methods: Within a larger U.S. national sample of 1,013 HIV-negative GBM, 438 (43.2%) met
objective CDC qualifications for PrEP eligibility and had engaged in condomless anal sex in the
past 90 days. These men provided demographic characteristics and completed scales
measuring depression, anxiety, internalized homophobia, anticipated HIV stigma, and
perceived social support.

Results: In bivariate analyses, those who did not consider themselves an appropriate
candidate for PrEP were significantlyolder (t(436) = 9.68, p < .01) and more likely to be in a
relationship (2(1) = 11.0, p < .01). No differences were noted for race/ethnicity or education.
In logistic regression analyses with demographic factors, mental health indicators (depression
and anxiety) and perceived social support entered as covariates, higher scores on anticipated
HIV stigma predicted not seeing oneself as an appropriate candidate for PrEP (OR = 0.96, p <
.01).

Conclusions: These results indicate that among GBM who are eligible for PrEP uptake, those
who have less social support and more HIV stigma are less likely to view themselves
as appropriate candidates, thus less likely to initiate a regimen or talk to a health care
provider about PrEP. It is important to target this group for intervention as they are
objectively at risk for HIV infection despite not viewing themselves as at risk. These findings
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2519

indicate that one way to increase uptake for this group may be to focus interventions on
building a social support network and destigmatizing HIV.

CORRESPONDING AUTHOR: Thomas H.F. Whitfield, MA, The CUNY Graduate Center / The
Center for HIV Educational Studies and Training, New York, NY, 10018;
[email protected]
S2520 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D082 10:00 AM-11:00 AM

TRENDS IN GLOBAL COGNITIVE FUNCTIONING AND HEALTH LITERACY IN PATIENTS WITH

CHRONIC OBSTRUCTIVE PULMONARY DISEASE

Irina Mindlis, MPH, Li Chen, MPH, Kimberly Muellers, MPH, Juan Wisnivesky, MD, DrPH

Icahn School of Medicine at Mount Sinai, New York, NY

Background

Poor cognition and low health literacy (HL) have been associated with chronic obstructive
pulmonary disease (COPD). However, limited information is available about the longitudinal
trajectory of cognition and HL in these patients and whether disease progression over time is
associated with worsening function.

Methods

We used data from 258 participants from a prospective cohort study of outpatients with
COPD. Participants were recruited from two major metropolitan hospitals in New York City,
NY and Chicago, IL. Eligible patients were English- or Spanish-speaking, 55 years old, and had
a physician diagnosis of COPD. Individuals were excluded if they had asthma or other chronic
respiratory disease, dementia, or any condition profoundly affecting cognition. Face-to-face
interviews were completed at baseline, 12 and 24 months. HL was measured using the Short
Test of Functional Health Literacy in Adults (S-TOFHLA), which assesses an individuals ability
to read and understand prose passages selected from commonly found healthcare-related
materials. The Mini-Mental State Examination (MMSE) was administered to assess global
cognitive functioning. Spirometry was used to determine percent predicted forced expiratory
volume at 1second (FEV1), a well-established marker of COPD severity. Unadjusted and
adjusted fixed effects models were used to analyze the trajectory of MMSE and STOFHLA over
the 24-month follow up period.

Results

The study cohort consisted of 258 participants, of whom 61% were women, with a
mean age of 67.5 (SD: 7.8) years. Almost half (48%) of the study population was Black, 38%
white, and 13% Hispanic. Unadjusted models showed no significant decrease in HL over the
24-month follow-up period (P=0.83). However, cognition significantly decreased over time,
with a baseline mean MMSE score of 25.4 (SD:3.4) and a decrease to mean=24.7 (SD:3.9) at
24 months (p1 over time was independently associated with a 2.5% increased odds (95% CI:
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2521

1.01-1.03) of impaired cognition.

Conclusions

COPD patients showed significant declines in cognition but not HL over a relatively short
follow-up period. Patients with worsening disease severity appear at the highest risk for
cognitive decline and may require close follow up and support interventions to maintain
adequate functioning and outcomes.

CORRESPONDING AUTHOR: Irina Mindlis, MPH, Icahn School of Medicine at Mount Sinai, New
York, NY, 11238; [email protected]
S2522 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D083 10:00 AM-11:00 AM

UNDERSTANDING AND PROMOTING ORGANIZATIONAL HEALTH LITERACY IN A PUBLIC

HEALTH SETTING

Natalie Kruzliakova, BS1, Kathleen J. Porter, PhD, RD1, Molly O'Dell, MD2, Eleanor S. Cantrell,
M.D.3, Melody Counts, M.D., M.H.M.4, Jamie Zoellner, PhD, RD1

1
Virginia Tech, Blacksburg, VA; 2Virginia Department of Health, Christiansburg, VA; 3Virginia
Department of Health, Wise, VA; 4Virginia Department of Health, Lebanon, VA

Title: Understanding and promoting organizational health literacy in a public health setting

Introduction: Organizational health literacy (HL) is the extent to which health organizations
and systems support individuals as they seek, receive, and use information and services.
Documented efforts to address organizational HL within a public health setting are lacking. To
address this opportunity, a collaborative HL team was formed between researchers and four
southwest Virginia Department of Health (VDH) districts serving federally designated
medically underserved counties. This study describes an organizational HL needs assessment.

Methods: This research is guided by the Agency for Healthcare Research and Quality (AHRQ)
HL Toolkit. VDH staff completed a 56-item survey adapted from the toolkit and based on four
domains of organizational HL. VDH customers completed a validated 3-item individual-level
HL screening measure (1=lowest HL, 14=highest HL) and seven questions on 4-point scale
(1=never, 4=always) pertaining to personal perceptions of VDH services. Analysis included
descriptive statistics and Spearmans correlations.

Results: Among 279 VDH staff, about 50% reported doing well across each HL domain;
however, a subset reported needs improvement or not doing across written
communication (31.4%), self-management and empowerment (23.9%), oral communication
(18.7%), and supportive systems (16.5%) domains. Among 177 VDH customers (82% female,
94% white, 40% high school education) HL scores averaged 12.8 (SD=1.7). Average ratings
pertaining to personal perceptions of VDH services ranged from 3.05-3.64, including following
instructions (M=3.05, SD=1.01), reference to personal history (M=3.25, SD=.81),
comprehension of forms (M=3.43, SD=.73), comprehension of written materials (M=3.51,
SD=.70), time spent with staff (M=3.54, SD=.68), encouragement of questions (M=3.55,
SD=.73), and explanation of services (M=3.64, SD=.60). Customer HL status was significantly
correlated (p< 0.05-0.01) with comprehension of forms (r=.50) and written materials (r=.49),
explanation of services (r=.26), following instructions (r=.16), and reference to personal
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2523

history (r=.16).

Implications: Findings reveal notable strengths in current HL practices, as well as

opportunities to improve customer experiences. This data has been shared with VDH and the
HL Team is developing an Organizational HL Improvement Plan, including implementation of
evidence-based resources from the AHRQ HL Toolkit. To capture changes over time, a 3-year
evaluation process has been established.

CORRESPONDING AUTHOR: Natalie Kruzliakova, BS, Virginia Tech, Blacksburg, VA, 24060;
[email protected]
S2524 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D084 10:00 AM-11:00 AM

WEBMDONT: CREATING A SCALE TO MEASURE ONLINE HEALTH INFORMATION SEEKING

BEHAVIOR

Allyson Hughes, M.A.1, Giovanna Perez, B.A.1, Osvaldo Morera, PhD2

1
University of Texas at El Paso, El Paso, TX; 2Latino Alcohol and Health Disparities Research
(LAHDR) Center, Department of Psychology, The University of Texas at El Paso, El Paso, TX

Individuals examine health information online due to the immediate access to medical
information and minimal costs. About 80 percent of internet users search for health
information online (Fox, 2011). A consequence of obtaining online health information is
inaccurate information. Specifically, reliability and trustworthiness of online health
information is a major concern and extremely hard to measure (Morris, 2008). Very few
studies have examined how online health information is evaluated by individuals seeking
medical advice. For this reason, this study aimed to examine how users assess online health
information. Research questions included: What predicts an individuals willingness to go the
internet to seek medical information? What criteria are individuals using to assess health
information online? What predicts an individuals confidence in the obtained medical
information? What barriers are related to an individual seeking health information online? We
proposed several inductively and deductively generated themes in response to the qualitative
questions. Proposed theoretically grounded themes included: high levels of self-reported
digital literacy, high levels of self-reported self-efficacy regarding internet usage and
information consumption, high levels of confidence in the obtained medical information, and
financial barriers to seeing a medical professional. 152 participants contributed qualitative
and quantitative data via Amazon Mechanical Turk. Qualitative prompts assessed health
information seeking behavior, specifically, what characteristics make the online information
reliable and what antecedents lead participants to pursuing medical information online.
Several themes were generated from the qualitative data. One third of the sample reported
that information found online is very trustworthy. Obtaining the same medical advice from
multiple sites, specifically WebMD, was a method used by participants to check for validity.
Severity and longevity of the health condition was directly linked to willingness to seek health
information online. Top barriers to seeing a healthcare professional included issues with
insurance and financial concerns. Participants reported using information found on google as
a decision making tool to decide whether or not they needed to go to the doctor. The results
from this study highlight how online information is being used to make important health
decisions. The results from this study will also be used to inform the development of a scale
to measure online health information seeking behavior.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2525

CORRESPONDING AUTHOR: Allyson Hughes, M.A., University of Texas at El Paso, El Paso, TX,
79902; [email protected]
S2526 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D085 10:00 AM-11:00 AM

FAMILY DIABETES HISTORY AND ITS INFLUENCE ON HEALTH OUTCOMES

Silvia Santos, Ph.D.1, Maria T. Hurtado-Ortiz, PhD2, Griselda Bernabe, B.A.3, Karina R. Aguilar,
BA4

1
California State University, Dominguez Hills, Carson, CA; 2California State University,
Dominguez Hills, Orange, CA; 3California State University, Dominguez Hills, South Gate, CA;
4
California State University, Dominguez Hills, Wilmington, CA

The present study is part of a larger NIH funded investigation that examined the influence of
family illness-history of diabetes on perceptions of disease threat and health outcomes among
Latino college students who are at-risk for diabetes acquisition due to genetic predisposition.
Type 2 diabetes is a disease previously seen mostly in older age groups; however, in recent
years its prevalence among teens and young adults has increased at an alarming rate. This
study aimed to test a theoretical model linking family history of diabetes (i.e., degree of
genetic risk, relatives illness severity, and distress caused by a relatives illness) and
perceptions of disease threat (i.e., illness seriousness and self-rated risk for disease
acquisition), to objective as well as psychological health status indicators (i.e., BMI, diabetes
risk score, health importance and self-rated health). The sample consisted of 156 Latino
college students who had a first-, second-, or third-generation relative afflicted with Type 2
Diabetes. A path-analysis using linear regression was used to test the proposed theoretical
model. The results revealed a direct path between genetic risk of disease to having a higher
diabetes risk score ( = .268). Similarly, psychological distress caused by a relatives illness was
also directly related to having a higher diabetes risk score ( = .304). Furthermore, a relatives
illness severity impacted health importance ( = .221) as mediated by disease threat ( =
.353). This suggests that the severity of a relatives illness influenced perceptions of diabetes
as a threatening illness, which impacted the value participants placed on their health. In
addition, the degree of psychological distress experienced due to a relatives diabetes was
associated with having a higher self-rated risk for disease acquisition ( = .156), which in turn
was related to having a higher BMI ( = .340) among participants. Lastly, familial genetic risk
was positively associated with having a higher perceived risk for acquiring the disease ( =
.255) among participants, which resulted in having a lower self-rated perceived health status
( = - .351). These results will be discussed in terms of the Health Belief Model and Common
Sense Illness Model of disease.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2527

CORRESPONDING AUTHOR: Silvia Santos, Ph.D., California State University, Dominguez Hills,
Carson, CA, 90747; [email protected]
S2528 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D086 10:00 AM-11:00 AM

LIFESTYLE INTERVENTIONS FOR PRIMARY PREVENTION OF TYPE 2 DIABETES IN U.S. HISPANIC

YOUTH: A SYSTEMATIC REVIEW

Jessica McCurley, M.S.1, Margaret A. Crawford, B.S.2, Linda C. Gallo, PhD3

1
SDSU/UC San Diego, San Diego, CA; 2SDSU/UC San Diego, Encinitas, CA; 3San Diego State
University, Chula Vista, CA

Prevalence of type 2 diabetes in youth has risen rapidly in recent decades along with increases
in childhood obesity. Disparities in both risk and prevalence of type 2 diabetes are evident in
Hispanic/Latino (hereafter, Hispanic) youth compared to their non-Hispanic counterparts.
Targeted diabetes prevention programs have been recommended to reduce risk factors prior
to adulthood in this vulnerable, high-risk population. The current systematic review explored
effectiveness of lifestyle-based diabetes prevention interventions for Hispanic youth following
Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)
guidelines.PubMed/Medline, PsychInfo, Web of Science, and Cochrane Central Register of
Controlled Trials were searched from database inception to August 2016 for studies
evaluating lifestyle-based diabetes prevention trials in U.S. Hispanic youth without diabetes.
Fourteen publications met criteria for inclusion. Ten of fourteen studies were randomized
controlled trials (RCTs); four were uncontrolled. Interventions were heterogeneous in
intensity, content, and setting. The majority of interventions were 12-16 weeks in duration
and were conceptualized as pilot studies. Mean age of participants ranged from 9.8 to 15.8
years, samples sizes were generally small (15-100 participants), and the majority of
participants were overweight [age-and sex-specific body mass index (BMI) 85th percentile].
Fewer than half of interventions (n = 5) resulted in changes in BMI, and 50% (n = 7) reported
changes in fasting glucose or insulin. Of the ten RCTs included, three (33%) reduced BMI and
five (50%) reported significant changes in fasting glucose or insulin. Study quality ranged from
moderate to strong, but effect sizes were generally small. Only four studies reported cultural
tailoring or adaptation of intervention content to improve engagement with the Hispanic
community. In conclusion, evidence for the effectiveness of lifestyle-based diabetes
prevention interventions targeting U.S. Hispanic youth remains limited. Few interventions
demonstrated success in reducing BMI and fasting glucose, effect sizes were small, and follow-
up times were brief. More studies are needed that recruit larger samples sizes, extend follow-
up times, explore innovative delivery modalities, measure and increase youth engagement,
and examine effectiveness across sex and age or pubertal stage groups.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2529

CORRESPONDING AUTHOR: Jessica McCurley, M.S., SDSU/UC San Diego, San Diego, CA,
92102; [email protected]
S2530 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D087 10:00 AM-11:00 AM

SECURE E-MAIL IN DIABETES CARE. A PILOT STUDY

Elin Brsund, PhD, RN1, Cecilie Varsi, MSc, RN1, Kre I. Birkeland, PhD, MD2, Milada
Cvancarova, PhD3, Cornelia M. Ruland, PhD, RN1

1
Center for Shared Decision Making and Collaborative Care Research, Oslo University Hospital,
Oslo, Norway, Oslo, Oslo, Norway; 2Department of Endocrinology, Morbid Obesity and
Preventive Medicine, Oslo University Hospital, Oslo, Norway, Oslo, Oslo, Norway; 3Oslo and
Akershus University College, Oslo, Norway, Oslo, Oslo, Norway

Background: Patient-provider e-mail communication has the potential to fill unmet needs of
patients with chronic conditions. E-mail use in diabetes care has been associated with
improved patient outcomes. This pilot RCT tested the effects of access to a practice-integrated
secure e-mail service (EMS), compared to usual care on problem areas such as anxiety,
depression, quality of life (QoL) satisfaction with treatment and changes in glycemic control
(HbA1C) in patients with Diabetes type I who were followed for 6 months.

Methods: 78 patients at a university hospital were randomized into the EMS group (n=39),
whose members could send messages to, and receive answers from nurses, physicians and
nutritionists at their hospital or the usual care group (n=39). Linear mixed models for repeated
measures were fitted to compare effects on patient outcomes over time, using an intention to
treat approach. In addition, interviews were conducted with four patients in the EMS group.

Results: There was a large range in time since diagnosis among the respondents (0-58 years)
and the usual care group members were significant older than the EMS group (49 years/39
years; p.041). In the EMS group 23% (n=9) sent e-messages (median 1, range 1-13). There
were no significant differences over time between the EMS group and the control group in
terms of problem areas in diabetes (PAID), anxiety (Hospital Anxiety and Depression Scale
(HADS)), depression (HADS), QoL (WHO-5), satisfaction with treatment (DTSQ) or HbA1c.
Interviews revealed that the patients were satisfied with having the opportunity to send
messages to their health care providers.

Conclusion: No differences were revealed between the group with access to e-mail compared
to the control group on self reported outcomes or changes in HbA1c. Lessons learned are to
recruit a more homogenous group in terms of time since diagnosis to be able to detect
differences. E-mail access to health care providers might be more beneficial to newly
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2531

diagnosed people with diabetes type I and thus address unmet information needs in this
patient group.

CORRESPONDING AUTHOR: Elin Brsund, PhD, RN, Center for Shared Decision Making and
Collaborative Care Research, Oslo University Hospital, Oslo, Norway, Oslo, Oslo, 0588;
[email protected]
S2532 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D088 10:00 AM-11:00 AM

SELF-CARE TRACKING AND BLOOD GLUCOSE STABILITY AMONG ONE DROP MOBILE APP
USERS

Tom Quisel, BS1, Luca Foschini, PhD2, David Kerr, MD FRCPE3

1
Evidation Health, Santa Barbara, CA; 2Evidation Health, San Mateo, CA; 3William Sansum
Diabetes Center, Santa Barbara, CA

Background: Self-tracking of weight, diet and exercise is growing in popularity in the United
States. These variables are also important metrics for the self-management of both type 1
(T1D) and type 2 (T2D) diabetes. However, the purported health benefits of consistent
tracking behavior using smartphone applications have been hard to quantify. Here, we
investigated the impact of self-care tracking on self-reported blood glucose (BG) levels in
individuals with T1D and T2D using the One Drop application, a diabetes management app for
people living with TD1, TD2 and Pre-Diabetes.

Methods: We analyzed self-care and BG data entered between January 2015 and January
2016 by 3,378 users with T2D and 999 users with T1D having at least 20 BG readings over a
90-day period. Self-care data included self-reported carbohydrate (carbs) intake,
medication/insulin doses, and physical activity. We aggregated BG data on a per-user-month
basis. For each user-month, we computed the percentage of out-of-range BG readings
(%OOR) (range defined as mg/dl below 70 or above 180) and the logarithm of the counts (log-
count) of each type of self-care behavior during the month. We used fixed-effects panel
regression analysis to model the association between users self-care and %OOR on a per-
month basis while controlling for inter-user variability. We included logged carbs,
medication/insulin taking, activity, and BG log-counts as explanatory variables.

Results: Users were 62% male with an average of 2140 and 1627 BG readings/month in the
T1D and T2D populations respectively. For the T1D cohort, the 37% of BG readings were out
of range compared to 20% out-of-range for the T2D cohort. We found an association between
increased BG logging and decreased %OOR among users with T2D. Doubling the amount of BG
logging was associated with a reduction in %OOR by 0.78 (95% CI [0.28 - 1.29], p = .002). Users
with T1D had a similar, but non-significant pattern, decreasing the %OOR by 0.59.

Conclusions: Increased BG logging with the One Drop app was associated with a lower
monthly percentage of BG readings outside the 70-180 mg/dl range, after controlling for
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2533

inter-user variability in self-care tracking. Future analyses will examine any relationships by
user age, insulin status, and duration of diabetes. Further research should assess whether the
increased stability in observed BG measurements reflects improved glucose control among
One Drop users.

CORRESPONDING AUTHOR: Tom Quisel, BS, Evidation Health, Santa Barbara, CA, 93101;
[email protected]
S2534 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D089 10:00 AM-11:00 AM

SOMATIC AND COGNITIVE ANXIETY IN TYPE 1 AND TYPE 2 DIABETES PATIENTS USING INSULIN

Irina A. Mason, B.A.1, Jennifer F. Ward, PhD2, Barbara A. Stetson, Ph.D.1

1
University of Louisville, Louisville, KY; 2University of Louisville, Hermitage, TN

Background: Anxiety is established as highly comorbid with many chronic medical conditions,
including obesity. Anxiety has been linked to increased number of diabetes mellitus (DM)
symptoms and complications, increased pain, and reduced quality of life. However, research
on anxiety levels in DM populations is scarce. Recent studies suggest that anxiety symptoms
are generally higher in individuals with DM than in the general population, but this
relationship is poorly understood.

Aims: This study aims to 1) describe somatic (sensations) and cognitive (thoughts) anxiety in
DM patients using insulin and (2) explore specific diabetes and health characteristics
associated with both types of anxiety.

Method: Forty three DM patients were recruited from a hospital-based diabetes outpatient
clinic (44% female, 93% White, M age = 42); 72% of the sample had Type 1 and 28% had Type
2 DM. All participants were using insulin, and 44% had a history of significant hypoglycemic
episodes; 72% were overweight or obese. Self-report measures assessed DM history, general
anxiety (Beck Anxiety Inventory, somatic and cognitive scales), and hypoglycemia-related
anxiety (Fear of Hypoglycemia Scale).

Results: Participants diabetes was moderately controlled (M HbA1c = 8.55, SD = 2.04). Nearly
half of the sample (46%) reported anxiety symptoms. Somatic levels were generally lower (M
= .41, SD = .38) than cognitive (M = .45, SD = .62). Anxiety was not associated with time since
diagnosis, frequency of blood sugar testing, injections, or perception of accuracy in estimating
blood sugar. Higher cognitive anxiety was associated with higher levels of worry about
hypoglycemia episodes (r = .51, p < .001) and higher somatic anxiety with higher BMI (r = .53,
p < .001). Somatic anxiety significantly differed by weight classification (ANOVA, F = 5.23, p =
.01) with planned comparisons revealing higher somatic anxiety in obese vs. overweight
participants (t (36) = -3.00, p = .005).

