The Neuropsychiatry of Epilepsy

The Neuropsychiatry of Epilepsy
Michael Trimble and Bettina Schmitz have assembled a multinational team of experts to review
the most recent ndings which explore the interface between epilepsy and behaviour disorders.
They begin by looking at the classications available and examine how adequate they are for
dening the subtleties of behavioural changes in patients with neurological disorders. Coverage
is broad-ranging, from related cognitive problems and the biological underpinnings, to clinical
aspects, including pseudoseizures and treatment issues.
There has been a great deal of research in this area over recent years, but limited published
reviews. This timely book covers the practical implications of ongoing research, and oers both
a diagnostic and a management perspective. It will be essential reading for all professionals
engaged in the treatment of epileptic patients.
Michael Trimble is Professor of Behavioural Neurology at the Institute of Neurology in

London.
Bettina Schmitz is based in the Epilepsy Research Group of the Deparment of Neurology,
Humboldt University, Berlin.
The Neuropsychiatry of
Epilepsy
Edited by
Michael Trimble
and
Bettina Schmitz

Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, So Paulo
Cambridge University Press

The Edinburgh Building, Cambridge , United Kingdom
Published in the United States of America by Cambridge University Press, New York
www.cambridge.org
Information on this title: www.cambridge.org/9780521813747
Cambridge University Press 2002
This book is in copyright. Subject to statutory exception and to the provision of

relevant collective licensing agreements, no reproduction of any part may take place
without the written permission of Cambridge University Press.
First published in print format 2002
-
isbn-13 978-0-511-06480-7 eBook (NetLibrary)
-
isbn-10 0-511-06480-2 eBook (NetLibrary)
-
isbn-13 978-0-521-81374-7 hardback
-
isbn-10 0-521-81374-3 hardback
-
isbn-13 978-0-521-00516-6 paperback
-
isbn-10 0-521-00516-7 paperback
Cambridge University Press has no responsibility for the persistence or accuracy of

s for external or third-party internet websites referred to in this book, and does not
guarantee that any content on such websites is, or will remain, accurate or appropriate.
Every effort has been made in preparing this book to provide accurate and up-to-date
information which is in accord with accepted standards and practice at the time of
publication. Nevertheless, the authors, editors and publishers can make no warranties
that the information contained herein is totally free from error, not least because
clinical standards are constantly changing through research and regulation. The authors,
editors and publisher therefore disclaim all liability for direct or consequential
damages resulting from the use of material contained in this book. Readers are strongly
advised to pay careful attention to information provided by the manufacturer of any
drugs or equipment that they plan to use.
Contents
List of contributors page viii
Part I Background
1 Introduction 3
M.R. Trimble and B. Schmitz
2 Neuropsychiatric disorders in epilepsy epidemiology and classication 5

E.S. Krishnamoorthy
3 Limbic connectivity: anatomical substrates of behavioural disturbances

in epilepsy 18
J. Engel Jr, C. Wilson and F. Lopez-Rodriquez
Part II Clinical aspects
4 The psychiatry of idiopathic generalized epilepsy 41

Dieter Janz
5 Epilepsy and learning disorders 62

Cesare Maria Cornaggia and Giuseppe Gobbi
6 Subtle cognitive and behavioural eects of epilepsy 70

Frank M.C. Besag
7 Aggression and epilepsy 81

L. Tebartz van Elst
8 Epilepsy and suicide: a neuropsychiatric analysis 107

Dietrich Blumer
9 Postictal psychoses, revisited 117

Kousuke Kanemoto
v
vi Contents
Part III Cognitive aspects
10 Dementia and epilepsy 135

Stephen W. Brown
11 The risk of cognitive decline in patients with refractory temporal

lobe epilepsy 152
Hennric Jokeit and Alois Ebner
12 Behavioural and neuropsychological aspects of frontal lobe epilepsy 164

Christoph Helmstaedter
Part IV Nonepileptic attacks
13 Epilepsy, dissociation and nonepileptic seizures 189

Richard J. Brown
14 Psychobiology of psychogenic pseudoseizures 210

J. Chris Sackellares and D. Kalogjera-Sackellares
15 Epilepsy and panic disorder 226

Howard A. Ring and Nuri Gene-Cos
Part V Treatment complications
16 The eects of antiepileptic drugs on behaviour 241

Bettina Schmitz
17 Antiepileptic drug treatment and epileptic seizures eects on

cognitive function 256
Albert P. Aldenkamp
18 Psychiatric eects of surgery for temporal lobe epilepsy 266

Steffi Koch-Stoecker
19 Vagus nerve stimulation and mood 283

Christian E. Elger and Christian Hoppe
Part VI Treatment
20 On the use of psychotropic drugs in patients with seizure disorder 299

M.R. Trimble and Anke Hensiek
vii Contents
21 The role of psychotherapy in the treatment of epilepsies 313

Martin Schndienst
22 Choosing measures to assess quality of life (QOL) in epilepsy 323

Caroline E. Selai, Katja Elstner and M.R. Trimble
Index 343
Contributors
Albert P. Aldenkamp Richard J. Brown

Secretary General National Hospitals for Neurology and
Instituut voor Epilepsiebestrijding Neurosurgery
Meer en Bosch Institute of Neurology
Postbus 21 Queen Square
2100 AA Heemstede London WC1N 3BG, UK
The Netherlands
Cesare Maria Cornaggia
Frank M.C. Besag Clinical Psychiatry
Learning Disability Service University of Milano Bicocca
Bedfordshire & Luton Community NHS San Gerardo Hospital
Trust Via Donizetti 106
Twinwoods Health Resource Centre Mona, Italy
Milton Road
Clapham Alois Ebner
Bedfordshire MK41 6AT, UK Epilepsy Centre Bethel
Epilepsy Surgery Program
Dietrich Blumer Maraweg 21
University of Tennessee D33617 Bielefeld, Germany
College of Medicine
Department of Psychiatry Christian E. Elger
135 North Pauline Department of Epileptology
Memphis TN38105, USA University of Bonn
Sigmund-Freud-Strasse 25
Stephen W. Brown Bonn, FRG-53105
Cornwall Healthcare Trust
Unit 10 Bodmin Business Park Katja Elstner
Harleigh Road Raymond Way Neuropsychiatry Unit
Bodmin Institute of Neurology
Cornwall PL31 1AH, UK Queen Square
London WC1N 3BG, UK
viii
ix List of contributors
Jerome Engel Dieter Janz

Chief Division of Epilepsy and Clinical Burgunder Strasse 8
Neurophysiology 14129 Berlin, Germany
Department of Neurology
Reed Neurological Research Centre Hennric Jokeit
UCLA School of Medicine Schweizerisches Epilepsie-Zentrum
710 Westwood Plaza Neuropsychologie
Los Angeles Bleulerstrasse 60
CA 90095-1769, USA CH-8008 Zurich
Switzerland
Nuri Gene-Cos
St. Bartholomews & The Royal London Dalma Kalogjera-Sackellares
School of Medicine Department of Neuroscience
London E1 1BB, UK University of Florida
Florida, USA
Giuseppe Gobbi
Department of Neuropsychiatry Kousuke Kanemoto
Maggiore Hospital Department of Neuropsychiatry
Bologna, Italy Aichi Medical University
21 Yazako-karimata, Nagakute
Christoph Helmstaedter 480-1195, Aichi, Japan
Universittsklinik fur Epileptologie
Sigmund-Freud-Str. 25 Ste Koch-Stoecker
Universitt Bonn Klinik Mara I
53105 Bonn, Germany Maraweg 21
D33617 Bielefeld, Germany
Anke Hensiek
Department of Neurology E.S. Krishnamoorthy
Addenbrookes Hospital National Hospitals for Neurology and
P.O. Box 165 Neurosurgery
Hills Road Institute of Neurology
Cambridge CB2 2HE, UK Queen Square
London WC1N 3BG, UK
Christian Hoppe
Department of Epileptology Howard A. Ring
University of Bonn Academic Department of Psychological
Sigmund-Freud-Strasse 25 Medicine
FRG-53105 Bonn Third Floor, Alexandra Wing
Germany Royal London Hospital
Whitechapel
London E1 1BB, UK
x List of contributors
F. Lopez-Rodriquez Caroline E. Selai

Department of Psychiatry and 126 Scott Ellis Gardens
Biobehavioural Sciences London
UCLA School of Medicine NW8 9HG, UK
710 Westwood Plaza
Los Angeles Ludger Tebartz van Elst
CA 90095-1769 Department of Psychiatry and Psychotherapy
USA Albert-Ludwigs-University
Haupstr.5
J. Chris Sackellares 79104 Freiburg, Germany
College of Medicine Department of
Neurosciences Michael R. Trimble
PO Box 100244 National Hospitals for Neurology and
Gainesville Neurosurgery
Florida 32610-0244, USA Institute of Neurology
Queen Square
Bettina Schmitz London
Neurologische Klinik und Poliklinik WC1N 3BG, UK
Charit, Virchow Klinikum
Humboldt-Universitt C. Wilson
Augustenburger Platz 1 Department of Neurology
13353 Berlin, Germany UCLA School of Medicine
710 Westwood Plaza
Martin Schndienst Los Angeles
Epilepsie-Zentrum Bethel CA 90095-1769, USA
Maraweg 21
33617 Bielefeld, Germany
Part I
Background
1
Introduction
M.R. Trimble1 and B. Schmitz2
1
Institute of Neurology, London, UK
2
Humboldt University, Berlin, Germany
In 1997, Dr Pete Engel, President of the International League Against Epilepsy,

invited us to establish a commission on psychobiology, and this book represents one
of the achievements of the commissions work. The task of the commission was to
explore the interface between epilepsy and behaviour disorders, from a biological
and social point of view. To these ends a number of subcommissions were estab-
lished, and their task was to explore the existing knowledge base of the discipline, to
suggest research strategies for interventions and to educate both patients and carers
about aspects of epilepsy which a number of people consider to have been neglected.
The present book is divided into several parts, covering a spectrum of clinical
topics which have been of concern to the commission. Some old chestnuts, for
example the interictal psychoses of epilepsy, have not been allocated specic chap-
ters, and a number of other areas, particularly relating to learning disability, cog-
nitive decline, dissociative attacks and vagus nerve stimulation have been included.
It is hoped that by expanding upon the literature on some of these less-well-dis-
cussed aspects of psychobiology in epilepsy further interest will be stimulated,
leading to both intellectual discussion and research endeavours.
We start our text with an introduction to the classication of psychiatric disor-
ders in epilepsy. The point is made that existing classications used in psychiatry
such as the DSMIV are quite inadequate when it comes to dealing with the sub-
tleties of the behaviour changes of patients with neurological disorders.
We then discuss the biological underpinnings of some behaviour problems, in
terms of exploring the limbic system and related structures that are aected by the
process of epilepsy and which are also related to behavioural disorders.
The part on clinical aspects explores in particular the problems of learning dis-
ability, and introduces the important area of state-dependent learning disabilities:
patients with cognitive decits that can be profoundly reversed by appropriate
treatment strategies. Other important areas covered include the ever-controversial
topic of aggression, the importance of suicide, and the group of psychoses that
occur postictally.
3
4 M.R. Trimble and B. Schmitz
The next part looks further at cognitive problems in patients with epilepsy,
examining whether the concept of dementia is relevant, discussing the question as
to whether or not there is cognitive decline in patients with various types of epi-
lepsy over time, and also examining the issue of frontal lobe epilepsies. The latter
have been well dened from the seizure point of view in recent years, but the behav-
ioural and cognitive associations have yet to be claried.
We make no apology for including a section on nonepileptic seizures. The fact
that many patients who are diagnosed as having epilepsy do not have epilepsy, but
have some form of nonepileptic attack disorder (pseudoseizure) is now well recog-
nized. This problem has been around for centuries, and such eminent neurologists
as Charcot have spent some considerable time attempting to dierentiate between
nonepileptic and epileptic seizures. However, this often still proves dicult. We still
have inadequate information as to the mechanisms for the development of nonepi-
leptic attack disorder, and these, and the possible biological associations, are taken
up in this section.
The nal sections deal with treatments and their side eects. Of importance in
this section are the references to surgery, not only temporal lobe resection, but also
more recent advances such as vagus nerve stimulation. The benecial and negative
psychiatric consequences of these treatments are at the present time being actively
explored, and some early work is presented here. However, in the context of
psychobiology, our treatment strategies must go beyond medication and surgical
interventions, and we include a discussion of psychodynamic principles in rela-
tionship to the management of epilepsy, and also a chapter on quality of life.
We, the editors, hope that the book will enliven the debate about the links
between epilepsy and behaviour, an area which is often not well discussed, partly
because of some worry that any association between psychiatry and epilepsy may
stigmatize patients with epilepsy even more than they already are. However, the
problems that we have identied are a reality not only in the clinic, but also for
patients and carers themselves. It is dicult to dene treatment and management
strategies if problems are ignored, and so our intention is to enliven this area with
these up-to-date reviews on behavioural and cognitive problems in epilepsy, and
their social and biological underpinnings.
2
Neuropsychiatric disorders in epilepsy

epidemiology and classification
E.S. Krishnamoorthy
Introduction
The association between epilepsy and psychiatric disorders has a long and che-
quered history. For centuries seizures were considered to be a form of demonic pos-
session. Beginning late in the nineteenth century, considerable attention has been
directed towards cataloguing, describing and understanding disorders at the inter-
face between epilepsy and psychiatry, particularly by European neurologists and
psychiatrists. However, it is only in the past few decades that any attention has been
paid to the epidemiology of these disorders. Similarly, aside from some early
attempts by European physicians, there have been no eorts to develop an opera-
tional classication of psychiatric disorders in epilepsy (Schmitz and Trimble, 1992
for a review).
The paucity of epidemiological research at this interface, and the failure to
develop an operational international system of classication, is in stark contrast
with developments both in epilepsy per se, and in mental health research. The epi-
demiology of epilepsy has been well studied in many countries and considerable
data (both descriptive and analytical) are now available. Indeed, epilepsy has been
subject to the gamut of epidemiological research including cross-sectional, case-
control and cohort studies (Hauser, 1998). Similarly, operational international
systems of classifying epilepsy and its syndromes have been developed both by the
Commission on Classication and Terminology of the International League
Against Epilepsy (1989), and the World Health Organization (1967), and are used
by epileptologists around the world.
Impressive developments have also taken place in the eld of mental health epi-
demiology. Eorts by the World Health Organizations Division of Mental Health,
and other pioneering organizations around the world, have led to a signicant
understanding of the epidemiology of psychiatric disorders. This has led to the
development of universally accepted classicatory systems in psychiatry, such as the
Diagnostic and Statistical Manual now in its fourth edition (DSMIV; American
5
6 E.S. Krishnamoorthy
Psychiatric Association, 1994), and the mental disorders component of the

International Classication of Diseases, now in its tenth edition (ICD10; World
Health Organization, 1992).
The commonly held conviction among epileptologists and neuropsychiatrists is
that psychiatric comorbidity is not only common in epilepsy, but that distinct and
unique forms of psychopathology are prevalent (Krishnamoorthy 2000, 2001). In
the past three decades attention has been directed towards discrete forms of
psychopathology in epilepsy such as the temporal lobe personality, interictal and
postictal psychosis, and interictal dysphoric disorder (Bear and Fedio, 1977;
Blumer 1995, 2000). This combined with the observation of similarities in behavi-
our during seizures and in psychopathological states has strengthened the notion
of an anity between epilepsy and psychiatric disorder. Yet, the evidence that
psychiatric disorders are overrepresented in epilepsy is far from convincing, with
conicting results in dierent studies.
In this chapter the epidemiology of psychiatric comorbidity in adult, non-learn-
ing-disabled patients with epilepsy will be reviewed. There is a considerable litera-
ture on children and the learning disabled that is being addressed elsewhere in this
book (Chapters 5 and 6). Some ideas on how psychiatric disorders in epilepsy may
be classied, and the work of the subcommission on classication of the
International League Against Epilepsy the Commission on Epilepsy and
Psychobiology in this regard, will also be discussed.
Epidemiology
A majority of studies in this area has been hospital- and institution-based. While
the contribution of these studies to the current understanding of psychopathology
in epilepsy has been invaluable, the strong selection bias in these studies does make
the extrapolation of their ndings to the majority of patients with epilepsy, who
live in the community, dicult.
There have been some population-based studies of psychiatric comorbidity that
are summarized here. Most studies have been cross-sectional and some have com-
pared cases with controls. By and large, save one or two exceptions, these studies
have generated crude estimates of prevalence, rather than more specic epidemi-
ological indices.
Population-based studies of psychiatric comorbidity in epilepsy

One of the earliest investigations to be carried out was that of Pond and Bidwell
(1959/60), who surveyed patients from 14 doctors surgeries in the south-east of
England. They found that 29% of 245 patients had psychological disorders of su-
cient severity to seek treatment, i.e. conspicuous morbidity. The main criticism lev-
elled against this study is its use of a social worker rather than a trained mental
7 Classification
health professional, and a lack of standardized techniques to assess patients with

epilepsy for psychiatric comorbidity. The strength of this study, however, lies in its
recognizing, four decades ago, the importance of an epidemiological approach.
Gudmundsson (1966) personally surveyed 987 patients with epilepsy living in
Iceland and reported that 512 (52%) had personality changes of various kinds. Of
these 271 (27.5%) were described as ixoid, 73 (7.4%) as ixothymic and 168
(17.0%) as neurotic. More men were ixoid and more women neurotic. While
Gudmundsson, unlike Pond, personally examined every subject, the clinical termi-
nology and classication used have few parallels today, and no attempts were made
to reduce bias. However, the high proportion of subjects with behavioural changes
in this community-based population is striking and worthy of note.
Edeh and Toone (1987) conducted a survey in doctors surgeries in south
London. They interviewed 88 adult patients with epilepsy drawn from doctors sur-
geries in the area, using the Clinical Interview Schedule, and reported that 48%
emerged as psychiatric cases. They also found that while patients with temporal
lobe epilepsy (TLE) and focal non-TLE did not dier in terms of psychiatric mor-
bidity, both groups were signicantly more impaired than patients with primary
generalized epilepsy. The techniques of ascertainment used in this study are com-
mendable. Subjects with epilepsy underwent both CT scans and EEG tests, in con-
rmation of their diagnosis. The study also used a validated instrument for
common mental disorder, the CIS-R (Lewis et al., 1992). In criticism, however, it
must be said that the study failed to examine matched population-based controls,
psychopathology specic to epilepsy was not examined, and while cases with
psychosis were identied, no validated diagnostic instrument for psychosis was
administered, the CIS-R being a validated instrument for common mental disor-
der alone.
Cockerell et al. (1996) conducted a nation-wide survey in the UK of acute
psychological disorders (APD) in patients with epilepsy using the British
Neurological Surveillance Unit. Sixty-four incident cases were ascertained over a
period of one year. Thirty-one were considered to have APD due to ictal or post-
ictal activity and 33 were interictal. In 30% of cases the APD was reported by the
referring physician to be secondary to the prescription of an antiepileptic drug
(AED). The drugs most commonly implicated were carbamazepine, lamotrigine
and vigabatrin. The broad psychiatric categories diagnosed included delirium
(25%), schizophreniform (31%), aective (30%), delusional (5%) and other dis-
orders (9%). The ndings of this study are of interest as it gives us crude incidence
gures of acute psychiatric disorder in epilepsy and highlights the importance of
antiepileptic drugs in precipitating comorbid psychiatric illness in epilepsy.
However, as the study used a reporting system, rather than a population-based
cohort, the results cannot be used to generate population-based incidence gures,
or be generalized.
Jalava and Sillanpaa (1996) examined a prospective population-based cohort

(mean follow-up of 35 years) of patients with epilepsy since childhood, for co-
morbid somatic, psychosomatic and psychiatric disorders. In comparison with
random controls, patients with epilepsy had a fourfold risk of psychiatric disorders
or combinations of somatic, psychosomatic and/or psychiatric disorders. Thus
patients with childhood-onset epilepsy demonstrated a higher risk for psychiatric
or psychosomatic disorders and this appeared to be related to epilepsy and not AED
administration.
This is perhaps the only cohort study of psychiatric comorbidity in epilepsy and
the ndings have great relevance. The results clearly indicate that subjects with epi-
lepsy are at higher risk of developing comorbid psychiatric illness, when compared
with population-based controls, and indicate the need for greater provision for
psychiatric treatment in primary care settings for epilepsy. However, as individual
cases were not ascertained in any systematic way, it is possible that the ndings do
not represent the true extent of comorbidity, with subtle nevertheless disabling
forms of psychopathology, or those not requiring medical attention or admission,
being missed. This is of relevance, as subtle forms of psychopathology that often do
not meet conventional diagnostic criteria may be overrepresented in epilepsy.
Bredkjaer et al. (1998) conducted a record-linkage study in Denmark between a
sample of people with epilepsy from the National Patient Register and from the
Danish Psychiatric Register. They found that the incidence of nonorganic nonaec-
tive psychoses including personality disorders that were broadly within the schizo-
phrenia spectrum was signicantly increased for both men and women with
epilepsy, even after excluding all people diagnosed as suering from a learning dis-
ability or substance misuse. The standardized incidence ratio was signicantly
increased for the entire schizophrenia spectrum (P108), nonaective psychosis
(P108) and schizophrenia alone (P0.0001).
In the absence of long-term prospective data, this study based on national regis-
ters provides evidence that disorders in the schizophrenia spectrum are clearly
overrepresented in epilepsy. The study enabled the calculation of more sophisti-
cated epidemiological indices, such as standardized incidence ratio, that have not
been estimated in previous studies. However, the methodological limitations of
reliance upon a case-register, i.e. the lack of standardization of ascertainment
methods, both for epilepsy and psychoses, and the exclusion of more subtle cases,
or those not requiring admission, do apply here.
Stefansson et al. (1998) conducted a case-control study comparing the preva-
lence of nonorganic psychiatric disorders among patients with epilepsy and con-
trols with other somatic diseases, both groups being of normal intelligence. The
two groups were drawn from a disability register of the State Social Security
Institute in Iceland. In this way, 241 index cases meeting inclusion criteria were
9 Classification
identied, and the ratio between subject (epilepsy) and control (somatic illness)
cases was 1:2. Psychiatric diagnosis was present among 35% of cases as compared
with 30% of controls, the dierence not being statistically signicant. Psychiatric
disorders were, however, signicantly more common in men with epilepsy, but not
in women, the dierence being due to a signicantly higher rate of psychosis, par-
ticularly schizophrenia or paranoid states, among men.
Some large hospital-based studies

Currie et al. (1971) surveyed 666 patients recorded to have features of temporal
lobe epilepsy in the hospital diagnostic index and the records of the neurology,
neurosurgery and EEG departments. They found 375 (56%) to be normal, 127
(19%) to be anxious, 71 (11%) to be depressed, 47 (7%) to be aggressive, 41 (6%)
to be obsessive and 38 (6%) to have a severe disturbance of aect.
Smith et al. (1986) studied 622 patients in the USA in a nation-wide cooperative
study spanning 10 Veterans Administration Medical centres, using a battery of
neuropsychological testing procedures. The majority of patients was not on anti-
convulsant drugs at the time of initial testing, and the few who were had subthera-
peutic levels on measurement. They found that patients with epilepsy scored
signicantly and consistently below the level of the 74 control subjects on all but
three behavioural measures. Dierences reaching statistical signicance were
found on tests of motor function (nger tapping, pegboard, colour naming),
cognitive-attention (digit symbol, discrimination reaction time, word uency) and
subtests of the Prole of Mood States (tension, depression, vigour and confusion).
These they felt provided a prole of behavioural characteristics of unmedicated
patients with epilepsy.
Gureje (1991) evaluated 204 unselected patients with epilepsy attending a neuro-
logical clinic using the Clinical Interview Schedule (Goldberg, 1972); 37% emerged
as psychiatric cases. Of these 53% had a neurosis, 29% had a psychosis and 7% were
diagnosed to have a personality disorder.
Mendez et al. (1993) conducted a retrospective investigation of neurology clinic
attenders. They found that interictal schizophrenic disorders occurred in 149
(9.25%) of 1611 patients with epilepsy as compared to only 23 (1.06%) of 2167
patients with migraine. They went on in the latter part of the study to compare 62
epilepsy and schizophrenia patients with 62 patients who had epilepsy alone on 6
seizure variables, and 62 patients with schizophrenia alone on 10 psychosis variables.
The epilepsy and schizophrenia group was found to have a later age of onset of
epilepsy with more complex partial seizures, more patients with auras and fewer
patients with generalized epilepsy. Except for increased suicidal behaviour, patients
with epilepsy did not dier from controls on psychosis variables; however, psy-
chotic symptoms often emerged with increased seizure activity. They felt that the
data supported a distinct association of schizophrenic disorders with epilepsy, par-

ticularly with seizures emanating from the temporal limbic system.
Manchanda et al. (1996) studied 300 consecutive patients refractive to treat-
ment, admitted for evaluation of their candidature for epilepsy surgery over a
6-year period. Of these, 231 had a temporal lobe focus, 43 had a nontemporal lobe
focus and 26 had generalized and multifocal seizure onset; 142 (47.3%) emerged as
psychiatric cases based on DSMIIIR criteria. A principal Axis I diagnosis was
made in 88 (29.3%). Anxiety disorders (10.7%) and schizophrenia (4.3%) were the
most common Axis I diagnoses. Dependent and avoidant personality traits were
frequent (18%) although patients rarely fullled criteria for a personality disorder.
Are psychiatric disorders commoner in epilepsy?
This question needs to be addressed from a public health perspective. Were psychi-
atric disorders to be commoner in patients with epilepsy, specic mental health
resources would need to be created in the community for this patient group. On
the other hand were there no excess in psychiatric comorbidity, when patients with
epilepsy were compared with other illness groups, matched for age, sex and disabil-
ity, and normal controls, such resources would not be required. Here we shall
examine the evidence, to see if depression and psychosis are commoner in epilepsy.
A majority of studies has shown depression to be common in epilepsy. Many of
these have employed the Minnesota Multiphasic Personality Inventory (MMPI).
Whitman et al. (1984) used a MMPI sequential diagnostic system (Goldberg, 1972)
to reanalyse 87 published proles of patients with epilepsy, other neurological dis-
orders and chronic physical illnesses, encompassing a total of 2786 patients. This
included 10 studies of epilepsy encompassing a total of 809 subjects. They found
that patients with epilepsy were at higher risk of psychopathology than normal
controls. However, no dierence was found between people with epilepsy and those
with other chronic disorders, or between people with TLE and those with general-
ized epilepsy. A similar investigation was also reported by Dodrill and Batzel
(1986), who found that patients with epilepsy demonstrated more psychopathol-
ogy than normal controls and patients with other neurological disorders, but that
there were no dierences in rates of psychopathology between TLE and other forms
of epilepsy.
Investigations using other instruments such as the Present State Examination
have also shown a higher prevalence of depression in epilepsy (Standage and
Fenton, 1975). However, other investigations have failed to demonstrate an
increased prevalence of depression in epilepsy. For a review of these studies and a
discussion of the phenomenology of depression in epilepsy see Lambert and
Robertson (1999).
11 Classification
Of all the dierent psychiatric disorders in epilepsy, it is psychosis for which

there is considerable evidence of overrepresentation. The prevalence of psychosis
in epilepsy is reported to be in the order of 4% (see Manchanda et al., 1996 for
example), sometimes rising as high as 10%. Psychotic disorders are 10 times more
common in epilepsy than in the general population, and this is borne out in well-
designed population-based cohort and case-control studies, reviewed here. For a
detailed review of studies of psychosis and of the nature and phenomenology of the
epileptic psychosis, see Trimble (1991).
Another reason for the diculty in answering this question is the selection bias
in most studies mentioned above. In this chapter we have deliberately concentrated
on population-based studies. Reviewing these (Table 2.1) it is apparent that there is
a considerable degree of psychiatric comorbidity in epilepsy, even in well-designed
cohort studies (see Jalava and Sillanpaa, 1996 for example). However, while the evi-
dence for a higher prevalence of psychotic disorders stands out, both in cohort
studies (Bredjkaer et al., 1998) and nested case-control studies (Stefannson et al.,
1998), the evidence for other psychiatric disorders, while present (see Jalava and
Sillanpaa, 1996 for example), is contradictory and not as compelling.
The classification of psychiatric disorders in epilepsy
The classication of psychiatric disorders in epilepsy has always been controversial.

