Nash N.

 Boutros

Standard EEG:
A Research
Roadmap for
Neuropsychiatry
Standard EEG: A Research Roadmap
for Neuropsychiatry
Nash N. Boutros

Standard EEG: A Research

Roadmap for
Neuropsychiatry

123
Nash N. Boutros
Department of Psychiatry
University of Missouri-Kansas City School
of Medicine (UMKC)
Kansas, MO
USA

and

The Center for Behavioral Medicine

University of Missouri-Kansas City School
of Medicine (UMKC)
Kansas, MO
USA

ISBN 978-3-319-04443-9 ISBN 978-3-319-04444-6 (eBook)

DOI 10.1007/978-3-319-04444-6
Springer Cham Heidelberg New York Dordrecht London

Library of Congress Control Number: 2013957811

 Springer International Publishing Switzerland 2013

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or
information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed. Exempted from this legal reservation are brief
excerpts in connection with reviews or scholarly analysis or material supplied specifically for the
purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the
work. Duplication of this publication or parts thereof is permitted only under the provisions of
the Copyright Law of the Publisher’s location, in its current version, and permission for use must
always be obtained from Springer. Permissions for use may be obtained through RightsLink at the
Copyright Clearance Center. Violations are liable to prosecution under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt
from the relevant protective laws and regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of
publication, neither the authors nor the editors nor the publisher can accept any legal responsibility for
any errors or omissions that may be made. The publisher makes no warranty, express or implied, with
respect to the material contained herein.

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

Foreword

Since the time of Berger, psychiatrists have searched for the clinical significance
of the electrical activity of the brain. Thus far they have learned little from the
EEG to relate to diagnosis, treatment, or course of illness. There is even the
likelihood that the standard EEG as a clinical and research tool in psychiatry might
be abandoned in favor of more sensitive electrophysiological and other brain
imaging techniques. However, this trend could well be premature as evidenced in
the present volume.
There is a general agreement that EEG findings do not conform to present-day
(DSM-based) diagnostic categories. Almost all the conditions and situations
described in this book can display a variety of EEG aberrations, including
increased fast activity, sharp waves and spikes, slowing of frequencies, and/or one
or more of the controversial waveforms. Or the EEG may be entirely within
normal limits although the latter category requires lengthy recordings in waking
and sleep to assure that deviations are not missed. There is also the problem of
referral bias wherein practitioners limit EEG studies to those patients suspected of
brain disorders. Despite these concerns, there are sound data showing significant
associations with symptomatology, family history, and other attributes which are
described in detail herein.
The issue of whether different diagnostic groups that display similar EEG
characteristics may have other features in common is largely unknown. This sit-
uation resembles a recent genetics study in which certain risk loci were identified
with shared effects on five major psychiatric disorders (Serretti and Fabbri 2013).
This was particularly remarkable since the syndromes were so different clini-
cally—schizophrenia, autistic disorder, bipolar disorder, major depressive disor-
der, and attention deficit hyperactivity disorder! Could it be that the EEG also is
indicative of some type of overlap?
Answers to these and other important research issues should still be sought from
standard EEG recordings with attention to the painstaking documentation of
current knowledge and questions for further investigation that Boutros has pro-
vided. Clearly, it is not yet the time to abandon standard EEG recordings in
psychiatry!
Joyce G. Small

v
vi Foreword

Reference

Serretti A, Fabbri C (2013) Shared genetics among major psychiatric disorders. Lancet
381(9875):1339–1341. doi:10.1016/S0140-6736(13)60223-8
Preface

The voluminous EEG literature relevant to psychiatry extends back to the mid-
1930s and is spread throughout a large number of journals of different specialties
as well as in textbooks and atlases. The psychiatrist interested in exploring this
literature faces a tedious task. This volume is designed to serve as a reference
source containing both historical and recent references with a special focus on the
existing gaps of knowledge regarding EEG deviations in psychiatric populations.
This book is not meant to be an exhaustive compendium of this vast literature, but
a guide to interested clinical researchers into the many unanswered questions
regarding standard EEG deviations in clinical psychiatry. The interested researcher
will find this book a good starting point with the most influential literature sum-
marized and the issues and questions highlighted. The researcher will then need to
further explore the literature particularly the areas not covered in this book.
It will become obvious to the reader that much of the literature reviewed in this
book is rather old. Many of these old works remain the most current work on the
particular topic. This is a testimony to the severe neglect this area of research has
experienced in the last few decades as the field of the clinical EEG in psychiatry
became an orphan field with minimal interest from both the fields of neurology and
psychiatry.
Structure and Organization of the Book: every chapter begins by outlining the
clinical issues then reviews available literature and concludes by highlighting; (a)
currently supportable findings, and (b) open research questions. In some chapters
the suggestions regarding the research design that will most likely lead to gen-
erating data that can move the field toward resolving unresolved issues are offered.
Some references are bolded. This signifies particular significance for the paper or
the textbook.
Part I of the book handles a number of general topics of relevance to the entire
field of psychiatry. This part begins with a historical account of why psychiatry
and the standard EEG are currently so far removed from each other. The chapter is
focused on the reasons for this current situation and then discusses some of the
issues that give the interpretation of the standard EEG in psychiatric settings a
special status, and the skills necessary for the adequate and skilled performance of
this task. The history chapter at the beginning of the Boutros et al. (2011) goes into
more details of the history of EEG in general. This part continues by tackling the
issue of the boundaries of the normal EEG and highlights the current lack of well-

vii
viii Preface

defined borders between patients with and without psychiatric problems. The two
kinds of abnormalities encountered in the standard EEG (the term sEEG is used
throughout the book to denote the visually inspected interpretation of the EEG
which is the sole focus of this volume) are discussed in separate chapters. Slowing
of the EEG rhythms or the appearance of abnormal slow rhythms occupies one
chapter while epileptiform activity occupies another. In the epileptiform chapter,
we also provide some data on the value of an animal model of the isolated
epileptiform discharge (IED). This part also covers areas related to the effects of
psychotropic medications on the sEEG within the context of differentiating gen-
eralized slowing of the EEG background or the superimposition of diffusely dis-
tributed slower rhythms secondary to the toxic effects of psychotropic medications
and diffuse slowing due to other general medical conditions. The book has not
attempted to cover the effects of psychotropics on the EEGs that do not render
them abnormal. This is the province of the pharmaco-EEG discipline.
Part II deals with various adult psychiatric conditions with Part III covering a
number of childhood and adolescent psychiatric conditions where increased
prevalence of EEG abnormalities have been documented. Part IV deals with the
difficult issue of the controversial waveforms. The five chapters included in this
part were the most difficult to write and I am deeply indebted to Prof. Frederick
Struve for his contribution to the Small Sharp Spike and B-Mitten chapters which
he developed while working with me at Yale University.
This book drew on a large number of outstanding sources most importantly the
‘‘Electroencephalography: Basic Principles, Clinical Applications and Related
Fields’’ edited by Ernest Niedermeyer, Fernando Lopes da Silva particularly the
latest two editions in 1987 and 2005. The ‘‘EEG and Evoked Potentials in Psy-
chiatry and Behavioral Neurology’’ by Hughes and Wilson (1983) remains an
important source of this literature. The Gibbs and Gibbs atlases are also important
and essential sources.
Throughout the time I was working on this volume, I was repeatedly advised to
include sections on the quantified EEG (QEEG). I elected to keep the volume
focused on the standard EEG. It is fully predicted that the major expansion in
psychiatric electrophysiology will come from the quantification of the signal
whether the signal was collected from evoked potential EEG or magnetoenceph-
alography (MEG) procedures. The inclusion of the already massive QEEG data in
psychiatric conditions would have completely drowned the important points being
made in this volume and would have resulted in increasing the cost of production
which also was felt to defeat the purpose. Finally, a number of excellent texts
devoted to QEEG have been already appeared. I would like to specially mention
the ‘‘Handbook of Quantitative electroencephalography and EEG Biofeedback’’
by Thatcher (2012). This volume is being periodically updated in print and online.
Finally, the volume also avoids the delving into neurological conditions like
dementia and delirium. The reason for that is again to keep the book focused on
main stream psychiatric disorders and the fact that these topics are extensively
covered in many EEG textbooks. An apparent obvious omission would be the issue
of epilepsy and psychiatric symptoms. Again, this topic has been the subject of
Preface ix

extensive investigations and a number of excellent volumes dedicated to this topic

are available. Relative to the many neurological disorders with psychiatric man-
ifestations is the skillful use of the standard EEG in conjunction with neuropsy-
chological testing. While some early work indicated correlations between the
measures, this literature remains limited and work on this interrelationship is now
all but abandoned in favor of the more advanced computer-based EEG analysis.
Janati (2010) argues against the premature discounting of the standard EEG in the
clinical practice of psychiatry.
All in all, my hope is that psychiatric electrophysiologists will find many ideas
worth pursuing in this book. In fact if the publication of this book results in a
single research project, I would be satisfied that I have succeeded in my mission.

References

Boutros NN, Galderisi S, Pogarell O et al (2011) Handbook of standard EEG in clinical

psychiatry. Blackwell, London
Hughes JR, Wilson WP (1983) EEG and evoked potentials in psychiatry and behavioral
neurology. Butterworths, Boston
Janati A (2010) Why the EEG should not be discounted. J Clin Neurophysiol 27(6):484 (letter)
Niedermeyer E, Lopes da Silva F (eds) (1987) Electroencephalography: Basic principles, clinical
applications and related fields. Urban & Schwarzenberg, Baltimore-Munich
Niedermeyer E, Lopes da Silva F (eds) (2005) Electroencephalography: Basic principles, clinical
applications and related fields. Urban & Schwarzenberg, Baltimore-Munich
Thatcher RW (2012) Handbook of quantitative electroencephalography and EEG biofeedback.
Anipublishing Company, St. Petersburg
Contents

Part I General Issues

1 Philosophical Differences in Standard EEG Interpretation

Between Neurology and Psychiatry: A Historical Perspective . ... 3
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... 3
The Renaissance of Electroencephalography in Psychiatry . . . . . ... 5
The American Psychiatric Electrophysiology Association:
History and Mission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
The Birth of the APEA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
AMEEGA-APEA Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . 8
APEA-AMEEGA Rapprochement . . . . . . . . . . . . . . . . . . . . . . . 8
The Birth of the EEG and Clinical Neuroscience
Society (ECNS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... 8
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... 14

2 What Constitutes a Normal EEG . . . . . . . . . . . . . . . . . . . . . . . . 15

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Discussion and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Supported Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Open Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

3 Special Electrodes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
T1 and T2 Electrodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Sphenoidal Electrodes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Supported Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Open Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

4 Effects of Psychotropic Drugs on the EEG. . . . . . . . . . . . . . . . . . 27

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
First Generation Psychotropic Agents . . . . . . . . . . . . . . . . . . . . . . . 28
Atypical Antipsychotic Medications . . . . . . . . . . . . . . . . . . . . . . . . 29

xi
xii Contents

Drug-Induced Paroxysmal EEG Activity. . . . . . . . . . . . . . . . . . . . . 30

Lithium. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Supported Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Open Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

5 The Special Case of Clozapine . . . . . . . . . . . . . . . . . . . . . . . . . . 35

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Early Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
More Recent Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Supported Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Open Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

6 Slowing of the EEG in Psychiatric Patients . . . . . . . . . . . . . . . . . 45

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Supported Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Open Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

7 Isolated Epileptiform Discharges in Nonepileptic

Psychiatric Patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
IEDs and the Tip of the Iceberg Concept . . . . . . . . . . . . . . . . . . 53
Are There Demonstrable Effects of IEDs? . . . . . . . . . . . . . . . . . 54
Etiology of IEDs in Nonepileptic Individuals . . . . . . . . . . . . . . . 55
Necessity for an Animal Model for IEDs . . . . . . . . . . . . . . . . . . 56
Behavioral and Cognitive Testing Procedures in Rodents . . . . . . . 57
Brief Description of Behavioral Observation Procedure . . . . . . . . 58
ANOVAs of Six Types of Behavior for Total Averages of
Each Subject (Vehicle and Treatment: N = 12 and N = 12) . . . . . 58
Effect of Environmental Stimuli on IEDs . . . . . . . . . . . . . . . . . . 60
Supported Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
Open Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

Part II Adult Psychiatric Disorders

8 Panic Attacks and Other Dissociative Disorders. . ............ 67

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . ............ 67
Panic as an Indicator of a Neurological Disorder . ............ 67
Does Laterality Play a Role in the Generation of
Fear or Panic Symptoms in Epileptic Patients? . . ............ 70
Contents xiii

EEG in Nonepileptic and Neurologically Intact Panic

Disorder Patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Heterogeneity of PD. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Supported Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Open Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Other Dissociative Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Open Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

9 Violence Aggression and Impulse Dyscontrol . . . . . . . . . . . . . . . . 83

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Episodic Aggression and Impulse Dyscontrol . . . . . . . . . . . . . . . . . 83
EEG and Episodic Dyscontrol . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Specific EEG Abnormalities . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Treatment Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Supported Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Open Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

10 Borderline Personality Disorder . . . . . . . . . . . . . . . . . . . . . . . . . 95

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Borderline Personality Disorder Links to Epilepsy . . . . . . . . . . . . 96
Standard EEG Abnormalities Reported in BPD Patients . . . . . . . . 96
Electrophysiological Profiles . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Quantitative EEG. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Supported Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Open Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

11 Psychotic and Affective Disorders . . . . . . . . . . . . . . . . . . . . . . . . 105

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
SEEG in Psychotic Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
SEEG in Schizophrenia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
SEEG Findings and Outcome . . . . . . . . . . . . . . . . . . . . . . . . . . 107
First-Episode Schizophrenia . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Affective Psychoses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Quantified EEG . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Supported Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Open Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110

12 EEG Role in Psychiatric Emergencies . . . . . . . . . . . . . . . . . . . . . 113

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Catatonia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
xiv Contents

Case Reports of EEG in Catatonia . . . . . . . . . . . . . . ......... 114

EEG Findings as State Indicators. . . . . . . . . . . . . . . ......... 118
The Patient Presenting with a Difficult to Assess Mental
Status (DAMS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
Ambulatory Nonconvulsive Status Epilepticus . . . . . . . . . . . . . . . . . 119
Supported Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Open Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122

Part III Childhood Psychiatric Conditions

13 Attention Deficit Disorder and Learning Disabilities . . . . . . . . . . 127

Introduction. . . . . . . . . . . . . . . . . . . . . . ......... . . . . . . . . . . 127
Standard-EEG in AD/HD Studies. . . . . . . ......... . . . . . . . . . . 128
Learning Disabilities . . . . . . . . . . . . . . . ......... . . . . . . . . . . 130
Supported Observations . . . . . . . . . . . ......... . . . . . . . . . . 132
Open Research Questions . . . . . . . . . . ......... . . . . . . . . . . 132
References . . . . . . . . . . . . . . . . . . . . . . ......... . . . . . . . . . . 133

14 Autistic Spectrum Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Evidence of Neurological Contributions to the Syndrome . . . . . . . 135
EEG Abnormalities in ASD . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
Electrical Status Epilepticus in Slow-Wave Sleep . . . . . . . . . . . . 141
Response to Anticonvulsant Treatment . . . . . . . . . . . . . . . . . . . . 141
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146

Part IV Controversial EEG Waveforms

15 Introduction to Controversial Sharp Waves or Spike Patterns . . . 151

Unusual but Little Studied in Psychiatric Populations Patterns. . . . . . 152
What does the Co-occurrence of These Waveforms Suggest? . . . . . . 155
A Hypothesis to be Tested . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157

16 Psychiatric Correlates of the B-Mitten EEG Pattern . . . . . . . . . . 159

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
The B-Mitten EEG Wave Form . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
Incidence and Age Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
Brain Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
Clinical Symptomatology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
Early Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
Affective Dysregulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Extrapyramidal Side Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
Contents xv

Tardive Dyskinesia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166

Supported Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
Open Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

17 Small Sharp Spikes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171

Wave Form Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
Activation Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
Age Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
Incidence Among ‘‘Normal’’ Control Subjects. . . . . . . . . . . . . . . 173
Brain Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
Clinical Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
Small Sharp Spikes in Unselected Psychiatric Patients . . . . . . . . . 176
The Affective Disturbance Connection . . . . . . . . . . . . . . . . . . . . 176
Autonomic and Neurovegetative Symptoms . . . . . . . . . . . . . . . . 180
Seizure Correlates Versus Nonseizure Correlates . . . . . . . . . . . . . 182
Supported Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Open Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184

18 Six/Second Spike and Wave Complexes, the Rhythmic

Mid-Temporal Discharges and the Wicket Spikes . . . . . . . . . . . . 187
The 6/s Spike and Wave Complexes . . . . . . . . . . . . . . . . . . . . . . . 187
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
Waveform Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
Incidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Clinical Correlates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
Supported Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
Open Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
The Rhythmic Mid-Temporal Discharges . . . . . . . . . . . . . . . . . . . . 191
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
Description of Pattern. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
Incidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
Clinical Correlates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
Supported Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
Open Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
Wicket Spikes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
Description of Pattern. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
Incidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
Clinical Correlates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
Supported Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
Open Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
xvi Contents

19 The 6–7 and 14 Positive Spikes . . . . . . . . . . . . . . . . . . . . . . . . . . 201

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
Description of Wave Form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
Incidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
Incidence in Normal Children . . . . . . . . . . . . . . . . . . . . . . . . . . 203
Incidence in Adults . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
Incidence in Pathological Conditions . . . . . . . . . . . . . . . . . . . . . 208
Psychiatric Populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
Incidence in Adult Criminals. . . . . . . . . . . . . . . . . . . . . . . . . . . 209
Incidence in Learning Disabilities . . . . . . . . . . . . . . . . . . . . . . . 210
Head Injury. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
Behavioral Correlates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
Additional Evidence of the Pathological Nature of the PS . . . . . . 217
Are PSs Simply Sleep Spindles Variants? . . . . . . . . . . . . . . . . . . 217
Treatment Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
Supported Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Open Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220

20 Some Final Thoughts for Clinical Researchers. . . . . . . . . . . . . . . 225

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
 Part I
General Issues
Chapter 1
Philosophical Differences in Standard
EEG Interpretation Between Neurology
and Psychiatry: A Historical Perspective

Introduction

An important historical fact is that the very beginning of the field of human
electroencephalography (EEG) emerged directly from the field of psychiatry. Hans
Berger, was a biologically oriented psychiatrist with strong interests in the rela-
tionships between mind and body (Gibbs and Gibbs 1950). In 1929, he launched
nearly a decade of landmark publications that essentially laid down the very
foundation of this new field. Early on most major academic departments of psy-
chiatry had clinical EEG laboratories. It is thus of great interest to try to under-
stand why today, there are hardly any departments of psychiatry housing clinical
EEG laboratories and why training in clinical EEG is not a part of the training
psychiatrists receive.
In less than a decade after Berger’s initial publication, the potential use of EEG
was being very actively explored in psychiatry and neurology. The strongest
correlations between EEG findings and clinical disease involved epilepsy, struc-
tural lesions, and encephalopathies. At the same time, several minor EEG
abnormalities were being found in much higher incidences in samples of psy-
chiatric patients as compared with normal control subjects, most of these EEG
findings lacked clear psychiatric diagnostic specificity or clear prediction of
responsiveness to treatment (i.e., the treatments available at the time these early
studies were conducted). These facts operated to reduce interest in EEG among
many psychiatrists. These two historical developments (important and influential
clinical findings pertinent to neurology and less clinically useful findings in psy-
chiatric populations) essentially moved electroencephalography closer to the dis-
cipline of neurology and away from its roots in psychiatry.
The impetus for using EEG in the study of seizure disorders began when
Frederic Gibbs became aware of an animal study by Fischer (1933) that showed
that high voltage discharges in the brain were produced when the animals were
thrown into seizures by administering convulsive drugs. Gibbs and his team soon
described the diffuse 3/s spike and wave discharges. The 3/s spike and wave
discharges proved to be an EEG signature for petit mal epilepsy and thus was

N. N. Boutros, Standard EEG: A Research Roadmap for Neuropsychiatry, 3

DOI: 10.1007/978-3-319-04444-6_1,  Springer International Publishing Switzerland 2013
4 1 Philosophical Differences in Standard EEG Interpretation

extremely useful and important (Gibbs et al. 1935). Many other discoveries related
to epilepsy as soon followed.
Of more specific relevance to psychiatry was the discovery a little over a decade
later of the focal anterior temporal EEG spike discharge which became recognized
as a diagnostically useful inter-ictal (i.e., occurring between seizures in epileptic
individuals) EEG finding in complex partial seizures (previously ‘‘psychomotor
epilepsy’’) (Gibbs et al. 1947). In this book we differentiate between inter-ictal
discharges (IIDs) and isolated epileptiform discharges (IEDs) (see Chap. 7 for
more details). This EEG finding (whether IIDs in known epileptic patients or IEDs
in non-epileptic individuals continues to be under-investigated in order to clearly
identify its differential diagnostic implications in the practice of psychiatry
(Boutros 2010). The description of the anterior temporal spike discharges by
Frederick and Erna Gibbs and their collaborators was followed by the description,
by the same group, of a number of EEG patterns that tend to be more common in
psychiatric patients and have come to be collectively known as the ‘‘controversial
EEG patterns.’’ Frederick Gibbs is widely regarded as the true father of clinical
electroencephalography.
Finally, the ceding of care of patients with organic brain involvements like in
delirium, dementia and epilepsy to internal medicine, family practice, or neurology
helped further increase the distance between psychiatry and the clinical (visually
inspected and unquantified) standard EEG (here to for called sEEG). Nonetheless,
the field of psychiatry has never stopped probing the EEG and other electro-
physiological measures like evoked responses and sleep studies for correlates of
psychiatric disorders but has focused significantly more on the computer quantified
and processed signal aided by the progressive evolution of computing capacity.
Unlike the field of epilepsy where sEEG had strong clinical roots and where it
still remains as a mainstream clinical investigative procedure, the assessment of
suspected structural lesions has largely moved away from EEG to embrace the
more recent imaging techniques which admittedly are more definitive. Today
practitioners in metropolitan areas would seldom consider EEG as a first referral
option for suspected brain tumor. Nonetheless, one should not forget that in some
of the more rural locations EEG may be quite a bit more available than the newer
and more expensive imaging techniques and it still will detect focal slowing in
nearly 90 % of tumors of the outer cortex (Gibbs and Gibbs 1964). This factor
almost eliminated the one bona fide clinical indication for EEG in psychiatry
which is to rule out ‘‘organicity.’’
The Neurology discipline appreciates hard EEG data with well-documented and
strongly supported diagnostic relevance. Psychiatry must, by necessity and at least
at the current state of knowledge, be concerned with EEG findings that are
associated with a variety of altered behaviors and not necessarily with diagnostic
categories defined by current classification systems. The value of EEG findings in
psychiatry must be determined from within the field of psychiatry and cannot be
evaluated in terms of the clinical conditions deemed important by other disciplines
like internal medicine, family practice, or neurology.
The Renaissance of Electroencephalography in Psychiatry 5

The Renaissance of Electroencephalography in Psychiatry

Interest in the electrophysiology of psychiatric disorders is undergoing a renais-

sance; especially for research purposes (an example is the increasing size of
chapters on electrophysiological endophenotypes of various disorders including
schizophrenia, major depressive disorder, and bipolar disorder (Boutros and Hatch
1990). Multimodal imaging refers to the ability (again aided by the availability of
powerful computing capabilities) of linking electrophysiology data with imaging
data thus being able to zero in on events occurring in different brain structures with a
good degree of accuracy both spatially and temporally. Initiatives to promote EEG
applications in Psychiatry are today found in the mission of scientific societies, such
as the EEG and clinical neuroscience society (ECNS) and the Psychophysiology
Section of the World Psychiatric Association.

The American Psychiatric Electrophysiology Association:

History and Mission

The American Psychiatric Electrophysiology Association (APEA) was established

in 1991 as a direct result of the announcement of the American Board of
Psychiatry and Neurology (ABPN) to adopt a qualifying examination in clinical
neurophysiology. For all practical purposes, psychiatrists were excluded from
practicing clinical neurophysiology. Thus, the immediate goal of the APEA was to
ensure the rights of psychiatrists to practice clinical neurophysiology (Boutros
2000).
Predecessors The American Medical EEG Association (AMEEGA) has tradi-
tionally housed the majority of psychiatrists practicing clinical neurophysiology.
AMEEGA has encouraged the development of clinical applications of electro-
physiology in psychiatry. A conscious effort on part of AMEEGA was made to
disseminate knowledge about clinical neurophysiology (particularly EEG quanti-
fied or unquantified as well as polysomnography) and to recruit more psychiatrists
to AMEEGA.
To accomplish these goals, a Committee for the ‘‘Promotion of Electrophysiology
in Psychiatry, PEP’’ was formed in 1990. The committee included Edward Reilly
(Chair), Nash Boutros, Turan Itil, Marshal Bradshaw, Maurice Rappaport, Monte
Buchsbaum, and Andrew Leuchter. The committee met for the first time in New
Orleans during the 1991 Annual Meeting of AMEEGA and developed plans for
increasing the visibility of AMEEGA in the psychiatric community. These plans
included sponsoring an educational booth during the Annual Meetings of the
American Psychiatric Association (APA), advertising AMEEGA’s meetings in
psychiatric journals, and preparing AMEEGA-sponsored Clinical Neurophysiology
presentations in psychiatric meetings.
6 1 Philosophical Differences in Standard EEG Interpretation

Merger negotiation between the AMEEGA and the American EEG Society was
underway during the 1990–1991 period. The PEP committee felt that if AMEEGA
merged with another society where the percentage of psychiatrists is even smaller,
that the representation of psychiatric electrophysiology would be weakened. The
idea that psychiatrists need their own electrophysiology organization and perhaps
their own certification process was discussed among members of the PEP com-
mittee. The worries were communicated to AMEEGA leadership.
A similar and an independent effort was already underway. Drs. Leuchter, Re-
illy, Weiner, and Reynolds were successful in lobbying the APA to form an interest
group with similar goals to those of the PEP committee. At this time, the ABPN
announcement was made. The ABPN, in concurrence with the American Board of
Medical Specialties (ABMS), established a committee on certification of added
qualification in clinical neurophysiology in October 1990 (ABPN 1998). The ori-
ginal announcement limited the eligibility to child and adult neurologists as well as
psychiatrists. A grand-fathering period was granted until 1999 during which all
trained clinical neurophysiologists can qualify for the examination by providing a
letter indicating the extent of their training and their competency. From 1999
onwards, an applicant will have to complete a 12-month American-or Canadian-
approved clinical neurophysiology fellowship. Such fellowships are by and large
limited to neurology departments without accessibility for psychiatrists. The
immediate and urgent goal at this time was preserving the right of psychiatrists to
practice Clinical Neurophysiology. Dr. Reilly, of the University of Texas in
Houston, took the lead in initiating a letter and telephone campaign to reverse the
ABPN decisions to limit the field to neurology. The campaign succeeded in
allowing a 5-year grandfathering period for those already trained (including psy-
chiatrists) to set for the examination. The more pronounced effect is that the ABPN
decision made it clear that the clinical neurophysiology discipline as it is organized
currently is indeed a neurology subspecialty. The need became manifest for either
expanding the scope of the field to embrace the ever-increasing knowledge
regarding the electrophysiological aberrations associated with psychopathology or
develop a new discipline solely focused on the clinical use of electrophysiological
technology in the diagnosis, prognosis, and even treatment of psychiatric disorders.

The Birth of the APEA

Prior to the committee’s second meeting and during the mail and phone campaign
the idea emerged for a society dedicated to psychiatric electrophysiology that is
focused on promoting the development of clinical applications, promoting clinical
research utilizing electrophysiological measures, and providing a forum for a more
intense scientific exchange among clinicians involved in this type of research than
is allowed in other larger meetings (e.g., APA or Biological Psychiatry). The new
organization, thus, had already two major and fundamental differences from the
parent organization, AMEEGA. Namely, the new organization had a clinical
The Renaissance of Electroencephalography in Psychiatry 7

research focus versus the clinical practice emphasis within AMEEGA and wel-
comed both MDs and PhDs as members (AMEEGA required MD for membership).
The Executive Committee of the proposed organization met for the first time in NY
during the 1992 Annual Meeting of the APA. The committee decided on the name
of the organization, APEA, and Dr. Turan Itil was elected as the first president.
Dr. Itil led the organization for 3 years. Dr. Monte Buchsbaum (at Mount Sinai
School of Medicine at the time) became the second president of APEA in May of
1995. He was succeeded by Dr. Martin Reite as the third president of APEA in May
1996. Dr. Norman Moore was elected to be the fourth president of APEA in May
1998. He remained president till the merger of APEA and AMEEGA. Dr. E. Roy
John, of New York University was president elect at that time. The founding
members of APEA were N. Boutros, M. Bradshaw, M. Buchsbaum, R. Cancro, M.
Fink, T. Itil, E.R. John, M. Rappaport, E. Reilly, C. Shagass, J. Small, and G. Ulett.
During the first 2 years, APEA grew to 150 members with 20 % from outside
the USA. The First Annual Meeting was held in San Francisco on 22 May 1993. In
addition to the annual meetings APEA held a satellite meeting in conjunction with
the Collegium International for Neuro Psychopharmacology (CINP). This meeting
was co-sponsored by the World Health Organization. This satellite meeting
attracted an audience from around the world. Dr. Itil organized and chaired the
meeting. During the 3 year as president of APEA Dr. Itil secured recognition from
both the APA and the Society for Biological Psychiatry as both assigned official
representatives to APEA’s Scientific Advisory Committee (SAC). SAC was
assigned the task of forming subcommittees to examine available literature and
give recommendations for the clinical applications of the different electrophysi-
ological testing modalities in psychiatry.
Five subcommittees were formed: unquantified EEG, quantified EEG, poly-
somnography, evoked potentials, and EEG neuro feedback. Dr. Joyce Small
chaired SAC for 2 years and was succeeded by Dr. John Crayton. The task of the
subcommittees proved to be extremely difficult and only the EEG and QEEG
subcommittees were able to complete a report (Hughes and John 1999). A number
of lessons could be learned from the experience of the subcommittees. First, each
subcommittee had a substantial literature to review. This was exemplified by the
extensive literature listing included in the Hughes and John (1999) paper. Second,
it bacame apparant that the current state of the literature (possibly with the
exception of QEEG) did not strongly support many clinical applications in psy-
chiatry. Indeed, the simple and absolutely essential task of keeping up with psy-
chophysiology literature has become daunting (Boutros and Hatch 1990). Third,
translating clinical research findings to clinical applications is made complex by
the fact that in psychiatry, in addition to issues of sensitivity and specificity,
concepts like prevalence of the disorder in the examined population and cost of
misdiagnosis both in terms of dollars and human suffering must be taken into
account (Boutros et al. 1997). Furthermore, the significant co-morbidities among
psychiatric disorders as well as the significant heterogeneity within the disorders
were then and continue to be significant obstacles holding back the development of
clinically useful diagnostic tests.
8 1 Philosophical Differences in Standard EEG Interpretation

AMEEGA-APEA Interactions

Between 1995 and 1997 (Drs. Buchsbaum and Reite presidencies), much inter-
action between the two organizations took place. Particularly regarding the issues
of certification. The ABEN was closely affiliated with AMEEGA. Collaborative
efforts were underway to expand the scope of the examination to include recent
advances in psychiatric electrophysiology and to develop a certification process for
Ph. D. clinical neurophysiologists. The first goal was accomplished with the
inclusion of a significant number of psychiatric electrophysiology questions in the
written part of the examination. Also, the ABEN brochure was modified to
emphasize that training by a psychiatrist electroencephalographer and in a psy-
chiatric institution would qualify a specialty board eligible or certified physician to
sit for the ABEN examination. In addition subspecialty examinations were
developed for added qualifications in QEEG and polysomnography.
The second goal was more complicated. A committee for Ph.D. certification
was formed within APEA and worked closely with ABEN and AMEEGA. This
committee was first chaired by Patricia Tueting of the University of Chicago who
was succeeded by Frederick Struve of the Louisiana State University.

APEA-AMEEGA Rapprochement

The continuing interaction between AMEEGA and APEA highlighted the com-
monality of goals between the two organizations. Dr. Norman Moore played a
pivotal role in paving the way for the merger. A major obstacle was the desire of
APEA to open AMEEGA membership to PhDs. The second obstacle was assuring
APEA members of the continued support for clinical research. AMEEGA mem-
bers, on the other hand, needed assurance that the new organization would con-
tinue to address the needs of the practicing clinical neurophysiologists. After
approximately 18 months of tireless work, Dr. Moore, working with both boards
of directors, eliminated all the obstacles.

The Birth of the EEG and Clinical Neuroscience Society

(ECNS)

During the Last Annual Meeting of AMEEGA held October 1998 in New Orleans,
with Dr. Moore presiding over both organizations, the merger was approved by the
AMEEGA board. Already approved by the executive committee of APEA, the
merger was completed and a new organization the ECNS was introduced. The
ECNS immediately faced many challenges.
The Renaissance of Electroencephalography in Psychiatry 9

For the purposes of this chapter, we will focus on the challenges inherited from
the APEA. The most important challenge is defining the role of psychiatrists in the
field of Clinical Neurophysiology. Two broad models exist. One model calls for
one discipline practiced by any physician who is trained and qualified in this field.
We will call this the One-Discipline Model (ODM). A second model is that of
developing a new discipline with a behavioral focus. We will call this the Two-
Discipline model (TDM) (Pogarell et al. 2005).
The ODM has the advantage that the discipline already exists; qualification
processes are already in place; and journals, scientific societies, and training
programs already well established. One additional plus for the ODM is that the
knowledge base for all persons practicing in this area is comparable. Many
problems exist with this model. Most notably, the major focus areas of the field, as
it stands today, are neurological in nature. In other words, a non-neurologist who is
interested in practicing clinical neurophysiology will need extensive clinical
neurology training. Also, given the availability of qualified neurologists, it is
unlikely that hospitals, clinics or patients themselves will feel comfortable with a
non-neurologist interpreting an EEG of a neurology patient that was ordered for an
evaluation of a neurological disorder. Alternatively, psychiatrists who would like
to provide this service for their own psychiatric patients may want to be qualified.
In this case, the currently established training programs by and large do not spe-
cifically focus on EEG abnormalities in psychiatric patients. Moreover, most of
currently available fellowships do not provide any training on many of the newer
technologies that offer promise for psychiatry (e.g., QEEG, event-related poten-
tials (ERPs), multimodality registration). It should also be noted that available
clinical neurophysiology training programs provide little or no training for sleep
disorders and polysomnography. The ODM necessitates that training programs be
designed to cover the entire scope of the field as well as being equally accessible to
both psychiatrists and neurologists. Under ordinary circumstances, psychiatrists
would not be accepted into these fellowships due to other clinical responsibilities.
The TDM presents some advantages and a set of different problems. Specific
qualifications different from those needed to practice clinical neurophysiology will
need to be developed. The current body of knowledge of the clinical applications
of electrophysiological testing in psychiatry is progressively expanding. While the
literature suggests that testing modalities like Q-EEG, ERP, and polysomnography
(PSG) hold significant promise for improving the practice of clinical psychiatry,
the usefulness of these tests have not been tested in well designed, large multi-
center studies. At the moment, such a field would be constituted of clinical EEGs
and limited applications of Q-EEG and sleep studies. Indeed such applications
alone would be sufficient to establish busy laboratories given the volume of psy-
chiatric patients. Academic electrophysiology laboratories in psychiatry depart-
ments could benefit from this model. The revenues generated from the limited
clinical applications can be channeled, at least in part, toward research endeavors.
In the current environment of extreme difficulty in securing research support, this
model could help this field propel it self. Indeed the volume of research necessary
to fully delineate the usefulness of each of the electrophysiological testing
10 1 Philosophical Differences in Standard EEG Interpretation

modalities is an immense task. In a recent review, we found that the boundaries of

normality for the conventional EEG, for psychiatric research purposes, have never
been adequately defined (Boutros et al. 2005). It should be noted that such
empirical or atheoretical research, requires significant funds, yet is unlikely to be
funded by most funding agencies.
Another major variance from the clinical neurophysiology discipline is the
inclusion of electrophysiological therapeutic modalities under the auspices of
psychiatric electrophysiology: electro-convulsive therapy (ECT), transcranial
magnetic stimulation (TMS), vagal nerve stimulation (VNS), transcranial electr-
ostimulation (TES) and biofeedback including EEG neuro feedback, as well as the
more investigative modalities like magnetic seizure therapy (MST), and deep brain
stimulation (DBS).
The ability to interpret the clinical implications of a deviance of an electro-
physiological measure in a psychiatric patient will require adequate knowledge of
the clinical syndromes, psychophysiology, psychopharmacology, as well as some
knowledge of psychology. None of these rather large bodies of knowledge are
needed for the competent practice of clinical neurophysiology. With the hope of
rapid expansion of the field, such laboratories could become an essential part of
every department of psychiatry and even with every group of practicing psychi-
atrists. The obvious obstacles here are the lack of training programs and a certi-
fication process. Both need to be developed. Indeed there is not a single textbook
in the market that outlines such field. L. Reilly wrote in a letter dated 6/1/1993 to
J. Scully Jr., then director of the office of education of the APA. ‘‘It is my opinion
that the APA office needs to vigorously, consistently and aggressively argue that
we should be able to train our own residents, in our own residency programs, if we
chose to do so. In addition psychiatrists should have an exam more suited to the
way psychiatrists practice in the field of neurophysiology.’’
Many obstacles face the ECNS in establishing either of the two models. The
most important obstacle is the lack of awareness among psychiatrists regarding the
richness of scientific findings of electrophysiological aberrations associated with
psychopathology. The second important obstacle is the territoriality between
Neurology and Psychiatry. In this day and age of limited resources, for a discipline
to give up even a small part of its business is extremely difficult. Such an obstacle
will be overcome only if the benefits of collaboration between the departments
become obvious. Third, scientific evidence that the use of either diagnostic or
therapeutic electrophysiological modalities is cost effective should be provided
through well-designed multicenter collaborative studies. Lobbying third party
payor, in an era of managed care, will prove to be an uphill battle.

Why Should Psychiatrists Pay Attention to the EEG of Their Patients?

The answer to this question is quite simple. This is a matter of cost benefit ratio. The
value of any test resides in the relative value of the information gained versus, the cost
of the test and its level of invasiveness (i.e., danger and inconvenience to patient).
The Renaissance of Electroencephalography in Psychiatry 11

In order to address this issue, it would be easier to address the cost and inconvenience
first. To start with, the EEG is a completely noninvasive test (perhaps with the
exception of the possible insertion of sphenoidal electrodes which will be discussed
in detail in the chapter on special electrode placements). EEG equipment has been
traditionally, and remains rather inexpensive. Most currently available commercial
EEG systems falls below 50 K and some are considerable less than that. EEG sys-
tems are durable and work for many years without need for much maintenance or
upgrades. Hence, the cost of the machine can be amortized over many years. The
actual major costs come from the personnel involved; the EEG technologist and the
EEG interpreter. Performing an EEG usually lasts less than 90 min. At $30/h a
reasonable estimate of the cost for the technologist/procedure should be under $50.
The clinical interpretation of a standard 30–40 min EEG consumes less than 15 min
from an expert EEGer. At $300/h, the cost for the interpreting physician should be
under $100. Assuming another $50 for all other overhead expenses (billing and
secretarial work), the actual cost of the procedure to the institution should be just
around $200. It is customary to charge $300–400/procedure thus in fact making a
profit for the EEG laboratory whether private or part of a larger institution. The main
issue being raised in this volume is the need for much research to better define the
medical usefulness of the data generated and hence improving the reimbursement
rates. To summarize, the standard EEG is both noninvasive and relatively inex-
pensive (Boutros et al. 2011a).
The more difficult issue to tackle is the actual value of the test. And here I
propose that given the low cost and noninvasiveness of the test that almost any
additional information the test yields will cause the balance to tip in favor of the test.
Between our prior volume ‘‘Standard EEG in Clinical Psychiatry’’ (Boutros et al.
2011b) and the many published chapters outlining the clinical utility of the sEEG,
the judgment on the cost-benefit analysis must be left for each clinician to decide.

Looking for Small Changes

During my 2 years of training on how to interpret the sEEG the concept of dis-
regarding minor EEG changes was deeply stressed. In the 1st year, the motto of
training was ‘‘If in Doubt/Throw it Out’’ meaning if the abnormality is not very
clear and undeniable, we must not worry about it too much. In the second year of
training, the motto was ‘‘if it does not Jump out at you, it is not there.’’ Further-
more, the standard that if an abnormality is seen only once (and unless it is
undeniably abnormal) it should be disregarded, further stresses the emphasis on
major and well-defined abnormalities. The issue here is simple. If the patient in fact
has epilepsy, the patient will sooner or later have a seizure. Hence, why base the
diagnosis that is stigmatizing on a questionable event on the EEG. The converse is
true for psychiatric conditions. Assume a patient with panic attacks exhibits a
single epileptiform discharge on the EEG. Disregarding this one episode would
12 1 Philosophical Differences in Standard EEG Interpretation

result in the EEG being interpreted as normal. The psychiatric clinician (being
mostly unaware of the serious limitations of the sEEG) would assume that in this
patient epileptiform activity do not exist and treat accordingly. It is also a tendency
that once an EEG is obtained and reported as normal, that a repeat EEG is almost
never obtained. If, on the other hand, the EEGer reports what is detected as
questionable and in need of further examination, the most that will happen is
repeating the EEG or obtaining an image (CT scan or MRI). It is also possible that
if the patient proves resistant to standard treatments (for example, a panic disorder
patient who is unresponsive to selective serotonin or serotonin/nor-epinephrine re-
uptake inhibitors), that a trial of an anti-epileptic drug (AED) may be attempted
given that the EEG was not interpreted as entirely normal. Moreover, the further
pursuant of the EEG deviation may yield significant information that would be
useful in a more complete bio-psychosocial formulation of the patient’s condition.

Stigmatization Versus Nonstigmatization

A rather crucial difference between psychiatry and neurology is the relative stig-
matizing influence of the different disorders. In the current day and age, psychiatric
disorders are by far much more stigmatizing than any neurological disorder
including epilepsy. In my experience as well as the experiences of many of my
colleagues, the identification of a biological abnormality (EEG or otherwise) in a
psychiatric patient is always received by both the patient and his/her family with
much welcome. In addition to stressing that the problem the patient is having is a
‘‘real’’ one and is indeed a brain problem, usually also means that some form of a
treatment may be based on the finding. While it is now customary to talk to
patients and families about ‘‘chemical imbalance’’ when talking about schizo-
phrenia and mood disorders, the actual demonstration of a brain abnormality
brings the point home in a much more real way. Thus, for psychiatry, giving the
patient the benefit of the doubt is to NEVER under-interpret the record. When in
doubt, more testing should be performed until the yield from such testing arrives at
the point of no additional value.

Changes Over Time

As is strongly stressed in the remainder of this book, many unanswered questions

remain (in fact much more questions exist than answers). With the advent of
psychiatry-based clinical EEG laboratories, much more data are likely to be
generated regarding the sEEG in the various psychiatric populations. Thus, it is
essential that psychiatric EEGers remain abreast of the knowledge in the field and
modify both technical and interpretive standards based on new knowledge
emerging in the field.
The Renaissance of Electroencephalography in Psychiatry 13

Cannot Stop Research Tell Correlates are Identified

Here I would like to discuss the term ‘‘nonspecific.’’ It is well known that iden-
tifying the precise behavioral or psychiatric correlates of any biological abnor-
mality is a rather challenging and difficult task. The most essential observation is
an increased prevalence of an abnormality in a group of psychiatric patients. Once
such is demonstrated, the observation cannot be called nonspecific. Assume the
abnormality is in fact seen equally prevalent in all psychiatric populations (there is
of course no such abnormality reported to date) then the abnormality must still
mean something specific to psychiatric population and it would be of significant
value to uncover the relationship. The simple disregarding of the observation as
nonspecific only helps to decrease the since of value of investigating such a
deviation.

Use All Data Obtained Including Artifact, Most Importantly Eye

Movement and Muscle Tension

The standard teaching in EEG interpretation is to disregard all recorded activity

believed to be of extra-cerebral sources like eye movement or muscle activity.
While the standard EEG is not suitable for adequate testing of these biological
activities, an observation made from the sEEG may be the impetus for further
obtaining more appropriate testing like examining saccadic, pursuit or even
blinking activity (Boutros et al. 1990). As these procedures progressively mature
toward clinical utility, the psychiatric EEGer must be aware of these related fields
as well.

All Psychiatric EEGers Must be Thoroughly Knowledgeable About

Quantification of the EEG

As is plenty evident from the already huge and rapidly growing literature on the
quantified EEG and evoked responses in psychiatric conditions, the major future
expansion in psychiatric electrophysiology will be in the quantification of the
signal. This important major advance is being facilitated by the fact that all current
standard EEG systems are digital. Thus, the recording of the EEG one time covers
both the standard visual interpretation and the further computer-based
quantification.

Concluding Remarks

One must wonder about the current status of sEEG in psychiatry given the massive
already existing literature and life works of such pioneers like Frederic Gibbs,
Russell Monroe, Jr., George Fenton, Riley TL, Ernest Rodin, William P. Wilson,
14 1 Philosophical Differences in Standard EEG Interpretation

Joyce C. Small, John R. Hughes, Edward Reilly as well as countless others?

Fenton and Standage (1993) already documented the significant usefulness of the
standard EEG in the practice of psychiatry. They found that 11 % of referrals to a
general EEG laboratory were psychiatric. In their practice, 37 % of these records
were definitely abnormal with an additional 19 % that were labeled ‘‘anomalous.’’
They concluded that 92 % of the EEGs to have been clinically useful to the
referring clinician. The fact is that literature remains very useful and should guide
future work in this field.

References

American Board of Psychiatry and Neurology Inc (1998) Information for applicants for
certification in the subspecialty of geriatric psychiatry, clinical neurophysiology, addiction
and forensic psychiatry, p 11–14
Boutros NN (2000) The American psychiatric electrophysiology association (APEA). Hist
Mission Clin EEG 31:67–70
Boutros NN (2010) Isolated epileptic discharges in non-epileptic psychiatric patients, a
(controversy in need for resurrection). J Clin Neurophysiol 27(6):484–485
Boutros NN, Hatch JP (1990) A bibliographical analysis of trends in clinical psychophysiology.
J Psychophysiol 4(2):194–195
Boutros NN, Fabian WA, Warner MD, Peabody CA (1990) Psychiatric correlates of repetitive
rhythmic blinking on routine EEG. Clin Electroencephalogr 21:196–199
Boutros N, Nasrallah H, Leighty R, Torello M, Tueting P, Olson S (1997) The mid-latency
auditory evoked responses clinical versus research applications. Psychiatry Res 69:183–195
Boutros NN, Mirolo HA, Struve F (2005) Normative data for the unquantified EEG: examination
of adequacy for neuropsychiatric research. J Neuropsychiatry Clin Neurosci 17(1):84–90
Boutros NN, Gjini K, Arfken CL (2011a) Advances in electrophysiology in the diagnosis of
behavioral disorders. Expert Opin Med Diagn 5(5):441–452
Boutros NN, Galderisi S, Pogarell O, Riggio S (2011b) Handbook of standard EEG in clinical
psychiatry. Wiley, Blackwell, Hoboken, NJ
Fenton GW, Standage K (1993) Clinical electroencephalography in a psychiatric service. Can J
Psychiatry 38(5):333–338
Fischer MH (1933) Elektrobiologische Auswirkungen von Krampfgiften am Zentralnervensystem.
Med Klin 29:15–19
Gibbs FA, Gibbs EL (1950) Atlas of electroencephalography, vol 1. Addison-Wesley Press,
Cambridge, pp 1–5
Gibbs FA, Gibbs EL (1964) Atlas of electroencephalography, vol 3. Addison-Wesley Publishing
Company, Reading, Boston, pp 337–393
Gibbs FA, Davis H, Lennox WG (1935) The electroencephalogram in epilepsy and in conditions
of impaired consciousness. Arch Neurol Psychiat 34:1133–1148
Gibbs FA, Gibbs EL, Fuster B (1947) Anterior temporal localization of sleep-induced seizure
discharges of psychomotor type. Tr Am Neurol A 79:180–182
Hughes J, John ER (1999) Conventional and quantified EEG in psychiatry. J Neuropsychiatry
Clin Neurosci 11(2):190–208
Pogarell O, Hegerl U, Boutros N (2005) Clinical neurophysiology service in psychiatry
departments. Psychiatr Serv Frontline Rep 56(7):871
Chapter 2
What Constitutes a Normal EEG

Introduction

The usefulness of the routine standard (i.e., visually inspected) electroencephalo-

gram (sEEG) in psychiatric practice and research is significantly hampered by the
reported prevalence of EEG abnormalities in ‘‘normal’’ adult populations, ranging
from 4 % to as high as 57.5 % (Struve 1985). This wide range likely reflects the lack
of clear and rigorous standards for choosing subjects for healthy comparison groups.
Blanc et al. (1964) documented that the inclusion of psychiatric patients in healthy
comparison samples contributes to increased prevalence of EEG abnormalities in
the examined sample. This observation was reported as early as 1939 (Davis and
Davis 1939) and remains unchallenged today. In order for this technique to be
reliable and useful, the boundaries of normality should be well defined.
The danger of reliance on a subject’s self-report of normalcy was highlighted
by Halbreich et al. (1989) who showed that in a sample of self-proclaimed
‘‘normal volunteers,’’ 16.5 % met criteria for diagnosis of a current mental dis-
order and of the subjects without a current mental disorder, as many as 35 % had
past histories and 39 % had family histories of mental illness. This is particularly
important because physiological differences have been found between ‘‘normal
subjects’’ with and without family histories of mental disorders (Schuckit 1984;
Wiesel et al. 1982). Inclusion and exclusion criteria used for patient selection have
also been progressively more restrictive, particularly for imaging and physiolog-
ical studies. The presence of general medical or neurological conditions that may
influence particular measurements are routinely used as exclusion criteria. Whe-
ther a subject is receiving medications that may affect brain functions is either
exclusionary or well controlled for in the majority of studies published in psy-
chiatric peer-reviewed literature during the last decade or longer and is an essential
requirement for most granting agencies. Furthermore, issues of drug abuse and
dependence, presence of Axis-I and Axis-II disorders are crucially important. In
order to assess the actual usefulness of currently established boundaries of nor-
mality for psychiatric investigations, an extensive search of the literature included
in Medline and PsychInfo databases for all articles listing EEG as a keyword was

N. N. Boutros, Standard EEG: A Research Roadmap for Neuropsychiatry, 15

DOI: 10.1007/978-3-319-04444-6_2,  Springer International Publishing Switzerland 2013
16 2 What Constitutes a Normal EEG

Table 2.1 Normality criteria used to evaluate published chapters

Number Historical or current absence of
criterion
1 Systemic disorders with CNS involvement (e.g., endocrine or metabolic
disorders)
2 Neurological disorders including traumatic brain injurya, history of childhood
neurological disorders, and dementing illnesses
3 Psychiatric disorders including all Axis-I disorders except alcohol and drug abuse
or dependence
4 Alcohol and drug abuse or dependence
5 Receiving CNS active medications including any psychotropic medications
6 First-degree relatives with family history of Axis-I psychiatric disorders
7 Axis-II disorders including personality disorders and mental subnormality
a
History of head injury is particularly problematic as definition of what is a ‘‘significant head
injury’’ is not known or agreed upon. Most current research groups require at least a momentary
loss of consciousness but different studies do require varied length of loss of consciousness

performed (Boutros et al. 2005). Textbook chapters discussing ‘‘normal EEG’’

were also examined for references (Niedermeyer 1993). These two sources were
the primary sources for references. The reference lists of each chapter or book
chapter were searched for older relevant chapters. Articles pertaining to quantified
or spectrally analyzed EEG, articles examining evoked responses or sleep EEG,
and articles that did not include normal subjects or did not study humans were
excluded from the study. Additionally, review articles and abstracts were exclu-
ded. Seven criteria that are commonly used in contemporary neuropsychiatric
research for selecting healthy comparison subjects were chosen as the bases for
this review (Table 2.1).
Criteria were ranked as present or absent for each article. If a criterion could not
be confidently rated as present, the criterion was marked as absent. As the majority
of studies identified by this search were performed prior to the publication of the
DSM criteria, the simple mention of a diagnosis was credited as correct.
A notably large number of studies simply stated that subjects were ‘‘healthy.’’
The simple assurance of normality was considered inadequate for our purpose, and
they were marked as lacking all seven normality criteria. While this is unlikely to
have been the case, we elected to err on the side of rigorous reporting. Among the
seven criteria were: presence of a general medical condition that may affect the
EEG, any neurological conditions including history of head injury leading to any
length of loss of consciousness, and history of any psychiatric disorders. History of
drug dependence or abuse (excluding tobacco or social use of alcohol), receiving
any CNS active medications, family history of psychiatric disorders, and presence
of an Axis-II disorder were also used as criteria for normality. The inclusion of
family history as an exclusionary criterion is based on data showing increased
EEG abnormality in family members of psychiatric patients (Chamberlain and
Russel 1952).
Introduction 17

A total of 39 articles examining the EEGs of ‘‘normal’’ individuals were

included. Articles either examined EEGs in ‘‘normal subjects’’ as the sole purpose
of the study or included a ‘‘normal’’ comparison group as a comparison for a
pathological group (Boutros et al. 2005).
A history or presence of a general medical condition, a neurological disorder, or
a major psychiatric disorder were three conditions considered to be most important
in determining the exclusion of normal subjects. All three criteria were considered
to have been met in a total of four studies. Five additional papers limited exclusion
criteria to medical and neurological disorders. Excluding subjects solely based on
history or the presence of a systemic disease that is likely to affect CNS functions
(e.g., hypo or hyperthyroidism) was indicated in two more studies. Exclusion
based solely on the presence or history of a neurological problem was seen in one
study (White et al. 1977). In this study, the criterion for excluding subjects based
on receiving CNS active medications was also met. Another study excluded
subjects based on psychiatric history but without specific criteria for neurological
or medical conditions (Chamberlain and Russel 1952). In this study, a family
history of a psychiatric problem was also an exclusionary criterion. A single study
screened for drug abuse (Buchtal and Lennox 1953). In two studies, subjects
receiving psychotropic medications were excluded (White et al. 1977; Nowak and
Marczynski 1981). None of the studies excluded subjects receiving such non-
psychotropic medications as steroids or centrally acting antihypertensive medi-
cations. Two papers reported collecting family history data and one other paper
reported screening for personality disorders, but these were not considered
exclusionary criteria. This is important, as the EEG literature is replete with
reports of abnormalities in association with different personality disorders, par-
ticularly antisocial, and borderline types. Standardized psychiatric scales (e.g., The
Structured Clinical Interview for the DSM-SCID) (Spitzer et al. 1992) were not
administered in any of the studies. Studies relied mainly on historical denial of
psychiatric symptomatology.
The most recent relevant paper we were able to find is that by Jabbari et al.
(2000). In this paper, the authors report on the EEGs of 100 male subjects ranging
in age from 18 to 45. These subjects were healthy controls for a number of
neurological studies. They were all active duty soldiers. History of seizures, head
injury (specifics of any loss of consciousness duration not provided), and history of
drug abuse (not specified if meeting DSM-III criteria), were the exclusion criteria.
It is worth noting that what would be a current day minimal requirement to exclude
psychiatric disorders (a structured or semi-structured diagnostic interview like the
SCID were not administered), nor a family h/o psychiatric disorders reported. The
results of this study are none the less very important as they found significant
abnormalities like focal or generalized slowing or unequivocal epileptic discharges
to be completely absent while 12 % had positive spikes and 11 % had small sharp
spikes. These EEG phenomena are discussed in much detail in later chapters.
18 2 What Constitutes a Normal EEG

Discussion and Conclusions

The literature review described above indicates that the overwhelming majority of
EEG normative studies were performed prior to the advent of normality criteria
currently applied by most research institutions in research endeavors involving
psychiatric or neuropsychiatric populations. In general, our search indicates that
the criteria for normality taken into consideration in currently available literature
for ‘‘normal’’ analog EEG range from poor to absent. A neurological history and
systemic pathologies impinging on the central nervous system were only implicitly
excluded in the majority of studies. The specific exclusion of significant head
injury was also lacking. Traumatic brain injury has an annual incidence of 370 per
100,000 (Kurtzke 1984). Traumatic brain injury can lead to personality changes
(McKinlay et al. 1981), affective disorders (Rutherford et al. 1977), or even
psychotic syndromes (Lishman 1987). Medications or psychoactive substances
contaminating the picture were not considered as a factor.
The largest studies establishing normality of EEG were conducted on Navy
candidates. These studies specifically addressed personality disorders. While
subjects were not excluded, data on personality disorders were collected but not
reported in the publications. The importance of rigorous exclusionary criteria was
demonstrated by Buchthal and Lennox (1953). In their sample, 5.4 % of candi-
dates who were refused admission to the Navy on psychiatric grounds had par-
oxysmal EEG abnormalities, while only 2.2 % who were admitted and completed
the training had paroxysmal abnormalities.
A number of additional factors that contribute to the diminished value of
currently existing norms also emerged from our search (specifically, small sample
sizes in a number of studies, lack of established criteria for EEG normality, and
disagreement among electroencephalographers regarding the significance of the
so-called controversial waveforms).
Based on the above findings, we conclude that the boundaries for normal sEEG
are not well defined for the purposes of neuropsychiatric clinical or research
endeavors. In order to be able to better define and study sEEG abnormalities in
neuropsychiatric populations, well-designed normative studies are needed. Spe-
cifically, future studies should not rely on a single normal routine EEG to conclude
lack of abnormalities. Similarly, the value of securing sleep tracings cannot be
overemphasized.
A major role of EEG is to reduce the heterogeneity of research studies (e.g.,
depression with or without localized abnormalities, aggression with or without
spikes). In Table 2.1, we have delineated what we consider adequate factors to be
taken into consideration for the inclusion of a subject as normal in a study.
Obviously, different types of studies will require different exclusion criteria. For
example, studies attempting to develop normative databases or criteria for general
use should observe the most stringent criteria. If a person or a group then deviates
from such norms, the cause or causes of the deviations can then be investigated in
subsequent studies specifically designed to isolate specific possible contributing
Discussion and Conclusions 19

factors. On the other hand, studies comparing specific patient populations to

normal comparison groups may wish to allow some of the factors (e.g., history of
drug abuse or head injury) based on study design or be more stringent (e.g., studies
examining genetics of an EEG pattern may extend the family history exclusion
beyond first-degree relatives).
Blanc et al. (1964), highlighted the problem of cross-sectional studies. Via case
examples, they pointed out that EEGs may change its characteristics at different
time points. They related these changes mainly to change in psychiatric status.
Additionally, Chamberlain and Russell (1952) pointed out that first-degree rela-
tives (particularly siblings) of schizophrenia patients may have higher prevalence
of EEG abnormalities. This finding suggests that family history of Axis-I disorders
(at least in first-degree relatives) should be an additional exclusion criterion for
normative studies.
The above review indicates that the high prevalence of abnormal EEGs in normal
populations, ranging from 5 to 20 %, is based on inadequate inclusion and exclusion
criteria for healthy comparison subjects. We conclude that the boundaries for EEG
normality are poorly defined as they currently stand and are invalid for drawing any
conclusions regarding prevalence or significance of EEG abnormalities in psychi-
atric populations. The EEGs of large samples of well-characterized healthy indi-
viduals meeting the criteria specified in Table 2.1 need to be examined in order to
provide more clearly defined boundaries of normality and to establish more uniform
criteria for abnormality. While all efforts were made to obtain all published papers
and book chapters addressing the development of normative EEG criteria, it is
unavoidable that a number of such publications were not securable (most of this
literature dates back 40–70 years). Similarly, it was not possible to contact the
individual investigators to verify normality criteria used (for the same reason).
Nonetheless, the data presented above strongly suggest the need for new research to
help define the normal boundaries of the unquantified EEG.

Supported Findings

(1) The exact boundaries of normality of the sEEGs are currently not well defined
for the purposes of psychiatric EEG research.
(2) There are no normative studies in healthy subjects with repeated testing over
time.

Open Research Questions

(1) What is an adequate EEG work-up in order to determine normality or

deviations?
(2) What are the boundaries of EEGs obtained from well-characterized healthy
children, adults, and elderly individuals?
20 2 What Constitutes a Normal EEG

References

Blanc C, Lafontaine E, Laplane R (1964) Meaning and value of electroencephalography in

aeronautical medicine. Aerosp Med 35:249–256
Boutros NN, Mirolo HA, Struve F (2005) Normative data for the unquantified EEG: examination
of adequacy for neuropsychiatric research. J Neuropsychiatry Clin Neurosci 17(1):84–90
Buchthal F, Lennox M (1953) The EEG effect of Metrazol and photic stimulation in 682 normal
subjects. EEG Clin Neurophysiol 5:545–558
Chamberlain HA, Russell GJ (1952) The EEGs of the relatives of schizophrenics. J Ment Disord
8:654–659
Davis H, Davis PA (1939) Active potentials of the brain in normal persons and in normal states of
cerebral activity. Arch Neurol Psychiatry 36:1214–1224
Halbreich U, Bakhai Y, Bacon KB (1989) The normalcy of selfproclaimed ‘‘normal volunteers’’.
Am J Psychiatry 146:1052–1055
Jabbari B, Russo MB, Russo ML (2000) Electroencephalogram of asymptomatic adult subjects.
Clin Neurophysiol 111(1):102–105
Kurtzke JF (1984) Neuroepidemiology. Ann Neurol 16:265–277
Lishman WA (1987) Organic psychiatry: the psychological consequences of cerebral disorder,
2nd edn. Blackwell Scientific, Boston
McKinlay WW, Brooks DN, Bond MR (1981) The short term outcome of severe blunt head
injury as reported by the relatives of the injured person. J Neurol Neurosurg Psychiatry
44:527–533
Niedermeyer E (1993) The normal EEG of the waking adult. In: Niedermeyer E, Da Silva FL
(eds) Electroencephalography: basic principles, clinical applications, and related fields, 3rd
edn. Williams and Wilkins, Baltimore, pp 131–152
Nowak SM, Marczynski TJ (1981) Trait anxiety reflected in EEG alpha response to stress.
Electroencephalogr Clin Neurophysiol 52:175–191
Rutherford WH, Merrett JD, McDonald JR (1977) Sequelae of concussion caused by minor head
injuries. Lancet 1:1–4
Schuckit MA (1984) Subjective responses to alcohol in sons of alcoholics and control subjects.
Arch Gen Psychiatry 41:879–884
Spitzer RL, Williams JBW, Gibbon M et al (1992) The structured clinical interview for DSM-III-
R (SCID), I: history, rationale, and description. Arch Gen Psychiatry 49:624–629
Struve FA (1985) Clinical electroencephalography as an assessment method in psychiatric practice.
In: Hall RC, Beresford TP (eds) Handbook of psychiatric diagnostic procedures, vol 2.
Spectrum Publications, New York, pp 1–48
White JC, Langston JW, Pedley TA (1977) Benign epileptiform transients of sleep: clarification
of the small sharp spike controversy. Neurology 27:1061–1068
Wiesel FA, Fyro B, Nyback H (1982) Relationships in healthy volunteers between secretion of
monoamine metabolites in urine and family history of psychiatric morbidity. Biol Psychiatry
17:1403–1413
Chapter 3
Special Electrodes

Introduction

The standard 10–20 international electrode placement system (American EEG

Society 1990) targets mainly the lateral surface of the cerebral hemispheres. Many
of the brain regions of particular salience to psychiatric symptomatology are
deeper structures like the medial and orbitofrontal regions, insular temporal,
hippocampal, or amygdalar areas. As many of these deeper regions are also
common sources for epileptiform discharges (EDs), there have been attempts to
develop special electrodes or special electrode placements that would improve the
odds of detecting EDs emanating from those regions. Most notably are the true
anterior temporal electrodes, nasopharyngeal, and sphenoidal electrodes.
By and large, none of these electrode placements or special electrodes has been
systematically examined in psychiatric populations. In this chapter, we will discuss
the noninvasive placements and the sphenoidal electrodes. No discussion of the
nasopharyngeal electrodes will be included as these are invasive electrodes ren-
dering patients uncomfortable even in the most experienced hands and thus it is
highly unlikely that their applications in psychiatric populations would be useful.
Documentation of epileptiform activity from the medial and deeper aspects of
the temporal lobes is frequently very difficult. An important study showed that
when repeated EEGs were obtained in a period over 1 year from patients with
documented temporal lobe epilepsy (TLE), epileptiform activity was document-
able in 98 % (Ajmone Marsan and Zivin 1970). Given the documented epilepsy in
these individuals, the reason and drive for persisting in trying to document the
abnormalities that are presumed to be there was understandable. This is much
more difficult when, as in patients with psychiatric presentations, it is not known
whether these individuals do have epileptic discharges. Studies attempting to
document the rate of detection of EDs in different psychiatric groups with repeated
recording over extended periods of time and utilizing special electrode placements
have never been performed.

N. N. Boutros, Standard EEG: A Research Roadmap for Neuropsychiatry, 21

DOI: 10.1007/978-3-319-04444-6_3,  Springer International Publishing Switzerland 2013
22 3 Special Electrodes

T1 and T2 Electrodes

The documented clinical significance of the anterior temporal regions has led to
interest in being able to record activity emanating from the temporal poles. It has
been postulated that the international 10–20 international electrode placement
system neglects the anterior temporal regions and specifically the temporal poles
because the T7 and T8 electrodes lie anterior to the location of the temporal poles
(Nowack 1988). Traditionally, placement of electrodes close to the temporal poles
has led to these electrodes being called T1 and T2. It should be emphasized that
these electrode locations are not routinely utilized. Hence, despite the modern day
ability to re-montage electronically after the recordings, if these electrode loca-
tions were not utilized to start with, data from these regions are simply not
obtainable.
Traditionally, a T1 electrode would be halfway the distance between from F7 to
T3 and the electrode would lie on the 10 % circumferential line around the head
(Jasper 1958). In the updated terminology there are no T1 or T2 and the positions
are called FT7 and FT8 (American EEG Society 1990). In the United States, a
common placement of the T1 and T2 electrodes was described by Silverman
(1965). The electrodes are placed one-third anterior to the external auditory canal
and one centimeter above the line between the auditory external canal and the
lateral canthus of the eye. The value of these electrodes in patients with pure
psychiatric presentations is currently entirely unknown.