Conclusion: The results support previous findings of high levels of anxiety in DM samples and
may point to important differences in anxiety symptom type. While cognitive anxiety may be
linked to worry about managing diabetes symptoms, such as hypoglycemia, somatic anxiety
may be related to comorbid conditions, such as obesity. Because anxiety has been linked to
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2535

poorer diabetes outcomes, further understanding of this comorbidity is needed and may be a
key to improved diabetes management.

CORRESPONDING AUTHOR: Irina A. Mason, B.A., University of Louisville, Louisville, KY, 40228;
[email protected]
S2536 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D090 10:00 AM-11:00 AM

THE RELATIONSHIP BETWEEN CHANGE IN HBA1C, DEPRESSION, AND DIABETES-DISTRESS

FOLLOWING A 6 MONTH, STEPPED CARE INTERVENTION.

Chelsey Solar, M.S./Ed.S1, Lesley Lutes, PhD2, Doyle M. Cummings, PharmD1, Bertha
Hambidge, M.D., MPH1, Alyssa Adams, MPH1, Marissa Carraway, Ph.D.1, Kerry Littlewood,
Ph.D., MSW3, Sheila Edwards, RN1, Peggy Gaitlin, B.S.1, Shiv Patil, MD, MPH, BC-ADM1

1
East Carolina University, Greenville, NC; 2University of British Columbia, Kelowna, BC,
Canada; 3AAJ Research and Evaluation, Seminole, FL

Objective: Diabetes and diabetes-related distress are commonly comorbid with Type 2
diabetes mellitus (T2DM) and both depression and diabetes-related distress have been found
to be positively correlated with HbA1c. The purpose of this project is to further explore the
relationships between HbA1c, depression, and diabetes-related distress following a 6 month
integrated, collaborative stepped care intervention for primary care patients with
uncontrolled T2DM and co-morbid distress or depression.

Methods: One hundred thirty nine (Mage = 52.6, 77.6% Female, 71% African American) with
uncontrolled T2DM (MHba1c = 9.6) were randomly assigned to intervention or control.
Intervention group participants were placed into one of three steps based on depression and
distress symptoms as assessed by Diabetes Distress Scale (DDS-17) and the Patient Health
Questionnaire (PHQ-9). Participants with low distress or depression received assistance with
self-management; those with moderate distress or depression received problem-solving
therapy; and those with high distress or depression received cognitive behavioral therapy.

Results: At 6 months there was no significant difference in HbA1c change, t(137) = .737, p =
.46, or PhQ9 change, t(107) = 1.9, p = .06 between intervention and control participants.
Participants in the intervention arm (M = -.75, SD = .92) experienced significantly greater
decrease in DDS as compared to controls (M = -.35, SD = .60), t(107) = 2.72, p = .008.
Reduction in DDS scores significantly predicted decreases in HbA1c within intervention
participants F (1, 52) = 9.28, p = .004 with an R2 = .15; however, there was no relationship
between change in DDS and change in HbA1c within control participants.

Conclusions: A collaborative, stepped-care intervention approach appears to reduce diabetes-

related distress in patients with T2DM and comorbid depression or distress.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2537

CORRESPONDING AUTHOR: Chelsey Solar, M.S./Ed.S, East Carolina University, Greenville, NC,
27834; [email protected]
S2538 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D091 10:00 AM-11:00 AM

WEIGHT GAIN IN YOUTH WITH TYPE 1 DIABETES COMPLETING INTERNET PSYCHO-EDUCATION

PROGRAMS

Garrett I. Ash, PhD1, Sangchoon Jeon, PhD2, Margaret Grey, DrPH, RN, FAAN2

1
Yale University, New Haven, CT; 2Yale University, West Haven, CT

Background: Prevalence of overweight among adolescents with type 1 diabetes has grown at
alarming rates since the 1980s. Our Internet-based behavioral interventions have successfully
maintained glycemic control and improved quality of life and diabetes management, but little
is known about the impact of such interventions on weight gain/maintenance.

Methods: We performed a secondary analysis for changes in body weight from the TeenCope
study, a multisite clinical trial where 320 youth, age 11-14 years; 37% minority; 55% female;
HbA1c 8.51.4% were randomized to one of two interventions; TeenCope or Managing
Diabetes. Research assistants retrieved height, weight, and HbA1c data from medical records
before and 12 months after the intervention for 223 participants, who comprised the cohort
for this analysis. We used mixed linear models to examine change in body weight and body
mass index percentile (BMI %ile) over time and whether such changes were influenced by
potential moderators [intervention assignment, change in HbA1c, therapy type (pump versus
injections), gender, and age of diabetes onset].

Results: Among the patients with normal weight at baseline (n=134, BMI5th and < 85th %ile),
average BMI %ile increased from 59.520.7 to 78.314.6 (p < 0.001) and 41% became
overweight. Among the patients with overweight at baseline (n=58, BMI85 th and < 95th
%'ile), average %'ile increased from 90.72.3 to 95.42.5 (p < 0.001) and 62% of them became
obese while virtually all others (36%) remained overweight. The patients with obesity at
baseline (n=31, 95th %'ile) did not change %'ile on average (97.71.4 to 98.51.6, p>0.05)
and virtually all (97%) remained obese. None of the potential moderators influenced weight
changes.

Conclusion: A substantive number of teens became overweight or obese during the

programs. Future studies should investigate behavioral factors moderating these abnormal
BMI increases over 12 months and ways our future interventions could address them.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2539

CORRESPONDING AUTHOR: Garrett I. Ash, PhD, Yale University, New Haven, CT, 06511;
[email protected]
S2540 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D092 10:00 AM-11:00 AM

A DIGITAL HEALTH INTERVENTION FOR WEIGHT LOSS IMPROVES DIET QUALITY AMONG
MEDICALLY VULNERABLE PATIENTS: RESULTS FROM AN RCT.

Dori M. Steinberg, PhD, RD1, Jasmine Burroughs, BS2, Sandy Askew, MPH1, Gary G. Bennett,
PhD1

1
Duke University, Durham, NC; 2Duke Global Health Institute, Durham, NC

Background: The Dietary Approaches to Stop Hypertension (DASH) dietary pattern is a proven
behavioral strategy to manage chronic disease; yet national compliance rates hover around
20%. We examined DASH dietary pattern compliance among high-risk, low-income patients
with diabetes, hypertension, and/or hyperlipidemia enrolled in a behavioral weight loss
intervention.

Methods: Participants (n=351) were enrolled in the Track study; a randomized controlled trial
examining a digital health weight loss intervention that included self-monitoring via mobile
technologies and provider counseling in community health centers, as compared to usual
care. At baseline and 12 months, we used the Block Food Frequency Questionnaire to assess
usual dietary intake. Using valid questionnaires (n=241), we calculated a DASH compliance
score based on recommended levels of 10 nutrients; carbohydrates, protein, total fat,
saturated fat, cholesterol, fiber, magnesium, calcium, potassium, and sodium. Total scores
ranged from 0-10, with higher scores indicative of greater adoption of the DASH dietary
pattern.

Results: Participants were on average 50.7 years old and obese (BMI 35.9 kg/m2). Most were
female (68%) and Black (54%); 12.5% were Hispanic. Half (51%) had an income $25,000/y
and 21% had all 3 chronic diseases. At baseline, average DASH scores were low (2.04 1.16)
and 48% had a score 1.5. DASH compliance improved among intervention participants
compared to usual care; although this was marginally significant (mean change: 0.281.3 vs. -
.011.2; p=.08). This difference was driven by group differences in sodium reduction
(intervention: -1011mg1168 vs control: -249mg 1244; < .0001). Among intervention
participants at 12 months, controlling for baseline, women had significantly higher DASH
scores compared to men (2.421.1 vs. 1.831.1; p=.012) and those with all 3 chronic diseases
reported poorer DASH scores compared to those with fewer (1.790.9 vs. 2.391.2; p=.062).
Further, improvements in DASH adherence were associated with greater weight loss (r=.27;
p=.004).
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2541

Conclusion: A digital health intervention marginally improved adoption of the DASH dietary
pattern among medically vulnerable patients; particularly among women, those managing
fewer chronic conditions, and those who lost weight. Despite these improvements, overall
compliance remains poor. Interventions should focus on improving diet quality to reduce
chronic disease burden.

CORRESPONDING AUTHOR: Dori M. Steinberg, PhD, RD, Duke University, Durham, NC, 27708;
[email protected]
S2542 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D093 10:00 AM-11:00 AM

DEPRESSION AND DIETARY FACTORS IMPLICATED IN OBESITY DEVELOPMENT: A REVIEW

Elizabeth A. Vrany, MS, Melissa A. Cyders, PhD, Adam T. Hirsh, PhD, Jesse C. Stewart, PhD

Indiana University-Purdue University Indianapolis, Indianapolis, IN

Objective: Although poor diet has been proposed as one pathway through which
depression promotes the development of obesity, no review has summarized the literature on
depression as a potential predictor of dietary factors. Accordingly, our aims were (1) to
summarize and evaluate the extant literature on depression and 21 dietary factors implicated
in obesity development and (2) propose a conceptual framework and formulate a future
research agenda for this area.
Methods: A total of 109 studies were included in this review, which was conducted between
March 3, 2016 and July 1, 2016. For 17 of the dietary factors, comprehensive literature
searches were conducted using PsycINFO, MEDLINE, and EMBASE databases. For the other
four dietary factors, results of recent meta-analyses and/or comprehensive reviews were
synthesized.
Results: Findings indicate that there is: (1) strong longitudinal evidence of greater depression
predicting greater eating disorder pathology; (2) strong cross-sectional evidence linking
greater depression to greater emotional eating, dietary restraint, rates of night eating
syndrome, and frequency of skipping breakfast; (3) modest cross-sectional evidence
suggesting that greater depression is inversely associated with Mediterranean and healthy
diets but is unrelated to Western diet and protein intake; (4) mixed cross-sectional evidence
regarding the associations of depression with energy intake, fat intake, carbohydrate intake,
fast food intake, fruit and vegetable intake, sweet food/beverage intake, and external eating;
and (5) insufficient evidence regarding the relationship between depression and the
remaining four dietary factors of adherence to dietary recommendations, meal preparation,
snacking, and portion size.
Conclusions: Based on these results, we developed a conceptual framework clarifying the
relationship between depression and dietary factors implicated in the development of
obesity. We also identify key knowledge gaps and propose a future research agenda for this
area. Of greatest importance, there is a critical need for prospective studies utilizing large,
representative samples. This review may increase researchers and providers awareness of
the numerous ways in which depression could influence dietary factors and, ultimately, the
development and progression of obesity.

CORRESPONDING AUTHOR: Elizabeth A. Vrany, MS, Indiana University-Purdue University

Indianapolis, Indianapolis, IN, 46237; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2543

D094 10:00 AM-11:00 AM

ENGAGING EFNEP AUDIENCES: AN ANALYSIS OF SOCIAL MEDIA POST CONTENT AND

PARTICIPANT ENGAGEMENT

David N. Cavallo, PhD, MPH1, Emily Foley, MS2, Lorelei Jones, M.Ed.3, Valerie Yen, N/A4,
Vanessa Y. Chen, N/A5, Alice S. Ammerman, DrPH6

1
Case Western Reserve University, Cleveland, OH; 2EFNEP, NC State University, Raleigh, NC;
3
NC State University, Raleigh, NC; 4Case Western Reserve University, Flushing, NY; 5Case
Western Reserve University, Canton, OH; 6University of North Carolina at Chapel Hill, Chapel
Hill, NC

Background: The prevalence of social media use among low-income internet users in the US
(56%) suggests that social media may be an appropriate platform for engaging and retaining
participants in programs such as the USDA Expanded Food and Nutrition Education Program
(EFNEP). Although a growing number of health promotion organizations such as EFNEP are
using social media, little formal research has been done to determine best practices in social
media message design around health topics. Identifying effective social media messaging
strategies could improve program engagement resulting in improved behavioral and health
outcomes.

Objective: to describe EFNEP social media campaign message characteristics and identify
those that are associated with greater levels of participant engagement.

Methods: We obtained objective server log data on participant interactions from the
Facebook Insights analytical platform for 406 moderator messages posted across 24 county
and state EFNEP Facebook fan pages over a 26-month period. Codes representing key
engagement strategies were developed through qualitative content analysis and from existing
categories (e.g., message type). Multiple coders then applied the appropriate set of codes to
each message. Using a Welchs adjusted ANOVA due to homogeneity of variance and
unbalanced sample sizes between groups, we examined the association between message
engagement strategies (e.g., messages including pictures, messages focusing on significant
others) and a measure of engagement that summed participant message interactions
(Comments + Shares + Likes + Link clicks + Photo views + Video plays + Other clicks).

Results: Categories of engagement strategies included message types [text only (3.9%), text
with links (74.4%), text with pictures (75.1%), and text with videos (6.7%)], and message
content [use of credible sources of information (21.2%), emphasizing significant others
S2544 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

(48.0%), encouraging discussion of a topic (21.4%), and posing rhetorical questions (35.0%].
The mean number of interactions was 14.84 (SD = 13.35). Engagement strategies positively
associated with interactions included having a picture F(1, 136.48) = 5.75, p < .05 and posing a
rhetorical question F(1, 271.03) = 5.26, p < .05. Negative associations were observed for
emphasizing significant others F(1, 400.42) = 4.54, p < .05; being related to a holiday F(1,
84.38) = 6.39, p < .05; being text only F(1, 15.52) = 15.87, p < .01;and containing motivational
language F(1, 39.49) = 4.90, p < .05.

Implications: Social media messages with a greater number of interactive elements can be
employed by organizations to increase engagement. In some cases, message characteristics
hypothesized to increase engagement produced the opposite effect, highlighting the
importance of continued research that can guide practitioners in the design of social media
campaigns that maximize engagement.

CORRESPONDING AUTHOR: David N. Cavallo, PhD, MPH, Case Western Reserve University,
Cleveland, OH, 44106; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2545

D095 10:00 AM-11:00 AM

FAMILY FUNCTIONING MODERATES ASSOCIATIONS BETWEEN MOMENTARY MOODS AND

BINGE EATING AMONG URBAN ADOLESCENTS

Anna Lee, MPH, MA, Kim Reynolds, PhD, Alan Stacy, PhD, Zhongzheng Niu, MS, MD, Bin Xie,
PhD

Claremont Graduate University, Claremont, CA

Background and purpose

This study builds upon prior research on associations between moods, family functioning (FF),
and binge eating behaviors, using ecological momentary assessment (EMA), an intensive
longitudinal data collection technique that involves participants reportage of data from
within their natural environments. Momentary associations between moods (e.g. positive and
negative emotions) and binge eating behaviors, as well as moderating effects of FF on these
associations, were examined among a nonclinical sample of urban adolescents.

Methods

This study included both time-invariant data reported on survey questionnaires, as well as
time-variant data reported on handheld PDA devices over a 7-day period. Time-invariant data
included demographics and five FF constructs adopted from FACES-IV (e.g. balanced
cohesion, balanced flexibility communication, satisfaction, and balance,); time-variant
data included moods (13 items from the Daily Affect Scale) and binge eating behavioral
responses (2 items from a subscale of the Eating Disorder Diagnostic Scale - binge eating
associated with embarrassment (BE1), and binge eating associated with a sense of loss of
control (BE2)). A multilevel modeling approach was employed to examine associations
between momentary moods and binge eating behaviors moderated by FF constructs, after
controlling for age, gender, ethnicity and obesity status.

Results

All negative emotions were significantly and positively associated with at least one form of
binge eating (BE1 or BE2; p-values < 0.05). The positive association between feeling tired and
binge eating was significantly buffered by family balanced cohesion (BE2: = -0.0016, p =
0.04), family satisfaction (BE2: = -0.0012, p = 0.03), and family balance (BE1: = -0.072, p
= 0.01; BE2: = -0.0071, p = 0.03). Such significant main and interaction effects were not
observed among positive emotions and binge eating (neither BE1 nor BE2).
S2546 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions

Findings indicate that FF buffers positive associations between negative emotions and binge
eating behaviors; this finding indicates the importance of inclusion of FF in the development
of eating behavior interventions for adolescents.

CORRESPONDING AUTHOR: Anna Lee, MPH, MA, Claremont Graduate University, Claremont,
CA, 91711; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2547

D096 10:00 AM-11:00 AM

HOUSEHOLD SOCIAL CONTRIBUTORS TO EVERYDAY DIETARY BEHAVIOR AMONG BLACK

AMERICAN ADOLESCENT FEMALES

Megan Winkler, PhD, RNC-NIC, CPNP-PC1, Sharron Docherty, PhD, PNP-BC, FAAN2, Gary
Bennett, PhD2, Leigh Ann Simmons, Ph.D., M.F.T.3, Debra Brandon, PhD, RN, CNS, FAAN3

1
University of Minnesota School of Public Health, Plymouth, MN; 2Duke University, Durham,
NC; 3Duke University School of Nursing, Durham, NC

Background: Black American adolescent females are disproportionately impacted by obesity

and exhibit high rates of obesity-associated dietary behaviors. However, given that food is
best understood within a social context, it is essential to examine food-related behaviors
situated within a persons household a place shared and created by its members.

Purpose: The purpose of this research was to deeply explore the household relationships and
social processes contributing to Black American adolescent females everyday food
consumption behaviors.

Methods: Data are from a mixed methods, multiple case study conducted with 14 mother-
adolescent female cases (7 adolescent females with obesity). Sixteen different data sources,
including repeated administrations of questionnaires, anthropometrics, interviews, daily
diaries, and field notes, were collected across four home visits for each case and analyzed
using within- and across-case techniques. A highly detailed view of dietary behavior, including
on average 45 distinct food consumption occurrences (more than 14 days of data), was
produced for each adolescent female, along with comprehensive descriptions and evaluations
of the social processes within her household.

Results: An explanatory configuration consisting of three sufficient factors (Mothers

Consistent Mindfulness of the Healthfulness of Foods, Daughters Agreement and
Accommodation to Household Food, Household Member Available and Willing to Plan,
Acquire, and Prepare Food) was developed to elucidate the various household social
contributors to a healthy profile of everyday food consumption among adolescent females.
The configurations explanatory power was estimated by illuminating properties for each
factor and examining their presence across each case (e.g. one of six properties for Mothers
Consistent Mindfulness is whether she frames healthy eating as a best practice for life).

Conclusions: This study is an initial step to understanding the multiple household social
processes contributing to various forms of healthy and unhealthy everyday food consumption
S2548 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

behavior among adolescents. By examining the nuanced ways the cases varied,
contextdependent knowledge essential to understanding the complicated health challenges
of obesity-associated dietary behaviors among Black American adolescent females was
produced; thus, informing both the development and refinement of household-level obesity
interventions.

CORRESPONDING AUTHOR: Megan Winkler, PhD, RNC-NIC, CPNP-PC, University of Minnesota

School of Public Health, Plymouth, MN, 55441; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2549

D097 10:00 AM-11:00 AM

A CONTENT ANALYSIS OF A BARIATRIC SURGERY SUPPORT PAGE ON FACEBOOK:

IMPLICATIONS FOR PROVIDERS

Dylan Jester1, Sarah Domoff, PhD2, Shanu Kothari, MD3, Afton Koball, PhD, ABPP4

1
Gundersen Medical Foundation, La Crosse, WI; 2Central Michigan University, Mt. Pleasant,
MI; 3Gundersen Health Systtem, La Crosse, WI; 4Gundersen Health System, La Crosse, WI

Abstract

Background: In-person support groups are an integral part of the post-operative lifestyle for
patients after bariatric surgery. Past research shows that patients who attend support groups
lose more weight. In todays technological revolution, online bariatric surgery support groups
have formed on social media platforms. Due to the novelty of an online format, a gap exists in
the literature on the content of these support groups. The current study hopes to bridge this
gap by examining the online posts on a Facebook bariatric surgery support page with over
42,000 members.

Methods: All posts from May 1st, 2016 to June 1st, 2016 were coded (N = 3,556 posts).
Content analysis was performed by coding the presence of thematic categories in individual
posts. Categories included seeking nutritional information, posting about progress after
surgery, and commenting about experiences of weight bias. Coders were trained to reliably
code the presence of thematic categories (all > .70).

Results: Content analysis indicated that seeking or providing recommendations was most
prevalent (66%), followed by writing supportive content to group members (37%) and sharing
postoperative progress (21%). When members sought recommendations, questions were
frequently related to nutritional (44%) or medical content (56%). When providing
recommendations, 23% of these posts suggested a specific product or tool (e.g., specific brand
of protein supplement). Patients often described changes in weight (69%) or provided weight
statistics before and after surgery (20%) when writing a progress post. Internalizing weight
bias or writing about experiencing weight bias was more common when posts centered
around changes in appearance (ps < .05).

Conclusion: Results indicate that online bariatric support groups are primarily used for asking
and answering questions, receiving support, and sharing experiences. These findings may be
useful to inform patients about the risks and benefits of joining these groups, and to develop
S2550 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

higher quality support for bariatric surgery patients. Future research examining the accuracy
of online support group content and the credentials of group administrators is warranted.

CORRESPONDING AUTHOR: Dylan Jester, Gundersen Medical Foundation, La Crosse, WI,

54601; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2551

D098 10:00 AM-11:00 AM

COLLEGE STUDENTS' USE OF MOBILE APPLICATIONS FOR PHYSICAL ACTIVITY

Matthew C. Whited, PhD, Licensed Psychologist1, John T. Freeman, B.A.1, Ansley Taylor
Corson, M.A.1, Jordan Ellis, MA1, Emily P. Midgette, Psychology, BA (Expected May 2017)2,
Sherry Pagoto, PhD3, Molly E. Waring, PhD3, Christine May, PhD4, Kristin Schneider, Ph.D.5,
Andrew Busch, PhD6, Brad Appelhans, PhD7, Eliza Varju, BS (Expected May, 2019)1, Laura K.
Nye, BA in Psychology, May 20171

1
East Carolina University, Greenville, NC; 2East Carolina University, Vanceboro, NC; 3University
of Massachusetts Medical School, Worcester, MA; 4University of Massachusetts Medical
School, Fiskdale, MA; 5Rosalind Franklin University of Medicine & Science, North Chicago, IL;
6
The Miriam Hospital, Rumford, RI; 7Rush University Medical Center, Chicago, IL

Physical activity (PA) is important for both mental and physical health. Mobile applications
(apps) have potential to increase PA, especially among heavy technology users such as college
students.

