ORIGINAL RESEARCH

Factors that influence exercise activity among

women post hip fracture participating in the
Exercise Plus Program

Barbara Resnick 1 Abstract: Using a social ecological model, this paper describes selected intra- and interpersonal
Denise Orwig 2 factors that influence exercise behavior in women post hip fracture who participated in the
Christopher D’Adamo 2 Exercise Plus Program. Model testing of factors that influence exercise behavior at 2, 6 and 12
Janet Yu-Yahiro 3 months post hip fracture was done. The full model hypothesized that demographic variables;
William Hawkes 2 cognitive, affective, physical and functional status; pain; fear of falling; social support for
exercise, and exposure to the Exercise Plus Program would influence self-efficacy, outcome
Michelle Shardell 2
expectations, and stage of change both directly and indirectly influencing total time spent
Justine Golden 2
exercising. Two hundred and nine female hip fracture patients (age 81.0 ± 6.9), the majority
Sheryl Zimmerman 4
of whom were Caucasian (97%), participated in this study. The three predictive models tested
Jay Magaziner 2 across the 12 month recovery trajectory suggest that somewhat different factors may influence
1
University of Maryland School exercise over the recovery period and the models explained 8 to 21% of the variance in time
of Nursing, 655 West Lombard
spent exercising. To optimize exercise activity post hip fracture, older adults should be helped
Street, Baltimore, MD,21201, USA;
2
University of Maryland School of to realistically assess their self-efficacy and outcome expectations related to exercise, health
Medicine, Howard Hall, Redwood care providers and friends/peers should be encouraged to reinforce the positive benefits of
Street, Baltimore MD 21201, USA;
3
Department of Orthopaedic
exercise post hip fracture, and fear of falling should be addressed throughout the entire hip
Surgery, Union Memorial Hospital, fracture recovery trajectory.
Baltimore, USA; 4 University of North Keywords: hip fracture, exercise, self-efficacy, outcome expectations, recovery
Carolina Chapel Hill, 301 Pittsboro
St., CB#3550, Chapel Hill, NC
27599-3550, USA
Introduction
While there has been limited work in the implementation of exercise activities
post hip fracture, there is some support to suggest important benefits for these
individuals. Specifically, for those who have sustained a hip fracture, regular
exercise (resistive and/or aerobic) improves mobility and quadriceps strength
(Tinetti et al 1999; Mangione et al 2005; Tsauo et al 2005; Jones et al 2006),
increases walking speed (Henderson et al 1992; Habris et al 1995; Jones et al
2006), and weight-bearing ability (Habris et al 1995). Despite the potential benefits
of exercise, however, the majority of older adults do not participate in sufficient
physical activity or exercise(Centers for Disease Control and Prevention Behavioral
Risk Factor Surveillance System 2006), including those who have sustained a hip
fracture.
A social ecological model is one of the most comprehensive approaches to
explaining exercise behavior in older adults (Sallis 2003; Sallis et al 2006; United
Correspondence: Barbara Resnick
University of Maryland School of Nursing, States Department of Health and Human Services 2000; Medley and Syme 2000).
655 West Lombard Street, Baltimore, MD, Specifically a social ecological model suggests that an individual’s behavior is
21201, USA
Tel + 410 706 5178
affected by a wide sphere of influences: intrapersonal, interpersonal, institutional/
Email [email protected] organizational, public policy, and the environment.

Clinical Interventions in Aging 2007:2(3) 413–427 413

© 2007 Dove Medical Press Limited. All rights reserved
Resnick et al

Intrapersonal factors Precontemplation, Contemplation, Preparation, Action and

Intrapersonal factors include such things as physical and Maintenance. Precontemplation occurs when the individual
cognitive status. Age-related dysfunction of frontal systems, has no intention to change behavior. Contemplation occurs
for example, can result in deficits in planning, organization, when the individual is thinking about changing behavior, but
self-control, and awareness of problems, which are likely to not committed to the behavior change. Preparation refers to
affect the ability to perform functional activities or engage the period when the individual intends to change behavior
in regular exercise (Sarkisian et al 2000; Norwalk et al 2001; sometime soon and is actively preparing. Action occurs when
Wang et al 2002). the individual has changed behavior recently (within the
Other intrapersonal factures, such as physical and mental past six months). Maintenance occurs when the individual
health status have been noted to influence self-efficacy and has maintained behavior change for a period longer than six
outcome expectations, with low mood disturbance and better months. These stages are directly related to exercise behavior.
overall mental health associated with stronger self-efficacy As individuals progress through the stages of change they
and outcome expectations (Gecht et al 1996; Kurlowicz report exercising more, are more fit based on physiological
1998; Perkins and Jenkins 1998). Mental health influences measures and have stronger self-efficacy expectations (Godin
exercise activity such that those who were depressed were et al 2004; Ackerman et al 2005). Likewise, self-efficacy
less likely to exercise (Oliver and Cronan 2002; Bonnet et al and outcome expectations increase from precontemplation
2005; Mangione et al 2005; Forkan et al 2006). Perceived to maintenance in older adults (Resnick and Nigg 2003;
physical health status also has been associated with adherence Schumann et al 2003; Godin et al 2004; Ackermann et al
to exercise in older adults (Sin et al 2002; Brown et al 2003; 2005; Riebe et al 2005).
Munneke et al 2003; Lee and Laffrey 2006). Further there is There is a relationship between self-efficacy and outcome
evidence that such things as gait and balance, functional sta- expectations with stage of change. Consistently, self-efficacy
tus, pain, or fear of falling may further influence an individu- and outcome expectations increase from precontemplation
als’ willingness to engage in exercise activities (Cumming to maintenance in older adults (Gorely and Gordon 1995).
et al 2000; Bruce et al 2002; Li et al 2003; Delbacre 2004; The older adult’s beliefs about his or her ability to exercise
Fletcher and Hirdes 2004; Martin et al 2005). and the benefits associated with exercise influences whether
or not the individual is willing to initiate and/or adhere to
Interpersonal factors an exercise program (ie, stage of change). Self-efficacy and
Two overriding theories help explain the interpersonal outcome expectations therefore can have both a direct and
interactions that can influence exercise behavior and behavior indirect effect on exercise through stage of change.
change as related to exercise. The first is social cognitive Another important interpersonal factor influencing
theory and the theory of self-efficacy (Bandura 1997)which participation in exercise is social support from friends,
suggests that the stronger the individual’s self-efficacy and family, and experts. Consistent with the theory of self-
outcome expectations, the more likely it is that he or she efficacy, when there is encouragement to exercise from
will initiate and persist with a given activity. Self-efficacy family, friends, and/or experts, older adults are more likely to
expectations are the individuals’ beliefs in their capabilities participate in regular exercise activities (Resnick et al 2002;
to perform a course of action to attain a desired outcome, Sharma et al 2005; Greene et al 2006; Lim and Laffrey 2006;
whereas outcome expectations are the beliefs that a certain Lippke and Ziegelmann 2006; Resnick et al 2006).
consequence will be produced by personal action. Both
self-efficacy and outcome expectations play an influential Institutional/organization
role in the adoption and maintenance of exercise behavior and environment and policy
in older adults (Brassington et al 2002; Gyurcsik et al 2003; The organizational structure and environment the older
Estabrooks et al 2005; Li et al 2005; McAuley et al 2006). adult lives in and the policies that impact their communities
The second theory is the transtheoretical model (TTM) can influence exercise activities as well (Takano et al 2002;
(Prochaska and Velicer 1997), an integrative model of Iwarsson 2005). Environments that facilitate function have
intentional behavior change. The central construct of the been noted to be important factors in prevention of functional
TTM is stage of change (SOC), which describes behavior decline (Takano et al 2002; Crews 2005; Iwarsson 2005)
change as a progression through a series of stages. Individuals and enabling people to achieve their highest level of func-
can be classified into one of the following five stages: tion and well-being (Humpel et al 2002; Takano et al 2002).