Sphenoidal Electrodes

In the standard International 10–20 system for EEG electrode placements, the
more inferior chains that cover the lateral temporal lobes do so anatomically at the
levels of the superior and middle temporal gyri. To electrically ‘‘see’’ activity that
is of a more basal temporal origin, perhaps activity arising from the inferior
temporal gyrus, the parahippocampal gyrus, or even hippocampal gyrus, electrodes
need to physically approach those areas (Schomer 2003). For that reason, elec-
troencephalographers introduced recording montages that included the earlobe
electrode, the tympanic membrane electrode, the nasopharyngeal electrode, the
zygomatic arch electrode, and the T1/T2 electrode. Any of these ‘‘basal’’ elec-
trodes is likely to be more sensitive to electrical activity originating in more basal
temporal structures if for no other reason than proximity.
Jones (1951) first reported the use of the ‘‘sphenoidal’’ electrode. Figure 3.1
depicts the location of the recording tip of the sphenoidal electrode. Jones (1951)
used a fine, insulated needle electrode that is inserted, under light local anesthesia,
through the notch in the mandible just under the zygomatic arch. This approach is
still used today and is labeled ‘‘anterior sphenoidal.’’ The recording tip of the
electrode rests at the base of the skull lateral to the foramen ovale. Marshall (1957)
Sphenoidal Electrodes 23

Fig. 3.1 The figure depicts the location of the recording tip of the sphenoidal electrode. As can
be seen the tip of the sphenoidal electrode lies at a much closer proximity to the medial temporal
lobe structures as compared to surface electrodes

then introduced a flexible piano wire instead of the rigid needle and thus allowed
physicians the option of recording from these electrodes over longer periods.
Subsequently, a variation on the standard sphenoidal electrode placement allowing
for two different locations, one just anterior to the pterygopalatine fossa and the
other inferior and lateral to the foramen ovale were introduced (Rovit et al. 1961).
Studies at that time have already shown that often the sphenoidal electrodes
showed abnormalities that were not seen on surface or nasopharyngeal electrodes.
Christopoulou (1967) further supported the earlier assertions regarding the use-
fulness of the sphenoidal electrodes. In his study of 104 cases of suspected TLE
with nondiagnostic routine EEGs they showed that the sphenoidal electrode gave
specific and diagnostic abnormalities in 54 cases.
Sperling and colleagues (1986) performed a large study that assessed the
interictal EEG findings in 45 patients with complex partial epilepsy and compared
three electrodes in their abilities to detect spikes (sphenoidal, earlobe, and naso-
pharyngeal). In 25 of the patients studied, interictal activity was detected using the
sphenoidal electrode; 23 patients were positive using the earlobe electrode, and 20
were positive using the nasopharyngeal electrode. Of a total of 875 individual
interictal spikes, 99 % were seen by the sphenoidal electrode while only 57 %
were seen on the nasopharyngeal electrode and 54 % on the earlobe. Their con-
clusion was that the sphenoidal electrode was far superior to the other two ‘‘basal’’
electrodes, the nasopharyngeal, and earlobe, in detecting spikes of mesial or
inferior temporal lobe origin. So et al. (1994) examined 101 patients suspected of
having TLE and who had initial normal scalp EEGs. These patients were examined
following sleep deprivation. In 11 subjects, EDs were detected only in the sphe-
noidal electrodes whereas only three had EDs in there scalp locations and not in
the sphenoidal electrodes.
24 3 Special Electrodes

There appears to be little disagreement that the sphenoidal electrode is superior

to the 10–20 international system scalp electrodes in detecting the interictal spike
discharges of focal TLE. It also appears to be better than some ‘‘basal’’ electrodes,
specifically the earlobe and nasopharyngeal electrodes. The latter are probably
influenced principally by neuronal activity originating in middle and superior
temporal gyri while the former record activity originating in the inferior temporal
gyrus, along the basal temporal lobe anteriorly, the anterior tip of the temporal
lobe, and perhaps from the posterior portions of the orbital frontal lobes. However,
it does not appear to be better than the surface electrodes placed at the T1/T2 site.
Given the documented importance of these regions to a variety of psychiatric
illnesses it is rather strange that neither the sphenoidal or the T1/T2 electrodes
were strongly exploited to investigate the electrical activity emanating from these
regions in individuals suffering from these illnesses. Figure 3.1 shows the relative
positions of the recording tip of the sphenoidal electrode versus the surface
electrodes. The sphenoidal electrodes can also be placed under the guidance of
fluoroscope thus assuring both accurate placement and decreasing the chances for
nerve injury or vessel perforation (Fenton et al. 1997).

Supported Observations

(1) Many of the basal electrodes as well as the sphenoidal electrodes may detect
epileptic discharges when the standard scalp electrodes miss them.
(2) The sphenoidal electrodes and the surface-scalp electrodes (including the T1
and T2 electrodes) are complementary and should be used together.

Open Research Questions

The actual value of sphenoidal or other special electrodes in improving the yield in
detecting isolated epileptiform discharges (IDEs) in non-epileptic psychiatric
patients remains entirely unknown.

References

Ajmone Marsan C, Zivin LS (1970) Factors related to the occurrence of typical paroxysmal
abnormalities in the EEG records of epileptic patients. Epilepsia 11:361–382
American EEG Society Electrode Nomenclature Committee (1990) American electroencepha-
lographic society guidelines for standard electrode position nomenclature. J Clin Neurophys-
iol 8:200–202
References 25

Christopoulou G (1967) Sphenoidal electrodes. Acta Neurol Scand 43:587–593

Fenton DS, Geremia GK, Dowd AM, Papathanasiou MA, Greenlee WM, Huckman MS (1997)
Precise placement of sphenoidal electrodes via fluoroscopic guidance. Am J Neuroradiol
18(4):776–778
Jasper HH (1958) The ten-twenty electrode system of the international federation. Electroen-
cephalogr Clin Neurophysiol 10:371–375
Jones DP (1951) Recording of the basal electroencephalogram with sphenoidal electrodes.
Electroencephalogr Clin Neurophysiol 3:100
Marshall C (1957) Sphenoidal electrodes: an effort toward their popularization. Electroencep-
halogr Clin Neurophysiol 9:379–382
Nowack WJ, Janati A, Metzer WS (1988) The anterior temporal electrode in the EEg of the adult.
Clin Electroencephalogr 19:199–204
Rovit RL, Gloor P, Rasmussen T (1961) Sphenoidal electrodes in the electrographic study of
patients with temporal lobe epilepsy. J Neurosurg 18:1512–1518
Schomer DL (2003) The sphenoidal electrode: myth and reality. Controversies Epilepsy Behav
4:192–197
Silverman D (1965) The anterior temporal electrode and the ten-twenty system. Am J EEG
Technol 5:11–14
So EL, Ruggles KH, Ahmann PA, Trudeau P, Weatherford K (1994) Yield of sphenoidal
recording in sleep-deprived outpatients. J Clin Neurophysiol 11:226–230
Sperling MR, Mendiuys JR, Engel J Jr (1986) Mesial temporal spikes: a simultaneous comparison
of sphenoidal, nasopharyngeal and ear electrodes. Epilepsy 27:81–86
Chapter 4
Effects of Psychotropic Drugs on the EEG

Introduction

The field of pharmacoEEG is concerned with examining the interaction between

pharmacological agents believed to influence brain activity and brain electro-
physiological measures like EEG, magnetoencephalography (MEG), evoked
responses and sleep studies. As such this clinical research field is rather vast
spanning anesthesiology, neurology, and psychiatry. The field has two main fun-
damental concerns; the identification of potentially therapeutic compounds (work
done by and large by the pharmaceutical industry), and predicting the clinical
response in an individual patient. The field of psycho-pharmacoEEG is rapidly
expanding thanks to the rapidly advancing computer analysis capabilities of
quantification of the ever increasing size of the EEG recorded signals from a large
number of dense electrode array systems. Much lesser emphasis is now placed on
the visual inspection of the EEG which is the sole focus of this volume. The
interested reader is referred to the website of the International PharmacoEEG
group (IPEG) (www.ipeg-society.org) for guidelines and recent literature. A
number of books outline research and clinical applications of pharmacoEEG in
psychiatry. For a comprehensive review the reader is referred to Drinkenburg et al.
(2004).
The focus of this chapter is on the common clinical situation where a patient
who is suffering from a chronic psychiatric disorder (and at times multiple co-
morbid psychiatric and medical conditions), and is receiving a number of CNS
active medications begins to exhibit signs of clinical deterioration. In such con-
ditions, it is crucially important to determine the possible factors contributing to
the worsening clinical condition. Causes of the deterioration could be disease-
related (breakthrough of symptoms or relapse due to noncompliance), medication-
related (toxicity), or other general causes like a concomitant medical condition.
Given the demonstrated sensitivity of the EEG to the neurotoxic effects of CNS
active drugs, knowledge of the effects of drugs on the EEG in therapeutic and in
toxic situations, and in the different age groups, would be important to assist in
such determinations and would significantly inform the clinical management of an

N. N. Boutros, Standard EEG: A Research Roadmap for Neuropsychiatry, 27

DOI: 10.1007/978-3-319-04444-6_4,  Springer International Publishing Switzerland 2013
28 4 Effects of Psychotropic Drugs on the EEG

individual patient. In the current day environment of ever-increasing psycho-

pharmacology armamentarium, and frequently in the face of polypharmacy, it
becomes crucial to characterize the effects of single and multiple medications on
the EEG to differentiate between nontoxic and toxic effects and possible disease-
related changes. Other than animal and human pharmacoEEG studies, conducted
by industry much of which remain unpublished, effects of psychotropic medica-
tions to date have been either case reports or small sample size studies. As is seen
from the review below, large, prospective, and multicenter studies are yet to be
performed. It is important to reiterate that it is not the purpose of this chapter to
review all pharmacoEEG studies conducted on the very many psychotropic
medications if the effects on the standard EEG do not render the record abnormal.
Where the EEG becomes abnormal (whether slowing of the background rhythms
or the appearance of paroxysmal activity) these effects are discussed.

First Generation Psychotropic Agents

The EEG is rather sensitive to the neurotoxic effects of psychotropic medications.

EEG changes secondary to phenothiazine treatment were investigated in the late
1960s and early 1970s. Two types of abnormalities were identified; abnormalities
associated with toxic effects of the drugs and abnormalities that are secondary to
pre-existing brain abnormalities that were made more manifest electroencephal-
ographically and clinically under the influence of the phenothiazine (Dasberg and
Robinson 1969a). Patients who developed severe EEG changes showed deterio-
ration in their clinical course. Patients with less severe and reversible EEG changes
eventually improved. In a prospective study, the EEGs of 20 patients suffering
from an affective disorder were examined before and after being started on thio-
ridazine treatment. All subjects had pretreatment normal EEGs and had no history
of any neurological problems. Here again, clinical improvement was observed in
association with mild or reversible EEG changes while unchanging rigid EEG
were concomitant with clinical stagnation and were indicative of unfavorable
outcome (Dasberg and Robinson 1969b). Dasberg and Robinson (1971) further
investigated the correlates of EEG abnormalities associated with treatment with
antipsychotic agents. In this study they included 30 patients with no neurological
problems (subjected to a thorough history and neurological examination) who
were suffering from anxiety and aggressive behavior, mood disorders, or schizo-
phrenia. Unmedicated patients with abnormal EEGs were excluded. Following
medication initiation, EEGs were repeated weekly for 4 weeks then biweekly for
four more weeks. In the serial EEG recordings, five types of developments were
observed. First, a majority of subjects exhibited gradual slowing of the alpha
activity with the appearance of sporadic theta waves. These changes were
observed early and were maximal by the 4th week of treatment. Second, there was
a group of patients where no changes were observed. The author did not consider
this to be a normal response and favored the response listed to occur in the first
First Generation Psychotropic Agents 29

group. A third group tended to lie between the prior two groups. The fourth group
(uncommon) had similar but exaggerated changes similar to the first group.
Finally, there was also a small group of patient who exhibited paroxysmal activ-
ities on serial recordings. They concluded that the changes characterizing their first
group are most predictive of a positive therapeutic response.
Van Sweden and Dumon-Radermecker (1982) reported a number of clinical
vignettes where sEEG was useful in detecting a neurotoxic reaction to medications
while a clinical picture of deterioration was evident. They specifically stressed that
such a scenario could occur at any time during the course of treatment as many
factors impact the patient and that the symptoms could be subtle. They stressed the
usefulness of an sEEG investigation when patients who are on long-term therapy
present with clinical deterioration. This is particularly true if the patient is known
to be taking the medications and that serum plasma levels are within therapeutic
range. The cases presented highlight the need for having baseline EEGs available
for comparison when patient presents with a clinical exacerbation. EEG norms as
currently defined are based on cross sectional evaluations and do not take into
account the dynamic nature of psychiatric disorders or the constantly changing
medications status (Helmchen 1974).

Atypical Antipsychotic Medications

Centorrino et al. (2002) compared the effects of a number of medications from

different categories on the EEG. They included 323 hospitalized psychiatric
patients; 30 were not on antipsychotics. The EEGs were graded blindly to diag-
nosis, condition, or medication. They showed a rate of induced abnormalities of
47.1 % for clozapine, 38.55 for olanzapine, 28 % for risperidone, 14.5 % for
typical neuroleptics, and 0 % for quetiepine. They also identified a number of risk
factors for inducing EEG abnormalities, summarized in Table 4.1 showing the risk
factor in descending order.
Pittman et al. (2000) compared 43 EEGs of patients receiving 10–25 mg/day
olanzapine in routine treatment blindly with EEG registrations from the same 43
patients with a different medication. There was no difference in epileptiform
activity between the conditions with and without olanzapine. However, EEG
slowing was significantly more frequent with olanzapine than under the other
condition. This difference could not be attributed to concomitant medication.
More recently, Wichniak and collaborators (2006) examined the prognostic
value of EEG slowing in olanzapine-treated schizophrenia patients. They com-
pared 54 so-treated patients to 54 unmedicated and 38 with olanzapine plus other
antipsychotics. All EEGs were rated blindly by the same examiner. They noted no
differences between good and poor responders on frequency or nature of EEG
patterns. Olnazapine-treated patients had significantly more slowing (70 %),
and sharp waves (22 %). Additionally, 15 % had paroxysmal slow wave bursts.
Co-treatment with another antipsychotic increased EEG abnormalities while
30 4 Effects of Psychotropic Drugs on the EEG

Table 4.1 Risk factors of increasing chances for drug-induced EEG abnormalities in descending
order
Risk factor Effect Comments
High blood pressure Increase
Atypical antipsychotics Increase
Bipolar Increase
Older age Increase Possibly more prominent with clozapine
Benzodiazepine co-treatment Lowered risk
Gender No effect
Treatment response No effect
Length of hospital stay No effect
Drug potency No effect
Daily dose No effect Except for clozaril

co-treatment with benzodiazepine diminished the olanzapine-induced EEG

abnormalities. They concluded that the risk for seizures when treated with olan-
zapine alone is low but that this risk could increase if olanzapine is combined with
other antipsychotic agents (Wichniak et al. 2006).

Drug-Induced Paroxysmal EEG Activity

Kuglar et al. (1979) published a large retrospective study of the effects of psy-
chotropic agents on the EEG. They concluded that ‘‘there is no doubt that par-
oxysmal EEG activity can be induced by administration of psychotropic drugs’’.
They examined 680 EEGs from 593 patients. They reported that the highest
proportion of abnormal EEGs was in clozapine (see next chapter) patients (59 %)
followed by lithium salt (50 %). The overall proportion of paroxysmal discharge
was 13 %. Actual seizures were witnessed with treatment using clozapine, lithium,
and maprotiline. A 54 years old woman with probable Alzheimer’s Disease
developing continuous myoclonic jerks (myoclonic status) just after adding
olanzapine to her medications (low doses of citalopram and donepezil) (Camacho
et al. 2005). Jerks coincided with spike and polyspike-wave complexes on EEG.
The Seizure activity as well as EEG abnormality dissipated once olanzapine was
stopped.

Lithium

Lithium continues to be widely used in the treatment of bipolar disorder as well

other episodic behavioral syndromes and at times aggressive tendencies. Scattered
case reports of neurotoxicity with lithium can be found in the literature (Garcia-
Solana et al. 2004). Lithium therapy can result in a confusional state by direct toxic
Lithium 31

effects, precipitation of a nonconvulsive status epilepticus (NCSE), or by inter-

acting with other CNS active medications to produce neuroleptic malignant syn-
drome or a serotonin syndrome. Lithium level and clinical examinations may not
be sufficient to differentiate among these possibilities and a standard EEG could be
useful (Kaplan and Birbeck 2006 for a review of this topic). A case report
described such a case where a bipolar patient was started on a combination of
lithium and risperdone and became encephalopathic within few days. This par-
ticular case also highlight that care must be taken in medicating patients with
pretreatment abnormal EEGs (Boora et al. 2008). Case reports also suggest that
utmost care should be exercised when prescribing lithium to an elderly patient.
Suda et al. (2009) described a case of a 77 years old female patient who went into
coma 7 days after starting lithium. EEG showed periodic sharp waves despite
being on carbamazepine. Lithium intoxication was diagnosed. Blood level of
2.14 mEq/L was found.
Two case reports attest to the possibility that lithium therapy can result in a
NCSE even at therapeutic doses. Bellesi et al. (2006) a 52 years old woman with
bipolar disorder who while on a therapeutic dose of lithium with serum levels
within normal range, began to develop episodic confusional states. EEG revealed
diffuse spike discharges. Immediate clinical and EEG resolution were noted with
intravenous diazepam and discontinuation of lithium therapy. Upon re-challenge
with lithium 2 months later, the same syndrome redeveloped. NCSE was also
reported as consequence of lithium overdose and the EEG played a crucial role in
making the correct diagnosis (Yip and Yeung 2007).
Small et al. (1972) examined the effects of lithium administered to normal
volunteers and groups of psychiatric patients. Normal volunteers had pre-lithium
normal EEGs. They found that lithium caused slowing of the background activity
in both groups. Paroxysmal activity occured in healthy subjects more frequently
than in psychiatric patients. The incidence of toxic delirium was 10 % in both
patients and normal volunteer groups. Small et al. (1998, 1999) examined clinical
EEG and quantitative EEG in bipolar patients and reported increased diffuse theta
activity as a predictor of poor lithium response. Reeves et al. (2001) reported that
patients presenting with mania and EEG abnormality, particularly sharp activity,
were statistically more likely to respond to valproate than to lithium (2001). The
authors also investigated the relationship between lithium response and EEG
findings in 27 bipolar patients. In this study none of five lithium responders had
EEG abnormality, whereas all of five patients with EEG abnormality (three had
epileptiform abnormalities and two had continuous abnormalities) were lithium
nonresponders. These studies suggest that EEG abnormality is a possible predictor
of lithium resistance in bipolar disorder (Ikeda et al. 2002; Ikeda and Kato 2003).
In the meantime, Dalen and colleagues distinguished bipolar patients with
normal EEG as likely to be familial cases whereas those with EEG abnormalities
were secondary to some form of central nervous system insult (Dalen et al. 1965).
Although a negative study was reported (Taylor 1980), subsequent researchers
also confirmed this relationship between EEG and family history (Kadrmas and
Winokur 1979; Cook et al. 1986; Small et al. 1997).
32 4 Effects of Psychotropic Drugs on the EEG

Supported Observations

The bulk of the evidence supports a number of conclusions:

(1) The presence of an unvarying EEG pattern is an indication of lack of response
to the medications being administered.
(2) The presence of epileptiform discharges may be suggestive of a lowered
seizure threshold and should cause the clinicians to monitor treatment more
closely or be vigilant for evidence of aura like symptoms.
(3) Clinical deterioration despite compliance may be indicative of a neurotoxic
effect of the medications. A significantly abnormal EEG with moderate to
severe diffuse slowing and/or paroxysmal activity should cause the treaters to
reconsider the medication regimen.
(4) Lithium therapy can result in a confusional state through a variety of mech-
anisms and a standard EEG can be helpful in the differential diagnosis.

Open Research Questions

(1) How cost-effective is obtaining baseline EEGs?

(2) When a patient is on lithium, what combinations of signs, symptoms, and
blood level should prompt obtaining an sEEG.
(3) What is the frequency of significant EEG changes (changes that should prompt
medication change) with lithium therapy? What are the effects of combina-
tions of medications, age, gender, and other co-morbid medical conditions?
(4) What are the nature and degree of EEG changes that should prompt medi-
cation changes? Then, how often should EEGs be repeated in follow-up?
(5) What are the relationships between EEG changes and blood levels of the
different psychotropic compounds?
(6) How do standard and quantified EEGs complement each other regarding
medicating psychiatric patients?

References

Bellesi M, Passamonti L, Silvestrini M, Bartolini M, Provinciali L (2006) Non-convlusive status

epilepticus during lithium treatment at therapeutic doses. Neurol Sci 26(6):444–446
Boora K, Xu J, Hyatt J (2008) Encephalopathy with combined lithium—risperidone adminis-
tration. Acta Psychiatr Scand 117(5):394–395
Camacho A, Garcia-Navvaro M, Marinez B (2005) Olanzapine induced myoclonic status. Clin
Neuropharmacol 28:145–147
Centorrino F, Price BH, Tuttle M, Bahk WM et al (2002) EEG abnormalities during treatment
with typical and atypical antipsychotics. Am J Psychiatry 159:109–115
References 33

Cook BL, Shukla S, Hoff A (1986) EEG abnormalities in bipolar affective disorder. J Affect
Disord 11:147–149
Dalen P (1965) Family history, the electroencephalogram and perinatal factors in manic
conditions. Acta Psychiatr Scand 41:527–563
Dasberg H, Robinson S (1969a) Isr Ann Psychiatry 7:185–200
Dasberg H, Robinson S (1969b) Isr Ann Psychiatry 7:201–212
Dasberg H, Robinson S (1971) Electroencephalographic variations following anti-psychotic drug
treatment (diagnostic and prognostic significance) Dis Nerv Sys July; 472–478
Drinkinburg WHIM, Ruigt GSF, Jobert M (eds) (2004) Essentials and applications of EEG
research in preclinical and clinical pharmacology. IPEG, Berlin
Garcia-Solana MI, Rodrigo-Sesma A, Garcia Rodriguez V et al (2004) Rev Neurol
39(6):595–597
Helmchen H (1974) Significance of psychotropic drug-induced abnormal EEGs. In: Turan Itil
(Ed) Psychotropic drugs and human EEG Mod. Probl. Pharmacopsychiat, vol 8.Krager, New
York, pp 317–326
Ikeda A, Kato D (2003) Biological predictors of lithium response in bipolar disorder. Psychiatry
Clin Neurosci 57:243–250
Ikeda A, Kato N, Kato T (2002) Possible relationship between electroencephalogram finding and
lithium response in bipolar disorder. Prog Neuropsychopharmacol Biol Psychiatry
26:903–907
Kadrmas A, Winokur G (1979) Manic depressive illness and EEG abnormalities. J Clin
Psychiatry 40:306–307
Kaplan PW, Birbeck G (2006) Lithium induced confusianal states: non-convulsive status
epilepticus or triphasic encephalopathy. Epilepsia 47(12):2071–2074
Kugler J, Lorenzi E, Spatz R, Zimmermann H (1979) Dug-induced paroxysmal EEG-activities.
Pharmakopsychiat 12:165–172
Pittmann F, Schlote K, Broich K, Marneros A (2000) Electroencephalogram alterations during
treatment with olanzapine. Psychopharmacology 150(2):216–219
Reeves RR, Struve FA, Patrick G (2001) Does EEG predict response to valproate versus lithium
patients with mania? Ann Clin Psychiatry 13:69–73
Small JG, Milstein V, Perez HC, Small IF, Moore DF (1972) EEG and neurophysiological studies
of lithium in normal volunteers. Biol Psychiat 5:65–77
Small JG, Milstein V, Medlock CE (1997) Clinical EEG findings in mania. Clin Electroencep-
halogr 28:229–235
Small JG, Milstein V, Malloy FW, Klapper MH, Golay SJ, Medlock CE (1998) Topographic
EEG studies of mania. Clin Electroencephalogr 29:59–66
Small JG, Milstein V, Malloy FW, Medlock CE, Klapper MH (1999) Clinical and quantitative
EEG studies of mania. J Affect Disord 53:217–224
Suda M, Kubota F, Aihara Y, Hiraoka T, Aoyama Y, Hattori S, Fukoda M, Mikuni M (2009) A
case of lithium intoxication with periodic sharp waves. Pharmacopsychiatry 42:122–123
Taylor MA, Abrams R (1980) Familial and non-familial mania. J Affect Disord 2:111–118
Van Sweden B, Dumon-Radermecker M (1982) The EEG in chronic psychotropic drug
intoxications. Clin EEG 13:206–215
Wichniak A, Szafranski T, Wierzbicka A et al (2006) Electroencephalogram slowing, sleepiness
and treatment response in patients with schizophrenia during olanzapine treatment.
J Psychopharmacology 20:80–85
Yip KK, Yeung WT (2007) Lithium overdose causing non-convulsive status epilepticus; the
importance of lithium level and the electroencephalogram in diagnosis. Hong Kong Med J
13:471–474
Chapter 5
The Special Case of Clozapine

Introduction

Clozapine is an atypical neuroleptic indicated for use in treatment-resistant

schizophrenia patients. In addition to its established effectiveness as an antipsy-
chotic agent clozapine possesses a profile of side effects that differ from other
typical or atypical neuroleptics. As many neuroleptics were shown to have effects
on the EEG, clozapine was similarly expected to affect the EEG. However, there
were no predictions made prior to the release of the drug in 1972 in Austria, to
wide clinical use regarding the significance or degree of effects. Even when
released in 1989 in the USA, a full understanding of the epileptogenic nature of the
drug was not fully recognized.

Early Observations

As early as the late 1970s, the effects of clozapine on the EEG were reported
(Koukkou et al. 1979). EEG effects and psychopathological changes during
treatment with haloperidol and clozapine were compared. Thirty-nine acute
schizophrenia patients were included; of them 20 patients received clozapine and
19 received haloperidol (Koukkou et al. 1979). Dosages were adjusted according
to patient’s clinical needs. EEG and clinical data were collected on days 0, 3, 10,
20, and 30. Paroxysmal EEG patterns during a 20 min recording were counted
blindly on six scalp locations. On days 3, 10, and 20 more clozapine patients
showed paroxysmal EEG patterns than those on haloperidol (P \ 0.05). Clozapine
patients with paroxysmal EEG patterns had received a lower clozapine dose on day
10 than those without paroxysm (P \ 0.02). Also clozapine patients with EEG
paroxysms showed a greater reduction of retarded and depressive symptoms on
day 20. Furthermore, the investigators found a negative correlation between the
number of paroxysms and intensity of depressive symptoms as well as a positive
correlation between the number of paroxysms and reduction of depressive

N. N. Boutros, Standard EEG: A Research Roadmap for Neuropsychiatry, 35

DOI: 10.1007/978-3-319-04444-6_5,  Springer International Publishing Switzerland 2013
36 5 The Special Case of Clozapine

symptoms with time. These results demonstrated a relation between clozapine

induced paroxysmal EEG patterns and its psychotherapeutic effects reminiscent of
a similar relation seen in patients with epileptic psychosis called forced normali-
zation. In clinical situations where forced normalization is seen, when the par-
oxysmal discharges of epileptic patients are strongly suppressed using
anticonvulsant medications, psychotic symptoms begin to emerge and get sup-
pressed again only when paroxysmal EEG activity is allowed to reappear (Trimble
and Schmitz 1998). The observation (see below) that the presence of paroxysmal
discharges may correlate with clinical improvement (particularly of mood symp-
toms) suggests that clozapine by generating epileptic activity may be functioning
as mini-ECT.
One of the serious side effects associated with clozapine is the higher risk of
dosage-dependent EEG abnormalities (Leppig et al. 1989) and seizures (Haller
1990). With dosage levels less than 300 mg/day the risk is 1–2 %, with medium
dosages (between 300 and 600 mg/day) the risk rises to 3–4 %. At higher doses
the risk seems to maximize at about 5 %.
Lu (1991), reported the side effects of clozapine from the largest sample to date
(Lu 1991; only abstract available in English). He reported a rate of 9 % of EEG
abnormalities, which is significantly lower than in any other report to date. He also
reported a 2 % rate of seizures from a sample of 7,921 patients. His reported rates
of abnormalities are rather low and without the full report it is hard to decide on
the accuracy of the data or how they relate to other reports. Indeed a lifetime
cumulative risk of seizures may be up to 10 % (Devinsky et al. 1991).
A number of risk factors have been identified. These risk factors include history
of seizures, history of head injury accompanied by loss of consciousness, presence
of abnormalities on the EEG, and history of recent ECT. The concomitant use of
drugs that are known to lower seizure threshold can similarly increase the risk for
seizures in clozapine-treated patients (Devinsky 1991). Around the same time,
Braun-Scharm and Martinius (1991) proposed that when the daily dose of cloza-
pine exceeds 300 mg, that the EEG should be monitored regularly. A rapid
titration toward a higher dose seems to be an additional risk factor for developing
seizures (Baker 1991). An important open question is what would be an optimal
monitoring schedule? Later on, Pacia and Devinsky (1994) reported that seizures
occur at low dosages during the titration phase and in association with high doses
during the maintenance phase.
Three cases of patients treated with clozapine and developed myoclonic sei-
zures were reported by Gouzoulis et al. (1991). All three schizophrenia patients
had no history of seizures prior to treatment. They developed paroxysmal EEG
patterns and generalized myoclonic jerks without alteration of consciousness.
These seizures were phenomenologically identical to those occurring in juvenile
myoclonic epilepsy. Again in 1995 a German group reported the occurrence of
myoclonic seizures and suggested that evoked potential investigations may be a
useful tool to assess such patients for the risk for seizures (Brogmus and Lesch
1995; report in German with only abstract available in English). In 1992, a report
appeared describing prolonged post-ictal encephalopathy in two clozapine treated
Early Observations 37

patients (Karper et al. 1992). This side effect was also associated with high doses
of clozapine.
The group in Munich reported on a relatively large group of clozapine-treated
patients with repeated EEGs (Gunther et al. 1993). They had 1,863 EEGs recorded
over time from 283 patients. While 61.5 % of the patients showed some abnor-
mality at some time, 22.5 % were considered minimal abnormalities with likely
little or no clinical significance. Significant slowing was seen in close to 40 % and
sharp waves in about 16 % (abnormalities can be combined). More importantly,
they reported that a nearly linear correlation was found in the range up to 300 mg/
day for diffuse slowing. Rate of abnormalities decreased slightly in doses between
300 and 600 mg then sharply rose in doses above 600 mg. They reported a rate of
actual seizures of only 1.1 %.
Welch et al. (1994) published their experience at New Hampshire Hospital.
They reported that seven of the first 35 patients treated with clozapine experienced
seizures (20 %). Seizures were dose-related. EEG changes were frequent with
clozapine, particularly as dosage increased. Of 35 patients, 26 (74 %) had EEG
abnormalities at some time during clozapine treatment. EEG is a sensitive means
of detecting drug clinical toxicity.
The concept of clinical toxicity is important to be elaborated upon. A drug may
cause clinical symptoms consistent with toxicity before reaching serum toxic
levels. This is most commonly due to factors that are either specific for the
individual or due to general factors that increases the sensitivity of the brain to the
toxic effects of drugs. Any factor that in itself can lead to encephalopathy (e.g.,
hepatic or renal problems), or factors known to accelerate a dementing process like
multiple head injuries or heavy alcohol and drug use, may all contribute to the
increased sensitivity to the toxic effects of drugs on the CNS. Welch et al. con-
cluded that when EEG abnormalities (slowing, dysrhythmia, or paroxysmal dis-
charges) are detected, immediately lowering the dose by at least 25–50 mg/day
and adjusting weekly until EEG returns to baseline can reduce the incidence of
seizures. Devinsky and Pacia (1994) suggested that more research on the value of
the EEG in predicting seizures as well the value in predicting clinical response is
needed. This suggestion remains as valid today as when it was made in 1994
attesting to the slow rate of research in this field.
In 1994, a report from the National Institute of Neurological Disorders and
Stroke (NINDS) appeared and deserves some commentary (Malow et al. 1994).
While only 10 subjects were included the report is significant. All EEGs were
evaluated by a blinded EEGer. All EEGs pre-clozapine were normal. All EEGs
showed increased slow-wave activity when on clozapine (mainly theta but some
also exhibited delta activity). Of the 10 subjects, five developed myoclonus (50 %)
and one a generalized tonic-clinic seizure (10 %). Of the 10 patients, seven had
spikes, polyspikes, or paroxysmal slow wave bursts. One patient developed a
photoparoxysmal response. All abnormalities decreased or disappeared following
dose reduction and/or the addition of the anti-epileptic drug (AED) valproic acid.
Sajatovic and Meltzer (1996) showed that patients on clozapine can develop
myoclonus in the absence of seizures or seizures in the absence of myoclonus.
38 5 The Special Case of Clozapine

As it is possible that the presence of paroxysmal activity to a certain degree may

facilitate the therapeutic response, much research is then needed regarding the
optimal combination of clozapine, the AED agent, and the EEG changes. More-
over, as not all AEDs have the same mechanisms of action, it is possible that
different agents may necessitate different combinations of the three factors for
optimal clinical response. Thus, a long and not so easy line of research remains
open and awaits dedicated psychiatric electrophysiologists to help provide some
answers to very real and important questions.
A group from Austria further investigated the relationship between serum clo-
zapine levels and effect on EEG (Haring et al. 1994). Twenty-nine inpatients were
included, all with pre-clozapine normal EEGs and were solely treated with clozapine.
Of the 29 patients 15 showed EEG changes. Discriminant analysis showed that EEG
changes are dependent on plasma level (p = 0.0009). The group of patients with
EEG changes had a plasma level of 235 ± 170 ng/ml while the group without EEG
abnormalities had levels of 82 ± 65 ng/ml. They found age, sex, weight, and
duration of treatment not to be significant factors. Olesen et al. (1995) found the only
side effect to correlate with serum level was EEG changes. Freudenreich et al. (1997)
also showed that clozapine-induced EEG changes are related to clozapine serum
levels in 50 chronic schizophrenia patients. Fifty-Three percent showed EEG
changes. Three patients had seizures (6 %). The one patient with no prior history of
seizures was on 900 mg and had a serum level of 320 ng/ml. The other two patients
had serum levels between 200 and 300 ng/ml, both had histories of prior seizure
disorders. They also found that a clozapine serum level between 350 and 450 ng/ml
led to more frequent and more severe slowing. The EEG slowing correlated with
observed sleepiness.
The epileptogenic tendencies of clozapine may also be paralleled by increased
cortical responsivity in these patients. In Jin et al. (1995), reported that clozapine
responders had a significantly greater increase in photic driving primarily in the
low alpha range (7.2, 8.3, 9, 9.6 Hz) but not higher frequencies. The increase in
photic driving was positively correlated with patient’s clinical improvement. The
difference in the resting EEG between responders and nonresponders did not reach
significance. Both paroxysmal activity and increased photic driving could be seen
as evidence of increased cortical excitability. It is possible then that increased
cortical excitability (which can be assessed by evoked potentials and transcranial
magnetic stimulation/TMS) would be predictive of favorable response to
clozapine.
Of major interest is the report by Risby et al. (1995) where in a group of eight
patients who developed EEG abnormalities on clozapine, clinical improvement
was evident while in a group of eight who did not exhibit EEG changes, no clinical
improvement was seen. This report raises a serious and important question
regarding the interrelationship between clinical effects and EEG effects, and the
possible value of the observations versus risk for seizures. In a similar vein, Pillay
et al. (1995) in a group of 86 patients treated with clozapine showed that whether
the EEG was abnormal or normal pre-clozapine did not predict response. How-
ever, when the group was divided by gender, females with abnormal pre-clozapine
Early Observations 39

EEGs improved more than those with pre-clozapine normal records. Similarly,
affective patients (Bipolar or Schizoaffective) with pre-clozapine abnormal EEGs
benefited more from clozapine therapy.
When dose was maintained at 300 mg and EEGs examined, a large % of EEG
abnormalities were still detected, but with no report of seizures (Treves and
Neufeld 1996). Of 11 patients completing the study, one exhibited epileptic
activity and diffuse slowing and five only showed diffuse slowing. EEG abnor-
malities were observed more frequently (albeit not-statistically significant) in those
who responded better to treatment.
These observations of EEG abnormalities predicting good response were not
shared by a report in neurological patients treated with clozapine (Duffy and Kant
1996). In a group of 16 patients with some neurological impairment, 25 %
developed confusional states, all had diffuse slowing on their pre-clozapine EEGs.
When 20 Parkinson’s Disease patients (who became psychotic on dopamine
agonist treatments) were given low dose clozapine, all improved clinically, only
five developed EEG slowing, and none had seizures (Neufeld et al. 1996). It is thus
of importance to establish the pre-clozapine EEG degree of deviation. The nature
and degree of pre-clozapine EEG abnormalities that predict poor response to
clozapine can thus be better determined.
Silvestri et al. (1997) examined the EEGs of 12 patients referred for an EEG
because of seizures (N = 8) or to assess for risk for seizures (N = 4) on clozapine
therapy. Six of the eight patients with seizures and two of the four without seizures
had epileptiform activity. One patient had a photo-convulsive response. A domi-
nance of left temporal focal abnormalities was observed.

More Recent Reports

Table 5.1 lists the risk factors for EEG abnormalities with clozapine treatment. In
Schuld et al. (2000), from Munich compared the effects of clozapine and olnaz-
apine on the EEG. While a small sample size (Nine subjects in each group), they
reported that clozapine induced significant EEG slowing in 78 % of patients and
definite epileptiform activity in 33 %. Olanzapine on the other hand produced less
pronounced EEG slowing and only in 44 %. Olanzapine had no effects on epi-
leptiform activity although in one subject an isolated sharp and slow-wave com-
plex was detected. They concluded that the effects of olanzapine on seizure
threshold is definitely significantly less than those of clozapine, but it deserved
further investigation for better characterization. Later on the same year, Pittman
and colleagues from Martin Luther University in Halle Germany, reported on the
EEGs of a larger group of olanzapine-treated schizophrenia patients (N = 43)
(Pittman et al. 2000). They reported very similar findings, with no epileptogenicity
but significant effects of diffuse EEG slowing.
Utilizing quantification of the EEG into the different frequency bands,
Joutsiniemi et al. (2001) showed convincingly that the slowing is not seen with
40 5 The Special Case of Clozapine

Table 5.1 Risk factors for Factor Effect

EEG abnormalities with
clozapine treatment Rapid upward titration Increase
Higher serum levels and higher dosages Increase
History of seizures Increase
History of factors that lower seizure Increase
threshold like head injury or stroke
Combined with other antipsychotic drugs Increase
Recent ECT Increase
Abnormal EEG Increase
Valproic acid treatment Decrease
Benzodiazepines Decrease

typical neuroleptics and that while diffuse, it is maximal on the frontal, central, and
parietal electrodes. They concluded that this EEG profile is specific for clozapine.
It should be noted that in this study, the clozapine spectral-EEG profile was not
compared directly to other atypicals also known to cause diffuse slowing like
olanzapine.
Whether co-medication of clozapine and an anticonvulsant should be used
prophylactically, be based on EEG abnormalities, or only if seizures do occur
remains an open question. Conca and colleagues showed that if such a course of
action is taken, close monitoring of blood levels becomes necessary particularly if
valproic acid is used (Conca et al. 2000). It should be noted that Wilson (1995)
provided data suggesting that the use of anticonvulsants may in fact hinder the
therapeutic effects of clozapine. The nature of this interaction (between clozapine
and anti-seizures agents) is in need of further and more controlled research.
Chung et al. (2002) examined the EEG effects of clozapine on 50 Korean
schizophrenia patients. All patients had normal baseline EEGs. Only two of the 50
(4 % had actual clinical seizures). Thirty-one (62 %) developed EEG abnormal-
ities with the majority of abnormalities comprised of diffuse EEG slowing, while
spikes (or spike and wave complexes) were rare. Their data suggested that the
probability of EEG abnormality is linearly dependent on the daily dose of clo-
zapine and patient’s age. They also concluded that in the majority of patients
exhibiting epileptic activity, seizures did not occur. They none the less did not
offer suggestions regarding what to do if such activity is detected or when spe-
cifically to obtain an EEG in a patient who is being treated with clozapine.
The relationship between clozapine level, spectral-EEG parameters, and per-
formance on vigilance and memory tasks was examined in a group of 17 chronic
schizophrenia patients being maintained on clozapine (Adler et al. 2002). There
was a negative correlation between clozapine serum level and the amount of high
frequency (beta) activity. A positive correlation was found between high frequency
EEG activity and memory performance. They concluded that clozapine treatment
brings about dose-dependent impairments of vigilance and memory. They also
concluded that the observed clozapine-induced reduction of high EEG frequencies
may be indicative of this effect on memory and vigilance.
More Recent Reports 41

Whether the slowing is indicative of unwanted side effect or reflects changes

correlating with efficacy is also an open question. Gross et al. (2003) suggested a
correlation between the increased EEG slowing and a decrease in the production of
reactive oxygen species (ROS) by peripheral blood monocytes which correlated
with clinical improvement. Their admittedly speculative suggestion that clozapine-
induced EEG slowing is a result of the modulatory action of the activated mi-
croglial cells in the CNS via production of the ROS or cytokines or both, is
intriguing and is deserving of further research.
As reviewed above, epileptiform EEG changes, myoclonus, and seizures are
reported in some patients treated with clozapine. Although these are undesirable
side effects, the excitation of specific neuronal networks by clozapine and other
neuroleptics may be important for the therapeutic effect of this class of agents.
Denney and Stevens (1995) worked on elucidating this relationship. In these
experiments, intraperitoneal clozapine 2–16 mg/kg produced dose-related myo-
clonic jerks in partially restrained rats. Paroxysmal slow waves and spike activity
were recorded from implanted electrodes in amygdala, hippocampus, and cortex
following higher doses of clozapine, but the EEG abnormalities were not corre-
lated with the myoclonic jerks. Myoclonus and seizures reflect increased excit-
ability of the central nervous system. It is possible that clozapine and other
neuroleptics exert a therapeutic effect by increasing excitability in critical sub-
cortical areas of the brain. As mentioned repeatedly in this volume, noninvasive
methodology to assess hyperexcitability in humans is rapidly advancing and
constitutes new avenues for investigating this relationship.

Supported Observations

(1) EEG abnormalities are common with clozapine treatment.

(2) EEG abnormalities under clozapine treatment are likely dose-dependent.
(3) Abnormalities tend to predict the occurrence of seizures but some level of
abnormalities may in fact indicate better prognosis or are associated with
better response.

Open Research Questions

(1) The exact role the sEEG should play in managing patients being treated with
clozapine is not known. Recommendations vary from very frequent moni-
toring to no monitoring.
(2) Is the nature of the EEG abnormality detected (slowing versus paroxysmal)
important for predicting response or side effects. The bulk of the literature
suggest that the presence of epileptiform discharges predicts a favorable
42 5 The Special Case of Clozapine

clinical response but also precedes the development of seizures. Hence, it is

very possible to speculate that the effect is a matter of how much paroxysmal
activity is detected. It is clear that while the rate of paroxysmal discharges may
be in the 20–30 %, the rate of seizures is only in the 1–6 %. Defining the
degree of abnormality that is consistent with a good response but not sug-
gestive of increased likelihood of seizures would be of interest.
(3) Elucidating the exact value for detecting and the degree of diffuse slowing
(highly prevalent) remains an open line of investigation.

References

Adler G, Grieshaber S, Faude V et al (2002) Clozapine in patients with chronic schizophrenia:

serum level EEG and memory performance. Pharmacopsychiatry 35:190–194
Baker R, Conly R (1991) Seizures during clozapine therapy. Am J Psychiatry 148:1265–1266
Braun-Scharm H, Martinius J (1991) EEG changes and seizures with clozapine medicationin
schizophrenic adolescents (German) Zeitschrift fur Kinder- und Jugendpsychiatrische
19(3):164–169
Brogmus KE, Lesch A (1995) Psychotropic drug-induced myoclonus. Psychiatr Prax 22(2):77–79
Chung SJ, Jeong SH, Alm YM et al (2002) A retrospective study of clozapine and electrographic
abnormalities in schizophrenia patients. Prog Neuropharmacol Biol Psychiatry 26(1):139–144
Conca A, Beraus W, Konig P, Waschgler R (2000) A case of pharmacokinetic interference in
comedication of clozapine and valproic acid. Pharmacopsychiatry 33(6):234–235
Denney D, Stevens JR (1995) Clozapine and seizures. Biol Psychiatry 37(7):427–433
Devinsky O, Honigfeld G, Patin J (1991) Clozapine related seizures. Neurology 41:369–371
Devinsky O, Pacia SV (1994) Seizures during clozapine therapy J Clin Psychiatry 55(9, suppl
B):153–156
Duffy JD, Kant R (1996) Clinical utility of clozapine in 16 patients with neurological disease.
J Neuropsychiatry Clin Neurosci 8(1):92–96
Freudenreich O, Weiner RD, McEvoy JP (1997) Clozapine-induced EEG changes as a function of
clozapine serum levels. Biol Psychiatry 42:132–137
Gouzouis E, Grunze H, von Bardleben U (1991) Myoclonioc epileptic seizures during clozapine
treatment: a report of three cases. Eur Arch Psychiatry Clin Neurosci 240(6):370–372
Gross A, Joffe G, Joutsiniemi SL et al (2003) Decreased production of reactive oxygen species by
blood monocytes caused by clozapine correlates with EEG slowing in schizophrenia patients.
Neuropsychobiology 47:73–77
Gunther W, Baghal T, Naber D et al (1993) EEG alterations and seizures during treatment with
clozapine. A retrospective study of 283 patients. Pharmacopsychiatry 26(3):69–74
Haller E, Binder RI (1990) Clozapine and seizures. Am J Psychiatry 147:1069–1071
Haring C, Neudorfer C, Schwitzer J et al (1994) EEG alterations in Patients treated with
clozapine in relation to plasma levels. Psychopharmacol 114(1):97–100
Jin Y, potkin SG, Sandman C (1995) Clozapine increases EEG photic driving in clinical
responders. Schizophr Bull 21(2):263–268
Joutsiniemt SL, Gross A, Appelberg B (2001) Marked clozapine-induced slowing of EEG
background over frontal, central, and parietal scalp areas in schizophrenic patients. J Clin
Neurophysiol 18:9–13
Karper LP, Salloway SP, Sibul JP, Krystal JH (1992) Prolonged postictal encephalopathy in two
patients with clozapine-induced seizures. J Neuropsychiatry Clin Neurosci 4(4):454–457
References 43

Koukkou M, Angst J, Zimmer D (1979) Paroxysmal EEG activity and psychopathology during
the treatment with clozapine. Pharmacopsychiatry 12:173–183
Leppig M, Bosch B, Naber D, Hippius H (1989) Clozapine in the treatment of 121 out-patients
Psychopharmacology 99(Suppl:S):77–9
Lu MK (1991) Clinical analysis of the main side effects of clozapine: inclosed 600 case reports
(Chinese). Chin J Neurol Psychiatry 24(2):71–74, 132
Malow B, Reese KB, Sato S et al (1994) Spectrum of EEG abnormalities during clozapine
treatment. Electroencephalogr Clin Neurophysiol 91(3):205–211
Neufeld MY, Rabey JM, Orlov E, Korczyn AD (1996) Electroencephalographic findings with
low-dose clozapine treatment in psychotic parkinsonian patients. Neuropharmacol
19(1):81–86
Olesen OV, Thpmsen K, Jensen PN et al (1995) Clozapine serum levels and side effects during
steady state treatment of schizophrenic patients: a cross sectional study. Psychopharmacology
117(3):371–378
Pacia SV, Devinsky O (1994) Clozapine-related seizures. Neurology 44:2247–2249
Pillay SS, Stoll AL, Weiss MK et al (1995) EEG abnormalities before clozapine therapy predict a
good clinical response to clozapine. Ann Clin Psychiatry 8(1):1–5
Pittmann F, Schlote K, Broich K, Marneros A (2000) Electroencephalogram alterations during
treatment with olanzapine. Psychopharmacology 150(2):216–219
Risby ED, Epstein CM, jewart RD et al (1995) Clozapine-induced EEG abnormalities and
clinical response to clozapine J Neuropsychiatry Clin Neurosci 7(4):466–470
Sajatovic M, Meltzer HY (1996) Clozapine induced myoclonus and generalized seizures. Biol
Psychiatry 39(5):367–370
Schuld A, Kuhn M, Haack M, Kraus T et al (2000) A comparison of the effects of clozapine and
olanzapine on the EEG in payients with schizophrenia. Pharmacopsychiatry 33(3):109–111
Silvestri RC, Bromfield EB, Khoshbin S (1997) Clozapine-induced seizures and EEG
abnormalities in ambulatory psychiatric patients. Ann Pharmacother 32(11):1147–1151
Treves IA, Neufeld MY (1996) EEG abnormalities in clozapine-treated schizophrenic patients.
Eur Neuropsychopharmacol 6(2):93–94
Trimble MR, Schmitz B (eds) (1998) Forced normalization and alternative psychoses of epilepsy.
Wrightson Biomedical Publishing, Bristol
Welch J, Manschreck T, Redmond D (1994) Clozapine-induced seizures and EEG chanages.
J Neuropsychiatry Clin Neurosci 6:250–256
Wilson WH (1995) Do anticonvulsants hinder clozapine treatment? Biol Psychiatry 37:132–133
Chapter 6
Slowing of the EEG in Psychiatric Patients

Introduction

Diffuse slowing constitutes 40 % of all electroencephalographic abnormalities in

psychiatric patients and thus is the most common abnormality detected the EEGs
obtained from psychiatric patients (Struve 1994). Correlations have been dem-
onstrated between the degree of slowing of the EEG and impairment of functions
such as awareness, attention, memory, and comprehension (Struve and Boutros
2005). As of the writing of this chapter, there are hardly any systematic and well-
controlled studies examining the clinical correlates of diffuse EEG slowing in the
various psychiatric populations.
Diffuse slowing of the EEG is considered as one of the hallmarks of the
presence of an encephalopathic process (Wise and Brandt 1992). This abnormality
is frequently attributed to the effects of psychotropic medications (Matsuura et al.
1994). Accepting this explanation without further investigation could lead to
missing cases of medication-induced encephalopathy or preexisting organic
cerebral pathology (Fink 1969). While slower activity (less than 8 Hz) can be
detected in the computer-quantified EEG (q-EEG) even after a single dose of a
psychotropic drug, the routine analog EEG remains normal, to visual inspection,
with much higher dosages of these medications. This phenomenon is supported by
the fact that the majority of psychiatric patients (most of whom are on psycho-
tropic medications) have normal analog EEGs (Struve 1994). The amount of slow-
wave activity has to increase substantially before its presence can be detected by
the naked eye. When the slow-wave activity is barely but consistently visible to the
naked eye, the record is usually interpreted as mildly or minimally abnormal. The
presence of much slower rhythms (less than 4 Hz) is usually necessary to consider
the EEG moderately or severely abnormal. These very slow rhythms (delta) are
rather easily detectable by visual inspection.
Engel and Romano (1959) emphasized the importance of the EEG in the
diagnosis of delirium and in the detection of minor degrees of impairment of
consciousness. A close correlation has been demonstrated between the degree of
slowing of the EEG and the degree of disturbance of consciousness as reflected in

N. N. Boutros, Standard EEG: A Research Roadmap for Neuropsychiatry, 45

DOI: 10.1007/978-3-319-04444-6_6,  Springer International Publishing Switzerland 2013
46 6 Slowing of the EEG in Psychiatric Patients

impairment of functions such as awareness, attention, memory, and comprehension

(Engel and Romano 1959). Johnson et al. (1970) showed that, following chronic
lithium administration, the presence and severity of EEG changes are highly
correlated with neurotoxicity. Dasberg and Robinson (1971) however, provided
data that suggest that the presence of diffuse or paroxysmal slowing in the EEG
following antipsychotic drug treatment may be indicative of a higher likelihood of
good clinical response. No recent studies have attempted to further explore the
clinical correlates of diffuse EEG (in the standard EEG) slowing in well-defined
groups of psychiatric patients.
It should be mentioned that focal slowing on the EEG usually indicates a focal
structural problem and should be considered indicative of a neurological focal
process and given the necessary attention until the cause of the focal finding is
known. There are no psychiatric causes for a focal slow-wave abnormality
detectable by the naked eye.
One immediate observation on this body of literature is the fact that discussion
of the different types of diffuse slowing is minimal if at all. There are two major
ways the sEEG can be determined to be diffusely slow. The background rhythm
can slow down. In this case the background rhythm, (which is supposed to be
composed mainly of alpha activity during relaxed wakefulness), is replaced by
slower rhythms (in psychiatric situations usually theta activity). Alternately, a
diffusely seen slower rhythm is superimposed on a relatively normal background
activity.
We now discuss in some detail a study that was conducted at the Ohio State
University in order to determine whether diffuse EEG slowing in hospitalized
psychiatric patients correlates with any of the hospitalization indices (length of
stay, diagnosis, and type and amount of psychotropic medications) (Boutros 1996).
Twenty five consecutive patients with diffusely slow EEGs were selected for
this study. The abnormal records were collected over a period of 1 year. The total
number of EEGs recorded during the year was 180. Twenty five control patients
were matched to the diffusely slow EEG patients on one-to-one bases. Once a slow
record was identified, the next age and gender matched patient with a normal EEG
was included in the control group. All subjects were hospitalized adult patients in a
state hospital with a 220 bed capacity. EEGs were recorded on an 11 channel Grass
machine utilizing a bandpass filter setting of 0.3–70 Hz. Twenty one electrodes
were applied according to the 10–20 international electrode placement system. All
records were at least 30 min in length and included photic stimulation and
hyperventilation. A number of standard montages (both referential and bipolar)
were used for all patients. One certified electroencephalographer interpreted all
records blinded of diagnosis, hospitalization, or medication status. Age and gender
of subjects were not blinded. Only records with diffuse slowing as the sole EEG
abnormality were included in the study. Records were interpreted as diffusely slow
when, while the patient was demonstrably awake (i.e., interacting with the EEG
technologist), the tracing was dominated by frequencies slower than alpha activity
(i.e.,\8 Hz). In addition, records with paroxysmal bursts of diffusely slow activity
superimposed on a normal background were also excluded (we believe this is a
Introduction 47

different abnormality that deserve it is own study). An equal number of patients on

clozapine were included in the two groups as clozapine is known to induce diffuse
EEG slowing (Welch et al. 1994) (see also Chap. 5 for more details). Finally,
patients with medical diagnoses who are likely to have contributed to the observed
slowing of the EEG were also excluded. These criteria were established a priori
and included any neurological, endocrine, or metabolic diagnoses appearing on the
patient’s problem list, or identified in progress notes.
Medical records were reviewed for the indications for which the EEGs were
obtained, the patients’ diagnoses, and the length of stay of each patient at time of
EEG recording. Presence or absence of history of drug abuse and personality
disorders was also noted. Urine screens for illicit drugs were done only if recent
drug use was suspected. The number of medications, the nature of the medications,
as well as the dosages, were recorded. Neuroleptic dosages were transformed into
chlorpromazine equivalents in milligrams. All diagnoses were made by the ward
physicians according to DSM-III or DSM-III-R criteria. Multiple pair-wise t tests
were performed comparing the different study variables between the two groups.
There was a highly statistically significant difference in length of stay between
the groups, with a mean of 278 days (ranging from 4 to 16 months) for the normal
EEG group and 754 days (ranging from 5 to 18 months) for the slow EEG group
(t = 3.20, df = 24, P \ 0.04, 2-tailed paired I tests). Only one patient in each
group was in the hospital for more than 10 years at the time of EEG study. With
exclusion of these two patients, the mean length of stay of the patients dropped to
48.2 days in the normal EEG group and 525.4 days in the slow EEG group. The
difference remained highly statistically significant at P \ 0.006. Two of the
patients with normal EEGs and five with slow EEGs were in the hospital for more
than 12 months at the time the charts were reviewed.
All patients in the normal EEG group were on psychotropic medications with
the exception of two patients. The number of medications per patient ranged from
1 to 7 (mean 2.8), with a total of 69 different psychotropic medications for the
group. In the slow EEG group, there were no records in which the slowing was
considered marked or severe. One patient in the slow EEG group was not receiving
medications at the time of the recording. In this group, the number of medications
per patient ranged from 1 to 8 (mean 3.8), with a total of 94 psychotropic medi-
cations for the group. This difference is statistically significant (2-tailed t test
P \ 0.05). Only two patients with diffuse EEG slowing were not on neuroleptic
medications as compared with six in the normal EEG group. In the slow EEG
group, the mean chlorpromazine dose was 512.2 mg. A mean of 334.8 mg per
patient was found in the normal EEG group. This difference was found to be
significant (F = 7.87, P \ 0.03) based on the one-way analysis of variance
(ANOVA). Additionally, 18 patients in the slow EEG group were on anticon-
vulsants as compared with II in the normal EEG group (df = 1, X2 = 4.023,
P \ 0.045). Lithium also tended to be more prevalent in the slow EEG group (10
patients vs. four in the normal EEG group) (df = 1, X2 = 3.571, P \ 0.06). There
were no differences between the groups in the number or dosages of antidepres-
sant, anticholinergic, or benzodiazepine medications. There were also no
48 6 Slowing of the EEG in Psychiatric Patients

differences in the nonpsychotropic medications (for example, vitamins, antibiotics,

and analgesics).
Correlational analysis was performed to study the relationship among the three-
dependent variables found to be significantly different between the two groups
(that is, length of stay, number of medications, and neuroleptic dosage). As
expected, a Pearson correlation coefficient of 0.43 (P \ 0.006) was found between
the number of medications and the neuroleptic dosage. No significant correlation
was found between the length of stay and either the number of medications
(P \ 0.9) or the neuroleptic dose (P \ 0.2). There were no significant differences
between the two groups in the diagnoses either on Axis I or Axis II. Drug and
alcohol use tend to be more prevalent in the normal EEG group (36 % vs. 12 % in
the slow EEG group, X1 = 3.95, P \ 0.07). Four of the patients in the slow EEG
group received their EEGs during their last week of hospitalization. None of the
four was suspected of drug use, so no drug screens were done. Of the 18 patients in
the normal EEG group who received their EEGs during their first week of hos-
pitalization, five were suspected of drug use. All five urine screens, however, were
negative for illicit drugs. The pattern of ordering EEGs also differed between the
groups. In the normal EEG group, 15 of the 25 records (60 %) were ordered for
work-up of a history of seizures or head injury, while most of the EEGs in the slow
EEG group were ordered for either no apparent reason (routine screening) or
because patients had vague neurological complaints (66 %).
The remarkably longer length of stay and the increased number of medications
and dosages used in the slow-EEG group may be a reflection of increased illness
severity in these patients. The possibility remains that the higher number and
dosages of medications, particularly neuroleptics, may have contributed to the
observed slowing in these patients. This possibility is supported by the observation
of Mellario (1964) that EEG tracings tended to be slower when several neuro-
leptics were administered simultaneously. The lack of correlation between the
neuroleptic dose, the number of medications, and the length of stay, however,
suggests that slowing of the EEGs cannot be completely explained by the nature or
dose of the medications received.
It should be noted that, although patients were admitted to the groups pro-
spectively as their EEGs were interpreted, no other measures of encephalopathy,
standardized rating scales, or follow-up EEGs were performed. Finally, routine
urine drug screens were not performed, so the possibility that drug withdrawal or
intoxication may have contributed to the findings cannot be confidently ruled out.
Nonetheless if intoxication or withdrawal contributed to the observed slowing then
the effects should have been observed more in the shorter stay and not in the longer
stay patients as observed. These short comings decreased the generalizability of
the findings. Larger studies employing multi measures of encephalopathy, clinical
rating scales, serum plasma medications levels, and sequential EEGs are necessary
to delineate the differences between drug-induced EEG changes that are not
clinically significant, and those that are indicative of either drug toxicity or an
independent encephalopathic process.
Supported Observations 49

Supported Observations

Diffuse slowing of the EEG is the most common EEG deviation reported in
psychiatric populations.

Open Research Questions

(1) Are there different types of diffuse EEG slowing and what are the clinical
correlates of each?
(2) How reversible are these changes?

References

Boutros N (1996) Diffuse electroencephalogram slowing in psychiatric patients: a preliminary

report. J Psychiatry Neurosci 21(4):259–263
Dasberg H, Robinson S (1971) Electroencephalographic variations following antipsychotic drug
treatment: diagnostic and prognostic significance. Dis Nerv Syst 32(7):472–8
Engel GL, Romano J (1959) Delirium, a syndrome of cerebral insufficiency. J chronic Dis
9:260–277
Fink M (1969) EEG and human psychopharmacology. Annu Rev Pharmacol 9:241–258
Johnson G, Maccario M, Gershon S, Korein J (1970) The effects of lithium on encephalogram
behavior and serum electrolytes. J Nerv Ment Dis 151:273–289
Matsuura M, Yoshino M, Ohta K, Onda H, Nakajima K, Kojima T (1994) Clinical significance of
diffuse delta EEG activity in chronic schizophrenia. Clin Electroencephalogr 25:115–121
Mellario F (1964) Electroencephalograph ie dans les intoxications aigues. Masson, Paris
Struve FA (1994) Selective referral versus routine screening in clinical EEG assessment of
psychiatric inpatients. Psychiatric Med 4:317–343
Struve FA, Boutros NN (2005) Somatic implications of generalized and/or focal slowing in
psychiatric patients. Clin EEG Neurosci 36(3):171–175
Welch J, Manschreck T, Redmong D (1994) Clozapine induced seizures and EEG changes.
J Neuropsychiatry Clin Neurosci 6:250–256
Wise MG, Brandt GT (1992) Delirium. In: Yudofsky SC, Hales RE (eds) The american
psychiatric press textbook of neuropsychiatry, 2nd edn. American Psychiatric Press,
Wahington, pp 291–293
Chapter 7
Isolated Epileptiform Discharges
in Nonepileptic Psychiatric Patients

Introduction

Although the existing EEG literature is replete with reports of abnormalities in

association with neuropsychiatric disorders, only a few generalizations can be
made between particular EEG patterns and disorders. The strong (and relatively
straight forward) correlation between EEG abnormalities and epilepsy has over-
shadowed the more complex relationship between EEG abnormalities and psy-
chiatric disorders. Moreover, the prevailing concept of ‘‘not treating the EEG’’ led
to further de-emphasizing such EEG deviations. A well-established fact is that the
incidence of EEG abnormalities; both slow wave abnormalities and epileptiform
discharges (EDs) are higher in psychiatric populations as compared to control
groups (Shelley et al. 2008). The implications and understandings of the interictal
ED have continued to evolve with research in the genetic and metabolic bases of
these events (Noebels 2003), intracranial electrophysiology, and the relationship of
EDs to neuropsychological functions (Aarts et al. 1984). Accumulating data, thus,
have begun to shift the older concept of the interictal EDs away from that of
electrophysiologic events that are unassociated with cognitive or behavioral con-
sequences to events that are the phenotypic expression of a variety of cellular
disorders (Fisch 2003).
The term Isolated epileptiform discharges (IEDs) in psychiatric patients refers
to the presence of epileptic discharges in nonepileptic individuals. IEDs are pos-
sibly not the same exact phenomenon as the interictal spikes (IIS) of individuals
suffering from an epileptic disorder. While no research available that compares the
biology, phenomenology, and clinical correlates of the two types of epilepstic
discharges one can assume that one type (IIS) is associated with seizures and thus
closely with epilepsy while the other is not. The interested reader is referred for a
recent review by Shelley et al. (2008). Bridgers (1987) confirmed the frequent
occurrence of IEDs in a population of nonepileptic hospitalized psychiatric
patients. The EEG findings were found to correlate with conditions such as
anorexia nervosa, depression, mania, personality disorders, suicidality without
depression, nonpsychotic explosive behavior, and effects of psychotropic

N. N. Boutros, Standard EEG: A Research Roadmap for Neuropsychiatry, 51

DOI: 10.1007/978-3-319-04444-6_7,  Springer International Publishing Switzerland 2013
52 7 Isolated Epileptiform Discharges in Nonepileptic Psychiatric Patients

medications. The epileptiform EEG abnormalities were documented in 2.6 %, and

consisted of focal temporal complexes, generalized spike-wave or polyspike-wave
discharges, and focal central/frontal complexes. This study did emphasize that
IEDs do occur in nonepileptic psychiatric populations and may reflect underlying
cerebral dysfunction without necessarily indicating a significant increase in the
liability for seizures. Notwithstanding the above generalization, the literature
points to the prevalence of IEDs to be highest in a number of conditions, most
prominently panic attacks and violence/aggression. Among childhood psychiatric
disorders Autism Spectrum Disorders (ASD), Tourette Syndrome, and Attention
Deficit/Hyperactivity disorders are prominent. This differential prevalence of IEDs
among psychiatric disorders strongly suggests that the increased IEDs in these
conditions are not a nonspecific finding (Rötig et al. 2005).
Hughes reported that IEDs were four times more common in panic patients than
in depressed patients (Hughes 1996). Shelley et al. (2008) indicated that about
25–30 % of panic attack patients had demonstrable EEG abnormalities, especially
so in atypical presentations of panic attacks. Some studies have documented an
epileptic pathophysiology to underlie atypical panic attacks (Weilburg et al. 1995).
Williams compared the EEGs of 206 habitual aggressors and 127 who committed
isolated acts of violence (Williams 1969). He reported a five-fold increase in EEG
abnormalities in habitual aggressors; 57 % as compared to 21 % in non-habitual
aggressors. Howard showed that patients who have committed violent offences
against strangers, as opposed to people known to them, tended to have bilateral
paroxysmal EEG features (Howard 1984). Seventy percent of subjects with
bilateral paroxysmal discharges have attacked strangers.
A critical issue is the ability of the scalp EEG to detect the isolated epileptic
discharge. Current guidelines for an outpatient EEG recording duration is 20 min
(AES 1994) but in fact in clinical practice the duration varies widely from as short
as 15 min to as long as 120 min (Cascino et al. 1996). Of major importance is the
need to attain and record stages one and two sleep when the purpose of the test is
to identify EDs. Sleep deprivation has been shown to increase the yield for
interictal epileptiform discharges over and above assuring the recording of EEG
during sleep (Leach et al. 2006). The over whelming EEG studies in psychiatric
populations where IEDS sought are usually comprised of a single recording with
lack of IEDs simply equated with absence of IEDs. In fact, in bonafide epileptic
patients only 29–56 % will have an IED on a single recording of a 30-min duration
(Goodin and Aminoff 1984). With repeating the EEGs up to six times, the yield
can reach 82 % (Marsan and Zivin 1970).
Two questions immediately emerge based on the above observations. First, how
important is the identification of an IED in a psychiatric patient? While the simple
discovery of an IED in a patient who has never had a clinical seizure cannot and
should not be grounds for diagnosing epilepsy, it is none the less an important
factor that should be taken into account when formulating the biopsychosocial
formulation of an individual case. Whether or not the presence of IEDs predicts
positive response of psychiatric symptoms to antiepileptic drugs (AEDs) is a wide
open research question. Available literature pertinent to this important clinical
Introduction 53

question is reviewed in several chapters. It should be highlighted that the wide use
of antiepileptic agents in treating psychiatric disorders is currently largely based
on the belief that these agents posses psychotropic capabilities. It should also be
emphasized here that while a diagnosis of epilepsy is stigmatizing in nonpsychi-
atric populations, it may be destigmatizing in psychiatric populations.
A second important issue to be raised is the current technology available to detect
IEDs. The only tangible advance has been the increased duration of monitoring:
either via ambulatory 24 h EEGs or in video-EEG monitoring in Epilepsy Moni-
toring Units. Magnetoencephalography (MEG) remains prohibitively expensive for
use in psychiatric practices and MEG equipment are indeed rare and only exist in
major research medical centers. Improving technology so smaller IEDs at the cortex
as well as IEDs originating from deeper cerebral locations (i.e., most of the cerebral
structures of vital importance for psychiatric disorders like the limbic or medial and
ventral frontal sources) can be detected at the scalp has not happened. It is my belief
that in the absence of pressure applied from the clinical world, such efforts will not
be forthcoming. A recent paper by Losey and Uber-Zak is worth highlighting (Losey
and Uber-Zak 2008). They examined the time to the first IED in extended recordings
from patients suspected of having epilepsy. They report that 47 % of all IEDs
occurred later than the 20-min duration currently officially recommended. At least
11 % of their records did not yield the IED till after 40 min. This work is bases for
recommending that the minimal recording duration in psychiatric patients, when
IEDs are suspected, should be no shorter than 60 min. In case of a negative study, a
repeat following an all night sleep deprivation is recommended with a longer
duration of 2–3 h (Losey and Uber-Zak 2008).