As part of an annual survey of college student health, students were asked: have you ever used a mobile application (app)
for physical activity or exercise? Users, those who reported using an app for PA, were asked about their most and least liked
feature for the most used app. Responses to these open-ended questions were categorized by 3 coders (percent agreement
ranged 71-82%).

Of total participants (n=1906; 67% female; 30% overweight/obese), less than half (39%) reported using an app for PA. Users
were more likely to be women (41% of women vs 35% of men; p=0.008), and more likely to have a BMI of less than 25 (41%
of normal BMI vs 35% of ovwt/obese, p=0.02).

Most users (43%) reported primarily using PA apps to track PA, 23% to increase PA, 11% for fun, 11% because an app was
recommended, and 8% to lose weight. Many (38%) reported using a PA app for less than one month, 25% 1-2 months, 18% 3-
4 months, 19% reported use of 5 months or longer. Length of use did not differ based on reason for use.

Most users (48%) liked tracking/self-monitoring features most followed by receiving tutorial or instruction (24%). Half of
students (50%) did not clearly identify least-liked app features in response to open-ended questions, 17% listed activity
tracking, 6% tutorial/instruction, and 6% financial or time cost.

Over one third of college students have used an app for PA, although most use them for 2 months or less. Just under half of
users were using apps to track a preexisting PA habit while far fewer were using apps to increase PA suggesting that
behavioral strategies in apps may not be attracting sedentary individuals. Almost 2/3 of users reported using the app for less
than 2 months with only 19% using these apps for 5 or more months. Further research is needed to identify how to enhance
PA app features to attract sedentary users for longer periods of time.
S2552 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Matthew C. Whited, PhD, Licensed Psychologist, East Carolina

University, Greenville, NC, 27858; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2553

D099 10:00 AM-11:00 AM

DIGITAL FOOD RECORDS: A MIXED-METHODS STUDY TO EVALUATE FEASIBILITY AND

ACCEPTABILITY FOR COMMUNITY-BASED INTERVENTIONS

Lauren A. Fowler, BS1, Gwenyth R. Wallen, PhD, RN2, Alyssa T. Brooks, PhD3, Leah R. Yingling,
BS4, Marlene Peters-Lawrence, RN5, Michael McClurkin, BS1, Kenneth L. Wiley Jr., PhD6,
Valerie Mitchell, BA7, Twanda D. Johnson, ABA8, Allan A. Johnson, PhD9, Avis P. Graham, PhD,
RD, LDN9, Lennox A. Graham, DM9, Kendrick Curry, PhD10, Tiffany Powell-Wiley, MD, MPH11

1
National Heart Lung and Blood Institute, National Institutes of Health,, Bethesda, MD;
2
Clinical Center, National Institutes of Health, Bethesda, MD; 3National Institutes of Health
Clinical Center, Bethesda, MD; 4National Heart, Lung and Blood Institute, Bethesda, MD;
5
Division of Intramural Research - Hematology Branch, National Heart Lung and Blood
Institute, Bethesda, MD; 6National Human Genome Research Institute, Bethesda, MD;
7
NIH/National Heart, Lung and Blood Institute, Bethesda, MD; 8NIH Clinical Center, Bethesda,
MD; 9Howard University, Washington, DC; 10Pennsylvania Ave Baptist Church, Washington,
DC; 11NHLBI, Bethesda, MD

Background

Optimized dietary assessment tools are needed for use in community interventions for
vulnerable populations. Digital Food Records (DFR), which use handheld devices to capture
dietary intake through text and photography, are a particularly promising mobile assessment
method. Less is understood about the acceptability and feasibility of DFR in community-based
participatory research (CBPR).

Methods

This CBPR-based pilot study was conducted among church-based populations in resource-
limited wards of Washington, D.C. as part of the D.C. Cardiovascular (CV) Health and Needs
Assessment (NCT01927783). Participants (n=17) received a mobile device (iPod Touch) to
photo-document dietary intake for a 3-day DFR using a mobile application (app), FitNinja
(Vibrent Health). After 2 weeks of use, qualitative data were collected in a moderated focus
group (n=8). Acceptability of DFR was explored through thematic analysis of verbatim
transcripts. Feasibility was evaluated by the percentage of participants complying with
instructions (i.e., capturing both before and after meal photos for at least two meals/day for
S2554 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

three days).

Results

Themes identified were related to: 1) feasibility and acceptability of the mobile device and
app, including issues in recording dietary information and difficulty with photo-
documentation; 2) suggestions for additional support and training experiences; and 3)
comparisons to other mobile apps. Overall, participants (all African American, 53% female, 28-
80 years old) were accepting of the photo-documentation tool, contributed to the discussion
of ways to improve the tool, and expressed interest in receiving expert feedback regarding
their diet. Fifteen of the 17 participants (88.2%) photo-documented at least one meal over the
study period. Most meals (89.4%) were recorded with at least one photo. Only 3 participants
(17.6%) provided before and after photos for at least two meals for three days.

Conclusion

This study demonstrates DFR as an acceptable tool in CBPR. Participation rates and feasibility
may be improved with the development of a tailored system of reminders to increase
before/after meal pictures. CBPR allows novel assessment methods to be optimized for use in
real-world settings by tailoring tools based on feedback from community members. Increasing
knowledge of the acceptability and feasibility of digital technology to measure dietary intake
in CBPR could improve interventions that promote healthy eating and reduce CV health
disparities.

CORRESPONDING AUTHOR: Lauren A. Fowler, BS, National Heart Lung and Blood Institute,
National Institutes of Health,, Bethesda, MD, 20892; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2555

D100 10:00 AM-11:00 AM

DIGITAL HEALTH TOOLS FOR SLEEP SELF-MANAGEMENT: WORKING MOTHERS USE CASE

Katarzyna Wac, PhD1, Laura Montanini, MSc2, Kimie Bodin Ryager, MSc3, Homero Rivas, MD4

1
Stanford University, Stanford, CA; 2Marche Polytechnic University, Italy, Stanford, CA;
3
University of Copenhagen, Denmark, Palo Alto, CA; 4Stanford University, USA, Palo Alto, CA

Ever sophisticated wearable activity trackers and mobile applications enable to assess
individuals daily life patterns and factors influencing their long-term health state. Working
mothers may be especially vulnerable to experience poor sleep hygiene as they try to
balance work with family obligations, suffering from chronic stress, and having poor social
support and insufficient personal resources to deal with daily challenges. Poor sleep hygiene
may be developing over a longer period, resulting in fatigue and exhaustion and may have
destructive long-term influences on the individuals psychological and physical health,
cognition, and behavior, leading to a burnout state.

We present an approach to operationalizing the major risk factors for the prevalence of poor
sleep hygiene in working mothers - including physical factors (e.g., physical activity, alcohol,
time to bed), psychological factors (e.g., stress, feeling 'lack of control'), social factors (e.g.,
family settings, social pressure for socializing) and environmental factors (e.g., light, noise,
room temperature). We have thoroughly examined these factors to identify which of those
can be operationalized via ICT-based tools and quantified by leveraging unobtrusive sensors,
including sensors embedded in personal smartphones and wearables. Such an
operationalization will enable to provide reliable assessment and prediction of risk exposure,
which, in turns, will permit early recognition and preventive interventions.

To support our research claims we have conducted an explorative pilot study with six working
mothers wearing BASIS PEAK activity and sleep tracker for a minimum of three months each.
We have collected both their subjective views (via interviews) and the underlying smartphone
and wearable device usage logs in situ. The results expose, on the one hand, the low physical
activity levels, late time to bed, high stress and social pressure as well as poor environmental
conditions of most of the mothers. On the other hand, the results indicate a high importance
of routines and predictability in the daily lives of mothers, which may enable better sleep
hygiene. Additionally, a wearable devices features like interaction design, battery lifetime and
fashion-ability; influence the quality of the data collected. Given the results and existing
research so far, we discuss design implications for monitoring and assessment of multi-
dimensional events, subjective symptoms as well as physiological and behavioral variables in
S2556 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

the natural daily settings of the working mothers for assurance of their sleep hygiene and
prevention of burnout condition in a long term.

CORRESPONDING AUTHOR: Katarzyna Wac, PhD, Stanford University, Stanford, CA, 94304;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2557

D101 10:00 AM-11:00 AM

EFFECTS OF AN ONLINE BOOSTER INTERVENTION ON BEHAVIORAL CHANGE IN A

LONGITUDINAL BONE HEALTH TRIAL

Eun-Shim Nahm, PhD, RN, FAAN1, Barbara Resnick, PHd, CRNP, FAAN, FAANP 2, Shijun Zhu,
PhD2, Michele Bellantoni, MD3, Kathleen Charters, PhD, RN4

1
University of Maryland School of Nursing, Baltimore, MD; 2University of Maryland, School of
Nursing, Baltimore, MD; 3Johns Hopkins School of Medicine, Baltimore, MD; 4Uniform Health
Services University, Columbia, MD

Osteoporosis is a serious public health problem affecting 52 million Americans. Although

several interventions have been shown to be effective in optimizing bone health, they have
not been fully incorporated into individuals daily lives. In an online dissemination study, our
research team assessed the impact of two online theory-based bone health programs on
selected health outcomes in older adults. The two interventions were the (1) Bone Power
program (8-week) and the (2) Bone Power Plus program, including a 10-month online booster
intervention followed by the 8-week Bone Power program. The aim of the study was to test
the impact of the online booster intervention on self-efficacy/outcome expectations for
calcium intake and exercise, and calcium intake and exercise behaviors. (8-week outcomes of
the study were published elsewhere.) The study was guided by social cognitive theory. This
was a 3-arm RCT with 4 observations (baseline, 8 weeks, 6 months, and 12 months).
Individuals were eligible if they were age 50 or older and could use the Internet/e-mail
independently. Participants were recruited from two online communities, SeniorNet and My
HealtheVet. The 8-week Bone Power program was comprised of learning modules, moderated
discussion boards, Ask-the-Experts, and a virtual library. Participants also had access to video
lecture libraries and a health toolkit. The booster intervention used for the Plus program was
a 10-month biweekly eHealth Newsletter. The newsletters included contents that highlighted
major health topics in the learning modules, recent research findings, calcium-rich recipes, an
interactive goal attainment form, and participants success stories. For the primary analysis,
linear mixed models were used to compare the intervention effects between intervention
groups on each of the outcomes. A total of 866 participants were randomized into either the
control (n=263), Bone Power (n=301), or Bone Power Plus (n=302) group (mean age, 62.8+8.5
years). At 12 months, participants in the Plus group showed substantially greater
improvement compared to the control group in calcium self-efficacy (t[1309]=1.99; p=.046),
exercise self-efficacy (t[1309]=2.1, p=.036), amount of exercise (t[1309]=2.38, p=.018) and
minutes of exercise (t[1309]=2.48, p=.013). Participants in the Bone Power group showed
greater improvement in calcium outcome expectation (t[1309]=2.68, p=.008) than those in
S2558 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

the control group. Although no statistical significance was observed, participants in the Plus
group tended to have a higher improvement than those in the Bone Power group in self-
efficacy for calcium and exercise and amount of exercise. Overall, the findings suggest
promising potential for using eHealth behavior interventions for large numbers of older
adults. The effects of booster interventions, however, need further investigation using varying
approaches.

CORRESPONDING AUTHOR: Eun-Shim Nahm, PhD, RN, FAAN, University of Maryland School of
Nursing, Baltimore, MD, 21201; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2559

D102 10:00 AM-11:00 AM

MERITORIOUS AWARD WINNER

ENGAGEMENT OVER ONE YEAR IN A BILINGUAL TEXT MESSAGING INTERVENTION PREDICTS

WEIGHT LOSS AMONG OVERWEIGHT/OBESE ADULTS: CONTXT

Margaret A. Crawford, BS1, Gina Merchant, PhD2, Matthew Allison, MD, MPH, FAHA2, Job
Godino, Ph.D.3, Kevin Patrick, MD, MS4

1
UC San Diego, Encinitas, CA; 2University of California San Diego, La Jolla, CA; 3UC San Diego,
La Jolla, CA; 4UCSD School of Medicine and The Qualcomm Institute, La Jolla, CA

Purpose: Text messaging is a promising weight-loss intervention tool because it allows for
continuous support, minimizes cost, and maximizes distribution. However, limited work has
investigated how participant engagement with study-delivered text messages impacts weight
loss, especially among diverse populations. We present a quantification of engagement and its
association with change in weight throughout a 12-month intervention.

Methods: 197 participants were randomized to receive the text messaging intervention (BMI
> 27 to < 40, ages 21-62, 128 female, 66 Hispanic). Our analyses include participants who had
complete information for the respective timepoints: 171 participants at 6 months and 159
participants at 12 months. Participants received 2-4 messages/day that were tailored on
baseline weight management strategies. Messages were iterative and interactive, enabling
goal setting and self-monitoring of physical activity and diet. Engagement was defined as: the
number of texts participants replied to / the number of texts received that asked for a reply
(interactive messages). Engagement was anchored to participants individual study start
dates, and was aggregated across two time periods: baseline-6 months and 6-12 months.

Results: Participants received a total of 210+24 and 278+36 interactive messages from
baseline-6 months and 6-12 months, respectively. Participants sent 230+98 and 230+140
replies from baseline-6 months and 6-12 months, respectively. Average engagement score
from baseline-6 months was 1.1+ .48, and 0.85+ .52 from 6-12 months. Average weight
change was -2.5 kg+5.79 from baseline-6 months, and .05 kg+4.73 from 6-12 months. In
multiple linear regression models that adjusted for baseline weight, age, sex, and ethnicity,
higher engagement was significantly associated with lower weight at 6 and 12 months. For
every SD increase in engagement, participants lost 1.63 kg of weight from baseline-6 months,
and 1.62 kg of weight from 6-12 months (p's < .001). Separate models assessed the
association of weight change with the number of replies sent, instead of the engagement
S2560 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

variable, and adjusted for number of interactive messages recieved, age, sex, and ethnicity. In
these models, the number of replies sent by the participant remained signficiantly associated
with weight loss at 6 and 12 months (p < .001).

Conclusion: An engagement score that accounts for replying to text messages and the number
of interactive messages received predicts weight loss in participants who received a 12-month
texting intervention. Our findings indicate that participants who have higher levels of
engagement with the texting system are more successful in losing weight. Future mobile
health interventions should include measures to monitor participant engagement during the
intervention and to include as a dosing variable when assessing intervention effects.

CORRESPONDING AUTHOR: Margaret A. Crawford, BS, UC San Diego, Encinitas, CA, 92024;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2561

D103 10:00 AM-11:00 AM

EVALUATION OF PRO MUJER, A FACEBOOK-BASED WORKSITE WELLNESS PROGRAM IN LATIN

AMERICA

Erica Levine, MPH1, Sandy Askew, MPH2, Gary G. Bennett, PhD2

1
Duke Global Health Institute, Durham, NC; 2Duke University, Durham, NC

Title: Evaluation of a Facebook-based Worksite Wellness Program in Latin America

Purpose: Chronic conditions account for 68% of mortality in Latin America. This is largely
attributed to increasing obesity, and adoption of a Westernized diet along with
inactivity. Latin America is also the fastest growing social media market; 94% of Internet users
in the region are on social media. Pro Mujer is a womens development organization that
offers financial services and healthcare in 5 countries in Latin America. In 2014, Pro Mujer
launched a Facebook page, Pro Mujer Salud, to help its employees adopt healthy eating and
physical activity habits. The page now has over 5,000 followers.

Methods: We conducted a mixed-methods evaluation to assess the relationship between the

Facebook page, employees self-reported adoption of healthy eating and physical activity, and
on social norms promoted by the page. The evaluation consisted of 2 focus groups (n=16) and
3 key informant interviews in Nicaragua, and a survey to all employees in Bolivia, Argentina,
Peru, Nicaragua, Mexico and the United States.

Results: Survey respondents (n=728; 36% of staff) were mostly female (73%), with a mean age
of 33 years and mean BMI of 26.7 4.7 kg/m2. The majority (80%) of employees with
Facebook follow Pro Mujer Salud; most followers (76%) interact with the page by clicking
"like" on posts or photos. After adjusting for BMI, country and education, there were
marginally statistically significant relationships between the length of time an employee
followed Pro Mujer Salud and the odds of reporting more exercise (OR [95% CI]: 1.16, [1.01,
1.34], p= .04), as well as eating healthier (OR [95% CI]: 1.15, [1.001, 1.32], p= .048). The
perception that coworkers were exercising more was a strong predictor of diet and exercise
change. For each increase in amount of agreement that coworkers exercise more as a function
of a Facebook page campaign called Sbado Saludable (Healthy Saturday), the odds of also
reporting more exercise nearly doubled (OR [95% CI]: 1.92 [1.45, 2.56], p< .0001), and the
odds of reporting diet related behavior change increased by 79% (OR [95% CI]: 1.79 [1.36,
2.34], p< .0001). Qualitative data analysis revealed high trust in the health information on the
Pro Mujer Salud page. It also revealed themes of increasing anxiety about a changing food
S2562 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

environment (increasing processed foods and a loss of connection to the local food landscape)
and low confidence in the ability to achieve recommended levels physical activity.

Discussion: This is the first evaluation of a Facebook-based worksite wellness program in Latin
America. This highly scalable intervention holds great promise for being able to influence
social norms, promote behavior change, and support a workplace culture of health.

CORRESPONDING AUTHOR: Erica Levine, MPH, Duke Global Health Institute, Durham, NC,
27708; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2563

D104 10:00 AM-11:00 AM

CITATION AWARD WINNER

EVIDENCE-BASED NOVELTY IN THE DELIVERY OF ONLINE CBT: THE IMPACT OF USER-DRIVEN

PROGRESSION THROUGH CBT CONCEPTS AND TECHNIQUES

Megan Jones Bell, PsyD1, Clare Purvis, PsyD2, Tess Brown, BA2

1
Lantern / Stanford University, San Francisco, CA; 2Lantern, San Francisco, CA

The efficacy of digital CBT-based interventions for common mental health disorders is clearly
demonstrated, with many studies showing that digital interventions are equally effective to
face-to-face treatment. Despite this, digital interventions are plagued by low engagement and
poor user retention; fully 50% of people using a digital health tool discontinue use within six
months. One contributing factor may be that digital health interventions have traditionally
followed fixed protocols and, as a result, have been unable to account for individual user
characteristics, situations, and needs - a mandate of effective evidence-based practice. We
propose that digital interventions can address this limitation by providing users a more
personalized experience. For example, by adopting a more flexible design, digital
interventions can allow users to select the content most relevant to their current concerns
and practice skills at a self-determined pace. This approach mirrors a flexibility within
fidelity approach to implementing empirically-supported treatments in face-to-face settings.

The present study investigates the effect of specific changes to the protocol and design of a
mobile app-delivered, guided CBT program, Lantern, on user engagement, retention, and
clinical outcomes. Two methods for delivering digital CBT were compared: a traditional
protocol in which CBT concepts and skills are delivered following a fixed sequence and pace
(standard CBT), versus an innovative delivery in which users determine the sequence and pace
at which they progress through CBT (nonlinear CBT). The same concepts and skills are
delivered across both conditions; users in both conditions also receive individual, in-app, text-
message based coaching.

Data from two ongoing deployments of Lantern with N=1000 adults (500 per condition) in
employer settings will be collected from January, 2016 to January, 2017 in quarterly cohorts.
All users have provided informed consent to data being used for research and product
development purposes. The first three cohorts of the N=1000 target have already been
recruited and are actively enrolled in Lantern. Outcome measures to be reported include
S2564 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

engagement (completion of sessions, techniques, coach messaging), retention, primary

symptom reduction (anxiety and/or stress). The study highlights the impact of an innovative,
evidence-based delivery of digital CBT.

CORRESPONDING AUTHOR: Megan Jones Bell, PsyD, Lantern / Stanford University, San
Francisco, CA, 94105; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2565

D105 10:00 AM-11:00 AM

FORMATIVE EVALUATION OF A TEXT MESSAGE-BASED INTERVENTION FOR ABNORMAL PAP

SMEAR FOLLOW-UP CARE TARGETING UNDERSERVED WOMEN

Erin K. Tagai, PhD, MPH1, Suzanne M. Miller, PhD2, Kuang-Yi Wen, PhD3, Enrique Hernandez,
MD, FACOG, FACS4

1
Fox Chase Cancer Center, Philadelphia, PA; 2Fox Chase Cancer Center/Temple University
Health System, Philadelphia, PA; 3Fox Chase Cancer Center, Philly, PA; 4Lewis Katz School of
Medicine at Temple University, Philadelphia, PA

Women of low socioeconomic status (SES) have poorer adherence to follow-up care after an
abnormal Pap smear. Text messaging is widely used among low SES populations and has been
reported as a feasible method to deliver health behavior messages. A formative evaluation
was conducted to assess the receptivity of a text message-based intervention guided by the
Cognitive-Social Health Processing Model (C-SHIP) for women of low SES to increase
adherence to abnormal Pap smear follow-up care. Women (N=28) attending a colposcopy
appointment for an abnormal Pap smear were recruited to complete a short demographic
survey and qualitative interview. Interview questions sought to identify participants interest
in a text message-based intervention and psychosocial factors. Participants evaluated 25
prototype counseling messages previously used in a telephone-based counseling intervention
and targeted to psychosocial barriers to adherence, with content guided by C-SHIP. The
messages were evaluated on a scale of 1 to 5 for personal helpfulness, understandability, and
culturally appropriateness. Participants were predominately African American (68%) and
Latino (21%), and had a household income below $15,000 (71%). Participants expressed
strong interest in receiving counseling messages to be delivered via text (68%). Most
participants (68%) felt nervous or worried before their follow-up appointment. However,
participants reported colposcopies as helpful (96%) and important (89%). Evaluations of
prototype messages indicated high favorability with mean scores of 4.7 (SD=0.4) for personal
helpfulness, 4.8 (SD=0.4) for understandability, and 4.6 (SD=0.6) for cultural appropriateness.
Women of low SES undergoing follow-up care for an abnormal Pap smear reported interest in
a text messaging-based intervention to increase adherence to follow-up care. High message
favorability indicates prototype messages guided by C-SHIP are appropriate for the target
population. Findings will be used to develop a text message-based intervention to increase
adherence to abnormal Pap smear follow-up.