414 Clinical Interventions in Aging 2007:2(3)

 Exercise Plus Program

Unfortunately, designated exercise space is generally limited A detailed description of eligibility and recruitment has
in home and facility based settings (Mihalko and Wickley been described elsewhere (Buie et al 2001). Briefly,
2003) and outside walkways, hallways, and common areas eligible patients were female, 65 years of age or older,
are seldom used to promote physical activity. While there community-dwelling at the time of fracture, had a non-
are general guidelines to encourage all adults to engage in pathologic fracture within 72 hours preceding admission,
30 minutes daily of physical activity (Centers for Disease and surgical repair of the hip fracture. Medical exclusions
Control and Prevention, Merck Institute of Aging and Health included evidence of symptomatic cardiovascular disease,
2004; Thompson 2003; National Blueprint for Increasing neuromuscular conditions limiting exercise, or other
Physical Activity 2002), there are no policies to promote this conditions that increased risk when exercising home alone.
and no specific guidelines post hip fracture. Participants had to be walking without human assistance
Despite existing knowledge on the factors and theories prior to the fracture and score ⱖ20 on the Folstein Mini
related to exercise, the fact remains that older adults do not fre- Mental Status Exam (Folstein et al 1975). Also, informed
quently exercise. Encouraging exercise is especially important consent and baseline measures had to be obtained within
for a post-hip fracture population, given that this is likely to 15 days of the fracture to be eligible for randomization.
optimize recovery. In recognition of this possibility, this in- Institutional Review Board approvals were obtained from
vestigative team undertook a clinical trial to motivate exercise the University of Maryland, School of Medicine as well as
behavior in older adults post-hip fracture, which included three the study hospitals, and all enrolled subjects provided their
treatment arms: a home-based exercise program (Exercise), a own informed consent.
motivational intervention (Plus), the combination of the two A total of 209 female hip fracture patients were consented
(Exercise Plus Program), compared with routine care. within 15 days of the hip fracture. The majority of the par-
The purpose of this paper is to describe selected intra- ticipants were Caucasian (97%), and the average age of the
and interpersonal factors that influence exercise behavior participants was 81.0 ± 6.9. Approximately one third (34%) of
in women post hip fracture who participated in this project. the participants were married. The remaining were widowed
Model testing of factors that influence exercise behavior (57%), never married (3%), or divorced or separated (6%). The
at 2, 6, and 12 months post hip fracture was done and average number of years in school was 12.2 ± 2.9.
consideration given to consistency and differences noted
between these models. The full model hypothesized that The intervention: The Exercise
demographic variables; cognitive, affective, physical and Plus Program
functional status; pain; fear of falling; social support for The Exercise Plus Program and theoretical premise of the
exercise, and exposure to the Exercise Plus Program would program has been described in detail elsewhere (Resnick
influence self-efficacy, outcome expectations, and stage of et al 2002a, 2007). Briefly, the Exercise component of the
change both directly and indirectly influencing total time Exercise Plus Program is a home based exercise intervention
spent exercising. The 89 hypothesized relationships are administered by exercise trainers which incorporates an
demonstrated in Figure 1. aerobic exercise program using a Stairstep (Yu-Yahiro et al
2001; Resnick et al 2007), a comprehensive strengthening
Methods program that covers all muscles groups, and stretching
Study design exercises which are part of the warm up and cool down
Data were derived from a randomized clinical trial using periods. Participants were encouraged to perform aerobic
a repeated measure two by two design with participants activity at least 3 days per week and strength training two
randomized to one of four groups: exposure to the Exer- days per week. The Plus component was also implemented
cise Plus Program (exercise plus motivation), the Exercise by an exercise trainer and included a self-efficacy based
only component of the Exercise Plus Program, the Plus intervention using education, verbal encouragement
(or motivational) only component of the Exercise Plus through goal setting and positive reinforcement, removal
Program, or routine care. of unpleasant sensations associated with exercise, and
individualized cueing (Resnick et al 2002a, 2007). In all
Sample treatment groups visits from the trainer were initially twice
Participants were recruited from 6 hospitals in the greater a week for the first three months, once a week for the next
Baltimore area between July 2000 and September 2004. three months, and then once a month in the final six months of

Clinical Interventions in Aging 2007:2(3) 415

Resnick et al

Pain Fear Trainer

age
Stage of
Self-Efficacy Change
Cognitive
Status

Physical
Status Outcome
Expectations Exercise
Time
Mental
Status

Charlson

Treatment
Social Group
Support
Friends
Social
Social Support Depression
Support Expert
Family
Figure 1 Full hypothesized model.

the program. On weeks when there was no face-to-face visit, Exercise Scale (SEE), the Outcome Expectations for Exercise
for those exposed to the Plus component, weekly telephone (OEE) scale, and the Stage of Change Questionnaire. The
calls were made to answer questions about exercise and Yale Physical Activity Survey was used to measure time spent
encourage adherence. exercising. A description of the measures and reliability and
validity is provided in Table 1.
Measures
Follow up data was collected at 2, 6, and 12 months post Data analysis
hip fracture. Measures addressing intrapersonal factors Descriptive statistics were done to describe the participants.
included demographic information, the Short Form Health Model testing was completed to establish the factors that
Survey (SF-36), the Centers for Epidemiologic Studies and influence exercise behavior at 2, 6, and 12 months post hip
Depression Scale, a single item fear of falling question, the fracture using structural equation modeling and the Amos
numeric rating scale for pain; interpersonal factors included statistical program. The sample covariance matrix was used
Social Support for Exercise Scale, the Self-efficacy for as input and a maximum likelihood solution sought. The