There are two main schools of thought. The rst is that the existing systems of clas-
sication in psychiatry, the current being the ICD10 and DSMIV, in its fourth
edition, have made adequate provision for organic conditions like epilepsy, and
further subsystems of classication would only add to their complexity. The second,
most often voiced by neuropsychiatrists with an interest in epilepsy, is that the exist-
ing systems of classication are hopelessly inadequate as far as neurological disor-
ders in general and epilepsy specically are concerned (Krishnamoorthy, in press).
One recurrent theme in reviewing the literature about psychiatric disorders in
epilepsy is that the failure to identify an excess of psychopathology is due more to
the instruments used (generic to mental disorder and not specic to mental disor-
ders in epilepsy), rather than a true nding. It was this that led Bear and Fedio
(1977) to develop their own instrument, and conduct a study of psychopathology
in patients with temporal lobe epilepsy. The traits that they looked for were those
identied by Gastaut, and later Geschwind, who described the constellation of per-
sonality traits that characterize patients with temporal lobe epilepsy, including
hypergraphia, hyposexuality, religiosity and emotional viscosity.
The study by Bear and Fedio (1977) showed that while the MMPI failed to iden-
tify dierences between patients with TLE and other patient groups, the dierences
were all too apparent when the responses to the instrument they developed were
Table 2.1. Important epidemiological studies of neuropsychiatric comorbidity in epilepsy
Investigators
Year (country) Results Comments
1960 Pond and Bidwell 29% of 245 patients had Study in 14 doctors surgeries
(UK) signicant morbidity Conducted by psychiatric social
worker
Instruments not standardized
1966 Gudmundsson 52% of 987 patients had personality Personal survey by expert
(Iceland) changes Instruments and diagnosis not
standardized
1987 Edeh and Toone 48% of 88 patients emerged as Primary care-based
(UK) cases Sophisticated case ascertainment
Standard instruments but not
epilepsy-specic
1996 Cockerell et al. 64 incident cases of acute Nation-wide survey
(UK) psychological disorder on AED Relied on reporting system
institution Crude data on incidence cannot be
generalized
1996 Jalava and Sillanpaa Patients with epilepsy fourfold Prospective cohort study with 35-
(Finland) risk of somatic, psychosomatic and/ year follow-up (only cohort study to
or psychiatric disorder in date)
combination compared with Results clearly indicate that subjects
population-based controls with epilepsy are at higher risk of
Results related to epilepsy and not developing comorbid psychiatric
antiepileptic drug administration illness
1998 Bredjkaer et al. Incidence of schizophrenia spectrum Record linkage study between a
(Denmark) psychoses signicantly increased for sample of people with epilepsy from
both men and women with epilepsy the National Patient Register and the
Standardized incidence ratio for the Danish Psychiatric Register
entire schizophrenia spectrum Enabled the calculation of
(P108), nonaective psychosis sophisticated epidemiological indices
(P108) and schizophrenia alone not estimated in previous studies
(P0.0001)
1998 Stefansson et al. Psychiatric diagnosis in 35% of 241 Patients with epilepsy, and controls
(Iceland) epilepsy cases as compared with 30% with other somatic diseases, both
of controls, the dierence not being groups being of normal intelligence
statistically signicant drawn from a disability register of
Signicantly higher rate of the State Social Security Institute
schizophrenia among men
13 Classification
analysed. A number of studies have been conducted since then with conicting
results (for reviews Shetty and Trimble, 1997; Trimble, 1991).
More recently, Blumer (1995, 2000) has drawn attention to the mood disorder
that is seen in patients with refractory epilepsy, particularly TLE that may occur in
conjunction with these personality traits. This interictal dysphoric disorder (IDD)
of epilepsy is described as being polymorphic, and characterized by a constellation
of eight symptoms, typically of short duration and occurring in dierent permu-
tations and combinations at dierent times.
Blumer argues that the personality traits seen in these patients, such as a serious
demeanour, deliberate speech, an ethical and spiritual orientation, especially when
subtle, can be positive attributes. However, when the IDD coexists, there may be a
paroxysmal venting of angry aects not normally characteristic of the person, fol-
lowed by a sense of remorse. This he contends can be the source of signicant dis-
ability.
A modied version of the BearFedio scale, the Neurobehavioural Inventory
(NBI; Blumer, 1995) is reportedly sensitive to this disorder, helping to identify
some of its typical features. However, as the psychometric properties of this instru-
ment have not been tested, and population-based studies have not been carried out,
the validity of this diagnosis has not been established.
Other dierences have also been reported to characterize psychopathology in
epilepsy and to dierentiate it from psychopathology in general. The interictal
psychosis of epilepsy is reported in many studies to be characterized by the preser-
vation of aect, religiosity and paranoid ideation, rather than the undierentiated,
or hebephrenic picture seen in schizophrenia. Slater rst observed this in his land-
mark study at the National Hospital, Queen Square (Slater and Beard, 1963).
Subsequently, many others have commented on these ndings (Trimble, 1991) and
indeed, they have to some extent been borne out in the population-based studies
reviewed here (Stefannson et al., 1998).
Further, while psychiatric classications such as the ICD and DSM tend to
subsume epilepsy under the organic umbrella, thus limiting the ability of such clas-
sication to be versatile, and clinically relevant, current classicatory systems in epi-
lepsy pay little or no attention to psychopathology. Besides, there is no place in
existing systems of classication for the psychiatric disorders that are reported to be
specic to epilepsy. A distinct classicatory system enables clearer phenomenologi-
cal descriptions of these disorders. It also lends itself more easily to empirical testing.
A convincing argument can therefore be made for a distinct classicatory system,
and this is currently the subject of a discussion document of the ILAE Commission
on Epilepsy and Psychobiology (Krishnamoorthy et al., in preparation).
As reviewed here, the neuropsychiatric disorders specic to epilepsy comprise
the gamut of neuropsychiatry. Included are the so-called organic mental disorders,
such as postictal confusional states and complex partial status with psychopatho-
logical manifestations; personality changes (the GastautGeschwind syndrome of
temporal lobe epilepsy, and the labile personality of juvenile myoclonic epilepsy);
a spectrum of psychoses with varying intensity, features and manifestations
depending on the temporal relationship with seizure(s) (Trimble, 1991); and a
spectrum of neuroses with predominantly aective features (Blumer 1995, 2000).
These disorders are also inexorably linked to their relationship with the seizure(s)
per se (preictal, postictal, interictal and perhaps periictal); their relationship to the
EEG (for example forced normalization of Landolt and alternative psychosis of
Tellenbach; Krishnamoorthy and Trimble, 1999); and their relationship to antiepi-
leptic drug (AED) therapy (the AED-induced neuropsychiatric disorders; Trimble,
1998). The inclusion of nonepileptic attack disorder (NEAD) in such a classica-
tion is rather more controversial, as there is a growing understanding that NEAD
is, in a number of people, the manifestation of a much wider spectrum of psycho-
pathology than that specic to epilepsy (Brown and Trimble, 2000).
Any classicatory system, to be viable, will need to take all these factors into con-
sideration. Further, it is important to acknowledge that patients with epilepsy could
like anyone, especially patients with chronic medical conditions, have comorbid
psychiatric disorders that match existing descriptions in the ICD10 and DSMIV.
It would serve little purpose to try and reclassify these disorders when associated
with epilepsy. The judgement about whether to record the illness in a given patient
as a comorbid disorder or as an epilepsy-specic disorder would be best left to the
clinician dealing with that individual case. It also goes without saying that such a
classicatory system should link closely with the ILAE Classication of Epilepsies
and Epileptic Syndromes.
While there is little doubt that classications grounded in aetiology and
pathophysiology are an ideal that we must aspire for in the long term, our under-
standing of causation and its mechanisms in psychiatry, even the neuropsychiatry
of epilepsy, is fairly rudimentary, and much ground needs to be covered before we
can move with any certainty towards an aetiological model. Aetiologically based
systems of classication also require specialized knowledge and access to suppor-
tive investigative techniques. Both of these are unavailable in a number of settings,
particularly in the developing world. Thus, at least at present, classicatory systems
that aim to be culture-free and acceptable across the board, would do well to adopt
a descriptive approach, based on a good history and clear clinical descriptions. Such
descriptive approaches mirror good clinical practice around the world, and make
few demands in terms of specialist expertise or investigation.
An ideal classication would be one that is simple, user-friendly, grounded in
good history taking and concise clinical descriptions, links with existing systems of
classication both in epilepsy and psychiatry, and one which is applicable across
15 Classification
the board. Such a classicatory system will be helped by its application in good
prospective population-based research leading eventually to its being operational
and valid in the years to come.
Undoubtedly, as our understanding of aetiology and pathophysiology in epi-
lepsy and neuropsychiatry improves, one would expect a radical alteration in such
descriptive classicatory systems, as we are beginning to see in other neurological
and indeed neuropsychiatric disorders. Until then, however, it would perhaps be
wise for us to continue in the classical descriptive tradition of good clinical medi-
cine.
Conclusions
Psychiatric disorders are common in epilepsy, and encompass the spectrum of

conditions, from those that are a direct consequence of epileptogenic activity, to
others that are simply comorbid.
There is considerable evidence from epidemiological research to suggest that the
psychoses are greatly overrepresented in epilepsy; the evidence for an overrepre-
sentation of other psychiatric disorders is less compelling.
While hospital-based data indicate the presence of epilepsy-specic psychopa-
thology, this has never been examined in the epidemiological setting. Further,
instruments such as the NBI that are supposedly sensitive to epilepsy-specic
psychopathology have not been validated in this setting.
Existing classicatory systems, both in epilepsy and in mental health, are inade-
quate, with regard to neuropsychiatric disorders in epilepsy. Eorts to build con-
sensus in this regard, and to develop a user-friendly classicatory system are
currently being made by the ILAE Commission on Epilepsy and Psychobiology.
To be relevant and applicable, a classicatory system must not be demanding in
terms of specialist knowledge or investigation. A descriptive approach that
mirrors good clinical practice is therefore recommended. Such a classicatory
system must also link in with existing systems in epilepsy and mental health.
Systematic population-based research using reliable methods of ascertainment,
and controls matched for age, sex, disability and ethnicity, based on the ILAE
classication of neuropsychiatric disorders in epilepsy, with long-term follow-up
of such cohorts, must be conducted in the future.
R E F E R E N C ES
American Psychiatric Association (1994). Diagnostic and Statistical Manual of Mental Disorders
(Fourth Revision) DSMIV. Washington, DC: American Psychiatric Association.
Bear, D.M. and Fedio, P. (1977). Quantitative analysis of interictal behaviour in temporal lobe
epilepsy. Arch Neurol, 34, 45467.
Blumer, D. (1995). Personality disorders in epilepsy. In Neuropsychiatry of Personality Disorders,
ed. J.J. Ratey, pp. 23063. Boston: Blackwell Science.
Blumer, D. (2000). Dysphoric disorders and paroxysmal eects: recognition and treatment of
epilepsy-related psychiatric disorders. Harv Rev Psychiatry, 8, 817.
Bredkjaer, S.R., Mortensen, P.B. and Parnas, J. (1998). Epilepsy and non-organic non-aective
psychosis: National Epidemiologic Study. Br J Psychiatry, 172, 2358.
Brown, R.J. and Trimble, M.R. (2000). Dissociative psychopathology, non-epileptic seizures and
neurology (Editorial). J Neurol Neurosurg Psychiatry, 69, 28591.
Cockerell, O.C., Moriarty, J., Trimble, M.R., Sander, J.W.A.S. and Shorvon, S.D. (1996). Acute
psychological disorders in patients with epilepsy: a nation-wide study. Epilepsy Res, 25, 11931.
Commission on Classication and Terminology of the ILAE (1989). Proposal for revised classi-
cation of epilepsies and epileptic syndromes. Epilepsia, 30, 38999.
Currie, S., Heatheld, K.W.G., Henson, R.A. and Scott, D.F. (1971). Clinical course and progno-
sis of temporal lobe epilepsy a survey of 666 patients. Brain, 94, 17390.
Dodrill, C.B. and Batzell, L.W. (1986). Interictal behavioural features of patients with epilepsy.
Epilepsia, 27 (Suppl. 2), S64S76.
Edeh, J. and Toone, B. (1987). Relationship between interictal psychopathology and the type of
epilepsy. Br J Psychiatry, 151, 95101.
Goldberg, L.R. (1972). Man versus mean: the exploitation of group proles for the construction
of diagnostic classication systems. J Abnorm Psychol, 79, 12131.
Gudmundsson, G. (1966). Epilepsy in Iceland a clinical and epidemiological investigation. Acta
Neurol Scand, Suppl 25, 43, 6490.
Gureje, O. (1991). Interictal psychopathology in epilepsy prevalence and pattern in a Nigerian
clinic. Br J Psychiatry, 158, 7005.
Hauser, A.W. (1998). Incidence and prevalence. In Epilepsy A Comprehensive Textbook, ed. J.R.
Engel and T.A. Pedley TA. New York: Lippincott-Raven.
Jalava, M. and Sillanpaa, M. (1996). Concurrent illnesses in adults with childhood-onset epi-
lepsy: a population-based 35-year follow-up study. Epilepsia, 37, 115563.
Krishnamoorthy, E.S. and Trimble, M.R. (1999). Forced normalization clinical and therapeu-
tic relevance. Epilepsia, 40 (Suppl. 10), S5764.
Krishnamoorthy, E.S. (2000). An approach to classifying neuropsychiatric disorders in epilepsy.
Editorial. Epilepsy Behav, 1, 3737.
Krishnamoorthy, E.S. (2001). Psychiatric issues in epilepsy. Curr Opinion Neurol, 14, 21724.
Lambert, M. and Robertson, M.M. (1999). Depression in epilepsy. Etiology, phenomenology,
and treatment. Epilepsia, 40 (Suppl. 10), S2147.
Lewis, G., Pelosi, A.J., Araya, R. and Dunn, G. (1992). Measuring psychiatric disorder in the com-
munity: a standardised assessment for use by lay interviewers. Psychol Med, 22, 46586.
Manchanda, R., Schaefer, B., McLachlan, R. et al. (1996). Psychiatric disorders in candidates for
surgery for epilepsy. J Neurol Neurosurg Psychiatry, 61, 829.
Mendez, M.F., Grau, R., Doss, R.C. and Taylor, J.L. (1993). Schizophrenia in epilepsy: seizure and
psychosis variables. Neurology, 43, 10737.
17 Classification
Pond, D.A. and Bidwell, B.H. (1959/60). A survey of epilepsy in fourteen general practices II.
Social and psychological aspects. Epilepsia, 1, 28599.
Schmitz, B. and Trimble, M.R. (1992). Epileptic equivalents in psychiatry: some 19th century
views. Acta Neurol Scand, 86 (Suppl. 140), 1226.
Shetty, T. and Trimble, M.R. (1997). The Bear Fedio Inventory: twenty years on. J Epilepsy, 10,
25462.
Slater, E. and Beard, A.W. (1963). The schizophrenia-like psychoses of epilepsy. Br J Psychiatry,
109, 95150.
Smith, D.B., Craft, B.R., Collins, J., Mattson, R.H., Cramer, J.A. and the VA Co-operative Study
Group 118 (1986). Behavioural characteristics of epilepsy patients compared with normal
controls. Epilepsia, 27, 7608.
Standage, K.F. and Fenton, G.W. (1975). Psychiatric symptom proles of patients with epilepsy:
a controlled investigation. Psychol Med, 5, 15260.
Stefansson, S.B., Olafsson, E. and Hauser, A.W. (1998). Psychiatric morbidity in epilepsy: a case
controlled study of adults receiving disability benets. J Neurol Neurosurg Psychiatry, 64,
23841.
Trimble, M.R. (1991). The Psychoses of Epilepsy. New York: Raven Press.
Trimble, M.R. (1998). New antiepileptic drugs and psychopathology. Neuropsychobiology, 38,
14951.
Whitman, S., Hermann, B.P. and Gordon, A.C. (1984). Psychopathology in epilepsy: how great
is the risk? Biol Psychiatry, 19, 21336.
World Health Organisation (1992). The International Classication of Mental and Behavioural
Disorders, Tenth Edition (ICD10). Geneva: WHO.
3
Limbic connectivity: anatomical substrates of

behavioural disturbances in epilepsy
J. Engel Jr, C. Wilson and F. Lopez-Rodriguez
UCLA School of Medicine, Los Angeles, USA
Introduction
Behavioural disturbances associated with epilepsy are due, in part, to a number of

important psychological and social factors; however, it is useful to acknowledge
that there are neurobiological factors as well. Not only does identication of
organic bases for behavioural aberrations help reduce the stigma they invariably
engender, but also it provides a rational scientic approach to prevent or reverse a
signicant cause of disability experienced by persons with epilepsy.
Cognitive and psychiatric disturbances associated with epilepsy may be due to
the same underlying pathological process that causes the epileptic condition, and
there is evidence that behavioural impairment may be related to the pathophysio-
logical nature of the underlying lesion, its eect on development and its location
(Engel et al., 1986; Engel and Shewmon, 1991). With respect to location, mesial
temporal and other limbic epileptogenic lesions are most likely to be associated
with behavioural disturbances (Engel et al., 1986; Engel and Shewmon, 1991).
Temporal lobe epilepsy is the most common form of human epilepsy, and hippo-
campal sclerosis is the most common human epileptogenic lesion (Engel, 1998).
Furthermore, mesial temporal lobe epilepsy with hippocampal sclerosis may be the
most refractory to antiepileptic drugs (Engel, 1998). Consequently, limbic system
dysfunction is undoubtedly the most important cause of behavioural disturbances
associated with epilepsy.
Epileptogenicity itself is known to cause enduring changes in brain function
and structure in a number of ways which can result in altered behaviour. For
instance, postnatally or in early childhood, abnormal recurrent synaptic bom-
bardment of distant projection areas by repeated ictal and interictal epileptiform
discharges can adversely aect the development of normal neuronal integration
(Engel and Shewmon, 1991; Morrell et al., 1995). In the mature brain such
abnormal synaptic activity can induce plastic changes through kindling mecha-
nisms (Corcoran and Mosh, 1998; Engel et al., 1986), and the naturally occur-
18
19 Anatomical substrates of behavioural disturbances
ring homeostatic seizure-suppressing mechanisms which maintain the interictal

state may have adverse consequences on normal neuronal function (Engel et al.,
1986; Engel and Shewmon, 1991). In addition, epileptic seizures are known to
disrupt sleep patterns (Shouse et al., 1997) and endocrine function (Herzog,
1997), which can alter behaviour; and severe epileptic activity results in neuro-
nal cell death and consequent synaptic reorganization (Sloviter and Damiano,
1981). Given the importance of the limbic system in the development and main-
tenance of cognition, emotion, motivation and other essential behaviours, it is
not surprising that epileptogenic lesions capable of disrupting limbic connectiv-
ity are associated with a great variety of clinically signicant behavioural abnor-
malities.
The limbic system
The term limbic system was coined by Maclean (1952) to designate a series of
structures originally delineated by Broca (1878) and Papez (1937), plus the amyg-
dala and its connections, which were believed to play a crucial role in mediating
the exchange of information between the thinking brain (cortex) and the more
primal animal brain (diencephalon and brain stem). Broca rst described the
limbic lobe as the area of brain making up the rim of the cortex, including the
hippocampus, cingulate cortex and certain frontal lobe structures. Papez more
precisely dened circuits that connected these structures, including the septal
area, mamillary bodies, and parts of the thalamus and hypothalamus. The addi-
tion of the amygdala and its connections by Maclean provided the functional sub-
strate for human emotional behaviours now attributed to the limbic system
(Figure 3.1). The hippocampus, amygdala and their projection elds are central
to the concept of the limbic system, and also are the most epileptogenic regions
of the mammalian brain. The perirhinal and piriform regions of the parahippo-
campal cortex, along with the central nucleus of the amygdala, are the most
rapidly kindled structures in the brain (McIntyre and Plant, 1993), and an area of
deep piriform complex, area tempesta, has the lowest threshold in the brain to sei-
zures induced by local application of antagonists to gamma amino butyric acid
(GABA), the primary inhibitory neurotransmitter in the brain (Piredda and Gale,
1985). The limbic system, therefore, is uniquely susceptible to the development
of epileptiform abnormalities, not only as a result of hippocampal sclerosis, but
due to any irritating disturbance. Furthermore, distant epileptogenic lesions
which preferentially project to mesial temporal structures also have a high pro-
pensity to induce mesial temporal epileptiform activity resulting in both tempo-
ral lobe seizures and their behavioural consequences (Williamson and Engel,
1997).
20 J. Engel Jr et al.
Cingulate SMA
Frontomedial
Fornix
Anterior cingulate
Thalamus
Fronto-lateral
Hypothalamus septum
MB Fronto-orbital
Hipp.
NA
Ento.
Parietal Parahipp.
Inferotemporal and
Post lateral anterior temporal
Insula
temporal
Mid-brain tegm. CC VHC DHC AC

reticular formation
Contralateral
hippocampal
formation
Second generalization
Figure 3.1. Diagrammatic representation of the human limbic system and its afferent and efferent
connections. Note the major connections between the hippocampus, entorhinal cortex
and amygdala (NA), and the components of the Papez circuit, described in the text. SMA,
supplementary motor area; HIPP, hippocampus; ENTO, entorhinal cortex; PARAHIPP,
parahippocampal gyrus; CC, corpus callosum; VHC, ventral hippocampal commissure;
DHC, dorsal hippocampal commissure; AC, anterior commissure. (With permission,
adapted from Wieser, 1988.)
Basket cell
Regio superior
(CA1)
Lateral
entorhinal
Schaffer cortex
collateral
Perforant
path
To
septum
Regio Medial entorhinal