CORRESPONDING AUTHOR: Erin K. Tagai, PhD, MPH, Fox Chase Cancer Center, Philadelphia,
PA, 19111; [email protected]
S2566 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D106 10:00 AM-11:00 AM

INSPIRATION OR THINSPIRATION: THE ASSOCIATION AMONG PROBLEMATIC INTERNET USE,

EXERCISE DEPENDENCE, AND EATING DISORDER RISK

Brian Cook, PhD1, Danika Quesnel, BHk, MSc 2, Kellen Murray, .1, Jessica Zamudio, .1

1
California State University Monterey Bay, Seaside, CA; 2University of British Columbia,
Winnipeg, MB, Canada

Introduction: Internet use provides the opportunity to access helpful information about
health behaviors, but excessive use of the Internet has been associated with engaging in
pathological health behaviors. Specifically, engagement and use of online social media
platforms may negatively impact health by promoting unhealthy behaviors and providing
inappropriate recommendations aimed at manipulating ones body shape and size.
Problematic Internet Use (PIU) has been associated with eating disorder (ED) risk factors and
pathological patterns of exercise (e.g., Exercise Dependence, EXD). Similarly, previous
research has demonstrated an association between EXD with increased risk of developing an
ED. However, no study to date has examined the co-occurrence among PIU, EXD, and ED risk
simultaneously. The purpose of this study was to examine the co-occurrence of PIU and EXD
in ED risk in a population-based sample of women. We hypothesize that PIU and EXD scores
will be higher in individuals with ED risk. Method: Participants were 898 women [M age =
27.61(10.22); 65.02% Caucasian; M Body Mass Index = 26.87(6.36)] who completed the
Exercise Dependence Scale, Eating Disorders Examination-Questionnaire (EDE-Q), and
Problematic Internet Use Questionnaire as part of a larger online survey study. First,
participants were grouped by ED risk status. PIU and EXD categories were also determined
using previously published algorithms. Next, differences in continuous PIU and EXD scores by
ED risk were assessed. Finally, Chi-square analyses were used to examine differences in PIU
and EXD risk categories were examined by ED risk status. Results: Overall, 14.32% reported
PIU, 1.96% reported EXD, and 35.07% reported EDE-Q scores > 2.3 indicating ED risk.
Independent samples t tests revealed significantly higher EXD [t(637) = -4.79, p=.01] and PIU
[t(673) = -8.40, p < .01] scores in individuals at-risk for ED. Chi-square analyses revealed
significant differneces in PIU [X2(2)=33.81, p.01] and EXD risk status by ED risk (X2(2)=8.71,
p=.01]. Conclusion: Excessive use of the Internet may allow for misinformation regarding
body shape satisfaction that encourages pathological health behaviors related to exercise and
eating. The main clinical implication of this study is that health professionals should be aware
of potential overlapping symptoms and consequences of Internet use, exercise, and ED
behaviors. The results also suggest a common factor may explain increased use of the
internet, exercise, and ED behaviors. Further research is needed to elucidate such factors.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2567

CORRESPONDING AUTHOR: Brian Cook, PhD, California State University Monterey Bay,
Seaside, CA, 93955; [email protected]
S2568 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D107 10:00 AM-11:00 AM

INTEGRATION OF PEER SUPPORT AND DIGITAL HEALTH: CONSIDERATIONS FOR CHRONIC

DISEASE MANAGEMENT INTERVENTIONS

Sarah Kowitt, MPH1, Rebeccah L. Sokol, B.S. 2, Nivedita Bhushan, MA3, Patrick Y. Tang, MPH4,
Edwin Fisher, PhD5

1
University of North Carolina at Chapel Hill, Gillings School of Global Public Health, Chapel Hill,
NC; 2Gillings School of Public Health, UNC Chapel Hill, Carrboro, NC; 3Gillings School of Global
Public Health - University of North Carolina at Chapel Hill, Chapel Hill, NC; 4Peers for Progress,
Gillings School of Global Public Health, UNC Chapel Hill, Durham, NC; 5The University of North
Carolina at Chapel Hill, Gillings School of Global Public Heealth, Chapel Hill, NC

Peer support and digital health are often used independently to address chronic disease
management barriers. Few studies have examined their integration in real-world settings. In
this study, we evaluated how integration occurred in the context of a feasibility trial that
combined health coaching and an app for patients with diabetes in a primary care practice.
Semi-structured interviews were conducted with 12 patients, 2 health coaches, and 5 project
staff (e.g., care coordinator, program manager) affiliated with the program. Coding of
transcripts used deductive and inductive codes and consensus among three coders to ensure
accuracy. Integration of the coaching and the app occurred in several domains. In terms of
organization of care, the health coaching and diabetes app provided distinctive, but
complementary roles for diabetes management; most patients reported using both program
components, while a minority reported only using one. A common theme of this integration
was that patients saw the diabetes management app as providing instrumental and
informational support to participants, whereas they saw the health coaching as providing
emotional and appraisal support. Both patients and coaches valued reports of blood glucose
measures and other indicators as anchoring their discussions. Further integration with the
electronic health records and general clinical care remain important opportunities. Lastly,
study staff discussed the potential role of digital health and peer support to increase
efficiency and reach of diabetes management interventions, carefully weighed with the need
for more resources required by such integration. These findings indicate that peer support
and digital health address distinctive, unique roles in diabetes management support and that
their integration may benefit future chronic disease care in diabetes and other areas.

CORRESPONDING AUTHOR: Sarah Kowitt, MPH, University of North Carolina at Chapel Hill,
Gillings School of Global Public Health, Chapel Hill, NC, 27516; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2569

D108 10:00 AM-11:00 AM

MIXED METHOD EXPLORATION OF USER ENGAGEMENT WITH A MOBILE APP-SUPPORTED

SMOKING CESSATION ATTEMPT: PRELIMINARY FINDINGS

Christian Cerrada, BS1, Kathleen Colao, BS, MPH2, Mojgan Sami, PhD3, Jimi Huh, PhD1

1
University of Southern California, Los Angeles, CA; 2University of Southern California Keck
School of Medicine, Venice, CA; 3University of California, Irvine, Irvine, CA

Introduction: Despite rapid development in the area of mobile health (mHealth), there
remains a dearth of research on user engagement, the specific ways and extent to which users
interact with a system over time. Although studies generally suggest that increased user
engagement leads to better behavior change outcomes, little is known about the variety of
patterns in which users might interact with mHealth apps to support behavior change. The
present study seeks to explore different user engagement patterns with a mobile smoking
cessation app, MyQuit USC (MQU), over the course of four weeks using a mixed methods
approach.

Methods: Data came from a preliminary subset of 18-25 year old Korean American smokers
(N=16) participating in an ongoing smoking cessation study. To explore patterns of
engagement with MQU, individual user interaction logs of the app during the study period
were used to generate engagement plots detailing number and type of app interactions each
day. Semi-structured interviews were conducted on the same participants shortly after
completing the study. Interview data were analyzed deductively using pre-identified codes
corresponding to different dimensions of user engagement of interest to researchers.

Results: Our data revealed notable heterogeneity in user engagement patterns and suggested
potential influential factors on different patterns. Individual engagement plots indicated
varying patterns differentiated by average number of daily app interactions over time
(min=1.1, max=11.2) and extent of daily variation in interaction volume (SD min=0.8,
max=6.1). Interview data shed light on several factors that appear to influence user
engagement patterns. App-related factors included novelty of intervention content and
whether a given interaction with MQU was user- or app-initiated. User-related factors
included users perceived role of MQU in their quit attempt, cessation progress and success,
and momentary mood.

Conclusions: In combination, our data shed light on diverse patterns of engagement with
MQU and factors that may influence an app-supported quit attempt. To improve future
S2570 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

versions of MQU, intervention content should be responsive to dynamic app- and user-related
factors, e.g. tailoring intervention contents to patterns of engagement or behavior change
success at a previous time point.

CORRESPONDING AUTHOR: Christian Cerrada, BS, University of Southern California, Los

Angeles, CA, 90014; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2571

D109 10:00 AM-11:00 AM

PREFERENCES IN WEIGHT LOSS APP FEATURES AMONG AFRICAN AMERICANS

Martinus Evans, B.S., M.S.1, Christine May, PhD2, Molly E. Waring, PhD1, Danielle E. Jake-
Schoffman, PhD3, Amanda Okpoebo, BS4, Sherry Pagoto, PhD1

1
University of Massachusetts Medical School, Worcester, MA; 2University of Massachusetts
Medical School, Fiskdale, MA; 3University of Massachusetts Medical School, Worcest, MA;
4
UMass Medical School- Preventive Behavioral Medicine, Leominster, MA

More than 75% of African Americans are overweight or obese. Research has shown that
culturally tailored interventions may be more effective in improving dietary habits and
increasing physical activity in African Americans, yet treatment models that require frequent
in-person visits have poor feasibility, especially among low-income African Americans. 56% of
African Americans own a smartphone, suggesting that a culturally-tailored weight loss mobile
app may be an effective way to reach this population with weight loss programming. This
study examined commercial weight loss apps used by African Americans, features they find
most and least useful, and features they would find most relevant. We conducted 4 in-depth
semi-structured interviews and 2 focus group discussions (N=10). We content analyzed
resulting open-ended question using an inductive method with pair coders. Participants self-
identified as African American, smart phone users, on average were 35 years old (SD=6.0), and
80% were college-educated. MyFitnessPal was the most commonly used app (kappa=1.0),
reported by 5 participants. Other apps included S Health (n=3), Fitbit (n=3), Weight Guru
(n=1), C25k (n=1), Apple Health (n=1), and LoseIt! (n=1). Of the apps they have used the
features they found most useful (kappa=.83) were tracking (n=4), auto-syncing to other
devices/apps (n=2), feedback about tracked data (n=2), ease of use (n=1), exercise suggestions
(n=1), diet suggestions (n=1), community (n=1) and calculate activity into calories burned
(n=1). Least useful app features (kappa=1.0) included missing relevant/ethnic foods (n=3) and
tracking is time consuming (n=3), issues with delayed syncing of steps and weight (n=1),
difficulty setting helpful calorie goal (n=1), notifications of progress w/o context (n=1), and ads
on free versions of apps (n=1). The feature requested for a culturally-tailored weight loss app
by the most participants (kappa=.91) was nutrition education including healthy alternatives to
ethnic food (n=4), stress management (n=2), music w/ exercise prompts (n=2), exercise
suggestions (n=2), community chat (n=1), voice to text (n=1), and diversity in visual
representation (n=1). Further research should examine which features are wanted by a larger
population of African Americans.
S2572 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Martinus Evans, B.S., M.S., University of Massachusetts Medical

School, Worcester, MA, 01655; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2573

D110 10:00 AM-11:00 AM

PROFILING APP USERS FROM NUTRITION, WEIGHT MANAGEMENT, AND FITNESS-RELATED

WEBSITES

Steriani Elavsky, Ph.D.1, David mahel, Ph.D.1, Hana Machackova, Dr.2

1
Faculty of Social Sciences, Masaryk University, Brno, Jihomoravsky kraj, Czech Republic;
2
Masaryk University, Brno, Jihomoravsky kraj, Czech Republic

The use of online communities and websites for health information has proliferated along
with the use of mobile apps for managing health behaviors such as diet and exercise. The
evidence available to date suggests that users of these websites and apps differ in significant
ways from non-users but most data come from US and UK-based populations. In this study we
recruited users of nutrition, weight management and fitness-oriented websites in the Czech
Republic to better understand their motives, expectations, and usage patterns with respect to
app use. The online survey was initiated by 1096 users and data were analyzed for
respondents aged 13-39 who provided information on app use (n=719; M(SD) age=24.0(5.2);
84%female). Respondents using apps for managing nutrition, weight and fitness (n=420, 58%)
were more likely to be female (OR 2.111, 95% CI 1.3643.268) and reported more expert
telephone skills (OR 1.450, 95% CI 1.3051.611). Among app users, male users reported less
frequent use of diet-oriented apps but more frequent use of weight-gaining apps as did
younger users (13-18yrs). Underweight app users reported more frequent use of weight loss
apps while obese users reported less frequent use of exercise-related apps (all p < .05). There
was a positive relationship between app use and drive for thinness (EDI-3), self-reported
excessive exercise (EPSI), internalizing, and social support, and for neuroticism with weight
loss apps and conscientiousness with exercise-related apps, respectively (all p < .05).
Significant differences were also found based on age, BMI, and gender in the frequency of use
and value of specific app functions such as those for monitoring, planning, comparing with
others, and communicating with others, along with differential impact of app use on self-
reported outcomes. These data unveil the motives and expectations of nutrition, weight
management and fitness website users with respect to smartphone apps, helping inform
subsequent development of mobile intervention strategies in this non-US based select
population.

CORRESPONDING AUTHOR: Steriani Elavsky, Ph.D., Faculty of Social Sciences, Masaryk

University, Brno, Jihomoravsky kraj, 602 00; [email protected]
S2574 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D111 10:00 AM-11:00 AM

REDUCING THE PATIENT INTERVAL FOR BREAST CANCER: THE DEVELOPMENT OF A DIGITAL
INTERVENTION USING THE PERSON-BASED APPROACH

Emma M. Carr, BA (Hons), HDip (Psych), MSc (HlthPsych) 1, Jane C. Walsh, BA, PhD2, AnnMarie
Groarke, BA, MA, HDipED, PhD, AFPsSI, CPsychol1

1
School of Psychology, National University of Ireland, Galway, Galway, Galway, Ireland;
2
National University of Ireland, Galway, Galwau, Galway, Ireland

Background

Early detection of breast cancer increases survival rates. The key to improving early detection
is decreasing the Patient Interval. This refers to (i) the time taken to interpret a bodily change
and (ii) the time taken to act upon that change and seek help from a health care professional
(HCP). The most successful interventions designed to decrease the Patient Interval to date
have been delivered by a HCP in a clinic setting. HCP delivered interventions however are
costly and time intensive. A novel way to decrease the Patient Interval is through a digital
intervention (DI). Research to date has established that DIs are a feasible and acceptable way
to deliver healthcare interventions. Furthermore, interventions incorporating public and
patient involvement throughout the design process are more successful. This study
represents phase one of the Person-Based Approach (PBA) to intervention development. This
approach is meant to complement the traditional evidence- and theory-based approaches to
intervention development while including end-user input at multiple stages of design process.

Objectives

The objective of this study is to complete phase one of the PBA by eliciting feedback from
potential end-users of the (currently) hypothetical DI that will then be used to shape the
design of the DI. The study aimed to facilitate broad discussion about the acceptability of a
hypothetical DI with end-users.

Research Design

A qualitative design was used to collect data via focus group interviews which were audio-
recorded, transcribed verbatim and thematically analysed. Participants (n=28) were sampled
purposively based on age, sex and occupation and were either individuals who would
potentially use the DI i)for their own healthcare purposes or ii) in an official capacity to
recommend to others e.g. Doctors.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2575

Results

Five themes emerged: uncertainty, interest, legitimacy, aesthetics/usability, efficiency.

Participants expressed uncertainty that a website could help them manage their health but
they were interested in trying it. Participants highlighted time as a primary barrier to using a
website for this purpose. It was stressed that the website would have to be aesthetically
pleasing and easy to use. Furthermore is was discussed that it should be obvious that it was
provided by a reputable source e.g. a university/hospital so that they could trust the
information provided.

Conclusion

Participants were eager and motivated to contribute to research at this early stage. It is now
accepted as good practice to elicit end-user/patient views at the design phase of
interventions. It is hoped that incorporating the views of theses participants will enhance the
acceptability and therefore success of the DI.

Future Plans

These results will be combined with a literature review of both qualitative and quantitative
research to allow the research team to design the DI.

CORRESPONDING AUTHOR: Emma M. Carr, BA (Hons), HDip (Psych), MSc (HlthPsych) , School
of Psychology, National University of Ireland, Galway, Galway, Galway, 0000;
[email protected]
S2576 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D113 10:00 AM-11:00 AM

ENHANCING CARE OF THE AGING AND DYING IN PRISONS (ECAD-P): USABILITY OUTCOMES
FOR AN END-OF-LIFE TECHNOLOGY TRAINING IN PRISONS

Valerie H. Myers, PhD1, Susan J. Loeb, PhD, RN, FGSA, FAAN2, Brenda L. Baney, MS2, Sophia
M. Strickfaden, BA1, Janice Penrod, PhD, RN, FGSA, FAAN2

1
Klein Buendel, Inc., Golden, CO; 2Pennsylvania State University, University Park, PA

About 18% of U.S prison inmates are 50 years or older. The growth of the older inmate
population increases burden of care on prisons. Inmates are significantly more likely to have
chronic conditions or disability, and despite advances in the free world, best practices for
managing geriatric issues and end-of-life care (EOL) have not be adapted for use in corrections
settings. Geriatric and EOL care training via technology has had a marked impact on
disseminating best practices. However, these technology approaches have been inaccessible
within prison walls due to the need to maintain care, custody, and control. Enhancing Care of
the Aging and Dying in Prisons (ECAD-P; R41AG049570) was a Phase I Small Business
Technology Transfer (STTR) project that examined the acceptability and feasibility of
developing a media-rich, technology-based program for training prison staff on geriatric and
EOL care. Previously developed and tested empirical materials for this setting and population
were adapted using instructional designapproaches and then programmed into an interactive,
web-based prototype. The prototype was evaluated by multidisciplinary prison staff on its
user interface, ease-of-use, perceived barriers, and for understanding and optimizing the
organizational culture for implementation. ECAD-P contained three modules each with
multiple interactive features (e.g., drag and drop, hover, click and reveal, video, etc.) and a
quiz. Usability and acceptability were tested to identify and rate the severity of potential
problems (n=16) with staff from two state correctional institutions (SCIs). Actions, eye
movements, utterances, and problems encountered were observed, and the System Usability
Scale (SUS), a validated tool for assessing the usability and acceptability of technology-based
products, served as the criterion measure.The SUS score was 87.8% (68% is above average)
suggesting a high level of acceptability and usability for visual design, functionality, and
content. Fishers exact and Wilcoxon Mann-Whitney tests were conducted to assess whether
there were any differences in usability outcomes between the two SCI facilities. Analyses
revealed non-significant findings (ps > .05) suggesting there were no significant differences
among usability testers regardless of their SCI affiliation.Additional usability results, the design
and development process for creating ECAD-P, and programming updates on the recently
funded Phase II project will be presented.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2577

CORRESPONDING AUTHOR: Valerie H. Myers, PhD, Klein Buendel, Inc., Golden, CO, 80401;
[email protected]
S2578 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D114 10:00 AM-11:00 AM

INTEGRATING THREE EVIDENCE-BASED PROGRAMS TO CONSTRUCT A RURAL COMMUNITY-

BASED CARDIOVASCULAR DISEASE PREVENTION PROGRAM

Rebecca Seguin, PhD, CSCS1, Meredith Graham, MS1, Lynn Paul, Ed.D., RDN2, Anna
Diffenderfer, MS, RDN, LN3, Emily Morgan, PhD, MPH1, Galen Eldridge, M.S.2, Miriam Nelson,
Ph.D.4, Sara Folta, PhD5

1
Cornell University, Ithaca, NY; 2Montana State University Extension, Bozeman, MT; 3Montana
State University, Department of Health and Human Development, Bozeman, MT; 4University
of New Hampshire, Durham, MA; 5Tufts University, Boston, MA

BACKGROUND: Strong Hearts, Healthy Communities (SHHC) is a lifestyle modification

program that aims to reduce body weight, improve dietary habits, and promote physical
activity among overweight/obese women aged 40 and older.

METHODS: Program development included in-depth community assessments and integration

of core concepts from three nationally disseminated evidence-based curricula. In 10 medically
underserved rural towns in Montana, 17 focus groups were held with 126 sedentary,
overweight/obese adults aged 40-91 years; 30 key informant interviews were conducted with
community leaders, educators, and healthcare providers; 10 community assessments were
conducted by local educators and researchers. The focus group and key informant topic
guides focused on barriers and enablers to health (nutrition, physical activity, and tobacco)
and access to healthcare services. Findings from the community assessments and qualitative
data informed the integration of three evidence-based programs to create the SHHC
curriculum.

RESULTS: The resulting SHHC is a 24-week program with twice weekly hour-long classes, plus
six out of class community-focused sessions. SHHC components include progressive strength
training, aerobic exercise, nutrition education, and civic engagement. The program has 13
evidence-based behavioral aims and is divided into three phases: (1) Get moving and eat well
(Weeks 1-8); (2) Assess individual progress and create supportive community (Weeks 9-16);
(3) Maintenance strategies and social support for healthy lifestyle (Weeks 17-24).