IEDs and the Tip of the Iceberg Concept

Epilepsy is defined as having had two unprovoked seizures. It is likely that a

certain amount of epileptic activity would be necessary for spontaneous seizures to
occur. Hence, it is possible that many individuals who may harbor EDs may never
have a clinically definable seizure. Given the pathological nature of such dis-
charges, it is highly unlikely that IEDs (i.e., spikes in the absence of seizures) are
simply incidental findings. It is much more likely that the presence of these dis-
charges signals abnormal brain activity.
Documentation of EDs from the temporal lobe is frequently difficult to dem-
onstrate. A classic study showed that when repeated EEGs were done over a period
of 1 year, temporal lobe epileptiform discharges could be seen in 98 % of patients
with documented temporal lobe epilepsy (Marsan and Zivin 1970). Similar effort
targeting psychiatric conditions was never attempted. Perhaps, the most obvious
difficulty is the lack of some form of gold standard to measure against. Given the
evidence that isolated epileptiform discharges (IEDs) in nonepileptic psychiatric
patients are not uncommon, and given the well-documented high false negative
rate for scalp EEG detection of epileptic discharges, it is possible that
54 7 Isolated Epileptiform Discharges in Nonepileptic Psychiatric Patients

Fig. 7.1 Tip of the iceberg. The tip of the iceberg represents panic patients with manifest
epilepsy and nonepileptic patients who exhibit epileptic discharges on the scalp recorded EEG.
The next larger (submerged) part of the iceberg represents patients who may have epileptic
activity that could only be detected by more extensive EEG monitoring, MEG, or invasive
recordings. The base of the iceberg represents patients who have brain tissue hyperexcitability
that is not enough to generate epileptic activity but enough to result in psychiatric symptoms

investigations of epileptic activity in psychiatric conditions only detects the tip of

an iceberg formed by the various degrees of brain tissue hyperexcitability as
shown in Fig. 7.1.

Are There Demonstrable Effects of IEDs?

In a study of patients who had repeated EEGs over many years, Hughes demon-
strated the ability of spike foci to create other brain foci (usually mirror images on
the contralateral side of the brain (Hughes 1985)). This work clearly underlines the
non-benign nature of EDs. An effect of EDs on perception and reaction times was
also clearly demonstrated by Shewmon and Erwin (Shewmon and Erwin 1988a).
Capitalizing on the capacity to automatically detect spike discharges they were
able to demonstrate that EDs temporarily disrupt cortical functioning. This effect
was anatomically specific (Shewmon and Erwin 1988b). Even when children were
carefully screened (and excluded) for neurological and psychiatric histories,
approximately 4 % still exhibited EDs. Half of these children proved to have
behavioral disturbances and/or psychomotor aberrations on further scrutiny
(Cavazzuti et al. 1980). Neurophysiological and functional neuroimaging evidence
suggests that IEDs may impact cognition through either transient effects on brain
processing mechanisms, or through more long-lasting effects leading to prolonged
Introduction 55

inhibition of brain areas distant from but connected with the epileptic focus (i.e.,
remote inhibitory effects) (Van Bogaert et al. 2011). This area of investigation,
particularly in individuals with IEDs (or less frequent EDs) remains extremely
limited. Spike detection programs designed to detect rare or low amplitude EDs
are virtually nonexistent.

Etiology of IEDs in Nonepileptic Individuals

The overwhelming majority of studies probing the origin of epilepsy focuses on

seizures. There is a severe paucity of studies attempting to understand the evo-
lution of IEDs in nonepileptic individuals. In this section, we offer some ideas
from the existing literature. Early childhood stress and lifetime assaultive violence
have been linked to cortical mal-development and increased electrophysiological
abnormalities. Several studies have reported that severe early stress and abuse
have the potential to alter brain development and cause limbic dysfunction during
specific sensitive periods of cortical maturation (Teicher et al. 2003). The cascade
of events is mediated through stress-induced neurohormones of the glucocorticoid,
noradrenergic, and vasopressin-oxytocin stress response systems which affect
neurogenesis, synaptic overproduction and pruning, and myelination. The aberrant
cortical development has been reported to involve the corpus callosum, left neo-
cortex, hippocampus, and amygdala (Teicher et al. 2004). During the last decade
studies have reported an emergence of EEG abnormalities in children with sexual
and psychological abuse. An increased prevalence of frontotemporal electro-
physiological abnormalities had been reported in abused children; with the
abnormalities tending to have a left-sided localization (Ito et al. 1993, 1998;
Teicher et al. 1997). These studies describe the neurobiological mechanisms
through which early abuse increases the risk of developing various psychopa-
thologies and its electrophysiological consequences. Head injury outside the
context of abuse is also common in psychiatric populations although the exact
prevalence in specific disorders is not known (Fleminger 2008).
IEDs in pediatric neurobehavioral disorders may represent an epiphenomenon
of cerebral dysfunction or underlying cortical morphofunctional abnormalities,
and/or reflect a brain neurophysiological disorder which is not sufficient to be
expressed as epilepsy. This may be due to the lack of properly functioning cor-
ticocortical fibers which restricts the spread of epileptiform activity from one brain
area to another and prevent its evolution to a clinical seizure. The subclinical
epileptogenesis in the developing maturing brain may also directly impair cog-
nitive-behavioral functioning by way of ‘‘transient cognitive impairment’’ mech-
anisms, well described by Binnie and his colleagues (Aarts et al. 1984; Binnie
et al. 1987; Martson et al. 1993).
56 7 Isolated Epileptiform Discharges in Nonepileptic Psychiatric Patients

Necessity for an Animal Model for IEDs

All currently available models of epilepsy emphasize acute seizures more than the
interictal discharges. As we have suggested above the IEDs occurring in the
complete absence of seizures, which is the condition of most psychiatric patients
exhibiting epileptic discharges, may be significantly different from the interictal
discharges occurring in the interval between seizures in epileptic patients and thus
are not suitable to simulate the psychiatric condition. Because of our findings that
the interictal discharges seem to be the driving force for many of the gene
expression, signaling, and synaptic changes in human epileptic cortex, we have
developed a model to study IEDs in more detail (Barkmeier et al. 2012). While this
model is clearly useful for studying epileptogenesis, it can also be used to study the
direct effects of IEDs on behavior. The model under development by the Loeb
group at Wayne State University (WSU) offers a unique opportunity to examine
the effects of IEDs on behavior in the absence of confounding effects of seizures.
A tetanus toxin model described by Jefferys and others appears to work best in the
somatosensory cortex to produce a minimally damaging lesion with predominantly
IEDs (Nilsen et al. 2005; Benke and Swann 2004; Jefferys 1996).
One of the major challenges in developing novel therapeutics for psychiatric
disorders is the need for animal models that faithfully replicate the human con-
dition. While there are a large number of models that have been used to develop
anti seizure drugs, at present, we have no effective antiepileptic drugs that block
the development of IEDs, IIS, or epilepsy. This may not be surprising since most
models used to screen for the drugs currently in use today consist of acute seizure
models. However, a number of studies show a dissociation between acute seizures
and the development of epilepsy, which is by definition a condition of chronic,
recurrent seizures that is often associated with IIS from the same focus (Herman
2002; Beghi 2003). In almost all patients, IIS is in fact far more frequent than
seizures; and, in our studies, most of the persistent molecular changes we found in
human neocortex correlate best with interictal rather than ictal activity (Rakhade
et al. 2007). In order to be useful in studying the interrelationship of IEDs and
behavioral changes an animal model of focal, neocortical epilepsy that satisfies the
need for a gradual development of IEDs while avoiding subsequent seizures is
necessary. Furthermore, research has shown that not all IIS are the same. In fact,
Keller et al. (2010) provided evidence that possibly there are upwards of 15
different categories of patterns of interictal discharges. They suggested that this
heterogeneity in single unit activity likely reflects the fact that IIS in epileptic
patients is not simple paroxysms of hypersynchronous excitatory activity, but
rather represent an interplay of multiple distinct neuronal types within a complex
neuronal network. This proposition is very likely applicable to IEDs in nonepi-
leptic psychiatric patients as well. The animal model described in this section or
other animal models of IEDs could facilitate the probing of this phenomenon.
Introduction 57

Fig. 7.2 Characterization of a focal model of neocortical epilepsy in the rat. a Injection of
tetanus toxin into the left somatosensory cortex is followed by the placement of six skull-based
screw electrodes at the indicated positions. b Within 1 week, small interictal discharges can be
detected with an expected electrical field centered over the injections site. c The rate of increase
of spike frequency can be optimized by the number of injections sites and the dose, with 100 ng
in a single site being optimal

Behavioral and Cognitive Testing Procedures in Rodents

A large variety of procedures have evolved over the years that allow testing of
behavior and cognition in laboratory rodents. These procedures vary in complexity
from open field observation (procedure chosen for current model) to complex
procedures like the rotarod test, water maze, and elevated plus maze. Such pro-
cedures were recently critically reviewed (Stafstrom 2006).
The goal of this model is to produce an infrequent form of epileptic discharges
that develops weeks after an initial insult that will give us both focal epileptic and
‘‘control’’ brain regions from the same animal, as well as additional controls from
sham-treated animals. To do this, the epilepsy group at Wayne State University
injected an irritant into the somatosensory cortex of rats as shown in Fig. 7.2.
While a variety of focal irritants are capable of producing epileptic discharges, we
have compared several of these and found that a tetanus toxin model appears to
work best in the somatosensory cortex to produce a minimally damaging lesion
58 7 Isolated Epileptiform Discharges in Nonepileptic Psychiatric Patients

with predominantly IIS, and rare seizures (Nilsen et al. 2005; Benke and Swann
2004; Jefferys 1996).
Tetanus toxin produces a transient inhibition of inhibitory neurons that is
known to lead to the generation of epileptic discharges that over time can produce
epileptic seizures. Within 1 week after injection, focal epileptic discharges
develop and increase in frequency over time. Eventually, symptomatic seizures are
detected in four to six weeks at which time the experiment is terminated. Thus far,
our preliminary results in our epilepsy work have shown that this model replicates
the same changes in molecular and synaptic reorganization found in human neo-
cortical epileptic foci.

Brief Description of Behavioral Observation Procedure

Open field activity (OFA) was recorded simultaneously with recording of electro-
physiology data. Each rat was placed in a clear plexiglass arena (17 9 17 9 12
inch). Infrared beam transmitters and receivers line the sides of the arena. In all, 16
infrared beam units are evenly spaced along the length and another 16 beams were
spaced along the width of the arena. OFA data collected from the arrays are sent to
the activity monitor data acquisition card and collection software (OFA Activity
Monitor, Med-Associates Inc., St. Albans, VT). Behavioral data are stored for offline
analysis. During each 1 h recording session movements are recorded over the time
course of the session for later analysis.

ANOVAs of Six Types of Behavior for Total Averages

of Each Subject (Vehicle and Treatment: N 5 12
and N 5 12)

Calculation of activity in the Activity Monitor data analysis module proceeds

offline, after data for an entire session are collected. Six behavioral measurements
were derived for further analysis: ambulatory activity, ambulatory movement
velocity, circular-rotational movement activity (clockwise and counter clockwise),
movement among four quadrants of the arena, and stereotypic activity. Ambula-
tory activity is interpreted as any activity that produces three or more sequential
beam breaks until a period of inactivity occurs whereas no beams are broken
(minimum 0.5 s inactivity period). Average velocity measures were derived from
the distance covered divided by time during the ambulatory episode. Rotational
activity during ambulatory episodes is defined by a curved path with small
(maximum 2.75 cm radius) clockwise or counter clockwise movements in the
arena. Stereotypy includes any movements that produces a beam break but does
Introduction 59

Fig. 7.3 The path (dark line) of animal position in the arena indicates ambulatory activity over
the course of the session. Movements are detected by beam breaks of the infrared sensor array
(light lines)

not qualify as ambulatory or rotational activity. Zone entry is calculated as the

number of times that the animal moves from one quadrant into another.
OFA data over the course of the session were analyzed according to the entire
session for several days prior to the onset of seizures. An example of a single
session for subjects from each condition (Fig. 7.3) demonstrates the changes in
subjects’ position in the arena over the course of the session for the second to last
day of the first week. The treated animal appears to cover more territory in the
arena during the session.
For each behavior type, averages over all sessions were generated for each
subject. Univariate ANOVAs were conducted for the six types of behavior. OFA
activity during the first week after surgery revealed several trends, but there were
no significant differences between control and activity groups. Univariate ANOVA
of the clockwise revolution for each subject revealed a significant effect of tetanus
toxin treatment F(1, 8) = 6.36, p \ 0.05. Treated animals made significantly more
clockwise revolutions than vehicle treated animals in the control condition.
ANOVA of the counterclockwise revolution did not show significant differences
between the groups F(1, 8) = 0.93, p = 0.36. The excess of clockwise rotational
activity for treated animals (Fig. 7.3) might reveal an underlying imbalance
between left and right hemispheres in the treated animals. Spiking activity
occurred in the left hemisphere in treated animals, and increased clockwise activity
could indicate an increase in motion in the direction of the right hemifield of body
space which is processed by the left hemisphere somatosensory cortex. An alter-
native explanation is that the IIS in the left hemisphere produced some motor
weakness on the right side of the body making strides on the right shorter and
hence leading to clockwise rotation. This is an unlikely explanation because such
weakness was not clinically apparent.
60 7 Isolated Epileptiform Discharges in Nonepileptic Psychiatric Patients

There were no differences between groups for average velocity F(1, 8) = 2.3,
p = 0.17, ambulatory activity F(1, 8) = 2.87, p = 0.13, stereotypy F(1, 8) = 1.5,
p = 0.26, and zone entries F(1, 8) = 0.44, p = 0.53. Overall, except for coun-
terclockwise revolution, treated animals had higher levels or rates of activity in
each of the behaviors measured. As more subjects are tested it will be important to
ascertain whether this pattern of increased activity is replicated.

Effect of Environmental Stimuli on IEDs

While it is well known that certain genetic strains of rodents have seizures in
response to environmental stimuli, such as loud noises (Faingold 1999), little is
known on the effects of similar environmental stimuli on IIS or IEDS. We have
begun to investigate these effects on IEDs in our unilateral spiking rats. We
observed that loud sounds evoke IEDs emanating from the region where tetanus
toxin is injected into somatosensory cortex in our rats. This is not seen in controls
or rats injected with saline. An example of these audiogenic-induced IEDS is
shown in Fig. 7.4. To our knowledge, this is the first evidence that environmental
stimuli can induce IIS or IEDs. Since IIS or IEDs themselves may produce
important behavioral abnormalities, the role of the environmental stimuli inducing
them could be a critical part of patient management.
The above preliminary data provide strong evidence that an animal model of
IEDs in the complete or almost complete absence of seizures can be developed,
and that such spikes not only can influence behavior but also seem to be sensitive
to environmental stimuli. Further exploration of this animal model particularly
with placing the IEDs in more behaviorally salient regions (e.g., hippocampus,
cingulate, or amygdale) could prove useful to this field including screening for
drugs that can modulate IEDs and hence modulate behavior.

Supported Findings

(1) At present the exact relationship between IEDs and psychiatric symptom-
atology in nonepileptic individuals is not known. Of course, if intervention at
an early stage can prevent a life-long disorder (e.g., in ASD children), even a
low yield may be well justified not only on economic bases but also for the
suffering of patients and their families that may be avoided.
(2) IEDs are significantly increased, as compared to healthy control and non-
neurological patient control groups, in some psychiatric patient groups.
(3) The increased incidence of IEDs in some psychiatric groups is specific to these
groups and is likely to indicate certain subgroups within this particular patient
population.
Introduction 61

Fig. 7.4 Audiogenic-induced interictal spiking. Rats in which spontaneous spikes were
generated from unilateral tetanus toxin injection into the left somatosensory cortex generated
focal spikes over the injections site with a field that spreads to involve the left frontal and left
central electrodes (left panel). In these same animals, but not controls, spikes with the same field
could be induced by a loud noise that startled the animal (right panel). Arrows indicate time
points when loud sounds were generated

Open Research Questions

(1) Does the fact that an epileptic discharge was not detected necessary mean that
there is no such activity going on ‘‘tip of the iceberg’’? One MEG study clearly
documented that MEG may be more sensitive in detecting such abnormalities
(Lewine 1999).
(2) Does inability to detect an abnormality in a single EEG recording means that
subsequent recordings will not reveal new findings. It is of interest that there
are no available studies of nonepileptic psychiatric patients with repeat or
prolonged EEG monitoring in order to ascertain the eventual yield from EEG
studies (as has been done for epilepsy patients). Can ambulatory EEG be of
use here and what would be the cost-effective duration and clinical situation of
its use (Schomer 2006)?
(3) It is also interesting that there are no studies where ASD children without
seizures or initial abnormal EEGs were followed-up with repeat EEGs to
ascertain the relationship between the occurrence of seizures and preseizures
EEG abnormalities.
(4) Can the technology for automated spike detection be further improved to allow
detecting IEDs and thus allowing the collection of longer samples of EEGs
(24 h ambulatory EEG for example). Furthermore, can the technology be
further improved to better detect smaller amplitude IEDs or IEDs emanating
from deeper brain sources particularly limbic and medial or orbital frontal
regions?
(5) Why is there a demonstrably higher prevalence of EDs in pediatric psychiatric
disorders (e.g., ASD, attention deficit/hyperactivity disorder, Tourette
62 7 Isolated Epileptiform Discharges in Nonepileptic Psychiatric Patients

Syndrome) as compared to adult psychiatric patients is a question in search for

answers. The currently accepted theory (without much evidence) is that the
brain heals with maturation. Verrotti et al. (1999) showed that from 40 non-
epileptic children who were referred for an EEG because of various com-
plaints (tics, ADHD, headaches, vertigo) and exhibited centrotemporal spikes
(CTS), 33 had no epileptic discharges when the EEGs were repeated few years
later. Simple explanations like skull thickness difference between childhood
and adult populations are rather easy to investigate. If skull thickness does
play a role in decreasing the power of detection of EDs then the need for
advancing technologies to deal with this obstacle becomes urgent.
(6) What are the values of repeated EEG testing over time in improving the
detection rates of IEDS in psychiatric conditions.
(7) Perhaps the most important unanswered question relative to IEDs is whether
their presence in an individual presenting with psychiatric symptoms (panic
attacks, repeated violence, autism) would predict a favorable response to
antiepileptic medications? Systematized and well-controlled studies are rare
(Van Bogaert et al. 2011). The few available studies are discussed in the
respective chapters.

References

Aarts JHP, Binnie CD, Smit AM, Wilkins AJ (1984) Selective cognitive impairment during focal
and generalized epileptiform EEG activity. Brain 107:293–308
American Elelctroencephalographic Society (1994) Guideline one: minimum technicaql
requirements for performing clinical electroencephalography. J Clin Neurophysiol 11:2–5
Barkmeier DT, Senador D, Leclercq K, Pai D, Hua J, Boutros NN, Kaminski RM, Loeb JA
(2012) Electrical, Molecular and Behavioral Effects of Interictal Spiking in the Rat. Neurobiol
Dis 47(1):92–101
Beghi E (2003) Overview of studies to prevent posttraumatic epilepsy. Epilepsia ee (Supplement
10):21–26
Benke TA, Swann J (2004) The tetanus toxin model of chronic epilepsy. Adv Exp Med Biol
548:226–238
Binnie CD, Kastelejin-Nolst Trenit_e DG, Smit AM, Wilkins AJ (1987) Interaction of epileptiform
discharges and cognition. Epilepsia Res 1:239–245
Bridgers SI (1987) Epileptiform abnormalities discovered in electroencephalographic screening
of psychiatric inpatients. Arch Neurol 44:312–316
Cascino GD, Trennerry MR, So EL (1996) Routine EEG and temporal lobe epilepsy: relationship
to long-term monitoring, quantitative MRI, and operative outcome. Epilepsia 37:651–656
Cavazzuti GB, Cappella L, Nalin A (1980) Longitudinal study of epileptiform EEG patterns in
normal children. Epilepsia 21:43–55
Fleminger S (2008) Long-term psychiatric disorders after traumatic brain injury. Eur J
Anaesthesiol Suppl 42:123–130
Faingold CL (1999) Neuronal networks in the genetically epilepsy-prone rat. Adv Neurol
79:311–321
Fisch BJ (2003) Interictal epileptiform activity: diagnostic and behavioral implications; 2002
ACNS presidential address. J Clin Neurophysiol 20:155–162
References 63

Goodin DS, Aminoff MJ (1984) Does the interictal EEG have a role in the diagnosis of epilepsy?
Lancet 8381:837–839
Herman ST (2002) Epilepsy after brain insult: targeting epileptogenesis. Neurology 59(Supple-
ment 5):S21–S26
Howard RC (1984) The clinical EEG and personality in mentally abnormal offenders. Psychol Med
14:569–580
Hughes JR (1985) Long-term clinical and EEG changes in patients with epilepsy. Arch Neurol
42:213–223
Hughes JR (1996) A review of the usefulness of the standard EEG in psychiatry. Clin
Electroencephalogr 27:35–39
Ito Y, Teicher MH, Glod CA, Harper D, Magnus E, Gelbard HA (1993) Increased prevalence of
electrophysiological abnormalities in children with psychological, physical, and sexual abuse.
J Neuropsych Clin Neurosci 5:401–408
Ito Y, Teicher MH, Glod CA, Ackerman E (1998) Preliminary evidence for aberrant cortical
development in abused children: A quantitative EEG study. J Neuropsych Clin Neurosci
10:298–307
Jefferys JG (1996) Chronic epileptic foci induced by intracranial tetanus toxin. Epilepsy Res
Suppl 12:111–117
Keller CJ, Truccolo W, Gale JT, Eskandar E, Thesen T, Carlson C, Devinsky O, Kuvinsky R,
Doyle WK, Madsen JR, Schomer DL, Mehta AD, Brown EN, Hochberg LR, Ulbert I, Halgren
E, Cash SS (2010) Heterogeneous neuronal firing patterns during interictal epileptiform
discharges in the human cortex. Brain 133:1668–1681
Leach JP, Stephan LJ, Slaveta C, Brodie MJ (2006) Which electroencephalogram (EEG) for
epilepsy? The relative usefulness of different EEG protocolsin patients with possible epilepsy.
J Neurol Neurosurg Psychiatry 77:1040–1042
Lewine JD (1999) Magnetoencephalographic patterns of epileptiform activity in children with
regressive autism spectrum disorders. Pediatrics 104(3):405–418
Losey ET, Uber-Zak L (2008) Time to first interictal epileptiform discharge in extended
recording EEGs. J Clin Neurophysiol 25(6):357–360
Marsan CA, Zivin LS (1970) Factors related to the occurrence of typical abnormalities in the
EEG records of epileptic patients. Epilepsia 11:361–381
Marston D, Besag F, Binnie CD et al (1993) Effects of transitory cognitive impairment on
psychosocial functioning of children with epilepsy: a therapeutic trial. Dev Med Child Neurol
35:574–581
Nilsen KE, Walker MC, Cock HR (2005) Characterization of the tetanus toxin model of
refractory focal neocortical epilepsy in the rat. Epilepsia 46(2):179–187
Noebels JL (2003) The biology of epilepsy genes. Annu Rev Neurosci 26:599–625
Rakhade SN, Shah AK, Agarwal R, Yao B, Asano E, Loeb JA (2007) Activity-dependent gene
expression correlates with interictal spiking in human neocortical Epilepsy. Epilepsia
48(Suppl 5):86–95
Rötig S, Pillman F, Blöink R, Haring A (2005) Is there evidence in the EEG for increased
epileptiform activity in ICD-10 acute and transient psychotic disorder? Psychopathology
38:281–284
Schomer DL (2006) Ambulatory EEG telemetry: how good is it? J Clin Neurophysiol
23(4):294–305
Shelley BP, Trimble MR, Boutros NN (2008) Electroencephalographic cerebral dysrhythmic
abnormalities in the trinity of nonepileptic general population, neuropsychiatric, and
neurobehavioral disorders. J Neuropsychiatry Clin Neurosci 20(1):7–22
Shewmon DA, Erwin RJ (1998) The effects of focal interictal spikes on perception and reaction
time I. General consideration. Electroencephalgraphy Clin Neurophysiol 69:319–337
Shewmon DA, Erwin RJ (1988) The effects of focal interictal spikes on perception and reaction
timeII. Neuroanatomic specificity. Electroencephalgraphy Clin Neurophysiol 69:338–352
64 7 Isolated Epileptiform Discharges in Nonepileptic Psychiatric Patients

Stafstrom CE (2006) Behavioral and cognitive testing procedures in animal models of epilepsy.
In: Pitkanen A, Schwartzkroin PA, Moshe SL (eds) Models of seizures and epilepsy, Elsevier,
London, pp 613–628
Teicher MH, Ito Y, Glod CA, Andersen SL, Dumont N, Ackerman E (1997) Preliminary
evidence for abnormal cortical development in physically and sexually abused children using
EEG coherence and MRI. Ann N Y Acad Sci 821:160–175
Teicher MH, Andersen SL, Polcari A, Anderson CM, Navalta CP, Kim DM (2003) The
neurobiological consequences of early stress and childhood maltreatment. Neurosci Biobehav
Rev 27:33–44
Teicher MH, Dumont NL, Ito Y, Vaituzis C, Giedd JN, Andersen SL (2004) Childhood neglect is
associated with reduced corpus callosum area. Biol Psychiatry 56:80–85
Van Bogaert P, Urbain C, Galer S, Ligot N, Peigneux P, De Tiège (2011) Impact of focal
interictal epileptiform discharges on behaviour and cognition in children. Neurophysioloie
Clin Clin Neurophysiol 42:53–58
Verrotti A, Greco R, Altobelli E, Domizio S, Sabatino G, Morgese G, Chiarelli F (1999) Centro-
temporal spikes in non-epileptic children: A long-term follow up. J Pediatr Child Health
35:60–62
Weilburg JB, Schachter S, Worth J, Pollack MH, Sachs GS, Ives JR, Schomer DL (1995) EEG
abnormalities in patients with atypical panic attacks. J Clin Psychiatry 56:358
Williams D (1969) Neural factors related to habitual aggression. Consideration of differences
between those habitual aggressives and others who have committed crimes or violence. Brain
92:503–520
 Part II
Adult Psychiatric Disorders
Chapter 8
Panic Attacks and Other Dissociative
Disorders

Introduction

Panic disorder (PD) is one of the most common anxiety disorders. According to the
Epidemiologic Catchment Area (ECA) study, PD affects around 2 % of the adult
population (Markowitz et al. 1989). Panic attacks can be disabling and may lead to
agoraphobia with devastating psychosocial and economic consequences. The
identification of the etiology of these attacks is important for the proper man-
agement of the disorder.
DSM-V describes a typical panic attack as a discrete period of intense fear or
discomfort, in which at least four of the following symptoms develop abruptly and
reach a peak within minute: palpitation, sweating, trembling or shaking, sensations
of shortness of breath, feeling of choking, chest pain or discomfort, nausea or
abdominal stress, feeling dizzy (or lightheaded, unsteady or faint), derealization
(feelings of unreality), depersonalization (being detached from oneself), fear of
losing control or going crazy, fear of dying, parethesias, chills or hot flashes. Many
of these symptoms have been reported by patients with well-documented complex
partial seizures particularly of temporal lobe origin, i.e., temporal lobe epilepsy
(TLE). It is of great interest that despite the many similarities between PD and
TLE, and the documented increase in panic and fear symptomatology in epileptic
populations, EEG investigations are not currently a routine recommendation for
the work up of (PD) patients.

Panic as an Indicator of a Neurological Disorder

As early as 1880, Hughlings Jackson recognized ‘‘isolated fear’’ as an aura for

TLE. A significant number of case reports and case series appeared mainly in the
Neurology literature describing fear in association with TLE of a variety of eti-
ologies (Macrae 1954). The Macrae (1954) paper described seven patients with
similar presentations. Four of Macrae’s patients had no clear associated

N. N. Boutros, Standard EEG: A Research Roadmap for Neuropsychiatry, 67

DOI: 10.1007/978-3-319-04444-6_8,  Springer International Publishing Switzerland 2013
68 8 Panic Attacks and Other Dissociative Disorders

neurological symptoms at the time of presentations. Common to all the cases

presented were lesions affecting the medial aspects of the temporal lobe.
Panic symptoms carry a significant resemblance to symptoms induced by
temporolimbic epileptic activity particularly those originating from the Sylvian
fissure region. Fear, derealization, tachycardia, diaphoresis, and abdominal dis-
comfort are characteristic symptoms of simple partial seizures with psychiatric and
autonomic symptomatology. Evidence from population surveys suggests that PD is
significantly more prevalent in epileptic patients than the general public (Pariente
et al. 1991). While 21 % of epileptic patients in their sample of 11 epileptic had
panic attacks, only 3 % of the control sample had similar symptoms. While this
study is suggestive of an association, the sample of epileptic patients is too small
for the data to be definitive. These authors did not find a similar correlation
between epilepsy and agoraphobia, social phobia, or generalized anxiety. On the
other hand, Spitz (1991) found no increase in the prevalence of PD among the
epileptic populations in his clinic. He reported that epileptics with panic symptoms
responded better to benzodiazepines as compared to regular anticonvulsants.
Edlund et al. (1987) reported six cases of what they termed ‘‘atypical’’ panic
attacks. In three of the cases, the panic episodes were followed by rage attacks.
Two cases responded to anticonvulsants temporarily. The third was lost to follow-
up before response to treatment could be assessed. One patient with history of
severe head injury preceding the panic attacks responded well to anticonvulsants.
All four patients had focal EEG abnormalities mainly in the temporal regions. The
last two patients had prior history of seizures or brain surgery. The authors of this
article neglected to specify the general frequency of EEG abnormalities in their PD
population in general.
Toni et al. (1996) compared symptomatology of 91 patients with panic-disor-
der-agoraphobia (PDA) and 41 patients with complex partial epileptic seizures
(CPE). They found much similarity and concluded that there may be a common
neurophysiological substrate linking CPS and PDA (Table 8.1).
In another report, Young et al. (1995) described five patients with brief simple
partial seizures that mimicked panic attacks. They concluded that the most com-
mon psychiatric disorder that must be differentiated from TLE is PD. In their
sample, seizures were briefer and more stereotyped than panic attacks. Addi-
tionally, aphasia and dysmnesia accompanied seizure activity in some patients.
This differentiation could be diagnostically challenging as patients with docu-
mented complex partial seizures of temporal lobe origin may have concomitant
non-ictal episodic emotional symptoms, including phobia, true panic attacks, and
anxiety (Signer 1988).
Dantendorfer et al. (1996a) provided evidence that EEG abnormalities in PD
patients predicts the presence of MRI abnormalities particularly in the septo-
hippocampal region. They reported 60.7 % of patients with EEG abnormalities to
have MRI abnormalities as compared to patients without EEG abnormalities
(17.9 %) and healthy controls (3.6 %). Alemayehu et al. (1995) described an
association between panic attacks and seizures originating from the parietal lobe in
two patients with right parietal lobe tumors. Intracranial monitoring documented
Introduction 69

Table 8.1 (Modified from Toni et al. 1996/permission obtained)

Feature PDA CPE V2 p
(n = 55) (n = 25) (df = 1)
Depersonalization (%)
Feeling of loosing self-control 63.0 48.0 1.6 NS
Feeling of estrangement from one’s body 37.0 24.0 1.3 NS
Feeling of a change in identity 46.3 32.0 1.4 NS
Diminution or loss of affection 38.9 20.0 2.8 NS
Sensation of being unable to control one’s own 14.8 12.0 0.1 NS
mental activities
Feeling that one’s own ideas or memories are 33.3 20.0 1.5 NS
extraneous
Feeling of observing one self from afar 35.2 16.0 3.0 NS
Feelings that one’s own limbs are unreal 22.2 0 6.5 0.01
Physical hypoesthesia 22.2 28.0 0.3 NS
Feeling that one’s own body has changed in 14.8 20.0 0.3 NS
consistency
Feeling that one’s own body has changed in 13.0 24.0 1.5 NS
appearance
Feeling that one’s own limbs has changed in 13.0 20.0 0.7 NS
appearance
Derealization (%)
Detachment from the environment 68.5 76.0 0.5 NS
Feeling the external world is unfamiliar 53.7 52.0 0.0 NS
Space disorientation 37.0 64.0 5.0 0.02
Temporal disorientation 22.2 68.0 15.4 0.001
Feeling nothing makes sense 16.7 16.0 0.0 NS
Feeling people are dead or transformed 13.0 16.0 0.1 NS
Loss of perspective 3.7 4.0 0.0 NS
Differential Diagnostic features in panic disorder/agoraphobia (PDA) and in complex partial
seizures (CPE) patients

correlations between the symptoms of fear and restricted regional parietal cortical
discharges. Surgical resections of the lesions (one total, one subtotal) resulted in
complete recovery or improvement.
Other investigators had similar observations. Lee et al. (1997) reported the first
case of an adolescent female who presented with PD with agoraphobia which was
a consequence of seizure activity. Careful diagnostic evaluation and correlation
with video-electroencephalography were important in distinguishing seizure
activity from PD.
There are a number of reasons why it has been difficult to establish a rela-
tionship between some forms of panic attacks and seizure activity. Most impor-
tantly is the fact that EEG is routinely examined from surface (scalp recorded)
samples. The electrodes used to record the EEG activity are located at a significant
distance from the brain, particularly deep temporolimbic structures, with tissues
like scalp, skull, and dura intervening, further attenuating signals from the brain.
The second important reason why the relationship is difficult to establish is the
70 8 Panic Attacks and Other Dissociative Disorders

need for a demonstration of simultaneous abnormal EEG activity and behavioral

syndromes of panic attacks. These difficulties have led to conclusions that even
though panic attack patients are known to have a high prevalence of psychosensory
disturbances, their EEGs may fail to demonstrate abnormalities (Udhe et al. 1985).
Udhe and colleagues reported two out of 17 panic patients to exhibit EEG devi-
ations (12 %) (Uhde et al. 1989). One of the two patients had low voltage irregular
EEG which is not considered abnormal by today’s standards. The other had focal
slowing. Neither patient had epileptiform activity. Serious problems plagued this
report. We were not told how many had sleep tracings during the recordings
(recording following sleep deprivation does not by itself guarantee the occurrence
of sleep during the recording), we are not told if hyperventilation was performed
and how well and for how long, and no blinding was used in interpreting the
records. Moreover, nasopharyngeal leads were used in eight of the subjects. These
leads are extremely uncomfortable and usually do not allow subjects to relax or
fall asleep thus further decreasing the chances that epileptiform discharges will be
emitted.
Based on the above, it is possible that a patient presenting with panic symptoms
may be suffering from a form of temporal lobe or partial complex epilepsy.
Lesions in other brain regions may also result in panic symptoms. It should be
emphasized that it is well documented that a single negative EEG does not rule out
epilepsy even in patients with well-documented seizures. Moreover, the necessity
for obtaining sleep during the recording cannot be overemphasized. Finally, it is
possible that EEG abnormalities can only be detected during actual panic episodes
requiring more prolonged ambulatory EEG monitoring.
An illustrative case report provided by Dantendorfer et al. (1996b) demon-
strates the need for thorough organic work ups of such patients in order to identify
possible structural abnormalities contributing to the patient’s syndrome. The
patient suffered from treatment-resistant sudden arousals from sleep for 30 years.
Patient eventually developed daytime panic attacks. Sleep studies revealed sudden
arousal from stage 4 sleep. Resting wakeful EEG was normal. A subsequent EEG
study following sleep deprivation revealed increased left temporal theta activity.
MRI showed left hemispheric atrophy. Complete remission of symptoms could be
obtained only with a combination therapy of carbamazepine and clonazepam.

Does Laterality Play a Role in the Generation of Fear

or Panic Symptoms in Epileptic Patients?

Several animal and human studies have provided evidence for right hemisphere
dominance over the cerebral control of heart rate and blood pressure modulation,
typically ascribed to sympathetic lateralization in the right hemisphere (Hilz et al.
2001). Only few case reports attempted to address this question. Sazgar et al. (2003)
reported that five consecutive patients with complex partial seizures and panic
Introduction 71

symptoms, all five had the epileptic focus on the right hemisphere. Reports of panic
symptoms associated with left hemisphere epileptic foci have also been reported
(Saegusa et al. 2004). Whether the localization of the epileptic activity is related to
the symptomatology of the panic attacks remains to be investigated.

EEG in Nonepileptic and Neurologically Intact Panic

Disorder Patients

A number of reports provide evidence that EEG abnormalities are not infrequent in
PD patients. However, the findings differ from study to study and range from
paroxysmal epileptiform discharges to asymmetric increases in slow wave activity
(Bystritsky et al. 1999). Weilburg et al. (1993) reported two patients with atypical
panic attacks while their EEGs were being monitored. Focal paroxysms of sharp
wave activity appeared on the EEG coinciding with the spontaneous onset of panic
attack symptoms in both patients. Consciousness was maintained during these
episodes. Later; the same group (Weilburg et al. 1995) reported on fifteen subjects
with atypical panic attacks who met DSM-IIIR criteria for PD and who underwent
a routine EEG followed by prolonged ambulatory EEG monitoring using sphe-
noidal electrodes. They found focal paroxysmal EEG changes consistent with
partial seizure activity and occurring during a panic attack in 33 % (N = 5) of the
subjects. It is important to note that multiple attacks were recorded before panic-
related EEG changes were demonstrated. Moreover, two of the five subjects with
demonstrated EEG abnormalities during panic attacks had perfectly normal
baseline EEGs. They concluded that it may be necessary to monitor the EEG
during multiple attacks to reveal an association between atypical panic attacks and
epileptiform EEG changes. In a larger study Lepola (1990) examined the EEGs of
54 patients with PD (both typical and atypical). They found slow wave abnor-
malities in 13 of the 54 patients (24 %). While they did not include a normal
control group, this prevalence of slow-wave abnormality is significantly elevated
as compared to what would be expected from a normal age matched group
(approximately 5 %, see Normative EEG chapter). In this study only 20 % of
patients were recorded during panic attacks. Moreover, EEG recordings did not
necessarily include sleep tracings. It should be noted that all night sleep studies
cannot replace routine EEG in attempting to detect epileptiform activity (mainly
due to the abbreviated montage and paper speed).
Jabourian et al. (1992) performed 24 h ambulatory EEG in a population of 300
nonepileptic outpatients with anxious and/or depressive pathology and subjects
with panic attacks. The recordings revealed a high prevalence of abnormalities in
subjects referred with PD. Two groups of 150 medication-free patients each have
been selected on the bases of DSM-III-R; one with PD, the other with depressive
patients without paroxysmal anxiety (DS). The results showed respectively 63.2 %
abnormal, 19.7 % normal, and 17.1 % dubious records in the PD group. In the DS
72 8 Panic Attacks and Other Dissociative Disorders

group, 74.5 % normal, 18.3 % abnormal, and 7.2 % dubious records. Epileptiform
abnormalities were four times more frequent in the PD group (80 %) than in the
DS group (20 %). Two nycthemeral peaks were found (5–8 pm and 3 h after
awakening). MRI has permitted the discovery of abnormal cerebral images in three
patients of the PD group (cyst of the insula, temporal, and parietal cryptic an-
giomas, sequelae of a parietal vasculo-cerebral stroke).
Reports such as those continue to appear in the literature indicating the need for
detailed work up of every PD patient (Gallinant and Hegerl 1999; Gumnit 1994)).
They reported that PD patients with abnormal EEGs responded well to valproic
acid therapy. Bystritsky et al. (1999) found 25 % of a sample of 21 PD patients to
have abnormal EEGs with 15 % having epileptiform activity. Furthermore, they
reported that panic patients without demonstrable clinical EEG abnormalities
tended to have less alpha power in the right temporal region suggesting temporo-
limbic abnormalities in these patients.
One negative report deserves some focused discussion. Stein and Uhde (1989)
reported that only 14 % of a group of 35 panic patients had EEG abnormalities and
that none of the abnormalities was of the epileptiform type. They concluded that in
the absence of seizures, a standard EEG would not be a useful investigation. A
number of methodological and conceptual issues need to be discussed to help
inform future studies aiming at defining the role of sEEG in managing PD. The
most important point to underline is that a single negative EEG does not rule out
the presence of epileptic discharges. Given that there is a significant degree of
subjectivity in the clinical interpretation of the EEG (even in the best of hands), it
is most crucial for the interpreters to be blinded to the patient group. A TLE
control group must be included as well as a normal control group. While many of
the patients in this study were sleep deprived, many also had nasopharyngeal
electrodes placed which is uncomfortable and tend to not allow the patient to fall
into sleep (in fact the % of patients that had sleep tracings was not given). Given
the seriousness of the issue, all the above safeguards are essential for the credi-
bility and generalizability of the data.
Another rather important aspect of the above paper (Stein and Uhde 1989) is
that they examined the predictive value of the presence of ‘‘psychosensory’’
symptoms for the presence of EEG abnormalities. They concluded that these
symptoms did not predict the presence of EEG abnormalities when in fact out of
20 patients without such symptoms none had definite EEG abnormalities while 4
out of 15 panic patients with such symptoms had abnormal EEGs. In fact this
difference is statistically significant (Fisher Exact two tail p value is =0.0260).
Studies of the use of antiepileptic drugs to treat PD patients had inconsistent
findings. When the EEGs were shown to be abnormal, there was a tendency for
patients to improve clinically on antiepileptic medications. Table 8.2 summarizes
available literature where an anticonvulsant was used to treat nonepileptic PD
patients with evidence of EEG abnormalities.
The literature reviewed above provides presumptive evidence that the subgroup
of PD patients who exhibit EEG abnormalities may be a distinct subgroup with
different treatment responses. Treatment of the subgroup of PD patients with
Table 8.2 Efficacy of antiepileptic treatment in Panic patients with abnormal EEGs or other evidence of structural brain pathology
Paper Patients Controls EEG findings AED used Resultsa Comments
Edlund et al.Atypical Panic Attacks N/A Three paroxysmal activity, Either CBZ* alone or in Two of the four had significant Case
(1987) N = 4 patients with an and one temporal combination with a clinical improvement Series
adequate trial of slowing benzodiazepine
CBZ
Reid et al. Generalized anxiety N/A Left anterior temporal spike Phenytoin Phenytoin improved EEG. Single
(1988) disorder with panic wave focus CBZ Primidone stopped panic case
attacks N = 1 Primidone attacks study
McNamara Typical Panic Attacks N/A Temporal lobe slowing Different AEDs were used AEDs led to complete cessation Case
and Fogel N =5 (N = 2) and paroxysmal at times in combination of attacks in all patients Series
(1990) activity (N = 3)
Weilburg Atypical panic attacks N/A Focal paroxysms of sharp Phenobarbital and None of those drugs stopped the Small
et al. N =2 wave activity phenytoin during panic attacks in either patient case
(1993) childhood. Alprazolam series
with clonazepam.
VPA** with CBZ
Nickell Typical panic attacks N/A A single EEG was normal Upon recurrence Addition of CBZ caused Single
(1994) responsive to but MRI revealed imipramine dose was complete remission of panic case
imipramine for multiple meningiomas. increased and attacks report
3 years followed by fluoxetine was added
recurrence with no effects
Dantendorfer Sudden arousal from N/A Increased left temporal theta Clonazepam & CBZ Only the combination decreased Single
et al. sleep N = 1 panic frequency case
EEG in Nonepileptic and Neurologically Intact Panic Disorder Patients

(1996b) study
Windhaber Panic attacks N/A Anterior temporal sharp Oxcarbazepine Complete remission of panic Single
et al. developing waves attacks case
(1997) following two
grand mal seizures
(continued)
73
Table 8.2 (continued)
74

Paper Patients Controls EEG findings AED used Resultsa Comments

Huppertz 7-year-old boy with N/A Sharp and slow wave CBZ Significant decrease in frequency Single
et al. fear attacks and complexes (detected of attacks case
(2002) visual only with video-EEG) study
hallucinations.
Gallinat et al. 54-year-old female N/A Sharp and spike-wave Doxepin and Only VPA significantly Single
(2003) with atypical panic discharges bitemporally. Paroxetine had no effects. decreased frequency and case
attacks without MRI and SPECT showed VPA intensity of attacks. Epileptic study
clear epileptic left temporal potentials decreased but did
symptoms abnormalities not stop
*CBZ Carbamazepine
**VPA Valproic acid
a
EEG and/or clinical improvements. When not specified improvement was clinical
8 Panic Attacks and Other Dissociative Disorders
EEG in Nonepileptic and Neurologically Intact Panic Disorder Patients 75

demonstrable EEG abnormalities has not been well examined. McNamara and
Fogel reported five cases of panic attacks with abnormal EEGs who seem to have
responded favorably to anticonvulsant treatment (McNamara and Fogel 1990). All
five patients had unequivocal abnormalities in their EEGs and all five had com-
plete remissions when placed on anticonvulsants. These five cases seem to be a
highly selected group. In this paper, the authors attempt to provide guidelines for
when an EEG should be obtained. They did not provide evidence that such criteria
will predict the presence or absence of EEG abnormalities to a significant degree.
Two uncontrolled studies deserve special mention. First is a case series of eight
subjects with recurrent panic attacks who were successfully treated with clona-
zepam (Beaudry et al. 1985). The main characteristic of patients in this series was
the lack of response to benzodiazepines (other than clonazepam) with rather robust
responses to clonazepam treatment. Another important characteristic was the fact
that in this group of subjects no external or environmental factors were found that
can explain the symptoms. PD, in these eight subjects, was judged to have been
‘‘endogenous.’’ While the paper reports that EEGs were obtained, no details of
EEG methodology were provided. The paper also states that presence of evidence
of organicity was exclusionary. Again we were not told how many were excluded
based on this criterion and what were the abnormalities found? None the less, the
paper concluded that given the failure to respond to traditional benzodiazepines
but responding to clonazepam, that the anticonvulsant effects of this drug may
have played an important role in mediating the therapeutic effect. It should be
noted that the duration of treatment necessary to document response varied from 1
to 16 weeks. The second paper described the clinical response to carbamazepine in
14 PD patients (Uhde et al. 1989). The authors conducted a study of carbamaz-
epine in the treatment of 14 patients with PD. There was a statistically significant
reduction in symptoms of anxiety on several measures but only one of the patients
was judged to have a marked and sustained clinical improvement while taking
carbamazepine. 40 % of the patients had a decrease in frequency of panic attacks
during carbamazepine treatment, 50 % had an increase, and 10 % showed no
change. The presence of either EEG abnormalities or prominent psychosensory
symptoms did not predict response to carbamazepine. Whether these findings will
hold with larger sample sizes as well as other anticonvulsants remains to be seen.
As discussed above, this report suffered from a number of shortcomings in the
reporting of EEG procedure and results. Moreover, the abnormality that would be
predictive of response to an anticonvulsant is the presence of epileptiform activity.
According to the report, patients with abnormal EEGs did not exhibit spikes or
sharp waves. There is no reason, at least theoretically, to predict that a slow or an
irregular EEG would predict responsiveness to anticonvulsant therapy. Thus, the
conclusion of this paper that ‘‘EEG abnormality did not predict response to car-
bamazepine’’ cannot actually be reached based on the data provided. It should also
be highlighted that failure to respond in a single trial of an anticonvulsant does not
necessarily mean that this patient will not respond to any anticonvulsant. It is well
known that bona fide epileptic patients do not respond uniformly to a single
antiepileptic agent.
76 8 Panic Attacks and Other Dissociative Disorders

Heterogeneity of PD

Goddard and Charney (1997) proposed a neuroanatomical circuit for PD. This
paper, as well as other papers, implies that all panic attacks are of similar path-
ophysiology. In this chapter we provide evidence from the literature, that in
addition to idiopathic panic attacks, at least one other variant may exist that the
proposed circuitry may not be as applicable. The authors propose a mechanism by
which psychotherapy works in patients with panic attacks, again failing to warn
the reader that a subgroup of these patients (no matter how small) may be suffering
from some form of an abnormal electrical discharge in their limbic system. It is
rather unlikely that these patients will respond well to psychotherapy. The work by
Toni et al. (1996) provided evidence for a broad similarity between the psycho-
sensorial phenomena experienced by epileptic and PD patients as well as the
significantly greater than chance association between epilepsy and panic attacks,
provide further support for the hypothesis that there may be a common neuro-
physiological substrate linking epilepsy and PDs.
A sizeable literature, suggests that there may be a subtype of PD patients who
while never having experienced an epileptic seizure, may have epileptic activity in
panic-related brain regions most notably the amygdalae and insular regions. The
amygdala has long been linked with the experience of fear. Electrical stimulation
of this medial temporal structure predictably induces experiential feelings mainly
fear (Gloor et al. 1982). Amygdala source panic is likely to be characterized by
dominance of fear with other autonomic symptoms following either secondary to
spread of epileptic activity or as a psychological reaction to the fear (Keele 2005).
When the amygdala is electrically stimulated in awake humans (usually during
epilepsy work-up or surgery), fear is the most commonly generated experience
(Meletti et al. 2006). Stimulation of even close-by structures like the hippocampus
(while avoiding amygdalar involvement) leads to much less fear reactions. The
insula is located deep within the Sylvian fissure beneath the frontal, parietal, and
temporal opercula. The insula has long been associated with visceral functions and
is responsible for integrating autonomic information. The insula has wide con-
nections with the neocortex, basal ganglia, thalamus, and the amygdala among
other limbic structures. This wide connectivity explains the varied symptoms of
seizures originating from this region (Nguyen et al. 2009). Epileptic activity
emanating from this small and deep region is unlikely to be detected by standard
scalp EEG. Insular source panic attacks would begin with the autonomic symp-
toms with fear following again either secondary to spread of epileptic activity or as
a conditioned response to the experience of the sudden and usually unprovoked
autonomic activation (Nguyen et al. 2009). The contribution of such epileptic
activity (usually isolated events of milliseconds durations) to the full-blown panic
attack (which is usually many minutes in duration) deserves further exploration.
An issue that deserves some discussion is the actual interpretation of an EEG
finding when one is detected in a PD patient. Hayashi et al. (2010) reported only
17 of 70 PD patients (24 %) had identified EEG abnormalities in a routine
EEG in Nonepileptic and Neurologically Intact Panic Disorder Patients 77

outpatient procedure. Of those 17 only2 had frank epileptic discharges. The other
15 had episodic (or occasional) slow waves in theta range. Nausea or abdominal
distress, derealization/depersonalization, and paresthesias were significantly rela-
ted to the presence of EEG abnormalities. The question to be answered is whether
the intermittent slow wave activity noted is in fact a reflection of a deeper par-
oxysmal or epileptiform activity. In fact, the value of utilizing sphenoidal elec-
trodes in probing panic attacks has not been investigated. These invasive
electrodes can be kept in place for a number of days thus allowing continuous
recording from the amygdala-hippocampal region in patients exhibiting frequent
panic episodes.
Computerized EEG promises further refining of the utility of EEG in detecting
abnormalities in PD patients as well as defining the diagnostic accuracy overall.
Abraham and Duffy (1991) were able to differentiate between panic patients and
control subjects with 92.5 % accuracy. An intriguing report by Enoch et al. (1995)
deserves a mention. They reported that low voltage alpha activity is a trait-
dependent variable that may be linked to the vulnerability to anxiety disorders
(including panic attacks) as well as alcoholism.
Finally, Gallinat and Hegerl (1999) summed up the literature examining the
relationship between PD and epilepsy by stating that a ‘‘subset of panic attacks
may be related to abnormal epileptiform neuronal activity in the limbic system.’’
The finding that anxiety is the most common experiential phenomenon produced
by electrical stimulation of the amygdala and hippocampus with depth electrodes
was the most convincing piece of evidence for them. The size of this subset is
difficult to determine because epileptiform discharges in the depth of the limbic
system often cannot be seen in the scalp EEG. They recommended that a trial of
anticonvulsants (they used valproic acid) should be attempted if standard phar-
macotherapy failed. Indeed the interrelationship between panic symptoms and
epilepsy remains not fully elucidated (Handal et al. 1995). Handal and colleagues
described three cases from their practice where in one PD was misdiagnosed as
complex partial seizure, in the second complex partial seizure was misdiagnosed
as PD, and the third patient with both disorders where one of the two disorders was
missed altogether.

Supported Findings

(1) EEG abnormalities are significantly increased in PD patients. A sizable pro-

portion, approximately 25–30 % of panic attack patients have demonstrable
EEG abnormalities indicative of a process other than an idiopathic PD.
(2) The presence of an EEG abnormality in a PD patient correlates with the
presence of other neurological abnormalities.
(3) Paroxysmal EEG abnormalities in PD patients could be indicative of a seizure
disorder.
78 8 Panic Attacks and Other Dissociative Disorders

(4) It is possible that these patients may be responsive to anticonvulsant treatment

(Guay et al. 1995). In the absence of large well-designed studies, patients with
panic attacks, and abnormal EEGs should have a trial of anticonvulsant
therapy.

Open Research Questions

(1) What is the actual prevalence of IEDs in PD patients?

(2) Are the IEDs detected in PD patients always of temporal lobe origin? Is there a
temporal lobe location that favors the appearance of fear symptoms?
(3) What are the treatment responses to AEDs of different chemical families in PD
patients exhibiting paroxysmal activity.
(4) Can specialized electrode placements improve the rate of detection of IEDs in
this population? A number of surface locations (Zygomatic and true anterior
temporal electrodes) as well as invasive electrodes like Sphenoidal electrodes
(see Chap. 3).
(5) What do focal slow-wave abnormalities mean in PD patients. When do surface
slow-waves reflect deeper epileptic activity?
(6) Large sample studies are necessary to further define the nature of neurological
processes that can masquerade as PD. The development of the clinical pre-
dictors of such outcome would be very useful to the clinician.
(7) It is also yet to be established whether or not the identification of an under-
lying neurological disorder does lead to a different clinical or therapeutic
outcomes.
(8) Do the panic symptoms of an individual relate to the laterality of the epileptic
focus? This is an important question as it could impact our understanding of
the circuitry mediating different forms of panic symptoms.

Other Dissociative Disorders

By and far, SEEG investigations have not been conducted in patients with other
dissociative disorders like fuge states, and multiple personality disorder.
In 1981 Marsel Mesulam published a detailed account of 12 patients with clinical
and EEG manifestations reminiscent of TLE (Mesulam 1981). In seven of these
patients the clinical picture was consistent with multiple personality disorder,
whereas the other five had the illusion of supernatural possession. The EEGs of the
majority of these patients revealed temporal lobe epileptic discharges. Subse-
quently, Devinsky and his colleagues reported six additional cases of multiple
personality disorder who received intensive EEG and video monitoring (Devinsky
Other Dissociative Disorders 79

et al. 1989). In none of these patients was the diagnosis of epilepsy substantiated. It
should be noted that all six patients were diagnosed with epilepsy prior to the EEG/
video monitoring. This report highlights the necessity of careful evaluation prior to a
diagnosis being assigned. The same group examined dissociative symptoms in 71
epileptic patients. The dissociative Experience Scale was administered. Partial
seizure patients with dominant hemisphere foci had higher depersonalization sub-
scale scores than those with nondominant foci (Devinsky et al. 1989).

Open Research Questions

(1) in fact this category of patients cannot be said to have been investigated
utilizing the standard EEG hence, the prevalence of EEG abnormalities either
slowing focally or diffusely, and focal paroxysmal activity are not known.
(2) The additional role of the QEEG is to be established.
(3) Response to various forms of treatment once an abnormality is detected is also
yet to be investigated.

References

Abraham HD, Duffy FH (1991) Computed EEG abnormalities in panic disorder with and without
premorbid drug abuse. Biol Psych 29:687–690
Alemayehu S, Berger GK, Barry E, Krumholz A, Wolf A, Fleming CP, Frear EJ Jr (1995) Panic
attack as ictal manifestations of parietal lobe seizures. Epilepsia 36:824–830
Beaudry P, Fontain R, Chouinard G, Annable L (1985) An open clinical trial of clonazepam in
the treatment of patients with recurrent panic attacks. Prog Neuropsychopharm Biol Psychiat
9:589–592
Boutros NN, Gjini K, Moran J, Chugani H, Bowyer S (2013) Panic vs epilepsy; a challenging
differential diagnosis. In Press, Clin EEG and Neurosciences
Bystritsky A, Leuchter AF, Vapnik T (1999) EEG abnormalities in nonmedicated panic disorder.
J Nerv Ment Dis 187(2):113–114
Dantendorfer K, Frayer D, Kramer J, Amering M, Baischer W, Berger P, Schoder M, Steinberger
K, Windhaber J, Imhof H, Katsching H (1996a) High frequency of EEG and MRI brain
abnormalities in panic disorder. Psychiatry Res 68:41–53
Dantendorfer K, Frey R, Maierhofer D, Saletu B (1996b) Sudden arousal from slow wave sleep
and panic disorder: successful treatment with anticonvulsants— a case report. Sleep
19:744–746
Devinsky O, Putnam F, Grafman J, Bromfield E, Theodore WH (1989) Dissociative states and
epilepsy. Neurology 39:835–840
Edlund MJ, Swann AC, Clothier J (1987) Patients with panic attacks and abnormal EEG results.
Am J Psychiat 144:508–509
Enoch MA, Rohrbaugh JW, Davis EZ, Harris CR, Ellingson RJ, Anderson P, Moore V, Varner
JL, Brown GL, Eckardt MJ, Goldman D (1995) Relationship of genetically transmitted alpha
EEG traits to anxiety disorders and alcoholism. Am J Med Genet 60:400–408
80 8 Panic Attacks and Other Dissociative Disorders

Gallinat J, Hegerl U (1999) Limbic ictus as a condition for anxiety attacks. Der Nervenartzt
70:206–215
Genton P, Bartolomei F, Guerrini R (1995) Panic attacks mistaken for relapse of epilepsy.
Epilepsia 36:48–51
Gloor P, Olivier A, Quesney LF, Andermann F, Horowitz S (1982) The role of the limbic system
in experiential phenomena of temporal lobe epilepsy. Ann Neurol 12(2):129–144
Goddard AW, Charney DS (1997) Toward an integrated neurobiology of panic disorder. J Clin
Psychiat 58 Suppl 2:4–11 (discussion 11–2)
Guay DR (1995) The emerging role of valproate in bipolar disorder and other psychiatric
disorders. Pharmacotherapy 15:631–647
Gunmit RJ (1994) The Epilepsy Handbook, 2nd edn. Raven press, New York, p 4–5
Handal NM, Masand P, Weilburg JB (1995) Panic disorder and complex partial seizures: a truly
complex relationship. Psychosomatics 36:498–502
Hayashi K, Makino M, Hashizume M, Nakano K, Tsuboi K (2010) Electroencephalogram
abnormalities in panic disorder patients: a study of symptoms characteristics and pathology.
BioPsychSocial Med 4:9
Hilz MJ, Dütch M, Perrine K (2001) Hemispheric influence on autonomic modulation and
baroreflex sensitivity. Ann Neurol 49:275–284
Huppertz HJ, Franck P, Korinthenberg R, Schulze-Bonhage A (2002) Recurrent attacks of fear
and visual hallucinations in a child. J Child Neurol 17(3):230–3
Jabourian AP, Erlich M, Desvignes C, el Hadjam M, Bitton R (1992) Panic attacks and 24-h
ambulatory EEG monitoring. Ann Med Psychol 150:240–244
Keele NB (2005) The role of serotonin in impulsive and aggressive behaviors associated with
epilepsy-like neuronal hyperexcitability in the amygdala. Epilepsy Behav 7(3):325–335
Lee DO, Helmers SL, Steingard RJ, Demaso DR (1997) Case study: seizure disorder presenting
as panic disorder with agoraphobia. J Am Acad Child Adolsc Psychiat 36(9):1295–1298
Lepola U, Nousiainen U, Puranen M, Riekkinen P, Rimon R (1990) EEG and CT findings in
patients with panic disorder. Biol Psychiat 28:721–727
Macrae D (1954) On the nature of fear, with reference to its occurrence in epilepsy. J Nerv Ment
Dis 120(5–6):385–393
Markowitz JS, Weissman MM, Ouellett R (1989) Quality of life in panic disorder. Arch Gen
Psychiat 46:984–992
McNamara ME, Fogel BS (1990) Anticonvulsant-responsive panic attacks with temporal lobe
EEG abnormalities. J Neuropsychiatry Clin Neurosci 2:193–196
Meletti S, Tassi L, Mai R, Fini N, Tassinari CA, Russo GL (2006) Emotions induced by
intracerebral electrical stimulation of the temporal lobe. Epilepsia 47(Suppl 5):47–51
Mesulam MM (1981) Dissociative states with abnormal temporal lobe EEG. Arch Neurol
38:176–181
Nguyen DK, Nguyen DB, Malak R, Bouthillier A (2009) Insular cortex epilepsy: an overview.
Can J Neurol Sci 36(Suppl 2):S58–S62
Nickell PV (1994) Panic attacks, complex partial seizures, and multiple meningiomas. Anxiety
1:40–42
Pariente D, Lepine JP, Lellouch J (1991) Life time history of panic attacks and epilepsy: an
association from a general population survey. J Clin Psychiatry 52:88–89
Reid TL, Raj BA, Sheehan DR (1988) Ictal panic/epileptogenic activity: treatment with
primidone. The Academy of psychosom med 29(4):431–433
Saeggusa S, Takahashi T, Moriya J, Yamakawa J, Itoh T, Kusaka K, Kawaura K, Kanda T (2004)
Panic attack symptoms in a patient with left temporal lobe epilepsy. J Int Med Res 32:94–96
Sazgar M, Carlen PL, Wennberg R (2003) Panic attack semiology in right temporal lobe epilepsy.
Epileptic Disord 5:93–100
Signer SF (1988) Seizure disorder or panic disorder? Amer J Psychiat 145(2):275–276
Spitz MC (1991) Panic disorder in seizure patients: a diagnostic pitfall. Epilepsia 32:33–38
Stein MB, Uhde TW (1989) Infrequent occurrence of EEG abnormalities in panic disorder. Am J
Psychiatry 146:517–520
References 81

Toni C, Cassano GB, Perugi G, Murri L, Mancino M, Petracca A, Akiskal H, Roth SM (1996)
Psychosensorial and related phenomena in panic disorder and in temporal lobe epilepsy.
Compr Psychiatry 37:125–133
Uhde TW, Boulenger JP, Roy-Byrne PP, Geraci MF, Vittone BJ, Post RM (1985) Longitudinal
course of panic disorder: clinical and biological considerations. Prog Neuro-Psychopharmacol
Biol Psychiatry 9(1):39–51
Uhde TW, Stein MB, Post RM (1989) Lack of efficacy of carbamazepine in the treatment of
panic disorder. Am J Psychiat 145:1104–1109
Weilburg JB, Schachter S, Worth J, Pollack MH, Sachs GS, Ives JR, Schomer DL (1995) EEG
abnormalities in patients with atypical panic attacks. J Clin Psychiatry 56:358–362
Weilburg JB, Schachter S, Worth J, Pollack MH, Sachs GS, Ives JR, Schomer DL (1993) Focal
Paroxysmal EEG changes during atypical panic attacks. J Neuropsychiatry 5:50–55
Windhaber J, Maierhofer D, Dantendorfer K (1997) Oxcarbazepine for panic disorder occurring
after two grand mal seizures: a case report. J Clin Psychiatry 58:404–405
Young GB, Chandarana PC, Blume WT, McLchlan RS, Munoz DG, Girvin JP (1995) Mesial
temporal lobe seizures presenting as anxiety disorders. J Neuropsychiat Clin Neurosci
7:352–357
Chapter 9
Violence Aggression and Impulse
Dyscontrol

Introduction

The thorough understanding of the biological mechanisms contributing to habitual

aggression is fundamental if effective preventive, diagnostic, and rehabilitative
programs are to be developed. Electrophysiological techniques, including con-
ventional and quantified EEG can be very helpful in advancing our knowledge of
this area. Of great interest is the serious gap between a large body of literature
attesting to the prevalence of EEG, and other brain imaging abnormalities, and the
actual utilization of this information in diagnosing and managing individuals
exhibiting such symptoms. Moreover, the dearth of more recent research in this
area further deepens this gap.