Mossy cortex
inferior fibres
(CA3) Buried Exposed
blade blade
of fascia of fascia
dentata dentata
Figure 3.2. Illustration of the intrinsic circuitry of the hippocampal formation. The primary focus of
this illustration is the trisynaptic circuit. The first synapse consists of the mossy fibre to
CA3 pyramidal cell connection derived from the granule cells of the dentate gyrus. The
second synapse is the connection of the CA3 cell output via the Schaffer collaterals to
CA1, and the third, not shown, is the CA1 axonal output to subiculum and entorhinal
cortex. Entorhinal cortex also provides the initial input to the dentate gyrus. (With
permission from OKeefe and Nadel, 1978.)
Hippocampus
Normal structure
The hippocampus is perhaps the most studied region of the mammalian brain, due
in part to its highly organized laminar structure, and its propensity for plastic
change, such as long-term potentiation and long-term depression, which may
underlie mechanisms of normal learning and memory (Gloor, 1997;
Schwartzkroin and McIntyre, 1997). The hippocampal formation consists of the
dentate gyrus and the hippocampus proper or cornu ammonis (CA), which is
divided into CA1, CA2 and CA3 (Figure 3.2). The major excitatory hippocampal
input is glutaminergic from the entorhinal cortex via the perforant path, which ter-
minates on the distal two-thirds of granule cell dendrites in the dentate gyrus. The
medial portion of the entorhinal cortex projects predominantly to the middle third
of the granule cell dendritic layer (middle molecular layer), while the lateral ento-
rhinal cortex projections, which also contain opioids, terminate on the distal third
of granule cell dendrites (outer molecular layer). These latter projections may play
a role in epileptogenicity due to the opioid disinhibitory eect on granule cell excit-
ability (Xie et al., 1992). The neuronal structure of the dentate gyrus provides the
substrates for strong feed-forward and feedback inhibition which resists the prop-
agation of epileptiform activity, leading to its designation as the dentate gate
(Lothman et al., 1991).
The classical excitatory trisynaptic pathway through the hippocampus is glutam-
atergic and utilizes NMDA and non-NMDA receptors, while inhibitory interneu-
rons are GABAergic. The pathway begins with the dentate granule cell axons, called
mossy bres, which project through the dentate hilus to area CA3. Mossy bres
contain two types of terminals: the rst type are the large, complex mossy terminals
which give the bres their name, and the second type are small terminals and lo-
podia extending from the mossy terminals. The primary target of the mossy termi-
nals are the CA3 pyramidal cells, with which they form multiple excitatory synapses,
while the primary target of the small terminals and lopodia are GABAergic inter-
neurons. The ratio of dentate granule axon terminals contacting GABAergic inter-
neurons to those terminals contacting pyramidal and mossy cells (excitatory
neurons in the dentate hilus) has been estimated at approximately 5:1, which means
that the major postsynaptic targets of granule cells are inhibitory neurons (Acsdy
et al., 1998). Although the large proportion of granule cell output which activates
interneurons has a signicant inhibitory eect, the mossy terminals exciting CA3
pyramidal cells are quite powerful. Because CA3 pyramidal cell intrinsic connec-
tions support recurrent excitation, focused discharge of specic CA3 pyramidal cells
can occur. Excitation is further supported by granule cell activation of mossy cells
in the dentate hilus, which in turn project back onto the proximal one-third of the
granule cell dendrites (inner molecular layer). This excitatory feedback loop is oset
by similar inhibitory feedback loops via a variety of hilar inhibitory neurons. CA3
pyramidal cells have a unique propensity to burst synchronously, thus excitatory
epileptiform input that traverses the dentate gate activates synchronously bursting
pacemaker cells in CA3, giving rise to interictal spikes (Wong and Traub, 1983).
The second segment of the hippocampal trisynaptic pathway consists of the
Schaer collaterals, CA3 pyramidal cell axons which terminate on CA1 pyramidal
neurons. Under epileptogenic conditions, synchronous bursting in CA3 can
produce continuous discharges in CA1, the usual site of ictal onset in the hippo-
campal slice preparation. CA2 is a transition zone that does not appear to par-
ticipate in the transmission of epileptiform discharges in the nonsclerotic
hippocampus. However, pyramidal cells in this region are relatively spared in the
process of neuronal loss that characterizes hippocampal sclerosis (Mathern et al.,
1997), and these remaining neurons may play a more important role in mesial tem-
poral lobe epilepsy (Williamson and Spencer, 1994).
The third segment of the hippocampal trisynaptic pathway is its primary output,
from CA1 to the adjacent subiculum. The subiculum in turn projects back to the
entorhinal cortex, completing a closed hippocampal excitatory loop. This pathway
is not absolutely unidirectional, however; there are also signicant connections
from entorhinal cortex to CA1, and some bres even reach CA3. The subiculum
predominantly projects to perirhinal and piriform cortex, but also to the mamil-
lary bodies. The perirhinal cortex has connections with frontal motor cortical
areas, whereas the piriform cortex projects reciprocally to olfactory areas and
the brain stem. These projections are predominantly responsible for mediating the
clinical manifestations of temporal lobe seizures; ictal discharges conned to the
hippocampus may have no overt signs and symptoms (Sperling and OConnor,
1990). Parahippocampal areas may also become an important site of seizure gen-
eration; success in eliminating seizures by amygdalohippocampectomy, performed
to treat medically intractable temporal lobe epilepsy, appears to depend upon the
amount of parahippocampal tissue removed (Siegel et al., 1990).
CA1 axons also project via the mbria and fornix along the traditional Papez
circuit to the septal area, midline thalamus, amygdala, hypothalamus and brain
stem autonomic centres. The amygdala, hypothalamic and brain stem projections
mediate aspects of emotional behaviours, and the hippocampus may play a role in
the adrenohypothalamicpituitary axis, in view of the fact that adrenal steroid
receptors are abundant in the hippocampus (Gould et al., 1991).
The hippocampus also receives direct cholinergic and GABAergic input from the
septal area, which mediates the classical hippocampal theta rhythm, and these pro-
jections may have disinhibitory as well as inhibitory inuences on epileptic activ-
ity. With respect to brain stem aerents, direct noradrenergic input to the
hippocampus from the locus coeruleus can be excitatory (Madison and Nicoll,
1986), while serotonergic input from the raph nucleus and dopaminergic input
from the substantia nigra and ventral tegmental area are predominantly inhibitory
(Andrade and Nicoll, 1987), but biogenic amine eects can be varied, depending
on the postsynaptic target. Indirect aerent inuences on the hippocampus, via
entorhinal cortex, derive from a large area of neocortex, including frontal limbic
regions responsible for goal-directed behaviour, cingulate cortex inuencing
memory, all sensory cortical areas for integration of polymodal sensory informa-
tion, and perirhinal and piriform cortex, as well as from amygdala.
Hippocampal sclerosis
Much is known about the structural and functional abnormalities that exist in the
epileptic hippocampus, due in part to the ready availability of human tissue from
epilepsy surgery programmes, and in part to numerous animal models of this con-
dition; however, the cause of these abnormalities, and the reasons why these abnor-
malities should result in spontaneous seizures, are not well understood. The
characteristic nding in hippocampal sclerosis is cell loss, most prominent in the
dentate hilus and CA1, but it is also marked in CA3 and, to a lesser extent, the
dentate gyrus and subiculum (Mathern et al., 1997). Although specic hilar inter-
neurons are lost, namely those containing somatostatin and neuropeptide Y, in
general there is relative preservation of hippocampal inhibitory interneurons, com-
pared to principal cells (Babb, 1992). The greatest proportion of hilar cell loss is
from mossy cell death, which results in loss of excitatory inputs to dentate granule
cell proximal dendrites, and consequent mossy bre sprouting. These collateral
granule cell axons, which can easily be seen with Timms stain, reinnervate the inner
molecular layer, but most likely end on inhibitory interneurons, as well as on
granule cell dendrites (Figure 3.3). There is also evidence of inhibitory interneuron
sprouting (Babb, 1992). Electrophysiological studies in patients have conrmed
that inhibition is actually increased, interictally, in the epileptiform sclerotic hippo-
campus (Wilson et al., 1998). This enhanced inhibition may act to suppress seizure
generation; however, it has been postulated that such enhanced inhibition together
with enhanced excitation, brought about by pathological neuronal reorganization,
results in a predisposition to abnormal neuronal hypersynchronization (Figure
3.4). In addition, single input bres sprout to innervate wider targets, and reduc-
tion of the dendritic domain places more excitatory input closer to the soma and
axon hillock. Hypersynchronization underlies interictal epileptiform EEG spikes,
but is also the hallmark of the typically hypersynchronous hippocampal ictal EEG
onset (Velasco et al., 2000). Consequently, enhanced inhibition in the sclerotic hip-
pocampus may contribute to its epileptogenicity (Engel et al., 1997). During the
process of epileptogenesis, once the local anatomical substrate for abnormal hyper-
synchrony has been established, these hypersynchronous discharges may act to
kindle distant structures responsible for the manifestation of ictal behaviour.
Amygdala
The amygdala has been implicated in a number of important primal functions,

including fear, aggression, learning, reproduction and appetitive drive. In contrast
to the hippocampus, it consists of a number of loosely organized subnuclei with
diverse neurons and cytoarchitecture, which have been divided into an olfactory
group, a centromedial group and a basolateral group (Gloor, 1997; Schwartzkroin
and McIntyre, 1997). The olfactory group has predominant reciprocal connections
with the hippocampus, piriform cortex and olfactory bulb, with less strong con-
nections to the septal area, brain stem nuclei, hypothalamus, thalamus and the
Normal
FD
SG Hilus
MML IML
OML
Mossy fibre reinnervation
FD
SG Hilus
MML IML
CML
Figure 3.3. Schematic diagram showing the reciprocal innervation between principal neurons of the
hilus (squares), and dentate granule cells of the hippocampal formation (circles). In the
normal situation shown above, granule cell axons (mossy fibres) synapse on principal
neurons of the hilus and CA3, while hilar neurons make synapses on proximal dendrites
of the granule cells. The reinnervation that occurs with neuronal loss, such as occurs with
hippocampal sclerosis, is shown below. When hilar input to granule cells is decreased, the
mossy fibres of surviving granule cells sprout and make contact with the vacated
postsynaptic membrane, creating monosynaptic recurrent excitatory circuits. FD, fascia
dentata; OML, outer molecular layer; MML, middle molecular layer; IML, inner molecular
layer; SG, stratum granulosum. (With permission from Engel, 1990.)
Figure 3.4. Reciprocal innervation of a hypothetical neuronal system is shown schematically above.
This is the typical synaptic organization of neocortex and hippocampus. Excitatory afferent
input terminates on the dendrites of principal neurons (filled triangles). Axon collaterals
from these principal neurons terminate on inhibitory interneurons (empty circles), which
in turn make hyperpolarizing synapses on the soma of the same, and adjacent (not
shown), principal neurons. With cell loss, a number of synaptic reorganizations are likely
to occur, as shown schematically on the right. Fewer afferent input fibres sprout to
innervate more principal neurons, predisposing to hypersynchronization. Because the
dendrites of principal neurons are shorter, these excitatory influences are closer to the
axon hillock and more likely to induce neuronal firing. Neuronal excitability is further
increased by the establishment of monosynaptic excitatory recurrent circuits, as also
shown in Figure 3.3. Although not definitively demonstrated, it is possible that inhibitory
interneurons sprout terminals to produce more powerful, and/or more extensive,
recurrent inhibitory influences, further enhancing the potential for hypersynchronization.
(With permission from Engel, 1990.)
nucleus accumbens of the basal forebrain. The centromedial group has reciprocal
connections with extensive areas of the brain stem, including the periaqueductal
grey, but also with the entorhinal cortex, hippocampus, hypothalamus, thalamus
and striatum. The basolateral group has strong reciprocal connections with soma-
tosensory and motor cortex, with only weak brain stem connections. It is likely,
therefore, that the centromedial nuclei, particularly the central nucleus of the
amygdala, are most important in mediating aective ictal and interictal symptoms
observed with amygdala involvement in the propagation of epileptiform dis-
charges.
Interhemispheric limbic connections
Due to the enormous expansion of the human cerebral cortex in comparison with
that of subprimate mammals, the limbic structures, with the exception of the cin-
gulate gyrus, have moved far from the midline. Interhemispheric connections are
necessarily increased in length and changed in trajectory by this major reorganiza-
tion (Wilson, 1995). Three pathways exist for interhemispheric communication
between limbic areas: corpus callosum, anterior commissure and hippocampal
commissure.
Corpus callosum
The corpus callosum carries bres from the cingulate gyrus, but also from the
medial orbital frontal cortex, the posterior two-thirds of the temporal cortex and
the posterior parahippocampal gyrus (Demeter et al., 1990). As a component of the
Papez circuit, cingulate cortex is in a position to relay output activity from limbic
structures to the contralateral hemisphere, as well as back to the ipsilateral ento-
rhinal cortex. The callosal connections of the orbital frontal cortex are signicant
for interhemispheric limbic interaction because of the major projection this area
receives from the amygdala (Amaral et al., 1984).
Anterior commissure
In primates, the anterior commissure carries bres from olfactory and other areas
of the orbital frontal cortex, piriform cortex, the anterior third of the temporal
cortex and posterior parahippocampal gyrus, and a few sparse bres from the cor-
tical and deep subnuclei of the amygdala. For the most part, the widespread corti-
cal projections of the amygdala are ipsilateral only, and the small interhemispheric
projection which exists terminates in contralateral temporal cortex, not in contra-
lateral amygdala (Demeter et al., 1990).
Hippocampal commissure
Phylogenetically, commissural bres have constituted a major group of extrinsic
connections joining the two hippocampi and their subelds, but the connections are
substantially reduced in the human brain. There are two components of the inter-
hemispheric pathway: the ventral hippocampal commissure, which carries bres
from the hippocampal formation, and the dorsal hippocampal commissure, which
carries bres from the entorhinal cortex, presubiculum and parahippocampal
gyrus. In the subhuman primate, the ventral hippocampal commissure is greatly
reduced and limited to the anterior hippocampus, while the dorsal commissure
carries primarily presubicular and parahippocampal gyrus bres to the contralat-
eral entorhinal cortex (Amaral et al., 1984). In the human brain, dissection of
autopsy specimens has revealed the presence of a dorsal hippocampal commissure,

but not the ventral (Gloor et al., 1993), and electrophysiology studies (Wilson et al.,
1990, 1991) provide additional functional evidence that hippocampal connections
are absent in humans.
The relative reduction of commissural interchange from lower animal to primate
may provide the basis for the hemispheric specialization unique to the human
brain, particularly relating to dierences in verbal versus spatial memory perfor-
mance, which is lateralized to the left and right temporal lobes, respectively
(Amaral et al., 1984; Demeter et al., 1985). The absence of any signicant inter-
hemispheric amygdala connections suggests that emotion and aect may have sim-
ilarly lateralized or hemispherically specialized components (Demeter et al., 1990),
which has important implications for patients with temporal lobe seizures.
Interictal and ictal behavioural and emotional manifestations of mesial temporal
lobe epilepsy can be very dierent depending on whether the epileptogenic abnor-
mality is on the right or left.
Possible anatomic substrates of some behavioural disturbances associated

with epilepsy
The behavioural disturbances most associated with epilepsy include depression

and other aective disorders and personality traits, schizophrenia, aggressivity and
anxiety.
Affective disturbances
The prevalence of aective disorders is higher among epileptic patients than the
general population, or neurological patients with the same degree of disability
(Mendez et al., 1986). No neurobiological mechanism has successfully explained
this correlation between aective disorders and epilepsy. However, several hypoth-
eses have been proposed.
Reciprocal connections between limbic structures and those brain stem nuclei
with biogenic amine and opioid projections to the forebrain provide ample path-
ways for the mediation of depression and other abnormalities of aect. In particu-
lar, endogenous opioids are known to be elevated during seizures (Vindrola et al.,
1981), and mu-opiate-receptor binding is increased in the human epileptic tempo-
ral lobe (Frost et al., 1988). It is postulated that opioid peptides act as endogenous
euphorogens, which may explain why electroconvulsive treatment is an eective
therapy for depression (Kline et al., 1977). Conceivably, patients with frequent
limbic seizures develop a dependence on high levels of circulating opioid peptides,
and the appearance of aective symptoms reects a form of withdrawal when sei-
zures do not regularly recur (Engel et al., 1991). Indeed, transient depression lasting
a few months is commonly observed in patients who become suddenly seizure-free

following successful surgical treatment (Glosser et al., 2000; Krahn et al., 1996).
Depression is more common among patients with complex partial seizures than
with generalized seizures (Devinsky and Vasquez, 1993). This nding is not sur-
prising, because complex partial seizures originate from those limbic structures in
which dysfunction has been strongly correlated with depression. For example, hip-
pocampal damage after stress, induced by elevated levels of cortisol, has been pro-
posed as a possible mechanism for depression (Stokes, 1995). Aective symptoms
have been associated with dysfunction in a circuit that travels from the anterior cin-
gulate and medial orbito-frontal cortices to the ventral striatum (nucleus accum-
bens and olfactory tubercle), the ventral pallidum, the medial dorsal nucleus of the
thalamus, and back to the anterior cingulate and medial orbito-frontal cortices
(Alexander et al., 1990; Ketter et al., 1996). Abnormalities in the components of this
circuit have also been described with complex partial seizures (Bromeld et al.,
1992; Savic et al., 1997). A common nding among patients with a primary aec-
tive disturbance, as well as in epileptic patients with depression, is decreased meta-
bolic activity of the prefrontal cortex (Bromeld et al., 1992; Rogers et al., 1998).
Whether the nding of decreased prefrontal metabolism is a consequence of the
depression in patients with complex partial seizures, or alternatively, complex
partial seizures have secondarily aected the activity of the prefrontal cortex, with
a consequent development of depression, is not clear. However, based on the exis-
tence of strong connections between the mesial temporal structures and the pre-
frontal cortex, this second hypothesis is anatomically and functionally plausible
(Drevets, 1998). Abnormalities of metabolism in the amygdala and anterior cingu-
late have also been repeatedly reported to be associated with depression (Drevets,
1998; Mayberg et al., 1994) and temporal lobe epilepsy (Henry et al., 1993a,b).
Therefore, complex partial seizures could induce functional changes in limbic areas
that overlap with those structures involved in the manifestations of aective
psychopathology.
Schizophrenia
The dopaminergic hypothesis of schizophrenia originated initially from two indi-
rect ndings: except for clozapine, all eective antipsychotics are blockers of D2
dopamine (DA) receptors, and stimulants that increase DA release in the limbic
system frequently induce psychosis (Farde, 1997). This hypothesis has been applied
mostly to the positive symptoms of schizophrenia such as hallucinations and delu-
sions, because typical antipsychotic medications, which act almost exclusively on
D2 receptors, are not very ecient in treating negative symptoms and thought dis-
organization. This last fact, together with the delay in action of antipsychotic med-
ications (although D2 antagonism is immediate), put the dopaminergic theory into
question (Goldstein and Deutch, 1992). However, recent studies have shown that
schizophrenic patients experience increased DA release after amphetamine chal-
lenge, when compared with control subjects, and this hyperactivity of DA trans-
mission was found to be associated with activation of psychotic symptomatology
(Laruelle and Abi-Dargham, 1999).
In cats, kindling of the dopaminergic ventral tegmental area of the brain stem,
which projects to the nucleus accumbens of the basal forebrain, does not produce
seizures, but alters behaviour in a way that has been interpreted to resemble aspects
of schizophrenia (Stevens, 1973). The nucleus accumbens and the ventral tegmen-
tal area both receive input from the mesial temporal limbic system, and are thus
subject to kindling-like aerent input from seizures originating from the hippo-
campus or perihippocampal regions. An enduring eect of limbic seizures on
dopaminergic function has been demonstrated experimentally by the fact that
amygdala kindling in cats enhances methamphetamine-induced stereotypy, sug-
gesting upregulation of DA receptors (Sato, 1983), although the location of these
receptors has not been identied.
Furthermore, disturbances of limbic structures have been associated with
schizophrenia (Weinberger et al., 1992). Pathological studies of brains of schizo-
phrenic patients revealed a disorganization of the typical laminar structure of the
hippocampus (Kovelman and Scheibel, 1984). Studies from discordant monozy-
gotic twins (i.e., one twin is aected by schizophrenia whereas the other is healthy),
have shown that the schizophrenic twin had smaller hippocampi, and that the size
of the hippocampi correlated with hypofrontality (assessed by blood-ow studies
during performance of the Wisconsin Card Sorting Test) in the aected twins
(Weinberger et al., 1992). This nding is consistent with the possibility that path-
ological abnormalities in the hippocampus might result in a widespread disruption
of other corticolimbic structures, such as the prefrontal cortex, with a consequent
development of schizophrenic symptoms.
Aggression
Kindling is known to induce postictal hyperreactivity (which has been mistaken for
aggression), as well as other postictal behaviours which can all be modulated by
pretreatment with opioid agonists and antagonists (Caldecott-Hazard and Engel,
1987). Some of these behaviours may be direct opioid eects, while others may rep-
resent withdrawal phenomena. Mesial limbic structures project to the brain stem
periaqueductal grey, a major source of forebrain opioid projections, and this area
has been implicated in an experimental model of aggression in cats termed defen-
sive rage. Electrical stimulation of this area, and of the lateral hypothalamus, pro-
duces a stereotypic Halloween cat response, with piloerection, arched back,
pupillary dilation and attack behaviour, lasting only as long as the duration of the
stimulus (Bandler, 1988). This behaviour can be modulated by lesions, or stimula-

tion, of hippocampus and amygdala.
Production of experimental temporal lobe epilepsy in cats, by hippocampal
injection of kainic acid which causes a lesion similar to hippocampal sclerosis, has
a profound eect on this behavioural model (Engel et al., 1991; Grith et al., 1987).
Immediately following kainic acid injection, cats experience continuous limbic sei-
zures lasting many hours, and during this time also exhibit spontaneous defensive
rage. This is followed by a latent period when the damaged hippocampus under-
goes epileptogenic neuronal reorganization and ultimately gives rise to spontane-
ous limbic seizures. During the latent period, which can last from weeks to months,
there are no spontaneous seizures and behaviour is normal. Once spontaneous sei-
zures occur, however, not only is the threshold to electrical stimulation-induced
defensive rage lowered, but cats also exhibit defensive rage in response to environ-
mental stimuli, in the absence of periaqueductal grey or lateral hypothalamic
electrical stimulation. Furthermore, the defensive rage reaction can be elicited only
when the cat is approached contralateral to the epileptic hippocampus (Figure 3.5).
Clearly, hippocampal epileptogenesis has induced distant structural changes, pre-
sumably involving amygdala, hypothalamus and periaqueductal grey, resulting in
enduring aberrant behaviour.
These experiments should not be interpreted to suggest that patients with mesial
temporal lobe epilepsy are susceptible to defensive rage or other aggressive behavi-
ours. Defensive rage is a hardwired species-specic behaviour of the cat, and the
brain stem and diencephalic structures responsible for its manifestation undoubt-
edly come under strong neocortical control in the human. A variety of human cor-
relates, other than rage, might be anticipated, including impulsivity, insecurity,
paranoia and perhaps even depression.
Anxiety
The amygdala is believed to provide an aective tone to perceptions and experi-
ences, and human lesion and neuroimaging studies have demonstrated a role for
the amygdala in verbal and nonverbal recognition of emotions (Adolphs et al.,
1998; Hariri et al., 2000), as well as in fear conditioning (Furmark et al., 1997) and
provoked anxiety (Davidson et al., 1999). In animals, there is also extensive evi-
dence to support a role for the amygdala in fear and fear conditioning (LeDoux,
1993). Other structures that have been found to be repeatedly correlated with
anxiety are the orbito-frontal cortex, anterior insula and anterior cingulate
(Malizia, 1999).
The projections of the central nucleus of the amygdala to the locus coeruleus and
hypothalamus appear to be responsible for the arousal and autonomic responses
associated with fear and anxiety (Goldstein et al., 1996) which, together with the
Figure 3.5. Demonstration of lateralized hypereactivity. Photographs were taken from a videotape
recording of a cat with a chronic, kainic acid-induced epileptogenic lesion in the left
hippocampus. Top: when the cat was approached on the left (ipsilateral) side, it
responded by rubbing up against the offered hand and purring. Bottom: when an attempt
was made to touch the cat on the right (contralateral) side, it assumed a defensive
posture (ears back, piloerection, and pupillary dilatation), hissed, and struck at the offered
hand with the right paw. (With permission from Engel et al., 1991.)
well-established function of the amygdala in emotion recognition and fear condi-