CONCLUSION: Integrating formative data with multiple evidence-based curricula into a

comprehensive, tailored program presents unique opportunities and challenges. Based upon
preliminary feedback, future implementation should consider the timing and framing of
content as well as dose adjustments.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2579

CORRESPONDING AUTHOR: Rebecca Seguin, PhD, CSCS, Cornell University, Ithaca, NY, 14853;
[email protected]
S2580 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D115 10:00 AM-11:00 AM

NURSE COORDINATOR ADOPTION OF AN ADAPTED, EVIDENCE-BASED WEIGHT LOSS

PROGRAM: A HYBRID EFFECTIVENESS-IMPLEMENTATION TRIAL

Paul A. Estabrooks, PhD1, Sarah Johnson, MCHES2, Fabio Almeida, PhD1, Anne Jessie, DNP,
RN3, Deirdre Rea, MSN, RN3, Wanda Bowman, RN3, Caroline Butt, RN3, April Vaughan, RN3,
Mark Greenawald, MD3, Sarah Frazier, RN2

1
University of Nebraska Medical Center, Omaha, NE; 2Virginia Tech/Carilion Clinic, Roanoke,
VA; 3Carilion Clinic, Roanoke, VA

The purpose of this study was to determine, if, when compared to a continuing medical
education (CME), a consultee-centered implementation strategy (IS) with goal setting and
feedback would improve the reach, effectiveness, adoption, implementation, and
sustainability of an adapted evidence-based weight loss intervention that included 20 patient
sessions over 12 months (Healthy Lifestyles). The IS included a 2.5 hour CME workshop and a
package of ready-to-deliver Healthy Lifestyles resources but added ongoing consultation at 1-
3-6, and 12-months to address patient recruitment and engagement in Healthy Lifestyles. In
addition, electronic medical record smart-phrases were used facilitate and document Healthy
Lifestyle session completion, fidelity, and costs. All nurse care coordinators in a regional
health care system participated (n=45) and 14 nurses in one region received the IS. Reach did
not significantly differ over 12 months (n=1421.9/CME nurse; n=2431.4/IS nurse), but IS
trained nurses were significantly more likely to adopt Healthy Lifestyles (100% vs 61%, p=5%
weight loss was higher for patients that received the intervention from IS trained nurses (14%
CME; 19% IS, p < .05). By 12 months there were no differences in weight loss (-2.15.9 CME; -
2.46.5 IS) or the proportion reaching a 5% weight loss (14% CME; 16% IS). Consultation and
action planning may accelerate adoption and organizational maintenance of evidence-based
weight loss programs compared to a CME approach while supporting a higher proportion of
patients to achieve a clinically meaningful weight loss, though a focus on weight loss
maintenance in future trainings appears warranted.

CORRESPONDING AUTHOR: Paul A. Estabrooks, PhD, University of Nebraska Medical Center,

Omaha, NE, 68198-4365; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2581

D116 10:00 AM-11:00 AM

RETURNING AGGREGATE RESEARCH RESULTS TO PARENTS OF YOUTH IN A DISEASE REGISTRY:

ADVANCING PATIENT-CENTERED OUTCOMES RESEARCH

Elissa R. Weitzman, ScD, MSc1, Kara M. Magane, MS2, Lauren E. Wisk, PhD1

1
Boston Children's Hospital and Harvard Medical School, Boston, MA; 2Division of
Adolescent/Young Adult Medicine, Boston Children's Hospital, Portsmouth, NH

Background. To improve the evidence base about pediatric onset chronic disease, sustainable
approaches are needed for engaging cohorts in providing patient-reported outcomes (PROs)
about child wellbeing, disease and treatment experiences. Returning research results has
been touted as a means for contextualizing healthcare decision-making and motivating
research engagement, areas of top national interest. We tested whether returning aggregate
research results (ARR) to parents of children with rheumatic disease (RD) would affect interest
in research participation and spur action on behalf of a child.

Methods. Parents of children enroled in a national clinical rheumatic disease registry provided
validated PROs about child quality of life, disease and treatment experiences, side effects and
problems (response rate: 69.2%). Subsequently, parents viewed ARR in the form of a slide-
deck of curated, annotated data about cohort patterns of these measures and then reported
their reactions and the perceived value of this information. We used principal components
analysis (PCA) to analyze reaction patterns elicited by ARR and multivariate logistic regression
to model the effect of reaction patterns on interest in future research participation and
planned next steps.

Results. Among 196 participating parents, 107 (54.6%) received ARR; 89 did not. Groups did
not differ on measures of child socio-demographics, clinical characteristics, or disease and
treatment PROs. Parents perceived ARR as very valuable for understanding their childs
condition and making care decisions; 76.6%, 57% and 41% reported ARR to be "very valuable"
for understanding medication side effects/problems, pain interference, and morning stiffness,
respectively. PCA identified two ARR reaction domains from 6 questions:
Affirmation/Validation (AV) of own experiences and Information Burden (IB). Overall, 54.2% of
parents reported being more interested (versus less or unchanged) in participating in future
research after seeing ARR. Parents with higher AV scores had twice the odds of reporting
more interest in future research compared to those with lower AV (AOR=2.04; 95%CI:1.20-
3.50). IB scores were related to planned actions: Compared to parents with lower IB scores,
those with higher IB were less likely to discuss ARR with their child (AOR=0.60; 95 CI:0.36-
0.98), and more likely to discuss ARR with providers (AOR=1.81; 95%CI:1.04-3.16).
S2582 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusion. Returning ARR to cohorts may foster a virtuous cycle of sustainable research
engagement to advance patient reported outcomes research and affect planned actions to
share and discuss ARR in support of a childs chronic disease care and treatment via a model
with high translation potential across pediatric chronic disease.

CORRESPONDING AUTHOR: Elissa R. Weitzman, ScD, MSc, Boston Children's Hospital and
Harvard Medical School, Boston, MA, 02115; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2583

D117 10:00 AM-11:00 AM

SETTING THE STAGE FOR ENGAGING COMMUNITY IN EVIDENCE-BASED PROGRAMS TO

REDUCE TYPE 2 DIABETES

Nicole Vaughn, PhD1, Crystal Wyatt, N/A1, Beatriz O. Reyes, DrPH2, Kimberly Kalupa, PhD3,
Kimberly Arnold, MPH4, Laura Hunter, MPH5, Longjian Liu, MD, PhD, MSc, FAHA1

1
Drexel University School of Public Health, Philadelphia, PA; 2Northwestern University,
Evanston, IL; 3University of New Mexico, Albuquerque, NM; 4Johns Hopkins University,
Baltimore, MD; 5GlaxoSmithKline, Philadelphia, PA

Diabetes is one of the top 10 leading cause of death in urban settings such as Philadelphia.
Unfortunately, African Americans have the highest rate of diagnosed diabetes. Thus, there
remains a critical role for understanding how community, academic and industry partnerships
can operate to engage with and offer evidence-based programs that align with a community's
need to reduce health disparities, particularly for minority and high-need populations. The
Getting People in Sync (GPS) study worked with lay facilitators to implement a diabetes
prevention program. Implementation strategies of the GPS study across the partners included
a multi-step process involving church identification, memorandums of understanding, a
detailed communication plan, non-disclosure agreement, as well as facilitator and data
training sessions. While this program demonstrated that intervention participants could lose
weight (-7.96 lbs., SD=10.62; t(56)=-5.66 p < .000), aspects of the partnership were critical to
the successful implementation in the community. Lessons learned from this multi-sectorial
partnership included communication, connectivity, seasonality and training needs. Additional
considerations for implementation included flexibility in timelines for legal documentation. In
conclusion, community, academic and industry partnerships can be models for
implementation of quality evidence-based programs that have measurable and clinically
meaningful outcomes however, more information discussing successes and challenges across
these types of partnerships is needed.

CORRESPONDING AUTHOR: Nicole Vaughn, PhD, Drexel University School of Public Health,
Philadelphia, PA, 19104; [email protected]
S2584 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D118 10:00 AM-11:00 AM

USING CBPR TO UNDERSTAND LOCAL COLLABORATIONS ADDRESSING CVD RISK

Phenesse Dunlap, MPH1, Leah Frerichs, PhD2, Brenda Boone, .3, Shirley McFarlin, BA4, Giselle
Corbie-Smith, MD, MSc1

1
UNC Center for Health Equity Research, Chapel Hill, NC; 2UNC Department of Health Policy
and Management, School of Public Health, Chapel Hill, NC; 3James McFarlin Community
Development, Rocky Mount, NC; 4James Mcfarlin Community Development, Inc., Rocky
Mount, NC

Introduction: Rural African Americans suffer from considerable cardiovascular disease (CVD)
disparities that no single agency can address alone. Inter-organizational networks and
collaboration can impact a communitys capacity to address CVD by influencing activities such
as coordination of services, diffusion of information, and adoption of evidence-based
interventions. In rural African American communities, grassroots community-based
organizations (CBOs) and faith-based organizations (FBOs) potentially play a critical role in
these activities; yet, we have limited understanding how they collaborate. The objective of
our study is to describe the inter-organizational networks among CBOs and FBOs in a rural,
African American community and explore factors that influence their collaboration around
CVD-related activities. Methods: We used a mixed-methods CBPR approach to develop a
survey and in-depth interview guide using existing network and collaboration assessment
tools. We used respondent-driven sampling to establish a census of organizations currently
collaborating on CVD-related activities and services. We completed surveys with an initial
seed sample of n=53 organizations and are currently in the process of conducting additional
surveys and interviews to obtain a more comprehensive understanding of collaboration.
Results: About half (n=25, 49%) of the seed sample of fifty-three CBOs and FBOs are currently
collaborating with at least one other organization on CVD-related activities. They most
commonly indicated collaboration with one of 11 healthcare service agencies (e.g., hospitals,
county health departments). Few (n=4) indicated collaboration with other CBOs/FBOs, and
preliminary results indicate current network density for CVD collaboration is low, 0.06. A non-
profit healthcare system and a community health center had the highest betweenness
centralities (i.e., are key connectors between other organizations in the community).
Conclusions: This CBPR approach was successful in identifying inter-organizational networks
of CBOs and FBOs and will improve understanding of factors that influence collaboration. We
will use the findings to identify strategies for creating stronger partnerships and will continue
to collect information over time in order to understand how networks influence CVD services
in the community such as adoption of evidence-based CVD interventions.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2585

CORRESPONDING AUTHOR: Phenesse Dunlap, MPH, UNC Center for Health Equity Research,
Chapel Hill, NC, 27599; [email protected]
S2586 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D119 10:00 AM-11:00 AM

CONCEPTUALIZATIONS OF PRIVACY AMONG EARLY ADOPTERS OF EMERGING HEALTH

TECHNOLOGIES

Cynthia Cheung, MPH/MA (c)1, Matthew J. Bietz, Ph.D.2, Kevin Patrick, MD, MS3, Cinnamon S.
Bloss, Ph.D.1

1
University of California - San Diego, La Jolla, CA; 2University of California, Irvine, Irvine, CA;
3
UCSD, La Jolla, CA

Introduction: Advances in health technology such as genome sequencing and wearable

sensors now allow for the collection of highly granular personal health data from individuals.
It is unclear how people think about privacy in the context of these emerging health
technologies. An open question is whether early adopters of these advances conceptualize
privacy in different ways than non-early adopters. Purpose: This study sought to understand
how early adopters of emerging health technologies conceptualize privacy. Methods:
Transcripts from in-depth, semi-structured interviews with early adopters of genome
sequencing and health devices and apps were analyzed with a focus on participant attitudes
and perceptions of privacy. Themes were extracted using inductive content analysis. Results:
Interviewees were proponents of sharing personal data to support scientific advancements,
however they still voiced concerns, as well as uncertainty about who has access to their data,
and for what purpose. In short, they were not dismissive of privacy risks. Key privacy-related
findings are organized into four themes: (1) contributing personal data to science; (2)
concerns about discrimination; (3) control over personal information; and (4) personal data
security. Conclusion: Early adopters of emerging health technologies appear to have more
complex and nuanced conceptions of privacy than might be expected based on their adoption
of personal health technologies and participation in open science. Early adopters also voiced
uncertainty about the privacy implications of their decisions to use new technologies and
share their data for research. Though not representative of the general public, studies of early
adopters can provide important insights into evolving conceptions of privacy in the context of
emerging health technologies and personal health data research.

CORRESPONDING AUTHOR: Cynthia Cheung, MPH/MA (c), University of California - San Diego,
La Jolla, CA, 92093-0811; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2587

D120 10:00 AM-11:00 AM

FITSPO AT EVERY SIZE?: A CONTENT ANALYSIS OF #CURVYFIT VS. #CURVYYOGA INSTAGRAM

IMAGES

Jennifer B. Webb, Ph.D.1, Victoria N. Clark, B. S. Student1, Elizabeth N. Burris, B. S. Student2,

Davina Y. Putz, B. S.1, Erin R. Vinoski, MPH, CHES1

1
UNC Charlotte, Charlotte, NC; 2North Carolina State University, Raleigh, NC

Recent scholarship indicates that fitspiration or fitspo content posted to popular Internet
sites frequently conflates notions of health and fitness with a thin-toned athletic
appearance ideal and conveys problematic themes emphasizing weight loss, body
objectification, and aspects of disordered eating. To extend this burgeoning line of research,
the present study examined how aspects of fitness were represented in more size-inclusive
social media contexts. The current analysis compared relevant themes depicted in images
posted to #curvyfit versus #curvyyoga on Instagram. Three hundred images (i.e., 150 per
hashtag) were systematically coded for type (i.e., visual, text, visual + text) and for the
following key thematic content domains: 1) image attributes (e.g., gender, race/ethnicity,
body size, and body shape), 2) health-exercise focus (e.g., health or fitness at any size,
functional body pride, problematic relationship with exercise, etc.), and 3) weight-appearance
focus (e.g., weight loss, body skin exposure, objectifying attire, etc.). Categorical tests of
association and ANOVA models were computed to address the primary study aims. Overall,
the findings indicate that #curvyyoga posts exhibit greater inclusion of individuals with higher
body weights (p < .001) and rounder body shapes (p < .001) and tend to represent health and
fitness in terms of body competence (e.g., endorsing health/fitness at every size: p < .001;
performing an action: p < .001; functional body pride: p < .001). Conversely, like traditional
fitspo content, #curvyfit images more often convey appearance-focused and objectified
attributes of fitness (e.g., weight loss: p < .001; bare midriff: p < .001; bare thighs/legs: p <
.01; visible cleavage: p < .001). Preliminary results help clarify the ways in which the Health at
Every Size health promotion paradigm is disseminated and differentially actualized among
peers in social media contexts who identify as curvy. Findings also contribute to illuminating
how such content could help transform current stereotypes regarding the embodiment of
fitness along with who can and does practice yoga.

CORRESPONDING AUTHOR: Jennifer B. Webb, Ph.D., UNC Charlotte, Charlotte, NC, 28223;
[email protected]
S2588 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D122 10:00 AM-11:00 AM

HEALTHCARE PROVIDER ADVICE ON GESTATIONAL WEIGHT GAIN: UNCOVERING A NEED FOR

MORE EFFECTIVE WEIGHT COUNSELING

Rebecca L. Emery, MS1, Rachel P. Kolko, PhD2, Michele D. Levine, PhD3

1
University of Pittsburgh, Pittsburgh, PA; 2University of Pittsburgh School of Medicine,
Pittsburgh, PA; 3University Of Pittsburgh, Pittsburgh, PA

Women who gain excessive gestational weight are at risk of serious health complications both
during and after pregnancy. In an effort to reduce the disease burden associated with
excessive gestational weight gain (GWG), the Institute of Medicine reissued guidelines for
GWG that are based on prepregnancy body mass index (BMI). However, limited research has
assessed whether women are adequately counseled on these revised GWG guidelines by their
healthcare providers. Participants for the present study were women who began pregnancy
overweight or obese (N = 116) and who were enrolled in a larger study assessing changes in
disordered eating symptoms and other psychological variables across the perinatal period.
Between 12 and 20 weeks gestation, women provided demographic information and were
asked to report whether they knew the recommended amount of gestational weight they
should gain during their current pregnancy. At 6 months postpartum, women were asked
whether a healthcare provider, such as a doctor, nurse, or midwife, ever advised them on
weight gain during pregnancy and how much weight they were told to gain. On average,
women were 27.08 5.26 years of age and 60% (n = 69) were black. At the beginning of
pregnancy, 48% (n = 56) of women reported knowing how much gestational weight to gain.
However, only 52% (n = 29) of these women reported a GWG within the recommended range
for their prepregnancy BMI. Women were more likely to have knowledge on appropriate
GWG if they were more highly educated, had a greater income, were white, or began
pregnancy overweight (ps < 0.02). Interestingly, only 35% (n = 41) of women received specific
advice about GWG from a healthcare provider, and only 39% (n = 16) of these women were
instructed to gain weight within the recommended range for their prepregnancy BMI.
Healthcare providers were more likely to appropriately inform women about GWG if they
began pregnancy overweight and tended to advise women who began pregnancy obese to
gain more weight than what is recommended (p = 0.03). Taken together, these findings
indicate that women who belong to a racial minority group, are socioeconomically
disadvantaged, or begin pregnancy obese are less likely to be knowledgeable on appropriate
GWG and that healthcare provider advice on GWG is largely insufficient and may be
inappropriate. Thus, there is room for behavioral interventions to aid both in the delivery of
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2589

education about GWG and in the management of weight gain during pregnancy among
overweight and obese women.

CORRESPONDING AUTHOR: Rebecca L. Emery, MS, University of Pittsburgh, Pittsburgh, PA,

15221; [email protected]
S2590 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D123 10:00 AM-11:00 AM

INFLUENCE OF PARENTAL HEALTH LITERACY ON REACH, RETENTION, AND OUTCOMES IN A

FAMILY-BASED CHILDHOOD OBESITY TREATMENT PROGRAM

Jamie Zoellner, PhD, RD1, Jennie L. Hill, PhD2, Wen You, Ph.D.3, Donna Brock, MS1, Madlyn
Frisard, PhD1, Ramine Alexander, PhD, MPH4, Fabiana Brito, BSN, MSPH, PhD5, Bryan Price,
BS6, Ruby Marshall, RN,BSN,CPC7, Paul A. Estabrooks, PhD2

1
Virginia Tech, Blacksburg, VA; 2University of Nebraska Medical Center, Omaha, NE; 3Virginia
TEch, Blacksburg, VA; 4UNC Center for Health Promotion and Disease Prevention, Durham,
NC; 5University of Nebraska Medical Center, La Vista, NE; 6Danville River District Consulting,
Danville, VA; 7Danville / Martinsville Physician Practices, Danville, VA

Introduction. Few interventions have evaluated the influence of parent health literacy (HL)
status on weight-related child outcomes. This study explores how parent HL impacts the
reach, attendance, retention, and outcomes from a three-month multi-component family-
based childhood obesity treatment program (iChoose).
Methods. This pre-post, quasi-experimental trial occurred in the Dan River Region, a federally
designated medically underserved area. iChoose research protocol and intervention strategies
were designed using a HL universal precautions approach. Validated measures and
standardized data collection techniques were used, along with generalized linear mixed effect
parametric models to determine the moderation effect of parent HL on outcomes.
Results. There was no statistically significant difference in HL scores between parents who
enrolled their child in the study [14.07 (1.52)] and parents who did not [14.13 (1.37)]. Of 94
enrolled parents, 34% were low HL, 49% income < $25,000, and 39% <high school education.
Of 101 enrolled children, 61% were black, and the mean age was 9.8 (1.3) years. Children with
low and high HL parents attended and were retained at similar rates. Significant main effect
improvements were observed for child BMI z-score [-0.05 (0.02)], ounces of SSB [-9.06 (3.17)],
and quality of life [3.00 (2.74)] and for parent BMI [-0.28 (0.04)], minutes of moderate to
vigorous physical activity [110.64 (40.74)], servings of fruits and vegetables [0.58 (0.27)], and
ounces of SSB [-6.63 (0.37)]. Parent HL status did not significantly influence improvements in
these outcomes, with the exception of child video game/computer screen time (coefficient
0.52, SE=0.11, p < 0.001; low HL decreased, high HL increased).
Conclusion. By incorporating design features that attended to the HL needs of
parents/caregivers, children with low HL parents engaged in and benefited from a family-
based childhood obesity treatment program similar to children with high HL parents.
Funding: NIH, National Institute on Minority Health and Health Disparities (R24MD008005)
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2591

CORRESPONDING AUTHOR: Jamie Zoellner, PhD, RD, Virginia Tech, Blacksburg, VA, 24061;
[email protected]
S2592 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D124 10:00 AM-11:00 AM

SELF-DETERMINATION THEORY, INTEREST IN AN MHEALTH INTERVENTION FOR IMPROVING

EXERCISE, AND EXERCISE MOTIVATION

Arturo Durazo, BA, Linda D. Cameron, Ph. D. , Roberto Corona, BA psychology, Martha P.
Blanco, BA

University of California, Merced, Merced, CA

With a mixed-methods approach, we examined the acceptability of HealthyYouTXT-Physical

Activity, an mHealth intervention developed by the US National Cancer Institute and self-
determination theorys (SDT) motivations as predictors of physical activity intentions. The
mHealth program is grounded in SDT, which predicts that intrinsic motivations, over extrinsic
motivation and amotivation, promotes intentions. The study was conducted in Californias
Central Valley, a majority-Hispanic region with high poverty and low education levels.
A community sample of 130 (61% female; M age=32, range=18-76), completed an initial
survey and structured interview in English (n=102) or Spanish (n=28). Participants received
and reviewed SMS text messages from the HealthyYouTXT-Physical Activity in English or
Spanish. Participants identified as Hispanic (62%), White (22%), Asian (9%) and Black (5%);
25% had 12th grade education or less; 37% were unemployed.
Thematic analyses of interviews revealed that most participants viewed the program
favorably. Many expressed strong interest in using it to overcome barriers such as
amotivation, work/family demands, lack of support, and weather conditions. Age-related
differences in receptivity to the program included younger adults being motivated by its utility
to assist in weight control and older adults being drawn to its potential to improve health and
longevity. Those who were already physically active perceived the program as a valuable
maintenance tool. Surprisingly, the program was viewed as a valuable source of extrinsic
motivation that would hold users accountable to adherence. Regression analyses revealed
that extrinsic motivation was the strongest predictor of physical activity intentions (=.26, p <
.01); intrinsic motivation and amotivation were not significant predictors. Interviews revealed
positive evaluations and demand for the HealthyYouTXT program among adults in rural areas.
Although the program is grounded in SDT and based on promoting intrinsic motivation,
extrinsic motivation emerged as the primary predictor of physical activity intentions. Future
research can further explore extrinsic motivation as a target of this and other mHealth
programs, particularly for ethnically diverse populations.