Episodic Aggression and Impulse Dyscontrol

The prevalence of abnormal EEGs in this clinical population vary widely among
studies ranging from as low as 6.6 % in patients with rage attacks and episodic
violent behavior (Riley and Niedermeyer 1978) to as high as 53 % in patients
diagnosed with antisocial personality disorder (Harper et al. 1972). Rare negative
studies, showing lack of significant EEG abnormalities in patients with rage
attacks or episodic violent behavior have also appeared. Riley (1979) examined the
EEGs of 212 patients. He found 14 subjects with abnormal EEGs for a rate of
6.7 %. In this study, 76 % of EEGs included sleep recordings and in some subjects
repeat EEGs were obtained (Riley 1979).

EEG and Episodic Dyscontrol

Bach-Y-Rita et al. (1971) reported the EEGs of 79 patients diagnosed with

‘‘episodic dyscontrol.’’ Thirty-seven of them were abnormal (close to the widely
reported 50 % incidence). Of the 37 abnormal records 20 showed spikes in the

N. N. Boutros, Standard EEG: A Research Roadmap for Neuropsychiatry, 83

DOI: 10.1007/978-3-319-04444-6_9,  Springer International Publishing Switzerland 2013
84 9 Violence Aggression and Impulse Dyscontrol

temporal region. They classified patients that were included in the study into four
categories; (1) patients already diagnosed with temporal lobe epilepsy; (2) patients
with epilepsy like episodes; (3) patients with ‘‘diffuse violence’’ with violent
outbursts at varied targets. These subjects constituted the largest group and
exhibited a significantly increased level of anxiety. The fourth group was that with
‘‘pathological intoxication.’’
Bennett et al. (1983) examined the EEGs of 48 children between the ages of 5.2
and 12.9 years who were hospitalized for aggressive, explosive, or conduct dis-
orders. EEGs were examined at baseline, on placebo, haloperidol, or lithium. They
reported a prevalence of 58.3 % abnormalities at baseline. Both haloperidol and
lithium caused the EEGs to look more abnormal (even in children who seem to be
responding to treatment). It should be noted that three EEGs that were found to be
normal on initial testing were found to be abnormal on subsequent testing later on
while patients were on placebo. This finding attests to the value of repeated EEG
testing.
Boelhouwer et al. (1968) was able to predict the presence of the 14 and 6
Positive Spikes (PS) by selecting a group of adolescents and young adults who
exhibited episodic aggressive outbursts (See Chap. 19 for more detail). Subjects
with PS had significantly higher histories of their mothers experiencing toxemia
during pregnancy with them. With extensive psychological testing the PS indi-
viduals had significantly more problems with judgement and success of control
mechanisms. It is of interest to note that these investigators found the PS to occur
independently of any diagnostic category listed by the American Psychiatric
Association at that time. Furthermore, subjects with PS had significantly more
anxiety and were more insightful and more ready to feel guilty and be self-critical
than subjects in a control group with similar behaviors but without PS. The
treatment implications of these psychological findings were not discussed but a
pharmacological investigation was reported. With a maximum of 8 week trials
comparing thioridazine, diphenylhydantoin, or combination of the two against
placebo. They reported that the PS group responded best to the combination of
drugs. These investigators then assessed the EEGs for the presence of the posterior
temporal lobe slowing (see below). They noted that patients with both abnor-
malities did least well on any treatment while patients with posterior temporal
slowing alone did best on the anticonvulsant as sole treatment.
Monroe (1989) followed 50 patients with episodic dyscontrol ðN ¼ 33Þ or other
episodic symptoms ðN ¼ 17Þ for 38 to 57 months. On the bases of alpha chlo-
ralose activated EEG results, an anticonvulsant was recommended for 39 patients.
Twenty patients received medications that raised the seizure threshold. Of these 20
patients, 10 reported marked and six reported moderate improvement. A relatively
more recent study utilizing spectral analysis of the EEG, reported an increase in
slow-wave activity (Delta) in the right frontal region (Bars et al. 2000) suggesting
that a frontal lobe abnormality could contribute to the picture of episodic
dyscontrol.
Episodic Aggression and Impulse Dyscontrol 85

Habitual Versus Sporadic Aggression

Williams (1969) compared the EEGs of 206 habitual aggressors and 127 who
committed isolated acts of violence. He reported a five-fold increase in EEG
abnormalities in habitual aggressors; 57 % as compared to 21 % in nonhabitual
aggressors. They also found more frontal region abnormalities in the habitual
aggressors but more diffuse and epileptic activity in the nonhabitual aggressors.
Another important finding reported by Howard (1984) is that patients who have
committed violent offences against strangers, as opposed to people known to them,
tended to have bilateral paroxysmal EEG features. 70 % of subjects with bilateral
paroxysmal discharges have attacked strangers. They found no statistically sig-
nificant associations between laterality of abnormality and personality variables, or
legal or diagnostic category.

EEG and Antisocial Personality Disorder

As early as the mid-1940s, it was recognized that criminals had a higher preva-
lence of EEG abnormalities. Among psychiatric populations the group of ‘‘psy-
chopaths’’ had the largest incidence of either borderline or frank abnormalities
which consisted mainly of diffuse background slowing (unmedicated patients) and/
or paroxysmal activity with or without spike components (Hill and Watterson
1942). Hill and Watterson (1942) examined the EEGs of 151 subjects with psy-
chopathic personalities. They reported 48 % of this group to exhibit abnormal
EEGs as compared to 15 % of a nonpatient control group. When they divided the
group into aggressive (N = 66) and nonaggressive ðN ¼ 38Þ they found 65 % of
aggressive patients and only 32 % of nonaggressive subjects to exhibit abnormal
EEGs. In this chapter they also reported a significant relationship between history
of head injury and presence of EEG abnormalities. They concluded that the more
aggressive the patient the more likely the EEG to be abnormal. Wong et al. (1994)
retrospectively examined the EEGs and CT scans of 372 male-patients in a
maximum-security mental hospital. Reviewers were blind to the specific history of
the individual. They reported that 20 % of the EEGs (and 41 % of CT scans) were
abnormal in the most violent patients as compared to 2.4 % (6.7 % for CT scans)
for the least violent patients.
Patients diagnosed with antisocial personality disorder frequently harbor
organic brain pathology that can be assessed with help of the EEG along with other
neuro-evaluative tools. Blake et al. (1995) performed detailed and thorough neu-
rological evaluations of 31 individuals awaiting trial or sentencing for murder.
EEGs, MRIs or CT scans, and neuropsychological testing were obtained from
most of the subjects. Neurological examination revealed evidence of ‘‘frontal’’
dysfunction in 20 (64.5 %). There were symptoms or some other evidence of
temporal lobe dysfunction in 9 (29 %). Specific neurologic diagnoses were made
in 20 (64.5 %). These diagnoses included borderline or full mental retardation in 9
and cerebral palsy in 2. Most importantly is that neuropsychological testing
86 9 Violence Aggression and Impulse Dyscontrol

revealed abnormalities in all subjects tested. There were EEG abnormalities in

eight of the 20 subjects who had EEGs. EEG abnormalities consisted mainly of
bilateral sharp waves with slowing. There were MRI or CT scans abnormalities in
nine of 19 subjects tested, primarily atrophy and white matter changes. There was
a documented history of profound and protracted physical abuse in 26 (83.8 %)
and sexual abuse in 10 (32.3 %). They concluded that prolonged, severe physical
abuse and neurological brain dysfunction interact with paranoia (all subjects had
evidence of paranoid ideations) to form the matrix of violent behavior.
It has also been shown that among groups of prisoners convicted of murder, the
highest incidence of EEG abnormalities (74 %) occurred in individuals whose
crimes were apparently motiveless or had minimal motives (Stafford-Clark and
Taylor 1949). In a well-designed relatively large study, Howard supported the
above findings (Howard 1984). In this study, the EEGs of 265 consecutive
admissions to a special hospital for offenders were examined in detail. Medication
status, nature of the offense as well as personality characterization were reported.
The EEGs were classified into; (a) normal (monorhythmic); (b) low voltage fast;
(c) choppy (i.e., dysrhythmic with excess theta); and (d) dysrhythmic with par-
oxysmal features. Their first major finding was that the prevalence of abnormal-
ities was not different between medicated and nonmedicated subjects, strongly
suggesting that abnormalities in this population are not secondary to medication
effects (64.8 % for medicated and 61.4 % for unmedicated patients). We should
note that Howard (1984) included ‘‘low voltage fast’’ as an abnormal EEG pattern.
Current day practice of standard EEG interpretation tends to discount this pattern.
None the less this pattern constituted less than 15 % of Howard’s subjects leaving
at least full 50 % of the subjects with clear EEG abnormalities. This indeed seems
to be the overall impression one gets from examining this complex body of lit-
erature. The close to 50 % prevalence of EEG abnormalities in association with
violence seems to be culturally independent. Okasha et al. (1975) reported a
prevalence of 43 % of EEG abnormalities in a group of Egyptian murderers.
Nelson and Boutros (1993) examined consecutive admissions to an inpatient adult
psychiatry unit with diagnosis of antisocial personality and found majority to have
evidence of organic brain involvement (majority with EEG abnormalities) war-
ranting changing the diagnosis in a majority of subjects to Organic Personality
Disorder according to the DSM-III.

EEG and Institutional Aggression (Including Aggression

in Schizophrenia)

Violence is also a problem in psychiatric institutions. With de-institutionalization,

the problem is inherited by inner city communities where most of these patients
tend to segregate. Barber et al. (1988) examined the clinical characteristics of 15
patients with repetitively assaultive behavior. These patients constituted 3.3 % of
the average daily census but accounted for 48.6 % of all assaultive incidents
during a 1-year period. As with the findings of Blake et al. above, 6 of the 10 with
Episodic Aggression and Impulse Dyscontrol 87

known IQs were either in the borderline or frankly subnormal ranges (Blake et al.
1995). Of the 15, 5 had abnormal EEGs and 6 had histories of significant head
injuries. Two patients had atypical complex partial seizures and one had both
generalized and partial seizures.
Sayed et al. (1969) reported the EEG abnormalities in a group of 32 murderers
who were deemed ‘‘insane’’ as compared, in a blind interpretation design, to a
group of nonpatient controls. They found an overall incidence of abnormalities not
far different from what has been reported in other studies (65.6 %) which was
approximately four times the incidence of EEG abnormalities in the control group.
What is interesting about this study is that the increased incidence in this group
seems to result from a higher prevalence of EEG abnormalities in the subgroup
diagnosed as ‘‘schizophrenics’’ with 73.4 % as compared to the nonpsychotic
psychopathic group (50 %). Three murderers with ‘‘psychotic depression’’ all had
abnormal EEGs. Again, the most frequently encountered abnormality was diffuse
slowing of the background rhythm (66 %) with paroxysmal abnormalities only in
four subjects (19 %). This was not the case in a group of children who committed
murders before the age of 16. Bender (1959) reviewed the clinical histories of 33
such children. EEGs were available for 15 of the children. Ten of the fifteen
records were abnormal. While no detailed EEG data were provided the author
indicated that three of the children with abnormal EEGs went ahead and developed
frank epilepsy some time after the fatal incident they were involved in. The other
abnormal EEGs were also suggestive of an epileptic process.

Specific EEG Abnormalities

Hemispheric Asymmetry

Convit et al. (1991) demonstrated that violence was very significantly related to
the hemispheric asymmetry in EEG for the frontotemporal regions. They provided
evidence that with increased levels of violence a greater level of delta power in the
left compared with the right hemispheres can be found. A relationship between left
hemisphere focal EEG abnormalities and increased violent tendencies was further
supported by Pillmann et al. (1999) who examined the EEGs of 222 offenders
referred for psychiatric evaluation. They found left hemisphere focal slowing to be
significantly related to higher numbers of violent offenses. The majority of focal
abnormalities were localized to the temporal lobe. They further confirmed that the
presence of mental retardation, epilepsy, and history of earlier brain damage were
contributory to the degree of violence subjects exhibited. These investigators
utilized highly conservative criteria for identifying EEG abnormalities (only def-
initely unambiguous abnormalities were identified with all controversial wave-
forms not included). They found an overall rate of abnormality of 40 %. Only one
subject had clear spike and wave discharges. Wong et al. (1994) examined the
EEGs of 372 inmates of a special hospital. They found a higher incidence of focal
88 9 Violence Aggression and Impulse Dyscontrol

abnormalities in a subgroup with the highest violence ratings with the majority of
abnormalities localized to the temporal lobes.

Frontal Lobe Slowing

EEG may also reveal other pathologies like frontal lobe abnormalities. Specific
cognitive deficits revealed via neuropsychological testing may be more amenable
to cognitive rehabilitation techniques rather than pharmacological treatment.

Posterior Temporal Lobe Focal Slow

Posterior temporal lobe focal slow wave abnormality has been described to be
more prevalent in populations that had higher propensity for violent or aggressive
acts (Fenton et al. 1974). This association was first noted in the early 1940s (Hill
1944). Later on, Hill reported a 12 % (N = 194 nonepileptic psychopaths) pos-
terior temporal lobe slowing in association with psychopathic personalities as
compared to 2 % in healthy control subjects. They also noted that this abnormality
tended to decrease with increasing age suggesting a maturational nature of the
abnormality (Hill 1952). Aird and Gastaut (1959) similarly noticed that children
with this abnormality tend to mature out of it. Rey et al. (1949) further confirmed
the increased prevalence of this abnormality in psychopathic patients. Fenton et al.
(1974) examined new admits and longer term residents of a special hospital for
patients with propensity to violence. They noted that a significant portion of
subjects with habitual aggression may continue to exhibit the posterior temporal
slowing on their EEGs. They suggested that the maturational theory does not fully
explain the association between the EEG abnormality and aggression. While in
their preliminary work they were unable to find a clear correlate (except a ten-
dency for subjects to be more violent) they suggested that further research was
necessary. None the less, the latest documented investigation of this phenomenon
was the (Fenton et al. 1974) paper. There is in fact no clear explanation for why
this line of research was abandoned given the seriousness and the severely stig-
matizing nature of the problem.

Treatment Implications

Whether the appearance of an IED in the standard EEG predicts a favorable

therapeutic response to anticonvulsant medications is currently unknown.
Table 9.1 lists studies where an anticonvulsant medication was used to treat
nonepileptic patients with aggression and impulse dyscontrol and abnormal EEGs.
Monroe (1975) showed that anticonvulsants can block electroencephalographic
epileptiform discharges and can lead to dramatic clinical improvement in
Table 9.1 Anticonvulsant treatment in nonepileptic patients with aggression and impulse dyscontrol and abnormal EEGs
Chapter Patient group Control group EEG findings AED use Results Comments
Hakola Violent N/A Three with focal paroxysmal, one CBZ Violent episodes decreased Case series
et al. (incarcerated) with diffuse slowing and four significantly in all eight
(1982) schizophrenic normal EEGs patients
women N = 8
Schiff Aggressive behavior N/A Bilateral frontal and temporal Propranolol Phenytoin plus CBZ were Single case
Treatment Implications

et al. N=1 slowing and spike and sharp Phenytoin ineffective. Lithium study
(1982) waves CBZ Lithium addition caused increased
violence attacks and more
spikes
Yassa Aggressive paranoid N/A Nonspecific EEG abnormality Chlorpromazie Aggressive behavior decreased Single case
et al. schizophrenic CBZ study
(1983) N=1
Luchins Violent nonepileptic Normal EEGs Not defined 6 weeks before Reduced aggression in all Controlled
(1984) schizophrenia N = 11 CBZ and patients. No significant study
patients with 6 weeks of difference based on EEG
abnormal EEGs CBZ were
N=8 compared
Monroe Various psychiatric N/A Temporal slowing, paroxysmal Two on CBZ Two patients improved Case series
(1986) disorders activity, increased delta-theta one on VPA clinically
accompanied by activity
aggression
N=3
(continued)
89
Table 9.1 (continued)
90

Chapter Patient group Control group EEG findings AED use Results Comments
Stone IED N = 1 N/A 6/s spike during hyperventilation. CBZ Relaxed but was hyper if Single case
et al. 6 and 12/s sharp spike during missed a dosage study
(1986) drowsiness
Mattes Intermittent patients without Mainly focal temporal slowing Propranolol EEG abnormalities did not Randomized
(1990) Explosive EEG versus CBZ significantly predict study
Disorder with abnormalities favorable response to CBZ
abnormal EEGs N = 60
N = 20
Reeves Axis-II disorders 22 VPA 8 of 22 VPA responders and 5 of VPA Presence of slow wave Correlational
et al. with aggression responders the nonresponders had abnormalities did not study
(2003) N = 42 and 20 VPA abnormal EEGs. All EEG predict response to VPA
nonresponders abnormalities were non
epileptiform
9 Violence Aggression and Impulse Dyscontrol
Treatment Implications 91

individuals exhibiting repeated and frequent aggressive behavior. An earlier study

by Boelhouwer et al. (1968) found adolescents or young adults exhibiting the 14
and 6 positive spikes to respond favorably to the combination of anticonvulsants
and antipsychotic medications. Tunks and Dermer (1977) reported a detailed case
where other than deafness, there were no obvious neurological abnormalities in a
female with episodic aggression who responded extremely well to carbamazepine
therapy. Neppe (1983) provided evidence that the addition of carbamazepine to the
treatment of schizophrenia patients, who also exhibit temporal lobe abnormalities
on the EEG and without a history of a seizure disorder, can be clinically useful.
Earlier, Hakola and Laulumaa (1982) noted a reduction of aggressive episodes
when carbamazepine was added to the neuroleptic regime of eight highly
aggressive women with schizophrenia who also had EEG abnormalities. A detailed
case report of a response of a severely aggressive patient with schizophrenia who
responded well to carbamazepine adjuvant therapy was reported by Yassa and
Dupont (1983). Mattes (1990) examined the clinical response to propranolol
versus carbamazepine in a group of 80 patients with rage outbursts. Fifty-one of
the subjects were randomized while patients with history of epilepsy (N = 11) were
assigned to the carbamazepine arm and patients with known allergy to one of the
medications were assigned to the other study medication. Of the 80 subjects 20 had
abnormal EEGs including 9 of the 11 with epilepsy. The diagnosis of attention
deficit disorder predicted a preferential response to propranolol while diagnosis of
intermittent explosive disorder predicted favorable response to carbamazepine. It
should be noted that the fact that all patients with epilepsy (majority with EEG
abnormalities) did not allow the examination of the predictive value of EEG
abnormality in predicting favorable response to carbamazepine.
On the other hand, other studies suggest that anticonvulsant therapy may have a
beneficial effect on aggressive tendencies irrespective of the presence or absence
of EEG abnormalities (Luchins 1984). Until definitive studies are performed
patients should be given the benefit of the doubt and a trial of anticonvulsant
should be performed when an EEG proves to be abnormal, particularly focally and
paroxysmally.

Supported Findings

The above-reviewed literature supports the following conclusions.

(1) EEG as well as other cerebral abnormalities (particularly neuropsychological)
are prevalent in populations that exhibit habitual aggressive behavior. We
should emphasize that the absence of identifiable cerebral abnormalities does
not necessary mean that one does not exist. The lack of identifiable surface
EEG changes may not necessarily reflect lack of epileptic activity in deeper
brain structures. Heath (1992) demonstrated the presence of spike and slow-
wave activity in the hippocampus and medial amygdala in nonepileptic
92 9 Violence Aggression and Impulse Dyscontrol

patients during rageful behavior without corresponding changes at the surface

scalp recording. These observations are very important and underline first the
need for careful evaluations as well as the need for more research on new
methodology to detect deep structures epileptic activity.
(2) We should also emphasize that the presence of an abnormality does not
necessary mean that the identified abnormality is fully responsible for the
behavior emitted. Violence is a biopsychosocial problem and nonbiological
determinants are also likely to play important roles in facilitating the emer-
gence of this behavior. The value of identifying any biological factors that
may have contributed to the behavior should be very useful in designing
treatment and rehabilitative plans for the individual subject.
(3) Such abnormalities could have significant treatment implications. Where the
investigators looked at the history of treatment in these populations it has been
the observation that most of the subjects did repeatedly seek help for their
behavioral problems and usually to no avail. This apparent lack of interest on
part of the health profession most likely reflects the lack of awareness that a
sizeable portion of this population have organic pathology that may be ame-
nable to cost effective treatment approaches. Large multicenter studies
examining the clinical correlates, particularly treatment responses, of patients
with various abnormalities are yet to be conducted.
(4) Finally, a number of studies documented the fact that epileptic activity
occurring in deep brain structures (e.g., limbic structures) are rarely reflected
on surface EEG recordings. Developing technology that can detect such
activity would significantly improve our diagnostic ability of this complex
biopsychosocial problem.

Open Research Questions

(1) Better definition of rate of EEG abnormalities and clear definition of the types
and severity of the deviations.
(2) More exact definition of the clinical-electrophysiological correlations.
(3) Examination in large and multicenter double-blind controlled studies of the
effects of various treatment modalities including various antiepileptic agents,
neuro feedback, and rTMS.
(4) Further definition of the interrelationship between the degree of violence and
severity of EEG deviations.
(5) Further definition of the interrelationship between the nature of the violence
and the nature of the abnormality (focal vs diffuse slowing and focal IEDs in
different brain locations.
References 93

References

Aird RB, Gastaut Y (1959) Occipital and posterior electroencephalographic rhythms. Electro-
encephalogr Clin Neurophysiol 11:637–656
Bach-Y-Rita G, Lion JR, Climent CE, Ervin FR (1971) Episodic dyscontrol: a study of 130
violent patients. Am J Psychiatry 127:473–1478
Barber JW, Hundley P, Kellogg E, Glick JL, Godleski L, Kerler R, Vieweg WVR (1988) Clinical
and demographic characteristics of 15 patients with repetitively assaultive behavior. Psychiatr
Q 59(3):213–224
Bars DR, Heyrend FL, Simpson CD, Munger JC (2000) Use of visual evoked-potential studies
and EEG data to classify aggressive explosive behavior of youths. Psychiatr Serv 52:81–86
Bender L (1959) Children and adolescents who have killed. Am J Psychiat 116:510–513
Bennett WG, Korein J, Kalmijn M, Grega DM, Campbell M (1983) Electroencephalogram and
treatment of hospitalized aggressive children with haloperidol and lithium. Biol Psychiat
18:1427–1440
Blake PY, Pincus JH, Buckner C (1995) Neurologic abnormalities in murderers. Neurology
45:1641–1647
Boelhouwer C, Henry C, Glueck BC Jr (1968) Positive spiking :a double-blind control study on
its significance in behavior disorders, both diagnostically and therapeutically. Am J Psychiat
125:473–480
Convit A, Czobor P, Volavka J (1991) Lateralized abnormality in the EEG of persistently violent
psychiatric inpatients. Biol Psychiat 30:363–370
Fenton GW, Tennet TG, Fenwick PBC, Rattray N (1974) The EEG in antisocial behavior: a study
of posterior temporal slow activity in special hospital patients. Psychol Med 4:181–186
Hakola HP, Laulumaa VA (1982) Carbamazepine in the treatment of violent schizophrenics.
Lancet 1:1358
Harper MA, Morris M, Bleyerveld J (1972) The significance of an abnormal EEG in
psychopathic personalities. Aust NZ J Psychiat 6:215–224
Heath RG (1992) Correlation of brain activity with emotion: a basis for developing treatment of
violent-aggressive behavior. J Am Acad Psychoanal 20(3):335–346
Hill D (1944) Cerebral dysrhythmia: its significance in aggressive behavior. Proc R Soc Med
37:317–328
Hill D (1952) EEG in episodic psychotic and psychopathic behavior. Electroencephalogr Clin
Neurophysiol 4:419–442
Hill D, Watterson D (1942) Electroencephalographic studies of psychopathic personalities.
J Neurol Psychiat 5:47–65
Howard RC (1984) The clinical EEG and personality in mentally abnormal offenders. Psychol
Med 14:569–580
Luchins DJ (1984) Carbamazepine in violent non-epileptic schizophrenics. Psychopharmacol
Bull 20:569–571
Mattes JA (1990) Comparative effectiveness of carbamazepine and propranolol for rage
outbursts. J Neuropsychiat Clin Neurosci 2:159–164
Monroe RR (1975) Anticonvulsants in the treatment of aggression. J Nerv Mental Dis
160:119–126
Monroe RR (1986) Treating atypical psychiatric disorders with anticonvulsants. Md Med J
35(9):757–761
Monroe RR (1989) Dyscontrol syndrome: long-term follow-up. Comp Psychiatry 30(6):489–497
Nelson D, Boutros N (1993) The organic personality disorder. Integr Psychiatry 9:140–144
Neppe VM (1983) Carbamazepine as adjunctive treatment in nonepileptic chronic inpatients with
EEG temporal lobe abnormalities. J Clin Psychiat 44:326–331
Okasha A, Sadek A, Abdel Moneim S (1975) Psychosocial and electroencephalographic studies
of Egyptian murderers. Brit J Psychiat 126:34–40
94 9 Violence Aggression and Impulse Dyscontrol

Pillmann F, Rohde A, Ullrich S, Draba S, Sannemuller U, Marneros A (1999) Violence, criminal

behavior, and the EEG: significance of left hemispheric focal abnormalties. J Neuropsychiat
Clin Neurosci 1:454–457
Reeves RR, Struve FA, Patrick G (2003) EEG does not predict response to valproate treatment of
aggression in patients with borderline and antisocial personality disorders. Clin Electroen-
cephalogr 34(2):84–86
Rey JH, Pond DA, Evans CC (1949) Clinical and electroencephalographic studies of temporal
lobe function. Proceed Royal Soc Med 42:891–904
Riley TL (1979) The Electroencephalogram in patients with rage attacks or episodic violent
behavior. Mil Med 144(8):515–517
Riley T, Niedermeyer E (1978) Rage attacks and episodic violent behavior: electroencephalo-
graphic findings and general considerations. Clin Electroencephalogr 9:131–139
Sayed ZA, Lewis SA, Brittain RP (1969) An electroencephalographic and psychiatric study of
thirty-two insane murderers. Brit J Psychiat 115:1115–1124
Schiff HB, Sabin TD, Geller A, Alexander L, Mark V (1982) Lithium in aggressive behavior. Am
J Psychiatry 139(10):1346–1348
Stafford-Clark D, Taylor FH (1949) Clinical and electro-encephalographic studies of criminals
charged with murder. J Neurol Neurosurg Psychiat 12(4):325–330
Stone JL, McDaniel KD, Hughes JR, Hermann BP (1986) Episodic dyscontrol disorder and
paroxysmal EEG abnormalities, successful treatment with carbamazepine. Biol Psychiatry
21(2):208–212
Tunks ER, Dermer SW (1977) Carbamazepine in the dyscontrol syndrome associated with limbic
system dysfunction. J Nerv Ment Dis 164:56–63
Williams D (1969) Neural factors related to habitual aggression. Consideration of differences
between those habitual aggressives and others who have committed crimes or violence. Brain
92:503–520
Wong MTH, Lumsden J, Fenton GW et al (1994) Electroencephalography, computed
tomography and violence ratings of male patients in a maximum-security mental hospital.
Acta Psychiatr Scand 90:97–101
Yassa R, Dupont B (1983) Carbamazepine in the treatment of aggressive behavior in
schizophrenic patients: a case report. Can J Psychiat 28:566–568
Chapter 10
Borderline Personality Disorder

Introduction

Accumulating electrophysiology evidence suggests that Borderline Personality

Disorder (BPD) patients, similar to many other psychiatric groups of patients, may
be a heterogeneous group (Boutros et al. 2003). The thorough understanding of the
neurobiology of BPD as well as defining its subtypes are essential steps for the
eventual development of effective preventive, therapeutic, and rehabilitative
approaches. While BPD is one of the most investigated personality disorders, the
neurobiological bases of this devastating disorder remain largely unknown. Evi-
dence for an organic basis for some BPD patients has been forthcoming since the
1980s (Andrulonis et al. 1980; Lahmeyer et al. 1989). Efforts to bring rapidly
advancing neuro-investigative technology to bear on the understanding of this
disorder are likely to contribute significantly to the unraveling of the underlying
pathophysiological processes and any possible biological subtypes of this disorder.
Based on available electrophysiology literature three possible subtypes can be
postulated (Reviewed in Boutros et al. 2003). The first postulated subtype may be
related to epilepsy. This is the major focus of this chapter. In a separate set of
reports BPD has been linked to affective disorders (Perry 1985). The high co-
occurrence of depressive symptoms in BPD patients has long been observed.
Indeed, longitudinal studies have found that even cases of apparently pure BPD,
when followed, have frequent suicide attempts and clear-cut affective episodes
(Pope et al. 1983). Steinberg et al. (1997) found an exaggerated depressive
response to physostigmine challenge in BPD patients as compared to patients with
other personality disorders.
In yet a third body of the literature, BPD has been linked to psychotic disorders
(Gunderson et al. 1981; Schulz et al. 1989). The partial clinical response to neu-
roleptics further supports a possible relationship to psychotic disorders (Brinkley
et al. 1979; Solof 1981). Early outcome studies have generally found similarities in
long-term adjustment of patients with BPD with that of schizophrenia patients
(Gunderson et al. 1981; Carpenter and Gunderson 1977). Later follow-up studies
of these patients showed that their long-term outcome is comparable with unipolar

N. N. Boutros, Standard EEG: A Research Roadmap for Neuropsychiatry, 95

DOI: 10.1007/978-3-319-04444-6_10,  Springer International Publishing Switzerland 2013
96 10 Borderline Personality Disorder

patients and overall more favorable than schizophrenia patients (McGlashan 1986;
Grilo et al. 1998).
In a review of available literature, Korzekwa et al. (1993) suggested that BPD is
an independent disorder with at least two possible biological subtypes; an affective
and a psychotic subtypes. Coccaro and Kavoussi (1991) suggested that the iden-
tification of such subtypes can be useful in guiding treatment choices.

Borderline Personality Disorder Links to Epilepsy

A number of electrophysiological studies linked BPD to complex partial seizures

(CPS) (Andrulonis et al. 1980; Muller 1992). Andrulonis et al. found 27 % of
adolescent BPD patients had evidence of brain dysfunction or current epilepsy
(Andrulonis et al. 1982). They also found history of head trauma, encephalitis, or
past seizures in 11 %. Several episodic or paroxysmal symptoms are common
between BPD and temporal lobe epilepsy: impulsivity, transient psychosis, and
intermittent experience of depersonalization and derealization (Fenwick 1981).
Carbamazepine has been shown to be effective in decreasing paroxysmal symp-
toms (Cowdry and Gardner 1988). Indeed, a number of case reports have described
complex partial seizures (CPS) in patients previously diagnosed as BPD (Snyder
and Pitts 1984; Cowdry et al. 1985; Messner 1989; Schmidt et al. 1989).

Standard EEG Abnormalities Reported in BPD Patients

Table 10.1 lists standard EEG studies in BPD patients.

Standard EEG studies have been carried out based on the hypothesis that
abnormal brain electrical activity and/or focal brain dysfunction, particularly in the
temporal lobes, play a significant role in the pathogenesis of BPD characterized by
impulsiveness and affective instability. As mentioned above, a number of case
reports described patients who were diagnosed with BPD who were subsequently
found to have complex partial seizures documented by epileptic discharges over
one or both temporal regions (Cowdry and Gardner 1988), and favorable clinical
response to anticonvulsant medications (Schmidt et al. 1989). As early as the mid-
1980s, the presence of significant EEG abnormalities in BPD patients was well
documented. Snyder and Pitts showed that patients with BPD have a significantly
higher rate of both definitive and less definitive EEG abnormalities when com-
pared to a group with dysthymic disorder (Snyder and Pitts 1984). In this report,
older BPD patients had more severe EEG abnormalities. Abnormalities (mainly
slowing) were most frequently bilateral and of frontal, temporal, or fronto-tem-
poral distribution. Similarly, in a study conducted as a collaboration between the
NIMH and Yale University Cowdry et al. (1985) examined the symptomatology
and EEG changes in 39 BPD patients. BPD patients showed a much higher
Table 10.1 EEG studies in BPD
Author (ref #) N (M/ Diagnostic Medications Control groups Co-morbidity Findings
F) system
Axis-I Axis-II
Introduction

Snyder and 37 (37/ DSM-IIIa None Dysthymia None None Increased slow-wave activity
Pitts (1984) 0)
(7)
Cowdry 39 (3/ DSM-IIIa NR Unipolar depression None current NR Posterior sharp waves
et al.(1985) 36)
(8)
Cornelius et al. 69 (17/ DIB None Non-BPD Axis-II None current None 5.8 % with severe EEG abnormalities as
(1986) (24) 52) compared to 0 % in a non-BPD AXIS-
II control group
Messner (1989) 1(1/0) DSM-IIIa None None None None Focal temporal lobe slowing
(9)
Archer et al. 16 DSM-IIIa None (1) Non-BPD Axis- None None 6.3 % had spike and wave discharges. Not
(1988) (23) II (2) Dythymia significantly higher than the other three
(3) mixed groups. No normal control group
diagnoses
Schmidt et al. 1(0/1) DSM-IIIa Antidepressant and None Depression None Routine EEG normal. Quantified EEG
(1989) (10) antipsychotic showed marked asymmetry
Drake et al. 6 MMPI NR None NR NR Normal EEGs in BPD patients with
(1992) (28) pseudoseizures
Ogiso et al. 18 (0/ DIB DSM- Anxiolytics, Non-BPD with Major NR Positive spikes and spike wave phantoms
(1993) (25) 18) III Antipsychotics, Axis-I disorders depression
Antidepressants and
substance
abuse
(continued)
97
Table 10.1 (continued)
98

Author (ref #) N (M/ Diagnostic Medications Control groups Co-morbidity Findings

F) system
Axis-I Axis-II
Dela Fuente 20 (6/ DSM-IIIR None None None current NR 40 % with diffuse slowing
et al. (1998) 14) DIB
(29)
DIB Diagnostic Interview for Borderline personality disorder. Gunderson score of at least seven
HAMD-Hamilton Depression Rating Scale
a
DSM-III: Indicates diagnoses based solely on clinical interviews without use of structured or semi-structured interview instruments
NR Not reported
10
Borderline Personality Disorder
Introduction 99

incidence of symptoms commonly seen in complex partial seizures or episodic

dyscontrol than a control group of unipolar depressed patients. Similarly, BPD
patients showed a much higher incidence of paroxysmal EEG activity particularly
posterior sharp waves. Archer et al. (1988) found 6.3 % of a group of 16 ado-
lescents with BPD to have bilateral spike and wave discharges while none of 10
subjects in a comparison group of other personality disorders had similar dis-
charges. Furthermore, minor abnormalities that are not suggestive of epilepsy but
may be contributing to episodic behavior (e.g., 14 and 6 positive spikes) were
found in 25 % of BPD patients, 30 % of other personality disorders and 18 % of
dysthymic patients. These percentages are worth contemplating and seriously
considering in view of the rarity of the 14 and 6 positive spikes in healthy adults.
When focusing on severe abnormalities only, higher incidence was found in the
BPD as compared to other personality disorders (Cornelius et al. 1986). They did
not include a normal control group.
Ogiso et al. (1993) provided evidence of different EEG correlates of BPD
symptom-profiles. Using subscales of the Diagnostic Interview for Borderline
(DIB) they found the presence of positive spikes (Chap. 19 for more detail) to
correlate with impulsivity, and the presence of spike and wave discharges to
correlate with high scores of interpersonal relationship dysfunction. Ogiso et al.
(1993) emphasized the observation that no one pattern characterized the entire
sample and that in some patients the EEG was perfectly normal. Similarly, Drake
et al. (1992) found BPD patients who presented with psychogenic seizures to have
normal EEGs.
On the other hand, De la Fuente et al. (1998) reported a 40 % incidence of
diffuse EEG slowing in a group of unmedicated BPD patients. None of their
patients received neuroleptic drugs for at least 2 months, and all other medications
were withdrawn for 10 days with 15 days for tricyclic antidepressants and
monoamine-oxidase inhibitors. Presence of active Axis-I comorbid disorders
including current drug abuse (as verified via repeated plasma testing) were ruled
out. Additionally, any history of neurological problems including seizures was
grounds for excluding subjects. Carbamazepine did not appear to modify the EEGs
of this patient sample. The finding of a high prevalence of slow-wave abnormal-
ities in BPD patients is in support of an earlier report by Tanahashi (1988). Based
on the English abstract of the paper (appearing in Japanese), Tanahashi (1988)
compared the EEG findings between patients with BPD and patients with
schizophrenia. He found a significantly higher incidence of slowing of background
activity among the BPD group (84.4 % of BPD versus 32 % of schizophrenia
patients). He also reported spike and wave complexes (of the 6/s variety) in 31 %
of BPD versus 4 % of schizophrenia patients. We have previously reported that the
presence of diffuse EEG slowing was correlated with the overall severity of
psychopathology (Boutros 1997).
100 10 Borderline Personality Disorder

Electrophysiological Profiles

We found two review articles addressing the contribution of the different elec-
trophysiological modalities to the study of BPD (Lahmeyer et al. 1989; Korzekwa
et al. 1993). As we summarized above, some literature linked BPD to temporal
lobe epilepsy (sEEG literature), a body of literature used the REM sleep changes to
link BPD to mood disorders, and a third set of studies used the event-related
potentials (particularly the P300 ERP) to link BPD to psychotic or thought dis-
orders (Boutros et al. 2003). Lahmeyer et al. (1989) concluded that sleep in general
and rapid eye movement (REM) sleep, in particular, are abnormal in BPD patients
as a group. They further suggested that several factors, including concurrent Axis-I
affective psychopathology, family psychopathology, and a personal past history of
depression predict REM sleep abnormalities, particularly shortness of REM
latency, in this group. Based on the limited number of evoked potentials available
at the time they concluded that a link between BPD and psychosis is likely to exist.
Korzekwa et al. (1993) concurred with Lahmeyer et al. that the consensus appears
to be that BPD have abnormal REM sleep, but this is most pronounced when
depression co-existed with BPD. They also agreed with Lahmeyer et al. (1989)
that evoked potential studies suggest a link between BPD and schizophrenia. Both
articles did not specifically focus on electrophysiological measures, did not
examine the composition of the groups studied, report on sample sizes, comor-
bidity, medications, or diagnostic systems used. Moreover, both articles did not
include reviews of clinical EEG data in this population.
Two conclusions can be reached based on the above review. The first con-
clusion is that electrophysiological investigation of BPD remains extremely lim-
ited. The available literature does not allow the development of specific
hypotheses regarding etiology, pathophysiology, or phenotypic expressions. This
conclusion is not unexpected given the small number of studies, the complexity
(and possible heterogeneity) of the disorder, the extreme problem with comor-
bidity both on Axis-I and Axis-II, the evolving diagnostic criteria, as well as the
confounding effects of pharmacotherapy.

Quantitative EEG

Finally, a number of other physiological measures were applied to the examination

of the BPD. The literature on these measures is extremely sparse but the data
generated were of interest. Cornelius et al. (1988) utilizing EEG spectral analysis
found no correlation between EEG spectra and depressive or transient psychotic
symptoms. They, on the other hand, found the mean frequency values to consis-
tently correlate with anxiety levels. Utilizing power spectral technology, Russ
et al. (1999) found theta activity to significantly correlate with pain ratings in BPD
patients with and without self-injurious behavior. Power spectral and high-density
Introduction 101

EEG technology allows the examination of more subtle EEG changes as compared
to standard EEG.
Finally, a very important aspect of BPD is the serious correlation with child-
hood abuse. Child abuse has been shown to result in clinical EEG (Rosenberg et al.
2000), evoked potentials (EPs), as well as quantified EEG and EEG coherence
abnormalities (Teicher et al. 1997; Ito et al. 1998). Similarly, child abuse has been
shown to adversely affect sleep (Glod et al. 1997). Child abuse has been strongly
linked to the development of BPD (Herman et al. 1989; Zanarini et al. 1997;
Figueroa and Silk 1997). Electrophysiological abnormalities were also reported in
association with other stress-related disorders like Posttraumatic Stress Disorder
(Metzger et al. 1997; McFarlane et al. 1993). More complete characterization of
such abnormalities may also shed more light on the interrelationships between
these variables.

Supported Findings

(1) This literature, as it exists, suggests that two types of standard EEG abnor-
malities may exist in this group of patients. First is the presence of epileptiform
discharges. This type of abnormality is likely to indicate decreased threshold for
seizure like activity or increased cortical excitability and may be predictive of
responsiveness to anticonvulsant therapy (Monroe 1975). The second type of a
standard EEG abnormality is the presence of diffuse EEG slowing. The presence of
diffuse slowing in unmedicated subjects indicates the presence of either a meta-
bolic or a degenerative brain disorder. Patients with mental retardation could also
exhibit diffuse slowing of the EEG. The presence of this abnormality should lead
to further work-up of the patient to identify causes of encephalopathy. The pres-
ence of a static (nonprogressive) and nonmetabolic-based diffuse EEG slowing
could be indicative of a more difficult group of patients who are less likely to
respond to pharmacotherapy (Boutros 1997). (2) Early EEG studies have under-
scored the fact that the clinical correlates of EEG abnormalities in BPD is unlike
the more straight forward correlations that can be seen in epilepsy patients and that
factor analyses of symptom-clusters are necessary to examine such correlates
(Archer et al. 1988; Cornelius et al. 1988).
(3) Only two studies reported controlling for Axis-I or Axis-II comorbidity. It is
not clear from the papers how careful these evaluations were. Furthermore, only
two studies included non-BPD personality-disorders control groups. Control for
medication-effects was reported in four of the seven papers reviewed in Table 10.1
(excluding the two case reports; Messner 1989; Schmidt et al. 1989).
Studies assessing the prognostic value of physiological changes are similarly
lacking. Electrophysiological investigations of BPD have the potential for con-
tributing to our understanding of the different pathophysiological processes that
may be aberrant in BPD patients. This is suggested by the findings in the reviewed
studies. Specifically, standard EEG studies can be useful in probing the
102 10 Borderline Personality Disorder

relationship of BPD to complex partial seizures, sleep studies can help probe the
relationship to mood disorders, and EPs can help elucidate BPD’s commonalties
with psychotic disorders.
In a single study more than one modality were applied in the same subjects
(Lahmeyer et al. 1989). Findings from this study suggest that the various elec-
trophysiological abnormalities (i.e., sEEG, REM sleep, and ERP) may not co-exist
in the same subjects and they may be indicative of different subtypes. In addition,
in none of the studies reviewed has the entire set of electrophysiological measures
been recorded in the same set of patients. Curiously, over the last several years,
there have been far fewer brain electrophysiological studies of patients with BPD.
However, over the same time frame, electrophysiological recording technology
has improved significantly. We now have the ability to study various components
of brain activity from a large number of scalp locations simultaneously. This
should improve the capacity to reliably detect brain abnormalities in this popu-
lation of patients. A comprehensive examination of the electrophysiological pro-
files of BPD patients and correlation with symptom clusters is likely to yield useful
information regarding both subtypes and treatment avenues for these patients. This
suggestion is also supported by the earlier conclusion by Lahmeyer et al. (1989)
suggesting that the use of a battery of tests could help define subgroups within this
disorder. The application of electrophysiological test batteries is more likely to be
useful than the application of a single test (Boutros et al. 1997), except when a
very specific hypothesis is being tested.

Open Research Questions

(1) Future electrophysiological investigations of BPD should strive to combine

the different test modalities available, and provide clinical rating scales
capable of elucidating the entire array of symptomatology exhibited by BPD
patients.
(2) Particular attention should be paid to the problem of comorbidity. Studies should
be explicit regarding the training and qualifications of the personnel involved in
the clinical characterization of study subjects. Specifically, such personnel
should be trained to a standard reliability criterion. Such designs will allow the
examination of any correlations between biological deviations and symptom
clusters. Studies should also include both healthy and patient control groups.
Furthermore, and based on the questions asked, researchers may consider
including Axis-II patient control groups. If medications are not withdrawn,
patient control group should additionally be matched for pharmacotherapy.
(3) Only a few studies utilized q-EEG technology to examine BPD. These studies
demonstrated the potential value of this technology in probing the EEG cor-
relates of the different symptom-dimensions of BPD patients (Cornelius et al.
1988; Russ et al. 1999).
Open Research Questions 103

(4) As is being asked relative to other disorders, does the presence of IEDs predict
a favorable response to AEDs? If so, does the chemical class of the AED and
dose/blood level play a role in determining responsiveness?
(5) Are there correlations between the history and nature of child abuse and the
various electrophysiological aberrations reported in this population?
(6) What are the clinical, diagnostic, and treatment implications of the presence of
diffuse EEG slowing when not medication induced?
(7) As is asked with other psychiatric disorders, particularly childhood disorders,
is there an increased incidence of the so-called controversial waveforms in
BPD? If so which ones and what are the implications for treatment and for
long-term prognosis?

References

Andrulonis PA, Glueck BC, Stroebel CF (1980) Organic brain dysfunction and the borderline
syndrome. Psychiat Clin N Am 4:47–66
Andrulonis PA, Glueck BC, Stroebel CF et al (1982) Borderline personality subcategories.
J Merv Ment dis 170:670
Archer RP, Struve FA, Ball JD, Gordon RA (1988) EEG in borderline personality disorder. Biol
Psychiatry 24:731–732
Boutros NN (1997) Diffuse electroencephalogram slowing in psychiatric patients: a preliminary
report. J Psychiatry Neurosci 21:259–263
Boutros N, Nasrallah H, Leighty R et al (1997) The mid-latency auditory evoked responses
clinical vs. research applications. Psychiatry Res 69:183–195
Boutros NN, Torello M, McGlashan TH (2003) Electrophysiological aberrations in borderline
personality disorder: state of the evidence. J Neuropsychiatry Clin Neurosci 15(2):145–154
Brinkley JR, Beitman BD, Freidel RO (1979) Low dose neuroleptic regiment in the treatment of
borderline patients. Arch Gen Psych 36:319–326
Carpenter WT, Gunderson JG (1977) Five-year follow-up comparison of borderline and
schizophrenia patients. Compr Psychiatry 18:567–571
Coccaro EF, Kavoussi JR (1991) Biological and pharmacological aspects of borderline
personality disorder. Hosp Comm Psychiatry 42:1029–1033
Cornelius JR, Brenner RP, Soloff PH (1986) EEG abnormalities in borderline personality
disorder: specific or non-specific. Biol Psych 21:977–980
Cornelius JR, Schulz C, Brenner RP (1988) Changes in EEG mean frequency associated with
anxiety and with amphetamine challenge in BPD. Biol Psychiatry 24:587–594
Cowdry RW, Pickar D, Davies R (1985) Symptoms and EEG findings in the borderline
syndrome. Intl. J. Psychiatry Med 15:201–211
Cowdry RW, Gardner DL (1988) Pharmacotherapy of borderline personality disorder. Arch Gen
Psychiatry 45:111–119
De la Fuente JM, Tugendhaft P, Mavroudakis N (1998) Electroencephalographic abnormalities in
borderline personality disorder. Psychiatry Res 77:131–138
Drake ME, Pakalanis A, Phillips BB (1992) Neuropsychological and psychiatric correlates of
intractable pseudoseizures. Seizure 1(1):3–11
Fenwick P (1981) EEG studies. In: Reynolds EH, Trimble MR (eds) Epilepsy and psychiatry.
Churchill Livingstone, New York
Figueroa E, Silk KR (1997) Biological implications of childhood sexual abuse in borderline
personality disorder. J Pers Disord 11:71–92
104 10 Borderline Personality Disorder

Glod CA, Teicher MH, Hartman CR, Harakal T (1997) Increased nocturnal activity and impaired
sleep maintenance in abused children. J Am Acad Child Adol Psych 36:1236–1243
Grilo CM, McGlashan TH, Oldham JM (1998) Course and stability of personality disorders.
J Prac Psych Behav Hlth 4:61–75
Gunderson JG, Kolb JE, Austin V (1981) The diagnostic interview for borderline patients. Am J
Psychiatry 138:896–903
Herman JL, Perry JC, Van der Kolk BA (1989) Childhood trauma in borderline personality
disorder. Am J Psychiat 146:490–495
Ito Y, Teicher MH, Glod CA, Ackerman E (1998) Preliminary evidence for aberrant cortical
development in abused children: a quantitative EEG study. J Neuropsychiat Clin Neurosci
10:298–307
Korzekwa M, Links P, Steiner M (1993) Biological markers in borderline personality disorder:
new perspectives. Can J Psychiatry 38(Supplement 1):S11–S15
Lahmeyer HW, Reynolds CF, Kupfer DJ et al (1989) Biologic markers in personality disorder: a
review. J Clin Psychiatry 50:217–225
McFarlane AC, Weber DL, Clark CR (1993) Abnormal stimulus processing in post-traumatic
stress disorder. Biol Psychiat 34:311–320
McGlashan TH (1986) The chestnut lodge follow-up study, III: long-term outcome of borderline
personalities. Arch Gen Psychiatry 43:20–30
Messner E (1989) Covert complex partial seizures in psychotherapy. Am J Orthopsychiatry
56:323–326
Metzger LJ, Orr SP, Lasko NB et al (1997) Evidence for diminished P3 amplitude in PTSD. Ann
N Y Acad Sci 821:499–503
Monroe RR (1975) Anticonvulsants in the treatment of aggression. J Nerv Ment Dis 160:119–126
Muller RJ (1992) Is there a neural basis for borderline splitting? Comp Psychiatry 33(2):92–104
Ogiso Y, Moriya N, Ikuta N et al (1993) Relationship between clinical symptoms and EEG
findings in borderline personality. Jap J Psych Neurol 47(1):37–46
Perry JC (1985) Depression in borderline personality disorder: lifetime prevalence at interview
and longitudinal course of symptoms. Am J Psychiatry 142:15
Pope HG Jr, Jonas JM, Hudson JI et al (1983) The validity of DSM-III borderline personality
disorder. Arch Gen Psych 40:23–30
Rosenberg HJ, Rosenberg SD, Williamson PD et al (2000) A comparative study of trauma and
posttraumatic stress disorder prevalence in epilepsy patients and psychogenic nonepileptic
seizure patients. Epilepsia 41:447–452
Russ JM, Campbell SS, Kakuma T et al (1999) EEG theta activity and pain insensitivity in self-
injurious patients. Psych Res 89:201–214
Schulz PM, Soloff PH, Kelly T et al (1989) A family study of borderline subtypes. J Pers
Disorders 3:217–229
Solof PH (1981) Pharmacology of the borderline disorder. Compr Psychiatry 22:535–543
Snyder S, Pitts WM Jr (1984) Electroencephalography of DSM-III borderline personality
disorder. Acta Psychiatr Scand 69:129–134
Schmidt PM, Handleman MJ, Bidder TG (1989) Seizure disorder misdiagnosed as borderline
syndrome. Am J Psychiatry 146:400–401
Steinberg BJ, Trestman R, Mitropoulou V et al (1997) Depressive response to physostigmine
challenge in borderline personality disorder patients. Neuropsychopharmacology 17:264–273
Tanahashi Y (1988) Electroencephalographic studies of borderline personality disorder.
Juntendoigaku 34:207–219 (In Japanese)
Teicher MH, Ito Y, Glod CA et al (1997) Preliminary evidence for abnormal cortical
development in physically and sexually abused children using EEG coherence and MRI. Ann
N Y Acad Sci 821:160–175
Zanarini MC, Williams AA, Lewis RE et al (1997) Reported pathological childhood experiences
associated with the development of borderline personality disorder. Am J Psych
154:1101–1106
Chapter 11
Psychotic and Affective Disorders

Introduction

Approximately 64–68 % of EEGs in psychiatric patients can provide evidence of

abnormal electrical activity (Hughes and John 1999). The issue is finding out what
such deviations mean diagnostically and therapeutically. Other than an abnormality
pointing to a medical condition like epilepsy or encephalopathy, most of the EEG
changes described in both schizophrenia and affective spectrum disorders do not
carry specific diagnostic value by today’s classification systems. Hence, the elu-
cidation and characterization of sEEG abnormalities in both psychotic and affective
disorders remain rudimentary and in need of much more systematized effort. As has
now been mentioned few times, much work exploring EEG deviations in these
disorders utilizing the power of computer analytic technology has been published
and continues to expand. The focus of the current chapter is on sEEG deviations and
possible significance as well as work that remains to be performed.

SEEG in Psychotic Disorders

Psychoses are abnormal conditions of the mind with a loss of contact with reality.
Major categories of symptoms include; positive symptoms (e.g., hallucinations and
delusions), negative symptoms (e.g., affective limitation and decreased initiative),
and cognitive symptoms (i.e., formal thought disorders). These symptoms are
unspecific and in addition to schizophrenia or severe mood disorders, can be
caused by various general medical conditions including substance abuse, neuro-
logical, or metabolic conditions.
The main role for sEEG as of the writing of this book has been the exclusion of
a general medical condition (specifically a neurological condition) contributing to
the symptoms. This role has been extensively covered in many excellent reviews
and textbooks. The purpose of the current chapter is to assess whether the sEEG
has a role in the differential diagnosis within functional psychiatric disorders or in
predicting treatment response in these conditions.

N. N. Boutros, Standard EEG: A Research Roadmap for Neuropsychiatry, 105

DOI: 10.1007/978-3-319-04444-6_11,  Springer International Publishing Switzerland 2013
106 11 Psychotic and Affective Disorders

SEEG in Schizophrenia

As was stated much earlier by Davis and Davis (1939) ‘‘Although the psychotic
individual cannot be recognized by his EEG, nevertheless, as a group the psychotic
individuals have a significantly larger percentage of abnormalities in their EEGs
than do normals’’, this statement remains true today. Nonetheless, sEEG is not
useful to diagnose schizophrenia and findings are not specific for this disorder. So
the main questions remaining are if such abnormalities are indicative of as yet
unidentified subtypes, carry prognostic or therapeutic predictive values in general
or for the individual patient.
In the early 1950s, before the introduction of antipsychotic drugs, five con-
trolled studies (reviewed in Torrey et al. 2002) reported a higher frequency of EEG
deviations in never-treated patients with schizophrenia, as compared with healthy
subjects (23–44 % in patients vs. 7–20 % in controls). Other reviews confirmed a
high frequency of these deviations (Itil 1977; Hughes and John 1999), and
extended the findings reporting the occurrence of EEG abnormalities in patients
with chronic schizophrenia, with overall reported frequency of EEG abnormalities
ranging from 20–60 %.
sEEG deviant patterns most frequently found in patients with schizophrenia
include ‘‘choppiness.’’ Choppiness has been defined as low amplitude, disorga-
nized fast activity, with reduced or absent alpha, and sometimes excess of slow
activity. Slowing of background activity as well as superimposed generalized
slowing is another relatively common pattern in schizophrenia patients. High
amplitude beta waves can also be seen. Finally, and most infrequently, spike and
spike-and-waves patterns can also be seen in schizophrenia patients. Again, as of
the writing of this book, clinical symptoms alone are not predictive of the presence
or absence of any of these patterns.
It is not known how the changing and evolving diagnostic boundaries of
schizophrenia have influenced the nature and incidences of sEEG findings in this
population (Small 1993). The introduction of the DSM criteria changed the
diagnostic boundaries of schizophrenia being more restrictive in consecutive
editions of the criteria. sEEG abnormalities reported in patients with schizophrenia
before the DSM-III probably included those found in affective and perhaps organic
psychoses. The presence of sEEG abnormalities predicted a change in diagnosis
with re-assignment to an affective disorder when applying more restrictive criteria
for schizophrenia. The slowing of alpha frequency was more severe in patients
with schizophrenia than in those with affective disorders, among those who did not
change diagnosis when using more restrictive criteria. A more recent study using
DSM-IV criteria for diagnosis found epileptiform variants (6 per second phantom
spike and wave, 14 and 6 per second cycle positive spikes, and small sharp spikes)
in affective disorders with psychotic features and schizoaffective disorder but not
in schizophrenia (Inui et al. 1998). These patterns are discussed in much detail in
later chapters.
SEEG in Schizophrenia 107

Investigated sEEG abnormalities in early studies were different from those

highlighted by more recent studies (e.g., epileptiform variants were investigated in
recent studies only), which makes findings difficult to compare; the most recent
studies often did not include a healthy comparison group and normative ranges of
the EEG features were poorly defined (Boutros et al. 2005; Shelley et al. 2008).
However, the overall picture seems to indicate that patients with schizophrenia
might present a lower rate of epileptiform abnormalities or variants than those with
an affective psychosis, while having a more severe alpha frequency slowing.
The association of sEEG abnormalities with presence or absence of family
history for schizophrenia remains controversial (Small 1993; Norman et al. 2007).
Moreover, studies which examined this issue did not distinguish between slowing
and epileptic abnormalities and did not characterize the clinical picture of patients
as to the presence of affective or negative symptoms.

SEEG Findings and Outcome

Apart from the slowing of alpha, the association between sEEG abnormalities and
outcome remains controversial. One study examined the clinical response of
eleven schizophrenia patients to carbamazepine in a double-blind placebo cross-
over design (Neppe 1983). All eleven patients had temporal lobe abnormalities on
their EEGs; nine with focal slowing and two with paroxysmal activity. Significant
improvement on carbamazepine compared to placebo (p \ 0.005) was reported. In
addition, the EEG abnormalities were said to also decrease while on carbamaze-
pine. It completely defies explanations why many more similar studies (using
different classes of anticonvulsants and with more uniform EEG abnormalities)
were not performed (Fig. 11.1).
In a subgroup of patients with affective symptomatology, early in the course of
the illness, a good outcome was predicted by normal EEGs and no slowing of the
alpha frequency. In patients without affective symptomatology a better outcome
was predicted by the presence of EEG abnormalities.
Schizophrenia is a heterogeneous disorder in which subgroups of patients
showing more abnormalities might have better outcome. Similar conclusions were
drawn on the basis of structural brain imaging studies (Kirkpatrick and Galderisi
(2008) and Galderisi and Maj (2009)). Whether these EEG and imaging deviations
correlate are also not known.

First-Episode Schizophrenia

First-episode patients with persistent psychosis were found to have more abnormal
sEEG as compared to remitted patients (Manchanda et al. 2005). The two groups
did not differ for severity of negative, positive or affective symptomatology.
108 11 Psychotic and Affective Disorders

Fig. 11.1 sEEG and

response to antipsychotic
drugs. SC, 26 years, paranoid
schizophrenia,
hypersynchronous EEG, and
poor response to
antipsychotic treatment. Eyes
closed EEG

In patients with a first-episode of schizophrenia sEEG abnormalities predicted a

negative outcome (Manchanda et al. 2005, 2008). These data need to be replicated
before any firm conclusions can be reached particularly that no attempts to dif-
ferentiate which kind of abnormalities predicted a poor outcome (i.e., epileptic
potentials or abnormal slowing) were made.

Affective Psychoses

Abnormal EEG-findings can be detected in about 20–40 % of patients suffering

from mood disorders (Taylor and Abrams 1981; Cook et al. 1986; McElroy et al.
1988). Early visual EEG-studies revealed an increase of small sharp spikes (SSS),
6/s spike-and-wave complexes, and positive spikes (especially in patients with
suicidal ideation) to be frequent in patients with mood disorders (Small et al. 1999;
Struve et al. 1977).
The characteristics of what has been termed subictal mood disorders include
brief euphorias, mixed bipolar episodes, brief severe depressive dips with
Affective Psychoses 109

impulsive suicide attempts, compulsive symptoms, irritability and hostile out-

bursts, and marked premenstrual worsening (Himmelhoch 1987). Silberman et al.
(1985) assessed the prevalence of what are termed ‘‘transient sensory, cognitive,
and affective’’ symptoms that are not uncommon in patients with complex partial
seizures. They compared 44 patients with affective illnesses and normal EEGs (or
minimally slow and nonspecific changes), to 37 with complex partial seizures, and
30 control subjects with hypertension. They did report that such symptoms did
occur frequently in patients with affective illnesses. Patients with subictal mood
disorders may also have paradoxical reactions to mood–active drugs (lithium and
antidepressants), with better response to anticonvulsants. A wide range of pleo-
morphic psychiatric conditions respond well to anticonvulsant medications
including patients with episodic dyscontrol (Monroe 1975) as well as patients
suffering from dissociative disorders with features resembling complex partial
seizures (Mesulam 1981). These disorders are characterized by evidence of CNS
disturbance or family history of epilepsy and by mental changes typical of the
interictal phase of temporal lobe epilepsy (Blumer et al. 1988). In the absence of
seizures, these disorders are identified as temporal lobe syndromes (Blumer et al.
1988). The interictal symptoms include mood lability, viscosity, hyperreligiosity,
episodic rage and anger, hyposexuality, and verbosity (Bear and Fedio 1977).
Monroe (1986) reported five cases that he termed ‘‘atypical’’ all with strong
positive responses to treatment with anti-seizure medications. Four of the five
cases had mood disorders (one cyclothymic and three with depressive features).
The fifth subject had explosive behavior. All five patients did not have frank
epileptic discharges on the EEGs but were characterized by episodic symptoms
that tended to be acute in onset, intense anger, fearfulness, and relatively stable
behavioral pattern between episodes. Stoll et al. (1994) suggested that atypicality
of presentation, history of seizures, abnormal EEGs, or history of head injury
predict favorable response to antiseizure medications in mood disorder patients.
Reeves et al. (2001) examined the EEGs of bipolar patients (N = 20) who had a
positive therapeutic response to valproic acid but not to lithium. All subjects had
mixed EEG abnormalities (both slowing and paroxysmal activities). They reported
that significantly more patients with EEG abnormalities responded better to val-
proic acid as compared to lithium responders (N = 20).
Nevertheless, there are no specific alterations in affective psychoses and sEEG-
recordings are particularly important in order to exclude organic brain functional
abnormalities as the underlying cause of affective disorders.

Quantified EEG

Using quantitative EEG-analyses several studies have shown an increase in alpha-

and/or theta-power in patients with depression (Monakhav and Perris 1980; Nieber
and Schleged 1992; Mucci et al. 2006). It has been shown that treatment with
110 11 Psychotic and Affective Disorders

antidepressants can reduce alpha-activity (Saletu et al. 1992). It is fully expected

that QEEG will play a major role in the diagnosis and management of mood
disorders in the not too far future.

Supported Findings

The number of standard EEG (sEEG) studies in schizophrenia has substantially

declined over time, with only a few papers published in the last 10 years, while, as
mentioned earlier, quantitative EEG methods have regained interest in psychiatry.
However,
(1) A high percentage of patients with schizophrenia show sEEG abnormalities
whose correlates are still not understood (Shelley et al. 2008; Boutros et al.
2009).
(2) sEEG abnormalities predict conversion to psychosis in subjects at risk
(Gschwandtner et al. 2009).
(3) The use of sEEG in cohort studies of first-episode patients indicated that EEG
abnormalities might predict a worse outcome (Manchanda et al. 2005, 2008).
(4) Some patients with affective disorders exhibit paroxysmal activity more likely
of the controversial category (see controversial EEG patterns chapters).

Open Research Questions

(1) Does the presence of episodic and transient symptoms predict the presence of
EEG abnormalities or the favorable response to anticonvulsant medications?
(2) Does the detection of a certain sEEG abnormality indicate a certain sybtype
(i.e., can EEG abnormalities be considered endophenotypes)?
(3) Are sEEG deviations state- or trait-dependent?
(4) What is the actual prevalence of each form of EEG abnormality in the various
psychotic and mood disorders?
(5) Does the detection of paroxysmal sEEG abnormalities predict favorable
therapeutic response to anticonvulsant medications?