tioning, imply a key role of the amygdala in the integration of behavioural and
neuroendocrine components of the stress response. Amygdala-kindled rats show
behaviours suggestive of increased fear, such as decreased exploration in the open
eld (Depaulis et al., 1997). Consequently, there are ample anatomical and patho-
physiological substrates that can be invoked to explain both ictal and interictal
anxiety as well as panic disorder in patients with temporal lobe epilepsy.
Acknowledgements
Original research reported by the author was supported in part by Grants NS-
02808, NS-15654, NS-33310, and GM-24839, from the National Institutes of
Health, and Contract DE-AC03-76-SF00012 from the Department of Energy.
R E F E R E N C ES
Acsdy, L., Kamondi, A., Sik, A., Freund, T. and Buzski, G. (1998). GABAergic cells are the major
postsynaptic targets of mossy bers in the rat hippocampus. J Neurosci, 18, 3386403.
Adolphs, R., Tranel, D. and Damasio, A.R. (1998). The human amygdala in social judgment [see
comments]. Nature, 393, 4704.
Alexander, G.E., Crutcher, M.D. and DeLong, M.R. (1990). Basal ganglia-thalamocortical cir-
cuits: parallel substrates for motor, oculomotor, prefrontal and limbic functions. Prog
Brain Res, 85, 11946.
Amaral, D.G., Insausti, R. and Cowan, W.M. (1984). The commissural connections of the
monkey hippocampal formation. J Comp Neurol, 224, 30736.
Andrade, R. and Nicoll, R.A. (1987). Pharmacologically distinct actions of serotonin on single
pyramidal neurones of the rat hippocampus recorded in vitro. J Physiol, 394, 99124.
Babb, T.L. (1992). GABA neurons, synapses and inhibition in human hippocampal epilepsy. In
Epilepsy and Inhibition, ed. E.-J. Speckmann and M.J. Gutnick, pp. 37597. Munich: Urban and
Schwarzenberg Press.
Bandler, R. (1988). Brain mechanisms of aggression as revealed by electrical and chemical
stimulation: suggestion of a central role for the midbrain periaqueductal gray region. In
Progress in Psychobiology, ed. A. Epstein & A. Morrison, pp. 67153. New York: Academic
Press.
Broca, P. (1878). Anatomie compare des circonvolutions cerebrales. Le grand lobe limbique et
la scissure limbique dans le serie des mammiferes. Rev Anthropol, Ser 2, 1, 385498.
Bromeld, E.B., Altshuler, L., Leiderman, D.B. et al. (1992). Cerebral metabolism and depression
in patients with complex partial seizures. Arch Neurol, 49, 61723.
Caldecott-Hazard, S. and Engel, J. Jr (1987). Limbic postictal events: Anatomical substrates and
opioid receptor involvement. Prog Neuropsychopharmacol Biol Psychiatry, 11, 389418.
Corcoran, M.E. and Mosh, S.L. (Eds) (1998). Kindling V. New York: Plenum Press.
Davidson, R.J., Abercrombie, H., Nitschke, J.B. and Putnam, K. (1999). Regional brain function,
emotion and disorders of emotion. Curr Opin Neurobiol, 9, 22834.
Demeter, S., Rosene, D.L. and Van Hoesen, G.W. (1985). Interhemispheric pathways of the hip-
pocampal formation. Presubiculum, entorhinal and posterior hippocampal cortices in the
rhesus monkey: the structure and function of the hippocampal commissures. J Comp Neurol,
233, 3047.
Demeter, S., Rosene, D.L. and Van Hoesen, G.W. (1990). Fields of origins and pathways of the
interhemispheric commissures in the temporal lobe of macaques. J Comp Neurol, 302, 29.
Depaulis, A., Helfer, V., Deransart, C. and Marescaux, C. (1997). Anxiogenic-like consequences
in animal models of complex partial seizures. Neurosci Biobehav Rev, 21, 76774.
Devinsky, O. and Vazquez, B. (1993). Behavioral changes associated with epilepsy. Neurol Clin,
11, 12749.
Drevets, W.C. (1998). Functional neuroimaging studies of depression: the anatomy of melancho-
lia. Annu Rev Med, 49, 34161.
Engel, J. Jr (1990). Functional explorations of the human epileptic brain and their therapeutic
implications. Electroencephalogr Clin Neurophysiol, 76, 296316.
Engel, J. Jr (1998). Etiology as a risk factor for medically refractory epilepsy: a case for early sur-
gical intervention. Neurology, 51, 12434.
Engel, J. Jr and Shewmon, D.A. (1991). Impact of the kindling phenomenon on clinical epilep-
tology. In Kindling and Synaptic Plasticity: The Legacy of Graham Goddard, ed. F. Morrell, pp.
195210. Cambridge, MA: Birkhuser Boston.
Engel, J. Jr, Caldecott-Hazard, S. and Bandler, R. (1986). Neurobiology of behavior: anatomic
and physiologic implications related to epilepsy. Epilepsia, 27(Suppl. 2), S3S11.
Engel, J. Jr, Bandler, R., Grith, N.C. and Caldecott-Hazard, S. (1991). Neurobiological evidence
for epilepsy-induced interictal disturbances. In Advances in Neurology, ed. D. Smith, D.
Treiman & M. Trimble, pp. 97111. New York: Raven Press. Vol. 55.
Engel, J. Jr, Dichter, M. and Schwartzkroin, P. (1997). Basic mechanisms of human epilepsy. In
Epilepsy: A Comprehensive Textbook, ed. J. Engel Jr and T.A. Pedley, pp. 499512. Philadelphia,
New York: Lippincott-Raven.
Farde, L. (1997). Brain imaging of schizophrenia the dopamine hypothesis. Schizophr Res, 28,
15762.
Frost, J.J., Mayberg, H.S., Fisher, R.S. et al. (1988). Mu-opiate receptors measured by positron
emission tomography are increased in temporal lobe epilepsy. Ann Neurol, 23, 2317.
Furmark, T., Fischer, H., Wik, G., Larsson, M. and Fredrikson, M. (1997). The amygdala and
individual dierences in human fear conditioning. Neuroreport, 8, 395760.
Gloor, P. (1997). The Temporal Lobe and the Limbic System. New York: Oxford University Press.
Gloor, P., Salanova, V., Olivier, A. and Quesney, L.F. (1993). The human dorsal hippocampal
commissure: an anatomically identiable and functional pathway. Brain, 116, 124973.
Glosser, G., Zwil, A.S., Glosser, D.S., OConnor, M.J. and Sperling, M.R. (2000). Psychiatric
aspects of temporal lobe epilepsy before and after anterior temporal lobectomy [In Process
Citation]. J Neurol Neurosurg Psychiatry, 68, 538.
Goldstein, M. and Deutch, A.Y. (1992). Dopaminergic mechanisms in the pathogenesis of
schizophrenia. FASEB J, 6, 241321.
Goldstein, L.E., Rasmusson, A.M., Bunney, B.S. and Roth, R.H. (1996). Role of the amygdala in
the coordination of behavioral, neuroendocrine, and prefrontal cortical monoamine
responses to psychological stress in the rat. J Neurosci, 16, 478798.
Gould, E., Woolley, C.S and McEwen, B.S. (1991). The hippocampal formation: morphological
changes induced by thyroid, gonadal, and adrenal hormones. Psychoneuroendocrinology, 16,
6784.
Grith, N., Engel, J. Jr and Bandler, R. (1987). Ictal and enduring interictal disturbances in emo-
tional behaviour in an animal model of temporal lobe epilepsy. Brain Res, 400, 3604.
Hariri, A.R., Bookheimer, S.Y. and Mazziotta, J.C. (2000). Modulating emotional responses:
eects of a neocortical network on the limbic system. Neuroreport, 11, 438.
Henry, T.R., Engel, J. Jr and Mazziotta, J.C. (1993a). Clinical evaluation of interictal uorine-18-
uorodeoxyglucose PET in partial epilepsy. J Nucl Med, 34, 18928.
Henry, T.R., Mazziota, J.C. and Engel, J. Jr (1993b). Interictal metabolic anatomy of mesial tem-
poral lobe epilepsy. Arch Neurol, 50, 5829.
Herzog, A.G. (1997). Disorders of reproduction and fertility. In Epilepsy: A Comprehensive
Textbook, ed. J. Engel Jr & T.A. Pedley, pp. 201319. Philadelphia, New York: Lippincott-Raven.
Ketter, T.A., George, M.S., Kimbrell, T.A., Benson, B.E. and Post, R.M. (1996). Functional brain
imaging, limbic function, and aective disorders. Neuroscientist, 2, 5565.
Kline, N.S., Li, C.H., Lehmann, H.E., Lajtha, A., Laski, E. and Cooper, T. (1977). Beta-endorphine-
induced changes in schizophrenic and depressed patients, Arch Gen Psychiatry, 34, 111113.
Kovelman, J.A. and Scheibel, A.B. (1984). A neurohistological correlate of schizophrenia. Biol
Psychiatry, 19, 160121.
Krahn, L.E., Rummans, T.A. and Peterson, G.C. (1996). Psychiatric implications of surgical treat-
ment of epilepsy. Mayo Clin Proc, 71, 12014.
Laruelle, M. and Abi-Dargham, A. (1999). Dopamine as the wind of the psychotic re: new evi-
dence from brain imaging studies. J Psychopharmacol, 13, 35871.
LeDoux, J.E. (1993). Emotional memory systems in the brain. Behav Brain Res, 58, 6979.
Lothman, E.W., Bertram, E.H. III and Stringer, J.L. (1991). Functional anatomy of hippocampal
seizures. Prog Neurobiol, 37, 182.
Maclean, P.D. (1952). Some psychiatric implications of physiological studies on frontotemporal
portion of limbic system (visceral brain). EEG Clin Neurophysiol, 4, 40718.
Madison, D.V. and Nicoll, R.A. (1986). Actions of noradrenaline recorded intracellularly in rat
hippocampal CA1 pyramidal neurones, in vitro. J Physiol (Lond), 372, 22144.
Malizia, A.L. (1999). What do brain imaging studies tell us about anxiety disorders? J
Psychopharmacol, 13, 3728.
Mathern, G.W., Babb, T.L. and Armstrong, D.L. (1997). Hippocampal sclerosis. In Epilepsy: A
Comprehensive Textbook, ed. J. Engel Jr & T.A. Pedley, pp. 13355. Philadelphia, New York:
Lippincott-Raven.
Mayberg, H.S, Lewis, P.J., Regenold, W. and Wagner, H.N.J. (1994). Paralimbic hypoperfusion in
unipolar depression. J Nucl Med, 35, 92934.
McIntyre, D.C. and Plant, J.R. (1993). Long-lasting changes in the origin of spontaneous dis-
charges from amygdala-kindled rats: piriform vs. perirhinal cortex in vitro. Brain Res, 624,
26876.
Mendez, M.F., Cummings, J.L. and Benson, F. (1986). Depression in epilepsy: signicance and
phenomenology. Arch Neurol, 43, 76670.
Morrell, F., Whisler, W.W., Smith, M.C. et al. (1995). Landau-Klener syndrome: treatment with
subpial intracortical transection. Brain, 118, 152946.
OKeefe, J. and Nadel, L. (1978). The Hippocampus as a Cognitive Map. Oxford: Clarendon Press.
Papez, J.W. (1937). A proposed mechanism of emotion. Arch Neurol Psychiatry, 38, 72543.
Piredda, S. and Gale, K. (1985). A crucial epileptogenic site in the deep prepiriform cortex.
Nature, 317, 62325.
Rogers, M.A., Bradshaw, J.L., Pantelis, C. and Phillips, J.G. (1998). Frontostriatal decits in uni-
polar major depression. Brain Res Bull, 47, 297310.
Sato, M. (1983). Long-lasting hypersensitivity to methamphetamine following amygdaloid kin-
dling in cats: the relationship between limbic epilepsy and the psychotic state. Biol Psychiatry,
18, 52536.
Savic, I., Altshuler, L., Baxter, L. and Engel, J. Jr (1997). Pattern of interictal hypometabolism in
PET scans with udeoxyglucose F18 reects prior seizure types in patients with mesial tempo-
ral lobe seizures. Arch Neurol, 54, 12936.
Schwartzkroin, P.A. and McIntyre, D.C. (1997). Limbic anatomy and physiology. In Epilepsy: A
Comprehensive Textbook, ed. J. Engel Jr and T.A. Pedley, pp. 32340. Philadelphia, New York:
Lippincott-Raven.
Shouse, M.N., Martins da Silva, A. and Sammaritano, M. (1997). Sleep. In Epilepsy: A
Comprehensive Textbook, ed. J. Engel Jr and T.A. Pedley, pp. 192942. Philadelphia, New York:
Lippincott-Raven.
Siegel, A.M., Wieser, H.G., Wichmann, W. and Yasargil, G.M. (1990). Relationships between MR-
imaged total amount of tissue removed, resection scores of specic mesiobasal limbic subcom-
partments and clinical outcome following selective amygdalohippocampectomy. Epilepsy Res,
6, 5665.
Sloviter, R.S. and Damiano, B.P. (1981). Sustained electrical stimulation of the perforant path
duplicates kainate-induced electrophysiological eect and hippocampal damage in rats.
Neurosci Lett, 24, 27984.
Sperling, M.R. and OConnor, M.J. (1990). Auras and subclinical seizures: characteristics and
prognostic signicance. Ann Neurol, 28, 3208.
Stevens, J.R. (1973). An anatomy of schizophrenia. Arch Gen Psychiatry, 29, 17789.
Stokes, P.E. (1995). The potential role of excessive cortisol induced by HPA hyperfunction in the
pathogenesis of depression. Eur Neuropsychopharmacol, 5, 7782.
Velasco, A.L., Wilson, C.L., Babb, T.L. and Engel, J. Jr (2000). Functional and anatomic correlates
of two frequently observed temporal lobe seizure-onset patterns. Neural Plasticity, 7, 4963.
Vindrola, O., Briones, R., Asai, M. and Fernandez-Guardiola, A. (1981). Brain content of leu 5-
and met 5-enkephalin changes independently during the development of kindling in the rat.
Weinberger, D.R. Berman, K.F., Suddath, R. and Torrey, E.F. (1992). Evidence of dysfunction of
a prefrontal-limbic network in schizophrenia: a magnetic resonance imaging and regional
cerebral blood ow study of discordant monozygotic twins. Am J Psychiatry, 149, 8907.
Wieser, H.G. (1988). Human limbic seizures: EEG studies, origin, and patterns of spread. In
Anatomy of Epileptogenesis, ed. B.S. Meldrum, J.A. Ferrendelli and H.G. Wieser, pp. 12738.
London: John Libbey.
Williamson, A. and Spencer, D.D. (1994). Electrophysiological characterization of CA2 pyrami-
dal cells from epileptic humans. Hippocampus, 2, 22637.
Williamson, P.D. and Engel, J. Jr (1997). Complex partial seizures. In Epilepsy: A Comprehensive
Textbook, ed. J. Engel Jr and T.A. Pedley, pp. 55766. Philadelphia, New York: Lippincott-
Raven.
Wilson, C.L. (1995). Functional pathways underlying ipsilateral and contralateral spread of tem-
poral lobe seizures. In Epilepsy and the Corpus Callosum, ed. A. Reeves, Vol. II, pp. 15373. New
York: Plenum Press.
Wilson, C.L., Isokawa-Akesson, M., Babb, T.L. and Crandall, P.H. (1990). Functional connec-
tions in the human temporal lobe: I. Analysis of limbic system pathways using neuronal activ-
ity evoked by electrical stimulation. Exp Brain Res, 82, 27992.
Wilson, C.L., Isokawa, M., Babb, T.L., Crandall, P.H., Levesque, M.F. and Engel, J. Jr (1991).
Functional connections in the human temporal lobe: II. Evidence for a loss of functional
linkage between contralateral limbic structures. Exp Brain Res, 85, 17487.
Wilson, C.L., Khan, S.U., Engel, J. Jr, Isokawa, M., Babb, T.L. and Behnke, E.J. (1998). Paired
pulse suppression and facilitation in human epileptogenic hippocampal formation. Epilepsy
Res, 31, 21130.
Wong, R.K.S. and Traub, R.D. (1983). Synchronized burst discharge in disinhibited hippocam-
pal slice. I. Initiation in CA2-CA3 region. J Neurophysiol, 49, 44258.
Xie, C.W., Morrisett, R.A. and Lewis, D.V. (1992). Mu opioid-receptor mediated modulation of
synaptic currents in the dentate granule cells of rat hippocampus. J Neurophysiol, 68, 111320.
Part II
Clinical aspects
4
The psychiatry of idiopathic generalized

epilepsy
Dieter Janz
Burgunder Strasse 8, Berlin, Germany
Temporal Lobe Epilepsy (TLE) is associated with an increase in psychiatric mor-

bidity compared to other types of epilepsy. This hypothesis is central to the discus-
sion of the relationship between epilepsy and psychopathology, according to the
statement of Anne Stub-Naylor, a young Danish doctor, in her thesis on Epilepsy
and psychiatric disorder a comorbidity study (1996). And she continues: Despite
the scarce evidence it appears to have become a dogma that such a relationship
exists. In contemporary medicine especially three schools have nourished this
view, represented by Elliot Slater, David Bear and Michael Trimble. Indeed, the
description of schizophrenia-like psychoses in patients with TLE (Slater et al.,
1963) has encouraged the expectation of nding a biological-based explanation
for schizophrenia. For this reason, the discussions on psychiatric disorder and espe-
cially those on psychoses in the case of epilepsy have long been dominated by the
limbic hypothesis.
It seems that the time has come to take a critical look at the above thesis, because
the understanding that other forms of epilepsy also show a clinical prole and a
biological quality is gaining ground. To have a stake in biological psychiatry, the
psychiatric aspects should be taken into consideration in regard to other forms and
syndromes of epilepsy as well.
For the following reasons, the syndromes of idiopathic generalized epilepsy
(IGE) are especially suitable for consideration. Grouped together they are at least
as frequent as the localization-related temporal and extra-temporal epilepsies.
Phenomenologically and electroencephalographically they are well dened; there-
fore they can easily be distinguished from focal epilepsies and symptomatic genera-
lized epilepsies. The coexistence of generalized and focal epilepsies is extremely
rare, although transitions from one to the other are possible due to a long and
severe course of disease, and they might well be relevant for the cause of psychiat-
ric disorder. The comparison of subsyndromes within families shows that the
various syndromes of idiopathic generalized epilepsy make up one genetic entity,
41
42 D. Janz
which diers clearly from genetic focal epilepsies in view of their phenotype (Janz,
1997) as well as from the molecular-genetic view (Sander et al., 2000).
According to the Commission on Classication and Terminology of the
International League against Epilepsy (1989), within IGE belong apart from
benign neonatal convulsions and benign myoclonic epilepsy in infancy, which
are irrelevant for our topic childhood absence epilepsy (CAE), juvenile absence
epilepsy (JAE), juvenile myoclonic epilepsy (JME) and epilepsy with grand mal sei-
zures (GTCS) on awakening (GMA). For these four syndromes of IGE of child-
hood and adolescence electroencephalographic regular or fast generalized
spike-wave discharges are the typical EEG patterns.
When scanning the literature for the frequency of psychiatric disorders in
various types of epilepsy, diculties of classifying become obvious. The earlier lit-
erature does not recognize the modern syndrome classication. Yet, in some recent
studies, especially those from the psychiatric side, the occurrence of generalized sei-
zures is equated with the supposition of generalized epilepsy (Mndez et al., 1993).
Epilepsy with exclusively grand mal seizures (TCS) can be taken for IGE, even
without EEG, if it concerns an epilepsy with grand mal predominantly on awaken-
ing (GMA). But such details are still frequently neglected, although the
International Classication has existed for 12 years.
Psychosis
Looking at the most serious, yet rarest, psychiatric disorder, the psychoses, and
paying attention to their frequency in generalized epilepsies in comparison to focal
epilepsies, most authors of the relevant literature (Table 4.1), found that psychoses
in focal epilepsies are slightly more frequent than in generalized epilepsies, but the
dierence proved to be signicant in only one study (Onuma, 1983). Two studies
using the International Classication did not nd the frequencies of psychoses in
IGE and in TLE to be dierent (Schmitz and Wolf, 1991, 1995; Sengoku et al.,
1997). With regard to specic epileptic syndromes, CAE was signicantly asso-
ciated with the presence of psychosis. CAE occurred in 7/28 (25%) of patients with
psychosis compared with 55/669 (8%) of patients without psychosis (Schmitz and
Wolf, 1991).
Relation to seizure type

With regard to specic seizure types, absences in combination with GTCS are asso-
ciated with an increased risk of psychosis. With this seizure combination psychosis
occurred in 11.3% (7/62) of patients in comparison with only 1% (1/91) of patients
with myoclonic jerks in combination with GTCS (Schmitz, 1988). This gure is in
good correspondence with ndings of Genton et al. (2000) who reported 5 (2.9%)
43 Psychiatry of idiopathic generalized epilepsy
Table 4.1. Controlled studies on the frequency of psychosis in different types of epilepsy
Studies Type of psychosis/population Results
Gastaut (1956) Mixed GE, TLE, non-TLE: not dierent

Small et al. (1962) Schizophrenia TLE, 24%; non-TLE, 12%: not
signicantly dierent
Stevens (1966) History of admission to GE, 29%; TLE, 31%; FOC non-TLE, 8%
psychiatric hospital
Mignone et al. (1970) MMPI: schizophrenia score Psychomotor vs. nonpsychomotor: not
dierent
Bruens (1971) Mixed GE, 3.3%; SIM FOC, 0%; COM FOC,
3.5%; GETLE, 19%
Standage and Fenton (1975) Present State Examination TLE vs. non-TLE: not dierent
Shukla et al. (1979) Schizophrenia GE, 4%; TLE, 17%
Onuma (1983) Paranoid symptoms PGE, 3.5%; TLE, 9.2%; non-TLE, 2.5%:
PGE vs. TLE, signicantly dierent
Sengoku et al. (1983) Mixed, epilepsy institution GE, 4.3%; TLE, 6.0%; non-TLE, 2.0%:
GE vs. TLE not signicantly dierent
Schmitz (1988) Mixed, Department of Neurology IGE, 3.5%; FOC,4.2%; GEFOC, 4%:
not signicantly dierent
Sengoku et al. (1997) Mixed, Department of Psychiatry IGE, 19.4%; TLE, 15.2%: not
signicantly dierent
Bredkjaer et al. (1998) Nonorganic Psychomotor epilepsy 3.57 SIR, petit
Nonaective psychosis mal epilepsy 2.3 SIR
Notes:
GE, generalized epilepsies; PGE, primary GE; IGE, idiopathic GE; TLE, temporal lobe epilepsies; non-TLE,
nontemporal lobe epilepsies; FOC, focal epilepsies; SIM FOC, epilepsies with simple focal seizures; COM
FOC, epilepsies with complex partial seizures; SIR, standardized incidence ratios; MMPI, Minnesota
Multiphasic Personality Inventory.
cases with psychotic disorders among 170 patients with JME. Unfortunately, we do
not know the inuence that other seizure variables, such as the length of the disease,
or frequency and intensity of the seizures, may have on these dierences between
the various IGE-syndromes.
In their retrospective studies of 697 and 611 consecutive patients with unselected
epilepsies (Wolf, 1976; Schmitz, 1988), Schmitz and Wolf (1995) found a puzzling
relation between absences and complex focal seizures regarding an increased risk
of psychosis. Contrary to this, patients with only GTCS had no increased risk of
44 D. Janz
psychosis. They speculated therefore that both seizure types testify to the ability of
a patients brain to produce epileptic events that are widespread but still limited. If
prolonged but strongly limited seizure discharge is indeed present during epileptic
psychoses of many types, as has been proposed, a brains ability to produce such a
discharge would be a prerequisite for psychosis. The brains of patients seem to have
learned the compromise in between if they cannot produce or give an epileptic all-
or-nothing response.
Ictal psychosis
The classical example of psychotic states due to epileptic events that are widespread
but still limited is the nonconvulsive status epilepticus. The symptoms which are
common to both types of nonconvulsive status, the so-called absence- (or petit
mal-) status and the focal status, are disturbance of consciousness and incomplete
or complete amnesia. This psychotic appearance of absence status is completed by
lack of responsitivity and stuporous behaviour. Absence status is by far more fre-
quent in IGE than its analogue, the complex focal status, in focal epilepsies.
Gibbs and Gibbs (1952) and Lennox (1960) reported a history of petit mal status
in 27 (2.6%) of their 1039 petit mal patients. We have seen it in 15 (5%) of our
approximately 300 patients with pyknoleptic petit mal (Janz, 1998), a percentage
which corresponds to the 5.8% of Loiseau and Cohadon (1970). Dalby (1969) has
reported petit mal status in 21 (6.2%) of his series of 346 patients with 3-Hz spike-
wave and 15 (9.3%) of the 160 patients with a history of petit mal.
Absence status which lasts for hours or days usually occurs only after adoles-
cence. As isolated events, they may even occur in middle or later age (Ja, 1962;
Lee, 1985). According to its clear clinical and electroencephalographic picture, the
almost negative interictal psychopathology, and the almost positive electric activ-
ity, the absence status can be characterized as a specic ictal psychosis of IGE.
Postictal psychosis
Furthermore, postictal and interictal psychoses are to be distinguished. Postictal
psychoses are clearly demarcated by a lucid interval from seizures themselves and
from immediate postictal confusion. The phenomenology is pleomorphic, with
uctuating combinations of delirium, delusions, hallucinations, thought disorders,
or/and aective change. Postictal psychosis is not specic for a type of epilepsy.
Dongier (1959/60) found a preponderance of generalized epilepsies in cases of post-
ictal psychosis. Others have found a higher rate among patients with localization-
related epilepsies (Logsdail and Toone, 1988; Savard et al., 1991). Devinsky et al.
(1995) noted in their case-control study of 20 patients with postictal psychoses no
signicant dierence between the percentage of patients with focal or primary gen-
eralized epilepsy in cases and controls. It seems as if long duration of epilepsy
(Kanemoto et al., 1996) and high number of lifetime GTCS (Devinsky et al., 1995)
contribute to the pathogenesis of postictal psychosis.
Interictal psychosis
Interictal psychotic episodes show a paranoid hallucinatory symptomatology.
According to some authors the delusions are less systematized than in true schizo-
phrenia (Janzarik, 1955; Tellenbach, 1965), personality deterioration is lacking
(Slater et al., 1963) and thought disorder unusual (Korzeniowski, 1965).
Unfortunately, there are no controlled data on the frequency of acute and chronic
interictal psychoses comparing IGE and TLE. According to Sengoku et al. (1997)
chronic psychoses can only be found in patients with TLE. Moreover, only in TLE
do psychoses relate to the frequency of seizures. Correspondingly, Schmitz (1992)
found that ve of six chronic psychoses occurred in patients with TLE, but she says
also this relation was not signicant due to small numbers and that psychoses in
TLE tend to be more severe than those in generalized epilepsy.
Unfortunately a comparison of symptomatology in Sengoku et al.s (1997) study
cannot be made for evaluation of interictal psychoses, as the semiological rating is
not dened. The preponderance of paranoid symptoms in TLE psychoses is sur-
prising, as is the frequent occurrence of perplexed behaviour, which is not dened
in the text. Similarly to Wolf (1982), Sengoku et al. (1997) have noted an associa-
tion between visual hallucinations and generalized epilepsies, whilst other authors
have also noticed auditory hallucinations in psychoses in TLE.
Unfortunately there is also no systematic comparison of the psychopathological
contents of interictal psychoses in both forms of epilepsy, but only impressions.
Thus Sengoku et al. (1983) wrote that with regard to the contents of psychosis, the
episodic paranoid states were most commonly found in the TLE group. In the gen-
eralized epilepsy (GE) group, however, the types of psychosis were more varied.
Wolf (1982) noted an association between olfactory hallucinations and TLE and
between visual hallucinations and generalized epilepsies, suggesting that the form
of the psychosis is inuenced by symptoms of the epileptic disorder.
Alternative psychosis with forced normalization of the EEG

Some of the interictal psychoses according to Schmitz (1988) 11%, to Wolf (1985)
15%, to Dongier (1959/60) 25% belong to the group of psychiatric disorders with
so-called forced normalization of the EEG (Landolt, 1955, 1958, 1963). With this
term Landolt described the condition that with spontaneous or, more frequently,
drug-induced disappearance of clinical and subclinical seizure manifestations,
psychiatric symptoms appear. These may be (Table 4.2) prepsychotic dysphoria
characterized by insomnia, anxiety, feelings of oppression and social withdrawal or
may proceed to psychotic symptoms characterized by hallucinatory paranoid states.
46 D. Janz
Table 4.2. Psychiatric syndromes in 44 cases of forced

normalization observed in 36 patients
Paranoid hallucinatory psychosis 19

Prepsychotic dysphoria 9
Hysterical episode 5
Hypochondrial episode 3
Depressive episode 2
Manic episode 2
Twilight state 1
Depersonalization 1
Source: Wolf (1984).
The phenomenon of an inverse relationship between well-being on one hand

and seizure and specic discharges in the EEG on the other occurs with both gen-
eralized and focal epilepsy. Landolts rst observations were with temporal lobe
epilepsy (1955). He felt himself supported by Gibbs (1951), who from his experi-
ence with Phenurone drew the conclusion that the epileptic and psychiatric com-
ponents of psychomotor epilepsy are physiologically antithetic.
Later, with the advent of the succinimides, the focus turned towards the gener-
alized epilepsies. The fundamental study of Tellenbach (1965) gives a detailed
description of as he called it alternative psychoses of a paranoid type in 12 cases,
9 with IGE, in 4 cases in combination with pyknoleptic (childhood) absences.
According to Wolf and Trimble (1985) alternative psychoses occurred three times
more frequently in IGE than in epilepsies with complex focal seizures and in IGE
have been the predominant type of psychoses. Nearly the same relationship holds
true in the series of Schmitz (1988) (Table 4.3).
In a subgroup patients with IGE with absences beyond childhood it may be
seen in as many as 8% of patients, depending on the medication given (Wolf, 1986).
According to Wolf s opinion (1986), in this group we seem to know some details of
the pathogenesis. He mentioned in this respect drug inuences, as for instance the
sleep deprivation and increase of light sleep with ethosuximide, and social status,
particularly regarding the frequent vocational disintegration. He points to the
observations of Tellenbach (1965) of the frequent familial conict situations, and
on the patients personality being aected by having to cope with disabling drug
eects. Tellenbach has argued that patients with awakening epilepsy are particularly
sensitive to a drug-induced process of slowing down, the feeling of being con-
strained, of being immured because of their personality trait of extraversion.
Table 4.3. Psychosis and related syndromes in patients of a neurological outpatient clinic (N 611)
Generalized Focal Epilepsies with Unclassied

epilepsies epilepsies generalized and focal epilepsies
Syndrome groups (n237) (n270) features (n46) (n58)
Ictal and peri-ictal 3 9 2 1