416 Clinical Interventions in Aging 2007:2(3)

 Exercise Plus Program

Table 1 Description of study measures

Measure Description Score range and Reliability and validity
interpretation
Self-efficacy for A nine item 0 (no confidence) Evidence of internal
Exercise: measure that to 10 (high consistency (alpha=0.93),
(Resnick and focuses on self- confidence). and validity based on a
Jenkins 2000) efficacy Higher scores significant relationship
expectations related indicate stronger between efficacy
to the ability to self-efficacy. expectations and
continue to exercise moderate exercise, and
in the face of confirmatory factor
barriers to analysis. (Resnick and
exercising. Jenkins 2000).
Outcome A nine item 1 (strongly Evidence of internal
expectations for measure that disagree) to 5 consistency (alphas
Exercise focuses on the (strongly agree). ranging from 0.88 to 0.93),
(Resnick et al perceived Higher scores and validity based on a
2000, 2001): consequences of indicate stronger significant relationship
exercise for older outcome between outcome
adults. expectations. expectations and
moderate exercise, and
confirmatory factor
analysis. (Resnick and
Jenkins 2000).
The SF-36 (Ware An eight dimension 0 to 14 for mental There is support for the
and Sherbourne measure of health health; and 0 to reliability (Chronbach’s
1992). status that focuses 100 representing alpha for subscales
on: physical the percentage of ranging from 0.75 to 0.86)
functioning, role- total possible score and validity of this
physical, bodily achieved. measure (based on
pain, general health, contrasting groups and
vitality, social factor analysis) when
functioning, role used with older adults
emotional, and (Stewart 1993, 1988;
mental health. The Walters and
8 subscales are Munro 2004).
combined to
constitute mental
and physical health
scores.
Yale Physical A five category 0 to 1440 minutes Evidence of test-retest
Activity Survey physical activity per week. reliability (r = 0.63,
(YPAS) survey that focuses p < 0.001), and validity
(DiPietro et al on time spent in: based on significant
1993) housework, correlations with
caregiving, physiological variables
yardwork, exercise, that are indicative of
and recreational habitual activity
activities performed (Dipietro et al 1993;
during a typical Pescatello et al 1994;
week. Only the Kolbe-Alexander et al
exercise subscale 2006).
was utilized in this
study.
Center for The possible range 0 to 5. Higher Prior use of these
Epidemiological of scores is 0 to 60. scores indicate measures provides
Studies more depressive evidence of their
Depression Scale symptoms. reliability and validity
(CESD) when used with older
(Continued)

Clinical Interventions in Aging 2007:2(3) 417

Resnick et al

Table 1 (Continued)
Measure Description Score Range and Reliability and Validity
Interpretation
(Radloff 1977; adults (Radloff 1977;
Turk and Okifuji Turk and Okifuji 1994
1994). Caracciolo and Giaquinto
2002; Bohannon et al
2003; ).
Numeric Rating A single item 0 (no pain) to 10 Evidence of test-retest
Scale (NRS) for measure that (the worst pain). reliability (Spearman
Pain (Herr and focuses on pain Higher scores rank correlations from
Mobily 1991) over the previous indicate more pain. 0.67 to 0.85) (Taylor et al
week. 2005), and concurrent
validity with other pain
measures (r = 0.56 to 0.90)
(Herr and Mobily 1993;
Herr et al 2004; Ware
et al 2006)
Fear of Falling A single item 0 (no fear) to 4 (a Evidence of validity with
(Jorstad et al measure that lot of fear). Higher fear of falling
2005 Resnick focuses on fear of scores indicate significantly associated
1998) falling. greater fear of with functional
falling. performance in older
adults (Resnick 1998
Jorstad et al 2005)
The Tinetti A 17 item 0 to 26. Higher Evidence of inter-rater
Mobility Scale performance scores indicate reliability (r = 0.90), and
(Tinetti 1986) measure that better mobility. construct validity with a
focuses on mobility significant relationship
and includes: nine between mobility and
balance maneuvers falls (Tinetti 1986).
and eight
assessments related
to gait.
The Social Includes three Possible ranges Evidence of internal
Support for separate subscales from 23 to 67. consistency (alphas
Exercise Habits of the same15 items Lower scores ranging from 0.61 to 0.91)
Scale (Sallis that reflect social reflect lower social and test retest reliability
et al 1987) interactions that support (r = 0.55 to 0.79). Evidence
might influence of validity was based on
exercise behavior statistically significant
from friends, relationships between the
family, and experts. social support scale and
exercise behavior (Sallis
et al 1986) (Resnick
et al 2002b).

chi-square statistic, the normed fit index (NFI), and Steigers is perfect model fit. The NFI is “normed” so that the values
Root Mean Square Error of Approximation (RMSEA) cannot be below 0 or above 1. The RMSEA is a population
were used to estimate model fit. The larger the probability based index and consequently is insensitive to sample size.
associated with the chi-square, the better the fit of the model An RMSEA of <0.10 is considered good, and <0.05 is very
to the data (Bollen 1989; Loehlin 1998). Since the chi-square good (Loehlin 1998). Path significance (ie, significance of
statistic is sample size dependent the chi-square divided by the Lambda values) was based on the Critical Ratio (CR),
degrees of freedom (df) was utilized to control for sample which is the parameter estimate divided by an estimate of the
size effects (Bollen 1989). The NFI tests the hypothesized standard error. A CR >2 in absolute value was considered
model against a baseline model and should be 1.0 if there significant (Arbuckle 1997).