References

Bear DM, Fedio P (1977) Quantitative analysis of interictal behavior in temporal lobe epilepsy.
Arch Neurol 54:454–467
Blumer D, Heilbronn M, Himmelhoch J (1988) Indications for carbamazepine in mental illness:
atypical psychiatric disorder or temporal lobe syndrome. Compr Psychiatry 5:108–122
References 111

Boutros NN, Mirolo HA, Struve F (2005) Normative data for the unquantified EEG: examination
of adequacy for neuropsychiatric research. J Neuropsychiatry Clin Neurosci 17(1):84–90
Boutros N, Lacono W, Galderisi S (2009) Applied electrophysiology, in comprehensive textbook
of psychiatry. In: Sadock BJ, Sadock VA, Ruiz P (eds). Lippincott Williams and Wilkins,
Philadelphia, pp 211–248
Cook BL, Shukla S, Hoff AL (1986) EEG abnormalities in bipolar affective disorder. J Affect
Disord 11(2):147–149
Davis AP, Davis H (1939) The electroencephalogram of psychotic patients. Am J Psychiatry
95:1007–1025
Galderisi S, Maj M (2009) Deficit schizophrenia: an overview of clinical, biological and
treatment aspects. Eur Psychiatry 24:493–500
Gschwandtner U, Pflueger MO, Semenin V, Gaggiotti M, Riecher-Rössler A, Fuhr P (2009) EEG:
a helpful tool in the prediction of psychosis. Eur Arch Psychiatry Clin Neurosci 259:257–262
Himmelhoch JM (1987) Cerebral dvsrhythmia, substance abuse and the nature of secondary
affective illness. Psychiatr Ann 17:710–727
Hughes JR, John ER (1999) Conventional and quantitative electroencephalography in psychiatry.
J Neuropsychiatry Clin Neurosci 11:190–208
Inui K, Motomura E, Okushima R, Kaige H, Inoue K, Nomura J (1998) Electroencephalographic
findings in patients with DSM-IV mood disorder, schizophrenia, and other psychotic
disorders. Biol Psychiatry 43:69–75
Itil TM (1977) Qualitative and quantitative EEG findings in schizophrenia. Schizophr Bull
3:61–79
Kirkpatrick B, Galderisi S (2008) Deficit schizophrenia: an update. World Psychiatry
7(3):143–147
Manchanda R, Norman RM, Malla AK, Harricharan R, Northcott S (2005) Persistent psychoses
in first episode patients. Schizophr Res 80:113–116
Manchanda R, Norman R, Malla A, Harricharan R, Northcott S (2008) EEG abnormalities and 3-
year outcome in first episode psychosis. Acta Psychiatr Scand 117:277–282
McElroy SL, Keck PE Jr, Pope HG Jr, Hudson JI (1988) Valproate in the treatment of rapid-
cycling bipolar disorder. J Clin Psychopharmacol 8(4):275–279
Mesulam MM (1981) Dissociative states with abnormal temporal lobe EEG. Arch Neurol
38:176–181
Monakhov K, Perris C (1980) Neurophysiological correlates of depressive symptomatology.
Neuropsychobiology 6(5):268–279
Monroe RR (1975) Anticonvulsants in the treatment of aggression. J Nerv Ment Dis 160:119–126
Monroe RR (1986) Treating atypical psychiatric disorders with anticonvulsants. Maryland Med J
35(9):757–761
Mucci A, Volpe U, Merlotti E, Bucci P, Galderisi S (2006) Pharmaco-EEG in psychiatry. Clin
EEG Neurosci 37:81–98
Neppe VM (1983) Carbamazepine as adjunctive treatment in nonepileptic chronic inpatients with
EEG temporal lobe abnormalities. J Clin Psychiatry 44:326–331
Nieber D, Schlegel S (1992) Relationships between psychomotor retardation and EEG power
spectrum in major depression. Neuropsychobiology 25(1):20–23
Norman RM, Manchanda R, Malla AK, Harricharan R, Northcott S (2007) The significance of
family history in first-episode schizophrenia spectrum disorder. J Nerv Ment Dis 195:846–852
Reeves RR, Struve FA, Patrick G (2001) Does EEG predict response to valproate versus lithium
in patients with mania? Ann Clin Psychiatr 13(2):69–73
Saletu B, Grunberger J, Anderer P, Linzmayer L, Semlitsch HV, Magni G (1992) Pharmaco-
dynamics of venlafaxine evaluated by EEG brain mapping, psychometry and psychophys-
iology. Br J Clin Pharmacol 33(6):589–601
Shelley BP, Trimble MR, Boutros NN (2008) Electroencephalographic cerebral dysrhythmic
abnormalities in the trinity of nonepileptic general population, neuropsychiatric, and
neurobehavioral disorders. J Neuropsychiatry Clin Neurosci 20:7–22
112 11 Psychotic and Affective Disorders

Silberman EK, Post RM, Nurnberger J, Theodore W, Boulenger JP (1985) Transient sensory,
cognitive and affective phenomena in affective illness. A comparison with complex partial
epilepsy. Br J Psychiatry 146:81089
Small JG (1993) Psychiatric disorders and EEG, in electroencephalography: basic principles,
clinical applications, and related fields. In: Niedermeyer E, Lopes da Silva F (eds). Williams
and Wilkins, Baltimore, pp 581–596
Small JG, Milstein V, Malloy FW, Medlock CE, Klapper MH (1999) Clinical and quantitative
EEG studies of mania. J Affect Disord 53(3):217–224
Stoll AL, Banov M, Kolbrener M, Mayer PV, Tohen M, Strakowski SM, Castillo J, Suppes T,
Cohen BM (1994) Neurologic factors predict a favorable valproate response in bipolar and
schizoaffective disorders. J Clin Psychopharmacol 14(5):311–313
Struve FA, Saraf KR, Arko RS et al (1977) Relationship between paroxysmal electroenceph-
alographic dysrhythmia and suicide ideation and attempts in psychiatric patients, in
psychopathology and brain dysfunction. In: Shagass C, Gershon S, Friedhoff AJ (eds).
Raven, New York, pp 199–221
Taylor MA, Abrams R (1981) Prediction of treatment response in mania. Arch Gen Psychiatry
38(7):800–803
Torrey EF (2002) Studies of individuals with schizophrenia never treated with antipsychotic
medications: a review. Schizophr Res 58:101–115
Chapter 12
EEG Role in Psychiatric Emergencies

Introduction

This chapter reviews the available literature on possible roles for the standard EEG
in psychiatric emergencies. As of the writing of this book, EEG hardly plays any
role in the diagnostic work-up or the management of patients in psychiatric
emergencies despite a significant literature as reviewed below. The emphasis of
this chapter, as in rest of the book, is to highlight the gaps in our knowledge that
are necessary to support an increased role of the EEG in psychiatric emergencies.
We begin with acute catatonia and then discuss conditions where it is difficult to
assess mental status and end with some discussion of the ambulatory noncon-
vulsive status epilepticus (ANCSE).

Catatonia

Catatonia has been associated with schizophrenia for a long time. Kahlbaum
described a wide range of motor abnormalities, some are classic but infrequent
(e.g., echopraxia, waxy flexibility) while others are common in psychiatric patients
(e.g., agitation, withdrawal) (Rao et al. 2012). Other signs of catatonia include
immobility, mutism, negativism, staring, stereotypy, verbigeration, echolalia,
pasturing, catalepsy, automatic obedience, rigidity, and refusal to eat or drink.
Catatonia may also be due to a nonpsychiatric medical condition (i.e., catatonia
due to a general medical condition (GMC).
Catatonia due to ‘‘GMC’’ is disallowed for an episode that occurs ‘‘exclusively
during the course of delirium.’’ The DSM requisite of assessing consciousness and
attention is problematic in the usually mute catatonic patient. This raises the issue
of the role of the sEEG in assessing patients in acute psychiatric settings presenting
with a difficult to assess mental status (DAMS) (see further down this chapter).
Clinical forms include: acute, chronic, periodic, when course is considered and
excited (better prognosis) and retarded forms, phenomenologically.

N. N. Boutros, Standard EEG: A Research Roadmap for Neuropsychiatry, 113

DOI: 10.1007/978-3-319-04444-6_12,  Springer International Publishing Switzerland 2013
114 12 EEG Role in Psychiatric Emergencies

The frequency of catatonic disorder due to general medical conditions

(CDGMC) may vary with the clinical setting. Rosebush and Mazurek (1996) found
that 9 % of the patients admitted to an adult psychiatric unit presented with this
syndrome. Two-thirds had associated medical conditions. In a study of patients
presenting with catatonia to a neurology ward over a period of 12 years, Barnes
et al. (1986) found that 20 % had catatonic syndromes due to general medical
conditions. Patients with this disorder have also been encountered on medical-
surgical units. However, determining which conditions caused catatonia proved
problematic in both of these studies. Several reports of systematic screening in
psychiatric admissions find an incidence of 7–17 %.
From the above it could be concluded that catatonia is a neuropsychiatric
syndrome, while commonly associated with mood disorders (Bipolar and major
depressive disorder (MDD)) as well as schizophrenia it may in reality present a
difficult diagnostic dilemma. Catatonic disorder due to general medical conditions
must be considered in every patient with catatonic signs since frank neurological
conditions like strokes and hematomas can present with new onset catatonia,
neurological work-up must be considered for such patients. The sEEG can be
useful in attempting to sort out possible etiologies for a particular patient pre-
sentation. As has been stressed previously, EEG is widely available, inexpensive,
and completely noninvasive. None the less, when catatonia is encountered in
psychiatric settings, standard EEG is rarely considered for the work-up.

Case Reports of EEG in Catatonia

Arias et al. (2003) reported a case of a 22-years-old woman admitted with a picture of
catatonic posturing, stupor, fever, rigidity, and seizures. She also had dysautonomic
symptoms (tachycardia and hypertension). CT, MRI, and CSF were normal. EEG
revealed diffuse slow waves and right frontotemporal paroxysmal activity. CK was
found to be elevated. Early in her presentation she did not respond to neuroleptics,
anticholinergics, nor antidepressants. Despite her EEG abnormalities, a course of
ECT was begun and she began to improve during the 19 treatments she received. This
case is illustrative of the possibility of an organic factor contributing to a functional
presentation. The presence of the EEG abnormalities did not preclude the consid-
eration of ECT and stressed that ECT is the most effective treatment option in situ-
ation of malignant catatonia.
Another illustrative case was reported by Swartz et al. (2002), where a
68-years-old man history of depression developed a catatonia-like syndrome.
Catatonic symptoms resolved with the administration of lorazepam. A subsequent
EEG revealed a continuing nonconvulsive status epilepticus (NCSE). Presumably,
lorazepam suppressed seizure in areas where it had caused clouding of con-
sciousness, but did not suppress all seizure activity. When lorazepam was stopped,
the catatonia-like delirium returned. In this case valproic acid was effective in
stopping both catatonic and EEG convulsive activity. This case illustrates the fact
Introduction 115

that reduction of signs of catatonia with benzodiazepine does not necessarily

confirm a diagnosis of functional catatonia.
Primavera et al. (1994) stated that although ictal catatonia as a manifestation of
NCSE has been described, reference to the occurrence of seizures in patients with
acute catatonic syndrome remains anecdotal. The case reported by De Entramba-
saguas et al. (2000), was that of a 24-years-old man presenting with a generalized
tonic–clonic seizure of focal onset. Within the following 48 h, he developed agi-
tation and paranoid ideations, which evolved into a severe catatonic syndrome. CT
and MRI scans as well as CSF, urine, and blood chemistries were all negative. EEG
showed theta and delta waves originating in the right frontal region which later
generalized. He remained in the ICU with severe catatonic stupor for 3 months.
During which time he was receiving a course of ECT. In this case 27 sessions were
necessary to resolve the catatonic syndrome.
Kanemoto and colleagues (1999) described a case presenting with catatonia as a
manifestation of absence status epilepticus following benzodiazepine withdrawal.
This elderly patient had long psychiatric history and developed acute catatonia
upon benzodiazepine withdrawal. EEG recording revealed a continuous state of
3 s spike and wave (absence status). The case suggests that new onset catatonia in
an elderly should be considered secondary until proven otherwise. Earlier, Louis
and Pflaster (1995) reported a case of a 24-years-old woman whose initial
examination was notable for a fixed stare, no response to voice or command, tonic
head posturing, gaze preference, constant stereotypic chewing movements, and
profuse foamy salivation. She responded dramatically to parenteral administration
of benzodiazepine. Based on the clinical examination and presentation patient was
diagnosed as being in a NCSE. EEG, however, was perfectly normal.
Orland and Daghestani (1987) reported a case of acute catatonia in a 51-years-
old male secondary to bacterial meningoencephalitis (EEG showed focal frontal
slowing). Catatonic and EEG changes associate with hyperparathyroidism (Cooper
and Schapira 1973), and acute intermittent porphyria (Arnott et al. 1972) were also
reported. Finally, the idea that catatonia can represent as an ANCSE is not new.
Lim et al. (1986) reported three cases of ictal catatonia. These three patients
responded dramatically to intravenous phenytoin. The EEG showed continuous
bilateral pseudoperiodic sharp waves and spike discharges in one, spike and wave
complexes were prominent on the right frontocentral region in another, and the
EEG of the third patient was dominated by periodic lateralized epileptiform dis-
charges during the catatonic state.
In rare occasions the EEG can point to a specific etiology. This is the case with
subacute sclerosing panencephalitis (SSPE). The EEG has a specific pattern in this
disorder consisting of short bursts of irregular, bilateral sharp, and slow wave
complexes, appearing periodically usually every 5–6 s (Cobb 1966). Koehler and
Jakumeit (1976) reported a case of SSPE presenting as catatonia. The EEG bore
the characteristic pattern of the disorder.
116 12 EEG Role in Psychiatric Emergencies

Case Series and Literature Reviews

The rate of abnormal EEGs in patients presenting with a catatonic syndrome

varied among studies. As early as 1942, it was noted that the rate of abnormal EEG
associated with a catatonic presentation is significant (Rao 2012). Walter reported
two out of six catatonic patients to have abnormal EEGs. Both of these patients
would have been diagnosed as catatonic disorder secondary to hyperthyroidism
using DSM-IV criteria. Abenson (1970) found that minor EEG abnormalities in
patients with catatonic schizophrenia are not uncommon (24 %). In this report, the
author reviewed the EEGs of 210 chronic schizophrenics, 78 diagnosed with
catatonic schizophrenia. Strict EEG criteria were used to evaluate abnormalities,
and patients were off neuroleptic medications for at least 3 months. Excluding one
pattern that was not clearly abnormal (choppy EEG-low voltage fast activity with
no visible alpha rhythm) the rate of abnormality was 12.6 %. This rate is com-
parable to the rate of EEG abnormalities in schizophrenia patients in general. The
clinical significance of these abnormalities in terms of predicting response to
treatment and long-term prognosis has not been investigated.
Suzuki et al. (1994) reported three cases of epileptic seizures superimposed on
psychiatric catatonic stupor, none of them had personal or family history of
neurologic disease and catatonia persisted after resolution of epileptic seizure with
phenytoin. They concluded that because catatonia can be caused by epileptic
seizures, EEG in patients presenting with catatonic stupor is indicated not only to
rule out status epilepticus but also to detect epileptic seizures superimposed on
catatonic stupor.
Rosebush and Mazurek (1996) reported five patients who became catatonic (age
range 53–88 years) following benzodiazepine withdrawal, illustrating that cata-
tonia can develop in the wake of benzodiazepine withdrawal and that the elder are
particularly vulnerable.
Twenty-nine patients with acute catatonic syndrome were reviewed to identify
those with seizure after the onset of catatonia. Patients were divided into four
diagnostic groups: affective (Louis and Pflaster 1995), schizophrenic (de Ent-
rambasaguas et al. 2000), toxic (Arias et al. 2003), and organic (Carrol et al. 1994).
Seizures occurred in four patients (13.8 %): two with dystonic seizure had viral
encephalitis and schizophrenia, one with complex partial seizure had viral
encephalitis, one patient with absence status had neuroleptic malignant syndrome
(NMS). These results indicate value of EEG in detection of epileptic activity in
patients with acute catatonia and to provide differential diagnosis between pseu-
doseizure and neuroleptic-induced acute dystonia.
Gjessing et al. (1967) attempted to correlate the state of catatonic stupor to the
EEG in patients with periodic catatonia. They observed that alpha frequency was
increased while amplitude was decreased in three patients with catatonia. This is
an important observation as it underlines the possibility that catatonia even when
not of an epileptic or encephalopathic etiology may still be attended with EEG
changes. The significance of these changes for the diagnosis and treatment are yet
Introduction 117

Table 12.1 General medical conditions associated with catatonic presentations with number of
subjects reported
Conditiona Number of subjects
Seizures or EEGs with epileptic discharges 25
Focal structural CNS damage 80
Encephalitis or other CNS infections 67
Systemic lupus erythematosus with or without cerebritis 8
Disulfuram 9
Phencyclidine 3
Corticosteroids 6
Generalized metabolic disorders 21
Porphyria 3
Other 39
a
In most of the conditions reported above a direct causal relationship was not proven but implied
and at times only as a factor, among many, contributing to the development of a catatonic
syndrome

to be explored. Rosebush et al. (1990), on the other hand, found most psychiatric
inpatients presenting with catatonia to have normal EEGs.
Carrol et al. (1994) reviewed the existing world literature regarding the various
etiologies of CDGMC and provided four detailed case reports of such cases. In
these four patients, general medical conditions associated with CDGMC included,
dystonia, HIV encephalopathy, encephalitis, and renal failure. The critical litera-
ture review concerning catatonia and associated nonpsychiatric medical conditions
only infrequently supported a causal relationship between a specific organic factor
and the development of catatonia. The majority of patients had multifactorial
etiologies. Table 12.1 lists a general medical condition that has been associated
with a catatonia presentation.
Carrol and Boutros (1995) further explored the nature of EEG abnormalities
and their clinical correlates in psychiatric patients with CDGMC, bipolar disorder
with catatonia, and catatonic schizophrenia. From among 82 episodes of catatonia
(obtained from 67 patients), a total of 42 EEG recordings were available from 26
catatonic episodes. EEGs that did not coincide with a catatonic episode were not
included in this study. There were 15 male and 11 female patients. In 50 % of the
patients a significant GMC was diagnosed: NCSE (3), metabolic encephalopathy
(1), dementia (1), Huntington’s disease (1), neuroleptic malignant syndrome (4),
frontal lobe syndrome (1), hypothermia (1), and CNS tumor (1). Sixteen of the first
26 EEGs were abnormal: diffuse slowing (12), focal slowing (3), bilateral spikes
(1). Patients receiving second or third EEGs were more likely to show abnormal
patterns. This observation raises the question of the value of repeated EEG testing
in catatonic patients. Four of six patients showed diffuse slowing on the second
EEG, and two of four had focal slowing on the third EEG. The presence of EEG
abnormalities was seen more frequently in patients over 40 and patients with more
than one medical condition. Psychiatric diagnosis, psychiatric family history, and
gender were not associated with the presence of EEG abnormalities.
118 12 EEG Role in Psychiatric Emergencies

EEG Findings as State Indicators

Hill (1974) discussed the clinical significance of EEG abnormalities in episodic

catatonia. The abnormal discharges were often paroxysmal patterns, particularly
fast spike-waves, which were closely associated with the catatonic episodes. Hill
hypothesized a process of homeostatic regulation between the catatonic phase and
epileptic convulsion in catatonic disorder. Ando and Ito (1959), in Japan described
same correlation between clinical phase and EEG findings, during the (prestupor)
phase, frequency and number of alpha waves decreased, during (stupor) phase, the
amplitude and frequency of alpha decreased and theta waves appeared, during the
(recovery) phase, the amplitude and number of alpha waves normalized.
Thus, catatonic schizophrenia often shows EEG abnormalities and studies indi-
cate that alteration of alpha and slow waves in relation to clinical phase is a common
EEG pattern; suppression of background activity and decreased alpha are seen
during the peak period of catatonia with resumption of alpha and appearance of slow
waves during the poststupor and recovery phase. The EEG abnormalities in periodic
catatonia could therefore be regarded as state indicators and suggest a role for EEG
monitoring even when the etiology of the catatonic syndrome is judged to be
functional (i.e., not epileptic or secondary to other neurological conditions).

The Patient Presenting with a Difficult to Assess Mental

Status (DAMS)

In the absence of frank catatonic symptoms, a patient may present with a DAMS
precluding the ability of clinicians to confidently establish whether or not the
patient is fully oriented. As delirium must be considered as one of the serious
conditions to be ruled out, one would have predicted the sEEG to play a role in this
evaluation. In fact, the following study from our laboratory at Wayne State Uni-
versity is the only one dealing with this issue in the published literature (Jav-
anbakht et al. 2012).
This study included three groups of subjects: 15 patients with DAMS who were
seen in the psychiatric emergency room, 15 patients oriented to time, place, and
person who were seen in the same psychiatric emergency room psychiatric control
(PC), and 10 healthy controls (HC). Reasons for difficulty in the assessment of
patients’ mental status in the DAMS group were lack of cooperation, mutism,
negativism, psychotic preoccupation, and severe disorganization. Exclusion cri-
teria included ability to make a judgment regarding orientation to person, place,
and time or the presence of a known cause for delirium (e.g., subject intoxicated or
known to be in withdrawal from substance), history of a known neurological
disorder like epilepsy, and history of an acute head trauma. The latter two
exclusion criteria necessitated waiting for recruitment until the patient was cleared
by a neurology consultant. Other exclusion criteria included an identified medical
The Patient Presenting with a Difficult to Assess Mental Status (DAMS) 119

condition which could cause the change in mental status. All subjects had to be
alert. Urine drug screens were not obtained as DAMS patients were by definition
uncooperative. As the aim of this study was the recognition of EEG frequencies
rather than localization of probable abnormal foci, we only used 16 channels for an
abbreviated EEG (AbEEG). The sole purpose of the AbEEG was to identify the
dominant background activity and any superimposed or intermixed rhythms. If a
patient was in a NCSE, a 10–15 min recording would at least suggest the need for
a full standard EEG. Moreover, the shortened length of recording would minimize
the possibility of interference of the study with routine clinical procedures.
Recordings were interpreted off-line by a certified EEG expert who was blinded to
all patient groups at the time of the EEG recording. Interpretations were not
performed immediately as there were no plans to use the data clinically. For each
subject, the slowest and the fastest frequencies were elicited. For statistical anal-
ysis in those with only one dominant frequency, that particular frequency was
considered as the slowest frequency for statistical analysis.
No subject was restrained for the purposes of obtaining the EEG recording.
Although some subjects were in physical restraints for clinical reasons as decided by
the on duty psychiatrist, no individual was put in physical restraint for study pur-
poses. Average minimum frequency for the DAMS group was 8.2 CPS (cycles per
second) compared to 9.5 and 9.95 for the PC and HC groups, respectively. The
average minimum EEG frequency in the DAMS group was significantly lower than
the minimum frequency in both the HC (post hoc two tail: t test, P \ 0.001) and the
PC (P \ 0.02) groups. Four subjects in the DAMS and none in the patient or the
healthy control groups exhibited background rhythms strongly suggestive of a
delirium state.
The preliminary data suggests that clinical examination of a DAMS person in
the psychiatric emergency room may benefit from EEG utilization. Perhaps
equally important is that an abbreviated EEG (shortest was 5 min) were obtainable
from the most uncooperative patients. All obtained EEG tracings allowed a con-
fident evaluation of the background activity. Future studies should also include
predictive validity of EEG findings by following the patients’ clinical course.

Ambulatory Nonconvulsive Status Epilepticus

ANCSE, once thought to be a relatively rare cause of altered mental status or

abnormal behavior, has been described increasingly in the neurology and epilepsy
literature (Riggio 2005). The clinical manifestations of ANCSE are characterized
by a change in behavior that can vary considerably from case to case. These
changes range from a mild departure from baseline, to psychotic or affective states
whether depression or manic like behavior. Little has been written in the psy-
chiatric literature about ANCSE. ANCSE is possibly one of the most frequently
missed diagnoses in patients who have a change in mental status. The diagnosis
may be missed in part because of the broad range of clinical presentations, and in
120 12 EEG Role in Psychiatric Emergencies

part because of limited awareness of this condition (Riggio 2005 for a compre-
hensive review). The hallmark of ANCSE is a change in behavior or mental status
that is associated with diagnostic EEG changes.
There are two main types of NCSE: absence status (AS), which is a primary
generalized process, and complex partial status (CPS), which is focal in origin. In a
prospective study of 198 patients, who had altered consciousness, some were
unresponsive and hence cannot be called ‘‘Ambulatory’’ but no clinical convul-
sions who were referred for emergency EEG. Privitera and Strawsburg (1994)
reported that 37 % showed EEG and clinical evidence of NCSE. NCSE has been
reported in all age groups from the very young to the very old without a clear
gender predominance (Flor-Henry 1969; Husain et al. 2003). Anywhere from 10 to
100 % of patients who present with NCSE do not have a history of seizure disorder
(Niedermeyer and Khalifeh 1965; Lee 1985). Absence status (AS) is a heteroge-
nous epileptic syndrome that can occur at any age, usually but not necessarily in a
context of prior epilepsy. Eleven cases of AS occurring in middle-aged patients
who had no history of epilepsy were retrospectively collected over a 10-year
period (10 women and one man; mean age, 58.6 years) (Thomas et al. 1992). Eight
patients were receiving high doses of psychotropic drugs. Clinical and EEG pre-
sentation was similar to AS occurring in patients with prior epilepsy. Evaluation of
precipitating factors revealed that AS coincided with benzodiazepine withdrawal
in eight cases. Cofactors included excessive use of other psychotropic drugs,
nonpsychotropic treatment, hypocalcemia, hyponatremia, and chronic alcoholism.
CT demonstrated mild cerebral atrophy in six cases. There was no recurrence, even
without chronic antiepileptic treatment. These data indicate that (1) most cases of
‘‘de novo’’ AS of middle age or late onset result from the addition of various
epileptogenic factors; (2) AS can be considered a new and uncommon compli-
cation of benzodiazepine withdrawal, and (3) long-term administration of anti-
convulsant medication may not be required (Thomas et al. 1992).
Despite ANCSE being readily diagnosed by the EEG, it remains completely
unknown how common is this problem and how frequently this condition is missed
in acute psychiatric settings. Fifty-two patients having suffered 60 episodes of
NCSE proven by electroencephalography between 1976 and 1986 were reported
by Rohr-Le Floch et al. (1988) from an acute but not a psychiatric setting to
demonstrate the utility of emergency sEEG in making the correct diagnosis.
According to electro clinical criteria, these cases of status epilepticus were clas-
sified into three groups: Petit Mal Status, Psychomotor Status, and Frontal Polar
Status. The exact diagnosis could not be accurately established by the clinical
examination alone. It required the analysis of the ictal EEG. However, some
clinical signs might suggest the correct diagnosis. Thus, a fluctuating confusional
state associated with myoclonus suggested a Petit mal status. A state of confusion
with alteration of the emotional sphere evoked especially a psychomotor status. A
confusional state associated with behavioral disorders of euphoric type and to
redirection difficulties was seen mainly in frontopolar status (Rohr-Le Floch et al.
1988). Furthermore, whether AS or CPS status is more common in psychiatric
settings is also not known in either children or adult patients.
Ambulatory Nonconvulsive Status Epilepticus 121

Supported Observations

(1) Available literature suggests that a patient presenting with catatonia may have
an acute neurological process or catatonia can be a representation of a purely
functional disorder. It seems, at least from the bulk of the literature that once
an organic etiology is ruled out, the routine EEG tends to be normal from a
neurological point of view but may still have spectral or topographical devi-
ations that may have diagnostic or therapeutic implications (Rosebush et al.
1990).
(2) Predicting EEG (or structural/neurological) abnormalities based on the pre-
sentation could be misleading (Patry et al. 2003) and clinicians should err on
the side of safety (again given the relative low cost of the EEG). A major
question remains if this is true even when the patient had a number of prior
presentations with functional catatonia (with documented normal EEGs).
(3) Advanced EEG analysis (i.e., spectral or coherence analysis) studies are sparse
but are likely to yield important information about the pathophysiology of
catatonia in functional disorders. The inescapable conclusion is that obtaining
an EEG in a person presenting with acute catatonia could be extremely
informative to the differential diagnostic process.

Open Research Questions

(1) What is the role of the sEEG and QEEG in the work-up of patients presenting
with catatonia?
(2) Does EEG (standard or quantified) have a role when the catatonia is known to
be of functional origin?
(3) Is there a role for EEG monitoring during the management of patients pre-
senting with catatonia whether of functional or neurological etiologies?
(4) Does the clinical setting (e.g., psychiatric versus nonpsychiatric emergency
settings) affect EEG findings in catatonic patients?
(5) When a patient has a history of functional catatonia should it be automatically
assumed that once functional always functional?
(6) What is the actual frequency of ANCSE in acute psychiatric settings? Are
there clinical indicators that should increase the level of suspicion of the
clinicians? Which of the two main subtypes of ANCSE (CPS or Absence) are
more likely to be encountered in acute psychiatric settings and is there
treatment and prognosis differences between them?
(7) What do the EEG changes noted in nondelirious DAMS patients (i.e., bor-
derline diffuse or generalized slowing of the EEG) tell us about the possible
clinical outcome and prognosis for each case?
122 12 EEG Role in Psychiatric Emergencies

References

Abenson MH (1970) EEGs in chronic schizophrenia. Br J Psychiatry 116:421–425

Ando M, Ito K (1959) Clinical and electroencephalographical studies on catatonia. Folia
Psychiatr Neurol Jpn 13:133–142
Arias M, Paramo M, Requena I, Sesar A, Robelo M, Peleteiro M (2003) Malignant catatonia as
paradigm of neuropsychiatric disease. Neurologia 18:107–111
Arnott G, Lehembre P, Lambert P, Dequiedt F (1972) Acute intermittent porphyria with cerebral
manifestations: generalized convulsions with focal EEG abnormalities in one case, catatonic
state in another case. Lille Med 17:857–862
Barnes MP, Saunders M, Walls TJ, Saunders I, Kirk CA (1986) The syndrome of Karl Ludwig
Kahlbaum. J Neurol Neurosurg Psychiatry 49(9):991–996
Carrol BT, Boutros NN (1995) Clinical electroencephalograms in patients with catatonic
disorders. Clin EEG 26:60–64
Carrol BT, Anfinson TJ, Kennedy JC, Yendrek R, Boutros M, Bilon A (1994) Catatonic disorder
due to general medical condition. J Neuropsychiatry Clin Neurosci 6:122–133
Cobb W (1966) The periodic events of subacute sclerosing leuconcephalitis. Electroencphalogr
Clin Neurophysiol 21:278–294
Cooper AF, Schapira K (1973) Case report: depression, catatonic stupor, and EEG changes in
hyperparathyroidism. Psychol Med 3:509–515
de Entrambasaguas M, Sanchez JL, Schonewille W (2000) Malignant catatonia. Revista de
Neurologia 30:132–138
Flor-Henry P (1969) Psychosis and temporal lobe epilepsy. A controlled investigation. Epilepsia
10:363–395
Gjessing LR, Harding GFA, Jenner FA (1967) The EEG in three cases of periodic catatonia. Br J
Psychiatry 113:1271–1282
Hill D (1974) Non-verbal behaviour in mental illness. Br J Psychiatry 124(0):221–230
Husain AM, Horn GC, Jacobson MP (2003) Non-convulsive status epilepticus: usefulness of
clinical features in selecting patients for urgent EEG. J Neurol Neurosurg Psychiatry
74(2):189–191
Javanbakht A, Amirsadri A, Arfken C, Dewald O, Boutros NN (2012) Standard EEG study of
acute psychiatric patients with difficult to assess mental status. Am Assoc Emerg Psychiatry
10(1):1–5
Kanemoto K, Miyamoto T, Abe R (1999) Ictal catatonia as a manifestation of de novo absence
status epilepticus following benzodiazepine withdrawal. Seizure 8:364–366
Koehler K, Jakumeit U (1976) Subacute sclerosing panencephalitis presenting as Leonhard’s
speech-prompt catatonia. Br J Psychiatry 129:29–31
Lee S (1985) Non-convulsive status epilepticus: ictal confusion in later life. Arch Neurol
42:778–781
Lim J, Yagnik P, Schraeder P, Wheeler S (1986) Ictal catatonia as a manifestation of non-
convulsive status epilepticus. J Neurol Neurosurg Psychiatry 49:833–836
Louis ED, Pflaster NL (1995) Catatonia mimicking non-convulsive status epilepticus. Epilepsia
36:943–945
Neidermeyer E, Khalifeh R (1965) Petit mal status, an electroclinical appraisal. Epilepsia
6:250–262
Orland RM, Daghestani AN (1987) A case of catatonia induced by bacterial meningoencephalitis.
J Clin Psychiatry 48:489–900
Patry L, Guillem E, Pontonnier F, Ferreri M (2003) Catatonia de novo, report on a case:
immediate vital prognosis and psychiatric prognosis in longer term. Encephale 29(1):72–79
(Article in French)
Primavera A, Fonti A, Novello P, Roccatagliata G, Cocito L (1994) Neurol Neurosurg Psychiatry
57(11):1419–1422
References 123

Privitera MD, Strawsburg R (1994) Electroencephalographic monitoring in the emergency

department. Emerg Med Clin North Am 12(4):1089–1100
Rao NP, Kasal V, Mutalik NR, Behere RV, Venkatasubramanian G, Varambally S, Gangadhar
BN (2012) Has Kahlbaum syndrome disappeared or is it under diagnosed? Reexamining the
nosology of catatonia. J ECT 28(1):62–63
Riggio S (2005) Non-convulsive status epilepticus: clinical features and diagnostic challenges.
Psychiatry Clin N Am 28:653–664
Rohr-Le Floch J, Gauthier G, Beaumanoir A (1988) (Confusional states of epileptic origin. Value
of emergency EEG). Rev Neurol (Paris) 144(6–7):425–436
Rosebush PI, Mazurek MF (1996) Catatonia after benzodiazepine withdrawal. J Clin Psychop-
harm 16(4):315–319
Rosebush PI, Hildebrand AM, Furlong BG, Mazurek MF (1990) Catatonic syndrome in a general
psychiatric inpatient population: frequency, clinical presentation, and response to lorazepam.
J Clin Psychiatry 51(9):357–362
Suzuki, Miura N, Awata S, Ebina Y, Takanoto T, Honda T, Shindo T, Matsuoka H (1994)
Epileptic seizures in patient with acute catatonic syndrome. J Neurol Neurosurg Psychiatry
57(11):1419–1422
Swartz CM, Bottum KM (2002) Salazar Jr. Suppression of catatonia-like signs by lorazepam in
non-convulsive status epilepticus without seizure termination. Am J Geriatric Psychiatry
10:348–350
Thomas P, Beaumanoir A, Genton P, Dolisi C, Chatel M (1992) ‘De novo’ absence status of late
onset: report of 11 cases. Neurology 42(1):104–110
 Part III
Childhood Psychiatric Conditions
Chapter 13
Attention Deficit Disorder and Learning
Disabilities

Introduction

Attention deficit/hyperactivity disorder (AD/HD) remains the most common

childhood school-related behavioral problem. Despite an extensive body of
research, it also remains one of the more controversial neuropsychiatric disorders
both regarding its diagnosis and treatment. Wolraich and Baumgaertel (1997)
included posttraumatic and postinfectious encephalopathies, fetal alcohol syndrome
(FAS), chronic lead poisoning, untreated phenylketonurea, pervasive develop-
mental disorders, and absence seizures among the disorders to be differentiated
from AD/HD. Additionally, other psychiatric disorders like Bipolar Disorder and
also non-organic behavioral problems should also be considered on the differential
diagnostic list. Given the non-invasive and relatively inexpensive characters of the
EEG, full consideration of its possible usefulness in helping the clinicians arrive at
an accurate diagnosis should be considered. In this chapter we review available
literature examining the sEEGs of children or adults with AD/HD.
Two important points need to be emphasized prior to discussing the role of the
routine EEG in the diagnosis and management of ADD and AD/HD. First, it seems
essential that each individual (child or adult) being considered for the diagnosis of
ADD or AD/HD should receive a full neurological history and examination, with
additional neurological work-up (EEG or imaging) if the exam gives any indica-
tion of abnormality (Millichap 2000; Niedermeyer and Naidu 1998). The second
important point to be emphasized is the differential diagnosis. Once the possibility
of AD/HD arises a differential diagnostic list needs to be developed. High on the
list is bipolar disorder and learning disabilities. The problem of co-morbidity
particularly with learning disability is a major confound of this literature.

N. N. Boutros, Standard EEG: A Research Roadmap for Neuropsychiatry, 127

DOI: 10.1007/978-3-319-04444-6_13,  Springer International Publishing Switzerland 2013
128 13 Attention Deficit Disorder

Standard-EEG in AD/HD Studies

Risser and Bowers (1993) reported elevated levels of poly-spike EEG activity in
AD/HD children. They did not define what they meant by polyspikes and did not
provide examples. Moreover a smaller number of their healthy control children
exhibited similar patterns causing difficulty in interpreting the study. Subse-
quently, Frank (1993) reported that 31 % of a sample of 7–12 years old children
diagnosed with AD/HD had abnormal routine EEG (21 out of 64). Of the 21
children with abnormal EEGs, 84 % had spikes or spike-wave discharges. The
others had slowing of the background in excess of what is expected for the age. In
1998, Boutros et al. reported an association between AD/HD and the 14 and 6
positive spikes in children and adolescents. This EEG pattern is controversial but
has been linked to a variety of behavioral abnormalities including episodic somatic
symptoms and hyperactivity, and emotional instability.
A significant number of studies found variable rates of EEG abnormalities in
children with AD/HD. Phillips et al. (1993) reported on routine EEG screening in
children hospitalized over an 18-month period for behavioral problems. Eighty-six
children were admitted for conduct disorder or conduct disorder plus AD/HD
(N = 75; breakdown not provided) and AD/HD alone (N = 11). They reported
that 91 % of the records were either normal or showed ‘‘normal variant patterns’’.
The specific ‘‘normal variant’’ patterns exhibited were not provided. Eight (9 %)
records showed definite abnormalities showing background slowing or paroxysmal
discharges. They concluded that EEG screening may be of limited value in
childhood behavioral problems without clinical evidence of neurological disorders.
The above report differs from conclusions reached in subsequent studies.
Hughes et al. (2000) examined the EEGs of 176 children with AD/HD. They
reported an overall rate of ‘‘definite noncontroversial epileptiform activity of
30.1 %, mainly focal (usually occipital or temporal). Less often the epileptic
activity was generalized, with bilaterally synchronous spike and wave complexes
seen in 11 children (approximately 5 %). It is of major interest that this is the only
available concrete estimation of the prevalence of petit mal or absence activity
(i.e., 3/s Spike and Wave discharges). In the entire group, only 27.8 % were
completely normal and an additional 18.8 % had positive spikes as the only
abnormality. They concluded that AD/HD is a condition often with organic
changes in the form of EEG abnormalities; at times these abnormalities are of
epileptiform character. Such activity could contribute to a deficit in attention or a
plethora of movements (Hughes et al. 2000). Independently, Millichap (2000)
reported on the EEG findings from 100 consecutive children with AD/HD. They
reported an incidence of 7 % of ‘‘definite abnormalities’’ suggestive of epilepsy
and an additional 19 % moderately abnormal dysrhythmias not diagnostic of
epilepsy. Based on their findings, they suggested six specific indications for when
to obtain an EEG in a child presenting with AD/HD.
Standard-EEG in AD/HD Studies 129

These indications were;

(1) the presence of personal or family history of seizures;
(2) inattentive episodes characterized by excessive ‘‘daydreaming’’, and/or peri-
odic confused states;
(3) co-morbid episodic, unprovoked temper or rage attacks;
(4) frequently recurrent headaches;
(5) a history of head trauma, encephalitis, or meningitis preceding the onset of
AD/HD; and
(6) abnormalities on neurological examination. Again of major interest is that
such suggestions were in fact never put to any form of testing in subsequent
studies.

Richer et al. (2002) collected sEEGs from 347 children between the ages of
5–16 who were diagnosed with ADD/AD/HD. Overall 62 of the 347 children had
abnormal EEGs, with 41 having non-epileptiform abnormalities (focal or diffuse
slowing). The paper focused mainly on the types of epilpetiform abnormalities
detected. The EEGs were coded as epileptiform in 6.1 ± 1.3 %. This rate was
found to be significantly higher than the 3.5 ± 0.6 % they found in ‘‘normal’’
school-age children (Chi square P \ 0.025). It should be emphasized that sleep
was not consistently obtained and sleep deprivation was not used as an activating
procedure in this study. Please also note the reservations raised in Chap. 2 of this
book regarding the rates of EEG abnormalities in ‘‘normal’’ individuals. Perhaps
more importantly was there observations that epileptiform abnormalities were only
noted, in majority of individuals who did exhibit them, with activation procedures
like hyperventilation and photic stimulation. They further reported that only three
of the 21 children who exhibited epileptiform activity went on and developed
seizures in follow-up. Based on this last observation they concluded that despite
the significantly increased rate of epileptifrom activity in this group, the clinical
utility of the finding is limited. This conclusion is based on the concept that an
isolated epileptic discharge in a non-epileptic (but otherwise symptomatic) indi-
vidual is not an indication for management with antiepileptic agents (Binnie 2003).
We are arguing strongly throughout this book that this is a serious issue that
remains open for investigation. Other than that severe patients had a more like-
lihood of exhibiting EEG abnormalities the specifics of the clinical picture could
not dictate which patients should undergo an EEG examination. Finally, two of the
21 patients who exhibited epileptiform discharges showed generalized discharges
of the 3 Hz spike and wave variety indicative of absence attacks. This is a rather
important observation as it could significantly influence diagnosis and treatment.
Relatively more recently, Holtman et al. (2003) examined the EEGs of 483 AD/
HD outpatients between 2 and 16 years of age (diagnosis based on DSM-IV).
Rolandic spikes were detected in the EEGs of 27 children (5.6 %); 22 boys and 5
girls. Seizure rate during follow-up tended to be larger in children with Rolandic
spikes. AD/HD children with rolandic spikes were brought for evaluation at an
earlier age than AD/HD children with normal EEGs. These children were also
130 13 Attention Deficit Disorder

noted to exhibit more hyperactive-impulsive symptoms. Aydin et al. (2003) found

an overall rate of abnormalities (both slowing and epileptiform) of 9.1 % of 49
children with AD/HD diagnosis. Three children had focal or multifocal spikes
(6.1 %) and three had slow wave abnormalities.
While the standard practice in the USA does not recognize the importance of
the routine sEEG in the work up of children presenting with AD/HD, this is not the
case in Europe. Schmidt et al. (2002) discusses this issue. The EEGs of 124
children with AD/HD before and during treatment with methylphenidate were
analyzed retrospectively. They did not see evidence of increased abnormalities due
to the therapy. They concluded that an EEG during therapy with methylphenidate
is not necessary but before commencing a planned methylphenidate therapy an
EEG should be performed. Again, this proposal was never re-tested in prospective
studies.
Finally, Hemmer et al. (2001) reported 15.4 % of 234 children (179 males/
9.1–3.6 years of age and 55 females/9.6–3.9 years of age) to have epileptiform
abnormalities. They did not include any of the controversial EEG waveforms like
the 14 and 6 positive spikes thus explaining (at least in part) the lower rate of
abnormalities. They also reported seizure incidence of 0.6 % in subjects with
normal pre-ritalin EEGs and 10 % in subjects with pre-Ritalin epileptiform EEGs.
This important finding is yet to be confirmed or refuted in prospective studies.
A much more difficult finding to interpret and base action upon is the finding of
one of the controversial waveforms (covered in much details in chaps. 15–19).
Boutros et al. (1998) gathered diagnostic information through structured inter-
views from four groups of psychiatric inpatients aged 4–17 years. The four groups
were; (1) patients exhibiting positive spikes (N = 25); (2) patients exhibiting frank
epileptic discharges (N = 29); (3) patients exhibiting slow wave abnormalities
(mainly focal) (N = 23); and (4) patients with normal EEGs (N = 25). AD/HD
symptoms were significantly more frequent in the PS group compared to the other
three groups combined. Anxiety symptoms showed a strong trend (P = 0.06) to be
more represented in the PS group as well.

Learning Disabilities

A surprise finding from our search is the presence of three reasonably well-designed
studies where individuals with different forms of learning disabilities and EEG
abnormalities responded well to AEDs (Porras-Kattz et al. 2010; Etchephareborda
2003) while one similarly well-designed study found the opposite (Ronen et al.
2000) (Table 13.1). While currently learning disability is not even considered an
indication to obtain an EEG, these observations suggest that further research in this
area is warranted. A rather limited literature addresses the observation that IEDs
(e.g., spikes, polyspikes or spike and wave discharges) may be detected in children
who are presenting solely with a learning disability (Becker et al. 1987). It is quite
 Learning Disabilities

Table 13.1 Treatment studies in LD patients with abnormal EEGs

Paper Patient group Control group EEG findings AED Results Comments
Ronen et al. Kids with behavioral or N/A Generalized or focal VPA or placebo Kids did better on Double blind, single
(2000) learning problems bursts of spike-slow placebo than VPA crossover study
N=8 waves
Binnie (2003) Patients with transitory N/A 3/s spike wave activity lamotrigine or Eight patients did better Cognitive
cognitive impairment levetiracetam on drugs than impairment
N = 10 placebo
Etchephareborda Groups of children with N/A Paroxysmal CBZ or VPA EEG improvement for Case series
(2003) LSD, SLD, AD/HD, all four groups
or MNS
N = 40
Porras-Kattz Children with learning Children on Interictal paroxysmal Magnesium Unchanged number of Double blind
et al. (2010) disability placebo for valproate or paroxysms but
N = 18 6 months placebo decreased theta
N=8 activity, increased
IQ measure on
medication
131
132 13 Attention Deficit Disorder

possible that sub-clinical IEDs emanating from certain areas important for the
learning functions can manifest solely as a learning disability. For more detail see
the chapter on IEDs.
As stated by Keck et al. (1992), the difficulties inherent in conducting con-
trolled studies of pharmacologic agents for the treatment of these syndromes may
rest in part with the uniqueness and poor generalizability of specific cases, the lack
of diagnostic homogeneity underlying the nonspecific but troublesome nature of
these behaviors, and ethical concerns regarding safety in placebo-controlled
designs. Evidently, this subject needs more research studies to be conducted to
have a better understanding of the efficacy of anticonvulsants on non-epileptic
patients.
Additional comments should be made about the study conducted by Porras-
Kattz et al. (2010). Although this study had a small sample size, it had the ideal
design for this line of investigation. The patient group was largely homogenous, all
having IEDs and they were randomly assigned an AED or a placebo. Finally,
improvement was measured utilizing standardized scales. By comparison, Reeves
et al. (2003) concluded that EEG does not predict response to valproate in the
treatment of aggression in patients with Axis-II disorders while none of their
patients exhibited frank epileptiform activities. It cannot be over emphasized that
different EEG abnormalities have different neurophysiological underpinnings and
only paroxysmal activity would logically be expected to predict a favorable
response to AED therapy.

Supported Observations

(1) A certain percentage of children with AD/HD exhibit clear EEG abnormalities
not infrequently of the paroxysmal epileptiform type.
(2) Based on current knowledge, it is not possible to predict which patient pre-
senting with AD/HD/ADD will in fact exhibit an EEG abnormality.

Open Research Questions

(1) The essential question of the impact of discovering epileptic activity in a child
presenting with AD/HD remains not completely answered.
(2) Whether the yield from an EEG is 6 or 30 %, the information seems important
for better formulating diagnosis and treatment. It is important to establish cost-
effectiveness of any procedure in well-designed prospective studies.
(3) What are the implications of identifying focal slow wave activity in AD/HD?
(4) What is the actual prevalence of petit mal episodes masquerading as AD/HD?
Learning Disabilities 133

(5) In what ways are sEEG and QEEG complimentary in the work-up of Children/
adolescents presenting with AD/HD.
(6) What is the role of EEG Neurofeedback in treating AD/HD with various EEG
abnormalities.
The same above research questions, perhaps with the exception of number 4,
are applicable to the problem of LDs.

References

Aydin K, Okuyaz C, Serdarglu A, Gucuyener K (2003) Utility of electroencephalography in the

evaluation of common neurologic conditions in children. J Child Neurol 18:394–396
Becker J, Velasco M, Harmony T, Marosi E, Landázuri AM (1987) Electroencephalographic
characteristics of children with learning disabilities. Clin Electroencephalogr 18(2):93–101
Binnie C (2003) Cognitive impairment during epileptiform discharges: is it ever justifiable to
treat the EEG? Lancet Neurol 2:725–730
Boutros N Fristad M Abdollohian A (1998) The fourteen and six positive spikes and attention-
deficit hyperactivity disorder. Biol Psychiatry 44(4): 298–301
Etchephareborda MC (2003) Treatment of children with a paroxysmal electroencephalogram
without seizures. Rev Neurol 37(3):293–297
Frank Y (1993) Visual event related potentials after methylphenidate and sodium valproate in
children with attention deficit hyperactivity disorder. Clin Electroencephalogr 24(1):19–24
Hemmer SA, Pasternak JF, Zecker SG, Trommer BL (2001) Stimulant therapy and seizure risk in
children with ADHD. Pediatr Neurol 24:99–102
Holtman M, Becker K, Kentner-Figura B, Schmidt MH (2003) Increased frequency of Rolandic
Spikes in ADHD children. Epilepsi 44:1241–1244
Hughes ER, DeLeo AJ, Melyn MA (2000) The electroencephalogram in attention deficit-
hyperactivity disorder: emphasis on epileptiform discharges. Epilepsy Behav 1:271–277
Keck PE, McElroy SL, Friedman LM (1992) Valproate and carbamazepine in the treatment of
panic and posttraumatic stress disorders, withdrawal states, and behavioral dyscontrol
syndromes. J Clin Pyschophar 12(1):36S–41S
Millichap JG (2000) Attention deficit-hyperactivity disorder and the electroencephalogram.
Epilepsy Behav 1:453–454
Niedermeyer E, Naidu SB (1998) Rett syndrome, EEG and the motor cortex as a model for better
understanding of attention deficit hyperactivity disorder (ADHD). Eur Child Adolesc
Psychiatry 7:69–72
Phillips BB, Drake ME Jr, Hietter SA, Andrews JE, Bogner JE (1993) Electroencephalography in
childhood conduct and behavior disorders. Clin EEG 24:25–30
Porras-Kattz E, Harmony T, Ricardo-Garcell J et al (2010) Magnesium valproate in learning
disabled children with interictal paroxysmal EEG patterns: preliminary report. Neurosci Lett
492:99–104
Reeves RR, Struve FA, Patrick G (2003) EEG does not predict response to valproate treatment of
aggression in patients with borderline and antisocial personality disorders. Clin Elelctroen-
cephalogr 34(2):84–86
Richer LP, Shevell MI, Rosenblatt BR (2002) Epileptiform abnormalities in children with
attention-deficit-hyperactivity disorder. Pediatr Neurol 26:125–129
Risser MG, Bowers TG (1993) Cognitive and neuropsychological characteristics of attention
dificit hyperactivity disorder children receiving stimulant medications. Percept Mot Skills
77:1023–1031
134 13 Attention Deficit Disorder

Ronen GM, Richards JE, Cunningham C, Secord M, Rosenbloom D (2000) Can sodium valproate
improve learning in children with epileptiform bursts but without clinical seizures? Dev Med
Child Neurol 42:751–755
Schmidt JK, Pluck J, Von Gontard A (2002) Verzicht auf eine EEG-diagnostik vor Beginn und
unter einer Therapie mit Methylphenidat: gefahrlich oder gerechtfertigt? (Waiver of EEG
diagnostics prior to and during methylphenidate therapy: dangerous or justifiable?).
Z. Kinder-Jugendpsychiar 30:295–302
Wolraich ML, Baumgaertel A (1997) The practical aspects of diagnosing and managing children
with attention deficit hyperactivity disorder. Clin Pediatr (Phila) 36(9):497–504
Chapter 14
Autistic Spectrum Disorders

Introduction

Pervasive developmental disorders (PDD) or autistic spectrum disorders (ASD)

encompass a heterogeneous group of individuals with early childhood onset of
deficits in social interaction and language development, a restricted repertoire of
interests and activities, as well as a wide range of cognitive difficulties. The DSM-
IV divides the PDDs into five behaviorally defined categories: autistic disorders,
Asperger syndrome, Rett Syndrome, disintegrative disorder, and pervasive disorder
not otherwise specified (NOS). In a significant minority (20–40 %) of these chil-
dren parents report observing deterioration in the child’s language skills, accom-
panied with deterioration of nonverbal communication and play skills. Behavior
often worsens and cognitive decline may also be observed. At least some of these
children are entirely healthy prior to the onset of the disorder. In particular Dis-
integrative Disorder refers to a subgroup of children on the PDD spectrum who
were developing entirely normally, including speaking in sentences, in whom
deterioration occurred after 2 years of age. In some of these children, onset of
deterioration may be as late as mid-childhood attesting to the significant hetero-
geneity among individuals diagnosed with Autism. Furthermore, while much
research into the biological factors contributing to the development of disorders on
the spectrum is underway, the constellation of factors necessary for the syndrome to
manifest remains far from clear.

Evidence of Neurological Contributions to the Syndrome

The initial report by Kanner (1943) where he described 11 children with autism did
provide significant hints for a significant neurological component to this syndrome.
One child had seizures, three were mute, and five were macrocephalic. Despite these
observations, only when the first report of EEG abnormalities in this population was
published, was a more serious consideration of an organic/neurological etiology

N. N. Boutros, Standard EEG: A Research Roadmap for Neuropsychiatry, 135

DOI: 10.1007/978-3-319-04444-6_14,  Springer International Publishing Switzerland 2013
136 14 Autistic Spectrum Disorders

entertained (Gubbay et al. 1970). In this initial report, from Western Australia, of 25
children with Autism, 30 % had seizures and a full 80 % had abnormal EEGs. It took
another 5 years before the next report appeared. Small and her collaborators (1975)
examined 147 children with autism and reported a 64 % prevalence of EEG
abnormalities. These early reports strongly suggested that the epileptiform activity
seen in association with ASD is not simply comorbid but is likely to be etiologic
(Levisohn 2007). In fact approximately one-third of children with ASD develop
epilepsy (Gillberg 1991). The occurrence of epilepsy in ASD has been intensively
investigated (Tuchman and Rapin 1997; Hrdlicka et al. 2004; Canitano et al. 2005;
Hughes and Melyn 2005). DeLong and Nohria (1994) obtained both complete
neurological assessment and psychiatric family history from 40 children with ASD.
Neurological evaluation included EEG, MRI, karyotyping, and positron emission
tomography (PET) as indicated. Twenty patients had positive neurological findings,
18 of which had negative psychiatric family histories. Fourteen of the 20 patients
without neurological findings had family histories of affective disorders. These
patients tended to be of higher function. These findings highlight the importance of
early clinical evaluation to identify the subgroup the patient belongs to as there
seems to be significant treatment and prognostic implications.
On the other hand, the literature also attests to the fact that a sizeable proportion of
ASD children who have never experienced a seizure may harbor isolated epilepti-
form discharges (IEDs). For a more expanded discussion regarding IEDs please see
Chap. 7. The clinical significance of the detection of IEDs in a child on the ASD as
well as the therapeutic implications are far from being will delineated, perhaps with
the most notable exception of Landau-Kleffner syndrome (LKS) where the IEDs are
considered etiologic for the development of the associated aphasia (Stefanatos et al.
2002). Whether or not expensive longitudinal studies may be necessary to find out if
and how much the bioelectrical abnormalities play a causal role in those subgroups
of children with various degrees of both language deterioration and the emergence of
autistic symptoms. One has to remember that it took nearly 40 years to fully
acknowledge the epileptic origin of aphasia in LKS and the milder acquired cog-
nitive problems in Rolandic epilepsies (Deonna and Roulet-Perez 2010). The fact
does remain that a strong link between IEDs, in the absence of seizures, and any form
of behavioral aberrations remains far from being established (So 2010). The purpose
of this chapter is to examine available literature and extract any possible clinical
recommendations as well as defining the areas of needed research.

EEG Abnormalities in ASD

A significant proportion of ASD children have abnormal EEGs even those who
never had seizures. These abnormalities can range from mild slow-wave abnor-
malities to frank epileptiform discharges. It is very important to point out very
early in this discussion that these epileptiform discharges may only be detected
during sleep and at times may require prolonged monitoring (Table 14.1).
Table 14.1 Percentages of IEDs among nonepileptic ASD children
Paper N (M/F) Percentage with IEDs Location Comments
Parmeggiani et al. (2010) 345 45.4 Temporal and central
Introduction

31.4 %. Much less

frontal
Yasuhara (2010) 1014 49 65.5 frontal regions
Ekinci et al. (2010) 57 (86 % male) 11 Overall rate of epilepsy in group
24 %
Giannotti et al. (2008) 34 regressed 17 Rate of epilepsy in the regressed
70 nonregressed 25 was 24 % while only 15 % in
the nonregressed group. The
two groups differed on the
percentage of subjects with
high-frequency spikes
Hara (2007) 130 18 %/nonepileptics Both focal and generalized Epileptiform EEGs predicted
68 %/epileptics subsequent seizures
Chez et al. (2006) 889(706/183) 60.7 R.Fr (21.5), Bitemporal No difference in rate of EEG
(20.2). Gen sp and wave abnormality between those
(16.2), left temp (15.2) with and without regression
Reinhold et al. (2005) 316 27 Temporal (38), frontal (28),
central (23), occipital (8)
Canitano et al. (2005) 46 22
Hashimoto et al. (2001) 86 43 76.6 % frontal Equal right and left side
Rossi et al. (1995) 106 18.9 Focal and multifocal. 45 % Photoparoxysmal response in some
centrotemporal children
Hughes and Melyn (2005) 59 75
Tuchman and Rapin (1997) 392 59 in epileptics and 8 % 50 % centrotemporal 14 % in nonepileptics but with
in nonepileptics history of deterioration
Hrdlicka et al. (2004) 77 55.6
137
138 14 Autistic Spectrum Disorders

Findji et al. (1979) described an 8-year-old child whose behavior alternated

between excitation and autism with stereotypes. The EEG showed 5 c/s tempo-
roparietal sharp wave discharges lasting from 1 s to 20 min. These discharges
(even the prolonged ones) were not associated with any clinical signs of epilepsy.
Tuchman and Rapin (1997) examined the EEGs (containing sleep) of 392
children on the autistic spectrum. The EEGs were epileptiform in 59 % among
epileptic children but only 8 % of 335 nonepileptic children. When historical
clinical deterioration is evident (as was the case in 155 nonepileptic children) the
rate of epileptiform EEG abnormalities was significantly higher (14 %) as com-
pared to the rate in children without obvious deterioration (6 %). In this study,
approximately half of the epileptic discharges were centrotemporal. A smaller
percentage of epileptiform activity was in the perisylvian region and this occurred
mainly in nonepileptic children with clinical deterioration. A rather important
point made in this paper is that the average child was first evaluated neurologi-
cally four years following the onset of deterioration. This point highlights the need
for studies conducted at much earlier stages of the illness.
Hashimoto et al. (2001) examined the EEGs (during sleep) of 86 autistic
children. Forty-three percent (37 cases) had epileptic discharges. Of these 37
patients 27 (73 %) had localized spikes, eight had multiple spike foci, one with
generalized, and one with both focal and generalized spikes. Forty-seven epileptic
discharging foci were identified in the 36 patients with focal abnormalities. Thirty-
six (76.6 %) of these foci were in the frontal region, one in the temporal, seven in
the centroparietal, and three in the occipital regions. Twenty (55.6 %) of the
frontal spikes were at midline (11 at FZ and 9 at CZ), eight on the left side, and
eight on the right side. The dipole of midline spikes was in the deep midline frontal
region. These results strongly suggest that frontal dysfunction is important in the
development of autistic symptoms.
Hrdlicka et al. (2004) examined 77 autistic children. They reported the EEGs to
be abnormal in 55.6 % with nonepileptiform abnormalities in 17.5 % and epi-
leptiform in 38.1 %.
Reinhold et al. (2005) found abnormal EEGs in 27 % of 316 children evaluated
for ASD. Of these abnormalities 65 % (55 children) were epileptiform and slowing
in 15 % (13 patients). The focality of the epileptiform activity was in the temporal
regions in 30 %, 28 % in the central region, 23 % in the frontal regions, and 8 % in
the occipital area. Also recently, Canitano et al. (2005) examined 46 consecutive
children with autism (34 boys and 12 girls, mean age 7.8 ± 2.7 years). Twenty-two
percent had epileptiform abnormalities without having history of seizures. They did
not report the rate of nonepileptiform abnormalities. On the other hand, Hughes and
Melyn (2005) found abnormal EEGs in 75 % of 59 children with autism. Twenty
percent of patients with spike discharges did not have clinical attacks.
A rather interesting study was conducted by Chez et al. (2004) where they
examined the rate of EEG abnormalities of age-matched siblings of autistic chil-
dren who had abnormal sleep EEGs. They report less abnormalities in these
children and conclude that genetic factors alone do not explain the higher fre-
quency of EEG abnormalities reported in ASDs.
Introduction 139

Lewine et al. (1999) utilized the magnetoencephalogram (MEG) to compare

patients with LKS and ASD. They made sure all subjects achieved phase III sleep
during the recording. They included six children with LKS, and 50 children with
ASD (with evidence of deterioration between 20 and 36 months). Sixteen of the
subjects met criteria for Autism and 34 for PDD-NOS. Five of the six LKS
children had complex partial seizures but only 15 of the ASD children had a
seizure disorder. The MEG of all LKS children had primary or secondary epi-
leptiform involvement of the left intraperisylvian region (all but one had
involvement of the right perisylvian region as well). In all LKS cases there were no
independent foci outside the sylvian regions. MEG identified epileptiform activity
in 41 of the 50 ASD patients (82 %). In contrast, simultaneous EEG revealed
epileptiform activity in only 68 %. This is a seriously important observation as it
highlights the fact that a negative test simply cannot mean the absence of
abnormality and the need for complete work-ups prior to declaring a syndrome
‘‘psychological’’. When epileptiform activity was present, the same intraperisyl-
vian areas like in LKS were involved in 85 % of cases. Whereas primary activity
outside of the sylvian regions were not seen in any LKS children, 75 % of the ASD
children with epileptiform activity demonstrated additional nonsylvian zones of
independent epileptiform activity. An additional important observation provided
by this group is the strong tendency to have multifocal abnormalities (i.e., both
perisylvian and extra temporal activities). Finally, majority of abnormalities were
detected during sleep. Despite the multifocal nature of the epileptiform activity in
the ASDs, neurosurgical intervention aimed at control has led to the reduction of
autistic features and improvement in language skills in 12 of 18 cases. Nass et al.
(1998) reported seven autistic epileptic children who failed anticonvulsant treat-
ments and were subjected to multiple subpial transactions (MST). While epilepsy
and EEG abnormalities improved in all seven, language, social and overall
behavior only improved moderately and improvement seemed to be temporary.
A well-designed study was conducted by Rossi and colleagues (1995) where
106 autistic patients without any evidence of congenital or acquired encephalop-
athies were examined. They reported a prevalence of 18.9 % of EEG abnormalities
in children without epilepsy. Epileptiform activity was focal and multifocal. In
45 % of cases this activity was typical of benign childhood partial epilepsy with
centrotemporal spikes. Table 14.1 lists the different percentages of abnormal
EEGs among autistic children published in full length papers in peer-reviewed
journals.
Chez et al. (2006) were able to review the EEGs of a large number of Autistic
children for the purpose of providing a definite account of the rate and nature of
EEG abnormalities in children with autistic symptoms who had no seizures or
identifiable genetic problems. Starting with 1,268 children with ASD syndromes
examined between 1996 and 2005, they began by excluding children with iden-
tified genetic disorders or tuberous sclerosis. They then excluded all children with
history of seizures. They were left with 889 children with no prior evidence of
epilepsy. These children then underwent 16 channel ambulatory sleep EEGs. They
reported a rate of EEG abnormalities of 60.7 %. One of their most important
140 14 Autistic Spectrum Disorders

findings is that, in this population, epileptiform abnormalities were detected only

during sleep. A second important finding is the report of different forms of epi-
leptic activities from focal temporal regions being the most common next to
generalized spike and slow-wave discharges. As is well known, the pathophysi-
ologies of these abnormal discharges are different. This suggests the possibility
that within the group of non-epileptic autistic children with epileptic discharges,
more specific phenotypes may exist. Moreover, in this rather large sample, the
frequency of frontal lobe epileptic activity was rather low. The significance of this
observation needs further exploration as it has been suggested that involvement of
the frontal lobes may be important for the eventual development of autistic
symptoms. Finally, 176 patients with epileptic discharges were placed on valproic
acid for treatment. In follow-up EEGs (on average 10 months later), 46.6 %
normalized, and an additional 17 % showed improvement. None had worsened.
The authors concluded that a more proactive approach to evaluation and treatment
was justified as some of the abnormalities may be reversible. This conclusion is
supported by the work of Pressler et al. (2005) where they show that treatment of
interictal epileptiform discharges can improve behavior in epileptic children with
behavioral problems.
An interesting relatively recent study from the University of Istanbul examined
a group of children with West Syndrome and compared the rate of emergence of
autistic features to the nature of EEG abnormalities (Kayaalp et al. 2007). Two
groups, all with West syndrome, were formed; West and Autism (N = 108) and
West alone (N = 123). All children were followed up with regular and video
EEGs. The number of patients with at least one hypsarrhythmic EEG at age one
year or later was significantly higher in the autistic group (86 %) than in the
nonautistic group (29 %). Frontal predominance of the primary foci on EEGs with
and without hypsarrhythmia was seen in 95.3 % in the autistic group but only in
28.8 % in the nonautistic group (p \ 0.001). Frontal abnormalities in the EEGs
which were mainly bilateral, and the persistence of hypsarrhythmia were signifi-
cantly related to the emergence of autistic behavior. The authors suggest that
paroxysmal discharges in cortical areas undergoing rapid maturation may be
involved in the development of autistic features.
In a relatively recent effort to develop guidelines for screening EEGs in ASD,
Kagan-Kushnir et al. (2005) conducted a comprehensive review of available lit-
erature. They concluded that seizures are common occurring in 20–30 % of
autistic children based on the majority of studies available at the time. Subclinical
EEG abnormalities (i.e., no epilepsy) were found in 6.1–31 %. Evidence for the
effectiveness of anticonvulsants and corticosteroids in reducing seizures and/or
autistic symptoms is based primarily on case series, with only one published
randomized trial at the time. They concluded that as of the time of the publication
(2005) there was insufficient evidence to recommend against the use of screening
EEGs in autistic patients. This conclusion remains current. They also concluded
that given the high frequency of seizure disorders in this population that a high
index of clinical suspicion should be maintained for subtle symptoms of seizures.
Introduction 141

Electrical Status Epilepticus in Slow-Wave Sleep

Electrical Status Epilepticus in Slow-Wave Sleep (ESES) is an epileptic syndrome

that occurs in childhood and may or may not be accompanied by seizures. This
syndrome is characterized by continuous slow spike-wave activity during non-
REM sleep. The electroclinical condition was described by Patry et al. (1971). It
occurs in children (mostly around age of 8 years) with chiefly nocturnal seizures
and mild mental retardation. The unique EEG manifestations during non-REM
sleep are coupled with rather minimal EEG abnormalities during wakefulness (and
perhaps accounting for what is possibly a lower rate of recognition). The dramatic
changes during sleep disappear during adolescence. It is very interesting and
important to note that when a child with ESES is awakened from such a severely
disturbed sleep, he/she is usually awake and in a normally refreshed state. Dete-
rioration of language and behavior is also seen in this condition which may need to
be differentiated from ASD (Beaumanoir et al. 1995). In ESES behavioral cor-
relates may range from a severe global deterioration compatible with the diagnosis
of disintegrative disorder or autistic disorder to minimal or absent behavioral
correlates (Rapin 1995).