Forced normalization 4 1
Intoxicated 1 2
Drug withdrawal 1
Unrelated to epilepsy 3 4 1
Multifactorial and unknown 2 1 1 1
Source: Schmitz (1988).
Anything which threatens their stability will immediately and recklessly be

expressed. In the number of conditions which in the case of forced normalization
of the EEG (forcierter Normalisierung des EEG) leads to psychiatric disorders,
Tellenbach includes special personality traits of patients with awakening epilepsy.
What did the scholars understand by this in 1965, and what do we know about it
currently?
Personality
The heart of the matter

When, in the beginning of the 1950s, the clinical picture of awakening epilepsy
became clear to me, including that of all epilepsies with seizures mainly after awak-
ening, i.e. todays IGE, I became more and more interested in the conditions which
led to the triggering of seizures. According to Niedermeyer (1996), in IGE the role
of arousal is the heart of the matter. The period after awakening, at whatever time
of awakening, is the most critical biological situation in any variety of IGE, both
for the appearing of generalized spike wave-discharges and for the manifestation of
the generalized tonic-clonic seizures, the absences and the myoclonic jerks.
Therefore the early-chosen term epilepsy on awakening (Janz, 1953, 1962) focuses
the attention to the crucial point which is pathophysiologically common to all
forms of IGE.
It has been noted that all seizures in IGE were frequently precipitated by exter-
nal factors such as emotional and physical stress especially in association with lack
of sleep. These factors seem to have little inuence on seizures in other forms of
epilepsy, as shown in comparison with sleep epilepsy (SE) (Table 4.4).
48 D. Janz
Table 4.4. Seizure precipitants in sleep epilepsy (SE) (n 127) and in awakening epilepsy
(AE) (n 90)
First seizures (%) Following seizures (%)
Precipitant SE AE SE AE
Sleep deprivation 5 25 6 42
Alcohol use 1 10 3 23
Abrupt wakening 12 28
Strong emotion 2
Physical stress 3 9 4 13
Menarche 2* 6*
Note:
* Female patients only.
Source: Janz (1953).
In JME in particular, we have reported unusual lack of sleep and sudden awak-
ening, and/or excessive alcohol consumption, as the most common precipitating
factors (Janz and Christian, 1957). Other triggering factors were less common.
Pedersen and Petersen (1998) registered, although retrospectively, alcohol con-
sumption in 51%, stress in 70% and sleep deprivation in 84% of patients as the
most frequent precipitating factors.
Lifestyle
Most patients are not in a position to sleep as much as they would like to every day;
they are forced to get up earlier than they would like to, and therefore suer from
a chronic sleep decit. Additional stress forces them to pass their limits, and sei-
zures result.
Everyone who knows patients with epilepsy on awakening can produce numer-
ous examples of seizures occurring after late nights spent in preparing for exam-
inations, attending parties, watching television or preparing to leave early for
holidays. Some of the causes are so extreme that one is tempted to say that it is no
wonder that seizures occur. However, the patients describe the events as unavoida-
ble, and one tends to pardon these excesses at rst. One becomes suspicious when
seizures continue to occur after repetitions of the same or similar events. It is sur-
prising that the patients often do not avoid the situations which provoke seizures,
and it is not clear why the patients are unable to learn from experience. One gets
the impression that the patients are unable to draw conclusions from their own
negative experiences and to change their lifestyles, and understands Niedermeyers
Table 4.5. Modes of falling asleep and awakening in patients with awakening epilepsy
(AE) and in patients with sleep epilepsy (SE)
AE SE
n % n %
Falling asleep Quick, unmolested 19 40 36 86

Delayed, dicult 28 60 6 14
Awakening Quick, unmolested 11 18 32 68

Delayed, dicult 50 82 15 32
Source: Janz (1953).
(1996) moaning: Unfortunately, some patients are incorrigible and all exhorta-
tions fall on deaf ears.
Sleeping habits
Of all the precipitating factors, sleep deprivation is certainly the most important, and
it would appear that other factors become eective in the presence of sleep depriva-
tion. The question arises what is the reason for this relation: is it a faulty behaviour
of the patients, is it a particular biological weakness or probably both? Patients with
IGE, particularly those with JME, tend to awake slowly and with diculty and remain
sleepy for some time afterwards (Table 4.5) (Janz and Christian, 1957; Janz, 1985,
2000). Nearly every patient can report that he or she is dicult to arouse, and would
love to stay in bed longer, that sleep still sits in their limbs, that they initially act
automatically, like in a dream, or are numb. Many have adjusted to this by sleep-
ing longer and having breakfast in bed. Others who cannot aord this, dunk their
heads in cold water or drink strong coee. All emphasize that they need much sleep
and most claim that they sleep very deeply. However, in our experience, these patients
do not admit or do not recognize that they often retire too late and that, therefore,
they often do not get enough sleep, i.e. that they easily suer from lack of sleep.
Interviews by questionnaire which have recently been carried out in two loca-
tions in Berlin indicate that there is a signicant dierence in well-being during
day-time between the patients with JME and those with TLE (Pung, 2000) (Table
4.6). JME patients feel less ecient in the morning than TLE patients. The fact that
perception of well-being in the evening met only partly but not signicantly with
expectations, might depend on a lack of acceptance or on an adaptation to social
conventions. It is useful to complete the interview of the patients together with
their near relatives or friends, because JME patients particularly tend to adapt their
answers to conventional attitudes.
50 D. Janz
Table 4.6. Answers of 20 patients with juvenile myoclonic epilepsy (JME) and 20 patients
with temporal lobe epilepsy (TLE)
Question Time period JME TLE
What time of the day do you feel at top?

1621.00 h 3 1
1016.00 h 16 12
810.00 h 1 6
58.00 h 1
Are you more a morning type or an evening type?

Morning 7 14
Evening 13 6
Source: Pung (2000).
Personality traits in patients with epilepsy on awakening

Phenomenology
Patients with IGE, especially those with JME, seem to lack a certain degree of self-
control. This is evident in the fact that they seem to neglect their physical needs and
their health. They are quite likable. They are not hypochondriacs, like many
patients with temporal lobe epilepsy, and do not behave hysterically, but it is impor-
tant to recognize that these patients fail to avoid situations which might provoke
seizures, and also forget to take their medications or even discontinue treatment.
This lack of self-control goes with a tendency to deny problems and conicts, and
besides a shortlasting depression postictally they tend to take seizures, symptoms
and complaints lightly, some even believing that they can avoid the disease by not
taking their tablets. These patients, particularly if teenagers, often become obsti-
nate, claiming something to the eect that I would not consider going to bed at
regular hours, abstaining from alcohol and living like a monk. I am not a child
anymore. If I cant do what I want, then I might as well hang myself . This is an
expression of an immature and childish attitude which must be met with compre-
hension and rmness rather than authoritarian exhortations.
Another characteristic that one often encounters in patients with IGE is their
impressionability. They comprehend and judge quickly, and they are versatile and
adaptable as well as impressionable, which facilitates contact with the physician but
endangers the continuity of the therapeutic relationship. Characteristics such as
impressionability, openness and awareness to the point of being easily distracted
make these patients appear alert and intelligent, but cause them problems when
tasks requiring resoluteness, patience and perseverance are concerned.
Patients with IGE are somewhat unstable and tend to lose their emotional
balance easily. They may be moody, though these phases do not last long. Lack of
self-condence leads them to underestimate themselves though overcondence
may also be observed.
Since my rst description (1953), I have pointed out repeatedly (1962, 1969,
1974, 1989, 1998) that patients with this epileptic syndrome are, as a rule, very
dierent from what had long been considered the typical personality of genuine
epileptics that is, slow, fussy, viscous, circumstantial and irritable, utterly pedan-
tic and obstinate, prone to hypochondria and practically incapable of changing the
subject of conversation. Some of these traits are now grouped together as the
Geschwind syndrome (Benson, 1991; Blumer, 1999) and associated with TLE.
Patients with IGE are inclined to the opposite more unstable, less reliable, incon-
siderate and neglectful of duties and self-interest, more ready to yield to the slight-
est temptation even against better judgement, insensitive to future consequences.
They behave sometimes, as Tellenbach put it, like an adult child or as Pellock
(1999) described JME patients, as perpetual adolescents.
Investigations
Peter Wolf (1985) wrote
The clinical observation of Janz was followed by a psychological investigation of Leder (1967),
who studied a group of 34 patients with GMA and 55 with GMS. From these, two groups of 10
persons each were selected for statistical evaluation. These groups were matched for sex, age,
seizure frequency, association with minor seizures, intelligence, absence of etiological clues and
social status. They were investigated with two personality tests, the Rorschach and Szondi tests,
and several signicant dierences between the groups were found. Their interpretation describes
patients with GMA as extroverts who have diculty in recognizing the limits of their own person
in relation to the external world. Often they have little ability to suppress, to contradict and to
renounce conicts; tensions and disinclinations are usually momentarily disposed often by
denial. These patients will follow simultaneously the most divergent and irreconcilable aims
without being aware of any diculties.
Wolf remarks that the Leder investigation, which has been conrmed by similar
studies in Japan (Aoki and Kawai, 1982; Kawai and Aoki, 1983; Yamashita et al.,
1984), has been criticized for its methodology, since many psychologists have res-
ervations about Rorschach and Szondi tests. However, his description will remind
unprejudiced readers of many of their own patients with IGE whose lack of disci-
pline is often an obstacle to successful therapy. Their unstable sleep behaviour may
precipitate seizures, and sleep withdrawal is often the cause of the rst seizure.
Therefore, it seems possible that unstable personalities of this kind are common in
GMA because such a personality is a factor which favours the manifestation of a pre-
disposition to seizures which may often remain latent in people with more regular
52 D. Janz
lifestyles. It is the opinion of Peter Wolf that comparative psychological investiga-

tions of patients and predisposed but unaected relatives could clarify this question.
Personality in patients with pyknolepsy (childhood absence epilepsy)

Phenomenology
In the case of two subsyndromes of IGE, namely in pyknolepsy or CAE and in JME,
the description of personality traits has its own tradition. Before the era of EEG, the
observation that children with pyknolepsy are usually lively to over-lively, mentally
active and markedly intelligent (lebhaft bis berlebhaft, geistig regsam und ausgespro-
chen intelligent; Rosenthal 1935), has often served as an argument that due to this
unepileptic behaviour, pyknolepsy is not an epileptic disease. But even later, when
the EEG had given an explanation for the true nature of the absences, the friendli-
ness and brightness of the children was emphasized, and they were often called
model children (Bridge, 1949), even though their positive traits could not hide the
critical ones. Bridge, for example, noted that children with pyknolepsy are over-
lively, never stop in order to relax and show no sign of fatigue when playing with
other children. At home and at school, they are very keen to full their duties due to
their high sensitivity to criticism and scolding. As a consequence, these children
strain their energies excessively. I have myself pointed out that often, in the course of
their further development, the parents expectations are disappointed (Janz, 1955,
1998). The increasing demands at school during their education reveal that the chil-
drens liveliness and cleverness hide a lack of profoundness and perseverance. The
performance at school decreases. The parents complain that their children do not
perform any better than average pupils even though they used to be among the best,
and they explain this as follows: despite their intelligence they are too light-headed,
they are not attentive and they have too many other things in their heads. What
once seemed like keen interest and brightness, then reveals itself to be a lack of con-
centration and carelessness, the liveliness is nervousness, and the quick perception a
lack of profoundness. This metamorphosis from the intellectual miraculous child to
the average pupil or even a loser reects a frequent misjudgement by the adults.
These observations of the psychological characteristics and the development of chil-
dren with pyknolepsy or childhood absence epilepsy show an important proximity
to the generally described psychological structure of awakening epilepsy.
Investigations
For this reason the experimentalpsychological investigations of absence epilepsy
which Mirsky and his group have pursued since the 1960s are of special interest.
They have found in absence epilepsy decits in sustained auditory and visual atten-
tion that are distinct from those seen in other seizure disorders (Mirsky et al., 1995).
The event-related potential studies indicate that an impaired sensory input, seen
1.0
0.5
0.0 Normal controls

(n 76)
Standard scores
0.5
Absence seizures
(n 76)
1.0
1.5 Complex partial seizures

(n 7)
2.0
2.5
3.0
Shift Focus/Execute Sustain Encode
Figure 4.1. Component scores extracted by principal components analysis of attention tests on
normal control subjects (white columns), patients with absence seizures (black columns)
and complex partial seizures (cross-hatched columns). The asterisks appearing below the
columns indicate that the group so designated differed significantly from the other two
groups on that component. The absence group performed significantly worse than the
other two groups on the sustain component, whereas the complex partial group was
impaired, in comparison with the other groups, on the shift and focus/execute
components. (Reproduced with permission from Mirsky et al., 1991.)
primarily in the auditory modality, not only occurs during spike-wave bursts, but
during interictal periods as well. The clinical impression of increased distractibil-
ity in absence patients is psychologically explained by a signicant score on a test
assessing the sustaining function of attention, whereas other components of the
complex process of attention like shift or focus/execute are signicantly more
impaired in patients with complex focal seizures (Mirsky et al., 1991) (Figure 4.1).
Levav (1991) has investigated the question posed by Wolf (1985), as to whether
the manifestations of IGE are associated with personality traits, that healthy rela-
tives do not show. She investigated the attention performance of children aected
with absence epilepsy and that of their rst-degree relatives. She found signicant
impairment in tests of sustained attention in comparison with unaected siblings;
moreover, particularly in female probands. That mothers tended to perform more
poorly than fathers, which is in accordance with the known maternally transmit-
ted inuence on seizure susceptibility (Ottmann et al., 1988) gave rise to the
assumption that measures of attention may serve as markers of vulnerability to the
disorder (Mirsky et al., 1995). This hypothesis needs investigating in relation to
subclinical seizure activity.
54 D. Janz
Personality traits in patients with juvenile myoclonic epilepsy

Phenomenology
With regard to the behavioural phenomenology of patients with JME, we described
the observations in our rst article (1957) as follows:
Their psychological style as was described for awakening epileptics is very frequently character-
ized by unsteadiness, lack of discipline, hedonism and indierence towards their disease, in con-
trast to typical epileptic conduct. Since patients with impulsive petit mal (nowadays called JME)
behave rather similar we feel that it is possible to establish a typology. Only one patient was men-
tally retarded. Most were intellectually average, and none was particularly gifted. But since they
were all quick to learn and judge, exible and adaptable, school and professional or occupational
training were easy for them. But they promise more than they deliver. Of 19 men whose occupa-
tional training was known to us in some detail, ten descended to the level of unskilled laborers after
they had been employees with business training, craftsmen, skilled workers and students. Although
the psychological and social eects of their disease itself were certainly responsible for this to some
extent, especially since the onset of the condition occurs very often during the period of occupa-
tional training, insucient motivation and endurance remain possible explanations for this nega-
tive development. The conduct of these patients often has an unfavorable inuence on therapy.
Although the patients are ready to swear that they have done everything that they were instructed
to do, they frequently fail to appear at follow-up visits or to take their medications regularly. Many
handle themselves with great assurance and demanding, but they may also be decidedly mistrust-
ful and shy, fearful and inhibited. Their labile feeling of self worth also leads them to be both eager
to help, to invite, to give, on the one hand and to be able to react in an exaggeratedly sensitive way
on the other hand. Their mood changes rapidly and frequently. This makes their contact both
charming and dicult. They are easy to encourage and discourage, they are gullible and unreliable.
Their suggestibility makes contacts easy but makes trust dicult. This personality prole plays
along a scale from a likable nonchalance or timidity, through a psychoasthenic syndrome to the
extremes represented by sensitive or reckless psychopathy behaviour (Janz and Christian, 1957).
Investigations
These impressions, gained only by observations, were completed in 1976 by
psychological investigations in Denmark (Bech et al., 1976, 1977; Lund et al., 1976;
Reintoft et al., 1976; Simonsen et al., 1976). In comparison to patients with idio-
pathic grand mal epilepsy no dierence could be found concerning social back-
ground, intelligence and education. However, patients with JME more often than
controls tended to experience a social decline, diculties in nding contacts and a
feeling of being discriminated against. Signs of character-neurosis, that were only
diagnosed in patients with JME, were associated to the psychological test ndings
of substability and subvalidity. These results have been integrated as signs for
instability and psychasthenie.
Since the introduction of JME to the Anglo-Saxon scientic orbit through
Asconap and Penry (1984) and Delgado-Escueta and Enrile-Bascal (1984), many
papers on JME have appeared, but comments on the psychological prole have
remained scarce and controversial. Psychiatric disorders in JME have been reported
in comparison with TLE to be relatively high (Vasquez et al., 1993) or relatively low
but higher than in controls (Perini et al., 1996). According to the latter authors, who
performed psychological tests, JME patients only showed one signicant dierence
from controls, by having high scores on a trait anxiety scale.
At the occasion of the 1989 meeting of the American Epilepsy Association in
Boston, I was asked by Kin Penry to outline before the young epileptologists the
personality prole of patients with JME from my view as a clinician.
At the end of my speech I raised two questions to the audience:
1. Is it possible to verify this behavioural pattern with psychological tests?
2. Is it possible to correlate this behavioural pattern with discrete brain structures?
And I cautiously gave the following answer:
Some of these characteristics such as limited rational self-control, suggestibility, distractibility