418 Clinical Interventions in Aging 2007:2(3)

 Exercise Plus Program

Results At six months post hip fracture (Figure 3), 12 of the 89

Of the 209 participants initially recruited, 165 women hypothesized paths were significant. Physical and mental
(79%) were available for 2-month assessments, 169 (81%) health, social support from an expert, and treatment group
were available for 6-month follow up, and 155 (75%) were all related to self-efficacy expectations such that those who
available for the 12-month follow up visits. One case was were exposed to any of the treatment groups, had better
deleted post-randomization due to being ineligible (no health, and less support from an expert to exercise, had
surgery was performed post hip fracture). Reasons for stronger self-efficacy expectations. Age, mental health, fear
loss to follow up have been reported elsewhere (Resnick of falling and social support from friends related to outcome
et al pers comm). The mean age of the participants was expectations for exercise. Those who were younger, had
80.7 (SD = 6.9), mean MMSE was 26.7 (SD = 2.8), and the better mental health, more support from friends for exercise,
majority were Caucasian (96%). and less fear of falling had stronger outcome expectations
The time from fracture to first intervention visit from the for exercise. Self-efficacy and outcome expectations were
trainer ranged from 28 to 200 days. While attempts were associated with stage of change such that those with stronger
made on the part of the trainers via weekly telephone calls efficacy expectations were more likely to be in higher stages
to initiate the intervention, participants generally were not of change such as action or maintenance. Stage of change and
willing to have a visit occur prior to 2 months post fracture. treatment group were the only variables to directly relate to
Only one participant had her first visit at 28 days post fracture. time in exercise, with higher stages of change and exposure
By two months, 22 (31%) of the participants had their first to treatment being associated with more time spent in exer-
visit, by three months 44 (62%) of the participants had their cise. All the other significant variables indirectly related to
first visit, and by four months 58 (82%) of the participants exercise time through self-efficacy or outcome expectations
had their first visit. and then stage of change. There was a fair fit of the model to
Table 2 provides descriptive statistics of the variables the data (χ2 = 110.6, df = 38, p = 0.00, ratio 2.9, NFI = 0.74,
under study by treatment (any of the three intervention arms) and RMSEA of 0.09), it explained 8% of the variance of
versus control group. Generally the participants had some exercise behavior at six months post hip fracture.
confidence they could exercise, believed in the benefits of At 12 months post hip fracture, nine of the 89 hypoth-
exercise and exercised about 1.5 to 2 hours weekly. Overall esized paths were significant (Figure 4). Physical health and
they were not depressed and reported fair mental and physical fear of falling related to self-efficacy expectations. Those
health, minimal pain and some fear of falling. with better health and less fear had stronger self-efficacy
Testing of the full 2 month model indicated that out of expectations. Self-efficacy expectations, social support from
the 89 paths hypothesized only 7 were statistically significant an expert, and fear of falling all related to outcome expec-
(Figure 2). Path coefficients for all models are shown in tations for exercise. Those who had stronger self-efficacy,
Table 3. Cognitive status and comorbidities related to self- more support from an expert, and less fear of falling had
efficacy expectations such that those who had better cogni- stronger outcome expectations for exercise. As noted in the
tive status and fewer comorbidities had higher self-efficacy 6 month model, self-efficacy and outcome expectations re-
expectations. Self-efficacy expectations and social support lated to stage of change, and those with stronger self-efficacy
for exercise from friends related to outcome expectations and outcome expectations were more likely to be in higher
such that those with higher self-efficacy expectations and stages of change for exercise. Stage of change and exposure
more support from friends to exercise had stronger outcome to treatment were the only variables directly related to time
expectations. Outcome expectations directly related to stage spent in exercise. All other variables indirectly related to time
of change such that those with stronger outcome expecta- in exercise through self-efficacy and outcome expectations.
tions were more likely to be exercising. Self-efficacy and There was a fair fit of the model to the data (χ2 = 59.7, df = 19,
stage of change directly related to time spent in exercise, as p = 0.00, ratio 3.1, NFI = 0.76, and RMSEA of 0.10), and the
those with stronger self-efficacy and a higher stage of change model explained 21% of the variance of exercise behavior
(eg, in maintenance versus precontemplation) spent more at twelve months post hip fracture.
time exercising. While this model showed a good fit to the
data (χ2 = 22.6, df =14, p = 0.07, ratio 1.6; NFI = 0.84, and Discussion
RMSEA of 0.05), it explained only 10% of the variance of The findings from this study support prior findings and add to
exercise behavior at two months post hip fracture. the understanding of the factors that relate to exercise behavior

Clinical Interventions in Aging 2007:2(3) 419

Resnick et al

Table 2 Means (SE) for selected outcome measures by treatment group (total n = 208; treatment group n = 157; control = 51)
Variable Mean Std. Deviation
Stage of change 2 months Control 1.8 1.6
Treatment 1.9 1.6

Stage of change 6 months Control 1.2 1.7

Treatment 1.9 1.8

Stage of change 12 months Control 1.0 1.5

Treatment 2.2 1.9

Outcome expectations 2 months Control 3.9 0.56

Treatment 3.9 0.64

Outcome expectations 6 months Control 3.8 0.64

Treatment 3.9 0.59

Outcome expectations 12 months Control 3.7 0.66

Treatment 3.9 0.61

Self-efficacy expectations 2 months Control 6.5 2.3

Treatment 6.5 2.8

Self-efficacy expectations 6 months Control 5.8 3.1

Treatment 7.2 2.5

Self-efficacy expectations 12 months Control 6.3 3.2

Treatment 7.4 2.4

CESD score at 2 months (larger = depressed) Control 12.2 9.3

Treatment 9.9 9.1

CESD score at 6 months (larger = depressed) Control 11.8 9.2

Treatment 9.2 8.7

CESD score at 12 months(larger = depressed) Control 9.0 7.7

Treatment 9.2 7.9

Summary gait and balance score 2 months Control 18.5 6.9

Treatment 20.6 4.5

Summary gait and balance score 6 months Control 17.3 5.7

Treatment 17.5 6.8

Summary gait and balance score 12 months Control 20.2 5.2

Treatment 20.3 5.4

Physical health status 2 months Control 31.3 11.8

Treatment 35.7 11.7

Physical health status 6 months Control 36.9 14.6

Treatment 40.8 13.6

Physical health status 12 months Control 40.3 15.6

Treatment 43.3 14.0

Mental health status 2 months Control 40.9 13.8

Treatment 45.9 9.9

Mental health status 6 months Control 47.8 12.0

Treatment 50.3 9.8

Mental health status 12 months Control 49.7 10.1

Treatment 50.9 9.3

Yale: total exercise time 2mo, hrs/wk Control 1.7 2.3

Treatment 1.8 2.2

Yale: total exercise time 6mo, hrs/wk Control 2.6 3.2

(Continued)

420 Clinical Interventions in Aging 2007:2(3)

 Exercise Plus Program

Table 2 (Continued)
Variable Mean Std. Deviation
Treatment 2.2 2.9
Yale: total exercise time 12mo, hrs/wk Control 0.92 1.4
Treatment 3.1 3.8