Response to Anticonvulsant Treatment

Gillberg and Schaumann (1983) described two cases of infantile autism without
clinical seizures, where EEG abnormalities were not discovered until relatively
late in the course of the psychiatric disorder. Anticonvulsant medications led to the
complete disappearance of psychotic symptoms and to simultaneous disappear-
ance of the pathological EEG changes (Table 14.2).
In 1997, Childs and Blair reported a case of twins with absence seizures and
autistic features. The two boys had autistic features prior to the onset of seizures
during their second year of life. By age three the twins were significantly delayed
in the areas of socialization, communication, and impulse control. Typical autistic
features were pronounced, including, nonpurposeful and self-stimulatory behavior,
lack of symbolic play, poor eye contact, echoic and noncommunicative speech,
and unresponsiveness to disciplinary efforts. While the absence seizures were
observed by the parents since age two, parents only recognized what they were
retrospectively after diagnosis was made at age 3 years and 1 month. Both boys
responded to valproic acid treatment with control of seizures and with a dramat-
ically accelerated rate of acquisition of both language and social skills. These
illustrative cases raise the possibility of a direct and causal affect between epilepsy
and the autistic features.
Hollander et al. (2001) conducted a retrospective pilot study to determine
whether valproic acid was effective in treating core dimensions and associated
features of autism. They included 14 patients with either autism, Asperger’s
Table 14.2 Clinical response to AEDs in ASD patients with EEG abnormalities
142

Paper Patient group Control group EEG findings AED Use Results Comments
Gillberg et al. (1983) Infantile autism N/A 3/s Spike and Wave in VPA, ethosuximide Both children passed Small case series
N=2 one, and slow and from the
sharp waves in the noncommunicative/
other bizarre area to the
quite normal area
Hollander et al. (2001) Autism: N = 10 N/A Three had abnormal Divalproex sodium 71 % had a sustained Case series Pilot
Asperger: N = 2 EEGs: sharp activity response. Divalproex study. Relation
PDD: N = 2 in two and focal slow is effective for to EEG
NT = 14 in one. Five with stabilizing mood, abnormalities
normal EEGs and aggression, social not discussed
EEGs not obtained in deficits, and repetitive
five behavior
Pressler et al. (2005) Controlled or Randomized. Spikes, sharp waves, Lamotrigine Suppressed discharges Double-blind,
mild epileptic AED multiple spikes and does not affect placebo-
children followed cognition P \ 0.05 controlled,
N = 61 by placebo Significant behavioral cross-over
or vice improvement study
versa
Canitano et al. (2006) Children with N/A Paroxysmal pattern, VPA, clobazam, or in EEG improved. language Small case series
ASD spike-wave combination after 5 years in one.
14

N=2 complexes, And behavior at 6 in

epileptiform another
discharges
(continued)
Autistic Spectrum Disorders
Table 14.2 (continued)
Paper Patient group Control group EEG findings AED Use Results Comments
Chez et al. (2006) ASD Patients N/A Abnormalities including VPA EEGs of 46.6 % DigiTrace
Introduction

With abnormal focal and bilateral normalized on VPA. ambulatory

EEGs sharp waves as well as Seventeen percentage digital EEG.
N = 176 focal slowing improved and 36.3 no Prospective
change and large study
Merhar et al. (2007) Autistic boys N/A Epileptiform discharges Trileptal, depakene, Lamictal improved skills Small case series
with lamictal and behavior in one of
Klinefelter the patients
syndrome
N=2
Frye et al. (2010) Children with N/A Sharp waves (61 %) and CBZ, VPA, OxCBZ, 80 % improved rapidly. Controlled study
ASD and focal/slow waves ethosuximide Twenty percentage open label
atypical lamotrigine, needed a second AED
cognitive levetricatem, IVIG
development
N = 20
143
144 14 Autistic Spectrum Disorders

syndrome or PDD-NOS. Subjects were included irrespective of history of seizures

or EEG abnormalities. Ten of the 14 patients who completed the trial (71 %) were
rated as responsive to treatment (mean dose 768 mg/day and range of
125–2,500 mg/day). The medication was generally well tolerated. Improvement
was noted in core symptoms of autism as well as the associated features of
affective instability, impulsivity, and aggression. Of note is that all patients with
abnormal EEGs were rated as responders.
A review by Tharp (2004) concluded that there is no justification for use of
anticonvulsant medications or surgery in children with PDD without seizures; that
is there is no evidence that treatments to eliminate EEG spikes will have a ther-
apeutic effect on the behavioral abnormalities of PDD and autism. In a similar
vein, a report of failure of antiepileptic medications to ameliorate symptoms of
autism in two children with Klinefelter syndrome (47, XXY) was published
(Merhar and Manning-Courteny 2007). This negative conclusion may not be fully
justified. The expectation that anticonvulsant will affect autistic behavioral
symptoms in a manner similar to the way they affect seizures is unsupportable as
the pathophysiologies are strikingly different. Whereby the development of seizure
activity in the brain can lead to seizures with varying intervening intervals, the
development of behavioral aberrations takes many years to develop. Hence, a fast
reversal with few weeks of treatment may not be realistic. Moreover, while failure
of epilepsy to respond to medications is only announced after many trials of
different classes of antiepileptic medications, researchers are ready to announce
lack of efficacy in ASD following usually a single medication trial. Concerted and
patient effort needs to be expended in conducting well-designed studies taking into
account, nature (focal vs. generalized) and severity of epileptic activity in choice
of medications to be tried. Moreover, it may be most crucial to attempt treatment
at the earliest indication of the illness rather after many years of damaging
interictal spikes before the onset of the first manifest seizure.
Several reports support the need for the above-proposed investigations. Anag-
nostou et al. (2006) provided limited pilot data showing that divalproex (valproic
acid) sodium acid can be useful to control the irritability associated with using
selective serotonin uptake inhibitors (SSRIs) to treat the behavior symptoms of
autism. While no EEGs were obtained, children with a documented history of
seizures were excluded from the study. Hollander and colleagues at Mount Sinai
(2006), included 13 individuals with ASD in an 8-week, double-blind, placebo-
controlled trial of divalproex (valproic acid sodium) versus placebo. There was a
significant group difference on improvement of repetitive behavior as measured by
the Children’s Yale-Brown Obsessive Compulsive Scale (C-YBOCS) (p \ 0.04)
with a large effect size of d = 1.6.
Most recently, Frye et al. (2010) retrospectively reviewed the charts of 22
children with atypical cognitive development that did not respond to standard
educational therapy and demonstrated discharges on the EEG. Children did not
demonstrate obvious signs of seizures. Majority of children had multifocal dis-
charges on the EEGs. Of the 20 patients treated with antiepileptic medications,
70 % demonstrated definite improvement within one clinic visit. The study
Introduction 145

suggests that children with EEG discharges and developmental cognitive disorders
demonstrate a unique pattern of symptomatology and discharges on EEG. They
further concluded that if a child with a developmental cognitive disorder and does
not respond to standard therapy may benefit from screening with an EEG and a
trial of antiepileptic medication if discharges are detected.

Supported Observations

(1) There is no doubt that the rate of epilepsy among ASD children is significantly
elevated.
(2) There is an extremely high prevalence of EEG abnormalities among the ASD
population as compared to other psychiatric disorders.
(3) A large proportion of ASD children with abnormal EEGs will never develop a
seizure disorder.
(4) Autism can develop in the absence of detectable epileptic spikes.
It is quite possible that Autism is a heterogeneous disorder much like all other
psychiatric disorders. The presence of epileptic discharges thus could present an
endophenotype that may help decrease the heterogeneity of the disorder.

Open Research Questions

(1) The cost-effectiveness of EEG work-up in children exhibiting ASD symptoms

has not been determined. Of course, if intervention at some stage can prevent
or ameliorate a life-long disorder, even a low yield may be well justified not
only on economic bases but also for the suffering of patients and their families
that may be avoided. It is important here to invoke the concept of the ‘‘tip of
the iceberg’’ as discussed in Chap. 7 on the isolated epileptic discharges.
(2) Does the fact that an epileptic discharge was not detected necessary mean that
there is no such activity going on?
(3) What is the role of MEG in view of its high cost?
(4) Does inability to detect an abnormality in a single EEG recording means that
subsequent recordings will never reveal new findings. It is of interest that there
are no available studies of autistic children without seizures or initial abnormal
EEGs with follow-up EEGs as their disease unfolds (of course until a seizure
does occur).
(5) How does the presence of IEDs affect the neural development in nonepileptic
ASD children?
(6) What is the impact of AEDs on current symptoms and on prognosis?
(7) Is there difference in prognosis relative to when AEDs were started (i.e., at
onset of ASD vs. onset of seizures)?
146 14 Autistic Spectrum Disorders

References

Anagnostou E, Esposito K, Soorya L, Wasserman S, Hollander E (2006) Divalproex versus

placebo for the prevention of irritability associated with fluoxetine treatment in autism
spectrum disorder. J Clin Psychopharmacol 26(4):444–446
Beaumanoir A, Bureau M, Deonna T. (eds) (1995) Electrical status epilepticus during sleep. In:
Continious spikes and waves during slow sleep. Libbey, London
Canitano R, Zappella M (2006) Autistic epileptiform regression. Funct Neurol 21(2):97–101
Canitano R, Luchetti A, Zappella M (2005) Epilepsy, electroencephalographic abnormalities, and
regression in children with autism. J Child Neurol 20(1):27–31
Chez MG, Buchanan T, Aimonovitch M, Mrazek S, Krasne V, Langburt W, Memon S (2004)
Frequency of EEG abnormalities in age-matched siblings of autistic children with abnormal
sleep EEG patterns. Epilepsy Behav 5(2):159–162
Chez MG, Chang M, Krasne V, Coughlan C, Kominsky M, Schwartz A (2006) Frequency of
epileptiform EEG abnormalities in a sequential screening of autistic patients with no known
clinical epilepsy from 1996 to 2005. Epilepsy Behav 8(1):267–271
Childs JA, Blair JL (1997) Valproic acid treatment of epilepsy in autistic twins. J Neurosci Nurs
29(4):244–248
DeLong R, Nohria C (1994) Psychiatric family history and neurological disease in autistic
spectrum disorders. Dev Med Child Neurol 36(5):441–448
Deonna T, Roulet-Perez E (2010) Early-onset acquired epileptic aphasia (Landau-Kleffner
syndrome, LKS) and regressive autistic disorders with epileptic EEG abnormalities: the
continuing debate. Brain Dev 32(9):746–752
Ekinci O, Arman AR, Isik U, Bez Y, Berkem M (2010) EEG abnormalities and epilepsy in
autistic spectrum disorders: clinical and familial correlates. Epilepsy Behav 17(2):178–182
Findji F, Harrison-Covello A, Lairy GC (1979) Long duration EEG studies in the case of a
psychotic child. Electroencephalogr Clin Neurophysiol 46(5):592–600
Frye RE, Butler I, Strickland D, Castillo E, Papanicolaou A (2010) Electroencephalogram
discharges in atypical cognitive development. J Child Neurol 25(5):556–566
Giannotti F, Cortesi F, Cerquiglini A, Miraglia D, Vagnoni C, Sebastiani T, Bernabei P (2008)
An investigation of sleep characteristics, EEG abnormalities and epilepsy in developmentally
regressed and non-regressed children with autism. J Autism Dev Disord 38(10):1888–1897
Gillberg C (1991) The treatment of epilepsy in autism. J Autism Dev Disord 21(1):61–77
Gillberg C, Schaumann H (1983) Epilepsy presenting as infantile autism? Two case studies.
Neuropediatrics 14(4):206–212
Gubbay SS, Lobascher M, Kingerlee P (1970) A neurologic appraisal of autitic children: results
of a western Australian survey. Dev Med Child Neurol 12:422–429
Hara H (2007) Autism and epilepsy: a retrospective follow-up study. Brain Dev 29(8):486–490
Hashimoto T, Sasaki M, Sugai K, Hanaoka S, Fukumizu M, Kato T (2001) Paroxysmal
discharges on EEG in young autistic patients are frequent in frontal regions. J Med Invest
48(3–4):175–180
Hollander E, Dolgoff-Kaspar R, Cartwright C, Rawitt R, Novotny S (2001) An open trial of
divalproex sodium in autism spectrum disorders. J Clin Psychiatry 62(7):530–534
Hollander E, Soorya L, Wasserman S et al (2006) Divalproex sodium vs. Placeboin the treatment
of repititive behaviours in autismspectrum disorders. Int J Neuropsychopharmacol
9(2):209–213
Hrdlicka M. Komarek V. Propper L. Kulisek R. Zumrova A. Faladova L. Havlovicova M.
Sedlacek Z. Blatny M. Urbanek T (2004) Not EEG abnormalities but epilepsy is associated
with autistic regression and mental functioning in childhood autism. Eur Child Adolesc
Psychiatry 13(4):209–213
Hughes JR, Melyn M (2005) EEG and seizures in autistic children and adolescents: further
findings with therapeutic implications. Clin EEG Neurosci 36(1):15–20
References 147

Kagan-Kushnir T, Roberts SW, Snead3rd OC (2005) Screening electroencephalograms in autism

spectrum disorders: evidence-based guideline. J Child Neurol 20(3):197–206
Kanner L (1943) Autistic disturbances of affective contact. Nervous Child 10:217–250
Kayaalp L, Dervent A, Satik S et al (2007) EEG abnormalities in West syndrome: correlation
with the emergence of autistic features. Brain Devt 29(6):336–345
Lewine JD (1999) Magnetoencephalographic patterns of epileptiform activity in children with
regressive autism spectrum disorders. Pediatrics 104(3):405–418
Levisohn PM (2007) The autism-epilepsy connection. Epilepsia 48 (Suppl, 9):33–35
Merhar SL, Manning-Courtney P (2007) Two boys with 47, XXY and autism. J Autism Dev
Disord 37(5):840–846
Nass R, Gross A, Devinsky O (1998) Autism and autistic epileptiform regression with occipital
spikes. Dev Med Child Neurol 40:453–458
Parmeggiani A, Barcia G, Posar A, Raimondi E, Santucci M, Scaduto MC (2010) Epilepsy and
EEG paroxysmal abnormalities in autism spectrum disorders. Brain Dev 32(9):783–789
Patry G, Lyagoubi S, Tassinari CA (1971) Subclinical ‘‘electrical status epilepticus’’ induced by
sleep in children. Arch Neurol (Chicago) 24:242–252
Pressler RM, Robinson RO, Wilson GA, Binnie CD (2005) Treatment of interictal epileptiform
discharges can improve behavior in children with behavioral problems and epilepsy. J Pediatr
146:112–117
Rapin I (1995) Acquired aphasia in children. J Child Neurol 10(4):267–270
Reinhold JA, Molloy CA, Manning-Courtney P (2005) Electroencephalogram abnormalities in
children with autism spectrum disorders. J Neurosci Nurs 37(3):136–138
Rossi PG, Parmeggiani A, Bach V, Santucci M, Visconti P (1995) EEG features and epilepsy in
patients with autism. Brain Dev 17(3):169–174
Small JG (1975) EEG and neurophysiological studies of early infantile autism. Biol Psychiatry
10:385–397
So EL (2010) Interictal epileptiform discharges in persons without a history of seizures: What do
they mean? J Clin Neurophysiol 27(4):229–234
Stefanatos GA, Kinsbourne M, Wasserstein J (2002) Acquired epileptiform aphasia: a
dimensional view of Landau-Kleffner syndrome and the relation to regressive autistic
spectrum disorders. Child Neuropsychol. 8(3):195–228
Tharp BR (2004) Epileptic encephalopathies and their relationship to developmental disorders:
do spikes cause autism? Mental Retard Dev Disabil Res Rev 10(2):132–134
Tuchman RF, Rapin I (1997) Regression in pervasive developmental disorders: seizures and
epileptiform electroencephalogram correlates. Pediatrics 99(4):560–566
Yasuhara A (2010) Correlation between EEG abnormalities and symptoms of autism spectrum
disorder (ASD). Brain Dev 32(10):791–798
 Part IV
Controversial EEG Waveforms
Chapter 15
Introduction to Controversial Sharp
Waves or Spike Patterns

The designation ‘‘controversial EEG wave forms’’ refers to a number of EEG

patterns that tend to be more frequently detected in psychiatric populations and
their association with epilepsy is weak if at all. Included under this category are
the Small Sharp Spikes (SSS), the Mitten pattern, the 6–7 and 14 positive spikes
(PS), the six per second spike and waves (6/s SpW), Wicket Spikes (WS), and the
Rhythmic Mid-Temporal Discharges (RMTD). It is of great interest that these
waveforms have generated such heated arguments and almost all of them have two
designations indicative of whether or not the speaker believes they have any
clinical relevance. For example, using the term Benign Epileptiform Transients of
Sleep (BETS) to refer to the SSSs would indicate that the speaker does NOT
believe that this wave form has any clinical relevance. For more expanded dis-
cussion please refer to Hughes and Wilson (1983) and Hughes (1994). In the next
four chapters examples of these patterns are given. The most important message is
that these patterns are relatively more difficult to detect particularly to the unex-
perienced or hurried EEG reader. Additional examples of these patterns can be
found in Gibbs and Gibbs (1964).
These six EEG patterns are consistently observed to be more prevalent in
psychiatric populations than either healthy or non-psychiatric patients control-
populations (Boutros et al. 2011). Despite the increased prevalence in psychiatric
patients, defining the patterns’ exact neurobiological basis and clinical correlates
has proved to be an elusive goal.
Three of the six patterns were described in full in Hughes and Wilson (1983)
volume; EEG and Evoked Potentials in Psychiatry and Behavioral Neurology.
Given that research in this area has all but stopped. Those chapters remain current.
We essentially provide updates on the three patterns (Small Sharp Spikes, 6–7 and
14 Positive spikes, and the 6/s Spike and Wave). Between the chapters in this book
and the chapters in the Hughes and Wilson Volume, all issues related to these
waveforms are covered. The reader is thus encouraged to try to acquire this
excellent and historic volume.
For reasons which are unclear, basic clinical and empirical research focused on
the clinical relevance of the various controversial EEG patterns has declined since
the mid-1970s and now seems to have been all but abandoned. This neglect is

N. N. Boutros, Standard EEG: A Research Roadmap for Neuropsychiatry, 151

DOI: 10.1007/978-3-319-04444-6_15,  Springer International Publishing Switzerland 2013
152 15 Introduction to Controversial Sharp Waves or Spike Patterns

unfortunate because much of the past work appears to point strongly to several
clinically relevant symptomatic correlates that stand a reasonable chance of being
substantiated if subjected to additional methodologically well-controlled investi-
gations. In the following four chapters, each of these patterns is discussed in detail.
Developing these chapters is allowed by the fact that a significant literature
examining the psychiatric correlates of these patterns exist.

Unusual but Little Studied in Psychiatric

Populations Patterns

A host of other EEG patterns labeled ‘‘unusual’’ have not been examined in
psychiatric populations. The interested reader is referred to a comprehensive
review and examples of these patterns (Westmorland and Klass 1990). As of now,
none of these patterns is included under the ‘‘controversial’’ category but the
physiological or pathological correlates of these patterns are not known and
whether they are more represented in psychiatric populations is also not known.
Here, we briefly mention the patterns to allow an interested investigator to begin
probing the nature and psychiatric correlates if they so chose.
Westmoreland and Klass (1990) classify these ‘‘unusual’’ patterns based on the
EEG frequency range they exhibit (e.g., alpha, beta…). The ‘‘Squeak’’ phenom-
enon represents a decreased reactivity of alpha frequency to eye opening (an
otherwise universal phenomenon). The observation was originally made by Bek-
kering et al. (1956) and has not been examined in psychiatric populations.
Retained alpha activity refers to alpha activity that persists in a focal area after
sleep. This pattern was originally described by Gibbs and Gibbs (1964) who
reported that it can be seen close to the site of a tumor, focal vascular lesion, a
skull defect, or the site of a traumatic brain injury.
Extreme spindles (Fig. 15.1) is another ‘‘unusual’’ pattern that deserves a special
mention. Again described by Gibbs and Gibbs (1962, 1964) and refers to a pattern of
sleep spindles that is continuous almost replacing the usual more episodic sleep
spindles seen in stage II sleep. This activity is seen maximally over the frontal-
central region and can wax and wane to some extent with frequencies ranging from 6
to 18 Hz. Extreme spindles have a more diffuse distribution and slightly higher
amplitudes compared to normal spindles. Extreme spindles are seen mainly in
children under five but have been reported in adults. The main clinical correlate
reported is an association with mental retardation. Individuals who show extreme
spindles during sleep may also exhibit the pattern during wakefulness. The pattern
can also be seen in association with medications that induce fast activity like ben-
zodiazepines and barbiturates. This raises another interesting question regarding
whether or not the appearance of this pattern in association with these anxiolytic
medications represents some form of a biological marker. Of great interest is that
Unusual but Little Studied in Psychiatric Populations Patterns 153

Fig. 15.1 This tracing was obtained from an intellectually disabled individual during sleep. It
shows an almost continuous sleep spindle termed ‘‘extreme spindle’’. (with permission from
Hughes 1994)

electrophysiological correlates of the many subgroups of mentally retarded indi-

viduals have not been seriously investigated.
A number of ‘‘unusual’’ patterns have faster frequencies (more in the beta
range). The fast alpha variant ranges between 16–20 Hz (Chatrian et al. 1974) and
co-exists with normal alpha and reacts similarly to eye-opening. The posterior
temporal fast activity in children was described in young children (1–3 years of
age) with a frequency mainly in the 20–26 Hz range and appearing in response to
novel or complex stimuli in the awake child (Chaloner and Pampiglione 1983).
The fast spiky spindles variant was detected in children with organic brain
pathology (Niedermeyer and Capute 1967). The pattern consists of sharply con-
toured sleep spindles during stages 2 and 3 with frequencies mainly in the
16–20 Hz but at times also 20–36 Hz.
A number of ‘‘unusual’’ patterns are in the theta range. The frontal arousal
rhythm is occasionally seen in children after arousal from sleep. The pattern
consists of trains of 7–10 Hz ranging from few seconds to 15–20 s runs. The
pattern usually disappears when the child is fully awake. This pattern was
observed in children with minimal brain dysfunction (White and Tharp 1974), but
has also been described in children with no identifiable brain damage. The midline
theta activity and the 4 Hz vertex rhythm are also of unknown clinical significance
(Westmoreland and Klass 1986; Daoust-Roy 1989).
Finally, some ‘‘unusual’’ patterns have delta range frequencies. The posterior
rhythmic slow-wave activity with eye-closure consists of brief trains of 2–3 Hz
154 15 Introduction to Controversial Sharp Waves or Spike Patterns

rhythmic, bilateral, and lasting 1–3 s that are seen over the posterior head regions
after eye closure. This pattern was originally called the Phi rhythm (Klass and
Fisher-Williams 1976) and is frequently associated with other EEG abnormalities
(PeBenito et al. 1983). This pattern has been described in association with a
number of neuropsychiatric conditions like Tourette syndrome and head injury but
has not been seriously investigated in psychiatric populations. Westmoreland and
Klass (1990) conclude that the pattern should be considered pathological but
nonspecific. This is in contrast to the subclinical rhythmic electrographic discharge
of adults (SREDA) which is considered to be of no clinical significance. It occurs
in adults over the age of 50 during wakefulness (Miller et al. 1985), and less
commonly during drowsiness and sleep. It consists of sharply contoured theta runs
of 5–7 Hz frequency with a widespread distribution but maximal over the parietal
and temporal regions. The duration of the episodes could be as short as few
seconds but could persist for few minutes. In an individual subject who exhibit
SREDA, the pattern can be seen in serial recordings. It would not be surprising if
the presence of this pattern signals some form of susceptibility or proneness to
some form of pathological brain responses.
EEG waveforms can be considered ‘‘controversial’’ for various reasons. In his
excellent review Hughes (1996) has suggested that the infrequent appearance of a
number of waveforms in EEGs obtained from many laboratories may have con-
tributed to the ‘‘controversial’’ designation eventually given to a particular wave
form. He notes, quite correctly, that electroencephalographers tend to be skeptical
about reported EEG findings that they do not (or only rarely) encounter in their
own laboratories. Certainly the incidence of these waveforms in populations of
normals, as well as many clinical groups, is known to be very low. Furthermore,
the appearance of a number of these waveforms in an EEG tracing is very heavily
dependent upon securing a drowsy and sleep tracing and not all laboratories utilize
sleep activation on a routine basis. These waveforms tend to be of low amplitudes
and because of this it may be difficult or impossible to record with several com-
monly used bipolar recording montages. Thus, it is not unreasonable to suspect
that at least some portion of those EEGs failing to detect these waveforms con-
stitute false negative tracings for these relatively little studied EEG findings.
All of the above considerations can conspire to reduce the visibility of these
waveforms in many laboratories. However, neither the infrequent occurrence of
these waveforms, their absence in the material from many laboratories, nor its low
incidence in several clinical populations constitute a reasonable basis for con-
cluding that they are without clinical relevance. With appropriate recording
montages, sleep activation, and a familiarity with the wave form, all of the so-
called controversial waveforms are in fact quite easily detected and can be sub-
jected to clinical investigation. The following four chapters review the evidence
for and against the clinical relevance of each of these waveforms and highlight the
research that remains necessary to fully understand the nature and any implications
these patterns may have.
Until the correlates of these patterns (physiological or pathological) are iden-
tified the assertion that these patterns are completely irrelevant to neuropsychiatric
Unusual but Little Studied in Psychiatric Populations Patterns 155

conditions is not fully supportable. At a minimum the appearance of such patterns

may give support to the organic nature of presenting symptoms. This assertion, in
our clinical experience, goes a long way in strengthening the alliance between the
patient and the treater against the diagnosed disorder as compared to the
assumption that patient is in control of the symptoms which tends to put the patient
and the treaters at opposite sides of the therapeutic effort. Whether the appearance
of these patterns predicts favorable response to anticonvulsant treatment is not
known and awaits well-designed prospective controlled studies. Until such time
patients harboring one of these patterns, and who have not responded well to the
standard treatment of the diagnosed disorder, should be given the benefit of the
doubt and allowed a trial of anticonvulsant treatment. If a clinician decides to
pursue this option, he/she must Keep in mind that failure to respond to one anti-
convulsant agent does not necessarily mean that no other anticonvulsant will be
beneficial.
The recent advances in the ability to localize the neural source generators of
activity recorded on the scalp can be illuminating once the clinical correlates of
these controversial patterns are elucidated. Further probing into the clinical cor-
relates of these EEG patterns can be facilitated by recent developments in stan-
dardized psychiatric rating scales and methods of examination of symptom-
clusters in psychiatric populations. The reader should be made aware that the use
of different terminologies to refer to the controversial waveforms complicates the
investigating of these waveforms.

What does the Co-occurrence of These Waveforms

Suggest?

If we assume that the six controversial patterns are rare incidental findings without
clinical relevance, the co-occurrence of two (and at times more) of these patterns
should indeed be extremely rare. Previous studies have indicated that patients
showing PS at times may also show two other controversial patterns: RMTD and/
or 6/s spike and wave complexes (Gibbs and Gibbs 1964). Assuming these are
incidental and unrelated findings, the chance for the co-appearance of these three
patterns in the same subject would equal the value resulting from multiplying the
incidences of the three patterns (in this case it would be 0.0000002). One example
of the former is a patient of Anderson and Vanderspek (1974) in status epilepticus
who showed the ‘‘psychomotor variant’’ pattern, who also showed PS. Silverman
(1967) collected a few cases of 6/s spike and wave and many with PS but, after
reinterpretation, concluded that the great majority of these patients had both
phenomena. Silverman claimed that the distribution of PS and 6/s SpW were
identical with the posterior quadrants (especially the posterior temporal area)
showing the maximal deflections. This author did point out that at times 6/s spike
and wave was maximal anteriorly, however, especially in patients who have
156 15 Introduction to Controversial Sharp Waves or Spike Patterns

clinical seizures, and thus suggested that the pattern with an anterior emphasis was
really a different pattern from the one maximal on the posterior region (see Chap.
18). Silverman had therefore suggested that the two patterns of PS and 6/s spike
and wave complexes are intimately related and that all the differences previously
noted between the two waveforms are related to the ‘‘maturational sequence’’ of
the PS. Although the two patterns have similarities and occasionally difficulty
ensues in differentiating the two, considerable differences do exist between these
two waveforms. For example, PS always show a maximal deflection on the pos-
terior temporal area, whereas the 6/s spike and wave complex is maximal on the
occipital areas or frontal regions. PS at any given moment are usually unilateral
while 6/s spike and wave complexes are usually bilateral and, in addition photic
stimulation, almost never elicits positive spikes, not infrequently will produce the
spike and wave complex. The age distribution, of course, is very different in that
the PS is a teenage phenomenon while the 6/s spike and wave is usually seen in
adults. More males are found with PS, while females predominate in the spike and
wave pattern. Finally, the symptomatology is different especially with regard to the
absence of clinical seizures in patients with PS and the not infrequent presence (at
least 36 %) in those with 6/s spike and wave pattern. The conclusion of the
reviewer is that the two patterns of the positive spikes and the 6/s spike and wave
may be intimately related, but they are likely not two variants of the same basic
waveform.
Currently, the rates of the co-occurrence or the clinical significance of the co-
occurrences are completely unknown. A more than chance co-occurrence of two or
more of these patterns would suggest a specific correlate whether physiological or
pathological.
In the following chapters the term ‘‘dysthrythmia’’ was avoided as much as
possible in favor for the term ‘‘pattern’’ to indicate the author’s neutrality toward
these phenomena. What is needed is well-collected prospective data in order to
define the correlates of these rather well-defined EEG patterns. Such studies should
be conducted by EEGers fully trained (to criteria) in detecting these patterns. All
EEG tracings should be examined by investigators blinded to the clinical or group
membership of the subjects and with at least two so qualified EEGers
independently.
Furthermore, a complete evaluation and characterization of the clinical syn-
drome being examined (as much as the state of knowledge allows at the time)
should be done to allow later correlational analyses with different groupings of the
various symptoms. Furthermore, it would be greatly beneficial to the field if cli-
nicians were to keep track of the efficacy of the various antiepileptic medications
when prescribed to individuals exhibiting one of these patterns. Either via case
reports, case series or a depot for such reports (e.g., a website for this purpose)
such data could accumulate rather fast and would lead to more controlled pro-
spective and definitive studies.
A Hypothesis to be Tested 157

A Hypothesis to be Tested

Increased cortical excitability is the hallmark of epileptic tissue. It is possible that

the various ‘‘controversial’’ patterns reflect some forms or degrees of increased
brain tissue excitability at the site of origin of the particular pattern. A number of
methodologies have now advanced enough to allow testing this hypothesis. Such
methodologies include transcranial magnetic stimulation, magnetoencephalogra-
phy, and magnetic resonance spectroscopy.

References

Anderson RL, Vanderspek HG (1974) Psychomotor variant status epilepticus. Clin Electroenceph
5:129–132
Bekkering D, Kamp A, Storm van Leeuwen W, Werre PF (1956) Example of use of the
spectrograph and the magnetograph the ‘‘Squeak’’ phenomenon. Electroencephalogr Clin
Neurophys 8:721(Abstract)
Boutros NN, Galderisi S, Pogarell O, Riggio S (2011) Standard elelctroencephalography in
clinical psychiatry. Wiley-Blackwell, Hoboken, pp 59–76
Chaloner J, Pampiglione G (1983) ‘‘Posterior temporal fast’’ EEG activity in childhood. Rev
Electroencephalogr Neurophysiol Clin 13:53–60
Chatrian GE, Bergamini L, Dondey M, Klass DW, Lennox-Buchthal M, Petersén I (1974) A
glossary of terms most commonly used by clinical electroencephalographers. Electroencep-
halogr Clin Neurophys 37:538–548
Daoust-Roy J (1989) A waking 4 Hz vertex rhythm: 4 cps vertex spindles re-visited. Am J EEG
Technol 29:147–163
Gibbs EL, Gibbs FA (1962) Extreme spindles: correlation of electroencephalographic sleep
pattern with mental retardation. Science 138:1106–1107
Gibbs FA, Gibbs EL (eds) (1964) Atlas of electroencephalography, vol III. Addison-Wesley
Publishing Company, Reading
Hughes JR (1994) EEG in clinical practice, 2nd edn. Butterworth-Heinemann, Boston
Hughes JR (1996) A review of the usefulness of the standard EEG in psychiatry. Clin EEG
27:35–39
Hughes JR, Wilson WP (1983) EEG and evoked potentials in psychiatry and behavioral
neurology. Butterworths, Boston
Klass DW, Fisher-Williams M (1976) Sensory stimulation, sleep and sleep deprivation. In:
Redmond A (ed) Handbook of electroencephalography and clinical neurophysiology, vol III.
Elsevior Scientific Publishing Company, Amsterdam, pp 5–73
Miller CR, Westmoreland BF, Klass DW (1985) Subclinical rhythmic EEG discharge of adults
(SREDA): further observations. Am J EEG technol 25:217–224
Niedermeyer E, Capute AJ (1967) A fast and spiky spindle variant in children with organic brain
disease. Electroencephalogr Clin Neurophys 23:67–73
PeBenito R, Cracco JB, Raymond CO, Knochler R (1983) The clinical significance of
paroxysmal rhythmic postyerior slow patterns. Electroencephalogr Clin Neurophys 56c:10P
Silverman D (1967) Phantom spike-waves and the fourteen and six per second positive spike
pattern: a consideration of their relationship. Electroencephalogr Clin Neurophys 23:207–213
Westmoreland BF, Klass DW (1986) Midline theta rhythm. Arch Neurol 43:139–141
Westmoreland BF, Klass DW (1990) Unusual EEG patterns. J Clin Neurophys 7(2):209–228
White JC, Tharp BR (1974) An arousal pattern in children with organic cerebral dysfunction.
Electroencephalogr Clin Neurophys 37:265–268
Chapter 16
Psychiatric Correlates of the B-Mitten
EEG Pattern

Introduction

Although the B-Mitten EEG wave form was described more than 60 years ago, it
failed to attract significant attention from within either psychiatry or clinical
neurology. There are three basic reasons for this. First, like many other contro-
versial paroxysmal discharges described over the years, the Mitten wave form
failed to display a clear relationship to neurological disease. Second, the behav-
ioral correlates which came to be postulated for this EEG finding were of insuf-
ficient specificity for use in psychiatric diagnoses. Third, the B-Mitten pattern can
only be recorded with a monopolar referential montage during relatively deep
sleep stages. Most EEG laboratories do not record long enough to reach the sleep
stages much beyond drowsiness and stages one and two.

The B-Mitten EEG Wave Form

The initial literature reports of B-Mittens in association with psychiatric dys-

function, primarily psychosis, began appearing more than 60 years ago (Lyketsos
et al. 1953; Gibbs et al. 1960; Halasz and Kajtor 1962; Gibbs and Gibbs 1963,
1964; Halasz and Nagy 1965). B-Mittens can only be detected in the EEG during
sleep recordings, particularly moderately deep to deep sleep levels (stages III–IV).
A referential monopolar linked ear montage is also essential for recording Mittens
because anterior to posterior bipolar runs and transverse linkages between
homologous electrodes will invariably cancel out the wave form and prevent its
registration. This finding was given the name ‘‘Mittens’’ because the wave form is
composed of a sharp transient of 1/10–1/12 s duration followed by higher voltage
slow wave which creates a complex resembling the outline of the ‘‘thumb and
hand’’ of a child’s mitten. Mittens occur bilaterally and synchronously over the
frontal-central cortex with the best representation of the wave form appearing at
prefrontal electrode sites (Figs. 16.1 and 16.2). Mittens appear to have an intimate

N. N. Boutros, Standard EEG: A Research Roadmap for Neuropsychiatry, 159

DOI: 10.1007/978-3-319-04444-6_16,  Springer International Publishing Switzerland 2013
160 16 Psychiatric Correlates of the B-Mitten EEG Pattern

Fig. 16.1 A-Beta Mittens pattern recorded from a 23-year-old woman diagnosed with
schizoaffective disorder

connection to sleep spindles and not infrequently the ‘‘thumb’’ component of the
complex is made up of the last wave of a frontal sleep spindle.
Two less common subtypes of the Mitten wave form have been identified based
on the frequency of the initial ‘‘thumb’’ component of the complex. The A-Mitten
has a ‘‘thumb’’ duration of 8–9 Hz making it slightly ‘‘slower’’ than the more
common and more extensively studied B-Mitten. Other than that the two types are
nearly identical in appearance and the frequency of the thumb component must be
measured to make the distinction. A-Mittens are not associated with psychiatric
symptomatology. Instead, early reports of the A-Mitten subtype (Winfield and
Sparer 1954) found a moderate association between this wave form and Parkin-
son’s Disease. For reasons little understood, this interesting lead has never been
followed up. A still slower Mitten, the A-1 Mitten, has an initial thumb component
duration of less than 8 Hz and when it is seen it is usually in association with
pathology involving deep brain structures, particularly with tumors involving the
thalamus or adjacent regions (Gibbs and Gibbs 1964).
The interEEG reliability of the B-Mitten wave form has been assessed on two
occasions (Struve and Becka 1968; Struve et al. 1972a) using blind and inde-
pendent readings of unmarked EEGs and in both instances the results were sta-
tistically adequate. Furthermore in an additional study (Kane et al. 1978), 12
tracings with Mittens and 13 tracings without Mittens were disguised by use of
The B-Mitten EEG Wave Form 161

Fig. 16.2 B-Mittens recorded from a 27-year-old woman with a mood disorder

code numbers and then reinterpreted after many months and there was complete
intrainterpreter agreement between the two sets of interpretations. The test–retest
reliability of the B-Mitten wave form over time has also been assessed and found
to be adequate (Struve et al. 1972a).

Incidence and Age Distribution

B-Mittens are only rarely encountered in the EEGs of normal individuals. Gibbs and
Gibbs (1964) obtained moderately deep sleep EEG recordings from over 900 normal
adults and reported a Mitten incidence of less than 3 %. They also reported a total
absence of the Mitten wave form in a sample of 2,000 normal controls below 20 years
of age. Although the control subjects in their series were described as ‘‘being free of
any significant disease or complaint,’’ it is doubtful that they were rigorously
screened by today’s standards for all current symptoms or past history of psychiatric
disorder (Boutros et al. 2005). As is shown below, B-Mittens appear to be confined to
psychiatric populations and are essentially absent among normal individuals.
The B-Mitten wave form occurs almost exclusively among adults. According to
Gibbs and Gibbs (1964), 93 % of all patients with Mittens are adults and this is in
162 16 Psychiatric Correlates of the B-Mitten EEG Pattern

agreement with the report a year later by Halasz and Nagy (1965). In a large
investigation of the age distribution of the B-Mitten wave form (Struve et al.
1973), 850 EEGs containing Mittens were drawn from a sample of 41,000 EEGs
secured from all age groups from childhood through geriatric subjects. Mittens
were virtually absent among children (0.12 % incidence below age 10), and begin
to appear gradually during adolescence (0.7 % incidence age 10–14; 8.0 % inci-
dence age 15–19) with over 75 % of the cases occurring in patients between 20
and 50 years of age. The Mitten incidence begins to show a decline in the fourth
decade with the decline becoming more dramatic in older individuals. Thus it is
essential for reports on the incidence of this pattern to clearly indicate the age
distribution of the included sample with the full expectation that inclusion of
children or elderly as well as failure to secure tracings during deep sleep will all
contribute to a decreased incidence in the study sample.

Brain Sources

Several lines of evidence converge to place the sources of the B-Mitten signal in
subcortical as opposed to cortical areas. As was mentioned, the Mitten morphology
and scalp distribution appear linked to that of thalamus-mediated sleep spindles. A
subcortical location is also supported by the earliest reports (Winfield and Sparer
1954) of the Mitten wave form (the A-Mitten subtype) occurring with a moderate
56 % incidence in Parkinson’s Disease. Others (Halasz and Nagy 1965) com-
mented that the scalp distribution of the B-Mitten pattern was compatible with a
signal location in mid-line, diencephalic nonspecific systems. In an influential
study (Gibbs and Gibbs 1973), the sleep EEGs of 261 patients with brain tumors
occurring in a variety of cortical and subcortical locations were analyzed for the
presence of the B-Mitten wave form. None of the cortical tumors were associated
with Mittens but subcortical neoplasms were frequently accompanied by this EEG
signal. Tumor locations associated with the highest incidence of B-Mittens were:
thalamus (66.7 %), ventricle (37.5 %), mid brain (33.3 %), and corpus callosum
(30 %). At this juncture it must be stressed that while tumors in certain brain
regions may be associated with the production of Mitten wave forms, the converse
is not at all true and less than 3 % of patients with B-Mittens have verified tumors
(Gibbs and Gibbs 1964).

Clinical Symptomatology

Early Studies

The B-Mitten wave form appears to occur almost exclusively among psychiatric
patients. The early series of papers describing this wave form (Lyketsos et al.
Clinical Symptomatology 163

1953; Gibbs et al. 1960; Halasz and Kajtor 1962; Gibbs and Gibbs 1963, 1964;
Halasz and Nagy 1965; Simon and DeVito 1976; Gibbs and Novick 1977) were
published years before modern day diagnostic criteria and symptom rating scales
became available. Consequently, these early papers reported findings in the lan-
guage and concepts typical of that era and it is not easy to extrapolate parallels to
the nosology and nomenclature used today. Furthermore, a perplexing range of
psychiatric conditions including, among others, sociopathic personality, alcoholic
psychosis, involutional depression, mania, paranoia, and schizophrenia were said
to be associated with the Mitten dysrhythmia. Halasz and Nagy (1965) summa-
rized their series of studies conducted in Hungary by stating somewhat cryptically
that B-Mittens were associated with ‘‘recidival neurotic, eventually psychotic
decompensations’’ which seemed to imply some kind of exacerbating clinical
course. Finally, two early papers (Tasher et al. 1970; Olson et al. 1970) reported a
high incidence of B-Mittens (39–40 %) in the EEGs of incarcerated criminals,
particularly those in a maximum security setting, but this finding seems primarily
related to the wide variety of underlying psychiatric diagnoses in this population
and not to the type of crime committed.
Taken as a whole, the early studies document a strong association between B-
Mittens and ‘‘psychiatric dysfunction.’’ However, this literature is disappointing in
its inability to substantiate relationships between the Mitten wave form and either
diagnostic categories on the broad end or more circumscribed symptoms or
symptom clusters on the narrow end. Although the highest incidence of Mittens
(42 %) was found among a patient population that Gibbs’s referred to as ‘‘epi-
leptics with psychosis’’ (Gibbs and Gibbs 1964), this interesting observation
becomes perplexing when one considers that Mittens correlate very poorly, if at
all, with pure seizure disorders. The early literature makes frequent reference to
‘‘schizophrenia’’ or ‘‘psychosis’’ and this provides a suggestion that Mittens might
be more strongly associated with more serious, as opposed to mild, psychiatric
disturbance. However, even this observation could reflect a methodological artifact
in that (1) all things being equal, psychiatric patients are more likely to be referred
for EEG study if their condition is viewed as serious and (2) almost all of the early
studies were conducted with inpatient hospitalized psychiatric patients which
would imply a heavy loading with psychosis and other major psychiatric distur-
bance. Perhaps a more profitable approach to assessing the clinical significance of
the B-Mitten pattern (and for that matter all other controversial wave forms) is to
examine the correlation with specific symptoms or symptom clusters (rather than
diagnoses) that may cut across broad diagnostic categories.

Affective Dysregulation

Beginning in the mid-1950s considerable research attention was focused on the

dichotomization of schizophrenia into ‘‘process’’ and ‘‘reactive’’ subtypes as a way
of resolving the heterogeneity of this disorder (Phillips 1953; Garmezy and Rodnick
1959; Herron 1962). ‘‘Process schizophrenics’’ were said to be characterized by an
164 16 Psychiatric Correlates of the B-Mitten EEG Pattern

insidious onset of the thought disorder with psychotic symptoms appearing in late
adolescence with little evidence of a precipitating event and a history of poor
premorbid social and sexual adjustment. Clinically, the psychotic presentation was
without affective turmoil. In contrast ‘‘reactive schizophrenia’’ emerged rapidly
during early adulthood, presented with considerable affective turmoil and often
followed a precipitating event superimposed upon a good premorbid history. A
pilot study (Struve and Becka 1968) found that 72.7 % of the reactive schizo-
phrenics displayed the Mitten wave form whereas only 10 % of the process
schizophrenics did so (p = 0.006). In this study premorbid history and EEG were
assessed blindly and without cross-communication and there were no significant
differences between the two groups on age, education, time between current
admission and testing, and total lifetime months of psychiatric hospitalization. A
larger second study (Struve et al. 1972b) using the same procedures as well as the
additional safeguards of equating process and reactive groups on intelligence,
chronicity, length of medication free state (minimum of 4 weeks), and total amount
of sleep EEG recording continued to demonstrate a strong (p = 0.0001) association
between B-Mittens and reactive schizophrenia.
Despite the above findings there were still a troubling number of reactive
schizophrenics without this EEG signal. Furthermore, Mittens were not specific to
reactive schizophrenia because they could be seen in 26.8 % of nonschizophrenics
(Struve et al. 1972a). Several clinical observations suggested that B-Mittens were
more closely related to affective components of illness than to schizophrenia.
Because of this question another study was conducted (Struve et al. 1972b) in
which 121 patients aged 19 and above (both schizophrenic and nonschizophrenic)
were divided into those with a primary and/or secondary diagnosis of either
schizoaffective psychoses, affective disorder, or depression versus those without
diagnostic indications of affective dysregulation and a two-way analysis of vari-
ance for proportions was performed using dichotomizing variables of (a) process-
reactive premorbid history and (b) presence versus absence of diagnoses indicative
of primary or secondary affective dysregulation. The results demonstrated sig-
nificant main effects for the positive association between both reactive premorbid
history and affective symptomatology and presence of the B-Mitten wave form.
In a later study (Struve and Klein 1976) a sample of 49 patients with definite B-
Mitten EEGs were assessed (blind to the EEG) for primary or secondary diagnostic
indications of affective dysregulation. While a primary diagnosis of affective
dysregulation occurred in 65.3 % of the cases, fully 85.7 % of the patients had a
significant affective component to their illness when secondary diagnoses of
affective dysregulation were included. Only 14 % of the B-Mitten patients were
without an affective overlay to their clinical presentation. In this same study
(Struve and Klein 1976), the relationship of B-Mittens to effective coloring was
further emphasized by considering the primary diagnosis of Character Disorder—a
diagnosis traditionally unassociated with either excess or inhibition of affective
factors. When 41 Character Disorder patients were divided by (a) presence or
absence of B-Mittens and (b) presence or absence of secondary diagnosis of
affective dysregulation it was found that 81.8 % of those with Mittens had a
Clinical Symptomatology 165

secondary affective component to their illness whereas only 25.8 % of those

without Mittens did so and the relationship of B-Mittens to affective dysregulation
within this primary diagnostic grouping was highly significant (p = 0.0018).
In summary B-Mittens wave forms do appear to be strongly associated with
symptoms of affective dysregulation, especially depression and dysphoric affect, in
ways that cut across broad diagnostic classifications. A situation seems to exist
where the presence of B-Mittens allows a fairly confident (roughly 86 %) pre-
diction that affective components will be found in the clinical presentation. The
reasons why B-Mittens relate to affective dysregulation spread over such a diverse
range of diagnostic classifications—from schizophrenia to clearly nonpsychotic
conditions with secondary depressive elements—remain obscure. The presumed
subcortical origins of the B-Mitten wave form may be viewed as consistent with
the affective symptoms seen with this EEG finding.

Extrapyramidal Side Effects

Rifkin et al. (1978) reported that paroxysmal EEG dysrhythmias were significantly
associated with the onset of EPS following procyclidine withdrawal in neurolep-
tically treated patients. In a subsequent study (Struve 1987), patients were
examined for EPS by specially trained examiners using the Simpson-Angus scale
for rating EPS symptomatology and all EPS ratings were made without knowledge
of EEG findings. A sample of 216 patients (85 with B-Mittens) were receiving
neuroleptic medication in combination with anti-EPS medication and a second
sample of 135 patients (50 with Mittens) received neuroleptic medication alone.
Subjects in the two comparison groups (who also received EPS examinations)
consisted of 55 patients who had never received neuroleptic medication and 78
patients who were also neuroleptic free but had received such medication in the
past. For the neuroleptically medicated patients, the EEGs were classified into
normal EEG, paroxysmal EEG (no mittens), and B-Mittens as the only finding.
EEGs were not noted for the comparison patients who were not being treated with
neuroleptic medication. Both B-Mittens (p = 0.01) and paroxysmal EEGs
(p = 0.01) were associated with a significantly greater amount of EPS than are
patients with normal EEGs who are also treated with neuroleptics. Similarly,
patients with B-Mitten being treated with neuroleptics alone without adjunctive
use of anti-EPS medication had a significantly (p = 0.025) elevated number of
EPS signs compared to similarly treated patients with normal EEGs.
The significant relationship between the B-Mitten wave form and increased
vulnerability to development of EPS would seem to argue that this controversial
wave form does represent some form of CNS vulnerability as of yet unknown.
Hence it is not likely to be a ‘‘normal’’ EEG wave form. However, although
statistically significant, the magnitude of the observed relationship is not strong
enough to allow individual prediction of EPS risk and the finding may be without
practical clinical relevance.
166 16 Psychiatric Correlates of the B-Mitten EEG Pattern

Tardive Dyskinesia

EEG studies of tardive dyskinesia (TD) have not been common and certainly they
have not focused on the Mitten wave form. The earliest study (Paulson 1968)
reported a 36.7 % incidence of EEG abnormality (generalized or focal slowing or
‘‘epileptiform’’ bursts) in a small sample of patients who were tested following the
diagnosis of TD. However, the reported incidence of EEG abnormality was not
unusual for a psychiatric population and since no control group without TD was
provided thus the finding could not be attributed to TD. In another study (Gardos et al.
1977), abnormal EEG (type unspecified) was one of five variables that contributed to
a discriminant function analysis which significantly identified patients with TD.
Unfortunately when the EEG findings were removed from the discriminant function
and analyzed separately, there were no significant differences in EEG abnormality
between TD and non-TD patients. Simpson et al. (1978) found no difference in the
incidence of abnormal EEGs between patients with and without dyskinesias and
Jeste and associates (1979) reported that EEG abnormalities were unrelated to the
persistence versus reversibility of TD. One team (Syvalahti et al. 1981) did report
seeing Mitten wave forms in some TD patients but interpreted this as no more than a
reflection of the prevalence of this EEG finding in psychiatric patients in general.
In a preliminary attempt (Struve et al. 1979) to study the relationship between the
B-Mitten wave form and early dyskinetic movements (e.g. not sufficient for a full
blown diagnosis of TD), ratings of dyskinetic movements using the Simpson Scale
(Simpson et al. 1979) were made independent of EEG evaluations for 34 normal EEG
and 39 Mitten EEG patients receiving neuroleptic medication. Both EEG groups
were dichotomized into ‘‘short duration’’ neuroleptic exposure (\6 months) versus
‘‘long duration’’ neuroleptic exposure ([6 months) and it was shown that patients
with the Mitten wave form had significantly more dyskinetic symptoms (p = 0.01)
than normal EEG patients for the long-duration neuroleptic condition but no sig-
nificant Mitten EEG effects were found for patients treated for less than 6 months.
Using the exact same methodology, this work was later expanded (Struve et al. 1982)
to include 184 neuroleptically treated patients equally divided between normal EEG
and Mitten EEG subjects. Patients with B-Mittens continued to have significantly
more dyskinetic symptoms (p = 0.025) than normal EEG patients for the long-
duration neuroleptic exposure condition. Again, this effect was not seen for patients
neuroleptically treated for only 6 months or less. The data also subjected to a two-
way analysis of variance with the number of positive signs on the Simpson scale
expressed as a joint function of both EEG category (Mittens vs. Normal) and length
of cumulative neuroleptic exposure (short vs. long). The results of this analysis
showed that there was a significant main effect for B-Mittens (p = 0.014) and, as
would be expected, duration of neuroleptic exposure (p = 0.0001) in producing
increased dyskinetic symptoms. When the analysis was repeated using as covariates
(a) predominant neuroleptic potency (high versus low) and (b) maximum 1 month
sustained dose in CPZ equivalents, a statistically significant main effect for B-Mit-
tens continued to be seen.
Clinical Symptomatology 167

The earliest report of B-Mittens and diagnosed TD was a small pilot study
(Wegner et al. 1977) in which the adequacy of the sleep EEG recording for B-
Mitten detection was assessed in 20 diagnosed cases of TD. Eleven patients had
adequate sleep EEGs and of these 10 (90.9 %) had the Mitten wave form. For three
of the patients, Mittens were recorded 12 months prior to the onset of TD and thus
could not have been a mere concomitant of the disorder. Subsequently (Struve et al.
1979), the sample of diagnosed TD cases with adequate sleep EEGs increased to 34
patients of which 93.1 % of the 29 cases below the age of 40 had the B-Mitten wave
form. It will be recalled from the above discussion of the Mitten age distribution
that the incidence of this finding drops off sharply in the fourth decade.
The exceptionally high incidence of the B-Mitten pattern among diagnosed TD
cases led to a subsequent investigation (Wegner et al. 1979) where a large number
of patients were screened to identify 21 patients with TD who could be success-
fully matched with 21 patients without TD on the variables of (a) age, (b) gender,
(c) diagnosis, (d) total lifetime cumulative neuroleptic exposure, and (e) maximum
neuroleptic dose in chlorpromazine equivalents. The diagnosis of TD was by
examination and required agreement by two independent examiners. Absence of
TD in controls was established by a combination of chart review, interviews with
treating physicians, and independent direct examination by two examiners. All
EEGs were secured and read prior to inception of the study. The results showed
that 20 (95.2 %) of the 21 TD patients had the B-Mitten wave form whereas this
EEG finding was present in only 33.3 % of controls and the association between
Mittens and TD was highly significant (v2 = 17.5259, df = 1, p \ 0.0001).
In further probing the issue (Struve and Willner 1983; Struve 1985), over 750
psychiatric patients ranging in age from 14 to 72 were entered into a 5-year
prospective study of tardive dyskinesia development and followed at approxi-
mately 3 month intervals to determine the relevance of a wide range of risk
variables to the development of this disorder. Not all of the patients were appro-
priate for an analysis of the contribution of the B-Mitten dysrhythmia. Criteria for
identifying a case of either Baseline Persistent TD (e.g., present on entry into the
study) or Prospective Persistent TD (e.g., one that developed during the course of
follow-up) required that the TD diagnosis be confirmed independently by three
examiners. Furthermore the case had to be one of ‘‘Persistent TD’’ and cases of
transient or withdrawal TD were removed from analysis. Control cases had to meet
the stringent criteria of survival without TD or questionable TD for a total
cumulative neuroleptic exposure period preceding TD onset experienced by 90 %
of TD patients (5 years of exposure). Of relevance to this chapter is the finding that
presence of the B-Mitten wave form was significantly associated with Baseline
Persistent TD (p \ 0.03) and prospectively developed Persistent TD (p = 0.04).
168 16 Psychiatric Correlates of the B-Mitten EEG Pattern

Supported Observations

(1) The B-Mitten EEG pattern has not received the degree of attention given to the
other controversial wave forms. Certainly it has never engendered the intense
and prolonged controversy that clouded the 6–7 and 14 per second positive
spike finding or, to a lessor extent, that surrounding rhythmic mid-temporal
discharges (psychomotor variant). In fact, most psychiatrists have never heard
of this EEG finding.
(2) The B-Mitten pattern is a well-defined wave form with a recognized age
distribution and it is easy to interpret given the proper montage and sufficient
depth of sleep activation.
(3) There is no real doubt that the finding is absent among normal individuals and
that when it is seen it is always in association with psychiatric dysfunction.
The lack of a clear correlation with an Axis I or II diagnoses has played a
major role in reducing interest in this interesting EEG signal.
(4) Despite this fact there is rather strong documentation that B-Mittens are
associated with affective dysregulation, primarily dysphoric affect, depression,
and emotional lability. However, the affective correlates to Mittens may be
either primary, secondary, or tertiary components to the clinical presentation
which is another way of saying that the affective correlates cut across broad
Axis 1 classifications.
(5) It would seem compelling that the significant statistical association between B-
Mittens and increased vulnerability to EPS, emergence of involuntary dyski-
netic movements and diagnosed tardive dyskinesia, combined with the
absence of this wave form among normals, would establish this EEG finding
as abnormal.

Open Research Questions

(1) What is the actual incidence in the various psychiatric populations diagnosed
based on most recent criteria and with a consensus diagnosis.
(2) What are the predictive values of detecting this pattern in various psychiatric
groups for treatment with anticonvulsant medications?
(3) What are the cerebral sources of this pattern and would these sources be
amenable to influence via focal treatments like TMS?
References 169

References

Boutros NN, Mirolo HA, Struve F (2005) Normative data for the unquantified EEG: examination
of adequacy for neuropsychiatric research. J Neuropsychiatry Clin Neurosci 17(1):84–90
Gardos G, Cole JO, Labrie RA (1977) Drug variables in the etiology of tardive dyskinesia:
application of discriminant function analysis. Prog Neuropsychopharmacol 1:147–154
Garmezy N, Rodnick EH (1959) Premorbid adjustment and performance in schizophrenia:
implications for interpreting heterogeneity in schizophrenia. J Nerv Ment Dis 129:450–460
Gibbs FA, Gibbs EL (1963) The mitten pattern: an electroencephalographic abnormality
correlating with psychosis. J Neuropsychiat 5:6–13
Gibbs FA, Gibbs EL (1964) Atlas of electroencephalography, vol 3: neurological and psychiatric
disorders. Addison-Wesley, Reading
Gibbs FA, Gibbs EL (1973) Tumor sites in cases of brain tumor with mitten patterns in the
electroencephalogram. Clin EEG 4:206–208
Gibbs FA, Novick RG (1977) Electroencephalographic findings among adult patients in a private
psychiatric hospital. Clin EEG 8:79–88
Gibbs EL, Gibbs FA, Tasher D, Adams C (1960) An electroencephalographic abnormality
correlating with psychosis. Electroenceph Clin Neurophysiol 12:265
Halasz P, Kajtor F (1962) ‘‘Mittens’’- a new form of electroencephalographic wave.(Hung.)
Ideggyogy Szemle 15:46–57
Halasz P, Nagy TA (1965) The mitten pattern- An EEG abnormality in sleep. Acta Med Acad Sci
Hung 21:311
Herron WG (1962) The process-reactive classification of schizophrenia. Psychol Bull 59:329–343
Jeste DV, Potkin SG, Sinha SF et al (1979) Tardive dyskinesia- reversible and persistent. Arch
Gen Psychiat 36:585–590
Kane J, Wagner J, Struve F (1978) The Mitten pattern as a potential EEG predictor of tardive
dyskinesia (proceedings). Psychopharmacol Bull 14(2):35–36
Lyketsos G, Belinson L, Gibbs FA (1953) Electroencephalograms of nonepileptic psychotic
patients awake and asleep. AMA Arch Neurol Psychiat 69:707–712
Olson WH, Gibbs FA, Adams CL (1970) Electroencephalographic study of criminals. Clin EEG
1:92–100
Paulson GW (1968) An evaluation of the permanence of the ‘‘tardive dyskinesias’’. Dis Nerv Syst
24:692–694
Phillips L (1953) Case history data and prognosis in schizophrenia. J Nerv Ment Dis 11:515–525
Rifkin A, Quitkin F, Kane J, Struve FA, Klein DF (1978) Are prophylactic antiparkinson drugs
necessary? A controlled study of procyclidine withdrawal. Arch Gen Psychiat 35:483–489
Simon R, DeVito H (1976) Alcohol activation of electroencephalographic abnormalities in
persons with a history of violence precipitated by drinking alcoholic beverages. Clin EEG
7:145–148
Simpson GM, Varga E, Lee JH et al (1978) Tardive dyskinesia and psychotropic drug history.
Psychopharmacology 58:117–124
Simpson GM, Lee JH, Zoubak B, Gardos G (1979) A rating scale for tardive dyskinesia.
Psychopharmacology 64:171–179
Struve FA (1985) Five-year prospective study of clinical EEG, neuropsychological, and
demographic risk variables for persistent tardive dyskinesia. In: Kemali D, Racagni G (eds)
Chronic Treatments in Neuropsychiatry. Raven Press, New York, pp 33–36
Struve FA (1987) Clinical electroencephalographic variables suggesting extrapyramidal side
effect risk. Clin EEG 18:173–179
Struve FA, Becka DR (1968) The relative incidence of the B-Mitten pattern in process and
reactive schizophrenia. Electroenceph Clin Neurophysiol 24:80–82
Struve FA, Klein DF (1976) Diagnostic implications of the B-Mitten EEG pattern: relationship to
primary and secondary affective dysregulation. Biol Psychiat 11:599–611
170 16 Psychiatric Correlates of the B-Mitten EEG Pattern

Struve FA, Willner AE (1983) A long term prospective study of electroencephalographic and
neuropsychological correlates of tardive dyskinesia: Initial findings at five year follow-up.
Clin EEG 14:186–201
Struve FA, Becka DR, Klein DF (1972a) The B-Mitten EEG pattern and process and reactive
schizophrenia: A replication. Arch Gen Psychiat 26:189–192
Struve FA, Becka DR, Klein DF (1972b) The B-Mitten EEG pattern in process and reactive
schizophrenia and affective states. Clin EEG 3:136–144
Struve FA, Burnett L, Becka DR (1973) Chronological age distribution of the B-Mitten EEG
pattern. Clin EEG 135–139
Struve FA, Kane JM, Wegner JT, Kantor J (1979) Relationship of Mitten patterns to neuroleptic
drug induced dyskinesias in psychiatric patients: Early investigative findings. Clin EEG
10:151–163
Struve FA, Ramsey PP, Kane JM, Willner AE (1982) Neuropsychological and electroenceph-
alographic correlates of neuroleptic induced involuntary movements: implications for tardive
dyskinesia. In: Malatesha RN, Hartlage LC (eds) Neuropsychology and Cognition, vol 2.
Martinus Nijhoff Publishers, The Hague, pp 674–709
Syvalahti E, Lehtinen I, Salokangas R (1981) Neuroleptic treatment and tardive dyskinesia-
clinical, neurophysiological and neuroendocrinological studies in schizophrenic patients.
Third World Congress of Biological Psychiatry, Stockholm
Tasher DC, Gibbs EL, Unrath DM (1970) Electroencephalograms of patients in a maximum
security hospital. Clin EEG 1:101–110
Wegner JT, Struve FA, Kane JM (1977) The B-Mitten pattern and tardive dyskinesia: a possible
association. Am J Psychiat 134:1143–1145
Wegner JT, Struve FA, Kantor JS, Kane JM (1979) Relationship between the B-Mitten EEG
pattern and tardive dyskinesia: A pilot control study. Arch Gen Psychiat 36:599–603
Winfield DL, Sparer PJ (1954) The electroencephalogram in paralysis agitans. Dis Nerv Syst
15:114–120
Chapter 17
Small Sharp Spikes

Wave Form Description

The small sharp spike (SSS) wave form was originally described by Gibbs and
Gibbs (1952, 1964) who considered the finding to be associated with seizure
disorders. This pattern is almost exclusively dependent on drowsy and sleep
activation for its detection, although reports of its rare occurrence in the awake
tracing have been made (Gibbs 1971; Struve and Pike 1974; Saito et al. 1987). The
pattern is also referred to as benign epileptiform transients of sleep (BETS). In
terms of its morphology and scalp distribution, the wave form consists of isolated
low amplitude (5–50 lV) negative or biphasic sharp spikes with a widespread
distribution involving frontal, anterior temporal, and mid-temporal electrode sites.
Although these discharges can be bilateral, they most often occur independently
over the left and right anterior cortex and they never appear as fixed interval
repetitive discharges. In some of the discharges a miniature low amplitude slow
wave follows the spike while in others this feature is not present. The morphology
of the SSS finding is best appreciated in monopolar-linked ear reference recordings
(Fig. 17.1).
Some departures from the classic wave form morphology and distribution
originally described by Gibbs and Gibbs (1952, 1964) have appeared. Although a
frontaltemporal distribution for this wave form appears to enjoy a near universal
acceptance in the field, some investigators (Reiher and Klass 1968) have stressed a
maximal amplitude over posterior temporal regions with a greater spread of the
wave form to posterior cortex than to anterior electrodes. Not unexpectedly, some
SSS discharges exceed the low voltage range and have amplitudes that are in excess
of 50 lV and this has prompted some electroencephalographers (Reiher and Klass
1968; Klass 1975; White et al. 1977) to quibble about using the adjective ‘‘small’’ in
labeling this wave form. In a thoughtful retrospective analyses of 250 EEGs con-
taining SSSs, Hughes and Olson (1981) were able to define two variations of this
wave form. The most common form included discharges that spread beyond the
temporal areas into other cortical regions while the less common form seen in 20 %
of the cases involved discharges that were confined to the temporal cortex.