CORRESPONDING AUTHOR: Arturo Durazo, BA, University of California, Merced, Merced, CA,
95340; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2593

D125 10:00 AM-11:00 AM

THE ROLE OF MENTAL ILLNESS IN THE SELECTION OF SOURCES FOR HEALTH INFORMATION

Jennifer A. Lueck, Ph.D.

Texas A&M University, College Station, TX

Depression and anxiety symptoms make up a large proportion of the global disease burden of
mental illness. Although there is evidence that depression and anxiety symptoms might make
coping and seeking help particularly difficult, less is known about the specific patterns in
which affected individuals engage with sources of health information patterns that could
inform effective health communication interventions. Using 2014 data from the nationally
representative Health Information National Trends Survey, the present research tested the
effects of depression and anxiety symptoms (e.g., Over the past 2 weeks, how often have you
been bothered by little interest or pleasure in doing things/feeling down, depressed, or
hopeless/feeling nervous, anxious, or on edge/not being able to stop or control worrying?)
on engagement with health information (The most recent time you looked for information
about health or medical topics, where did you go first?). Confirming previous findings in the
literature, preliminary results from binary logistic regression analyses indicated that more
severe symptoms of depression showed a negative relationship with seeking health
information from health professionals (OR=.81, 95%CI=.67-.98), indicating a preference for
internet sources, whereas mild symptoms of depression showed a positive relationship with
seeking health information from health professionals (OR=1.32, 95%CI=1.12-1.56). Anxiety
symptoms, on the other hand, showed a strong positive relationship with seeking health
information from health professionals (OR=1.23, 95%CI=1.03-1.47) and a negative relationship
with seeking health information from internet sources (OR=.80, 95%CI=.69-.93). An
encouraging result emerged from the data neither depression nor anxiety symptoms
showed a negative relationship with seeking health information from interpersonal sources,
suggesting potentially effective pathways for reaching highly vulnerable individuals with
health information. Implications for the design and dissemination of such interventions will be
discussed.

CORRESPONDING AUTHOR: Jennifer A. Lueck, Ph.D., Texas A&M University, College Station,
TX, 77845; [email protected]
S2594 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D126 10:00 AM-11:00 AM

A MOBILE APP TO PROMOTE ART ADHERENCE AND FRAMED IN THE INFORMATION-

MOTIVATION-BEHAVIOR SKILLS MODEL: BASELINE DATA

Marcia Holstad, PhD, FNP-BC1, Melinda Higgins, PhD2, Igho Ofotokun, MD, MSc3, Eugene
Farber, PhD, ABPP3, Drenna Waldrop-Valverde, PhD1

1
Emory University, Atlanta, GA; 2Emory University, Decatur, GA; 3Emory University School of
Medicine, Atlanta, GA

Rural HIV + populations often have difficulties adhering to antiretroviral therapy (ART). The
Music for Health Project (MFH) is a randomized controlled trial that tests the efficacy of a
smart phone application (app) on ART adherence and symptom self-management in rural
Georgia. The intervention app contains a music-based messaging program and manual
designed to educate and motivate HIV+ rural dwelling persons to adhere to ART and self-
manage symptoms and side effects. The control app contains a general education program in
a comparable format. Eligibility criteria are HIV+ and: ART nave or ART change due to drug
resistance or detectable VL or ART prescribed. We analyzed baseline data of all 149 enrolled
to examine associations of self-report ART adherence (3 scales) self-efficacy (2 scales), ART
motivation, depressive symptoms (CESD-R), and HIV health literacy with the components of
the Information-Motivation-Behavior (IMB; Life Windows Scale) model and determine if
differences existed by eligibility groups and demographics. Analyses included descriptive
statistics, Spearmans rho correlations, Mann Whitney tests and Kruskal-Wallis ANOVA. The
average age of the group was 39.1 years old, 66% were male, 77% African American (AA) and
48% self-reported as heterosexual. The majority (80%) lived with others versus alone; and
16% lived with another HIV+ person. The regimen change and treatment nave groups had
higher Information scores (p=.048). AA had higher Behavior skills scores (p=.041). Higher ART
Information, Motivation and Behavior skills scores were significantly correlated with better
self-reported adherence (p < .001), higher self-efficacy for symptom self-management (p <
.05) and ART adherence (p < .001), and higher intrinsic motivation for adherence (p < .001).
Higher HIV health literacy was associated with higher Information scores (p=.015). Depressive
symptoms were associated with lower Motivation and Behavior scores (p < .001). These data
support use of the IMB model for ART adherence behavior technology interventions for rural
dwelling HIV+ persons.

CORRESPONDING AUTHOR: Marcia Holstad, PhD, FNP-BC, Emory University, Atlanta, GA,
30322; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2595

D127 10:00 AM-11:00 AM

ACCEPTABILITY OF A SMOKING CESSATION AND EMOTION REGULATION INTERVENTION FOR

ADULTS LIVING WITH HIV

Jessica N. Coleman, BA1, Allison K. Labbe, PhD2, Julianne G. Wilner, M.A.3, Samantha M.
Marquez, MSW4, Jesse D. Kosiba, BA5, Michael J. Zvolensky, PhD6, Jasper Smits, PhD7, Conall
O'Cleirigh, PhD2

1
Massachusetts General Hospital, Cambridge, MA; 2Massachusetts General Hospital/Harvard
Medical School, Boston, MA; 3Boston University, Boston, MA; 4Fenway Health, Boston, MA;
5
Syracuse University, Syracuse, NY; 6Department of Psychology, University of Houston,
Houston, TX; 7Department of Psychology, University of Texas at Austin, Austin, TX

Background: The prevalence of cigarette smoking among persons living with HIV (PLWH) is
three times that of the general population. PLWH who smoke experience greater morbidity
and mortality than non-smokers and tobacco use is associated with poorer HIV-related health
outcomes among this population. Similarly, mood and anxiety disorders are especially
common among PLWH. We developed a novel 8-week, integrated, transdiagnostic cognitive-
behavioral intervention for PLWH aimed at (1) promoting successful smoking cessation, and
(2) improving emotion regulation skills. The aim of this study was to assess the acceptability of
the intervention via qualitative interviews and preliminary outcome data.

Methods: HIV+ smokers were recruited from a community health center and an infectious
disease clinic at a large medical center in Boston, MA. Exit interviews were conducted with a
subsample of participants (N = 10, 90% Male, 50% Black/African America and 40% White,
Mcpd = 17.64) who completed through the 6-month follow-up visit. Interviews were
recorded, transcribed, and analyzed using an iterative approach.

Results: The majority of participants indicated that quitting smoking, becoming healthier, and
finishing what they started were primary motives for starting and completing the
intervention. Participants reported that the homework, handouts, and cognitive restructuring
skills were helpful in quitting smoking, whereas thought tracking and awareness aspects were
reported as helpful for managing distress. Five of the participants described using intervention
components to maintain abstinence at their 6-month follow-up, and almost half of the
participants reported using skills learned in treatment to manage distress. Participants
described some aspects of the treatment as challenging (i.e., using the patch consistently,
delaying smoking a cigarette, practicing certain skills), but also indicated that these aspects
were helpful. All participants reported that they would recommend the treatment to others.
S2596 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Five participants had quit completely and four participants reduced the amount of cigarettes
smoked per day by the 6-month follow-up.

Discussion: Overall, participants found this newly developed treatment to be acceptable.

Preliminary outcome data suggest the intervention was successful in helping participants to
quit or reduce smoking at the 6-month follow-up. Further randomized controlled studies are
needed to test the efficacy of this novel, transdiagnostic treatment.

CORRESPONDING AUTHOR: Jessica N. Coleman, BA, Massachusetts General Hospital,

Cambridge, MA, 02139; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2597

D128 10:00 AM-11:00 AM

ADHERENCE TO PRE-EXPOSURE PROPHYLAXIS FOR HIV AND SEXUAL RISK COMPENSATION

AMONG MEN WHO HAVE SEX WITH MEN AND TRANSGENDER WOMEN

Michael J. Li, M.P.H.1, Joel Milam, Ph.D.1, Michael Dube, MD1, Jessica Tobin, M.A.1, Katya
Corado, MD2, Eric Daar, MD2, David J. Moore, Ph.D.3, Sheldon Morris, MD, MPH4

1
Keck School of Medicine, University of Southern California, Los Angeles, CA; 2Los Angeles
Biomedical Research Institute at Harbor-UCLA Medical Center, Torrance, CA; 3UCSD HIV
Neurobehavioral Research Program, San Diego, CA; 4UCSD Antiviral Research Center, San
Diego, CA

Introduction: There is mixed evidence suggesting that use of pre-exposure prophylaxis for HIV
prevention predicts increases in sexual risk behaviors, also known as sexual risk
compensation. It is unclear whether the degree of change in risk behavior depends on how
well a PrEP user adheres to the regimen.

Objective: The purpose of this study is to test whether adherence to PrEP affects changes in
sexual risk behavior among a cohort of men who have sex with men initiating PrEP.

Methods: Participants (N = 377, mean age 35.1 yrs; 28.9% Latino, 11.4% Non-Hispanic Black,
and 9.3% other non-White) were obtained from multiple clinics in Southern California taking
part in a PrEP demonstration study and were followed for 60 weeks following PrEP initiation.
Self-reported PrEP adherence and sexual risk behaviors in past 4 weeks were assessed after
baseline at 4 , 12, 24, 36, 48, and 60 weeks, with participants who self-reported having ever
missed a dose within the prior 4 weeks considered non-adherent. We used negative binomial
mixed modeling, adjusting for age, race (White or non-White), drug abuse, clinic site, and
study arm to test whether adherence to PrEP predicts changes in prior-4-week sexual risk
behavior over 60 weeks.

Results: Over 60 weeks, participants overall reported increased unprotected receptive anal
intercourse (URAI) acts (p < .001). Participants who never missed a dose (ranging 66.6-80.1%
across all waves) had a significantly greater increase in URAI (p = .037) and UIAI (p = .002) in
the prior 4 weeks compared to participants who reported imperfect adherence in the prior 4
weeks. Participants who reported imperfect adherence did not significantly change in any
sexual risk behavior.

Discussion: Sexual risk compensation behaviors increased following PrEP initiation in those
that reported they did not miss a dose in the past 4 weeks. This suggests that PrEP users
S2598 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

conscientiously adhere to F/TDF when they they have increased sexual risk behaviors to
minimize HIV acquisition.

CORRESPONDING AUTHOR: Michael J. Li, M.P.H., Keck School of Medicine, University of

Southern California, Los Angeles, CA, 90032; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2599

D129 10:00 AM-11:00 AM

BODY IMAGE AND SEXUAL RISK AMONG SEXUAL MINORITY MEN LIVING WITH HIV

Benjamin M. Rooney, B.A.1, Shadi Gholizadeh, M.Sc., M.S., M.P.H. 2, Erin L. Merz, PhD, MPH3,
Vanessa L. Malcarne, Ph.D.1, Steven A. Safren, PhD4, Aaron Blashill, Ph.D. 1

1
San Diego State University, San Diego, CA; 2SDSU/UCSD Joint Doctoral Program in Clinical
Psychology, San Diego, CA; 3California State University, Dominguez Hills, Carson, CA;
4
Department of Psychology, University of Miami, Coral Gables, FL

Background: It is unclear under what conditions body image impacts sexual risk behaviors
(SRB). It is possible that body dissatisfaction only predicts increased sexual risk for individuals
who also have high appearance investment. However, to date, there are no known studies
that have examined the interaction of body dissatisfaction and appearance investment in
predicting SRB.

Methods: Participants were 105 sexual minority men living with HIV who reported having sex
with a man within the previous 12 months. Participants completed a battery of self-report
measures, including the Appearance Orientation (AO) subscale from the Multidimensional
Body Self-Relations Questionnaire (MBSRQ) as a marker of appearance investment, the
Muscle Dysmorphic Disorder Inventory (MDDI) as a marker of body dissatisfaction, and SRB
(i.e., a count of the number of condomless anal intercourse acts with HIV-negative or HIV-
unknown status male partners in the past 3 months).

Results: A generalized linear model with a Poisson distribution (to account for the count SRB
outcome variable) was employed, with the main effects of appearance investment, body
dissatisfaction, and their interaction term entered as independent variables. The main effects
of appearance investment and body dissatisfaction were non-significant. A significant
appearance investment by body dissatisfaction interaction was revealed (b = .08, SE = .03,
95% CI: .007, .159, Wald 2 = 4.5, p = .033). To probe this significant interaction, simple slope
analyses were conducted, setting values of appearance investment at 2 SD from the mean.
Results suggested that body dissatisfaction negatively predicts SRB at low levels of
appearance investment (b = -.07, t = -1.89, p = .06), whereas it positively predicted SRB at high
levels of appearance investment (b = .10, t = 1.9, p = .05).

Conclusions: Appearance investment significantly moderated the relationship between body

dissatisfaction and SRB, underscoring the multidimensional impact of body image in SRB for
sexual minority men living with HIV. Further understanding of the mechanisms through which
S2600 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

body image concerns impact SRB has the potential to inform future screening and treatment
efforts meant to promote psychological well-being and reduce HIV sexual transmission risk
behaviors.

CORRESPONDING AUTHOR: Benjamin M. Rooney, B.A., San Diego State University, San Diego,
CA, 92120; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2601

D130 10:00 AM-11:00 AM

EXAMINING THE ACCEPTABILITY OF MHEALTH TECHNOLOGY IN HIV PREVENTION AMONG

HIGH-RISK DRUG USERS IN TREATMENT

Roman Shrestha, MPH1, Frederick Altice, PhD2, Michael Copenhaver, PhD3

1
University of Connecticut Health Center, Willimantic, CT; 2Yale University, New Haven, CT;
3
University of Connecticut, Storrs, CT

Introduction: Despite promising trends of the efficacy of mobile health (mHealth) based
strategies to a broad range of health conditions, very little if any studies have been done in
terms of the examining the use of mHealth in HIV prevention efforts among people who use
drugs (PWUDs) in treatment. Thus, the goal of this study was to gain insight into the real-
world acceptance of mHealth approaches among high-risk PWUDs in treatment.

Methods: A convenience sample of 400 HIV-negative drug users, who reported drug- and/or
sex-related risk behaviors, were recruited from a methadone clinic in New Haven, CT.
Participants completed standardized assessments of drug- and sex-related risk behaviors,
neurocognitive impairment NCI), and measures of communication technology access and
utilization, and mHealth acceptance. Chi-square and t-tests were conducted to examine the
statistical differences on ownership or utilization of communication technologies and mHealth
acceptance between individuals with and without NCI.

Results: We found a high prevalence of current ownership and use of mobile technologies,
such as cell phone (91.5%) including smartphone (63.5%). Participants used mobile
technologies to communicate mostly through phone calls (M = 4.25, SD = 1.24), followed by
text messages (M = 4.21, SD = 1.29). Participants expressed interest in using mHealth,
particularly text messaging service, for medication reminders (72.3%), to receive information
about HIV (65.8%), and to assess drug-related (72.3%) and sex-related behaviors (64.8%).
Participants who were neuro-cognitively impaired were less likely to own a cell phone (OR =
0.453, 2 = 4.976, p = 0.026), particularly smartphone (OR = 0.476, 2 = 11.251, p < 0.05) and
more likely to use cell phone without internet [t(398) = -2.747, p = 0.006, d = -0.301] .
Furthermore, individuals who were cognitively impaired were significantly more likely than
those without NCI to show interest in using mHealth to remind them to take medication(s)
[t(398) = -2.490, p = 0.013, d = -0.262], to receive information about HIV [t(398) = -4.220, p <
0.001, d = -0.437], to assess drug-related risk behaviors [t(398) = -2.003, p = 0.046, d = -
0.242], and sex-related risk behaviors [t(398) = -2.617, p = 0.009, d = -0.312].
S2602 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

Conclusions: The findings from this study provide empirical evidence that mHealth-based
programs, specifically cell phone text messaging-based health programs, may be acceptable to
this high-risk population. Researchers are encouraged to conduct formative research to
explore opportunities to integrate cell phone text messaging into HIV prevention programs
designed for implementation, followed by practical development, implementation, and
evaluation among PWUDs in treatment.

CORRESPONDING AUTHOR: Roman Shrestha, MPH, University of Connecticut Health Center,

Willimantic, CT, 06226; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2603

D131 10:00 AM-11:00 AM

HIV TREATMENT OPTIMISM AS IT RELATES TO PERCEIVED SEXUAL TRANSMISSION RISK IN

MSM

Matthew Sullivan, B.A., Seth Kalichman, PhD

University of Connecticut, Storrs, CT

Greater perceived risk of contracting HIV or other sexually transmitted infections has been
linked to increased use of protective health measures such as condoms. In the US, men who
have sex with men (MSM) are at particularly high risk for HIV infection, and thus these
perceptions may be especially consequential for preventing risky sex behavior within this
population. Recent research has found that optimism about HIV treatment advances may
reduce the perceived risk associated with acts such as unprotected anal sex, thereby
potentially increasing risky sex behavior. Thus, treatment optimism may be an important
consideration in understanding and reducing sexual risk-taking. This study sought to examine
the relationship between HIV status, treatment optimism, and the perceived reduction of HIV
transmission risk associated with antiretroviral (ARV) treatment in sample of 318 sexually
active MSM recruited at a 2015 Atlanta Gay Pride event (mean age=35.17 years; 96%
Caucasian; 17.3% tested HIV positive). To examine this relationship, a mediation model was
tested by bootstrapping the indirect effect. As hypothesized, a direct effect of HIV status on
perceived risk reduction was found (b=-0.48, 95% Confidence Interval: -0.88 to -0.05, p=0.03),
such that HIV-positive men perceived a greater risk reduction associated with ARV treatment
than HIV-negative men. However, this effect was not significantly mediated by treatment
optimism. One potential reason for this lack of a mediation effect is that treatment optimism
was relatively high across all individuals surveyed (mean score of 12.8 (range of 4-16)). When
these relationships were examined within HIV status subgroups, no relationship between
treatment optimism and perceived risk reduction was found among HIV-positive individuals.
However, among nonpositive individuals, greater treatment optimism was associated with
greater perceived risk reduction (p=0.057). This finding suggests that optimistic nonpositive
individuals may be more likely to perceive HIV treatment as reducing the risk of unprotected
anal intercourse with an HIV-positive partner. Such beliefs may predispose such individuals to
a greater likelihood of engaging in risky sex behaviors.

CORRESPONDING AUTHOR: Matthew Sullivan, B.A., University of Connecticut, Storrs, CT,

06269; [email protected]
S2604 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D132 10:00 AM-11:00 AM

PROGRESO EN SALUD, AN HIV RISK REDUCTION INTERVENTION FOR LATINA SEASONAL

FARMWORKERS: PRELIMINARY FINDINGS

Mariano Kanamori, PhD1, Mario De La Rosa, PhD2, Stephanie L. Diez, MSW2, Jessica
Weissman, MS1, Mary Jo Trepka, MD3, Alicia Sneij, MA1, Peter Schmidt, PhD4, Patria Rojas,
PhD5

1
Center for Research on U.S. Latinos HIV/AIDS and Drug Abuse, Robert Stempel College of
Public Health & Social Work, Florida International University, Miami, FL; 21 Center for
Research on US Latino HIV/AIDS and Drug Abuse, Florida International University, Miami, FL;
3
Department of Epidemiology, Florida International University, Miami, FL; 4Department of
Political Science, University of Gieen, Gieen, N/A, Germany; 5Health Promotion and Disease
Prevention. Florida International University., Miami, FL

Introduction. Latina seasonal workers in the US are an underserved population with limited
access to health care and health education. In 2015, the HIV diagnosis rate in South Florida, a
region where many Latina seasonal farmworkers reside and work, was the highest among all
metropolitan areas in the US (38.8 per 100,000). The lack of culturally tailored evidence-based
HIV prevention programs for Latina seasonal workers led to the development and
implementation of Progreso en Salud, which is a cultural adaptation of the CDCs evidenced
based intervention VOCES/VOICES. We will present preliminary findings of our adapted
intervention.

Methods. We measured sex behaviors at baseline and 6-month follow-up among 114 Latina
seasonal workers who participated in the Progreso en Salud intervention. Analyses included
McNemars test, Related Samples Marginal Homogeneity Test, Related Samples Wilcoxon
Signed Rank Test, and Mediation Analysis.

Results. A higher percentage of women at 6-month follow-up used condoms during the past
30 days every time (13.5% vs. 9.6%) or sometimes (24.0% vs. 17.3%) when they had vaginal
and/or anal sex. There was also a decrease in the percentage of women who never used
condoms (57.7% vs. 49.0%) when they had sex. There was a significant increase in HIV testing
from baseline (15.0%) to 6-month follow-up (27.9%). There were also significant increases in
the percentages of HIV/AIDS-related communications with friends, HIV knowledge, condom
use self-efficacy, and correct use of condoms. Condom use self-efficacy had a positive effect
on the relationships between: (a) HIV knowledge and condom use; and (b) Communication
with friends regarding HIV prevention and condom use.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2605

Conclusions. Our results suggest that existing best practice interventions (e.g. VOICES/VOCES)
can be tailored to improve HIV knowledge, HIV testing, and condom use among Latina
seasonal workers. For such interventions to be successful, they should address Latino cultural
values and be conducted in an environment that allows Latina seasonal workers to feel
comfortable discussing these culturally taboo and sensitive topics. Our findings suggest that
Progreso en Salud may have helped participants increase their knowledge about HIV, assisted
with practices regarding condom use, and increased rates of HIV testing. Also, many
participants maybe in the process of forming their condom use self-efficacy as a result of the
intervention. Future research is needed to support our findings.