Pain 2 months Control 3.7 2.2

Treatment 4.1 2.8

Pain 6 months Control 4.0 3.0

Treatment 3.6 3.0

Pain 12 months Control 3.7 2.9

Treatment 3.1 2.9

Fear 2 months Control 2.6 1.4

Treatment 2.3 1.4

Fear 6 months Control 2.4 1.4

Treatment 2.0 1.5

Fear 12 months Control 2.1 1.4

Treatment 1.9 1.4

Social support experts 2 months Control 17.5 6.4

Treatment 18.9 4.1

Social support experts 6 months Control 17.8 4.5

Treatment 26.1 8.2

Social support experts 12 months Control 17.5 6.4

Treatment 18.9 18.9

Social support friends 2 months Control 17.4 2.6

Treatment 17.6 3.6

Social support friends 6 months Control 18.2 3.8

Treatment 18.0 3.6

Social support friends 12 months Control 17.4 2.6

Treatment 17.6 3.6

Social support family 2 months Control 22.2 7.8

Treatment 22.2 6.0

Social support family 6 months Control 19.6 4.7

Treatment 20.9 6.5

Social support family 12 months Control 21.1 8.9

Treatment 19.8 5.6

in older adults, particularly those who have sustained a hip of trainer effect during any of the testing time points. Thus, the
fracture. The three predictive models tested across the 12 month benefits of encouraging exercise are not trainer-specific, and the
recovery trajectory suggest that somewhat different factors may skills to be an effective trainer may be easily learned.
influence exercise over the recovery period. At two months post Similar to prior studies with community dwelling older
hip fracture the participants were just beginning to be exposed to adults (Litt et al 2002; Resnick and Nigg 2003; Benjamin
the intervention, which may explain why treatment group status et al 2005; Stiggelbout et al 2006), self-efficacy and outcome
was not related to exercise behavior. However, 6 and 12 months expectations related to stage of change for exercise, which
post fracture the exposure to treatment did relate to time spent directly related to exercise behavior. However, with the
doing exercise; this finding speaks well to the effort of encourag- exception of the two month testing time point, self-efficacy and
ing exercise. Although there were five different trainers providing outcome expectations had no direct relationship with exercise.
treatment during the course of the study, there was no evidence Instead, they indirectly related to exercise through stage of

Clinical Interventions in Aging 2007:2(3) 421

Resnick et al

Table 3 Path coefficients for significant paths in hypothesized models

Path Tested 2 Month Model 6 Month 12 Month
Model Model
Cognitive status → Self-efficacy 0.29(0.00)
Comorbidities → Self-efficacy –0.15(0.04)
Physical health → Self-efficacy 0.20 (0.04) 0.37(0.00)
Mental health → Self-efficacy 0.27(0.01)
Social Support Experts → Self-efficacy –0.20(0.01)
Treatment group → Self-efficacy 0.20(0.01)
Fear → Self-efficacy –0.25(0.00)
Age → Outcome expectations –0.20 (0.01)
Mental Health → Outcome expectations 0.28(0.01)
Social Support Friend → Outcome expectations 0.29(0.00) 0.19(0.01)
Social Support Experts → Outcome expectations 0.15(0.04)
Fear → Outcome expectations –0.23(0.00) –0.23(0.00)
Self-efficacy → Outcome expectations 0.39(0.00) 0.39(0.00)
Self-efficacy → Stage of change 0.24(0.00) 0.26(0.00)
Outcome expectations → Stage of change 0.44(00) 0.33(0.00) 0.36(0.00)
Self-efficacy → Exercise time 0.25(0.00)
Stage of change → Exercise time 0.15(0.04) 0.20(0.01) 0.42(0.00)
Treatment group*→ Exercise time 0.20(0.01) 0.17(0.02)

*Exposure to any component of the intervention (Exercise only, Motivation only, Exercise Plus Motivation) versus routine care

change. Although there have been multiple studies supporting measure and had a high mean score of 3.9 (SD = 0.69 and
a direct relationship between self-efficacy and/or outcome range of 1 to 5) (Resnick 2006).
expectations with exercise (Booth et al 2000; Rhodes et al Age, cognitive status, and comorbidities had a limited
2001; Brassington et al 2002; Litt et al 2002; Conn et al 2003a, indirect relationship with exercise behavior. The participants
2003b; O’Connor 2004; Cress M 2005; Sharma et al 2005; were, however, all older adults and had to meet specific cog-
Taylor-Piliae and Froelicher 2005; Lee and Laffrey 2006; nitive criteria and be free of a large number of comorbidities
Wilcox et al 2006), this relationship has not been consistent to be eligible to participate in the study. Consequently, the
among older adults. In a recent study (Stiggelbout et al 2006) homogeneity of the sample may have influenced findings.
of community dwelling older adults involved in exercise Nonetheless, results suggest that in a similar population of
programs, self-efficacy influenced intention to exercise but older adults with hip fractures, age, cognition, and health
not actual exercise behavior. The lack of relationship between are not a deterrent to exercise. Perceptions of physical and
self-efficacy and actual exercise behavior was also noted in a mental health status were noted to relate to self-efficacy
sample of older adults participating in a home-based exercise and/or outcome expectations at 6 and 12 months post frac-
program viewed on television (Hopman-Rock et al 2005) ture, with those in better perceived health having stronger
and among older adults post stroke (Resnick pers comm). self-efficacy or outcome expectations, which then indirectly
Thus, the results of this study add to the evidence suggesting influenced time spent in exercise. Clinically, it is important
that interventions might best be targeted at encouraging self- to recognize that those with poorer health are likely to have
efficacy related to readiness to adopt exercise behavior, after lower self-efficacy and outcome expectations associated with
which time doing exercise will increase. exercise and may benefit from interventions to strengthen
The lack of a direct relationship between outcome expec- those beliefs. In particular it is critical that individuals with
tations and exercise behavior in older adults post hip fracture perceptions of poor health status understand and believe that
may be due to the sample studied and a ceiling effect of the it is safe for them to exercise and that there will be a benefit
measure. That is, these individuals had all volunteered to to doing so (Resnick et al 2005).
participate in an exercise intervention study and therefore The relationship between self-efficacy and outcome
were likely to have high outcome expectations related to expectations for exercise and fear of falling noted in this study
exercise. Indeed, at baseline (2 months post hip fracture), has not been reported in prior research. The study findings
the majority of participants agreed with the positive benefits suggest that the relationship between fear and exercise may
associated with exercise on the outcome expectations be mediated by self-efficacy and outcome expectations, as

422 Clinical Interventions in Aging 2007:2(3)

 Exercise Plus Program

Cognitive
Status
0.29

Self-efficacy
0.15
0.25
Charlson
0.39

Exercise
Time
Outcome
Expectations 0.15
0.43

Stage of
Change
0.29

Social Supports
Exercise
Friends

Figure 2 Two month model significant paths only.