and indierence to physical needs suggests involvement of frontal regions of the brain. This con-
clusion is compatible with the fact of accentuated bilateral frontal or fronto-central spike-wave
activity in this type of epilepsy. Also morphological ndings of bilateral cortical microdysgene-
sis which is probably more pronounced in the frontal regions (Meencke, 1985; Meencke and Janz,
1984, 1985) correlate with the described psychological syndrome, which therefore one might des-
ignate as a mild frontal lobe syndrome.
Since then two groups have studied this correlation with specic frontal psycholog-
ical tests.
Swartz et al. (1994) examined memory function in 9 patients with JME, 15
patients with frontal lobe epilepsy (FLE) and 15 controls. The key test was the
delayed match to sample, which they refer to as a measure of a primary or working
memory. The JME group performed intermediate between the controls and the
patients with clear frontal pathology. The authors explained their ndings as
related to impairment in selective attention. In a second study with identical tasks
they performed a FDG positron emission tomography (PET) comparing JME
patients and normal controls (Swartz et al., 1996). Controls showed activation in
the dorso-lateral prefrontal area while in JME patients there was a decrease in this
area. Instead, JME patients showed increased activations in the brain stem, the
medio-temporal region and the lateral orbital cortex. The authors explained this as
an attempt to compensate for the failure of activation in the orbito-frontal region.
Devinsky et al. (1997) compared the scores of a number of frontal tasks taken
from 16 patients with JME with those from 15 patients with TLE, matched for IQ.
The results showed evidence for a pattern of impaired frontal functioning in
patients with JME. This was signicantly dierent from the frontal functioning in
patients with TLE. It is the higher order cognitive functions of planning, reasoning
56 D. Janz
and exibility of thought that are compromised, and to lesser extent, the ability to
focus attention and inhibit habitual but inappropriate responses (Devinsky et al.,
1997).
Taken together, these data do suggest subtle but nonetheless real frontal impair-
ments of psychological functions in people with JME (Trimble, 2000).
This raises some questions:
Are there essential dierences in the personality traits of dierent subsyndromes of
IGE? According to the information we have up to now, the dierences seem to
be less important than compared with focal epilepsies, especially TLE.
If we assume certain psychological characteristics, what is then the common basic
disorder?
Is it a disturbance of the function of attention which is noted in the sustained atten-
tion test and in the visual working memory test ?
What is the relation between the assumed psychological basic disturbance and the
circadian sleepwaking cycle and maybe even the ultradian restactivity changes?
And nally, what is the relation between the assumed disturbance of a cognitive
performance that of continuous attention and a biological performance
that of a rhythmic change of rest and activity in view of the development of
those two performances during adolescence?
I think that such questions help best to prevent the aected from being stigma-
tized. If we see that these patients often behave like adult children (Tellenbach,
1965), perpetual adolescents (Pellock, 1999), or better, in some way like still ado-
lescents, then we can nd ways to understand their weaknesses and to help them
eectively.
R E F E R E N C ES
Aoki, K. and Kawai, I. (1982). Awake epilepsy and sleep epilepsy: a psychological study with
Rohrschach test. Rohrschach Res, 24, 10117.
Asconap, J. and Penry, J.K. (1984). Some clinical and EEG aspects of benign juvenile myoclonic
epilepsy. Epilepsia, 25, 10814.
Bech, P., Kjaersgard Pedersen, K.K., Simonsen, N. and Lund, M. (1977). Personality traits in epi-
lepsy. In Epilepsy, the Eighth International Symposium, ed. J.K. Penry, pp. 257263. New York:
Raven Press.
Bech, P., Kjaersgard Pedersen, K., Simonsen, N. and Lund, M. (1976). Personality traits in epi-
lepsy. A multidimensional study of personal traits ad modum Sjbring. Acta Neurol Scand, 54,
34858.
Benson, D.F. (1991). The Geschwind syndrome. In Advances in Neurology, ed. D. Smith, D.
Treiman and M. Trimble, Vol. 55, pp. 42141. New York: Raven Press.
Blumer, D. (1999). Evidence supporting the temporal lobe epilepsy personality syndrome.
Neurology, 53 (Suppl. 2), S912.
Bredkjaer, S.R., Mortensen, P.B. and Parnas, J. (1998). Epilepsy and non-organic non-aective
psychosis. National epidemiological study. Br J Psychiatry, 172, 2358.
Bridge, E.M. (1949). Epilepsy and Convulsive Disorders in Children. New York: McGraw-Hill.
Bruens, H.J. (1971). Psychoses in epilepsy. Psychiatr Neurol Neurochir, 17492.
Commission on Classication and Terminology of the International League Against Epilepsy
(1989). Proposal for revised classication of epilepsies and epileptic syndromes. Epilepsia, 30,
38999.
Dalby, M.A. (1969). Epilepsy and 3 per second spike and wave rhythms; a clinical, electroen-
cephalographic, and prognostic analysis of 346 patients. Acta Neurol Scand, 45 (Suppl. 40),
1183.
Delgado-Escueta, A.V. and Enrile-Bascal, F. (1984). Juvenile myoclonic epilepsy of Janz.
Neurology, 34, 28594.
Devinsky, O., Abramson, H., Alper, K. et al. (1995). Postictal psychosis: a case control series of 20
patients and 150 controls. Epilepsy Res, 20, 24753.
Devinsky, O., Gershengorn, J., Brown, E., Perrine, K., Vasquez, B. and Luciano, D. (1997). Frontal
functions in juvenile myoclonic epilepsy. Neurol Neuropsychol Behav Neurol, 10, 2436.
Dongier, S. (1959/60). Statistical study of clinical and electroencephalographic manifestations
of 536 psychotic episodes occurring in 516 epileptics between clinical seizures. Epilepsia, 1,
11742.
Gastaut, H. (1956). Colloque de Marseille, 1519 Octobre 1956. Compte rendu du colloque sur
ltude lectroniques en dehors des crises cliniques. Revue Neurol, 95, 587616.
Genton, P., Glisse, P. and Thomas, P. (2000). Juvenile myoclonic epilepsy today: current deni-
tion and limits. In Juvenile Myoclonic Epilepsy: The Janz Syndrome, ed. B. Schmitz and T.
Sander, pp. 1132. Peterseld, UK and Philadelphia, USA: Wrightson Biomedical Publishing.
Gibbs, F.A. (1951). Ictal and non-ictal psychiatric disorders in temporal lobe epilepsy. J Nerv
Ment Dis, 113, 5228.
Gibbs, F.A. and Gibbs, E.L. (1952). Atlas of Electroencephalography. Cambridge, MA: Addison-
Wesley.
Ja, R. (1962). Ictal behaviour disturbance as the only manifestations disorder: Case report. J
Nerv Ment Dis, 135, 4705.
Janz, D. (1953). Aufwach-epilepsien (als ausdruck einer den nacht- oder schlaf epilepsien
gegenberstehenden verlaufsform epileptischer erkrankungen. Arch Psychiat Nervenkr, 191,
7398.
Janz, D. (1955). Die klinische Stellung der Pyknolepsie. Dtsch med Wchschr, 80, 13924,
1399400.
Janz, D. (1962). The grand mal epilepsies and the sleepingwaking cycle. Epilepsia, 3, 69109.
Janz, D. (1969). Die Epilepsien. Spezielle Pathologie und Therapie. Stuttgart: Thieme (1998), 2. unv
Au.
Janz, D. (1985). Epilepsy with impulsive petit mal (juvenile myoclonic epilepsy). Acta Neurol
Scand, 72, 44959.
58 D. Janz
Janz, D. (1989). Juvenile myoclonic epilepsy. Epilepsy with impulsive petit mal. Cleveland Clin J
Med, 56 (Suppl.), 52333.
Janz, D. (1997). The idiopathic generalized epilepsies of adolescence with childhood and juvenile
age of onset. Epilepsia, 38, 411.
Janz, D. (2000). Epilepsy with grand mal on awakening and sleep-waking cycle. Clin Neuro-
physiol, 111 (Suppl. 2), S10310.
Janz, D. and Christian, W. (1957). Inpulsiv-Petit mal. Dtsch Z Nervenheilk (J Neurol), 176,
34686. English (1994) in Idiopathic Generalized Epilepsies, ed. A. Malafosse, P. Genton, E.
Hirsch, C. Marescaux, D. Broglin and R. Bernasconi, pp. 22951. London: John Libbey.
Janzarik, W. (1955). Der wahn schizophrener prgung in den psychotischen episoden der epilep-
tiker und die schizophrene wahnwahrnehmung. Fortschr Neurol Psychiat, 23, 53346.
Kanemoto, K., Kawasaki, J. and Kawai, I. (1996). Postictal psychosis: a comparison with acute
interictal and chronic psychoses. Epilepsia, 37, 5516.
Kawai, I. and Aoki, K. (1983). Primary generalized epilepsy and temporal lobe epilepsy: a psycho-
logical study using Rohrschach test. Folia Psychiat Neurol Jap, 37, 24551.
Korzeniowski, L. (1965). Les problemes diagnostiques concernant les psychoses paranoiaques
schizophreniformes en pilepsie. Ann Md-psychol, 123, 3542.
Landolt, H. (1955). ber verstimmungen, dmmerzustnde und schizophrene zustandsbilder
bei epilepsie. Ergebnisse klinischer und elektroencephalographischer untersuchungen.
Schweiz Arch Neurol Psychiatr, 76, 31321.
Landolt, H. (1958). Serial electroencephalographic investigations during psychotic episodes in
epileptic patients and during schizophrenic attacks. In Lectures on Epilepsy, ed. A.M. Lorentz
de Haas. Folia Psychiatr Neurol Neurochir, (Suppl. 4), pp. 91133. Amsterdam: Elsevier.
Landolt, H. (1963). Die dmmer- und verstimmungszustnde bei epilepsie und ihre elektroen-
cephalographie. Dtsch Z Nervenheilk, 185, 41130.
Leder, A. (1967). Zur psychopathologie der schlaf- und aufwachepilepsie (eine psychodiagnos-
tische untersuchung). Nervenarzt, 38, 43442.
Lee, S.I. (1985). Nonconvulsive status epilepticus: Ictal confusion in later life. Arch Neurol, 42,
77881.
Lennox, W.G. (1960). Epilepsy and Related Disorders. Boston, MA: Little & Brown.
Levav, M.L. (1991). Attention performance in children aected with absence epilepsy and their
rst-degree relatives. Doctoral dissertation. University of Maryland, College Park.
Logsdail, S.J. and Toone, B.K. (1988). Postictal psychosis: a clinical and phenomenological
description. Br J Psychiatry, 152, 24652.
Loiseau, P. and Cohadon, F. (1970). Le Petit Mal et ses Frontires. Paris: Masson.
Lund, M., Reintoft, H. and Simonsen, N. (1976). Eine kontrollierte soziologische und psycholo-
gische untersuchung von patienten mit juveniler myoklonischer epilepsie. Nervenarzt, 47,
70812.
Meencke, H.-J. (1985). Neuron density in the molecular layer of the frontal cortex in primary
generalized epilepsy. Epilepsia, 26, 4504.
Meencke, H.-J. and Janz, D. (1984). The signicance of microdysgenesis in primary generalized
epilepsies: a study of eight cases. Epilepsia, 25, 821.
Meencke, H.-J. and Janz, D. (1985). The signicance of microdysgenesis in primary generalized
epilepsy: an answer to the considerations of Lyon and Gastaut. Epilepsia, 26, 36871.
Mndez, M.F., Doss, R.C., Taylor, J.L. and Arguello, R. (1993). Relationship of seizure variables
to personality disorders in epilepsy. J Neuropsychiatry Clin Neurosci, 5, 2836.
Mignone, R.J., Donelly, E.F. and Sadowsky, D. (1970). Psychological and neurological compari-
son of psychomotor epileptic patients. Epilepsia, 11, 34559.
Mirsky, A.F., Anthony, B.J., Duncan, C.C., Ahearn, M.B. and Kellam, S.G. (1991). Analysis of the
elements of attention: a neuropsychological approach. Neuropsychol Rev, 2, 10945.
Mirsky, A.F., Duncan, C.C. and Levav, M.L. (1995). Neuropsychological and psychophysiologi-
cal aspects of absence epilepsy. In Typical Absences and Related Syndromes, ed. J.S. Duncan and
C.P. Panayiotopoulos, pp. 11219. London: Churchill Communications.
Niedermeyer, E. (1996). Primary (idiopathic) generalized epilepsy and underlying mechanisms.
Clin Electroencephalogr, 27, 121.
Onuma, T. (1983). Limbic lobe epilepsy with paranoid symptoms: Analysis of clinical features
and psychological tests. Folia Psychiat Neurol Jpn, 37, 2537.
Ottmann, R., Annegers, J.F., Hauser, W.A. and Kurland, L.T. (1988). Higher risk of seizures in
ospring of mothers than of fathers with epilepsy. Am J Hum Genet, 43, 25764,
Pedersen, S.B. and Petersen, K.A. (1998). Juvenile myoclonic epilepsy: Clinical and EEG features.
Acta Neurol Scand, 97, 1603.
Pellock, J.M. (1999). JME: The elusive epilepsy. Perspect Pediat Neurol, 1, 16.
Perini, G.I., Tosin, C., Carraro, C. et al. (1996). Interictal mood and personality disorders in tem-
poral lobe epilepsy and juvenile myoclonic epilepsy. J Neurol Neurosurg Psychiatry, 61, 6015.
Pung, T. (2002). Schlaf-wachrhythmus und persnlichkeitbei juveniler myoklonischer epilepsie
und patienten mit temporaler epilepsie. Thesis, Humboldt Universitt Berlin.
Reintoft, H., Simonsen, N. and Lund, M. (1976). A controlled sociological study of juvenile myo-
clonic epilepsy. In Epileptology, ed. D. Janz, pp. 4850. Stuttgart: Thieme.
Rosenthal, C. (1935). Die gehuften kleinen Anflle des Kindesalters (Pyknolepsie). Ergebn inn
Med Kinderheilk, 48, 77124.
Sander, T., Schulz, H., Saar, K. et al. (2000). Genome search for susceptibility loci of common
idiopathic generalised epilepsies. Hum Mol Genet, 9, 146572.
Savard, G., Andermann, F., Olivier, A. and Remillard, G.M. (1991). Postictal psychosis after
partial complex seizures: a multiple case study. Epilepsia, 32, 22531.
Schmitz, B. (1988). Psychosen bei epilepsie. Eine epidemiologische untersuchung. Thesis, Freie
Universitt, Berlin.
Schmitz, B. (1992). Psychosis and epilepsy: the link to the temporal lobe. In The Temporal Lobes
and the Limbic System, ed. M.R. Trimble and T.G. Bolwig, pp. 14967. Peterseld, UK and
Bristol, PA, USA: Wrightson Biomedical Publishing.
Schmitz, B. and Wolf, P. (1991). Psychosis in epilepsy. In Epilepsy and Behaviour. Frontiers of
Clinical Neuroscience, ed. O. Devinsky and W.H. Theodore, Vol. 12, pp. 97128. New York:
Wiley-Liss.
Schmitz, B. and Wolf, P. (1995). Psychosis in epilepsy: frequency and risk factors. J Epilepsy, 8,
295305.
60 D. Janz
Sengoku, A., Yagi, K., Seino, M. and Wada, T. (1983). Risk of occurrence and psychosis in rela-
tion to the types of epilepsies and epileptic seizures. Folia Psychiat Neurol Jpn, 37, 2216.
Sengoku, A., Toichi, M. and Murai, T. (1997). Comparison of psychotic states in patients with
idiopathic generalized epilepsy and temporal lobe epilepsy. Epilepsia, 38 (Suppl. 6), 225.
Shukla, G.D., Srivastava, O.N., Katiyar, B.C. et al. (1979). Psychiatric manifestations in temporal
lobe epilepsy: A controlled study. Br J Psychiatry, 135, 41117.
Simonsen, N., Mollgaard, V. and Lund, M. (1976). A controlled clinical and electroencephalo-
graphical study of myoclonic epilepsy (Impulsiv-Petit-Mal): preliminary report. In
Epileptology, ed. D. Janz, pp. 418. Stuttgart: Thieme.
Slater, E., Beard, A.W. and Glithero, E. (1963). The schizophrenia-like psychoses of epilepsy. Br J
Small, J.G., Milstein, V. and Stevens, J.R. (1962). Are psychomotor epilepsies dierent? Arch
Neurol, 7, 18794.
Standage, K.F. and Fenton, G.W. (1975). Psychiatric symptom proles of patients with epilepsy:
a controlled investigation. Psychol Med, 5, 15260.
Stevens, J.R. (1966). Psychiatric implications of psychomotor epilepsy. Arch Gen Psychiatry, 14,
46171.
Stub-Naylor, A. (1996). Epilepsy and Psychiatric Disorder A Comorbidity Study. Copenhagen:
FADL Publishers.
Swartz, B.E., Halgren, E., Simpkins, F. and Syndulko, K. (1994). Primary memory in patients with
frontal and primary generalized epilepsy. J Epilepsy, 7, 23241.
Swartz, B.E., Simpkins, F., Halgren, E. et al. (1996). Visual working memory in primary general-
ized epilepsy. An FDG-PET study. Neurology, 47, 120312.
Tellenbach, H. (1965). Epilepsie als anfallsleiden und als psychose. ber alternative psychosen
paranoider prgung bei forcierter normalisierung (LANDOLT) des elektroencephalo-
gramms epileptischer. Nervenarzt, 36, 190202.
Trimble, M. (2000). Cognitive and personality proles in patients with juvenile myoclonic epi-
lepsy. In Juvenile Myoclonic Epilepsy: The Janz Syndrome, ed. B. Schmitz and T. Sander, pp.
1019. Peterseld, UK and Philadelphia, USA: Wrightson Biomedical Publishing.
Vazquez, B., Devinsky, O., Luciano, D., Alper, K. and Perrine, K. (1993). Juvenile myoclonic epi-
lepsy: clinical features and factors related to misdiagnosis. J Epilepsy, 6, 2338.
Wolf, P. (1976). Psychosen bei Epilepsie. Ihre Bedingungen und Wechselbeziehungen zu Anfllen.
Habilitationsschrift. Berlin: Freie Universitt.
Wolf, P. (1982). Halluzinationen in Rahmen epileptischer Psychosen. In Halluzinationen bei
Epilepsien und ihre Dierentialdiagnose, ed. K. Karbowski, pp. 5866. Bern: Huber.
Wolf, P. (1984). The clinical syndromes of forced normalization. Folia Psychiatr Neurol Jpn, 38,
18792.
Wolf, P. (1985). Epilepsy with grand mal on awakening. In Epileptic Syndromes in Infancy,
Childhood and Adolescence, 2nd edn (1992), ed. J. Roger, M. Bureau, Ch. Dravet, F.E. Dreifuss,
A. Perret and P. Wolf, pp. 32941. London: John Libbey.
Wolf, P. (1986). Forced normalization. In Aspects of Epilepsy and Psychiatry, ed. M.R. Trimble and
T.G. Bolwig, pp. 10112. Chichester: John Wiley & Sons.
Wolf, P. and Trimble, M.R. (1985). Biological antagonism and epileptic psychosis. Br J Psychiatry,
146, 2726.
Yamashita, K., Aoki, K. and Kawai, I. (1984). Primary generalized epilepsies and temporal lobe
epilepsies: a psychological study using PF study. J Jap Clin Psychology, 2, 3843.
5
Epilepsy and learning disorders

Cesare Maria Cornaggia1 and Giuseppe Gobbi2
1
University of Milano Bicocca, Milan, Italy
2
Maggiore Hospital, Bologna, Italy
Introduction
In the practice of epilepsy, the association between epilepsy and learning disorder
is very important, even if sometimes underestimated. A correct and early diagno-
sis of a learning disorder in a person with epileptic seizures will often determine
future development and prognosis.
Definitions
Cognitive function denes the capacity of the human brain to process all informa-
tion coming from the outside and internal world of the individual, and program
ongoing behaviour (Aldenkamp and Bronswijk, 1999). This capacity involves the
ability to remain in contact with the outside world (through the function of vigi-
lance), to select and focus information (through the function of attention), and to
memorize data (through the function of memory). In this way, cognitive function
gives humans the opportunity of becoming aware of themselves and to solve prob-
lems something that we also call intelligence. The latter is the generic capacity of
using all of the elements of thinking necessary to recognize, plan and solve new
problems in a directed and correct manner.
More than one cortical area of the human brain is involved in cognitive function
and related processes, and impaired cognitive function has been observed in the
presence of a lesion or stable dysfunction in the temporal, frontal or parietal lobes
of the dominant and nondominant hemispheres. An impairment in cognitive func-
tion may be seen as a reduction in the capacity of learning of children or reduction
in the intellectual ability of adults. We normally refer to this as learning disability in
children (which, at least in some countries, is dierent from mental retardation),
and deterioration or dementia in adults or people who have already acquired mental
capacity.
From the terminological point of view, learning disorder is dierent from mental
retardation. Actually, in some countries, such as the UK, learning disability is used
62
63 Epilepsy and learning disorders
for dening both a mental retardation and the learning disorders. But, in the
Diagnostic and Statistical Manual of Mental Disorders (DSMIV; American
Psychiatric Association, 1994) learning disorders (DSMIV 315.00315.09) include
a signicant disturbance in academic achievement or daily living activities that
require reading, mathematical or writing skills. In contrast, mental retardation is
dened as a not necessarily lifelong disorder characterized by the combination of a
low cognitive ability and diminished social and adaptive competence (DSMIV
317319) due to signicantly subaverage intellectual functioning, with an onset
before the age of 18 years.
Actually, in most parts of the world, mental retardation denes a situation
involving an abnormal IQ, and a learning disorder one involving a normal IQ.
On the other hand, there are some doubts about the denition of mental retar-
dation, especially in childhood epilepsy. In childhood epilepsy there is a risk of
diagnosing the presence of mental retardation and failing to identify patients who
are only aected by a specic learning disorder. Thus, it is known that an epileptic
focus may lead to the dysfunction of a given specic performance, that coincides
with the cortical area involved, without causing mental retardation. But, at a devel-
opmental age, when cognitive performance is progressing, a dysfunction in a spe-
cic cognitive area may cause a disequilibrium of the complex system of mental
(cognitive) development, causing an apparent mental retardation, that may com-
pletely resolve if this specic cognitive dysfunction is adequately treated.
The term retardation itself may suggest that it is possible to make up the leeway,
but we know that in many or perhaps the majority of cases of so-called mental
retardation this is not to be expected, especially in the presence of brain lesions or
some special epileptic syndromes. In some countries (including Italy), in fact, the
term mental deciency is used instead of mental retardation for dening situations
assuming a denite decit.
This is the case, also, with the term disability. Disability means a restriction of
the ability to perform an activity and it is generally considered to be stable and irre-
versible, at least in some countries such as Italy. In the case of epilepsy it could be
more appropriate to use the word disorder because this opens up the possibility that
the loss of capacity may be transitory or reversible, so-called state dependency as
discussed by Besag (1989, 1995, Chapter 6).
Learning disorders are linked to various types of other behaviour disturbances
that must be separately classied: for example, the Attention Decit/Hyperactivity
Disorder (ADH; DSMIV 314.00314.01), which may or may not be associated
with behavioural or mood disturbances (DSMIV 312.8). We also know that learn-
ing disorders are associated with other disorders, such as conduct disorders, major
depressive disorders, dysthymic disorders and opposition deant disorder, in 10 to
25% of the cases.
64 C.M. Cornaggia and G. Gobbi
Learning and behavioural disorders are linked both in childhood and in adults.
Although it is true that behavioural or mood disorders may be seen as a sympto-
matic expression of a cognitive impairment during epilepsy, it is also true that
behavioural or mood disorders may lead to reduced learning ability.
Prevalence
There is no doubt that a disproportionate percentage of children with epilepsy have

learning and behavioural problems (Bourgeois, 1998), and children with behav-
ioural disturbances are certainly a minority in respect to the number of those with
learning disorders.
The concept that epilepsy is invariably associated with progressive intellectual
deterioration both in children and in the adult population, which prevailed during
the rst half of the twentieth century (Bourgeois, 1998), cannot be accepted any
more. In fact, a large proportion of children with epilepsy have some schooling
diculties (Ross et al., 1980, Sillanp et al., 1999), but fewer than 1% attend
special epilepsy schools. Nakken and Brodtkorb (1999) have recently reported that
about 20% of the epilepsy population are mentally retarded (Forsgren et al., 1990,
1996), and about 20% of mentally retarded people have epilepsy (Wester, 1982). In
reality, only a small subgroup of children with epilepsy shows stable decreases in
IQ, as recently shown by Besag (1995), who demonstrated that the observation of
a declining IQ at any given moment does not necessarily indicate a loss of skills, but
simply implies that the child is not developing at the expected rate.
According to some authors, this lag between mental and chronological age is
more pronounced in older children, and in children with an earlier seizure onset,
a higher lifetime total number of seizures, or multiple seizure types (Seidenberg et
al., 1986). There is some evidence indicating that children aged less than ten years
at the time of assessment gain less than those with an age at onset of more than ten
years (Neyens et al., 1999).
Causes of learning disorders
There are three specic ways in which epilepsy and learning disorders may be
related. First, epilepsy and learning disorders are the result of brain damage and/or
permanent brain dysfunction. Secondly, epilepsy may cause brain damage or per-
manent brain dysfunction, which in turn leads to a learning disorder. Status epi-
lepticus would be the classical example, as there is no doubt that epilepsy can cause
deterioration in children experiencing prolonged bouts of status epilepticus.
Thirdly, epilepsy may cause learning disorders directly, without brain damage or
permanent brain dysfunction, such as in the case of continuous epileptiform EEG
discharges or subtle seizures, and there is also some evidence that certain sources
of seizure discharge are associated with selective cognitive decits (Stores, 1985)
for example the LandauKlener syndrome.
Classification of learning disorders
Learning disorders occurring during epilepsy can be classied into two categories:
state-dependent (potentially treatable and reversible) and permanent.
The prevalence of state-dependent or permanent learning disorders is not satis-
factorily known. Brain damage or stable brain dysfunction is mainly associated
with permanent learning disorders. Epilepsy itself or the medication used to treat it
may lead to state-dependent learning disorders. The picture is further complicated
by associated mood disorders or psychoses, a low level of self-perception and
expectation and reduced learning opportunities.
Regarding epilepsy itself, various situations can be recognized as leading to a
state-dependent learning disorder.
Direct eects of seizures. There is evidence that single complex partial and secon-
darily generalized seizures are associated with neuronal damage (Rabinowicz et
al., 1996), and that brain extracellular glutamate may build up to neurotoxic
levels in partial seizures (During and Spencer, 1993). Perhaps, not surprisingly, a
reduction in seizure frequency over time is associated with an improvement in
cognitive functioning (Seidenberg et al., 1981), and the presence of discharges
aects discrete scholastic performances (Kasteleijin-Nolst Trenite et al., 1988).
Special epileptic syndromes. Typical examples are West syndrome (WS) and
LennoxGastaut syndrome (Gokyigit and Caliskan, 1995; Oguni et al., 1996).
Ictal changes. Nonconvulsive status epilepticus, which may continue for months
or years is an example.
Peri-ictal changes. This situation refers to patients with so many seizures every day
that they do not have time to recover from one to another. Such patients are eec-
tively in a continuous postictal state, which may be misinterpreted as a perma-
nent learning disorder.
The presence of frequent subtle seizures and the direct eects of interictal EEG epi-
leptiform discharges. Children with epilepsy who show epileptiform discharges
during IQ tests do less well than those who do not (Siebelink et al., 1988). The
Transitory Cognitive Impairment (TCI) due to the direct eects of frequent
interictal EEG epileptiform discharges or frequent subtle seizures, is a typical
example. If these are very frequent, they may represent a nonconvulsive status
epilepticus.
In fact, state-dependent learning disorders may be a consequence of TCI, asso-
ciated with either focal or generalized specic epileptiform EEG discharges. In 1939
Schwab demonstrated that generalized spike-wave discharges without any apparent

clinical concomitants may be associated with a simultaneous decline in cognitive
function. TCI was also reported in 1957 by Kooi and Hovey, who studied a case
involving focal discharges. These subclinical or larval discharges are accompanied
by a TCI which, if not recognizable from the patients spontaneous behaviour, can be
detected by appropriate psychological tests during EEG monitoring (Binnie, 1979).
TCI is most readily demonstrated by dicult tasks and during generalized
regular spike-wave discharges lasting more than 3 seconds, but it can also be found
with briefer and focal discharges. In this last case, the errors are more specic: left-
sided focal discharges are more likely to produce errors in verbal tasks, whereas
right-sided discharges are associated with the impaired handling of non-verbal
material. This means that TCI is not only a consequence of global inattention, and
it has been hypothesized that the errors may be due to a working memory disorder
(Aarts et al., 1984). The eect of a TCI accompanying subclinical EEG discharges
on everyday functioning is still uncertain and needs further examination, but there
is experimental evidence that subclinical discharges may be accompanied by a dis-
ruption of educational skills in children, and the AED suppression of discharges is
concomitant with cognitive improvement.
It has been suggested that, if allowed to continue for a long time, state-dependent
learning disorders may produce permanent learning disorders. Strong evidence for
this comes from some rare models of epilepsy, e.g. LandauKlener syndrome and
continuous spike-wave in slow wave sleep (CSWS).
Certainly, the extent to which subtle seizure manifestations aect learning and
behaviour remains unknown. This suggests that there are major controversial ques-
tions to be asked regarding the importance of subtle manifestations of epilepsy on
learning and behaviour. Further, it is possible that some children diagnosed as
having ADHD or autism may actually be aected by undiagnosed manifestations
of epilepsy.
It is certainly a common experience to observe that many normally underesti-
mated or undiagnosed severe epileptic conditions, such as some frontal lobe epi-
lepsies with frequent specic EEG activity (including rapid activity lasting between
one and one-and-a-half seconds), are subsequently associated with the onset of
severe behavioural disturbances (sometimes with psychotic symptoms), or schizo-
phrenia-like or autism-like syndromes.
The existence of situations such as those reported above underlines the need to
clarify the causative role of epilepsy or the presence of nonconvulsive EEG epilep-
tiform discharges in childhood and adult behavioural problems. We must not be
afraid to use the EEG in children and in adults showing behavioural or learning dis-
turbances. Similarly, we believe that the issue of intervening to ameliorate interic-
tal EEG discharges has to be reconsidered. Of course, that needs care and has to be
evaluated in each single patient, since it is dicult to ensure that the drugs them-
selves do not disrupt cognitive processing.
The role of treatments
The data available in the literature indicate that the administration of antiepileptic
drugs (AEDs) can impair cognitive function and aect the behaviour of children
and adults, but the extent and severity of this is still unknown. Moreover, AEDs may
cause state-dependent learning disorders, but evidence is anecdotal.
Some recent studies have compared the ecacy and presence of the cognitive
and behavioural side eects of various AEDs. For example, it has been found that
the ecacy of vigabatrin, lamotrigine and gabapentin is similar, but that vigaba-
trin is associated with a higher incidence of behavioural problems (Bhaumik et al.,
1997). On the other hand, an overall improvement in attention and school perfor-
mance has been reported in a group of children treated with vigabatrin monother-
apy in respect to a carbamazepine monotherapy control group (Gobbi et al., 1999).
The results of other studies support the hypothesis that the cognitive prole of
lamotrigine is similar to that of carbamazepine (Aldenkamp et al., 1997).
In general, an AED may improve learning by reducing the number of EEG dis-
charges or the frequency of seizures but, at the same time, it may impair learning as a
result of its side eects sleepiness, slowed reactions, attention decit and so on. It is
important not to forget clinical observations and manage each single case separately.
Finally, learning disorders need to be considered in the preoperative evaluation
and postoperative rehabilitation of patients with epilepsy.
In the preoperative phase, it is important to consider the eect that the surgical
ablation of the area of the epileptic focus may have on the processes of learning.
The ablation of an epileptogenic area generally produces positive eects because it
eliminates the negative consequences of the epileptic discharges that go from this
to other parts of the brain. However, if the area of the epileptogenic focus has not
completely lost its own primary function, the same ablation may lead to a func-
tional decit. In this case, even a satisfactory improvement in seizure frequency
may lead to very negative postoperative results.
Postoperative rehabilitation must take into account the developmental level of
the individual, and may take weeks or months of intensive specialist input but, pro-
vided that this is carefully planned and provided, the outcome can be very good.
Conclusions
Tentatively, we may summarize the actual picture as follows. Epilepsy may be asso-
ciated with learning disorders, and also behavioural or mood disorders. Behaviour
and learning are often linked; behavioural changes may result from temporary or
permanent cognitive decits. Learning disorders in epilepsy may be state depen-
dent or permanent; state-dependent learning disorders may produce permanent
learning disorder. The prevalence of both conditions is not known. Strong evidence
now shows that very frequent overnight and day-time epileptiform discharges may
eventually result in permanent learning disorder. Epilepsy itself, AEDs and social
factors may inuence the onset of learning disorders. Although a large percentage
of children with epilepsy have school diculties, only a minority shows a loss of
skills or a decreasing IQ. For children in whom actual loss of skills occurs, it is
essential to look for a specic cause, and to plan an ad hoc management pro-
gramme.
R E F E R E N C ES
Aarts, J.H.P., Binnie, D.J., Smit, A.M. et al. (1984). Selective cognitive impairment during focal
and generalized epileptiform EEG activity. Brain, 107, 293380.
Aldenkamp, A.P. (1997). Eect of seizures and epileptiform discharges on cognitive function
Epilepsia, 38 (Suppl. 1), 525.
Aldenkamp, A.P. and Bronswijk, K. (1999). Cognitive side eects as an outcome measure in anti-
epileptic drug treatment: the current debate. In Epilepsy and Mental Retardation, ed. M.
Sillanpaa, L. Gram, S.I. Johannessen and T. Tomson, pp. 135146. Peterseld: Wrightson
Biomedica Publishing.
(Fourth edition) (DSMIV). Washington, DC: APA.
Besag, F.M.C. (1989). Epilepsy, learning and behaviour. Educ Child Psychol, 6.
Besag. F.M.C. (1995). Epilepsy, learning, and behavior in childhood. Epilepsia, 36 (Suppl. 1), 5863.
Bhaumik, S., Branford, D., Duggirala, C. and Ismail, I.A. (1997). A naturalistic study of the use
of vigabatrin, lamotrigine and gabapentin in adults with learning disabilities. Seizure, 6,
12733.
Binnie, C.D. (1979). Direction of transitory cognitive impairment during epileptiform EEG dis-
charges: problems in clinical practice. In Epilepsy and Behaviour, ed. B.M. Kulig, H. Meinardi
and G. Stores. Lisse: Swets & Zeitlinger.
Bourgeois, B.F.R.D. (1998). Antiepileptic drugs, learning, and behavior in childhood epilepsy.
Epilepsia, 39, 91321.
During, M.J. and Spencer, D.D. (1993). Extracellular hippocampal glutamate and spontaneous
seizure in the conscious human brain. Lancet, 341, 160710.
Forsgren, L., Edvinsson, S.-O., Blomquist, H.K., Heijbel, I. and Sidenvall, R. (1990). Epilepsy in
a population of mentally retarded children and adults. Epilepsy Res, 6, 23448.
Forsgren, L., Bucht, G., Eriksson, S. and Bergmark, L. (1996). Incidence and clinical character-
ization of unprovoked seizures in adults: a prospective population-based study. Epilepsia, 37,
2249.
Gobbi, G., Pini, A., Bertani, G. et al. (1999). Prospective study of rst-line vigabatrin monother-
apy in childhood partial epilepsies. Epilepsy Res , 35, 2937.
Gokyigit, A. and Caliskan, A. (1995). Diuse SpikeWave status of nine year duration without
behavioural change or intellectual decline: Epilepsia, 36, 21013.
Kasteleijin-Nolst Trenite, D.G., Bakker, D.J., Binnie, C.D., Buerman, A. and Van Raaij, M. (1988).
Psychological eects of subclinical epileptiform EEG discharges. I Scholastic skills. Epilepsy
Res, 2, 11116.
Kooi, K.A. and Hovey, H.B. (1957). Alterations in mental functions and paroxysmal cerebral
activity. Arch Neurol Psychiatry, 1978, 26471.
Nakken, K.A. and Brodtkorb, E. (1999). Epilepsy services for the patient with mental retardation
in Norway. In Epilepsy and Mental Retardation, ed. M. Sillanp, L. Gram, S.I. Johannessen and
T. Tomson, pp. 193200. Peterseld: Wrightson Biomedical Publishing.
Neyens, L.G.J., Aldenkamp, A.P. and Meinardi, H.M. (1999). Prospective follow-up of intellec-
tual development in children with a recent onset of epilepsy. Epilepsy Res, 34, 8590.
Oguni, H., Hayashi, K. and Osawa, M. (1996). Long-term prognosis of LennoxGastaut syn-
drome. Epilepsia, 37 (Suppl. 3), 447.
Rabinowicz, A.L., Carreale, J., Buotros, R.B., Coulwell, W.T., Henderson, C.W. and DeGiorgio,
C.M. (1996). Neuron-specic enolase is increased after single seizures during inpatient video-
EEG monitoring. Epilepsia, 37, 1225.
Ross, E.M., Peckham, C.S. and West, P.B. (1980). Epilepsy in childhood: ndings from the
National Child Development Study. Br Med J, 1, 20710.
Schwab, R.S. (1939). A method of measuring consciousness on petit-mal epilepsy. J Nerv Ment
Dis, 89, 6901.
Seidenberg, M., OLeary, D.S., Berent, S. and Boll, T. (1981). Changes in seizure frequency and
test-retest scores on Wechler adult intelligence scale. Epilepsia, 22, 7583.
Seidenberg, M., Beck, N. and Geisser, M. (1986). Academic achievement of children with epi-
lepsy. Epilepsia, 27, 7539.
Siebelink, B.M., Bakker, D.J., Binnie, C.D. and Kasteleijn-Nolst Trenite, D.G. (1988).
Psychological eects of subclinical epileptiform EEG discharges in children. II general intelli-
gence tests. Epilepsy Res, 2, 11721.
Sillanp, M., Gram, L., Johannessen, S.I. and Tomson, T. (1999). Epilepsy and mental retarda-
tion. Peterseld: Wrightson Biomedical Publishing.
Stores, G. (1985). Clinical and EEG evaluation of seizures and seizures-like disorders. J Am Acad
Child Psychiatry, 24, 1016.
Wester, K. (1982). Epilepsi hos psykik utviklingshemmede. Tidsskr Nor Laegeforen, 102, 2258.
6
Subtle cognitive and behavioural effects of