CORRESPONDING AUTHOR: Mariano Kanamori, PhD, Center for Research on U.S. Latinos
HIV/AIDS and Drug Abuse, Robert Stempel College of Public Health & Social Work, Florida
International University, Miami, FL, 33139; [email protected]
S2606 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D133 10:00 AM-11:00 AM

PSYCHOSOCIAL CORRELATES OF INSULIN RESISTANCE IN AN ETHNICALLY AND RACIALLY

DIVERSE SAMPLE

Violeta Rodriguez, MSEd, Deborah Jones, PhD, Phillip Szewc, Mahendra Kumar, PhD

University of Miami Miller School of Medicine, Miami, FL

Introduction: Methamphetamine (MA) use has increasingly common in the United States and
is prevalent among MSM in South Florida (18%). Among HIV infected MSM who report drug
use, 43% reported trading sex for MA, and overall, MSM are at great risk for MA use. Long
term MA use has been implicated in insulin resistance, which in turn has been associated with
sleep disorders, antiretroviral use, trauma, and obesity. Despite the high prevalence of MA
use among MSM, insulin levels and associated risk factors have not been assessed in this
population. This study examined the relationship between MA use, HIV status, and insulin
resistance among ethnically and racially diverse MSM living in South Florida.
Method: Participants (n = 335) were men 18-55 years of age recruited in South Florida, and
comprised three groups: 1) HIV uninfected-non-MA-using (HIV-MA-; n = 136, 2) HIV-infected,
MA-using, MSM (HIV+MA+; n = 147) and 3) HIV-uninfected, MA using, MSM (HIV-MA+; n =
52). Participants completed assessments on demographics, depression, trauma history, sleep
disturbance, and drug use, provided a fasting blood sample to assess glucose and insulin. Body
mass index (BMI) was derived from measurement of height and weight and insulin resistance
was calculated using the Homeostasis Model.
Results: Participants were aged 18 - 55 years (mean = 37, SD = 9.91). The majority were
Hispanic, 60% of participants had completed at least high school, 60% were overweight or
obese, and 50% had a monthly income of less than US$500. Groups differed with regard to
age, race, income, BMI, insulin resistance and severity of sleep disturbance. Polydrug use was
associated with insulin resistance in bivariate analyses, but in multivariable analysis, insulin
resistance levels were highest among obese men in comparison with normal weight men,
highest among HIV-MA- men in comparison with other groups, and highest among Hispanics
in comparison with non-Hispanic whites.
Discussion: Insulin resistance levels among this sample of ethnically and racially diverse
methamphetamine-using HIV-negative men were higher than among HIV-infected men, likely
due to decreased hunger associated with stimulant use and those participants either
overweight or obese in the non-MA group, many of whom were Hispanic. Future studies
should explore the association of insulin resistance with a history of marijuana use in
comparison to polydrug use, as decreased levels of insulin resistance may also be specific to
marijuana use, and continue to examine the relationship between antiretroviral use and
insulin levels. Funded by a grant from NIDA, R01DA031201.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2607

CORRESPONDING AUTHOR: Violeta Rodriguez, MSEd, University of Miami Miller School of

Medicine, Miami, FL, 33136; [email protected]
S2608 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D134 10:00 AM-11:00 AM

SOCIAL ISOLATION AND AFFECT IN OLDER ADULTS LIVING WITH HIV: NEW INSIGHTS USING
MOBILE TECHNOLOGY

Alexandra Rooney, B.A.1, Christopher N. Kaufmann, PhD, MHS2, Brent Mausbach, PhD3, Colin
Depp, PhD3, Anne Heaton, N/A4, Jeremy Delgadillo, N/A5, Sheena Dev, B.S.5, David J. Moore,
Ph.D.6, Raeanne C. Moore, PhD7

1
University of California San Diego, San Diego, CA; 2UC San Diego, La Jolla, CA; 3University of
California San Diego, La Jolla, CA; 4UCSD HIV Neurobehavioral Research Program, Del Mar, CA;
5
San Diego State University, San Diego, CA; 6UCSD HIV Neurobehavioral Research Program,
San Diego, CA; 7University of California, San Diego, San Diego, CA

Older adults with HIV are more likely to live alone than their peers without HIV, which puts
them at risk for loneliness and social isolation. Social isolation is a major and prevalent health
problem, and is an under-assessed condition in older HIV+ adults. The repeated sampling of
daily experiences in a naturalistic setting, known as ecological momentary assessment (EMA),
offers a potentially useful method for examining social isolation in relation to other health-
related factors. We examined the real-time relationship between social interactions with
concurrent and future affect and physical symptoms in a sample of older HIV+ adults. Twenty
older adults infected with HIV (Mean (M) age=59; 85% male; 70% Caucasian; M current CD4:
476.21; M duration of HIV: 20.4 years) participated in one week of intensive daily monitoring
via smartphones. Participants provided five real-time reports per day of their current social
functioning (alone vs. not alone and number of social interactions since last report), as well as
their current affective state (sadness, stress, happiness), tiredness, and pain. HIV disease
characteristics were obtained in-person prior to the EMA portion of the study. Participants
spent an average of 63% of their time alone. HIV disease characteristics were unrelated to
being alone or number of social interactions. Mixed-effects models showed being alone was
related to lower concurrent happiness (Wald F=10.22, p < 0.01); however, being alone was
unrelated to concurrent sadness, stress, tiredness, or pain. The average number of social
interactions over the week was unrelated to affect, tiredness, or pain. Within-day time-lagged
analyses were conducted with alone vs. not alone at one time point predicting affect and
physical symptoms at the next time point. Lagged analyses found participants felt more pain
(coef=0.027, p=0.02), were more tired (coef=0.17, p < 0.001) but had less stress (coef= -0.05, p
< 0.001) over the course of the day. Increases in social activity predicted greater increases in
tiredness. Our findings support a framework for understanding happiness in the context of
social interactions among older HIV+ adults. While participants became more tired and had
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2609

increasing reports of pain over the course of the day, they experienced greater happiness
when spending time with others as compared to being alone. Clinical interventions that
promote social engagement, despite the presence of pain and fatigue, may positively impact
well-being in this growing population.

CORRESPONDING AUTHOR: Alexandra Rooney, B.A., University of California San Diego, San
Diego, CA, 92103; [email protected]
S2610 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D135 10:00 AM-11:00 AM

THE RELATION OF HEROIN USE, A REGULAR SOURCE OF HEALTH CARE, AND AMBULATORY
CARE UTILIZATION IN HOMELESS ADULTS

Jared Israel, Ph.D.1, Kamila White, Ph.D.2

1
Washington University School of Medicine, Seattle, WA; 2University of Missouri - St. Louis, St.
Louis, MO

The physical and mental health care needs of homeless adults are extensive, however, those
needs regularly go unmet. Substance use in homeless adults is associated with worse health
outcomes and decreased access to health care services. Overall, health care patterns in
homeless adults are characterized by elevated rates of emergency department and inpatient
services coupled with low levels of ambulatory care. Studies have consistently found that
having a regular source of health care is related to improved ambulatory care utilization. Little
research has been conducted to examine the relationships between substance use, a regular
source of health care, and rates of ambulatory care. This study investigated the association
between heroin use in the past year and ambulatory care use during that same period while
controlling for whether participants had a regular source of health care.

Data were collected at a community mental health clinic in Midtown St. Louis. The sample
consisted of 131 homeless adults, the majority of whom were male (70%) and African-
American (70%). Mean age was 39 years 4 months (SD = 12.2). Because ambulatory care use
in the past year was a dichotomous variable (one or more ambulatory care visits vs. none), a
hierarchical logistic regression was used. Use of heroin in the past year and having a regular
source of health care were entered as independent variables. At Step 1, regular source of care
was a significant predictor (b = 1.43, Wald Statistic = 13.87, p < .01). At Step 2, regular source
of care remained a significant predictor (b = 1.44, Wald Statistic = 13.87, p < .01), and heroin
use was a significant predictor (b = -1.45, Wald Statistic = 4.28, p =.04).

These results fit with previous findings that show 1) heroin use in homeless adults decreases
the likelihood of ambulatory care use and 2) a regular source of health care increases the
likelihood of ambulatory care use. These results also demonstrate that while having a regular
source of health care may buffer the negative affect of heroin use on ambulatory care access,
additional efforts are required to ensure that this vulnerable group receive sorely needed
ambulatory care services.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2611

CORRESPONDING AUTHOR: Jared Israel, Ph.D., Washington University School of Medicine,

Seattle, WA, 98122; [email protected]
S2612 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D136 10:00 AM-11:00 AM

A CUSTOMIZED ADHERENCE ENHANCEMENT INTERVENTION + LONG ACTING INJECTABLES

FOR HIGH RISK PATIENTS WITH SERIOUS MENTAL ILLNESS

Jennifer B. Levin, PhD1, Luis Ramirez, MD2, Edna Fuentes-Casiano, MSSA3, Jamie Cage, MS4,
Curtis Tatsuoka, Ph.D.5, Ashley Bukach, BS5, Kristin A. Cassidy, MA5, Martha Sajatovic, MD6

1
Case Western Reserve University School of Medicine/UH Cleveland Medical Center,
Cleveland, OH; 2Department of Psychiatry, Case Western Reserve University School of
Medicine/University Hospitals Cleveland Medical Center, Cleveland, OH; 3Case Western
Reserve Univsersity, Cleveland, OH; 4University Hospitals Case Medical Center, Cleveland, OH;
5
Case Western Reserve University, Cleveland, OH; 6Case Western Reserve University, Gates
Mills, OH

Aims: Long-acting injectable antipsychotic medication (LAI) can be a practical treatment

option to optimize adherence for high-risk groups such as homeless individuals with serious
mental illness (SMI). The authors of the current study developed a Customized Adherence
Enhancement (CAE) intervention and combined it with LAI to improve treatment adherence
of recently homeless individuals with schizophrenia (SZ) or schizoaffective disorder (SA). For
this analysis, preliminary data from 2 prospective, 6-month open-label, uncontrolled studies
of LAI + CAE were pooled. Haloperidol decanoate as used in Study 1 and paliperidone
palmitate was used in Study 2. Since the literature suggests that SA is more complex to treat
because of the added mood component, this analysis investigated differences in baseline and
follow up (6 months) between the two diagnostic groups, SZ and SA.

Methods: Both trials combined CAE + LAI in recently homeless individuals with SZ (N=16) or
SA (N=44) for a total combined sample of 60. Clinical outcomes included medication
adherence using the Tablets Routine Questionnaire (TRQ), LAI injection frequency and
psychiatric symptoms measured by the Positive and Negative Syndrome Scale (PANSS), Brief
Psychiatric Rating Scale (BPRS) and global psychopathology (Clinical Global Impressions /CGI).

Results: Mean combined age of the sample was 42.7 years (SD=9.0), majority male (56.7%),
minorities (88.3% African-American), single/never married (71.2%) with a mean of 11.4 years
of education. Baseline rate of substance abuse within the past year was 25.0% and rate of
incarceration within the past 6 months was 21.6 %. Non-adherence at screen was 55% of
missed doses in the past week for SZ and 57% in SA. No demographic differences were found
between the SZ and SA group at baseline aside from a larger number of Hispanic ethnicity in
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2613

the SZ group (20%) than the SA group (2.3%) (x2=5.56, p=.02) and no difference in severity of
clinical symptoms or medication adherence (t=-.200, df=55, p=.84). There was a trend of more
homeless days spent outdoors for SAs than SZs at baseline (x2=20.2, p=.09). There was no
difference between the groups on response to treatment with both groups improving
adherence to oral medications (t=.2.64, df=4, p=.058 and t=3.02, df=20, p=.007 for SA and SZ,
respectively) and no difference in injection frequency between the two groups at 6 months
(t=-.47, df=48, p=.64).

Conclusion: Highly symptomatic SMI is common in the homeless population and LAI
combined with a targeted adherence enhancement approach may be a useful therapeutic
approach in these high-risk individuals, both for those with SZ and those with SA. Despite the
added burden of mood symptoms for the latter, they are as likely to benefit from this
combined treatment and show an equally strong response.

CORRESPONDING AUTHOR: Jennifer B. Levin, PhD, Case Western Reserve University School of
Medicine/UH Cleveland Medical Center, Cleveland, OH, 44139; [email protected]
S2614 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D137 10:00 AM-11:00 AM

A MIXED-METHODS EXPLORATION OF THE ROLE OF PARTNERS IN BREAST CANCER

SURVIVORS ADHERENCE TO AROMATASE INHIBITORS

Moriah Brier, MA1, Dianne Chambless, PhD2, Susan Zhang, BA3, Jun Mao, MD, MSCE4

1
University of Pennsylvania, Seattle, WA; 2University of Pennsylvania, Philadelphia, PA; 3NIH,
Bethesda, MD; 4Memorial Sloan Kettering Cancer Center, New York, NY

Background:The relationship between marital status and cancer-specific mortality is

particularly strong among breast cancer patients. It has been suggested that partner support
provided in marriages may improve adherence rates to treatment, thereby increasing survival
rates. The purpose of this study was to explore whether partner specific support for
aromatase inhibitors (AIs), an adjuvant hormonal therapy for breast cancer survivors, was
associated with increased AI adherence. Method: We used both qualitative and quantitative
methods to explore this topic. In the qualitative study, using a grounded theory approach, we
interviewed 25 breast cancer survivors and 15 partners about the role of partner support in
decision-making around AI initiation and continuation, and management of AI side effects.
The sample represented both survivors who were still on treatment and survivors who had
prematurely discontinued. In the quantitative study (n = 240), we examined cross-sectional
associations between different types of received support (informational, tangible, emotional,
and affection), and satisfaction with support. Results:In both studies, we found that breast
cancer survivors placed greater value on emotionally nurturing support than on solution-
focused support. In the qualitative study, partners of adherent survivors were more
frequently aware of the impact of AI treatment on survivors self-concept and provided more
support around sexual dysfunction caused by AIs than partners of non-adherent survivors.
Quantitative findings showed that after accounting for depression, received affection had the
strongest association with womens satisfaction with AI-specific support from partners (sr =
.30, p < .001), followed by emotional/esteem support (sr = .11, p = .052). Neither AI-related
tangible support nor informational support was significantly associated with support
satisfaction. These relationships were the same among women with both high and low levels
of AI side effects. Conclusion: The findings of these studies suggest that spousal support,
especially in the form of emotional support and affection, is relevant to the experience of AIs
and persistence with treatment. Survivors with adverse effects, who are vulnerable to
premature discontinuation of treatment, may benefit from interventions that not only target
personal coping, but also their romantic relationships, to enhance couples abilities to be
emotionally expressive and partners abilities to provide support that is responsive to
survivors needs.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2615

CORRESPONDING AUTHOR: Moriah Brier, MA, University of Pennsylvania, Seattle, WA, 98118;
[email protected]
S2616 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D138 10:00 AM-11:00 AM

AVOIDANCE AND DISEASE MODIFYING TREATMENT (RE)INITIATION AMONG MS PATIENTS

Rachel Schoor, MA1, Jared Bruce, Phd2, Amanda S. Bruce, Ph.D.3, Kathy Goggin, Ph.D.4,
Bethany R. Schanfarber, Bachelors of Arts in Psychology5, Andrea Bradley- Ewing, MPA, MA6,
Joanie M. Thelen, M.A.7, Morgan Glusman, Ph.D.7, Sharon Lynch, MD8, Lauren Strober, Ph.D.9,
Delwyn Catley, Ph.D.10

1
University of Missour-Kansas City, Kansas City, MO; 2Umkc, Leawood, KS; 3University of
Kansas Medical Center, Kansas City, KS; 4Children's Mercy Hospital & University of Missouri -
Kansas City, Kansas City, MO; 5University Missouri Kansas City, Los Angeles, CA; 6Children's
Mercy, Kansas City, MO; 7University of Missouri - Kansas City, Kansas City, MO; 8University of
Kansas MEdical Center, Kansas city, KS; 9Kessler Foundation, West Orange, NJ; 10Children's
Mercy Hospital & University of Missouri-Kansas City, Kansas City, MO

Disease modifying therapies (DMTs) reduce annualized relapse rates and disease progression
among those with relapsing-remitting MS (RRMS), yet the majority of MS patients choose not
to take DMT or prematurely discontinue taking them. Anecdotal evidence from a successful
randomized trial of Motivational Interviewing Cognitive Behavior Therapy to promote
initiation of DMT suggested avoidant coping was the main reason for not taking DMT.

The aims of this study were to follow up on these observations to (1) determine whether
avoidant patients could by reliably identified, (2) compare the clinical and psychological
characteristics of these patients to those with other reasons for not taking DMT, and (3)
examine whether their outcomes differed from those with other reasons for not taking DMT.

Audio recordings (N=77) from the first of 5 counseling sessions from the trial were examined.
Using a grounded theory approach researchers identified cost barriers, mild course, side
effects, and avoidance as the main reasons participants were not taking DMT. Avoidant
participants tended to downplay the seriousness of MS, and provided irrational justifications
for not taking medications characterized by distorted reasoning and exaggerated or paranoid
concerns. Criteria were developed and two raters independently classified each participant
according to their main reason (K= .88). Coding data was analyzed with baseline and end of
treatment surveys of psychological and clinical characteristics as well as follow-up
assessments of motivation and confidence to initiate and self- reported initiation.

Coding indicated that 33 participants (42.9 %) were avoidant. Chi-square and one-way
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2617

ANOVAs revealed few differences between groups on clinical and self-reported psychological
characteristics other than those that validated the categorization (e.g., those in the cost
barriers group were more likely to report not being able to afford medication, those in the
side effects group were more likely to report side effects being intolerable). One-way ANOVAs
and post hoc tests indicated that compared to each of the other groups avoidant individuals
self-reported lower motivation to take DMTs at the end of treatment (p < .001), but did not
differ in confidence (p=.78). Chi squares analyses revealed that avoidant participants were less
likely to initiate DMTs (24% compared to 52.5% of non-avoidant participants) by week 10
follow-up, p=.023, X2(3, 69)= 5.142.

Patients that appeared to be avoidant were not distinguishable with self-report clinical and
psychological measures and likely present a particular challenge for intervention because they
appeared less responsive to a generally effective intervention.

CORRESPONDING AUTHOR: Rachel Schoor, MA, University of Missour-Kansas City, Kansas

City, MO, 64134; [email protected]
S2618 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D139 10:00 AM-11:00 AM

DISPARITIES IN MEDICATION ADHERENCE PERSIST AFTER CONTROLLING FOR

SOCIODEMOGRAPHICS AND LITERACY/NUMERACY SKILLS

Abel B. Belachew, MA1, Andrea E. Lagotte, MA2, Lyndsay A. Nelson, PhD2, Mariya P. Kristeva,
BS3, Arathi S. Nandyala, BS, BA4, Chandra Y. Osborn, PhD, MPH5

1
Marshall University, Huntington, WV; 2Vanderbilt University Medical Center, Nashville, TN;
3
University of Central Florida College of Medicine, Kissimmee, FL; 4University of South Florida
College of Medicine, Tampa, FL; 5Informed Data Systems Inc., Nashville, TN

Background:

Taking antihyperglycemic agents and/or insulin improves glycemic control, preventing

diabetes complications. Several factors facilitate medication adherence such as sufficient
literacy and numeracy skills, and ones motivation, self-efficacy, financial/practical support
and resources. Populations with fewer facilitators and more barriers to adherence (e.g.,
limited diabetes knowledge, negative beliefs about medication) often have high rates of non-
adherence, and, in turn, high rates of suboptimal glycemic control. We used the Adherence to
Refills and Medications Scale for Diabetes (ARMS-D) to assess the relationship between race
and adherence among adults with type 2 diabetes (T2D).

Methods:

We analyzed cross-sectional data from English-speaking adults with T2D recruited for a
medication adherence trial at Vanderbilt University Medical Center in Nashville, TN. We
collected self-reported demographics (age, gender, race, education, income), diabetes
characteristics (diabetes duration, insulin status), and responses to the Brief Health Literacy
Screen (BHLS), Subjective Numeracy Scale (SNS), and ARMS-D. We performed a hierarchical
regression model to examine the association between race and medication adherence
adjusting for demographics, diabetes characteristics, and health literacy and numeracy skills.

Results:

Participants (N=151) were 55.310.9 years old, 60% female, and 24% non-White (69% were
Black/African American), with 15.21.8 years of education, 25% annual incomes < $40K, and
43% on insulin. In the unadjusted model, being non-White was significantly associated
with less adherence ( =-.31 p=-.29 p < .01), explaining 9.4% of the variance in adherence.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2619

After adjusting for covariates, being non-White remained significantly associated with being
less adherent (=-.29 p < .01).

Conclusions:

Non-Whites were less adherent to diabetes medications than Whites after accounting for
sociodemographics, diabetes characteristics, and literacy/numeracy skills often associated
with medication adherence. Disparities in adherence may contribute to persistent disparities
in glycemic control and other diabetes outcomes.

CORRESPONDING AUTHOR: Abel B. Belachew, MA, Marshall University, Huntington, WV,

25701; [email protected]
S2620 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D140 10:00 AM-11:00 AM

LINGUISTIC ACCULTURATION AND ADHERENCE TO CARDIAC MEDICATIONS AMONGST

HISPANICS IN THE UNITED STATES

Noor Habboosh, B.A.1, Marin Kautz, B.A.1, Carmela Alcantara, PhD2, Ian Kronish, MD, MPH3

1
Center for Behavioral Cardiovascular Health, Columbia University Medical Center, New York,
NY; 2Columbia University, New York, NY; 3Columbia University Medical Center, Center for
Behavioral Cardiovascular Health, New York, NY

Background: According to the Hispanic Paradox, Hispanics in the US have a lower rate of
cardiovascular disease mortality and longer life expectancy among the three largest
racial/ethnic groups despite a higher prevalence of CVD risk factors. Differential adherence
among racial/ethnic groups to cardiovascular medications might help explain this paradox. We
tested the independent association of Hispanic ethnicity and linguistic acculturation with
medication adherence.