N. N. Boutros, Standard EEG: A Research Roadmap for Neuropsychiatry, 171

DOI: 10.1007/978-3-319-04444-6_17,  Springer International Publishing Switzerland 2013
172 17 Small Sharp Spikes

Fig. 17.1 An example of the small sharp spikes (SSS) which appears here in the right
frontotempolar region during stage-2 sleep (normal looking sleep spindles appear in the figure).
This 24-years-old woman suffered from recurrent major depression and had no history of
seizures, head trauma, or encephalopathy. She did complain of repeated headaches and infrequent
paresthesias

Activation Techniques

As indicated above, the use of drowsy and sleep activation is essential if SSSs are
to be detected with any regularity (Gibbs and Gibbs 1964; Gibbs 1971; Reiher and
Klass 1968; Small 1970; Struve and Pike 1974; Koshino and Niedermeyer 1975).
Furthermore, sleep activation obtained following sleep deprivation may have some
additional incremental value in eliciting this wave form (Jabbari et al. 2000). In a
study that compared sleep EEGs obtained from 63 adult psychiatric patients fol-
lowing 24 h of sleep deprivation with standard sleep activated EEGs from 60
nonsleep-deprived patients (Milstein et al. 1979), sleep deprivation was associated
with a significant increase in the incidence of EEG variants including SSSs, 6/s
spike and wave, and 14 and 6/s positive spikes.
Other activation techniques described in the EEG literature have not been
useful in increasing the detection of this wave form. Photic stimulation does not
elevate the incidence of this finding (Wacaser 1969; Small 1971) although the
Wave Form Description 173

presence of SSSs in the sleep EEG tracing may be higher among patients with a
positive photoconvulsive response than among those with EEGs refractory to
photic activation (Small 1971).

Age Distribution

There is a wide spread consensus that the SSSs are almost exclusively confined to
adult populations. In their original report, Gibbs and Gibbs (1964) stated that 80 %
of the 1,025 patients with this wave form were aged 20 and above and nearly all
were above age 15. Furthermore, they also reported that among 3,476 normal
controls this wave form was only encountered in subjects aged 20 and above and it
was totally absent among the 2,857 controls below age 20 (Gibbs and Gibbs 1964).
In a psychiatric population involving patients starting below age 10 and continuing
throughout adulthood, Small (1970) recorded SSSs in patients aged 21 and above
but reported an absence of this finding among 480 patients below age 21. Later
Koshino and Niedermeyer (1975) reviewed SSS records drawn from a population
of unselected patients referred for EEG study and found that 78.3 % of the cases
were aged 21 or older. Of a significant note is a study from Japan (Saito et al.
1987) in which the age distribution for SSSs was given separately for epileptic
patients and patients without seizures. Of 109 nonepileptic patients with this SSSs,
only 2.8 % were younger than age 16, again confirming the dominant adult age
range for this EEG finding. In contrast, however, when 125 seizure patients with
SSSs were examined, 17.6 % of the cases occurred in children below age 16. This
finding, if replicated, could call into question the currently proposed lack of
association between the presence of SSSs and epilepsy particularly in children.

Incidence Among ‘‘Normal’’ Control Subjects

The incidence of the SSSs among normal control subjects is summarized in

Table 17.1. While a slight increase in incidence with advancing age is apparent,
three of the four studies (Gibbs and Gibbs 1964; Hughes and Cayaffa 1977; Small
et al. 1978) confirm the generally very low incidence of this wave form among
nonclinical control populations. In sharp contrast, White et al. (1977) report an
exceptionally high incidence of this wave form including a 24 % overall incidence
as well as a 4.5-fold increase for ages 50–59 and a 6-fold increase for ages 30–39
as compared with the other studies. The elevated incidence findings of these
authors have been harshly criticized by Gilbert (1979) who argued that the criteria
they used to identify this wave form deviated significantly from the earlier criteria
established in the literature. This is a serious issue as significant training, to
competency criteria, in detecting these relatively small and difficult, to detect
phenomena seems lacking. Emphasis on the need for accurate detection and
174 17 Small Sharp Spikes

Table 17.1 Incidence of small sharp spikes among normal controls

Study Age range
\10 10–19 20–29 30–39 40–49 50–59 60+
Gibbs and Gibbs (1964) N 2,188 669 275 116 89 76 139
Percentage (%) 0.0 0.0 1.1 6.0 7.9 6.5 5.8
a
Hughes Cayaffa (1977) N 60 60 60 60 60 60 60
Percentage (%) 3 3 6 6 17.5 10 26
White et al. (1977) N --- 21 26 25 17 26 5
Percentage (%) - - - 9.5 7.7 36 23.5 38.5 40
6–12
Small et al. (1978) N 16
Percentage (%) 0.0
a
Actual percentage figures are not available in the cited text because the percentage data was
only presented in graphed form. The present authors obtained the percentage figures shown in this
table by estimating (as carefully as possible) percentage data from the published graphs

reporting of these discharges particularly in psychiatric populations cannot be over

emphasized.
In their paper, White et al. (1977) partially attribute their high incidence of
SSSs among normal control subjects to the routine placement of nasopharyngeal
electrodes which presumably allowed them to detect these wave forms when they
failed to appear with standard electrode placements. However, Hughes (1983)
expressed skepticism that abnormal discharges of any type will appear very often
confined to nasopharyngeal leads while being absent in all scalp derivations. Such
skepticism is supported by earlier work (de Jesus and Masland 1970) demon-
strating that out of 789 EEGs (495 of which were abnormal) only 7 tracings were
found with an abnormal discharge confined to nasopharyngeal leads (0.9 % of all
patients screened). Others have also reported that nasopharyngeal electrodes
provide little or no additional information beyond standard scalp electrodes when
used with psychiatric patients (Struve and Feigenbaum 1981). Moreover, naso-
pharyngeal leads are usually uncomfortable and would be difficult to place par-
ticularly in not fully cooperative psychiatric patients. Furthermore, the chances for
relaxing, and falling asleep while these invasive electrodes are in place are low and
thus likely to decrease instead of increase in the yield from the studies in psy-
chiatric populations. At the current time, one cannot see justification for use of
nasopharyngeal leads in psychiatric EEGs.
Most recently, Santoshkumar et al. (2009) reported an incidence of 1.85 %
of 35,249 referrals for a routine EEGs of what they called ‘‘benign sporadic
sleep spikes.’’ They justifiably concluded that these discharges do not predict
the subsequent occurrence of seizures but without much justification concluded
that they were incidental findings without any clinical relevance (psychiatric or
otherwise).
Wave Form Description 175

Brain Sources

The exact brain locus (or loci) responsible for the generation of the SSS pattern is/
are not known. Using simultaneous scalp and depth electrode placements with two
patients being evaluated for surgical treatment of epilepsy, SSS pattern was
recorded from widespread deep intra cranial locations involving temporal, frontal,
and parietal lobes (Westmoreland et al. 1979). However, the amount of brain
accessed with depth electrodes was said to be quite limited and inadequate for
specification of detailed anatomic generators for this wave form. In another depth
electrode study involving one patient (Fernandes de Lima et al. 1980), spike
discharges that were detected in the amygdaloid nuclei were shown to correspond
to SSS pattern simultaneously registered with scalp electrodes. This report
underscores the reported correlation between SSSs and mood disorders (see
below). Lebel and associates (1977) used simultaneous 32 channel scalp lead
recordings in an effort to map the topography of the wave form and reported an
unusually widespread distribution producing a dipole axis that was uninformative
regarding possible wave form generators. Although currently available source
localization technology using noninvasive 64, 128, or 256 channel EEG recordings
could make a significant contribution to the study of this EEG finding, such studies
have not been undertaken. In the same vain, magnetoencephalography (MEG) and
the rapidly advancing multimodality registration techniques linking EEG/MEG
data with functional imaging like functional MRI (fMRI) and positron emission
tomography (PET) scanning and capable of both accurate temporal and spatial
resolutions can also be useful in this regard.

Clinical Considerations

The great majority of accepted abnormal EEG discharges can result from multiple
causes and because of this they are often nonspecific for exact etiology. This is
especially true for the group of controversial EEG wave forms which include the
SSS pattern. Historically, the fact that the presence or absence of this wave form
that fails to contribute to the establishment or exclusion of a psychiatric diagnosis
has constituted one reason for the view that it is without clinical or psychiatric
relevance (Reiher and Klass 1968, 1970; Reiher et al. 1969; White et al. 1977).
However, this insistence on precise diagnostic relevance may be considerably
shortsighted because it discourages consideration (and continued clinical investi-
gation) of potential stronger relationships between this EEG finding and discrete
symptoms or symptom clusters which may cut across standard Axis 1 or Axis II
diagnoses. As is described below, there are suggestions that some of these
symptoms may involve affective disturbance, autonomic neurovegetative symp-
toms, and even a moderate association with seizures.
176 17 Small Sharp Spikes

Small Sharp Spikes in Unselected Psychiatric Patients

The incidence of SSSs in the general population of essentially unselected psy-

chiatric patients is exceedingly low with figures that are comparable to the low
incidence found among normal controls. In essence, the presence of this EEG
wave form does not permit one to make a distinction between psychiatric and
nonpsychiatric subjects. This is not surprising in view of the lack of a relationship
between this wave form and Axis 1 psychiatric diagnoses. The incidence of SSSs
among psychiatric patients is given in Table 17.2.
Based on a grand weighted average across studies in this table, SSSs occur with
an incidence of 2.96 % among 7,604 patients tested. Furthermore, the incidence is
relatively consistent across studies and ranges from 1.2 % to around 8 or 9 %. The
one exception showing a 14.3 % incidence (Small et al. 1978) is derived from 21
hyperkinetic children for which some type of positive EEG finding was one of the
inclusion criteria employed in the study.
Most of the studies reported in Table 17.2 were based on EEGs (either referred
or routine) from inpatient psychiatric facilities. Although in two studies (Tasher
et al. 1970; Olson et al. 1970) prison populations were used, the psychiatric
diagnostic composition was quite comparable to the other studies. There was no
indication that this wave form related in any way to the reason for incarceration. In
a large study of prisoners omitted from Table 17.2 because it did not provide
adequate psychiatric diagnostic information (Conte and Demeter 1976), SSSs
occurred in 6.8 % of 248 prisoners with wake and sleep EEGs and the finding was
not related to any of the measures of criminal activity which were obtained.

The Affective Disturbance Connection

Table 17.3 provides a brief synopsis of some of the salient findings from studies that
have at least some bearing on the possible relationship between SSSs and affective
disorders. Some of the studies offer reasonable support for such an association while
with others the support is limited or absent. Furthermore, the studies vary consid-
erably in the degree of experimental rigor (from controlled comparisons to inci-
dental data derived from general surveys) with which this issue was addressed.
The first chapter that suggested a possible association between SSSs and
affective disorder (primarily depression) among a population of psychiatric
patients appeared in 1970 (Small 1970). This study was important because it
carefully contrasted 50 patients with SSSs as the only EEG finding (note: some
EEGs may contain two or more EEG abnormalities thus confounding attempts to
establish clinical correlations) with 50 age-matched patients with pure normal
EEGs and all EEGs contained drowsy and sleep stages necessary for the detection
of this wave form. Fifty-four percent of the patients with EEGs containing SSSs
were diagnosed with an affective disorder (manic-depressive reaction or psychotic
Wave Form Description 177

Table 17.2 Incidence of small sharp spikes (SSS) in patients with psychiatric diagnoses referred
for EEG study
Study Diagnoses Age N Number Percentage (%)
SSS SSS
Small (1970) Mixeda Adults 1,300 50 3.8
Struve and Honigfeld Mixeda 14–60 790 16 2.0
(1970)
Tasher et al. (1970) Mixedb Adults 122 5 4.1
Olson et al. (1970) Mixedb Adults 104 3 2.9
Struve and Pike (1974) Mixeda 14–70 4,000 46 1.0
Koshino and Niedermeyer Mixeda Adolescent- 208 3 1.4
(1975) Adult
Gibbs and Novick (1977) Mixeda Adult 1,000 94 9.4
Small et al. (1978) Hyperkinetic 6–12 21 3 14.3
childrenc
Rau et al. (1979) Compulsive Adult 59 5 8.5
eatingc
Grand weighted Average: 7,604 225 2.96
a
Patients drawn from inpatient psychiatric service
b
Psychiatrically diagnosed patients drawn from a prison population
c
Outpatients participating in funded studies

depressive reaction) versus 28 % of patients with normal EEGs (p \ 0.01). A

second important paper from this team (Small et al. 1975) reported that 43 % of 60
bipolar patients displayed this EEG wave form with the incidence of SSSs being
much higher among both female (p \ 0.001) and male (p \ 0.01) bipolar patients
than among age- and gender-matched patients from the general psychiatric pop-
ulation. This paper also presented family history data suggesting a possible genetic
contribution to the association. While these two early papers suggest a positive
association between this wave form and affective disturbance, another paper from
their lab (Small et al. 1997) found only a 17 % incidence of SSSs in a sample of
163 patients tested during a manic state. The authors suggest that the dramatic
increase in the use of anticonvulsant medications for bipolar disease which
occurred since their earlier 1970 and 1975 papers may have contributed to the
marked drop in the incidence of this EEG finding in their 1997 series of patients.
The other studies included in Table 17.3 were not focused specifically on the
hypothesis that this specific EEG wave form was related to primary diagnoses of
affective disorder or to some of the discrete symptoms often seen with affective
disorders. In a very large survey of wake and sleep EEG findings in 1,000 hos-
pitalized psychiatric patients (Gibbs and Novick 1977), 67 % of the 94 patients
with SSSs had some kind of affective disorder diagnosis (depressive neurosis,
psychotic depression, manic-depressive psychosis) and this wave form was the
most common EEG finding (22.2 %) among suicidal patients. While this study
seems to lend support to the findings from Small and collaborators discussed
above, others offer little or no support. In one survey (Saito et al. 1987), SSSs
178 17 Small Sharp Spikes

Table 17.3 Summary of salient findings relating small sharp spikes (SSS) to affective
dysfunction
Study Results relevant to affective disorders
Small (1970) (1) 50 psychiatric patients with SSS as only EEG finding: 54 % had Dx
of Affect Disorder. 50 matched patients with normal EEG: 28 %
had Dx of Affect Disorder. (V2 = 5.953, df = 1, p = 0.015)
(2) SSS associated with increased incidence of psychomotor
retardation, mood swings, feelings of hopelessness, concentration
difficulty, sleep disturbance
(3) SSS patients had higher scores on MMPI depression scale and
Zung Rating Scale for Depression although the difference was not
statistically significant
Milstein and Small Psychiatric patients with SSS (n = 42) did not differ from normal EEG
(1971) patients (n = 64) on the Zung Depression Scale or the Raskin
Mood Scale
Struve et al. (1973) Study of suicide in patients with and without paroxysmal EEGs (total
n = 225). A reanalysis for this table contrasts presence of suicide
attempts among patients with SSS (n = 10) versus other
paroxysmal EEG findings (n = 75, including 14 and 6 positive
spikes, diffuse parox. slow, 6/s spike and wave, focal spikes,
RMTD). Suicide attempts were made by 100 % of SSS patients
versus 69 % for all other paroxysmal EEG patients (Fishers Exact
Test, p = 0.056)
Small et al. (1975) (1) Reported 43 % incidence of SSS among 60 bipolar manic-
depressive patients
(2) Incidence in female bipolar patients higher than in age-matched
general female population (55 % vs. 12.7 %, p \ 0.001). Incidence
in bipolar males higher than in age-matched general male
population (32 % vs. 11.2 %, p \ 0.01)
(3) Patients with SSS had more consecutive generations of families
with affect disorder than did patients without the SSS EEG finding
Koshino and Large survey of 3,377 EEGs from a neurology service. 36 patients Dx
Niedermeyer (1975) manic-depressive. 0 % with SSS
Gibbs and Novick (1) A survey of EEGs in 1,000 hospitalized psychiatric patients. Of the
(1977) 94 patients with SSS, 67 % had an affect disorder
(2) SSS was the most common EEG finding among suicidal patients
(22.2 % incidence)
Struve et al. (1977) Large study of relationship of EEG findings to suicide behavior among
psychiatric patients involving 168 nonsuicidal controls and 491
who displayed suicide ideation and/or attempts. Of 15 patients with
the SSS EEG pattern, all (100 %) displayed significant suicide
ideation and 7 (46.7 %) had made one or more suicide attempts
Saito et al. (1987) Large survey of 5,875 patients referred for EEG. Of 188 patients with
diagnosis of affect disorder, 3.2 % had an EEG with SSS
Small et al. (1997) SSS occurred in 17 % of 163 patients tested during a manic state

occurred in only 3.2 % of a subsample of 188 patients with a diagnosis of affective

disorder referred for EEG study (out of a population of 5,875 EEGs) and Koshino
and Niedermeyer (1975) could find no evidence of SSSs in a subsample of 36
patients with a manic-depressive diagnosis culled from 3,377 EEGs secured in a
Wave Form Description 179

neurology service. It is quite difficult to assign a weight or value to the negative

incidence findings in these later two surveys because the meager data relevant to
the issue of SSSs and affective disorder appeared to be tangential to the main thrust
of these surveys. One of the papers (Koshino and Niedermeyer 1975) involved a
survey of EEGs from a neurology service while the other (Saito et al. 1987) was
sharply focused on the relationship between this EEG finding and seizure disor-
ders. In addition, the small subsample of patients with manic-depressive or
affective disorder diagnoses in these two surveys were patients presumably
referred to a neurology service for neurological reasons and hence they may not be
typical of affective disorder patients without suspected neurological involvement.
Niedermeyer and Lopes da Silva (2005) concluded that considering SSS as per-
fectly normal variants is not strongly supported and that it may carry some level of
epileptogenicity and thus may be indicative of some form of cortical
hyperexcitability.
The studies in Table 17.3 by Struve et al. (1973, 1977) were concerned with an
exploration of a variety of paroxysmal EEG wave forms in relationship to several
aspects of suicide behavior and data involving the SSS pattern offer only incidental
support to the findings of Small and her team. In one (Struve et al. 1973), 100 % of
10 patients with SSSs made serious suicide attempts versus a 69 % incidence of
attempts in patients with paroxysmal EEGs other than the SSS finding with the
difference almost significant (Fisher’s Exact Test, p = 0.056). Later, (Struve et al.
1977) it was found that 100 % of 15 patients with SSSs displayed suicide ideation
and 7 of them (46.7 %) had made one or more suicide attempts. Although these
EEG suicide studies do not come close to providing an appropriate test of the
relationship between SSSs and affective disturbance, the high loading of suicide
behavior within the sample of patients with SSSs does appear to be at least
consistent with an affective disturbance behavioral correlate for this EEG finding.
Psychological testing has not been useful in clarifying the issue of depressive
symptomatology in patients with this EEG wave form. In her initial study (Small
1970) patients with SSSs, as contrasted with normal EEG patients, had higher
(although not statistically significant) scores on the MMPI depression scale and the
Zung Rating Scale for Depression. Later, however, patients with SSSs were found
not to differ from normal EEG patients on psychological test measures relevant to
depression or mood (Milstein and Small 1971).
In conclusion, it appears that the strong initial suggestions of an affective
symptom correlate for this wave form provided by Small and her team have simply
not been actively followed up by others in the field. Partly, this neglect may stem
from the unfortunate influence of early strong statements from neurology that this
wave form is little more than a ‘‘Benign Epileptiform Transient of Sleep or BETS’’
without demonstrable relevance (White et al. 1977) or a pattern of ‘‘Doubtful
Clinical Significance’’ (Reiher and Klass 1968; Reiher et al. 1969). However,
neither of these papers contained a rigorous experimental study of affective
symptom correlates to this wave form and hence the conclusions were offered
without empirical substantiation of any kind. Reluctance to study this EEG finding
in psychiatry may also have been a partial spin-off of the major controversy in
180 17 Small Sharp Spikes

neurology over the relationship of SSSs to seizure disorders. Moreover, it should

not be very surprising that differences among studies existed in view of the sig-
nificant heterogeneity of mood disorders.
There is little doubt that if experimental studies of this EEG pattern were
repeated using modern research techniques the relationship of this wave form to
psychiatric behavior could be greatly clarified and possibly even fully substanti-
ated. One approach of particular promise might be to separately examine the
association between SSSs and a variety of discrete symptoms and behaviors
common in affective disorders in order to determine which symptoms or symptom
clusters might be most strongly expressed in patients with this EEG wave form. For
example, while roughly half of psychiatric patients with affective disorders type
diagnoses may have SSSs in their EEG (Small et al. 1975), other studies (Struve
et al. 1973, 1977) suggest that nearly all psychiatric patients with this EEG finding
may display suicidal behavior. Thus, the possibility that the presence of this EEG
wave form in psychiatric patients with affective disorders denotes a subgroup of
patients with elevated suicide risk represents one hypothesis worth some research
time. In addition the possibility, as suggested more recently by Small and her
associates (1997), that anticonvulsant medication may significantly attenuate or
reduce the expression of SSS discharges in the EEG of affective disorders patients
represents another problem easily amenable to empirical investigation.

Autonomic and Neurovegetative Symptoms

A previous review of the Small Sharp Spike EEG finding (Hughes 1983) pointed
to a variety of neurovegetative symptoms, in addition to seizures, as one of the
symptom clusters associated with this wave form. This type of clinical correlate
was first suggested by Gibbs and Gibbs (1964) who reported a 49 % incidence of
headaches and a 32 % incidence of dizzy spells among approximately 287 patients
with SSSs referred for EEG study who were found not to have seizures. However,
it must be noted that based on inspection of their graphed data this incidence may
not have been significantly greater than that seen in normal EEG patients also
obtained from their laboratory (i.e., no statistical tests were performed). Other
neurovegetative symptoms occurred with less than a 20 % incidence and among
epileptic patients with this EEG finding symptoms of headaches and dizziness
occurred with less than a 10 % incidence. Although a number of incidence studies
followed the Gibbs’ report, detailed well-controlled experimental investigations of
this electrophysiological symptom association have never been carried out.
In a large study, which is one of the better incidence reports (Hughes and Olson
1981) two topographical subtypes of the Small Sharp Spike EEG wave form were
identified and various neurovegetative symptoms (i.e., headache, dizziness, ver-
tigo, blackouts) occurred in approximately 48 % of 200 patients with diffuse SSSs
and approximately 58 % of 50 patients in which the SSSs were confined to
temporal electrodes (note: the incidence figures given here had to be estimated
Wave Form Description 181

‘‘by eye’’ from the published graphed data through use of a protractor and hence
they constitute ‘‘estimates’’). Importantly, in each of these two topographical
subtypes of SSSs the incidence of neurovegetative symptoms was reported to be
significantly higher (p \ 0.005) than the corresponding incidence found in each of
three separate control groups that did not contain patients with SSSs. In a sub-
sequent study (Hughes and Gruener 1984), neurovegetative symptoms were
reported in 62 % of patients with this EEG wave form.
Other incidence studies reported in the literature are flawed because of serious
methodological deficiencies. Neurovegetative symptoms (syncope, dizzy spells,
transient weakness) were found in 18.5 % of nonseizure patients with pure SSSs
(and no other finding) in the EEG (Koshino and Nidermeyer 1975). However, the
symptoms appear to have been tabulated from chart reviews only and this could
easily have caused a significant under representation of their true incidence. It is
interesting that these authors also found the incidence of such symptoms to drop to
only 5 % when patients with this wave form coexisting with other EEG abnor-
malities were studied. In their EEG survey of 1,000 hospitalized psychiatric
patients (Gibbs and Novick 1977) the authors present a large multipage table
consisting of a cross tabulation of various EEG findings versus a number of
diagnostic classifications and several specific symptoms. Reading through this
large data set one finds that 20.2 % of patients with SSSs complained of headaches
and the incidence of dizzy spells and nausea were both 6.3 %. However, again it is
almost certain that chart review data were used and that patients were not indi-
vidually interviewed by study examiners regarding the presence, absence, severity,
and history of various neurovegetative symptoms. Not infrequently vegetative
symptoms may be viewed by the treating clinician as incidental to the primary
presenting diagnosis and thus they may not be adequately detailed in the EEG
consult, chart, case history, or discharge summary. Saito et al. (1987) reported an
11.9 % incidence of a ‘‘diagnosis’’ of headaches among 109 patients with this
wave form who did not have seizures. Under the heading of diagnosis they also
listed an ‘‘Other’’ category which occurs with a 34.8 % incidence among patients
with this EEG finding. However, there is no specification provided for what
symptoms are subsumed under the category of ‘‘Other.’’ Furthermore, the critique
of using a methodology which is incapable of estimating the true incidence of
neurovegetative symptoms in patients with this EEG wave form applies to this
study as it does to the others contained in this paragraph.
Rating scales have only been rarely used in assessing potential neurovegetative
symptoms in patients with this wave form. As a subanalyses of one study (Small
1970), it was reported that seven out of eight subjects with SSSs more strongly
endorsed Cornell Medical Index questions regarding dizzy spells, fainting, and
headaches than did matched controls. However, no statistics were presented. In a
later paper (Milstein and Small 1971), items on the Cornell Medical Index failed to
differentiate patients with SSSs from normal EEG patients or patients with other
types of paroxysmal EEG discharges.
Neurovegetative symptoms are not uncommon as associated symptoms coexis-
ting with other major psychiatric diagnostic categories. When infrequent in
182 17 Small Sharp Spikes

occurrence or mild in severity they may be viewed by patients as inconsequential to

the larger issues of psychiatric disturbance requiring treatment. However, when they
occur frequently and with moderate or greater severity they may become treatment
relevant. As with the probable association between SSSs and affective disorders
discussed above, the possible association between this EEG finding and elevated
autonomic neurovegetative symptomatology is entirely amenable to conclusive
experimental resolution given the milieu of research methods available today.

Seizure Correlates Versus Nonseizure Correlates

While seizures constitute a primary, yet still controversial, clinical correlate of the
SSS wave form this issue is outside the scope of this book. However, a careful
reading of the literature suggests that the association is, in fact, genuine and of at
least moderate strength. This is a conclusion stated by Niedermeyer (2001).
Nonetheless, when one attempts to consider the SSS literature from a purely
psychiatric standpoint affective disorders and, to a lessor extent, neurovegetative
symptoms loom as possible clinical correlates whereas moderate incidences of
seizures are reported in papers from laboratories that focus on medical and neu-
rological referrals. Furthermore, the two classes of correlates do not appear to mix
well. When hospitalized psychiatric patients diagnosed with bipolar or psychotic
depressive disorders have SSSs in their EEGs they tend not to have seizures or
questions of seizures. Conversely, those epileptic patients with this EEG finding
are seldom classified as having a coexisting psychiatric affective disorders,
although it remains possible that in some of the later cases the seizure disorders
takes precedence over what may be coexisting psychiatric symptoms in the minds
of the treating clinician. Even so, it appears that seizures represent one class of
sequelae of this wave form while nonseizure symptoms represent a second and
separate class of sequelae. In this respect, Hughes and Gruener (1984) noted that
when neurovegetative symptoms were present in SSS patients they occurred far
more often (75 %) in patients without seizures than in (26 %) patients with sei-
zures (p \ 0.0001). Although the underlying reasons for this two-category class of
clinical sequelae remains obscure, it should not allow one to conclude that the
wave form is somehow without relevance. Each class of correlates is capable of
being substantiated independently of the other.

Supported Findings

(1) The SSS wave form has had a history that has suffered from the undue early
influence of a minority of widely quoted papers (Reiher and Klass 1968, 1970;
Reiher et al. 1969; White et al. 1977) attempting to document its lack of
Supported Findings 183

clinical relevance even though these critiques appeared to be based primarily

on discussion as opposed to any presentation of empirical or experimental
refutation of published positive findings. This has had the unfortunate effect of
contributing to a state of affairs leading to a reduced interest in and, as a result,
reduced investigation of this wave form, its clinical sequelae, and its potential
treatment relevance. More recent opinion or review articles continue to appear
stressing the lack of clinical relevance of the SSS (Santoshkumar et al. 2009).
(2) SSSs do not appear to contribute to Axis I psychiatric diagnosis and in this
respect they are not very different from the majority of EEG findings. How-
ever, there is ample reason to suspect that their detection in the clinical EEG
may, pending further confirming research, contribute to a more subtle
understanding of depressive affective disorders and even risk for suicide.
(3) As mentioned above, the modest relationship between this wave form and
complex partial seizures may suggest that some exploration and consideration
of seizure equivalents might be of value in the selected patient and if EEG
clinical correlations appear convincing or strong an empirical anticonvulsant
trial might receive consideration. Similar considerations might apply to the
patient where neurovegetative symptoms are severe enough to warrant clinical
treatment.

Open Research Questions

(1) What is the optimal EEG work-up to detect the SSSs?

(2) What are the exact incidences in the different psychiatric entities particularly
those with affective components like mood disorders (bipolar, major depres-
sion with and without psychotic features, dysthymia, and cyclothymia) as well
as other disorders like borderline personality disorder and schizoaffective
disorders?
(3) What are the effects of the different classes of anticonvulsant medications and
are the effects the same irrespective of the clinical presentation or discharge
forms and locations?
(4) What are the cerebral sources of the SSSs and could they be amenable to the
effects of focal stimulation or inhibition via transcranial magnetic stimulation?
(5) Finally, it must be said that the SSS wave form is easily detected in the EEG as
long as referential recordings are made during drowsiness and sleep. Fur-
thermore, the previously identified potential clinical correlates of depressive
affective disorders, neurovegetative symptoms, and seizures remain com-
pletely amenable to continued investigation and there is little doubt that much
of the early work could profitably be followed up using modern research
methods and designs.
184 17 Small Sharp Spikes

References

Conte WR, Demeter CR (1976) Electroencephalograms correlated with commitment data on 270
offenders. Clin Electroencephalogr 7:35–48
De Jesus PV, Masland WS (1970) The role of nasopharyngeal electrodes in clinical
electroencephalography. Neurology 20:869–878
Fernandes de Lima V, Chatrian GE, Ojemann GA, Lettich E (1980) ‘‘Small sharp transients’’
‘‘benign elipeptiform transients’’). EEG recordings from the scalp and the amygdaloid nuclei
of a human subject. In: Wada JA, Penry JK (eds) Advances in epileptology, 10th epilepsy
international symposium, New York, Raven Press, p 538
Gibbs FA (1971) Objective evidence of brain disorder in cases of whiplash injury. Clin
Electroencephalogr 2:107–110
Gibbs FA, Gibbs EL (1952) Atlas of electroencephalography, vol 2. Wesley, Cambridge
Gibbs FA, Gibbs EL (1964) Atlas of electroencephalography, vol 3. Wesley, Reading
Gibbs FA, Novick RG (1977) Electroencephalographic findings among adult patients in a private
psychiatric hospital. Clin Electroencephalogr 8:79–88
Gilbert GJ (1979) (Lttr. to Editor) Clin Electroencephalogr 10:54–56
Hughes JR (1983) A review of small sharp spikes. In: Hughes JR, Wilson WP (eds) EEG and
evoked potentials in psychiatry and behavioral neurology. Butterworths, Boston, pp 347–359
Hughes JR, Cayaffa JJ (1977) The EEG in patients at different ages without organic cerebral
disease. Electroencephalogr Clin Neurophysiol 42:776–784
Hughes JR, Gruener G (1984) Small sharp spikes revisited: Further data on this controversial
pattern. Clin Electroencephalogr 15:208–213
Hughes JR, Olson SF (1981) An investigation of eight different types of temporal lobe discharges.
Epilepsia 22:421–435
Jabbari B, Russo MB, Russo ML (2000) Electroencephalogram of asymptomatic adult subjects.
Clin Neurophysiol 111:102–105
Klass DW (1975) Electroencephalographic manifestations of complex partial seizures. In: Penry
JK, Daly DD (eds) Advances in neurology. Raven Press, New York, pp 113–140
Koshino Y, Niedermeyer E (1975) The clinical significance of small sharp spikes in the
electroencephalogram. Clin Electroencephalogr 6:131–140
Lebel M, Reiher J, Klass D (1977) Small Sharp Spikes (SSS): Electroencephalographic
characteristics and clinical significance. Electroencephalogr Clin Neurophysiol 43:463 (abst)
Milstein V, Small JG (1971) Psychological correlates of 14 & 6 positive spikes, 6/s spike-wave,
and small sharp spike transients. Clin Electroencephalogr 2:206–212
Milstein V, Small JG, Sharpley P, Golay S (1979) All night sleep deprivation in psychiatric
patients: Relation to mood. EEG and psychiatric diagnosis. Clin Electroencephalogr 10:25–30
Niedermeyer E, EEG pattern review (2001) Marginal paroxysmal patterns and their significance.
Clin Electroencephalogr 32:44 (abst)
Niedermeyer E, Lopes da Silva F (eds) (2005) Electroencephalography: basic principles, clinical
applications and related fields, Baltimore-Munich, Urban & Schwarzenberg
Olson WH, Gibbs FA, Adams CL (1970) Electroencephalographic study of criminals. Clin
Electroencephalogr 1:92–100
Reiher J, Ham O, Klass DW (1969) EEG characteristics and clinical significance of small sharp
spikes- a reappraisal. Electroencephalogr Clin Neurophysiol 26:635
Reiher J, Klass DW (1968) Two common EEG patterns of doubtful clinical significance. Med
Clin North Amer 52:933–940
Reiher J, Klass DW (1970) ‘‘Small sharp spikes’’ (SSS): electroencephalographic characteristics
and clinical significance. Electroencephalogr Clin Neurophysiol 28:90–105
Saito F, Fukushima Y, Kubota S (1987) Small sharp spikes: possible relationship to epilepsy. Clin
Electroencephalogr 18:114–119
References 185

Santoshkumar B, Chong JJ, Blume WT, McLachlan RS, Young GB, Diosy DC, Burneo JG,
Mirsattari SM (2009) Prevalence of benign epileptiform variants. Clin Neurophysiol
120(5):856–861
Small JG (1970) Small sharp spikes in a psychiatric population. Arch Gen Psychiat 22:277–284
Small JG (1971) Photoconvulsive and photomyoclonic responses in psychiatric patients. Clin
Electroencephalogr 2:78–88
Small JG, Milstein V, Jay S (1978) Clinical EEG studies of short and long term stimulant drug
therapy of hyperkinetic children. Clin Electroencephalogr 9:186–194
Small JG, Milstein V, Medlock CE (1997) Clinical EEG findings in mania. Clin Electroencep-
halogr 28:229–235
Small JG, Small IF, Milstein V, Moore DF (1975) Familial associations with EEG variants in
manic-depressive disease. Arch Gen Psychiat 32:43–48
Struve FA, Feigenbaum ZS (1981) Experience with nasopharyngeal electrode recording with
psychiatric patients: a clinical note. Clin Electroencephalogr 12:84–88
Struve FA, Pike LE (1974) Routine admission electroencephalograms of adolescent and adult
psychiatric patients awake and asleep. Clin Electroencephalogr 5:67–72
Struve FA, Saraf KR, Arko RS, Klein DF, Becka DR (1973) Further, investigation of
electroencephalographic correlates of suicide ideation and attempts: Preliminary results. In:
Fourth annual conference of the Indian psychiatric society-West Zone, Bombay, India, 29–30
Sept 1973
Struve FA, Saraf KR, Arko RS, Klein DF, Becka DR (1977) Relationship between paroxysmal
electroencephalographic dysrhythmia and suicide ideation and attempts in psychiatric
patients. In: Shagass C, Gershon S, Friedhoff AJ (eds) Psychopathology and brain
dysfunction. Raven Press, New York, pp 199–221
Tasher DC, Gibbs EL, Unrath DM (1970) Electroencephalograms of patients in a maximum
security hospital. Clin Electroencephalogr 1:101–110
Wacaser L (1969) Photic activation of the electroencephalogram. Clin Electroencephalogr
1:32–35
Westmoreland BF, Reiher J, Klass DW (1979) Recording small sharp spikes with depth
electroencephalography. Epilepsia 20:599–606
White JC, Langston JW, Pedley TA (1977) Benign epileptiform transients of sleep: clarification
of the small sharp spike controversy. Neurology 11:1061–1068
Chapter 18
Six/Second Spike and Wave Complexes,
the Rhythmic Mid-Temporal Discharges
and the Wicket Spikes

The 6/s Spike and Wave Complexes

Introduction

As with other controversial EEG patterns the 6/s Spike and Wave pattern (6/s SpW)
has two names. The other name here is the Phantom Waves, which also denotes that
the author(s) do not believe it has clinical significance. This pattern has also been
referred to as ‘‘miniature spike and wave’’ as in general the entire wave complex
looks like a miniature reproduction of the 3/s spike and wave discharges which are
the EEG correlates of petit mal epilepsy. The 6/s SpW pattern was described as a
distinct electrographic pattern by Gibbs and Gibbs (1952).

Waveform Description

Two forms were later described by Hecker et al. (1979) and Hughes (1980). Both
publications strongly emphasized the need to differentiate between the two forms
in research studies addressing clinical correlates.
(1) The WHAM form appearing in the Waking record, High in amplitude,
Anterior in location, and seen more in Males (thus the acronym WHAM form),
it is associated mainly with seizures (Fig. 18.1).
(2) The FOLD form seen in Females, Occipital in location, Low in amplitude,
noted in Drowsiness (thus the acronym FOLD form). The FOLD form is
associated mainly with neurovegetative and psychiatric complaints
(Fig. 18.2). The 6/s SpW pattern is difficult to recognize since the discharges
are often very low in amplitude.
Silverman (1967) stressed similarities to the 6/s component of the 14 and 6
positive spikes (Chap. 19). Silverman (1967) also noted a gradual transition from
the 6/s SpW to the 14 and 6 positive spikes with deepening drowsiness. While the

N. N. Boutros, Standard EEG: A Research Roadmap for Neuropsychiatry, 187

DOI: 10.1007/978-3-319-04444-6_18,  Springer International Publishing Switzerland 2013
188 18 Six/Second Spike and Wave Complexes

Fig. 18.1 The 6/s Spike and wave. This form is recorded during wakefulness with usually a high
amplitude anteriorly located and more common in males hence the acronym (WHAM) form.
(From Hughes 1994 with permission)

6/s SpW pattern can be recorded during wakefulness and light non-REM sleep,
drowsiness appears to be the optimal recording state.

Incidence

Marshall (1955) reported the 6/s SpW pattern in 0.9 % of 2,000 unselected EEG
laboratory referrals. Thomas (1957) confirmed an incidence of 0.7 % of all
unselected patients referred for an EEG, but an incidence of 1.1 % in epileptic
patients. Similarly, Gibbs and Gibbs (1964) reported its frequency in what they
deemed healthy control subjects as 0.8 %. The 6/s SpW is reported as the sole
discharge in the EEG in only 0.4 % of usual referrals for a neurology-based EEG
laboratory (Hughes et al. 1965), and 0.8 % (Thomas and Klass 1968).
Most recently Santoshkumar et al. (2009) reported an incidence of 1.02 % of
35,249 referrals for routine EEGs. They justifiably concluded that the 6/s SpW
discharges do not predict the subsequent occurrence of seizures but without much
justification concluded that they were incidental findings.
In contrast, Small (1968) reported this same pattern in 4.5 % of 1,100 con-
secutive EEG recordings in a psychiatric population. Subsequent reports of the
prevalence of this pattern in mixed psychiatric populations varied widely from as
low as 1.5 % (Gibbs and Novick 1977) to Struve (1977) who stated that the 6/s
SpW was the second most common abnormality (20.8 %) among psychiatric
patients.
The 6/s Spike and Wave Complexes 189

Fig. 18.2 The 6/s spike and wave; a second form of this pattern that is seen more in Females,
maximal in the Occipital regions with Low amplitudes and detected during Drowsiness hence the
acronym (FOLD) Form. (From Hughes 1994 with permission)

The 6/s SpW pattern is frequently called phantom spike-waves (as discussed
earlier in the introduction chapter to this section), because the amplitude of the
spike component is often under 25 lV and/or that the spike component is partially
concealed as a notch in the attendant slow wave component. Hughes (1980)
highlighted the uneven sex distribution, which is rather unique as no other EEG
abnormality shows any gender preference.

Clinical Correlates

According to Niedermeyer (2005), 50–60 % of individuals with this pattern have

frank epilepsy with the remainder having syncopal episodes, post-traumatic symp-
toms, or other psychiatric manifestations. Olson et al. (1971) found that patients with
190 18 Six/Second Spike and Wave Complexes

6/s SpW scored high on the MMPI hysteria scale than in patients with other EEG
abnormalities. Boutros et al. (1986) compared a group of patients with the 6/s SpW
discharges to two groups of patients; one with RMTD (see RMTD section below) and
one with psychiatric complaints but normal EEGs. Although the 6/s SpW group had
similar overall psychopathology to the normal EEG group, they had fewer psychi-
atric hospitalizations and suicidal attempts, most probably signifying differences in
the underlying psychopathology. Small (1968) reported similar observations, in
addition to finding higher Institute for Personality and Ability Testing (IPAT) scores
on the anxiety scale, differentiating her 6/s SpW patients from two other groups of
psychiatric patients with and without EEG abnormalities. The 6/s SpW group had
more anxiety-related problems, such as phobias and panic disorders, than did the
other three groups, but this did not reach statistical significance.
Based on reports from the literature, it seems likely that the FOLD form of the
6/s SpW does indicate significant psychopathology, with a propensity toward
anxiety and anxiety-related disorders, although possibly with a better prognosis
than patients with normal EEGs. Psychiatric patients with normal EEGs, on the
other hand, seem to have different aspects of psychopathology that are more
evenly represented and with an overall severity greater than patients with EEG
abnormalities (Small 1968; Olson et al. 1971).
Like the other controversial waveforms they have been correlated with increased
impulsivity and neurovegetative symptoms. Olson et al. (1971) reported that patients
with this pattern scored high on the MMPI hysteria scale. Kocher et al. (1975) were
struck by the occurrence of the 6/s SpW complexes in the abstinence or withdrawal
phase of drug-dependent individuals. It is plausible that the discharges may reflect
some form of increased cortical excitability and may have implications to long-term
abstinence and relapse. Hecker et al. (1979) reported that the occipital (FOLD) form
is often related to drug dependence (Barbiturates) and withdrawal. Research on this
potentially informative correlation has been minimal.

Supported Observations

(1) The prevalence of the 6/s SpW pattern is significantly higher in psychiatric
populations as compared to any other control group.
(2) There are two forms of the 6/s SpW pattern which may have different clinical
correlates.

Open Research Questions

(1) What are the actual incidences of the two forms of the 6/s SpW patterns in
diagnostically well-defined psychiatric populations?
Open Research Questions 191

(2) What is the clinical correlate of each of the two forms?

(3) What are the cerebral sources of the two forms?
(4) What are the treatment implications (particularly possible responsiveness to
antiepileptic medications) of detecting these patterns in psychiatric patients?

The Rhythmic Mid-Temporal Discharges

Introduction

As with other ‘‘controversial waveforms’’ The Rhythmic Mid-Temporal Dis-

charges (RMTD) has two labels. The second label for RMTD is psychomotor
variant which is the original term ‘‘psychomotor variant pattern’’ (Gibbs et al.
1963) which indicated the Gibbses belief that it had some similarity or some
relevance to epilepsy.

Description of Pattern

The morphological features of RMTD were described in some detail by Egli et al.
(1978) with a frequency of 5.5–6.5/s and monophasic and regular with occa-
sionally interposed 12/s activity. The theta activity shows a well-defined negative
sharp component which stresses the paroxysmal nature of the bursts. Localization
is commonly mid-temporal, often spreading to anterior, seldom posterior regions.
RMTD is usually bilateral and can be simultaneous or alternating sides. They are
closely linked to the drowsy state, occurring at the transition from drowsiness to
stage II sleep and arising from a fairly desynchronized EEG. RMTD are com-
monly seen within REM periods, which are markedly fragmented with inter-
spersed periods of drowsy patterns, during which the RMTD are seen.
Occasionally they are strictly related to slow eye movements and periodic respi-
ration. Slow and fast wave sleep in subjects with RMTD are disturbed. Both of
them, especially the fast wave sleep are reduced in favor of markedly increased
stages of drowsiness with RMTD, which sometimes last several minutes. In spite
of such abnormal organization of sleep the subjects feel recovered in the morning
and sleep disturbances are not reported. RMTD could therefore be considered as a
‘‘bioelectrical sleep disorder’’ (Fig. 18.3).
192 18 Six/Second Spike and Wave Complexes

Fig. 18.3 Rhythmic mid-temporal discharges (psychomotor variant) during drowsy recording.
Left mid-temporal focus. The patient is a male 20 years of age with episodic temper dyscontrol,
frequent intense homicidal urges, paresthesias, headaches, abdominal pain, and suicide plans (as a
way of avoiding acting on strong homicidal impulses). Patient responded well to phenytoin
(Dilantin) therapy

Incidence

Egli et al. (1978) reported that out of 50,000 EEG’s those of 38 subjects contained
RMTD, corresponding to an incidence of 0.1 %. Most recently, Santoshkumar
et al. (2009) reported an incidence of 0.12 % of 35,249 referrals for a routine
EEGs. They justifiably concluded that RMTD does not predict the subsequent
occurrence of seizures but without much justification concluded that it is an
incidental finding.
The Rhythmic Mid-Temporal Discharges 193

Fig. 18.3 (continued)

Clinical Correlates

Arfel et al. (1978) reported unilateral or bilateral electrographic temporal dis-

charges at 5 Hz and discusses their morphological characteristics, their topogra-
phy, and their evolution in eleven subjects. The clinical context, in which there
was no evidence of epileptic disorder, was also analyzed. They considered the
discharges to be an infra-clinical phenomena. Egli et al. (1978) did not find any
correlation between RMTD and clinical findings, in particular not with psycho-
motor or any other form of epilepsy. Maione et al. (1981), reported two cases of
patients whose EEGs showed localized rhythmic seizure activity in the mid-
temporal regions of one or both hemispheres, unaccompanied by any clinical
symptoms: the patients’ histories differed: one was of classic migraine and the
other complex partial epilepsy. The frequency and morphology of the paroxysmal
anomalies were identical in the waking state and in sleep. Reports of such a pattern
are rare in the European literature and nonexistent in the Italian literature, facts
which make an ordinary interpretation of the phenomenon difficult.
RMTD have been associated with increased somatization (Boutros et al. 1986).
One third to half of the subjects exhibiting this pattern were found to have psy-
chiatric problems, particularly symptoms of anxiety. One early study reported a
36 % incidence of seizures in patients with RMTD and the possible association of
194 18 Six/Second Spike and Wave Complexes

this pattern with ictal phenomena remains unanswered (Hughes and Cayaffa 1973).
One documented case of psychomotor variant ‘‘status epilepticus’’ with associated
prolonged periods of confusion has also been reported (Anderson and Vanderspec
1974). While several efforts (Gibbs and Gibbs 1964; Garvin 1968; Lipman and
Hughes 1969; Eeg-Olofsson and Petersen 1982) have been made to link this EEG
pattern with behavioral symptoms such as temper dyscontrol, personality disorder
or autonomic phenomena, and others (Neidermeyer and Lopez da Silva 1987) have
not been able to find such associations.
Hughes (2001) reported a case of a 15-year-old male who was sent for an EEG
because of possible staring spells and a learning disability. At times, written
examinations in school were successfully completed and at other times the patient
would write only his name. In two EEGs 1 month apart during hyperventilation,
RMTDs were activated and appeared continuously for 84 min and in the second
instance for 140 min. No maneuvers could modify the pattern and no clinical
changes were reported by the patient. In this admittedly unusual case, Hughes
(2001) asserts the need to test such patients in more rigorous ways in order to be
able to identify any subtle clinical changes that may be occurring.
Relatively more recent work by Lin et al. (2003) attempted to evaluate the
source location and clinical significance of RMTD by magnetoencephalography
(MEG) in nonepileptic and epileptic patients. They conducted simultaneous MEG
and EEG recordings with a whole-scalp 306-channel neuromagnetometer in three
patients: one with right temporal lobe epilepsy (TLE), one with right frontal lobe
epilepsy (FLE), and one with tension headache. They visually detected the RMTD
activity and interictal spikes, and then localized their generators by MEG source
modeling. MEG measurements were repeated 3 months after right anterior tem-
poral lobectomy (ATL) in the TLE patient; 3 months after anticonvulsant medi-
cation in the FLE patient. In epileptic patients, RMTD activities were found during
drowsiness over the left temporal channels of both MEG and EEG recordings, and
their generators were localized to the left posterior inferior temporal region. In the
patient with tension headache, RMTD was localized in the right inferior temporal
area. When the epileptic patients became seizure free with disappearance of epi-
leptic spikes, RMTD was still found over the left temporal channels. Besides,
some bursts of RMTD appeared also in the right temporal channels in the TLE
patient after ATL. These results indicate that the source of RMTD activity is
located in the fissural cortex of the posterior inferior temporal region. Here authors
concluded that RMTD is a physiologic rhythm related to dampened vigilance, and
has no direct relation to epileptogenic activity.
One compromise for the name of this pattern might be Rhythmic Mid-Temporal
Epileptiform Activity (RMTEA), avoiding the term discharge, which implies a
seizure state that has not yet been established. Niedermeyer and Lopes da Silva
(2005) concluded that the RMTD pattern should not be considered a normal
drowsiness pattern and considered it a mild abnormality.
Supported Findings 195

Supported Findings

(1) RMTD is a rare pattern that occurs mainly during drowsiness.

(2) Various associations with psychiatric symptoms have been reported over the
years.

Open Research Questions

(1) What is the incidence in various well-characterized psychiatric populations?

(2) What are the clinical correlates in the different psychiatric populations?
(3) What is the predictive value of RMTD in responsiveness to anticonvulsant
treatment?

Wicket Spikes

Description of Pattern

EEG wicket rhythms are 6–11 Hz medium-to-high voltage bursts. Wicket spikes
(WS) usually occur in adults over 50 years of age during drowsiness and light
nonrapid eye movement (NREM) sleep. No data exist on the precise distribution of
this activity during all the different sleep stages, particularly during rapid eye
movement (REM) sleep. Asokan et al. (1987), described this pattern as consisting
of mixed 2–7 and 8–14/s activity with intermingled minor sharp transients
(occasionally even frank spikes) over the anterior temporal–mid-temporal region
and, in the vast majority (84 % in this report) predominantly on the left side. This
pattern is most prominent in early drowsiness, and may change to rhythmical spiky
discharges in light NREM sleep (‘‘wicket spikes’’). WS can appear isolated or in
trains, sometimes being difficult to differentiate from more definite epileptic dis-
charges. One of WS cardinal features is a changing mode of occurrence through
any single recording: from intermittent trains of more or less sustained, arciform,
discharges resembling mu rhythm, to sporadic, single spikes. When occurring
singly, WS can be mistaken for anterior or middle temporal spikes, since they
predominate in either area, and since they share with them other characteristics
such as amplitude (60–210 lV), polarity (surface negative) duration, and config-
uration. There is some predilection for being left sided (Niedermeyer and Lopes da
Silva 2005). Batista et al. (1999) asserted that while WS can appear sporadically,
196 18 Six/Second Spike and Wave Complexes

Fig. 18.4 The Wicket Spikes can occur singly or in runs. (With permission Hughes 1994)

they, in general, present a smooth increase in amplitude, ranging from 60 to

200 lV, followed by an abrupt decrease in amplitude, to finally reach the pre-burst
baseline. When appearing as isolated spikes, it tends to have a negative potential
resulting in the phase reversal appearance which is usually taken as evidence of
pathogenicity (Fig. 18.4).

Incidence

Santoshkumar et al. (2009) reviewed EEG records of 35,249 neurology-based EEG

laboratory referrals. They observed an incidence of 0.03 % for the WS. On the
other hand, and a seemingly more careful study where only subjects with awake
and asleep records were included Reiher and Lebel (1977) from 4,458 also neu-
rology-based laboratory referrals through the years 1969–1975, WS were detected
in 39 patients (0.9 %).
Batista et al. (1999) reviewed 2,000 EEG’s from EEG laboratory referrals,
found 65 with WS (3.25 %). This is the highest rate reported in the literature.
These authors suggested that there is a left hemisphere preference for WS.
Wicket Spikes 197

Clinical Correlates

Crespel et al. (2009) presented a patient with WS and idiopathic generalized

epilepsy who had been wrongly diagnosed with focal epilepsy, which led to the
prescription of carbamazepine with severe aggravation of generalized tonic–clonic
seizures. She was referred for presurgical assessment of refractory focal epilepsy
but long-term video-EEG showed sharp theta waves over the temporal regions
during awakening, with a typical aspect of WS during drowsiness, nREM sleep
stages I and II, and REM sleep. There were a few generalized spike-waves during
sleep but interictal changes were increased in frequency at awakening with bursts
of fast-generalized spike-waves. Carbamazepine was progressively withdrawn and
the patient was progressively switched to zonisamide. The patient no longer
complained of generalized tonic–clonic seizures. At 1 year follow-up, this patient
receives zonisamide with valproate. She has remained seizure free. This case
demonstrates the need to be able to accurately identify WS.
Krauss et al. (2005) determined the clinical and EEG features of patients with
wicket rhythms who had been incorrectly diagnosed with epilepsy. Electroen-
cephalographers at an epilepsy center re-read EEGs for patients referred for epi-
lepsy management and identified patients with wicket rhythms. On further
evaluation, the majority (54 %; 25/46) of these patients were found not to have
epilepsy. We were not told if the patients with epilepsy had additional EEG
deviations? The authors compared the clinical and EEG features for the 25 patients
with WS and nonepileptic episodes with those of age- and sex-matched patients
with partial-onset epilepsy using univariate and multivariate analysis. Several
features distinguished patients with EEG wicket patterns and nonepileptic episodes
(n = 25) from age- and sex-matched patients with epilepsy (n = 25): mid-adult
age at onset of episodes (mean 38.4 years vs. 19.8 years), prolonged clinical
episodes (mean 155 min vs. 2.3 min), and long duration of EEG wicket patterns
(mean 0.66 s vs. 0.11 s spikes). After controlling for other factors, patients without
major confusion during episodes were unlikely to have epilepsy. They concluded
that this distinctive EEG pattern should be considered in patients with clinical
episodes atypical for epilepsy.
Gélisse et al. (2003) reported five observations of persistence of WS during
REM sleep. Only one patient was over 50 years of age. The authors found a
predominant expression on one temporal side, but inconsistently on the left side
(three on the left side vs. two on the right side). They have an identical mor-
phology during drowsiness or stage II sleep. There were no changes in their
location. The authors found no correlation with the tonic or phasic phases of REM
sleep.
Batista et al. (1999) while stating that WS are a benign EEG variant, they also
asserted that they are seen in many clinical situations. They compared WS patients
with 65 patients with normal EEG without WS. There was statistically significant
association between WS and age over 33. Age over 65 was significantly correlated
with the presence of WS. Adolescent age was correlated to absence of WS. There
198 18 Six/Second Spike and Wave Complexes

was an inverse correlation between WS and epilepsy. A significant association

with cerebrovascular disorders disappeared after controlling for age; a significant
correlation with headache was also related to age. While more females had WS,
the difference with males was not statistically significant. There was a great variety
of clinical situations associated with WS. Of the 65 WS patients, 4 had histories of
psychiatric problems (again it should be noted that psychiatric symptomatology
per se are not cause for referrals for an EEG) while none of the 65 control subjects
without WS had such histories. They concluded that WS are nonspecific normal
variant of the EEG that is age-related. Furthermore, Batista et al. (1999) noted a
left temporal preponderance of WS as well as documenting phase reversal in that
region indicating a surface negativity similar to most pathological EEG abnor-
malities. Reither and Lebel (1977) had already documented the surface negativity
of the WS.
Klass and Westmoreland (1985) described the WS as among EEG patterns with
a distinct morphology. Asokan et al. (1987), on the other hand, highlighted the
clinical significance of temporal minor slow and sharp activity (TMSSA) on the
basis of then available literature and their own personal observations (209 patients,
227 EEG records). The origin of TMSSA is unclear. There is reason to believe that
hippocampic ischemia might be the underlying substratum but the evidence
remains tenuous. Vertebrobasilar artery insufficiency states may result in TMSSA
since the hippocampus largely depends on this vascular system (via posterior
cerebral artery). The occurrence of TMSSA usually represents a mild abnormality
of potentially considerable clinical significance.
Reiher and Lebel (1977) from an analysis of the electroencephalograms of
4,458 patients who underwent recording during wakefulness and sleep, through the
years 1969–1975, WS were detected in 39 patients (0.9 %). They concluded that
WS should not be considered interictal abnormalities; as they do not correlate with
epilepsy. It should be highlighted though that eight of the 39 WS subjects had
history of seizures. On the other hand the conclusion that WS do not correlate with
‘‘any particular symptom complex’’ is hardly justified given that 13 subjects had
syncopal episodes and an additional 11 had ‘‘anxiety reaction or tension head-
aches.’’ Furthermore, no standardized psychiatric evaluative scales or even evi-
dence that competent psychiatric examinations were performed. Of the 39 EEG
records with WS, 31 were otherwise normal indicating that WS may be the only
indication of deviation if the pattern eventually proves to be pathological.
It is of great interest that it is currently indisputable that normal activity of
cerebral origin can sometime resemble epileptiform discharges (Klass and West-
moreland 1985). This observation has not been systematically investigated among
psychiatric populations. Is it possible that the appearance of epileptiform-like
activity is an indication of changes in the cortical excitability? The advances in
methodologies to assess cortical excitability like transcranial magnetic stimulation
and magnetic resonance spectroscopy, makes investigating this observation
feasible.
Supported Findings 199

Supported Findings

(1) Wicket Spikes (WS) are relatively rare in usual neurology-based EEG labo-
ratory referrals.
(2) Incidence and clinical correlates in psychiatric populations are not known.
(3) There seems to be a correlation with age with most patients exhibiting WS are
above 30.

Open Research Questions

(1) Is there an increased incidence of WS in psychiatric populations?

(2) If so, in which populations and what are the clinical correlates?
(3) Does the location of the WS influence significance or correlates?
(4) Does the frequency of WS influence the behavioral correlates if any?

References

Anderson RL, Vanderspek HG (1974) Psychomotor variant status epilepticus. Clin Electroenceph
5:129–132
Arfel G, Leonardon N, Bureau M, Isman ML, Naquet R (1978) Unexplained electrographic
temporal discharges (author,s transl). Rev Electroencephalogr Neurophysiol Clin
8(3):335–340 (Article in French)
Asokan G, Pareja J, Niedermeyer E (1987) Temporal minor slow and sharp EEG activity and
cerebrovascular disorder. Clin Electroenceph 18(4):201–210
Batista MS, Coelho CF, de Lima MM, Silva DF (1999) A case-control study of a benign
electroencephalographic variant pattern. Arq Neuropsiquiatr 57(3A):561–565
Boutros NN, Hughes JR, Weiler M (1986) Psychiatric correlates of rhythmic midtemporal
discharges and 6/second spike and wave complexes. Biol Psychiatry 21:94–99
Crespel A, Velizarova R, Genton P, Coubes P, Gélisse P (2009) Wicket spikes misinterpreted as
focal abnormalities in idiopathic generalized epilepsy with prescription of carbamazepine
leading to paradoxical aggravation. Neurophysiol Clin 39(3):139–142
Eeg-Olofsson O (1982) Petersén I Rhythmic mid-temporal discharges in the EEG of normal
children and adolescents. Clin Electroenceph 13(1):40–45
Egli M, Hess R, Kuritzkes G (1978) The significance of rhythmic mid-temporal discharges
(author’s transl). EEG EMG Z Elektroenzephalogr Elektromyogr Verwandte Geb 9(2):74–85
(Article in German)
Garvin JS (1968) Psychomotor variant pattern. Dis Nerv Syst 29:59–76
Gélisse P, Kuate C, Coubes P, Baldy-Moulinier M, Crespel A (2003) Wicket spikes during rapid
eye movement sleep. J Clin Neurophysiol 20(5):345–350
Gibbs FA, Gibbs EL (1952) Atlas of Elelctroencephalography, vol 2. Addison-Wesley,
Cambridge
Gibbs FA, Rich CL, Gibbs EL (1963) Psychomotor variant type of seizure discharge. Neurology
13:991–998
Gibbs FA, Gibbs EL (1964) Atlas of elelctroencephalography, vol 3. Addison-Wesley, Reading
Gibbs FA, Novick RG (1977) Electroencephalographic findings among adult patients in a private
psychiatric hospital. Clin Electroenceph 8:79–88
200 18 Six/Second Spike and Wave Complexes

Hecker A, Kocher R, Ladewig D (1979) Das minature-spike-wave-muster. Das EEG Labor

1:51–56
Hughes JR (2001) The continuous rhythmic mid-temporal discharge. Clin Electroenceph
32(1):10–13
Hughes JR, Cayaffa JJ (1973) Is the ‘‘psychomotor variant- rhythmic mid temporal discharge’’ an
ictal pattern? Clin Electroenceph 4:42–52
Hughes JR, Schlagenhauff RE, Magoss M (1965) Electro-clinical correlation in the 6/sec spike
and wave complexes. Electroenceph Clin Neurophysiol 18:71–77
Hughes JR (1980) Two forms of the 6/sec spike and wave complex. Electroenceph Clin
Neurophysiol 48:535–550
Hughes JR (1994) EEG in clinical practice, 2nd edn. Butterworth-Heinemann, Boston
Klass DW, Westmoreland BF (1985) Nonepileptogenic epileptiform electroencephalograhic
activity. Ann Neurol 18:627–635
Kocher R, Scollo-Lavizzari G, Ladewig D (1975) Miniature-spike-wave-muster: Electroenceph-
alographisches Korrelat in der Abstinen-Zphase bei Medikamentenabhängigkeit. Z EEG-
EMG 6:78–82
Krauss GL, Abdallah A, Lesser R, Thompson RE, Niedermeyer E (2005) Clinical and EEG
features of patients with EEG wicket rhythms misdiagnosed with epilepsy. Neurology
64(11):1879–1883
Lin YY, Wu ZA, Hsieh JC, Yu HY, Kwan SY, Yen DJ, Yiu CH, Ho LT (2003)
Magnetoencephalographic study of rhythmic mid-temporal discharges in non-epileptic and
epileptic patients. Seizure 12(4):220–225
Lipman IL, Hughes JR (1969) Rhythmic mid-temporal discharges: an electroclinical study.
Electroenceph Clin Neurophysiol 27:43–47
Maione R, Montanari E, Manzoni C, Mancia D, Terzano MG (1981) The problem of subclinical
localised paroxysmal rhythmic discharges (psychomotor variant discharges). Report of two
cases. Ital J Neurol Sci 2(4):343–349
Marshall C (1955) Some clinical correlates of the spike and wave phantom. Electroenceph Clin
Neurophysiol 7:633–636
Niedermeyer E (1987) Abnormal EEG patterns (epileptic and paroxysmal). In: Niedermeyer E,
Lopes da Silva F (eds) Electroencephalography: basic principles, clinical applications and
related fields, 2nd edn. Urban & Schwarzenberg, Baltimore-Munich
Niedermeyer E (2005) Abnormal EEG patterns (epileptic and paroxysmal). In: Niedermeyer E,
Lopes da Silva F (eds) Electroencephalography: basic principles, clinical applications and
related fields, pp 263–265. Urban & Schwarzenberg, Baltimore-Munich
Olson SF, Arbit J, Hughes JR (1971) Psychological testing in patients with the 6/sec spike and
wave complex: a controlled study. Clin Electroenceph 2:202–205
Reiher J, Lebel M (1977) Wicket spikes: clinical correlates of a previously undescribed EEG
pattern. Can J Neurol Sci 4(1):39–47
Santoshkumar B, Chong JJ, Blume WT, McLachlan RS, Young GB, Diosy DC, Burneo JG,
Mirsattari SM (2009) Prevalence of benign epileptiform variants. Clin Neurophysiol
120(5):856–861
Silverman D (1967) Phantom spike-wave and the fourteen and six per second positive spike
pattern: a consideration of their relationship. Electroenceph Clin Neurophysiol 23:203–217
Small JG (1968) The six per second spike and wave—a psychiatric population study.
Electroenceph Clin Neurophysiol 24:561–568
Struve F (1977) In: Shagass C, Gershon S, Friedhoff AJ (eds) Psychopathology and brain
dysfunction. Raven Press, New York, pp 199–221
Thomas JE (1957) A rare electroencephalographic pattern: the six-per-second-spike and wave
discharge. Neurology (Minneap) 7:438–442
Thomas JE, Klass DW (1968) Six-per-second spike and wave pattern in the electroencephalo-
gram. Neurology (Minneap) 18:587–593
Chapter 19
The 6–7 and 14 Positive Spikes

Introduction

Henry (1963) and Hughes (1965) published reviews on the topic. Both reviews
suggested that the 6–7 and 14 PS had some pathological correlations possibly
including some forms of epilepsy. Subsequently, a number of reports were pub-
lished that asserted a lack of correlation with epilepsy. This conclusion was based
on the observation that the detection of this pattern did not predict the occurrence
of any form of seizures recognized by the epilepsy clinical or research commu-
nities. The serious problem occurred when this lack of a close relationship between
PS and epilepsy or any other neurological conditions was generalized to mean that
the pattern in fact had no pathological correlates at all. A number of reports
appeared that concluded that the PS was an entirely normal phenomenon void of
any clinical significance. As is obvious, there is a rather large logical gap between
the two assertions. By the mid-1970s a number of reports began appearing that
related the 6–7 and 14 PS to behavioral and neurovegetative symptoms. Among
the six ‘‘controversial’’ patterns described in this section, the PS proves to be the
most controversial.
The goal of this chapter is to highlight the available literature supporting each
of the following three possibilities; (a) 6–7 and 14 PS are entirely normal EEG
variants; (b) 6–7 and 14 PS could appear normally at a certain age, to a certain
degree and in a certain stage of sleep, and are pathological if they appear outside
these parameters. If indeed this is the case for the PS, the possibility that when the
PS appear in seemingly healthy individuals that the presence of the pattern may be
indicative of some form of vulnerability or susceptibility to developing some form
of psychiatric or behavioral symptoms under proper circumstances, will need to be
investigated. As the field of Psychiatry becomes progressively more interested in
developing preventive policies and methodologies, this possibility begins to take
on more significance. In no where is this more important than in children and
adolescents where this pattern is most frequently observed; and (c) they are
entirely pathological.