epilepsy
Frank M.C. Besag
Bedfordshire and Luton Community NHS Trust, Bedford, UK
Introduction
Both cognitive and behavioural problems are common in people with epilepsy.
Epidemiological studies have indicated that around 50% of children with epilepsy
have some schooling diculties and that many have behavioural problems (Besag
et al., 1999; Pazzaglia and Frank-Pazzaglia, 1976; Ross et al., 1980; Sillanp, 1992).
The National Child Development Study in the UK (Ross et al., 1980) revealed that
only 67% of children with epilepsy were attending a mainstream school at age 11.
Using this broad-brush measure it was evident that a large proportion of the chil-
dren with epilepsy had signicant schooling problems. The work of Pazzaglia and
Pazzaglia in Cesena, Italy, (Pazzaglia and Frank-Pazzaglia, 1976) also conrmed
that a high proportion of children were underachieving at school. The excellent
epidemiological studies of Sillanp (Sillanp, 1992) in Finland showed that
31.4% of this unselected sample of children had mental retardation.
In a recent study carried out in the London Borough of Lambeth Besag et al.
(1999) surveyed 127 children with epilepsy. The parents perceived that the children
had schooling problems or learning disability in 65% and 48% were disturbed on
Rutter Behavioural Scales. Sixty-two per cent of the scores indicated a signicant
impact on quality of life which was moderate or great in 35%. These studies suggest
that schooling and behavioural diculties constitute a major problem in child-
hood epilepsy. There is a lack of data, however, to indicate the causes of the prob-
lems encountered, either in children or adults.
A systematic framework for assessing behaviour
The current author has suggested that, when faced with an individual who has epi-
lepsy and behavioural problems, the best approach is to use a systematic framework
for assessing the possible cause or causes of the behavioural disturbance (Besag et
al., 1989a; Besag, 1995). This framework consists of ve main categories: the epi-
70
71 Subtle effects of epilepsy
lepsy itself, treatment of the epilepsy, reactions to the epilepsy, associated brain
damage/dysfunction and causes equally applicable to people without epilepsy.
The rst category, the epilepsy itself, may be broken down into the peri-ictal phe-
nomena of prodrome, aura, automatism and postictal changes, interictal psycho-
ses, focal discharges and frequent absence seizures. This list is not necessarily
comprehensive.
Examples of the way in which the epilepsy itself may aect behaviour have been
provided elsewhere(Besag et al., 1989a). A subtle cognitive or behavioural eect of
epilepsy may be dened as an eect that is not immediately obviously attributable
to an epileptic seizure. This does not imply that the epileptic activity itself is neces-
sarily subtle. For example, prodromal mood change preceding a tonic-clonic
seizure could be regarded as a subtle eect of the epilepsy, although the seizure itself
is not at all subtle.
On the other hand, the subtle cognitive and behavioural eects of epilepsy may
be the result of subtle seizure activity. In this context, subtle seizure activity is taken
to mean a clinical manifestation of epileptiform activity that is not immediately
obviously identiable as a seizure. Absence seizures and many complex partial sei-
zures would fall into this category as would transitory cognitive impairment. There
is a growing realization of the fact that epilepsy may aect cognition and behavi-
our in a variety of ways (Aldenkamp, 1997; Deonna, 1995; Stores and Hart, 1975;
Stores, 1978; Trimble, 1988). There is often an overlap between cognitive and
behavioural eects of epileptiform activity.
Frequent frontal discharges may lead to a high degree of social disinhibition and
consequent behavioural disturbance.
A teenager had a long history of gross behavioural disturbance. The EEG showed very fre-
quent left frontal discharges, typically occurring every second. He underwent a left frontal
lobectomy, following which the epileptiform discharges were abolished, his behavioural dis-
turbance resolved and he returned to his pleasant premorbid personality.
Frequent left temporal discharges may be associated with aggressive behaviour.

There is evidence from the literature suggesting that young men who have left tem-
poral discharges and subsequently undergo a left temporal lobectomy show
improvements both in seizure control and in behaviour (Falconer, 1973).
Further examples of both cognitive and behavioural disturbance that can result
from subtle manifestations of epilepsy will be discussed in the next section.
State-dependent cognitive impairment
It is important to distinguish permanent cognitive impairment on one hand from

state-dependent cognitive impairment on the other. The concept of permanent
72 F.M.C. Besag
cognitive impairment is readily understood. This may arise from a wide variety of
causes of permanent brain damage or dysfunction that may be prenatal, perinatal
or postnatal. The concept of state-dependent cognitive impairment is less widely
acknowledged (Besag, 1994).
What is state-dependent cognitive impairment? State-dependent cognitive
impairment may be dened as cognitive impairment that depends on current
factors aecting the individual, for example antiepileptic medication or epileptic
activity, that are not necessarily permanent. State-dependent cognitive impairment
is potentially reversible and treatable. Failure to treat state-dependent cognitive
impairment is failure to provide an adequate service to the patient. Some might put
the case even more strongly, suggesting that failure to recognize and treat state-
dependent cognitive impairment is a reection on professional competence.
The rst step in managing state-dependent cognitive impairment is to think of
the diagnosis. Regrettably, professionals often fail to take this rst step. The result
is that the condition is neither recognized nor treated. State-dependent cognitive
impairment may be divided into two broad categories: drug-induced and epilepsy-
induced. As already indicated, the epilepsy itself may cause state-dependent cogni-
tive impairment in a number of dierent ways.
Frequent absence seizures

Frequent absence seizures, by interrupting awareness, can aect both cognitive per-
formance and behaviour. People who are having frequent absence seizures may
present as having withdrawn behaviour, fragmented thought processes which may
be mistaken for a psychosis, attention-decit disorder with motor overactivity or,
if the frequency of the seizures is variable, attention-seeking behaviour. The last of
these behaviours tends to be seen when the person emerges from a bout of very fre-
quent absence seizures. It is almost as if the child is making up for lost time in
being badly behaved when he has the opportunity of doing so, having been unable
to misbehave when the absence seizures were very frequent.
In some cases the absence seizures may become so frequent as to amount to non-
convulsive status epilepticus. The person is eectively cut o from his or her sur-
roundings because he or she does not have the opportunity to function adequately
between the absence seizures. Complex partial seizure status epilepticus may also
cause a continuous state of behavioural and cognitive change until the status epi-
lepticus is either treated or resolves spontaneously.
It is clear that absence seizures can aect both performance and condence in a
major way, particularly if they are occurring frequently. It is not possible to count
absence seizures by simple observation alone. To overcome this diculty, an auto-
matic spike-wave monitor, the Monolog was developed (Besag et al., 1989b). The
Monolog is so-called because it monitors and logs spike-wave episodes automat-
ically. Two recording electrodes and a reference electrode are attached to the scalp.
The leads are brought down to a small solid-state recording device, which is in a
belt-worn pouch.
This device provides three pieces of hard data: the total duration of the spike-
wave episodes over the period of recording, the number of spike-wave episodes
and, by playing the solid state circuitry into a chart recorder, the position of each
spike-wave episode in time. In addition, there is an optional plug-in lapel badge
with a ashing light and bleeper, activated by the presence of the spike-wave epi-
sodes. The teacher or carer is alerted by the ashing light/bleeper and he or she will
realize that the child is having a spike-wave episode; it may be appropriate for the
teacher/carer to repeat what has been said to the child during that episode. The light
and bleeper also serve to alert the teachers and carers to the high frequency of
absence seizures occurring in some children. Because absence seizures are often
quite subtle in their manifestation, the teacher may be surprised to note how often
the badge indicates that spike-wave discharges are occurring. This in itself may
result in an alteration of attitude by the teacher who may have an increased incli-
nation to repeat information if the child has not obviously taken it in on the rst
occasion. Using this monitor it has been possible to show that some children have
not only hundreds but thousands of spike-wave episodes daily. The device also con-
rms whether treatment is eective or not.
A 13-year-old boy with a history of epilepsy was reported as having reached the stage at
which the epilepsy was not a problem for him. He was having relatively infrequent overt sei-
zures. However, he presented as a withdrawn child who would not join group activities. He
preferred to sit in the corner of the room sucking his thumb. In the clinic the examiner
uttered the childs name on four occasions in succession. On the first three occasions there
was no response because he was momentarily unconscious in a subtle absence seizure. On
the fourth occasion, when he was not having an absence seizure, he responded promptly
and was more than willing to attend to whatever task the examiner asked him to do.
Prolonged EEG monitoring revealed that around three thousand spike-wave episodes
occurred daily. He responded very well to treatment.
Some individuals have both more obvious seizures, such as tonic-clonic or tonic
seizures and absence seizures. The antiepileptic medication may improve both the
obvious and the subtle seizures. However, in some cases the obvious seizures may
not be aected by medication while the subtle seizures are greatly reduced. Such
individuals may become bright, alert and in control of their lives as a result of the
medication change, even though the obvious seizures have not been reduced in fre-
quency. When patients are assessed in the outpatient clinic the doctor will typically
ask how many seizures have been recorded or will consult the patients seizure diary.
If there has been no reduction in the obvious seizures the doctor may choose to dis-
continue the medication with which the patient has been treated and try another
74 F.M.C. Besag
(a) Overt seizures (b) Spike-wave events

120 (a) in 3 months 4000 (b) in 24 hours
100
Number of seizures
3000
Number of events
80
60 2000
40
1000
20
0 0
Baseline 13 46 79 Baseline 2 4 8 10
Months Weeks
Figure 6.1. The effect of lamotrigine administration on number of overt seizures (a) and spike-wave
events (b) experienced by Subject A.
antiepileptic drug. However, on this basis the doctor might stop a drug that has
been highly eective in treating subtle absence seizures, even though there has been
no major eect on the obvious seizures. An example is provided in Figures 6.1 and
6.2. In Case A there has been a reduction both in the obvious seizures and in the
spike-wave episodes with the addition of lamotrigine. In Case B, however, the
obvious seizures have not decreased. The spike-wave episodes, on the other hand,
have been reduced by over 2000 per day. The parents of this teenager were delighted
with the transformation. He was able to relate much more readily to the world
around him and appeared much more alert. It is important not to discontinue
medication on the basis that obvious seizures have not responded if the individual
has had a good response in terms of reduction of subtle seizure activity.
Transitory cognitive impairment
A number of papers describing this phenomenon have been published by Binnie

and coworkers (Aarts et al., 1984; Marston et al., 1993). The basic concept is that
an epileptiform discharge that does not appear, on simple observation, to be mani-
festing as a seizure, may nevertheless cause a transitory impairment of cognitive
function. It has been shown that discharges on the left side may impair language
function and on the right side may impair visuo-spatial skills. The quality of the
impairment is not always straightforward. For example, a child who is having epi-
(a) Overt seizures in 3 months (b) Spike-wave events in 24 hours

120 3000
100 2500
Number of seizures
Number of events
80 2000
60 1500
40 1000
20 500
0 0
Baseline 13 46 79 2 4 6 10 32 48
Months Baseline Weeks
Figure 6.2. The effect of lamotrigine administration on number of overt seizures (a) and spike-wave
events (b) experienced by Subject B.
leptiform discharges during reading may pause or may read even more quickly but
make additional mistakes when discharges are occurring.
Binnie and his co-workers have shown that reduction of epileptiform discharges
can result in improvement of psychosocial functioning (Marston et al., 1993),
although this study was confounded by the fact that obvious seizures were
improved as well as the discharges causing the transitory cognitive impairment.
Observation of video tapes taken during testing of young people who have transi-
tory cognitive impairment leaves no doubt about the fact that they nd their lapses
in performance irritating and frustrating. It is highly likely that this phenomenon
aects self-esteem and self-condence. This implies that, although the phenome-
non itself is transitory, there may be an ongoing negative eect on the attitude and
behaviour of the individual.
Frequent localized discharges
Reference has already been made to frequent frontal discharges and frequent left
temporal discharges aecting behaviour. Localized discharges can also aect
ongoing cognitive performance. This eect is more than just transitory if the dis-
charges are frequent.
A girl had a history of a benign left-sided temporal lobe tumour. She had very frequent epi-
leptiform discharges arising from the left temporal lobe. She underwent left temporal lobec-
tomy at 13 years of age. The epileptiform discharges disappeared after the neurosurgery. In
a single year her speech development improved from around the 4-year level to the 7-year
76 F.M.C. Besag
level. The temporal lobectomy released the remaining brain from the effect of very frequent
epileptiform discharges, allowing rapid progress to be made.
Frequent hemispheric discharges
The extreme example of frequent localized discharges is provided by individuals

who have an abnormal hemisphere which is the source of such discharges. A
number of dierent pathological conditions can give rise to this phenomenon,
including hemimegalencephaly, a porencephalic cyst and the unilateral cerebral
atrophy that accompanies Rasmussens encephalitis. These individuals often
improve markedly after hemispherectomy. Not only are the seizures and epilepti-
form discharges abolished but the behaviour also improves greatly(Goodman,
1986). If there is a gross structural abnormality that is causing very frequent epi-
leptiform discharges then early surgery should be considered. If the lesion respon-
sible has been present from a very early age, for example if it is a congenital
porencephalic cyst, then the outcome is likely to be good.
A dilemma arises in the case of the child who develops Rasmussens encephalitis
that is causing progressive hemiatrophy of one hemisphere. The function subserved
by the aected hemisphere is likely to worsen and the longer the operation is
delayed the less opportunity there will be for the healthy hemisphere to take over
function. However, hemispherectomy usually results in loss of nger function on
the aected side and there is a tendency to delay the operation until it is clear that
the nger function has already been lost, implying that the surgery will not impair
function further.
Is it appropriate to wait? Because brain plasticity is greatest when the child is
young, there is a better chance that the remaining hemisphere will take over func-
tion if the surgery is performed early. There is a strong argument, in such cases, for
carrying out the surgery sooner rather than later if the assessment indicates that the
disease in the aected hemisphere is progressive.
Postictal state-dependent cognitive impairment
At rst sight it may appear trivial to discuss postictal cognitive impairment because
this is such an obvious phenomenon. However, it is not always as obvious as one
might imagine.
An apparently dementing teenager sat rocking in his chair, not knowing where he was,
barely able to carry out a rudimentary conversation and unable to perform his skills of daily
living. He had been accepted as a boarding pupil in a special residential centre for children
with epilepsy and other needs. The staff felt that he had been placed inappropriately. They
commented that he could not learn and that he should not have been accepted. At that
stage he was having between three and five seizures a day. Following a medication review
he became seizure-free. He was then fully orientated and began progressing very well in his
educational programmes. Both his parents and the staff were delighted with his progress.
The staff then commented how appropriate the placement was. They had previously
assumed that his cognitive impairment was permanent whereas, in fact, a large component
of the cognitive impairment was state-dependent, treatable and reversible. When he had
been having frequent daytime seizures he had not had the chance to recover from one
seizure before he had another. He was in a constant postictal state. When he emerged from
this constant postictal state he was able to learn again and progressed rapidly.
It is important to be aware of the possibility of reversible, state-dependent cogni-

tive impairment in an individual who is in a constant postictal state from frequent
seizures. Reducing the seizure frequency can improve cognitive function markedly.
Overnight video-telemetry has revealed that some people have very frequent
unsuspected and unobserved nocturnal seizures. These seizures may be brief, silent
and easily missed, even by awake night sta. In a series of 15 patients examined by
the author and his coworkers, large numbers of nocturnal seizures were recorded
by video-telemetry, that had been unobserved by awake night sta. In one case over
200 brief nocturnal tonic seizures were recorded. Some individuals are woken by
the seizures. Frequent nocturnal seizures may aect daytime performance not only
because of the direct after-eects of the seizures themselves but also because they
may cause a very broken nights sleep.
Electrical status epilepticus of slow wave sleep
The LandauKlener syndrome of acquired epileptic aphasia is classically asso-

ciated with electrical status epilepticus of slow wave sleep (ESES) (Beaumanoir,
1992; Landau and Klener, 1957). However, a series of six cases of ESES examined
by the author and colleagues conrmed that language is not necessarily the func-
tion that is impaired. Although some of these children clearly had language impair-
ment in association with the ESES, others did not.
A boy with a right congenital porencephalic cyst and an accompanying left hemiparesis had
ESES. He had good language skills but his visuo-spatial skills became increasingly impaired.
He was unable to find his way from his bedroom to the bathroom at home. Following neuro-
surgery his visuo-spatial skills improved greatly.
The LandauKlener syndrome provides an important model of state-dependent

cognitive impairment. Some of the older textbooks suggested that antiepileptic
treatment was only of value in treating the seizures but would not aect the cogni-
tive impairment. Such statements were probably based on burnt-out cases in which
the damage had become permanent. It is quite clear that early, energetic treatment
78 F.M.C. Besag
may improve the cognitive impairment in some children with LandauKlener syn-
drome. The implication is that not only may cognition be improved by early treat-
ment but that permanent impairment may be avoided.
Treatments include antiepileptic medication such as sodium valproate and
lamotrigine, steroid treatment and neurosurgery. Multiple subpial transection,
pioneered by Morrell (Morrell et al., 1989), has been particularly eective in some
cases. This can abolish the ESES and allow the language function to return. The
results obtained from multiple subpial transection in this condition have made it
clear that past statements that antiepileptic treatment could not help cognitive loss
were wrong. It is worth repeating that early treatment may not only result in a
return of these skills but may also prevent long-term cognitive impairment.
The issues raised by ESES have led to the suggestion that any child who loses skills
should be investigated fully and, if another cause is not found, investigations
should include overnight EEG monitoring. It should be noted, in this context, that
it is said that around 25% of the children with LandauKlener syndrome do not
have a previous history of obvious seizures (Beaumanoir, 1992).
Preventing cognitive and behavioural problems
In the examples just given, it was emphasized that some children with ESES may
present with impaired cognition. They often also present with markedly disturbed
behaviour. Both the cognitive and behavioural disturbance are reversible with early
treatment.
It has become evident that a number of children who present with markedly
autistic features have unsuspected epileptiform discharges either during the day or
at night (ESES or continuous spike-wave in slow wave sleep CSWS). Treating
these children allows them to emerge from their apparently autistic state.
However, follow-up of teenagers who had very frequent absence seizures in
childhood has indicated that some longer-term problems in social interaction or
behaviour may be evident. Some teenagers who have had very frequent epilepti-
form EEG discharges earlier in life may subsequently still have autistic features
when the epileptiform discharges are no longer present. It would appear that these
young people have been unable to interact adequately with the world around them
during a critical developmental phase because of the frequent epileptiform dis-
charges at that time. If they have not had the opportunity to develop two-way social
interaction during this developmental phase they may present with a very
Asperger-like picture during teenage years. This suggests that early treatment of the
epileptiform discharges, so as to allow the child to gain fully from the early devel-
opmental years, may have avoided these social impairments.
In summary, it appears that there is now an increasing body of evidence to
suggest that frequent epileptiform discharges should not be allowed to continue for
very long periods because they might cause not only permanent cognitive impair-
ment but also permanent social/behavioural problems.
Conclusions
Systematic assessment of cognitive and behavioural problems often allows the

cause to be found and rational management to be provided. It is very important to
consider the possibility of state-dependent cognitive impairment, especially in the
child who has lost skills. Unless another obvious cause is found, the investigations
should include overnight EEG monitoring. If frequent epileptiform discharges are
found in association with cognitive or behavioural problems then early, energetic
treatment should be initiated. Such treatment may not only improve the impair-
ments at the time but may also prevent permanent cognitive or social impairments.
R E F E R E N C ES
Aarts, J.H., Binnie, C.D., Smit, A.M. and Wilkins, A.J. (1984). Selective cognitive impairment
during focal and generalized epileptiform EEG activity. Brain, 107, 293308.
Aldenkamp, A.P. (1997). Eect of seizures and epileptiform discharges on cognitive function.
Epilepsia, 38 (Suppl. 1), S525.
Beaumanoir, A. (1992). The LandauKlener Syndrome. In Epileptic Syndromes in Infancy,
Childhood and Adolescence, ed. J. Roger, M. Bureau, Ch. Dravet, F.E. Dreifuss, A. Perret and P.
Wolf, pp. 23143. London: John Libbey.
Besag, F.M.C. (1994). Epilepsy, education and the role of mental handicap. In Epilepsy, ed. E.M.
Ross and R.C. Woody, pp. 56183. Baillieres Clin Paediatr Int Practice Res, 2, 56183.
Besag, F.M.C. (1995). Epilepsy, learning, and behavior in childhood. Epilepsia, 36, S5863.
Besag, F.M.C., Loney, G., Waudby, E., Fowler, M. and Brooks, N. (1989a). A multidisciplinary
approach to epilepsy, learning diculties and behavioural problems. Educational Child
Psychol, 6, 1824.
Besag, F.M., Mills, M., Wardale, F., Andrew, C.M. and Craggs, M.D. (1989b). The validation of a
new ambulatory spike and wave monitor. Electroencephalogr Clin Neurophysiol, 73, 15761.
Besag, F., ONeill, C. and Ross, E. (1999). A comparison between children with epilepsy in an
inner-city region and those within a special centre, using measures of educational diculty,
behavioural problems and quality of life. Epilepsia, 40, 243.
Deonna, T. (1995). Cognitive and behavioral disturbances as epileptic manifestations in children:
an overview. Semin Pediatr Neurol, 2, 25460.
Falconer, M.A. (1973). Reversibility by temporal-lobe resection of the behavioral abnormalities
of temporal-lobe epilepsy. N Eng J Med, 289, 4515.
Goodman, R. (1986). Hemispherectomy and its alternatives in the treatment of intractable epi-
lepsy in patients with infantile hemiplegia. Dev Med Child Neurol, 28, 2518.
80 F.M.C. Besag
Landau, W.M. and Klener, F.R. (1957). Syndrome of acquired aphasia with convulsive disorder
in children. Neurology, 7, 52330.
Marston, D., Besag, F., Binnie, C.D. and Fowler, M. (1993). Eects of transitory cognitive impair-
ment on psychosocial functioning of children with epilepsy: a therapeutic trial. Dev Med Child
Neurol, 35, 57481.
Morrell, F., Whisler, W.W. and Bleck, T.P. (1989). Multiple subpial transection: a new approach
to the surgical treatment of focal epilepsy [see comments]. J Neurosurg, 70, 2319.
Pazzaglia, P. and Frank-Pazzaglia, L. (1976). Record in grade school of pupils with epilepsy: an
epidemiological study. Epilepsia, 17, 3616.
Ross, E.M., Peckham, C.S., West, P.B. and Butler, N.R. (1980). Epilepsy in childhood: ndings
from the National Child Development Study. Br Med J, 280, 20710.
Sillanp, M. (1992). Epilepsy in children: prevalence, disability, and handicap. Epilepsia, 33,
4449.
Stores, G. (1978). School-children with epilepsy at risk for learning and behaviour problems. Dev
Med Child Neurol, 20, 5028.
Stores, G. and Hart, J.A. (1975). Proceedings: reading skills of children with generalized and focal
epilepsy attending ordinary school. Electroencephalogr Clin Neurophysiol, 39, 42930.
Trimble, M.R. (1988). Cognitive hazards of seizure disorders. Epilepsia, 29 (Suppl. 1), S1924.
7
Aggression and epilepsy