Methods: Between 2013 and 2015, we enrolled a consecutive cohort of 654 English and
Spanish-speaking patients who were evaluated for suspected acute coronary syndrome (ACS)
in the emergency department of an academic medical center in New York City. Participants
self-reported ethnicity and first language at enrollment. Medication adherence was assessed 1
month after discharge by asking How much of the time do you take your heart medications?
Participants who reported missing their medications at least some of the time were
categorized as nonadherent. Logistic regression was used to determine the association
between Hispanic ethnicity and language with and with adjusting for age, sex, race (white vs
non-white), education, comorbidity (Charlson score), prior history of established coronary
artery disease, discharge diagnosis (ACS vs non-ACS), and depressive symptoms (8-item
Patient Health Questionnaire score).

Results: The mean age was 62 13 years, 43% were women, 73% were White, 55% were
Hispanic, and 35% had less than a high school education. Among Hispanics, 19% identified
English as their first language. Overall, a lower proportion of Hispanics than non-Hispanics was
nonadherent (11.9% vs 18.3%, p=0.03). Amongst Hispanics, a greater proportion of those who
spoke English as a first language versus those who did not speak English as a first language
were nonadherent (23.1% vs 9.3%, p=0.002). In unadjusted models, Hispanic patients were
less likely to be nonadherent (OR 0.60, 95%CI 0.39, 0.94; p=0.03). In a model adjusting for first
language and other covariates, first language as English (AOR 2.30, 95% CI 1.12, 4.76; p=.02),
but not Hispanic ethnicity (AOR 1.09, 95%CI 0.55, 2.16; p=0.80) was associated with increased
likelihood of nonadherence. In an adjusted model limited to participants who did not speak
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2621

English as a first language, increasing ability to speak English well was associated with
increased likelihood of nonadherence (p=0.002).

Conclusion: Hispanics with increasing English language acculturation are at increased risk for
nonadherence to cardiovascular medications.

CORRESPONDING AUTHOR: Noor Habboosh, B.A., Center for Behavioral Cardiovascular

Health, Columbia University Medical Center, New York, NY, 10032;
[email protected]
S2622 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D141 10:00 AM-11:00 AM

MHEALTH TECHNOLOGY INTERVENTIONS FOR ANTIHYPERTENSIVE MEDICATION ADHERENCE:

A SCOPING REVIEW

Cheryl Griffith, MSN RN-BC1, Rebecca J. Bartlett Ellis, PhD, RN, ACNS-BC 2

1
Indiana University School of Nursing, Brownsburg, IN; 2Indiana University School of Nursing,
Indianapolis, IN

The purpose of this scoping review was to identify and describe how mHealth technologies
have been used in interventions to support antihypertensive medication adherence in adult
patients. Google Scholar, Ovid, and EBSCOhost databases were searched to find primary
studies of adult patients published in English between January 2010 and June 2016. Extracted
data included: descriptions of mHealth interventions; measurement of medication adherence;
whether adherence data were shared with patients and/or providers; and patient satisfaction
with mHealth intervention.

Ten studies met inclusion criteria and were included in this review. Reviewed studies used
different combinations of Smartphone technologies as an intervention: unidirectional text
messaging where patients only received messages (n=8); bi-directional text messaging (i.e.
patients received and sent messages) (n=3); scheduled reminders and alerts (n=6); real-time
reminders (n=2); voice messaging (n=2); email (n=4); app (n=1); and internet of things (IoT)
(n=2). A combination of self-reporting (n=6) and/or electronic/manual pill counts (n=1/n=4)
were used to measure medication adherence and this data was emailed to providers either on
a weekly or bi-monthly basis (n=5). Adherence data was used to send text-message reminders
for patients to pick up medication refills (n=1) or notify patients who had not accessed their
electronic pillbox within a prescribed timeframe (n=1). Provider satisfaction of shared data
was measured in one study. Patient satisfaction with the intervention was measured in five
studies.

This review found relatively few published studies that used mHealth interventions for
antihypertensive medication adherence. Of those published, the mHealth interventions varied
in the technologies used to support adherence. Further studies and meta-analysis linking
behavior change techniques with mHealth technology components would be useful as the
knowledge base of mHealth interventions grow. Future research is needed to determine
which mHealth technologies and technology strategies are the most effective and for whom,
along with exploring how mHealth can be leveraged to provide patients with feedback
regarding their medication adherence behavior.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2623

CORRESPONDING AUTHOR: Cheryl Griffith, MSN RN-BC, Indiana University School of Nursing,
Brownsburg, IN, 46112; [email protected]
S2624 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D142 10:00 AM-11:00 AM

THE ASSOCIATION OF MARITAL CONFLICT WITH ADHERENCE TO LONG-TERM OXYGEN

THERAPY AMONG CHRONIC LUNG DISEASE PATIENTS

Ellen M. Kinner, MA1, Karin F. Hoth, PhD2, Frederick Wamboldt, MD3, Jeff Swigris, DO, MS 3,
Robert Sandhaus, MD, PhD3, Mark Aloia, PhD4, Kristen Holm, PhD3

1
University of Iowa, Iowa City, IA; 2University of Iowa Carver College of Medicine, Iowa City,
IA; 3National Jewish Health, Denver, CO; 4National Jewish Health and Philips, Denver, CO

The Association of Marital Conflict with Adherence to Long-term Oxygen Therapy among
Chronic Lung Disease Patients

Ellen M Kinner1, Karin F Hoth2, Fred Wamboldt3, Jeff Swigris3, Robert Sandhaus3, Mark
Aloia3,4, Kristen Holm3

1
Department of Psychological and Brain Sciences, University of Iowa; 2 Department of
Psychiatry, University of Iowa; 3Department of Medicine, National Jewish Health; 4Philips

Background: Long-term oxygen therapy (O2) is often a critical component for the treatment of
chronic lung diseases such as chronic obstructive pulmonary disease (COPD) and interstitial
lung disease (ILD). Rates of adherence to long-term O2 therapy are below 50%. No studies
have yet examined the potential impact of interpersonal relationships on O2 adherence. The
aim of this study is to examine the impact of marital conflict on O2 adherence in patients with
chronic lung disease.

Methods: Patients with COPD or ILD who were married and had a prescription for 24-hour
oxygen use completed self-report questionnaires regarding their demographic and medical
history, interpersonal relationships, mood, and oxygen adherence. Conflict was assessed using
the conflict subscale of the Quality of Relationship Inventory (QRI). Oxygen adherence was
coded as adherent or non-adherent based on any admission of non-adherence on six
questionnaire items. Logistic regression was used to determine whether conflict was
associated with adherence to oxygen after adjusting for age, education, smoking history, type
and duration of lung disease, symptoms of depression and anxiety, and length of marriage.

Results: Data were analyzed from 132 patients (mean age = 70.14; 75% COPD, 25% ILD; mean
duration of oxygen use = 8.13 years). 76% of the sample was categorized as being adherent.
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2625

Conflict had a negative association with O2 adherence (OR = .36, 95% CI = .18 .75, p = .006)
after adjusting for demographic and medical characteristics as well as length of marriage and
symptoms of depression and anxiety.

Conclusions: Patients who reported less conflict in their marital relationship were more likely
to be adherent to their O2 than patients who reported higher levels of conflict. These findings
suggest that patients with higher levels of conflict may be at risk for poorer adherence. Given
that marital conflict is a potentially malleable factor, it may be an important target for
psychosocial interventions that aim to promote adherence to O2 in patients with chronic lung
disease who are married.

CORRESPONDING AUTHOR: Ellen M. Kinner, MA, University of Iowa, Iowa City, IA, 52242;
[email protected]
S2626 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D143 10:00 AM-11:00 AM

ACUPUNCTURE AND MINDFULNESS-BASED STRESS REDUCTION AMONG ADULT SURVIVORS: A

RANDOMIZED WAIT-LIST CONTROLLED STUDY

Catherine Dempsey, PhD1, Margaret A. Chesney, PhD2, Lixing Lao, PhD, MB3, Mary Bahr-
Robertson, BA4, Patti Vegella, MS, MA5, Brian Berman, MD6, Elizabeth Kimbrough, PhD7

1
Uniformed Services University, Bethesda, MD; 2UCSF School of Medicine, Menlo Park, CA;
3
University of Hong Kong, School of Chinese Medicine, Pokfulam, Hong Kong, N/A, China
(People's Republic); 4Center of Integrative Medicine, Baltimore, MD; 5Henry Jackson
Foundation, Washington, DC; 6University of Maryland Baltimore, Baltimore, MD; 7Center of
Integrative Medicine, Department of Family and Community Medicine, University of Maryland
School of Medicine, Baltimore, MD

Background. Adult survivors of childhood sexual abuse (CSA) often suffer psychological
consequences: Post-Traumatic Stress Disorder (PTSD) and depression. To relieve these
symptoms, this study compares efficacies of Mindfulness-Based Stress Reduction,
Acupuncture, and Wait-List Control in CSA survivors.

Methods. Randomized were 83 CSA survivors to Mindfulness-Based Stress Reduction, 2.5

hours weekly/ 8 weeks; Acupuncture, 35-40 minutes bi-weekly/ 8 weeks; or Wait-List Control.
Outcomes were assessed at baseline, four, eight, and 12 weeks, post-randomization. Primary
outcomes were measured using the PTSD Checklist and Beck Depression Inventory. Secondary
outcome assessment utilized the Pittsburgh Sleep Quality Index and the Mindful Attention
Awareness Scale.

Results. Large effect sizes were indicated with Acupuncture for PTSD symptoms, versus the
controls (Cohens d = 1.38). Moderate effects for depression with Mindfulness-Based Stress
Reduction (Cohens d = 0.99) were reported. Both treatments significantly improved
mindfulness, sustainable for 12 weeks (F = 5.85; p = .00101). Acupuncture significantly
improved PTSD symptoms (F =3.37; p = .0052) and sleep quality (F = 3.74; p = .0076) for up to
12 weeks. Although not as sustainable, both treatments also improved depressive symptoms
(F = 2.78; p = .0165). This study demonstrates and compares the benefits of Acupuncture and
Mindfulness-Based Stress Reduction to conventional treatments.

CORRESPONDING AUTHOR: Catherine Dempsey, PhD, Uniformed Services University,

Bethesda, MD, 20814; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2627

D144 10:00 AM-11:00 AM

COMORBIDITY AND EATING BEHAVIOR IN INDIVIDUALS WITH SYMPTOMS OF

AVOIDANT/RESTRICTIVE FOOD INTAKE DISORDER (ARFID)

Hana F. Zickgraf, MA1, Jordan Ellis, MA2, Martin E. Franklin, Ph.D.3, Matthew C. Whited, PhD,
Licensed Psychologist2

1
University of Pennsylvania, Philadelphia, PA; 2East Carolina University, Greenville, NC;
3
University of Pennsylvania Perelman School of Medicine, Psychiatry, Philadelphia, PA

Avoidant/Restrictive Food Intake Disorder (ARFID) is an eating disorder (ED) diagnosis for
individuals whose eating behavior leads to failure to meet nutritional/energy needs and/or
significant psychosocial impairment. ARFID was added to DSM-5 to capture impairing eating
pathology without compensatory behavior or shape/weight concerns. DSM-5 describes three
types of eating disturbance that can lead to ARFID symptoms: picky eating (PE), low appetite/
interest in eating, and fear of negative consequences from eating (e.g., choking, vomiting).

A 2016 study found that adults whose PE led to the nutrition/energy or psychosocial
symptoms of ARFID reported just as much distress and comorbidity as individuals screening
positive for risk of anorexia or bulimia (AN/BN).The current study extends these findings to
ARFID symptoms due to other ARFID manifestations.

Participants were screened for ARFID due to any of the three eating behaviors via self-report
measures. Analyses were replicated in two samples: 1219 college students; 450 adults
recruited specifically for the eating difficulties of interest. Participants responded online to the
Nine-item ARFID screen (NIAS), which assesses the three eating restrictions, and the ARFID
symptom checklist (ARFID-cl), which assesses DSM-5 criteria to establish a possible diagnosis.
Four groups were created: 1) No restrictions, 2) restrictions/no ARFID symptoms, 3)
subclinical ARFID, and 4) full ARFID criteria. Groups were compared on psychological
comorbidities (anxiety, depression, eating quality of life) with bootstrapped ANCOVAs,
covarying self-reported symptoms of AN/BN.
S2628 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

In each sample, ARFID symptom group was associated with anxiety and depression. Eating
quality of life impairment was assessed in the second sample and significantly differed across
the ARFID symptom groups. Severity of symptoms/impairment increased with level of ARFID
symptomatology with small-moderate to large effect sizes observed (p2 = .04 .27). Post-hoc
comparisons in the general sample revealed differences in reported symptoms/behaviors
between ARFID-symptom levels, with the most consistent differences between groups with
no ARFID symptoms (groups 1 and 2) and the groups with some degree of ARFID symptoms
(groups 3 and 4). In the selected sample, only groups on either end of the continuum (groups
1 and 4) differed on symptoms of anxiety and depression. In this sample, each group differed
significantly from all other groups on quality of life impairment.

These findings highlight the need for more recognition of, and research on, ARFID. ARFID is an
ED that is associated with a distinct pattern of eating behavior and a high level of comorbidity
and distress.

CORRESPONDING AUTHOR: Hana F. Zickgraf, MA, University of Pennsylvania, Philadelphia, PA,

19104; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2629

D145 10:00 AM-11:00 AM

COMPARISON OF SELF-RATING OF COGNITION AND DEPRESSION IN PATIENTS WITH MAJOR

DEPRESSION

Kevin Bera, Bachelor of Science

University of Southern California, Irvine, CA

Background: Traditionally clinicians have utilized neuropsychological tsts to evaluate cognitive

impairment in patients with major depression. However, the burden of many of these
instruments to both clinicians and patients make these tools difficult to access for the
majority of patients. These burdens include the time it takes for these tests to be
administered, lack of coverage for these tests by insurance carriers and the difficulty in finding
a trained clinician who can administer these test. The recently developed Massachusetts
General Hospital Cognitive and Physical Functions Questionnaire (CPFQ) is a 7 item self-
questionnaire that has been found to have strong internal consistency. To our knowledge this
is the only self-Rating scale for cognitive function in existence. Many self-rating scales for
depression have been in existence for many years. The Beck Depression Scale (BDI) is one of
the most common in use. We set out to assess patients responses when both scales were
administered and to see if there were any particular correlations between items on these two
scales.

Methods: 100 adult patients with Major Depression will be evaluated during their regular
scheduled outpatient appointment in a mental health clinic. Diagnosis will be based on their
medical record. These patients did not carry any other primary psychiatric disorder on Axis I.
Everyone will be administered both the CPFQ and BDI. We will also collect patients age, sex
and time since first diagnosed with major depression. To date 36 patients have been
evaluated and it is estimated that by November 15, 2016, data from 100 patients will have
been collected and ready for full analysis.

Results: We will evaluate responses to all 7 items on the CPFQ and all 21 items on the BDI
initially to see if there is any correlation of items with patients age, sex and duration of major
depression. Then direct analysis and comparison of the items on the CPFQ and BDI will occur
to determine if there are particular relationships that emerge between cognitive items and
depressive items. All analysis complete by December 15, 2016 for presentation at the Society
of Biological Medicine Meeting in San Diego from March 29 to April 1, 2017.

Conclusion: We will discuss our findings comparing patients with major depression self-
response on the CPFQ and self-response on the BDI. We will determine with items on each
scale have a high correlation with each other. We will also determine how user friendly the
S2630 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CPFQ is in a general outpatient mental health clinic. This will help clinicians determine the
utility of a self-rating cognition scale in major depression patients. Most importantly clinicians
will now be able to assess cognition in a rapid fashion and determine if the treatment being
prescribed is resulting in improved psychosocial functioning as it is now felt that cognition
improvement is a more important marker than depression with regards to patient outcome.

CORRESPONDING AUTHOR: Kevin Bera, Bachelor of Science, University of Southern California,

Irvine, CA, 92618; [email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2631

D146 10:00 AM-11:00 AM

EVENING SALIVARY CORTISOL, DEPRESSION, AND SYMPTOMS OF CHRONIC FATIGUE

SYNDROME/ MYALGIC ENCEPHALOMYELITIS

Sara Milrad, M.S.1, Daniel L. Hall, Ph.D.2, Devika Jutagir, M.S.3, Emily Lattie, Ph.D.4, Sara Czaja,
Ph.D.5, Dolores Perdomo, Ph.D.6, Mary Ann Fletcher, Ph.D.7, Nancy Klimas, MD8, Michael
Antoni, Ph.D.3

1
University of Miami, Miami, FL; 2Harvard Medical School/MGH/BIDMC, Wellesley, MA;
3
University of Miami, Coral Gables, FL; 4Northwestern University, Chicago, IL; 5University of
Miami Department of Psychiatry and Behavioral Sciences, Miami, FL; 6University of Miami,
miami, FL; 7Nova Southeastern University, Fort Lauderdale, FL; 8Nova Southeastern
University, Ft. Lauderdale, FL

Introduction: Chronic fatigue syndrome/ myalgic encephalomyelitis (CFS/ME) is commonly

comorbid with depression. Individuals suffering from CFS/ME and depression can also share
some similar somatic symptoms and markers of neuroendocrine dysfunction, such as fatigue
and elevated evening cortisol. Depression is correlated with elevated evening cortisol in
otherwise healthy individuals, but less is known about cortisol associations with depression
and CFS/ME symptoms in CFS/ME patients.

Methods: Baseline data were drawn from diagnosed CFS patients (N = 150) participating in
studies testing the efficacy of a remotely-delivered cognitive behavioral stress management
(CBSM) intervention. CFS/ME symptom and depression questionnaires and salivary cortisol
were used for multiple regression analyses using SPSS.

Results: The majority of the sample (69%) had clinically significant depressive symptoms (CES-
D 16). Depressive symptoms (= 0.212, p= 0.019) and CDC CFS Symptom Severity (= 0.309,
p= 0.036) were positively correlated with evening cortisol, while perceived stress, fatigue,
tension/anxiety, and sickness impact were not (all ps>0.05).

Discussion: Results highlight the importance of considering neuroendocrine processes that

are associated specifically with both CFS/ME symptoms and depression. Further mechanism-
based, longitudinal research might elucidate the underlying etiology of both CFS/ME and
depression symptoms.
S2632 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

CORRESPONDING AUTHOR: Sara Milrad, M.S., University of Miami, Miami, FL, 33146;
[email protected]
ann. behav. med. (2017) 51 (Suppl 1):S1S2867 S2633

D147 10:00 AM-11:00 AM

EXPLORING DEMORALIZATION AMONG PATIENTS WITH MEDICAL ILLNESS

Jay T. Wiles, M.A. in Clinical Psychology1, Linda Baum, Ph.D.2, Judith Johnson, Ph.D.3

1
Regent University, Richmond, VA; 2Regent University, Virginia Beach, VA; 3Regent University,
Norfolk, VA

The experience of demoralization in patients with medical illness has been extensively
documented; however, limited research has highlighted the bidirectional relationship
between demoralization and medical illness. The current study examined the relationship
between medical illness and experience of demoralization (as measured by the MMPI-2-RF).
Specifically, the purpose of this study was to explore the multifaceted relationship between
demoralization, psychological symptoms, and physical illness. The study sample consisted of
279 adults who completed a valid MMPI-2 or MMPI-2-RF at a university-based outpatient
psychological services center. Of the 279 participants, 154 comprised a medical patient
sample. Results of this study suggest significant differences in mean Demoralization (RCd) T-
scores for medical patients with comorbid depressive or anxiety related psychological
disorders when compared to medical patients without comorbid depressive or anxiety related
disorders. Medical patients with comorbid psychological disorder(s) (i.e., inclusive of all
clinical disorders) were found to have significantly higher mean Demoralization (RCd) T-scores
when compared to those without a comorbid psychological disorder. The results of this study
help to expand the understanding and conceptualization of the interconnectedness between
medical illness and experience of demoralization. Based on the results of this study,
implications and future research directions are discussed.

CORRESPONDING AUTHOR: Jay T. Wiles, M.A. in Clinical Psychology, Regent University,

Richmond, VA, 23233; [email protected]
S2634 ann. behav. med. (2017) 51 (Suppl 1):S1S2867

D148 10:00 AM-11:00 AM

FACEBOOK FOR SUPPORTING LIFESTYLE INTERVENTIONS FOR PEOPLE WITH MENTAL ILLNESS

John Naslund, MPH1, Kelly Aschbrenner, PhD2

1
Dartmouth College, Lebanon, NH; 2Dartmouth College, Nashua, NH

Objective: Facebook could support efforts to address elevated obesity rates, high
cardiovascular risk, and resulting early mortality that disproportionately impact people with
mental illness. In this exploratory study, we examined whether actively engaging in a
Facebook group as part of a community-based lifestyle intervention was associated with
clinically significant weight loss and fitness outcomes among individuals with mental illness.

Methods: We invited participants with mental illness and obesity enrolled in a 6-month group
lifestyle program delivered through a community mental health center to join a private
Facebook group to promote group support for achieving shared physical activity and weight
loss goals. Two volunteer peers moderated the Facebook group. Participants interactions in
the Facebook group were defined as active engagement including posts, comments, or likes.
We measured the proportion of participants who achieved 5% weight loss or improved
fitness (increase >50 meters on the 6-Minute Walk Test). We also explored content of the
Facebook posts.

Results: Participants (n=25) were 52% female and had major depressive disorder (44%),
bipolar disorder (36%), or schizophrenia (20%), a mean age of 49.2 years (SD=11.8), and a
mean BMI of 37.3 kg/m2 (SD=8.1). Nineteen (76%) participants joined the Facebook group.
These participants contributed 956 interactions consisting of 326 unique posts, 275
comments, and 355 likes. Participants who achieved 5% weight loss (37%) were more likely
to interact in the Facebook group (mean=23.0 interactions, SD=8.3) compared to participants
who did not achieve 5% weight loss (mean=3.9 interactions, SD=5.9; t=-2.9; df=17; p <
0.01). Clinically significant improved fitness was not associated with interactions in the
Facebook group. Facebook posts contained content about healthy eating (30%), program
reminders (19%), personal sharing about successes an