N. N. Boutros, Standard EEG: A Research Roadmap for Neuropsychiatry, 201

DOI: 10.1007/978-3-319-04444-6_19,  Springer International Publishing Switzerland 2013
202 19 The 6–7 and 14 Positive Spikes

Fig. 19.1 6–7 and 14 Hz positive spikes, independent left and right mid-temporal posterior-
temporal-occipital areas. a Fourteen per-second variety; female patient, 13 years of age. b Six-
per-second variety; male patient, 24 years of age. An electroencephalograph may contain either
variety alone or both combined (Hughes 1994 with permission)

Description of Wave Form

While most commonly referred to as the 6 and 14 positive spikes (PS), work by
Hughes (1960) suggested a more accurate term to be the 6–7 and 14 PS (hereto for
called PS). The PS have a characteristic comb-like shaped waveform with a neg-
ative smooth component and a positive sharp component (Sullivan 2010)
(Fig. 19.1). This pattern was originally described by Gibbs and Gibbs (1951) who
interpreted the presence of the slow (6–7 Hz) and the fast (14 Hz) components as
evidence of hypothalamic and thalamic epilepsy. Reither and Carman (1991)
suggested that two additional patterns related to the 6–7 and 14 PS exist; a min-
iscule 28/s positive spikes and a large N-shaped potential. They ascribed no
additional significance to these patterns, they appear under the same conditions and
in the same population as the PS. Hence we will not discuss these additional
patterns further in this chapter. According to Shimoda et al. (1969), PS show a more
diffuse or wider distribution in the EEG when the clinical symptoms of the patients
become more prominent, and are relatively restricted to the posterior temporal
regions during clinically symptom-free intervals. This important early observation
Description of Wave Form 203

Fig. 19.2 Queen Square Montage

underscores the possibility that this pattern could appear in apparently healthy
children but remains limited in intensity (amplitude and preponderance) and outside
these, not clearly defined, boundaries it may be indicative of pathology.
Given the low amplitude of the PS it usually requires a combination of refer-
ential and bipolar montages to be able to more readily and confidently detect this
pattern. More specifically, the Queen Square montage (Fig. 19.2) was developed
with runs of temporal bipolar links followed by a single long referential lead
linking the temporal lobe with the contralateral ear. Figure 19.3 highlights how
difficult it is to see the pattern (particularly in a busy clinical EEG laboratory) and
how the referential leads strongly highlight the pattern.

Incidence

Incidence in Normal Children

The discussion here is qualified by the discussion regarding EEG in normal

individuals fully discussed in Chap. 2. Hence, all data presented below must be
understood in this view to be representing individuals with no gross psychiatric
204 19 The 6–7 and 14 Positive Spikes

Fig. 19.3 Effects of using combined referential and bipolar montages on facilitating the
detection of the PS. From Hughes 1994 with permission

disorders like schizophrenia or autism. More subtle disorders like attention deficit
hyperactivity disorder (ADH/D), learning disability, or some forms of less-per-
vasive behavioral problems are likely to be represented in the groups of patients
included in the studies reviewed below.
One of the major reasons many electroencephalographers concluded that PS
were of little or no clinical significance came from a few reports indicating a high
incidence in seemingly normal subjects. Of note, one particular report by
Lombroso et al. (1966) strongly suggested the normality of the pattern, since 58 %
of 155 students of a private boarding school were reported as showing PS. These
investigators indicated that a few subjects required a relatively long duration of
sleep before the pattern appeared. It should be noted that prolonged recordings to
assure the inclusion of sleep are not customary in commercial EEG laboratories.
Lombroso et al. (1966) left the impression that if the record is long enough, nearly
all subjects will eventually show this waveform. This rate of incidence in seem-
ingly healthy children reported by Lombroso et al. (1966) and later by the same
group Schwartz and Lombroso (1968) have never been independently replicated
and stands as an outlier in this literature. Despite providing reasons why low
incidence values are reported by other investigators, independent replication
simply have not been forthcoming. The reasons that were given for the lower
incidence rates reported by other investigators included (1) the use of medication
to induce sleep, (2) fast progression to deep sleep stages, (3) too short duration of
sleep, (4) insufficient use of referential recordings to see the positive spikes when
Incidence 205

Table 19.1 Incidence of positive spikes in normal children (modified from Hughes and Wilson
1983)
Paper Sample size Incidence (%) Age range
Millen and White (1954) 30 0 Infants to adolescents
Kellaway et al. (1959) 1000 2.3 Adolescents
Gibbs and Gibbs (1964) 384 20.8 10–14
Wiener et al. (1966) 24 8 15–16
Lombroso et al. (1966) 155 58 13–15
Lombroso and Schwartz (1968) 77 55 8–16
Eeg-Olofsson (1971) 599 16.2 Adolescents
Hughes (1971) 606 17 8–11
Bosaeus and Sellden (1979) 222 17.5 5–17

they did occur, and (5) different (inappropriate) age groups. Table 19.1 shows the
different studies that have been published, and lists the sample sizes and age
groups included, on the incidence of PS in seemingly normal subjects and ranges
from as low as 1.5 to 28 %, excluding the two studies by the Lombroso group with
a literature mean of 12.5. Some of the reported papers included what was labeled
as questionable PS underlying the fact that there is some level of subjectivity in
identifying the PS.
The studies of Eeg-Olofsson (1971), Petersen and Eeg-Olofsson (1981), Hughes
(1971), and Gibbs and Gibbs (1964), dealing with nearly 2000 subjects would
argue strongly that the incidence in normal children is approximately about 17 %.
One of the larger studies included EEGs from 743 children aged 1–15 years and
185 subjects aged 16–21 years chosen from among 1,177 children considered
normal on the basis of 13 different criteria (Petersen and Selldén 1981). Then 222
children were randomly chosen from the sample and further examined by a child
psychiatrist (Bosaeus and Sellden 1979). One important conclusion was that ‘‘the
most common EEG pattern in children with previous behavior disorders or other
clinical symptoms was normal resting EEG and 6–7 and 14 Hz PS in sleep.’’ A
rather important observation to be emphasized, for our purposes in this book, is
that these children were initially considered normal, but upon careful investigation
were described as having behavioral disorders. Other important findings were
positive correlations with stubbornness and poor concentration (p \ 0.001), dis-
turbed peer relationships, and school problems (p B 0.01). The closest correlation
with symptoms such as aggressiveness, anxiety, or abdominal pain appeared in
children with PS. A replication utilizing the currently available clinical rating
scales is sorely needed (see recommendations for further studies below).
The data above, not withstanding the caveat of the difficulty in defining nor-
mality, strongly suggest that PS can appear in seemingly healthy children and
opens the door for a new field of investigation of the significance of the pattern in
healthy children (i.e., what kind of susceptibility, if any, do they indicate?).
206 19 The 6–7 and 14 Positive Spikes

Incidence in Adults

There is fairly unanimous agreement that the pattern is highly age related. It is
rarely encountered in the very young and increases to a peak prevalence during
adolescence after which the pattern becomes exceedingly rare throughout adult-
hood (Torres et al. 1983; Cervone and Blum 2007; Chaloner and Pampiglione
1983). In the Gibbs control series of 616 normal adults aged 20 and above (Gibbs
and Gibbs 1963), the prevalence of PS drops rapidly to 1.3 % by age 25 and
reaches zero percent by age 40. Because PS in normal individuals become
increasingly rare throughout adulthood, it may be easier to demonstrate associa-
tions between this pattern and clinical phenomena by studying older people—a
research strategy that has seldom been pursued. For example, Wegner and Struve
(1977) compared the incidence of PS in the routine wake and sleep EEGs of 2,888
consecutively admitted psychiatric patients aged 20 and above (all psychiatric
admissions were required to have an EEG examination) with the above-referenced
normal control series of Gibbs and Gibbs (1963). When this was done they found
the elevated incidence of PS among adult psychiatric admissions compared to the
adult normal population to be overwhelmingly significant at all age categories (age
20–24, 27.4 vs. 8.7 %, X2 = 31.8, p \ 0.0000001; age 25–29, 20.9 vs. 1.3 %
X2 = 17.9, p = 0.00002; age 30–39, 13.3 vs. 0.9 %, X2 = 14.8, p = 0.002; age 40
to 49, 7.4 vs. 0.0 %, X2 = 6.97, p = 0.01, and age 50 to 59, 4.6 vs. 0.0 %,
X2 = 3.67, p = 0.056). In another study specifically addressing the issue of PS in
an adult population (Hughes and Cayaffa 1978), 460 adults aged 30 to over 80 with
PS were contrasted with a variety of comparison groups matched for age, gender,
and race. Neurovegetative or psychiatric symptoms, either occurring separately or
combined, were present in 85 % of the adults with PS and the authors concluded
by arguing for the clinical significance of PS when encountered in the adult
patient. One of the important features of PS in the adult is that no one has found a
single case over 41 years of age in a normal control group (Gibbs and Gibbs 1964;
Hughes and Cayaffa 1977; Petersen and Selldén 1981). These data argue for a
greater significance for the pattern in this older age group than in early
adolescence.
Crowley and Liske (1967) reported on the incidence of PS in aircrew personnel
whose ages ranged from 19–47 years with a mean of nearly 30. The incidence of
this waveform among these ‘‘normal’’ subjects was 0.6 % and among medical
referrals was a similar 0.7 %. Only 36 patients showed PS in the two groups, and
within those two groups 12 reported headaches, a similar number with abdominal
symptoms or loss of consciousness, including seizures, and a few had some kind of
dizziness episodes. Since over 500 patients had been referred for loss of con-
sciousness and only 10 patients with these spikes were found with this type of
symptom, the authors concluded that the correlation was no more than 1.6 %.
Also, since over 200 patients had been referred for headache and only 12 patients
had headaches and PS, the correlation was less than 1 %. The authors concluded
that the clinical correlations were weak for loss of consciousness, headaches,
Incidence 207

abdominal complaints, and character disorders, so that this waveform may thus be
a normal variant.
Walsa (1968) reported on 66 patients with PS from the ages of 19–63 years
with a mean age of 35 years. Half of the patients showed paroxysmal clinical
findings, 19 had typical epileptic seizures and 12 reported various neurovegetative
symptoms associated with anxiety. Headaches seemed to be the most common
symptom, but disorders of appetite and sleep were also found. The most common
etiology was head injury, and the slow form (i.e., 6/s) was found in all of these
adults whereas the faster form (i.e., 14/s) was found in only 17 of the 66 adult
patients. Walsa (1968) concluded that the presence of this pattern did support a
diagnosis of atypical epilepsy.
Gibbs and Gibbs (1973) reported on 207 patients with PS over the age of
29 years. Since these patients came from a referral group of 3452, the incidence of
6 % was given for the pattern in the adult, but this percentage was 20 times the
incidence in controls (0.3 %). Emotional instability was found in the PS patients
with an incidence of 24 %, which was 1.7 times the incidence found in those with
normal records (14 %). The instability showed itself in the form of rage, episodic
irritations, and depression. Headaches were found in slightly over half, nausea and
vomiting in one-fifth, and questionable epilepsy in 28 % of patients with these
spikes, but the latter values were not significantly different from the controls. Since
59 % of the patients had symptoms less than 3 years prior to the EEG and 41 %
less than 1 year, these symptoms were not considered to be an adolescent pattern
that continued into adulthood. No more than 15 % of the patients had symptoms
before adulthood. These authors reported that nearly half (46 %) of the PS were
seen in the waking state. The same authors had previously made the point that
younger children with the pattern frequently show these spikes during deeper
stages of sleep and, as age increases, they appear in lighter and lighter stages
(Gibbs and Gibbs 1964). Thus in adults PS frequently appear in the waking record.
Hughes and Cayaffa (1978) investigated a large number of adults with PS. This
study involved 460 adults from the third to the eighth decade with PS, found after
screening more than 50,000 patients. Control groups were matched for age, sex,
and race and included (1) a random hospital population, (2) those with normal
EEGs, (3) those with abnormal EEGs, and (4) patients without evidence of organic
brain disease. In the PS group with a female to male ratio slightly greater than 2 to
1, neurovegetative (61 %) and psychiatric (41 %) types of complaints were often
found, one or the other, in 85 % of the adults with this waveform. The incidence of
neurovegetative complaints was significantly greater than in all control groups
except those with normal EEGs, and the number of psychiatric complaints was
greater than in the random hospital controls and less than in the normal EEG
group. Clinical seizures were found in 11 %, but the majority of the latter patients
had other paroxysmal abnormalities in their record to account for the attacks. In
adult females, neurovegetative complaints including dizziness/vertigo were found
especially in the 20s and psychiatric complaints in the 50s, mainly anxiety,
depression, and psychosis. A significantly increased incidence of slow waves and
paroxysmal sharp wave or spike abnormalities was found in the PS group,
208 19 The 6–7 and 14 Positive Spikes

compared to the EEG controls without organic brain disease. Arguments for and
against the clinical significance of PS in the adult were presented in this paper,
which concluded that sufficient evidence exists in favor of its significance to
warrant at least further attention to this waveform.
Beun et al. (1998) examined the all-night EEGs of 60 adults (18–50 Years of
age) said to be ‘‘healthy.’’ Six subjects exhibited PS (10 %). This is one of the
studies where carefully evaluated subjects for normality were included. None the
less, there are no indications that they had an adequate psychiatric evaluation, or
that minor psychiatric problems like ADHD or possibly Axis-II personality dis-
orders were excluded. Furthermore, family history was not assessed (see Chap. 2).
None the less, the report clearly documents the presence of this pattern in adults.

Incidence in Pathological Conditions

General incidence in EEG laboratory populations was reported in a number of

studies. EEG laboratory populations are by and large individuals who are referred
for the purpose of ruling out epilepsy. A much smaller percentage of individuals
are referred for complaints of headache or for psychiatric symptoms (again to rule
out epilepsy as a root cause for the behavioral problem). The paper by Wang et al.
(1991) is a good example. Wang et al. (1991) reported EEG data from 2,026
Chinese patients ranging in age from 1 to 18 years old. They found an overall
incidence of 2.52 % of PS with the highest incidence in the 6–10 years range.
Twenty five of the 51 patients exhibiting PS had episodic attacks of headaches,
abdominal pain, and other autonomic symptoms. Of those 25, 19 were eventually
diagnosed as having autonomic epilepsy. These findings are in large agreement
with much earlier studies. Millen and White (1954) examined the EEG records of
547 patients referred for an EEG study. They also examined 30 additional children
who were not referred for an EEG and were considered ‘‘healthy.’’ They reported a
rate of 4 % in the EEG-referral population and zero % in the healthy group.
Radhakrishnan et al. (1999) reported on 1,778 subjects examined over a 3-year
period, included individuals ranging in age from 2 months to 89 years. They only
included EEG records that included both wakeful and drowsy/sleep tracings. For
children ages 6–15 years, they had an incidence of 18.5 % with a relatively higher
incidence in girls (21.3 vs. 15.5 % for boys). The highest incidence was found in
the age range from 16 to 25 years of age (32.9 %) with a significant reversal of the
gender difference with the pattern being seen in men at 52.9 % while remaining
21.4 % in women. A much more recent report examined 35,249 individuals from a
similar EEG laboratory population (Santoshkumar et al. 2009). This sample was
collected over a span of 35 years. They report an incidence of 0.52 % for PS.
While this is one of the largest sample sizes reported, it is also one of the lowest
rates of incidence reported. As mentioned above, the population studied has
minimal psychiatric representation as is the case in most neurology-based EEG
laboratories.
Incidence 209

Psychiatric Populations

A number of studies have dealt with the incidence of PS in patients with psy-
chiatric disorders. For example, Struve and Honigfeld (1970) determined the
incidence of these spikes among over 800 patients in a psychiatric hospital,
especially with acute psychiatric conditions with an age range of 14–60 years, but
with the majority between 18 and 25 years. The incidence of 28 % was relatively
high when the age range is considered. Later Struve and Pike (1974) enlarged this
particular study and reported on over 4,000 psychiatric patients whose ages were
above 14 years. The incidence of patients with PS in this study was 20 %, again
somewhat higher than anticipated because of the age range. The latter value was
above the 14 % of Gibbs and Novick (1977), whose patients were all adults with
psychiatric disorders, but this incidence represented the most common abnormality
in the group, very different from the incidence in controls (3 %). The incidence of
PS was 21–24 % in each of the four groups with emotional instability, personality
changes, hallucinations, or anxiety neurosis.
Gianturco et al. (1972) studied over 100 teenage patients who had (1) psy-
chological disturbances, (2) neurovegetative symptoms, or (3) the combination of
these latter two kinds of complaints. A psychological or behavioral disturbance
was found in 19 %, a neurovegetative or dysautonomic disorder was found in
29 %, but a combination of these later complaints was found in 56 % of patients
with PS, more often males. It was this combination of symptoms that showed the
greatest statistically significant difference (p \ 0.001) from the control group.
Gianturco and colleagues reported a clinical trend that the major difference
between the patients with PS and the control group was with regard to the presence
in the former group of psychoneurosis, especially neurotic depressive reactions.
Gibbs and Novick (1977) recorded EEGs of 1,000 adult psychiatric patients. They
found positive spikes in 28 % of patients with dizziness, 24 % with abdominal
pain, 23 % with headaches, and 21 % of patients with nausea and vomiting.

Incidence in Adult Criminals

Olson et al. (1970) studied over 100 inmates of an Indiana state prison whose ages
were between the third and eighth decades, mainly in their 30s. Five patients
showed PS and accounting for 5 % in that age group. The authors point out that
this latter percentage is greater than for the normal population at that age, namely
0.9 %. Since so few patients were studied with the spikes, no definite conclusions
can be drawn from this study. Tasher et al. (1970) studied over 100 males who had
committed violent criminal acts whose ages were mainly in the 30s, but ranged
from slightly under 20 to slightly over 50 years. Seven of these patients showed
PS, and similar to the above study no definite conclusions could be drawn
210 19 The 6–7 and 14 Positive Spikes

regarding a possible increased incidence in this group, although the data may be
suggestive. The more commonly found waveform in this latter group was the
mitten pattern, previously described by Gibbs and Gibbs (1964) (see Chap. 16).

Incidence in Learning Disabilities

Positive spikes have been described by a number of investigators in dyslexic

patients, but a considerable range in incidence has been reported. For example, in
Hughes and Park (1968) a 21 % incidence was reported, in Knott et al. (1965)
30 %, in Bryant and Friedlander (1965) 35 %, and in Muehl et al. (1965) 55 %.
Torres and Ayers (1968) found no significant difference in the PS in dyslexics
compared to those in controls, nor did Bryant and Friedlander, who reported a
28 % incidence in the control group. In a study on learning disorders with 214
underachievers compared to matched controls, the increased incidence of PS in the
former group did not reach statistical significance (Hughes 1971).
Milstein et al. (1978) reported a significant relationship between the presence of
PS and low scores on the Concept Level Analogy Test, more significant than any
other EEG finding. However, when the study was repeated on only four PS
patients, eliminating possible distracting influences, the positive findings could not
be replicated.
As learning disabilities are many and heterogeneous, expecting any pattern or
abnormality to correlate highly with the entire group is unlikely. In fact certain
correlations have been found between PS and various types of learning disabilities.
In a study of dyslexics (Hughes and Park 1968), children with PS were the
brightest among four different EEG groups but had the greatest difference between
potential and actual achievement in reading. From studying children with these
spikes, Small et al. (1968) reported on various trends referring to certain results
with probability values between 0.05 and 0.10. Children with PS showed a trend
toward scholastic performance below the age-expected level, in addition to family
history of epilepsy, positive neurological signs, and more questionable brain
damage. However, in the investigation by Muehl et al. (1965), PS did not correlate
with any psychological test score.
The general conclusion from these investigators was that the occurrence of PS
in children was significantly associated with strong evidence of brain damage,
based on the presence of three or more clinical or laboratory criteria. Further
evidence of this general point came later (Small et al. 1977) when autistic children
were studied. These investigators found that the spikes were significantly associ-
ated with neurological deficits, as defined by the Brain Damage Index of DeMyer
(DeMyer et al. 1977).
Smith et al. (1968) found that, in patients with LD and PS, ethosuximide
improved the verbal and full-scale IQ but did not change the personality and motor
performance test scores in these patients.
Incidence 211

Head Injury

As shown by Gibbs and Gibbs (1987) the relationship is complex. Gibbs and Gibbs
(1977) studied 2,640 patients involved with medico-legal suits and found PS as the
most common EEG finding (23.0 %). Among patients with skull fractures, those
with basal fractures more often (64 %) showed these spikes than those with
depressed (50 %) or linear fractures (35 %). The authors also reported that the
incidence of asymptomatic cases at all ages was higher (19 %) among the patients
with normal EEGs than among those with PS (12 %). It was somewhat surprising
to find that in the 2–9 year age range, negative spike foci were reported as more
commonly asymptomatic than positive spiking.
The potentially causal role of head trauma was suggested by a large EEG study
of over 2,000 head injury cases by Gibbs and Gibbs (1977). This study, which was
originally presented with percentage data, was later subjected to rigorous statistical
analyses (Struve and ramsey 1977) which substantially increased the support for
the role of head trauma as one probable cause for this controversial dysrhythmia.
Specifically, PS incidence was shown to increase while normal EEG incidence
decreased as severity of the skull fracture increased from linear to depressed to
basal (Cochran X2 for linear regression, p \ 0.05) and the incidence of positive
spiking also increased whereas the incidence of normal EEGs decreased as the
severity of the immediate post-traumatic state increased from conscious to dazed
to unconscious (Cochran X2 for linear regression, p = 0.0003). Furthermore, when
for each discrete symptom (headache, dizziness, nausea, temper dyscontrol, etc.) a
separate X2 contingency was formed contrasting the presence or absence of the
symptom with the presence or absence of positive spiking, symptom presence was
always significantly associated with positive spiking (X2 values ranged from 28.6
to 83.3 with p \ 0.0001 for all comparisons). Struve and Ramsey (1977) con-
clusions were that the incidence of PS was directly and linearly related to the
increasing severity of fractures and significantly different from chance. If the
patients with normal records or those with focal spikes were included into a non-
PS group, then the increase in the pattern from retained consciousness (33 %) to a
dazed condition (46 %) and to unconsciousness (48 %) is associated with a high
level of significance (p = 0.0003). Other findings were a very significant linear
decrease in asymptomatic cases, related to the increasing amount of positive
spiking and an association of the pattern with symptoms. The authors concluded
that ‘‘the results of the data analysis presented are incompatible with the view of
PS as a spontaneous normal electrical wave of brain origin.’’

Behavioral Correlates

Gibbs and Gibbs (1977) presented data on a less investigated variable, namely the
amount of spiking in the record. As the voltage and number of bursts increased, the
212 19 The 6–7 and 14 Positive Spikes

incidence of asymptomatic cases decreased (14–10 %); thus the authors showed
that an increasing amount of positive spiking raises the probability of symptom-
atology associated with those spikes. In patients who were conscious after a head
injury, the incidence of PS was 33 %, in contrast to those who were dazed (46 %)
or unconscious (48 %) (Struve and Ramsey 1977). The incidence of PS increased
from 19 to 36 % in patients unconscious from minutes to 23 h but then diminished
to 22 % in those unconscious from 3 to 6 days. For patients 5–9 years of age with
a normal EEG, 23 % had temper tantrums, in contrast to 68 % of those with PS.
Other differences included episodes of blurred vision, behavior problems, episodic
pain, and emotional instability, seen more often in patients with these spikes
(Gibbs and Gibbs 1977).
Greenberg and Pollack (1966) examined the clinical correlates of the presence
of the PSs in a group of patients (N = 60) between the ages of 16 and 35 con-
secutively admitted for inpatient psychiatric care. All EEGs included sleep trac-
ings. Of the 60 patients, nine exhibited PSs (15 %). Of these nine patients, eight
carried the diagnosis of schizophrenia. An age and gender-matched control group
with similar diagnosis but had normal EEGs was also examined. Patients with PSs
were rated as having less affective relatedness, less understanding of the conse-
quences of their actions, greater suicidal tendencies, more aggressive behavior,
and management problems, and were more often discharged as clinically
unimproved.
An intriguing observation that is of significant clinical importance and is in
need of replication was reported by Rau et al. (1979) related to compulsive eating.
Fifty-nine patients (51 women) mainly in their late 20s and early 30s were
examined as part of probing the effects of phenytoin in the treatment of com-
pulsive eating. The most important finding was an extremely high rate of EEG
abnormalities in general (64.4 %). Of the abnormal EEGs, 57.9 % exhibited PS
(37.3 % of the entire sample).
Millen and White (1954) provided a detailed list of symptoms exhibited by
children with PS. From 82 symptoms recorded from 22 patients, the largest per-
centage was psychiatric in nature (38 of 82) with erratic behavior, restlessness,
emotional outbursts, and temper tantrums being most common. A host of other
symptoms like headache, mental retardation, vague pains, and blackouts or
fainting spells were also reported.
Overall, two major sets of behavioral correlates have been reported in associ-
ation with the 6–7 and 14 PS; a group of symptoms labeled neurovegetative
symptoms and psychiatric symptoms including, albeit less commonly, violence
and aggression.

(a) Neurovegetative Symptoms: Many of the early reviews on positive spikes

(Gibbs and Gibbs 1964; Henry 1963; Hughes 1965; Eeg-Olofsson 1971)
pointed to the dysautonomic or neurovegetative symptoms frequently found in
patients with this waveform. Subsequent studies added further evidence. For
example, Takahashi (1966), studied over 200 cases with PS and divided the
records into the slow 6/s, fast 14/s, and the ultra fast 28/s form. The nature of
Incidence 213

the clinical symptoms in 22 % of these patients led Takahashi to use the term
autonomic epilepsy. Autonomic seizures refer to recurrent episodes of auto-
nomic symptoms when associated with paroxysmal EEG discharges (Kurata
1982). Hotta and Fujimoto (1973) suggested that if the accompanying EEG
changes include epileptic discharges with and without PS can be called
autonomic seizures while if the episodic symptoms are only accompanied by
the PS to call them nonepileptic. Symptoms common to autonomic seizures
are the neurovegetative symptoms reported in association with the PS—
migraines, dizziness, abdominal pain, and episodic vomiting. Of the 52 sub-
jects examined by Kurata (1982) with episodic autonomic symptoms, 13 had
frank epileptic discharges alone, 11 had both frank epileptic discharges and
PS, and 16 had PS alone with 12 patients exhibiting no EEG abnormalities.
Domenici et al. (1991) concluded based on the literature and their experience
working in a school environment that the association between the 6–7 and 14
PS and vegetative symptoms is undeniable. They conducted a study of 617
consecutive EEGs (all including wake and asleep recordings) from children
aged 5–16 years of age who were referred to the Neurology outpatient clinic.
All periodic symptoms were carefully assessed. All EEGs included relatively
extended sleep tracings (at least 25 min) and all EEGs were interpreted
independently by two EEGers. One important finding was the need for com-
bined bipolar and monopolar montages to adequately record the PS (see
Fig. 19.2). Of the 617 subjects 109 exhibited PS (17.6 %), 63 of them pre-
sented with vegetative periodic symptoms. Of the 510 cases without PS, 91
had episodic autonomic complaints. These data translated to PS having a
sensitivity of 40.9 % and a specificity of 90.1 % and a predictive value of
57.8 % for the diagnosis of an autonomic periodic syndrome.
Silverman (1967) reported that the most frequent symptoms in his patients who
had both PS and 6/s spike and wave complexes were headaches or syncope, found
in 37 % of the patients. Hughes et al. (1965) showed a significant relationship of
neurovegetative symptoms to patients with behavior disorders and PS. As stated
earlier, Bosaeus and Selldén (1979) examined 222 seemingly healthy children and
report a significant relationship of PS with food problems (p \ 0.05), but espe-
cially with sleep disturbances (p \ 0.001) as possible manifestations of neuro-
vegetative disorders. Consistent with these findings is another study (Rau et al.
1979), which reported a 37 % incidence of PS among compulsive eaters, repre-
senting 60 % of the patients with some positive EEG finding. In all age groups up
to the sixth decade, a higher incidence of these spikes was found than in psy-
chiatric inpatients or controls.

(b) Psychiatric Correlates

Gianturco et al. (1972) compared patients with PS and those without this
waveform with regard to psychological symptoms. Though the majority of
both groups had personality disorders, it was the presence of psychoneurosis
214 19 The 6–7 and 14 Positive Spikes

that distinguished the group with PS compared to the controls. However, this
latter change was considered only a trend.
Sullivan et al. (1963) reported on a seven-member family with the mother
having a normal EEG and the other six members with abnormalities. Four of these
patients showed PS and excessive posterior slow waves, while three showed
temporal lobe abnormalities. A high level of impulsivity was found in all of the
family members (except for the mother, whose record was normal) and specifically
in controlling their anger. Compared to their mothers’ score, the psychological test
scores were clearly different in those with EEG abnormalities, but the patients with
PS did not significantly differ from those with temporal lobe dysfunction. Cas-
tellotti and his colleagues (1966) studied only five patients with PS but concluded
that the symptoms seen in these patients often were an emotional disturbance of
paroxysmal character. Andy and Jurko (1972) reported a co-existence of the PS
deep in the diencephalon and paroxysmal behavior disorders in the form of
emotional instability. They suggested the term hyper-responsive syndrome,
especially since the one characteristic which appeared to permeate all the symp-
toms of this patient was the element of hyper-reactivity. These symptoms could be
considered an exaggerated manifestation of otherwise normal or usually occurring
physiologic functions. Struve et al. (1972) attempted to predict the presence of
these spikes based on an interview with the patients as they were being prepared
for EEG recording and, in particular, an assessment of their affect. Two techni-
cians successfully predicted the presence of PS in 50 and 58 % of the 120 patients
and also correctly predicted that this pattern would not be seen in 81 and 82 % of
the patients who showed no spikes. These values reflect a statistically significant
difference from chance at a probability value less than 0.025. A physician also
made predictions based on lengthy comprehensive professional interviews and
successfully predicted the presence of PS in 75 % of patients and predicted
accurately the absence of these spikes in 71 %. The values reflect a p value below
0.02. Thus, these successful predictions may well reflect the association of certain
psychological or behavioral patterns with positive spikes.
Weisz et al. (1978) reported on the successful prediction of PS based on clinical
criteria. In addition to behavior disorder, hyperactivity, and learning disability, the
neurovegetative symptoms of chronic headaches and episodic abdominal pain
were also used for prediction among 1,096 children and teenagers. For this total
group, successful prediction was 89 % (p \ 0.01), and especially for the 491
patients under 15 years of age, this value was 93 %. The most sensitive parameters
were behavior disorder, hyperactivity, and learning disability for those under
15 years old. The authors concluded that PS were highly predictable in a select
population with the aforementioned symptoms and may be an EEG manifestation
of a central nervous system disturbance reflected in these patterns.
Boutros et al. (1998) gathered diagnostic information through structured
interviews from four groups of psychiatric inpatients aged 4–17 years. The four
groups were; (1) patients exhibiting PS (N = 25); (2) patients exhibiting frank
epileptic discharges (N = 29); (3) patients exhibiting slow wave abnormalities
Incidence 215

(mainly focal) (N = 23); and (4) patients with normal EEGs (N = 25). ADHD
symptoms were significantly more frequent in the PS group compared to the other
three groups combined. Anxiety symptoms showed a strong trend (p = 0.06) to be
more represented in the PS group as well.
Other studies have concluded that the relationship between PS and psycho-
logical behavioral disorders is unclear or nonexistent. During the 1960s, Engelhart
and Knott (1964) studied 111 psychiatric patients, 38 with PS, 27 with other
abnormalities, and 46 with normal records, and compared these groups on the basis
of MMPI scores. No meaningful MMPI differences were related to the EEG
categorization, and the authors maintained that claims for explicit personality
characteristics of patients with PS seemingly must continue to rest on impres-
sionistic grounds. Loomis (1965) compared two groups of patients. One group was
referred because of delinquency and the other group was randomly selected from a
training school. Positive spikes were seen in 30 % of the referral group and 20 %
of the randomly selected group, but these values were not considered significantly
different. Wiener et al. (1966) studied 80 delinquent and 70 nondelinquent male
adolescents. Since 13 % of the delinquents showed PS and the same percentage of
the nondelinquents showed the same waveform, the authors concluded that there
was no relationship between the presence of this pattern and juvenile delinquency.
Avery (1968) also studied the PS in patients with behavioral problems. Utilizing a
projective test battery including 51 personality factors, the patients with PS
showed significantly lower scores (p \ 0.01) on judgment and success of control
mechanisms. This author pointed out that from the total of 113 scoring evaluators,
only three items showed a statistically significant difference which, according to
Avery, appeared to be on a random basis. Although all the patients with PS had
frequent lapses in control or chronic dyscontrol problems and significantly more
showed this finding than in the control group, the author concluded that the finding
was on a random basis. Thus the conclusion was that PS were not a separate
clinical entity represented by a particular pattern of results on psychological tests.
Finally, Pollack et al. (1969) reported on 70 psychiatric patients and 100 siblings
of those same patients. PS were found in 24 % of the patient group and 25 % of
the siblings, but there was no clear relationship between the presence of these
spikes in a given patient and the ratings of psychological abnormality. The authors
concluded that these studies failed to ascribe clinical significance of PS in psy-
chiatric patients. This conclusion is not surprising given the age of the study, and
the availability of methodology to cluster symptoms for effective examination of
correlations. Also, without age- and gender-matched healthy control group it is
hard to draw definite conclusions.
In the 1970s, further negative studies also appeared. Milstein and Small (1971)
studied 36 patients with PS and compared them to controls with a normal EEG and
of the same age, sex, and educational level. All patients and control subjects came
from a psychiatric institution. The patients with PS were different from the con-
trols with regard to the Cornell Medical Index Subscale describing musculoskel-
etal signs, anger, and other indicators of hostility; they also differed in frequency of
illness and scores on the Minnesota Multiphasic Personality Inventory (MMPI)
216 19 The 6–7 and 14 Positive Spikes

paranoia subscale. Surprisingly, however, the matched controls with normal EEGs
had higher values. There was no difference in the Institute for Personality and
Ability Testing (IPAT) Anxiety, Raskin Mood, and Zung Depression scales. Thus
very few differences were found in the extensive neuropsychological battery
among psychiatric patients divided into those with normal records and those with
PS. The authors did point out that ‘‘the significance of such waveforms, at least in
a psychiatric population, may be quite different than in a neurologic or general
clinical group of subjects.’’ Gibbs (1972) maintained that in children with PS, no
more than 2 % have rage attacks, and of these no more than a quarter would likely
inflict injury on someone else, based on their previous history. Small et al. (1978)
found 7 of 21 hyperkinetic children (AD/HD mainly hyperactive by today’s
diagnostic system) to have positive spikes, but this incidence was not significantly
different from the controls. Furthermore, Riley and Niedermeyer (1978) found
only 9 % of teenagers with episodic behavioral manifestations with these spikes,
an incidence below the normal control values of most studies. This again high-
lights the necessity for establishing the firm boundaries of normalcy as discussed
in Chap. 2.
Relatively more recently, Small et al. (1997) examined the EEGs of 202
patients admitted to a psychiatric hospital in a state of acute mania. All patients
had admissions EEGs every time they were admitted (thus there were 131 repeat
EEGs). Patients were carefully diagnosed based on the DSM version of the time
(from DSM-III to DSM-IV in addition to the Schedule for affective Disorders and
Schizophrenia life time version (SADS-L) from appropriate informants. EEGs
were on the average 60 min and included both referential and bipolar montages.
All EEGs were interpreted blind of the diagnosis. Sixteen percent exhibited sig-
nificant EEG abnormalities beyond what can be ascribed to the effects of medi-
cations. PS were detected in only 13 (8 %) of the patients.
The above studies underscore the difficulties in finding ready correlations and
do not justify simply abandoning this line of investigation. Most of the above-
mentioned studies reported EEG records for presence or absence of PS when the
literature strongly suggests that a number of variables could be significant. Among
these the frequency, amplitude, stage of arousal, and age of subjects are important.
The complexity of the relationship is further exemplified by the work by Gibbs and
Gibbs (1987) examining the relationship between PS, severity of head injury, and
durations since the injury is an example of how the relationship could be complex
and defies simple correlation studies. Gibbs and Gibbs (1987) provided evidence
that PS is a delayed reaction to mild head injury in children but can also be seen in
children recovering from more serious injuries.
In 1987 and driven by their own clinical experience, DeLong and colleagues
compared the clinical symptoms of two sets of children, 100 subjects each with
one group exhibiting PSs and the other with normal EEGs, all referred for a
clinical EEGs. The presence of the PSs correlated significantly with the presence
of behavioral problems and aggression. Disturbances of temper, mood, attention,
learning, and sleep were the major correlates of the PSs.
Incidence 217

Additional Evidence of the Pathological Nature of the PS

Small et al. (1968) also pointed to other EEG abnormalities in patients with PS.
Specifically, generalized or multifocal ‘‘epileptoid’’ EEG abnormalities such as
spike and wave complexes, spikes or sharp paroxysms were reported as very
prominent in the records of patients with PS, mainly children. This latter finding
was one of the reasons why Small and her colleagues concluded that the occur-
rence of PS in children was significantly associated with strong evidence of
organic brain damage. Hoshika et al. (1981) reported a 38 % incidence of gen-
eralized spike and wave complexes (or high-voltage slow wave bursts), a 14 %
incidence of focal spikes and slow waves predominantly on the side where the
positive spikes were more often found. As one negative finding, Small (1971)
reported that PS were not significantly associated with the presence of photo-
convulsive or photomyoclonic responses.

Are PSs Simply Sleep Spindles Variants?

Since PS have a morphological similarity to normal sleep spindles, one obvious

interpretation of these spikes is that they are spindles slightly modified in their
configuration and polarity. However, various findings argue against this sugges-
tion. For example, Gibbs and Gibbs (1971b) pointed out that 5 % of patients with
PS show these patterns only during the waking record and 5 % more show the
pattern during both wake and sleep; thus, 10 % demonstrate the pattern during
wakefulness. The presence of this pattern in a waking record with clear alpha,
noted mainly in adults, argues strongly against the possibility that PS are simply a
modified sleep spindle. On the other hand, Friedlander (1969) reported that there
was a significant difference between normal sleep spindles and 14/s PS with regard
to the average frequency, but not between the spindle and the 6/s form. The author
thus suggested that the 6/s form, and not the 14/s PS, may be related to the sleep
spindle. The results of other studies would argue against this possibility. For
example, Okuma et al. (1968) showed that PS increase in their incidence during
the rapid eye movement (REM) stage of sleep, and the incidence of these spikes
was relatively low in the spindle phase. Their conclusion was that PS could not be
related to spindles since the highest incidence occurred when spindles were absent.
Also, Tsuzuki (1967) carefully studied five patients with nocturnal sleep. The
author concluded that, since PS were noted especially during REM sleep, this
finding would cast serious doubt on the view that the spikes are modified spindles.
Moreover, In 1963 Walker and Marshall recorded PS in the depth and found no
relationship between sleep spindles and positive spikes either in time or in
localization within the brain.
218 19 The 6–7 and 14 Positive Spikes

Treatment Implications

The clinical symptoms of neurovegetative or dysautonomic disorders and behav-

ioral or psychiatric complaints are commonly found in many patient populations,
and an important practical question is what type of medication, if any, may be
efficacious in these patients. Gibbs and Gibbs (1971a) reported on 1,462 with PS,
half of whom were treated and the other half untreated with an average 4-year
follow-up. Without medication 10 % improved greatly and 28 % became
asymptomatic, totaling 38 %, or about one-third, showing an improvement without
treatment. This percentage contrasts with the 21 % who improved greatly on
anticonvulsants in addition to the 45 % who became asymptomatic, totaling 66 %,
or two-thirds improving with treatment. While one of the earliest studies exam-
ining the predictive ability of the presence of PS to a favorable therapeutic
response to anticonvulsant medication therapy, it is also the largest and most
definitive. Gibbs and Gibbs also reported on 25 patients who began medication and
then discontinued the drugs; the symptoms reappeared but later disappeared with
reinstitution of drug therapy (Gibbs and Gibbs, 1971a). The general conclusion
from these investigators was that anticonvulsant drugs were especially effective
against symptoms that appeared paroxysmal.
Gianturco et al. (1972) reported that anticonvulsants had no appreciable effect
on the psychological or behavioral disorders but that the dysautonomic or neu-
rovegetative symptoms were often dramatically improved. In the study on com-
pulsive eaters, Rau et al. (1979) reported that phenytoin improved 89 % of the
patients with positive spikes, significantly different from the 42 % with other
abnormalities or 35 % with normal records. Kurata (1982) examined the thera-
peutic effects of carbamazepine (CBZ) on episodic autonomic symptoms in 52
patients. Of 40 patients with autonomic seizures without generalized epileptic
seizures, 90 % became free of pain following treatment. In the 12 patients with
both types of seizures 83 % became free of headaches.
Boelhouwer et al. (1968) reported a double-blind control study on the efficacy
of various types of medication for behavior disorders with PS. Two groups were
studied, one with and one without this pattern; those with the waveform were
found to have significantly less success with control mechanisms. The Minnesota
Hartford Personality Assay (MHPA) test separated the two groups on the basis of
five factors; the patients with PS were more ready to feel guilty, more severe with
self-criticism, and had more dissociative concerns and thought disorders. Also,
these patients showed greater physical destructiveness, assaultiveness, and excit-
ability, in addition to poor impulse control. The authors then tried phenylhydantoin
and thioridazine, phenylhydantoin with a placebo, and thioridazine with a placebo
in a double-blind study. In nine ways the scores on the MHPA test were improved
(reduced) for the PS group with the combination of phenylhydantoin and thio-
ridazine. Improvements on test scores from the other drug regimens for either
group were less than with the latter combination in the PS patients.
Supported Observations 219

Supported Observations

(1) The PS may be associated with certain types of discrete symptoms which cut
across psychiatric diagnostic boundaries and can be found in people with
almost any condition, or for that matter, no psychiatric diagnosis at all. Taken
as a whole, the available literature suggests that PS-related symptoms cluster
in two basic categories. The first grouping involves various physiological and/
or autonomic symptoms such as headache, stomach ache, nausea, flushing,
spells of dizziness, paresthesias and the like with headache being the most
common followed by abdominal symptoms. The second symptom cluster
involves temper dyscontrol and related phenomena such as irritability and
emotional lability. More recently, a possible contribution of the PS to ADHD
syndromes was suggested (Boutros et al. 1998). Prototypes for the first
symptom cluster are the well-argued papers on paroxysmal pain and auto-
nomic disturbances by Kellaway et al. (1959) and Sheeby et al. (1960).
(2) Symptom correlates of positive spiking may range in severity from being
mildly expressed where they may cause the individual little difficulty to being
so strongly expressed that they form the basis for clinical complaint.
(3) It has been estimated that approximately 20 % of individuals with this EEG
pattern may be asymptomatic in that none of the expected symptoms are
present. The implications, if any, of the appearance of PS in healthy subjects is
currently entirely unknown.
(4) While only a few old studies support the efficacy of using AEDs in symp-
tomatic patients with PS, no recent studies have been performed to further
support or refute the finding. It should be stressed that in the absence of
additional controlled studies, the literature stands to support the trial of anti-
convulsants in symptomatic patient with PS.

Open Research Questions

(1) What is the actual prevalence of PS in well-characterized healthy children and

adults?
(2) What would be the contribution of including REM sleep tracings in improving
the detection of the PSs. REM sleep is hardly ever included in routine EEG
studies.
(3) When PS are detected in healthy individuals, what are their characteristics in
term of frequency (more slow or more fast), prevalence in the record,
amplitude, and laterality?
(4) What are the relationships among frequency of occurrence in the record,
amplitudes, topography, and clinical symptoms?
220 19 The 6–7 and 14 Positive Spikes

(5) Given the significant advances in source localization methodology, it would be

of interest to ascertain whether all PS (fast or slow) emanate from the same
cerebral sources and whether or not the same sources are active both in
symptomatic and asymptomatic individuals with PS?
(6) What is the incidence and relevance to eating disorders?
(7) What is the significance of PS in children with different forms of learning
disabilities?
(8) What is the response to AEDs in the different clinical situations when PS is the
sole deviation detected on the EEG. These studies are particularly difficult and
must be performed in controlled, double-blind, and randomized manner con-
trolling for the EEG finding.
(9) What are the prognostic values of PS in the different psychiatric conditions
where they may be detected?

References

Andy OJ, Jurko MF (1972) Focal thalamic discharges with visceral disturbance and pain treated
by thalamotomy. Clin Electroencephalogr 3:215–223
Avery C (1968) A psychological study of patients with behavior problems and 6 and 14 per
second positive spikes in their electroencephalograms. Am J Psychiatry 24:171–173
Beun AM, van Ende Boas W, Dekker E (1998) Sharp transients in the sleep EEG of healthy
adults: a possible pitfall in the diagnostic assessment of seizure disorders. Electroencephalgr
Clin Neurophys 106:44–51
Boelhouwer C, Henry CE, Glueck BC (1968) Positive spiking—a double-blind control study on
its significance in behavior disorders both diagnostically and therapeutically. Am J Psychiatry
25:473–481
Bosaeus E, Selldén U (1979) Psychiatric assessment of healthy children with various EEG
patterns. Acta Psychiatr Scand 59:180–210
Boutros NN, Fristad M, Abdollohian A (1998) The 14 and 6 positive spikes and attention deficit
hyperactivity disorder. Biol Psychiatry 44(4):298–301
Bryant N, Friedlander WJ (1965) ‘‘14 ? 6’’ in boys with specific reading disability.
Electroencephalogr Clin Neurophysiol 19:322 (Abstract)
Castellotti V, Cermibori A, Pittaluga E (1966) Observations on the electroencephalographic
pattern of the 6–14 c/sec positive spikes. Sist Nerv 18:82–99
Cervone R, Blum A (2007) Normal variant EEG patterns. In: Blum AS, Rutkove SB (eds) The
clinical neurophysiology primer. Humana Press, Totowa, pp 83–100
Chaloner J, Pampiglione G (1983) ‘Posterior temporal fast’ EEG activity in childhood. Rev
Alectroencephalogr Neurophysiol Clin 13:53–60
Crowley WJ, Liske E (1967) Fourteen and six per second positive spiking—an EEG finding in
some air crew personnel. Aerospace Med 38:851–855
DeLong GR, Rosenberger PB, Hildreth S, Silver I (1987) The 14&6-associated clinical complex:
a rejected hypothesis revisited. J Child Neurol 2:117–127
DeMyer MK, DeMyer W, Norton JA (1977) Brain damage index for young children. Clin
Electroencephalogr 8:35
Domenici R, Meossi C, Stefani G, Castelli S (1991) A diagnostic controversy: the significance of
14-6/sec positive spikes in clinical electroencephalography. Pediatr Med Chir 13(4):417–22,
(Italian)
References 221

Eeg-Olofsson O (1971) The development of the electroencephalogram in normal children from

the age of 1 through 15: 14 and 6 Hz positive spike phenomenon. Neuropaediatrie 2:405–427
Engelhart RS, Knott JR (1964) An inquiry into MMPI correlates of the 14 and 6/sec positive
spike phenomena. Electroencephalogr Clin Neurophysiol 17:467 (Abstract I)
Friedlander WJ (1969) Relationship of the average wave frequency of sleep spindles and
temporally related 14 and 6/sec positive spikes. Electroencephalogr Clin Neurophysiol 26:118
(Abstract)
Gianturco DT, Wilson WP, Musella L (1972) Effect of psychiatric and autonomic symptoms on
the incidence of fourteen and six per second positive spikes among adolescents. Clin
Electroencephalogr 3:55–59
Gibbs EL, Gibbs FA (1951) Electroencephalographic evidence of thalamic and hypothalamic
epilepsy. Neurology 1:136–144
Gibbs EL, Gibbs FA (1973) Clinical significance of 14 and 6 per second positive spikes in the
electroencephalograms of patients over 29 years of age. Clin Electroencephalogr 4:140–144
Gibbs FA (1972) Violent behavior and the electroencephalogram. Clin Electroencephalogr 3:209
(in discussion)
Gibbs FA, Gibbs EL (1963) Fourteen and six per second positive spikes. Electroencephalogr Clin
Neurophysiol 15:553–558
Gibbs FA, Gibbs EL (1964) Atlas of electroencephalography, vol 3. Addison-Wesley, Reading
Gibbs FA, Gibbs EL (1971a) Anti-epileptic treatment of patients with 14 and 6 per second
positive spikes in the electroencephalograms. Clin Electroencephalogr 2:52–55
Gibbs FA, Gibbs EL (1971b) How much do sleep recordings contribute to the detection of seizure
activity? Clin Electroencephalogr 2:169–172
Gibbs FA, Gibbs EL. 1977. Electroencephalography in post-traumatic legal cases. Clin
Electroencephalogr J 8:156–164
Gibbs FA, Novick RG (1977) Electroencephalographic findings among adult patients in a private
psychiatric hospital. Clin Electroencephalogr 8:79–88
Gibbs FA, Gibbs EL (1987) Electroencephalographic study of head injury in children. Clin
Electroencephalogr 18(1):10–11
Greenberg IM, Pollack M (1966) Clinical correlates of 14 and 6/sec positive spiking in
schizophrenic patients. Electroencephalogr Clin Neurophys 20:197–200
Henry CE (1963) Positive spike discharges in the EEG and behavior abnormality. In: Glaser GH
(ed) EEG and behavior. Basic Books, New York, pp 315–344
Hoshika A, Matsuno T, Ogihara M, Miyajuma T, Aritaki S, Honda T, Oana Y, Miura S (1981)
Clinical EEG study of fourteen and six/sec positive spikes followed by slow waves.
Electroencephalogr Clin Neurophysiol 52:S110
Hotta T, Fujimoto I (1973) A study on abdominal epilepsy. Yanago Acta Med 17(3):231–239
Hughes JR (1960) The 14 and 7 per second positive spikes-a reappraisal following a frequency
count. Electroencephalogr Clin Neurophysiol 12:495–496
Hughes JR (1965) A review of the positive spike phenomenon. In: Wilson WP (ed) Applications
of electroencephalography in psychiatry. Duke University Press, Durham, pp 54–101
Hughes JR (1971) Electroencephalography and learning disabilities. In: Myklebust HR (ed)
Progress in learning disabilities, vol 2. Grune and Stratton, New York, pp 18–55
Hughes JR (1994) EEG in Clinical Practice, 2nd edn. Butterworth-Heinemann, Boston
Hughes JR, Cayaffa JJ (1977) The EEG in patients at different ages without organic cerebral
disease. Electroencephalogr Clin Neurophysiol 42:776–784
Hughes JR, Cayaffa JJ (1978) Positive spikes revisited—in the adult. Clin Electroencephalogr
9(2):52–59
Hughes JR, Park GE (1968) The EEG in dyslexia. In: Kellaway P, Petersen I (eds) Clinical
electroencephalography in children. Almqvist and Wiksell, Stockholm, pp 307–327
Hughes MR, Means ED, Stell B (1965) A controlled study on the behavior disorders associated
with the positive spike phenomenon. Electroencephalogr Clin Neurophysiol 18:349–353
Hughes JR, Wilson WP (1983) EEG and Evoked Potentials in Psychiatry and Behavioral
Neurology. Butterworths, Boston
222 19 The 6–7 and 14 Positive Spikes

Kellaway P, Crawley JW, Kagawa N (1959) A specific electroencephalographic correlate of

convulsive equivalent disorders in children. J Pediatr 55:582–592
Knott JR, Muehl S, Benton AL (1965) Electroencephalograms in children with reading
disabilities. Electroencephalogr Clin Neurophysiol 18:513 (Abstract)
Kurata S (1982) Clinical trials of carbamazepine for autonomic seizures with and without
generalized epileptic seizures. Brain Dev 4:81–86
Lombroso CT, Schwartz IH, Clark DM, Muench H, Barry J (1966) Ctenoids in healthy youths:
controlled study of 14-and 6—per-second positive spiking. Neurology 16:1152–1158
Loomis SD (1965) EEG abnormalities as a correlate of behavior in adolescent male delinquents.
Am J Psychiatry 121:1003–1006
Millen FJ, White B (1954): Fourteen and six per second positive spike activity in children.
Neurology 4: 541-549
Milstein V, Small JG (1971) Psychological correlates of‘ 14 ? 6 positive spikes, 6/sec spike-
wave and small sharp spike transients. Clin Electroencephalogr 2:206–212
Milstein V, Small JG, Golay S, Niggl D (1978) 1Conceptual thinking and the EEG in psychiatric
patients. Clin Electroencephalogr 9:96–100
Muehl S, Knott JR, Benton AL (1965) EEG abnormality and psychological test performance in
reading disability. Cortex 1:434–440
Okuma T, Kuba K, Matsushita T, Nakao T, Fujii S, Shimoda Y (1968) Study on 14 and 6 per
second positive spikes during nocturnal sleep. Electroencephalogr Clin Neurophysiol
25:140–149
Olson WH, Gibbs FA, Adams CL (1970) Electroencephalographic study of criminals. Clin
Electroencephalogr 1:92–100
Petersén I, Eeg—Olofsson O (1981) Longitudinal study of the EEG in normal children and
adolescents. Electroencephalogr Clin Neurophysiol 52:S61 (Abstract)
Petersén I, Selldén U (1981) On the need to collect EEG data from so-called normal individuals.
In: Stalberg E, Young RR (eds) Clinical neurophysiology. Butterworths, Woburn, pp 325–346
Pollack M, Jaffe R, Woerner MG, Klein DF (1969) Fourteen and six per second positive spikes in
psychiatric patients and their sibs. Electroencephalogr Clin Neurophysiol 27:669–670
Radhakrishnan K, Santoshkumar B, Venugopal A (1999) Prevalence of benign epileptiform
variants observed in an EEG laboratory from South India. Clin Neurophysiol 110:280–285
Rau JH, Struve FA, Green RS (1979) Electroencephalographic correlates of compulsive eating.
Clin Electroencephalogr 10:180–189
Reiher J, Carmant L (1991) Clinical correlates and electroencephalographic characteristics of two
additional patterns related to 14 and 6 per second positive spikes. Can J Neurol Sci
18:488–491
Riley T, Niedermeyer E (1978) Rage attacks and episodic violent behavior. Electroencephalo-
graphic findings and general considerations. Clin Electroencephalogr 9:131–139
Santoshkumar B, Chong JJ, Blume WT, McLachlan RS, Young GB, Diosy DC, Burneo JG,
Mirsattari M (2009) Prevalence of benign epileptiform variants. Clin Neurophysiol
120(5):856–861
Schwartz IH, Lombroso CT (1968) 14 and 6/second positive spiking (ctenoids) in the
electroencephalograms of primary school pupils. J Pediatr 72:678–682
Shimoda Y, Yoshino Y, Tanaka K (1969) Semiological relations of the frequency and
distribution of six and fourteen per second positive spikes. Electroencephalogr Clin
Neurophysiol 27:668 (Abstract)
Silverman D (1967) Phantom spike—waves and the fourteen and six per second positive spike
pattern: a consideration of their relationship. Electroencephalogr Clin Neurophysiol
23:207–213
Sheeby BN, Little SC, Stone JJ (1960) Abdominal epilepsy. J Pediatr 56:355–363
Small JG (1971) Photoconvulsive and photomyoclonic responses in psychiatric patient. Clin
Electroencephalogr 2:78–88
Small JG, Milstein V, DeMyer MK, Moore JE (1977) Electroencephalographic (EEG) and
clinical studies of early infantile autism. Clin Electroencephalogr 8:27–35
References 223

Small JG, Milstein V, Jay S (1978) Clinical EEG studies of short and long term stimulant drug
therapy of hyperkinetic children. Clin Electroencephalogr 9:186–194
Small JG, Milstein V, Medlock CE (1997) Clinical EEG findings in Mania. Clin Electroencep-
halogr 28(4):229–234
Small JG, Sharpley P, Small IF (1968) Positive spikes, spike-wave phantoms and psychomotor
variants. Arch Gen Psychiatry 18:232–238
Smith WL, Philippus MJ, Guard HL (1968) Psychometric study of children with learning
problems and 14–6 positive spike EEG patterns, treated with ethosuximide (Zarontin) and
placebo. Arch Dis Child 43:616–619
Struve FA, Honigfeld A (1970) Routine electroencephalograms of psychiatric patients awake and
asleep. Clin Electroencephalogr 1:80–283
Struve FA, Feigenbaum ZS, Farnum CD (1972) Prediction of 14 ? 6/sec positive spikes in EEGs
of psychiatric patients. Clin Electroencephalogr 3:60–64
Struve FA, Pike LE (1974) Routine admission electroencephalograms of adolescent and adult
psychiatric patients awake and asleep. Clin Electroencephalogr 5:67–72
Struve FA (1977) Ramsey PP. Concerning the 14 and 6 per second positive spike cases in post-
traumatic medical legal EEGs reported by Gibbs and Gibbs: a statistical commentary. Clin
Electroencephalogr 8:203–205
Sullivan L (2010) Waveform window #16; 14 and 6 hertz positive spikes. Am J Electroneu-
rodiagnostic Technol 50:67–72
Sullivan FW, Gentile K, Boelhouwer C (1963) Relationship of clinical symptomatology to
abnormal EEG findings: a family study. Am J Psychiatry 124:554–559
Takahashi T (1966) Fourteen and six per second positive spikes—further consideration of its
clinical significance. Folia Psychiatr Neurol Jpn 20:181–194
Tasher DC, Gibbs EL, Unrath DM (1970) Electroencephalograms of patients in a maximal
security hospital. Clin Electroencephalogr 1:101–110
Torres F, Ayers FW (1968) Evaluation of the electroencephalogram of dyslexic children.
Electroencephalogr Clin Neurophysiol 24:287 (Abstract)
Torres F, Faoro A, Loewenson R (1983) The electroencephalogram of elderly subjects revisited.
Electroencephalogr Clin Neurophysiol 56:391–398
Tsuzuki H (1967) The 14 and 6 per second positive spikes during paradoxical sleep. Folia
Psychiatr Neurol Jpn 21:181–188
Walker AE, Marshall C (1963) Depth recording of the 14 and 6 pattern. Electroencephalogr Clin
Neurophysiol 15:162–163
Walsa R (1968) Our experience with the 14 and 6/sec. spike activity. Electroencephalogr Clin
Neurophysiol 25:594
Wang PJ, Tseng CL, Lin LH, Lin MY, Shen YZ (1991) Analysis and clinical correlates of the 14
and 6 Hz positive electroencephalographic spikes in Chinese children. Acta Paed Sin
32:272–279
Wegner JT, Struve FA (1977) Incidence of the 14 and 6 per second positive spike pattern in an
adult clinical population: an empirical note. J Nerv Ment Dis 164:340–345
Weisz R, Klem G, Henry C (1978) 6 and 14 Hz positive spikes—a predictable abnormality.
Electroencephalogr Clin Neurophysiol 45:28P
Wiener JM, Delano JG, Klass DW (1966) An EEG study of delinquent and non-delinquent
adolescents. Arch Gen Psychiatry 15:144–150
Chapter 20
Some Final Thoughts for Clinical
Researchers

As is obvious from the preceding chapters, defining the exact clinical correlates
and the clinical value of identifying sEEG abnormalities in psychiatric population
has proven to be a difficult task. Most of the studies referenced in this book were
conducted in the course of clinical work in busy clinical EEG laboratories. Few
studies were in fact supported by national federal or even private grant funding. In
fact during my EEG training with John R. Hughes at the University of Illinois in
Chicago, I noticed that despite his prolific academic and scholarly productivity, he
was not investing any time in writing grant proposals. The laboratory was orga-
nized so that data can be collected prospectively or records could be retrospec-
tively minded. This is crucially important as without busy psychiatry-based EEG
laboratories, most of the questions raised in this book will have to await grant
funding an increasingly uphill battle in an ever tightening economic situation. In
fact there are enough clinical indications with significant literature support to start
such laboratories. These laboratories would in fact be significant sources of
income to the Departments where they are housed, a major teaching resource, an
invaluable clinical service as well as being a major source of data that would help
answer many questions surrounding the sEEG place in the clinical psychiatry
world (Pogarell et al. 2005).
In view of the significant heterogeneity within each psychiatric entity and the
significant overlap as well as comorbidity among psychiatric syndromes, signifi-
cant variability should be expected in any study looking at biological markers.
When such a marker is shown to differ between a target patient group and healthy-
matched control subjects and the finding is replicated by independent research
groups, then the finding is significant and cannot be said to be ‘‘non-specific.’’
Such finding, as is the case with all currently identified biological markers in
psychiatry, are never sensitive or specific enough for establishing a diagnosis in an
individual patient. Nonetheless, the finding most likely indicates that subgroup,
imbedded within the patient group, is the source of this signal. Research focused
on identifying such subgroup must proceed until the subgroup is identified and the
exact clinical and therapeutic correlates are defined.

N. N. Boutros, Standard EEG: A Research Roadmap for Neuropsychiatry, 225

DOI: 10.1007/978-3-319-04444-6_20,  Springer International Publishing Switzerland 2013
226 20 Some Final Thoughts for Clinical Researchers

Particular attention should be paid to the problem of comorbidity. Studies

should be explicit regarding the training and qualifications of the personnel
involved in the clinical characterization of study subjects. Specifically, such per-
sonnel should be trained to a standard reliability criterion. Such designs will allow
the examination of any correlations between biological deviations and symptom
clusters. Studies should also include both healthy and patient control groups.
Furthermore, and based on the questions asked, researchers may consider
including Axis-II patient control groups. If medications are not withdrawn, patient
control group should additionally be matched for pharmacotherapy.
One of the major questions repeatedly raised throughout the book is the pre-
dictive value of detecting paroxysmal EEG activity, in a nonepileptic individual
presenting with psychiatric symptoms, for a positive response to anti-epileptic
drug treatment. The study conducted by Porras-Kattz et al. (2010) came rather
close to an ideal design for such work despite the small sample size. The patient
group was largely homogenous, all having IEDs and they were randomly assigned
an AED or a placebo. Finally, improvement was measured utilizing standardized
scales. By comparison, Reeves et al. (2003) concluded that EEG does not predict
response to valproate in the treatment of aggression in patients with Axis-II dis-
orders while none of their patients exhibited frank epileptiform activities. It cannot
be over emphasized that different EEG abnormalities have different neurophysi-
ological underpinnings and only paroxysmal activity would logically be expected
to predict a favorable response to AED therapy.
As stated by Keck et al. (1992), the difficulties inherent in conducting con-
trolled studies of pharmacologic agents for the treatment of these syndromes may
rest in part with the uniqueness and poor generalizability of specific cases, the lack
of diagnostic homogeneity underlying the nonspecific but troublesome nature of
these behaviors, and ethical concerns regarding safety in placebo-controlled
designs. Evidently, this subject needs more research studies to be conducted to
have a better understanding of the efficacy of anticonvulsants on nonepileptic
patients.
The issue of the controversial waveforms is most difficult to tackle. What is
needed is well-collected prospective data in order to define the correlates of these
rather well-defined EEG patterns. Such studies should be conducted by EEGers
fully trained (to criteria) in detecting these patterns. All EEG tracings should be
examined by investigators blinded to the clinical or group membership of the
subjects and with at least two so qualified EEGers independently.
Furthermore, a complete evaluation and characterization of the clinical syn-
drome being examined (as much as the state of knowledge allows at the time)
should be done to allow later correlational analyses with different groupings of the
various symptoms.
References 227

References

Keck PE, McElroy SL, Friedman LM (1992) Valproate and carbamazepine in the treatment of
panic and posttraumatic stress disorders, withdrawal states, and behavioral dyscontrol
syndromes. J Clinical Pyschopharmacology 12(1):36S–41S
Pogarell O, Hegerl U, Boutros N (2005) Clinical neurophysiology service in psychiatry
departments. Psychiatric Services; Frontline Reports. 56(7):871
Porras-Kattz E, Harmony T, Ricardo-Garcell J et al (2010) Magnesium valproate in learning
disabled children with interictal paroxysmal EEG patterns: preliminary report. Neurosci Lett
492:99–104
Reeves RR, Struve FA, Patrick G (2003) EEG does not predict response to valproate treatment of
aggression in patients with borderline and antisocial personality disorders. Clin Elelctroen-
cephalogr 34(2):84–86