was demonstrated at 6 and 12 months post hip fracture. It an increase or maintenance of efficacy expectations in the
is of note that the impact of fear seems most prevalent at treatment group and a decline in the control group. The lack of
12 months post fracture rather than in the more immediate a significant impact on self-efficacy and outcome expectations
post fracture period (eg, 2 months post fracture). It seems may in part be due, as indicated previously, to the sample
likely that individuals further along in the recovery trajectory included in the study (ie, volunteers in an exercise intervention
may be engaging in more activity, and thereby reconsider- study) and ceiling effects of these measures. It is also possible
ing their fear in the face of that activity. This suggests that that, post hip fracture, older adults may evaluate their self-
ongoing efforts might be made to address the fear of falling efficacy and outcome expectations based on their prior health
experienced by individuals well after their initial fracture. status, not current status post hip fracture. Consequently, as
Exposure to the intervention did not directly relate to self- noted in this study, a self-efficacy based intervention improved
efficacy and outcome expectations, as was anticipated (Resnick exercise behavior but did not influence self-efficacy or outcome
et al 2002a) although there were non-significant trends of expectations in the first year post hip fracture.

Clinical Interventions in Aging 2007:2(3) 423

Resnick et al

Social Support
Exercise Experts

Treatment
Physical

–0.20
Health 0
0.2 0.2
0
Self-efficacy

0.20
27 Expectations

0.2
Mental Stage of
0.

4
Health Change

0.33
0.
20
0.2
8

–0.20
Outcome
Age Expectations Exercise
Time
0.19

–0.23

Social Support
Exercise Friends Fear

Figure 3 Six month model significant paths only.

Ongoing research is needed to explore the measurement Practitioners should consider the use of peers to strengthen
of self-efficacy and outcome expectations post hip fracture, beliefs and thereby improve exercise behavior in older adults
and establish ways to help older adults carefully evaluate their post hip fracture as was done in a recent study testing a group
self-efficacy and outcome expectations related to exercise in based exercise program for older adults post hip fracture led
the face of an acute clinical change. This is important because by an older adult trainer (Jones et al 2006).
self-efficacy based interventions may be even more effective Social support for exercise from experts (anyone per-
when the participant realistically appreciates his or her true ceived by the participant to be an expert) was negatively
efficacy expectations. related to self-efficacy expectations at 6 months post hip
Social support for exercise from friends related to self- fracture. Although it was anticipated that social support on
efficacy for exercise at 2 and 6 months post hip fracture. This the part of the experts would increase time spent in exercise
finding has been inconsistent in prior research with social there are several possible explanations for the negative rela-
support for exercise from friends relating to exercise behavior tionship. Social support for exercise did increase from two
among some samples of community dwelling older adults months to six months post fracture in the treatment group
(Booth et al 2000; Resnick et al 2002b), but not others (Eyler (18.9 to 26.1) while staying essentially unchanged in the
et al 1999; Brassington et al 2002). It is possible that inter- control group (17.5 to 17.8). It is possible that this increase
actions with peers, possibly peers who themselves exercise in social support from the experts was not related to exercise
(and may have experienced a hip fracture), has a positive behavior, with other factors taking on a greater precedence.
influence on self-efficacy related to exercise post hip fracture. It is also possible, since the intervention did not control

424 Clinical Interventions in Aging 2007:2(3)

 Exercise Plus Program

Physical Health
Status 0. Intervention
37
Group
Self-efficacy
.25
–0
0.17
Fear
–0 0.39
.23
Exercise

0.2
Time

6
Outcome
Expectations

2
0.4
0.3
6

Stage of
Change
0.15

Social Supports
Exercise
Experts

Figure 4 Twelve month model significant paths only.

the interactions between the participants and any of their significant hypothesized predictors further support the challenges
health care providers (primary care physicians, nurses, nurse associated with increasing exercise activity among older adults,
practitioners, or surgeons), that some negative interchanges particularly those who have sustained a hip fracture. Specifically,
related to exercise may have occurred. We had experiences, pain, depressive symptoms, and gait and balance consistently
for example, in which some participants were told not to had no direct or indirect influence on exercise behavior. Using
exercise by their orthopedist or primary medical doctor the social ecological model of behavior, possible factors that
(Resnick 2005). Future research should seek to understand might influence exercise behavior among older adults but were
the ongoing exchanges between patients and providers for not considered in this study could be added to future work,
this reason. including environmental and policy considerations (eg, providing
While the revised models with significant paths had a fair financial incentives for participation in exercise or establishing
to good fit with the data, they only explained a small amount of safe walking paths within communities) (Booth et al 2000),
the variance in exercise behavior (8% to 20%). The many non- whether or not the individual had to stop exercise for a period

Clinical Interventions in Aging 2007:2(3) 425

Resnick et al

due to an acute event (Stiggelbout et al 2006), or consideration Buie V, Orwig B, Resnick B, et al. 2001. Frail Elderly women post-
hip fracture: Recruitment and retention into a 12-month exercise
of life events such as the death of a loved one, pet, or change in intervention study. The Gerontologist, 41:41–57.
location (Wilcox and King 2004). Caracciolo B, Giaquinto S. 2002. Criterion validity of the center for epidemi-
ological studies depression (CES-D) scale in a sample of rehabilitation
inpatients. Journal of Rehabilitation Medicine, 34:221–5.
Limitations Centers for Disease Control and Prevention, 2006. behavioral risk factor
This study was limited in that the sample size was small and surveillance system 2006 [online]. Accessed on December 12, 2006.
URL:https://www.cdc.gov/brfss.
homogenous which influenced model fit results and the likely Centers for Disease Control and Prevention, Merck Institute of Aging and
replicability of the findings. However, despite these limitations, Health. 2004. The State of Aging and Health in America, 2004: 21–6
the findings provide some guidance for future work in the area of Conn VS, Burks KJ, Pomeroy SL, et al. 2003a. Are there different predictors
of distinct exercise components? Rehabilitation Nursing, 28:87–91.
developing interventions to increase exercise post hip fracture, as Conn VS, Tripp-Reimer T, Maas ML. 2003b. Older women and exercise:
well as measurement challenges for social cognitive constructs theory of planned behavior beliefs. Public Health Nursing, 20:153–63.
Cress MB, Prohaska T, Rimmer J, et al. 2005. Best practices for physical
post hip fracture (ie, accurate assessment of self-efficacy and activity programs and behavior counseling in older adult populations.
outcome expectations). In addition to helping older adults post Journal of Aging and Physical Activity, 13:61–74.
hip fracture realistically readjust their self-efficacy and outcome Crews D. 2005. Artificial environments and an aging population: De-
signing for age related functional losses. Journal of Physiological
expectations related to exercise, the research team recommends Anthropology, 24:103–9.
that health care providers and friends/peers reinforce the positive Cumming RG, Salkeld G, Thomas M, et al. 2000. Prospective study of the
impact of fear of falling on activities of daily living, SF-36 Scores, and
benefits of exercise post hip fracture, and continue to address fear nursing home admission. Journal of Gerontology, 55:299–305.
of falling throughout the entire hip fracture recovery trajectory, as Delbacre K, Crombez G, Vanderstraeten G, et al. 2004. Fear related avoid-
well as explore additional factors that may influence time spent ance of activities, falls and physical frailty: A prospective community
based cohort study. Age and Ageing, 33:368–73.
in exercise post hip fracture. Dipietro L, Caspersen CJ, Ostfeld AM, et al. 1993. A survey for assessing
physical activity among older adults. Medicine Science Sports and
Acknowledgments Exercise, 25:628–42.
Eyler AA, Brownson RC, Donatelle RJ, et al. 1999. Physical activity social
Support for this project was provided by National Institute support and middle- and older-aged minority women: results from a
on Aging (NIA) grants R37 AG09901, R01-AG18668, R01 US survey. Social Science Medicine, 49:781–9.
Fletcher P, Hirdes J. 2004. Restriction in activity associated with fear of
AG17082, and the Claude D. Pepper Older Americans Inde- falling among community-based seniors using home care services. Age
pendence Center P60-AG12583. Authors also would like to Ageing, 33:273–9.
Folstein MF, Folstein SE, Mchugh PR. 1975. “Mini-Mental state” – A practi-
thank Thera-Band Academy for their generous contribution
cal method for grading the cognitive state of patients for the clinician.
of Thera-Band® resistive bands used by study participants; Journal of Psychiatric Research, 12:189–98.
hospitals and personnel participating in the Baltimore Hip Habris R, O’Hara P, Harper D. 1995. Functional status of geriatric reha-
bilitation patients: a one year follow-up study. Journal of the American
Studies; and research staff who worked with study patients Geriatric Society, 36:890–6.
and their families. Authors also would like to thank hip Henderson SA, Finlay OE, Murphy, et al. 1992. Benefits of an exercise
class for elderly women following hip surgery. Ulster Medical Journa,
fracture patients and their families for volunteering their
61:144–50.
time and information for this work. Herr K, Mobily P. 1993. Comparison of selected pain assessment tools for
use with the elderly. Applied Nursing Research, 6:39–46.
References Herr K, Spratt K, Mobily P. et al. 2004. Pain intensity assessment in older
Arbuckle J. 1997. Amos users’ guide version 3.6, Chicago, Small Waters adults: use of experimental pain to compare psychometric properties
Corporation. and usability of selected pain scales with younger adults. Clinical
Bandura A. 1997. Self-efficacy: The Exercise of Control., New York, New Journal of Pain, 20:207–19.
York, W.H. Freeman and Company. Herr KA, Mobily PR. 1991. Complexities of pain assessment in the elderly.
Benjamin K, Edwards NC, Bharti VK. 2005. Attitudinal, perceptual, and norma- Clinical considerations. Journal of Gerontological Nursing, 17:12–9.
tive beliefs influencing the exercise decisions of community-dwelling physi- Hopman-Rock M, Borghouts JA, Leurs MT. 2005. Determinants of
cally frail seniors. Journal of Aging and Physical Activity, 13:276–93. participation in a health education and exercise program on television.
Bohannon RW, Maljanian R, Goethe J. 2003. Screening for depression Preventive Medicine, 41:232–9.
in clinical practice: reliability and validity of a five-item subset of Humpel N, Owen N, Leslie E. 2002. Environmental factors associated with
the CES-Depression. Perceptions and Motor Skills, 97:855–63. adults’ participation in physical activity: A review. American Journal
Booth ML, Owen N, Bauman A, et al. 2000. Social-cognitive and perceived of Preventive Medicine, 22:188–99.
environment influences associated with physical activity in older Iwarsson S. 2005. A long-term perspective on person-environment fit and ADL
Australians. Preventive Medicine, 31:15–22. dependence among older Swedish adults. The Gerontologist, 45:327–36.
Brassington GS, Atienza A, Perczek et al. 2002. Intervention-related Jones G, Jakobi J, Taylor A, et al. 2006. Community exercise program for
cognitive versus social mediators of exercise adherence in the elderly. older adults recovering from hip fracture: A pilot study. Journal of
American Journal of Preventive Medicine, 23:80–6. Aging and Physical Activity, 14:439–55.
Bruce DG, Devine A, Prince RL. 2002. Recreational physical activity levels Jorstad E, Hauer K, Becker C, et al. 2005. Measuring the psychological
in healthy older women: the importance of fear of falling. Journal of outcomes of falling: a systematic review. Journal of the American
the American Geriatrics Society, 50:84–9. Geriatrics Society, 53:501–10.

426 Clinical Interventions in Aging 2007:2(3)

 Exercise Plus Program

Kolbe-Alexander TL, Lambert EV, Harkins JB, et al. 2006. Comparison Sharma M, Sargent L, Stacy R. 2005. Predictors of leisure-time physical
of two methods of measuring physical activity in South African older activity among African American women. American Journal of Health
adults. Journal of Aging and Physical Activity, 14:98–114. Behavior, 29:352–9.
Lee Y, Laffrey S. 2006. Predictors of physical activity in older adults with Stewart A, King A, Haskell W. 1993. Endurance exercise and health-related
borderline hypertension. Nursing Research, 55:110–20. quality of life in 50–65 year-old adults. The Gerontologist, 33:782–9.
Li F, Fisher J, Harmer P, et al. 2003. Fear of falling in elderly persons: Stewart AL, Hays RD, Ware JE. 1988. The MOS short-form general health
association with falls, functional ability, and quality of life. Journal of survey: Reliability and validity in a patient population. Medical Care,
Gerontology, 58:283–90. 26:724–735.
Litt M, Kleppinger A, Judge J. 2002. Initiation and maintenance of exercise Stiggelbout M, Hopman-Rock M, Crone M, et al. 2006. Predicting older
behavior in older women: Predictors from the social learning model. adults’ maintenance in exercise participation using an integrated social
Journal of Behavioral Medicine, 25:83–97. psychological model. Nursing Research, 21:1–14.
Loehlin J. 1998. Latent variable models. New Jersey, Lawrence Erlbaum Takano T, Nakamura K, Watanabe. 2002. Urban residential environments
Associates. and senior citizens’ longevity in megacity areas: the importance of
Mangione KK, Craik RL, Tomlinson SS, et al. 2005. Can elderly patients walkable green spaces. Journal of Epidemiology and Community
who have had a hip fracture perform moderate- to high-intensity exercise Health, 56:913–8.
at home? Physical Therapy, 85:727–39. Taylor-Piliae R, Froelicher E. 2005. Measurement properties of Tai Chi
Martin FC, Hart D, Spector T, et al. 2005. Fear of falling limiting activity in exercise self-efficacy among ethnic Chinese with coronary heart
youngold women is associated with reduced functional mobility rather disease risk factors: a pilot study. European Journal Cardiovascular
than psychological factors. Age and Ageing, 34:281–7. Nursing, 3:287–94.
Mihalko S, Wickley K. 2003. Active living for assisted living: Promoting Taylor L, Harris J, Epps C, et al. 2005. Psychometric evaluation of
partnerships within a systems framework. American Journal of Preven- selected pain intensity scales for use with cognitively impaired
tive Medicine, 25:193–203. and cognitively intact older adults. Rehabilitation Nursing,
National Blueprint for Increasing Physical Activity. 2002. National 30:55–61.
Blueprint for increasing physical activity among adults 50 and older, Thompson P, Buchner D, Pina IL, et al. 2003. American Heart Association
Creating a strategic framework and enhancing organizational capacity Council on Clinical Cardiology Subcommittee on Exercise,
for change. Journal of Aging and Physical Activity, 9:S5–28. Rehabilitation, and Prevention; American Heart Association Council
O’Connor BP, Rousseau FL, Maki SA. 2004. Physical exercise and experienced on Nutrition, Physical Activity, and Metabolism Subcommittee on
bodily changes: the emergence of benefits and limits on benefits. Interna- Physical Activity. Exercise and physical activity in the prevention
tional Journal of Aging and Human Development, 59:177–203. and treatment of atherosclerotic cardiovascular disease: a statement
Pescatello L, Dipietro L, Fargo A, et al. 1994. The impact of physical activity from the Council on Clinical Cardiology (Subcommittee on Exercise,
and physical fitness on health indicators among older adults. Journal Rehabilitation, and Prevention) and the Council on Nutrition, Physical
of Aging and Physical Activity, 2:2–13. Activity, and Metabolism (Subcommittee on Physical Activity).
Radloff LS. 1977. The CES-D Scale: A self-report depression scale for research Circulation, 107:3109–16.
in the general population. Applied Psychological Measures, 1:385–401. Tinetti ME. 1986. Performance oriented assessment of mobility prob-
Resnick B. 1998. Functional performance of older adults in a long term care lems in elderly patients. Journal of the American Geriatrics Society,
setting. Clinical Nursing Research, 7:230–46. 34:199–206.
Resnick B. 2006. The Baltimore Hip Studies: Hip 4 and Hip 5. University Tinetti ME, Baker DI, Gottschalk M, et al. 1999. Home-based multicomponent
of Maryland Unpublished Data Analysis. rehabilitation program for older persons after hip fracture: a randomized
Resnick B, Jenkins LS. 2000. Testing the reliability and validity of the trial. Archives of Physical Medicine and Rehabilitation, 80:916–22.
Self-Efficacy for Exercise scale. Nursing Research, 49:154–9. Tsauo J, Leu W, Chen Y, et al. 2005. Effects on function and quality of life of
Resnick B, Magaziner J, Orwig D, et al. 2002a. Evaluating the components postoperative home-based physical therapy for patients with hip fracture
of the Exercise Plus Program: Rationale, theory and implementation. Archives of Physical Medicine and Rehabilitation, 86:1953–7.
Health Education Research, 17:648–59. Turk D, Okifuji A. 1994. Detecting depression in chronic pain patients:
Resnick B, Michael K, Shaughnessy, M, et al 2007. Testing a Comprehensive adequacy of self-reports. Behavioral Research and Therapeutics,
Conceptualization of Physical Activity in Stroke Patients. Personal Com- 32:9–16.
munication (Manuscript submitted). Walters S, Munro J. 2004. Using the SF-36 with older adults: a cross-
Resnick B, Nigg C. 2003. Testing a theoretical model of exercise behavior sectional community-based survey. Age and Ageing, 30:337–43.
for older adults. Nursing Research, 52:80–8. Ware L, Epps C, Herr K , et al. 2006. Evaluation of the revised faces pain
Resnick B, Orwig D, Magaziner J, et al. 2002b. The effect of social support on scale, verbal descriptor scale, numeric rating scale, and Iowa pain
exercise behavior in older adults. Clinical Nursing Research, 11:52–70. thermometer in older minority adults. Pain Management in Nursing,
Resnick B, Orwig D, Wehren L, et al. 2005. The Exercise Plus Program 7:117–25.
for Older Women Post Hip Fracture: Participant Perspectives. The Wilcox S, Dowda M, Griffin SF, et al. 2006. Results of the first year of active
Gerontologist, 45:539–44. for life: translation of 2 evidence-based physical activity programs
Resnick B, Orwig D, Yu-Yahiro J, et al. 2007. Testing the Effectiveness of for older adults into community settings. American Journal of Public
the Exercise Plus Program in Older Women Post Hip Fracture. Annals of Health, 96:1201–9.
Behavioral Medicine, 34:67–76. Wilcox S, King A. 2004. The effects of life events and interpersonal loss
Resnick B, Orwig D, Zimmerman S, et al. 2005. The Exercise Plus Program on exercise adherence in older adults. Journal of Aging and Physical
for Older Women Post Hip Fracture: Participant Perspectives. The Activity, 12:117–30.
Gerontologist, 45:539–44. Yu-Yahiro J, Gardner A, Resnick B, et al. 2001. Challenges of designing
Rhodes RE, Martin AD, Taunton JE. 2001. Temporal relationships of self- an exercise program for use with frail elderly hip fracture patients. The
efficacy and social support as predictors of adherence in a 6-month Gerontologist, 41:57.
strength-training program for older women. Perception and Motor
Skills, 93:693–703.
Sallis J, Grossman R, Pinski R, et al. 1986. The development of scales to
measure social support for diet and exercise behaviors. Preventive
Medicine, 16:825–36.

Clinical Interventions in Aging 2007:2(3) 427

Reproduced with permission of copyright owner. Further reproduction
prohibited without permission.