L. Tebartz van Elst
Institute of Neurology, London, UK and Albert-Ludwigs-University, Freiburg, Germany
Introduction
Human aggression is an important social and clinical problem (Fenwick, 1986; Saver
et al., 1996; Swartz et al., 1998; Trimble, 1996). One diculty of studying aggression
is its phenomenological and probably neurobiological heterogeneity, leading to di-
culties in assessment and classication. An important distinction has emerged
between the terms violence and aggression. Following Treiman, violence has been
dened as forceful iniction of abuse or damage onto another individual or object,
but which is not necessarily the result of intentional aggression. In contrast, aggres-
sion is considered an oensive action directed toward another individual or object
with the intent to harm, threaten or control (Treiman, 1991). But even this distinc-
tion leaves researchers with the problem of assessing intentionality, which in clinical
practice often is impossible. Therefore, it is important to dene phenomenological
criteria of the specic behavioural syndrome of interest before embarking on
research or discussion into the neurobiology and psychology of aggression.
Classification of aggression
In the animal kingdom aggressive behaviour has been classied according to the
context in which it can be observed. Table 7.1 summarizes dierent subtypes of
aggressive behaviour in animals following Moyer (1987). In contrast to animal
behaviour, human behaviour is less preformed and it depends less on external cues.
Thus a simple transfer of this classication to human aggressive behaviour is not
valid (Kalin, 1999). Nevertheless, there is general agreement that at least two dier-
ent phenomenological and neurobiological subtypes of aggressive behaviour can
be dierentiated in humans: predatory and defensive aggression (Goldstein, 1974;
Kalin, 1999; Moyer, 1987; Mungas, 1983; Vitiello and Sto, 1997).
Predatory aggression is characterized phenomenologically as a well-structured
and goal-directed behaviour performed in an emotionally calm and concentrated
state of mind. In contrast, defensive aggression is typically being seen in the context
81
82 L. Tebartz van Elst
Table 7.1. Subtypes of animal aggression
Aggression type Example of animal behaviour Behavioural characteristic
Predatory aggression Predator kills prey Calm, concentrated, goal directed, well
structured
Intermale (territorial) Fight for supremacy of two lions Arousal, concentrated, goal directed, well
aggression structured
Maternal aggression Swan protects ospring from Arousal, concentrated, goal directed, well
predator structured
Sex-related aggression Mantis kills male after copulation Calm, behavioural stereotypes
Fear-induced aggression FlightFightReaction Arousal, vocalization, less well structured,
Gnu ghts predator diverse behavioural pattern, reactive
behaviour
Source: Moyer (1987).
of high emotional arousal associated with vocalizations and signs of fear or anger.
The behaviour itself is less structured and defensive (Valzelli, 1981). Apart from the
planned and goal-directed aggression often conducted by persons with antisocial
personality disorder (predatory aggression), most forms of human aggression are
thought to be a reaction toward a perceived threat, be it adequate or not (Albert et
al., 1993). Obviously, the perception as to whether a stimulus is threatening or not
is decisive in the information processing leading to the aggressive behaviour.
Clinical relevance of aggression
Aggressive behaviour can be observed in the context of dierent medical, neuro-

logical and psychiatric disorders and diseases. It is a common problem in patients
with mental retardation, possibly due to impaired social perception or decits in
expressing personal needs (Barratt et al., 1997; Gunn, 1977; Kligman and Goldberg,
1975; Saver et al., 1996). Aggressive behaviour in the context of organic brain
disease like frontal or hypothalamic brain tumours, neuro-degenerative disease,
delirium or drug abuse is often malstructured, defensive and tends to occur in the
context of states of confusion and diuse emotional arousal. Goal-directed and
well-planned acts of aggression can occur on the background of psychiatric dis-
orders like psychosis with delusional states, attention-decit hyperactivity disorder
(ADHA) or bipolar disorder. It is frequently observed in patients with antisocial
personality disorder (APD) where it is part of the characteristic trait-like behaviour
(Barratt et al., 1997; Miller et al., 1997; Stein et al., 1993).
83 Aggression and epilepsy
Table 7.2. Functional relevance of different brain structures for aggressive behaviour
Brain structure Assumed function
Frontal lobe Inhibitory function

Suppression of aggressive behavioural drive
Amygdala and limbic circuits Emotional evaluation of multimodal sensory and cognitive
input; emotional drive and arousal
Hypothalamus Control of brain stem behavioural programs; regulation of
internal environment; coordination of behavioural programs
and internal environment in ightght situations
Brain stem structures i.e. Evolutionary performed behavioural ightght programs
periaqueductal grey
The only clinical syndrome of aggression that has been included in an interna-
tional classicatory system is intermittent explosive disorder (IED) according to the
guidelines of DSMIV (American Psychiatric Association, 1994). IED is character-
ized by several discrete episodes of failure to resist aggressive impulses that result
in serious assaultive acts or destruction of property. The behaviour is out of pro-
portion to any precipitating psychosocial stressor and is not due to substance abuse,
another mental disorder like personality disorder, any other rst axis psychiatric
disorder or a general medical condition like head trauma or neuro-degenerative
diseases. The phenomenological criteria are basically those of episodic dyscontrol, a
psychopathological entity that has been described by many authors in the past
(Bach-Y-Rita et al., 1971).
Neurobiology of aggression
Various brain structures have been implicated in the mediation of aggressive

behaviour in animals and humans, the most important being the periaqueductal
grey (Behbehani, 1995; Brandao et al., 1994), the hypothalamus (Andy and Jurko,
1972), the amygdala and associated limbic structures (Aggleton, 1993; Dicks et al.,
1969; Halgren, 1992; Kling and Brothers, 1992; Rolls, 1992), and the frontal lobes
(Damasio et al., 1990; Miller et al., 1997; Raine et al., 1998).
While an understanding of the complex interplay of these dierent brain circuits
in regulating dierent aggressive behaviours is still poor, the rst elements of a
functional anatomy of aggression can be outlined (see Table 7.2). In animals, brain
stem structures like the periaqueductal grey are crucial for the activation of evolu-
tionary preformed behavioural programs like attacking or defensive behaviour
(Behbehani, 1995; Brandao et al., 1994). These structures are controlled by higher
neuronal centres in the hypothalamus (Bhatnagar and Dallman, 1998; Van de Poll
and Van Goozen, 1992) which in addition to controlling these behavioural brain-
stem programs adjust the internal endocrinological and immunological environ-
ment to aggressive behaviour in ightght situations (Luo, 1998; Reis, 1969;
Shaikh, 1997; Zanchetti, 1968).
Frontal lobe functions are known to be crucial for the ability to suppress behav-
ioural impulses. Consequently patients with frontal lobe lesions lose their ability to
suppress aggressive impulses and thus might present with severe aggressive and
violent psychopathology (Damasio et al., 1990; Krakowski and Czobor, 1997; Petty
et al., 1996; Stein et al., 1993).
Within the network of critical brain structures for aggressive behaviour, the
amygdala are thought to play a crucial role for the mediation of fear-induced
aggression, a subtype of defensive aggression (Aggleton, 1993; Charney and
Deutch, 1996; Gallagher and Chiba, 1996; LeDoux, 1995). They receive extensive
input from various levels of sensory information processing and project to most of
the other critical brain structures i.e. the brain stem, hypothalamus, thalamus and
frontal lobe (Alheid et al., 1995; Amaral et al., 1992). From a neurophysiological
point of view they are in a predestined position for the aective evaluation of multi-
modal sensory input. Thus pathology within the circuits aecting the amygdala
might lead to mental states where the misinterpretation of sensory input as threat-
ening leads to aggressive outbursts. In agreement with this assumption, electrical
stimulation of the amygdala can lead to experiences like fear, anxiety or anger
(Chapman et al., 1954; Gloor et al., 1982) and lesioning of the amygdala severely
impairs fear conditioning in animals (Davis et al., 1994) and humans (LaBar et al.,
1995). Furthermore, in an open retrospective study of 481 cases of bilateral amyg-
dalotomies performed for the control of conservatively untreatable aggressiveness
moderate to excellent improvement of aggressive behaviour was reported in
7076% (Ramamurthi, 1988).
Aggression and epilepsy
The relationship between epilepsy and aggressive behaviour is a particularly con-

troversial issue (Geschwind, 1975). While in patients with episodic aective aggres-
sion, a history of epilepsy is reported to be more common (Bach-Y-Rita et al., 1971;
Elliot, 1982) most of the community-based studies did not nd an increased prev-
alence of aggressive behaviour in patients with epilepsy (Kligman and Goldberg,
1975; Lishman, 1998).
The prevalence of aggression in epilepsy in general, not regarding the specic
epileptic subsyndrome, varies between 4.8% (Rodin, 1973) and 50% (Gastaut et al.,
1955). In a large survey of 666 patients with TLE, Currie et al. (1971) reported
aggression in 7% of the patients. Falconer reviewed 100 patients from Londons
Maudsley Hospital referred for temporal lobectomy and found a prevalence of out-
bursts of aggressive behaviour in 27% of their patients (Falconer, 1973). However,
these studies were hampered by selection bias and the real prevalence of aggressive
behaviour in epilepsy remains controversial (Lishman, 1998).
In epilepsy three dierent types of aggressive behaviours should be distinguished
on the basis of their relationship toward the seizures: ictal, postictal and interictal
aggression. Ictal aggression occurs with extreme rarity (Gunn, 1971; Saver et al.,
1996). In a large survey of several thousand seizures documented on video-
telemetry an incidence of about 1/1000 seizures was found (Delgado-Escueta et al.,
1981). In ictal aggression hostile and verbal or physical aggressive behaviour is often
directed towards nearby objects or persons and may be provoked or not. The
patients are usually amnestic for these aggressive episodes and express remorse for
their behaviour (Devinsky and Bear, 1984; Fenwick, 1989).
Postictal aggression is more common than ictal aggression but it is still believed
to be rare (Treiman, 1991). It often occurs following a cluster of complex partial
seizures or very severe secondary generalized seizures. Ictal pain or dysphoria may
predispose individuals to the development of postictal aggressive behaviour
(Gerard, 1998). Postictal aggression is frequently observed in the context of post-
ictal confusional states or postictal psychosis but it also occurs without any signs of
delusion or hallucination (Kanemoto, 1999; Lancman, 1999; Szabo and Lancman,
1996). If postictal aggression is part of a postictal confusional state the disruptive
behaviour immediately follows the seizure without a lucid interval. The violent
behaviour tends to be resistive, poorly structured and patients usually are very
aroused, angry and fearful (Kanemoto, 1999; Lancman, 1999).
Postictal psychosis follows a cluster of complex partial and secondary general-
ized seizures after a lucid interval, which might vary in duration between hours up
to days (Fenwick, 1986; Kanemoto, 1996; Logsdail and Toone, 1988; Trimble,
1991). Aggressive behaviour in the context of a delusional state may be well-
structured and goal-directed and even though patients often feel angry and
aroused, they may appear calm and concentrated to the observer (Kanemoto, 1999;
Szabo and Lancman, 1996). Kanemoto et al. (1999) recently pointed out that well-
directed and self-destructive behaviour was not a feature of epileptic psychosis in
general but the hallmark of postictal psychosis in particular.
Finally, interictal aggression can be seen in the context of an antisocial personal-
ity disorder which, in turn, might be the consequence of the sometimes dicult
psychosocial background of patients with epilepsy. It might also be part of a psy-
chotic episode (Logsdail and Toone, 1988) but as Kanemoto (1999) pointed out
this is rare compared with postictal psychosis.
Interictal aggression is frequently seen in patients with epilepsy and mental

handicap, but in these patients aggressive behaviour is often a result of poor social
and communication competence in expressing personal needs, and very rarely
results in severe violence (Gunn, 1977). Apart from that, an interictal syndrome of
episodic aective aggression, independent of observable ictal activity, major
psychiatric disorder or antisocial personality disorder, is well described and has
been referred to as episodic dyscontrol (Bach-Y-Rita et al., 1971; Elliott, 1984;
Leicester, 1982; Maletzky, 1973; Ratner and Shapiro, 1979; Stone et al., 1986).
Episodic dyscontrol is characterized by several discrete episodes of extreme
unprovoked arousal and rage resulting in severe aggressive and violent behaviour.
The behaviour is out of proportion to any precipitating psychosocial stressor and
patients often feel remorse for their deeds. The phenomenological criteria are those
of DSMIV intermittent explosive disorder (IED) described above (Elliott, 1984).
Because of the emotional arousal typically seen in episodic dyscontrol, this behav-
ioural syndrome has to be classied as a subtype of aective, defensive aggression.
The high level of arousal with signs of anxiety or fear is also the reason why amyg-
dala pathology is thought to contribute to this behavioural syndrome (Elliot, 1992;
Fenwick, 1986; Trimble et al., 1996).
Intermittent explosive disorder in epilepsy
In the past, few studies have addressed the relationship between dierent psycho-
biological factors like brain pathology, IQ and demographic background, and
aggression in epilepsy. While Rodin found more evidence of organic brain disease
(Rodin, 1973), and Falconer (1973) reported an increased incidence of mesial tem-
poral lobe sclerosis in aggressive patients with temporal lobe epilepsy (TLE),
Herzberg and Fenwick did not nd any relationship between specic electroen-
cephalography (EEG) or computerized tomography (CT) pathology and aggres-
sion in patients with TLE (Herzberg and Fenwick, 1988). All three studies found a
relationship between low IQ and aggression and two reported an association
between male sex and aggression (Falconer, 1973; Rodin, 1973).
Since none of these studies used modern magentic resonance imaging (MRI)
techniques to assess the relationship between brain pathology and aggressive
behaviour in patients with epilepsy, in two recent projects we studied this relation-
ship using quantitative MRI (Tebartz van Elst et al., 2000a,b; Woermann et al.,
2000). The aims of our studies were rst, to investigate amygdala pathology, and
second, to assess cortical abnormalities in patients suering from TLE and addi-
tional aective aggression, specically IED.
The most common pathology underlying TLE in general is hippocampal scler-
osis (HS) often in the context of mesial temporal sclerosis (MTS) (Gloor, 1991;
Margerison and Corsellis, 1966; Wieser, 1983). A radiological in vivo diagnosis of

mesial temporal lobe sclerosis is possible by demonstrating atrophy of the mesial
temporal lobe structures on T1-weighted anatomical MRI images and increased
signal on conventional spin echo T2 MRI sequences (Duncan et al., 1996; Jackson
et al., 1990; Woermann et al., 1998). Since HS seems to be diuse rather than focal
in most of the cases (Kim et al., 1995) an involvement of the amygdala by the patho-
logical process underlying hippocampal sclerosis might be expected, and indeed is
reported in the literature (Hudson et al., 1993; Miller et al., 1994). In vivo identi-
cation of amygdala sclerosis by measuring the amygdala T2 relaxation time has
been reported in patients with TLE (Kalviainen et al., 1997; Van Paesschen et al.,
1996). Furthermore amygdala volumetry has been validated as a reliable method
(Cendes et al., 1993; Kalviainen et al., 1997; Soininen et al., 1994; Watson et al.,
1992).
In our study we hypothesized that, in patients with TLE and intermittent aec-
tive aggression, amygdala sclerosis in the context of hippocampal sclerosis would
be more common as compared to control patients. Furthermore, we wanted to test
if there is an association between aggression on the one hand and hippocampal
sclerosis, low IQ, and poor social adjustment on the other hand in patients with
TLE.
In a second step we analysed cortical grey matter abnormalities in these patients
in order to investigate frontal lobe pathology, since from a phenomenological point
of view episodic dyscontrol or IED can be characterized as a hyperarousaldyscon-
trol syndrome.
Amygdala pathology in patients with TLE and IED

Twenty-ve patients with TLE and IED diagnosed according to the DSMIV crite-
ria described above and 25 control patients with TLE without any psychopathol-
ogy were recruited from a tertiary referral centre (National Hospital for Neurology
and Neurosurgery and the associated Chalfont Centre for Epilepsy). The clinical
syndrome of interest was dened as complex partial seizures with a semiology, EEG
and MRI ndings compatible with TLE. On the basis of the discharge summaries
patients with TLE with and without a history of aggression were identied, con-
tacted and seen by a neuropsychiatrist (LTVE). Patients with extratemporal or gen-
eralized epilepsy were excluded as were those with a history of mental handicap or
psychoses. Patients with TLE with and without a history of IED diagnosed accord-
ing to DSMIV criteria were included in the study.
A neurological and psychiatric history and examination were obtained, as well as
routine EEG investigations and neuropsychological investigations. Full (FIQ), verbal
(VIQ), and performance IQ (PIQ) were measured and patients with a FIQ below 70
were excluded from the study to avoid selection bias. All patients were asked to ll in
the Beck Depression Inventory (BDI13) and the State Trait Anxiety Inventory
(STAI). Both questionnaires are self-rating instruments for depression and anxiety
respectively (Thomson, 1989a, b). In order to assess aggression, carers were asked to
ll in the Social Dysfunction and Aggression Scale (SDAS21), a well validated and
recommended questionnaire (European Rating Aggression Group, 1992; Mak and de
Konning, 1995). Twenty healthy volunteers, matched for age and sex were scanned
and measured twice in order to assess the reliability of the volumetric measurements.
The MRI images were obtained at the Chalfont Centre for Epilepsy on a 1.5 T GE
Signa scanner (GE Medical Systems, Milwaukee, USA) using a T1weighted
inversion-recovery prepared volume acquisition (IRSPGR: TI/TR/TE/ip
450/15/4.2/20; 121.5 mm thick contiguous coronal slices; matrix 256192,
24 cm18 cm FOV), and a conventional spin echo sequence (TR/TE1/TE2/NEX
2000/30/120/1, 256192 matrix, 2418 cm FOV, 5 mm slices) for computation
of T2 values. Volumetric measurements were performed using a locally developed
interactive software program MRreg (Lemieux et al., 1998; Moran et al., 1999) fol-
lowing the established protocol described by Watson et al. (1992). The rater (LTVE)
was blind to the subject grouping. The amygdala volumes were corrected for total
brain size by division by the intracranial volume. Intrarater reliability gures were
carefully assessed and proved to be satisfactory. Amygdala atrophy was dened as a
volume smaller than 3 standard deviations (SD) below the average amygdala
volume of the control group. Amygdala T2 values were measured using DispImage
image analysis software (Plummer, 1992) by placing the largest possible elliptic
region of interest within the amygdala while avoiding anatomical boundaries.
Amygdala T2 signals were dened as pathological if they were greater than 2 SD
above the mean of the control population. Details of the methodology have been
published elsewhere (Tebartz van Elst et al., 1999, 2000b).
The demographic data of both groups are summarized in Table 7.3. The two
patient groups were matched for age, sex, demographic background, duration of
epilepsy and seizure severity. There was no signicant group dierence regarding
the history of birth complications, febrile convulsions or status epilepticus.
However, the incidence of encephalitic brain disease (Fishers Exact Test: P0.05)
and left-handedness (Chi-square Test: P0.05) was signicantly increased in
aggressive patients. There was less right-sided focal EEG abnormality and more
bilateral EEG abnormality in the aggressive group compared to nonaggressive
patients with TLE. Left- as well as right-sided hippocampal sclerosis was signi-
cantly less common in patients with TLE and IED. Other left temporal pathology,
including three patients with amygdala pathology (amygdala sclerosis, amygdala
glioma, amygdala DNT), two patients with multiple small temporal infarctions and
two patients with diuse left temporal atrophy of unknown origin, was signi-
cantly more common in patients with TLE plus IED (see Table 7.4).
Table 7.3. Demographic and historical data of patient groups with temporal lobe epilepsy
(TLE), with TLE and intermitttent explosive disorder (IED) or without IED (TLE alone)
(n 25 in each group)
TLE and IED TLE alone
Age (in years) [range] 30.1 [1849] 33.8 [1956]

Sex: f/m 8/17 10/15
Work: number unemployed 21 15
Living: number living independently 14 10
Income: number on social support 18 16
Social: number living in stable relationship 8 8
Therapy: monotherapy polytherapy 322 322
Mean duration of TLE (in years) [range] 22.4 [545] 24.5 [746]
Mean seizure frequency (estimated frequency 13.4 [0.560] 21 [1.5190]
per month) [range]
Birth complications 9 7
Febrile convulsions 5 9
Status epilepticus 0 2
Left-handed 7 1
History of encephalitis 5 0
Source: Tebartz van Elst et al. (2000a).
Table 7.4. Radiological MRI pathology on visual assessment
Other left
Overall signicance: No Right Left Bilateral temporal
P0.002 pathology hemisphere hemisphere hemispheres pathology
TLE and IED 10 1 4 3 7

TLE alone 6 8 10 1 0
Signicance ** ** **
Notes:
TLE, temporal lobe epilepsy; IED, intermittent explosive disorder.
Other left temporal pathology: amygdala sclerosis1, amygdala glioma1, amygdala DNT
1; multiple small temporal infarctions2; generalized left temporal atrophy2.
Closed test procedure: *P0.5, **P0.01).
Source: Tebartz van Elst et al. (2000b).
There was no evidence of an increased amygdala T2 relaxation time in the

aggressive group. Using the denition of amygdala T2 pathology as a T2 time
greater than 2 SD above the mean of the control group, there was no signicant
group dierence in amygdala pathology.
A group comparison did not reveal a signicant overall dierence in amygdala
volumes (right side: aggressive1893 mm3, SD435 mm3; nonaggressive1909
mm3, SD231 mm3; left side: aggressive1840 mm3, SD398 mm3; nonaggres-
sive1868 mm3, SD290 mm3). However, in the aggressive patients, a subgroup
of ve patients (20%) showed amygdala atrophy as compared to only one in the
nonaggressive group (Chi-square: P0.04). Two patients had left-sided amygdala
atrophy, two had bilateral atrophy and the only patient who exhibited right-sided
amygdala atrophy was left-handed. An increased incidence of encephalitis (Chi-
square Test: P0.05; Fishers Exact Test: P0.1) was the only clinical feature that
distinguished patients with amygdala atrophy from those with normal amygdala
volumes. Since there was no overlap between the ve patients with severe amygdala
atrophy and those seven patients with other amygdala pathology, in a total number
of 12 out of 25 aggressive patients there was some evidence of amygdala pathology
as compared to only one in the nonaggressive group (see Figure 7.1).
There was a highly signicant group dierence in IQ gures with the verbal IQ
(VIQ), the performance IQ (PIQ) and hence the full IQ (FIQ) all being lower in the
aggressive group (Table 7.5). The verbal IQ diered more than the performance IQ.
As expected there was a highly signicant group dierence in the Social
Dysfunction and Aggression Scores (SDAS 9, SDAS 21) since this was the criterion
for group denition (Table 7.5). There was a signicant group dierence in BDI
and STAI scores with the aggressive group rating much higher in depression
(P0.05) state (P0.05) and trait anxiety (P0.01).
Cortical abnormalities in patients with TLE and IED

In order to detect possible subtle cortical brain pathology, that was not present on
visual assessment of the MRI scans, in a second approach we analysed the same
groups of patients using a voxel-by-voxel-comparison of cortical grey matter. After
automated segmentation of cerebral grey matter from T1-weighted MRI, the objec-
tive technique of statistical parametric mapping (SPM) was applied to study the
patient groups described above and comparison made with 35 healthy control sub-
jects (Woermann, 1999). Both TLE patient groups were compared with each other
and with the control subjects on a voxel-by-voxel basis for increases and decreases
of grey matter. SPM 96 characterizes signicant regional dierences in image
parameters, while allowing for global dierences to be taken into account. Details
of the methodology are published elsewhere (Woermann, 1999; Woermann et al.,
1999). The resulting signicant dierences through the 3D image space were dis-
Figure 7.1. Amygdala pathology in patients with TLE with and without IED. AGG aggressive.
played collapsed into three orthogonal planes (glass brain, see Figure 7.2). Regions
of signicant dierence were overlaid on normalized T1-weighted images to facil-
itate correlation with anatomy (see Figure 7.3).
In patients with TLE with IED compared as a group with healthy control sub-
jects, reductions of grey matter were found over large areas of the left extra-
temporal neocortex, maximal in the left frontal neocortex; one maximum
dierence projection had a Z score of 5.67 at Talairach coordinates x58, y36, z
9 mm (left anterior frontolateral cortex), the other a Z score of 4.78 in a more pos-
terior left frontal lobe location (Talairach coordinates x66, y0, z28 mm,
Figures 7.2 and 7.3). Patients with TLE who did not have IED showed no signicant
decrease of cortical grey matter compared with control subjects. Patients with TLE
with IED also had reduction of left frontal grey matter, compared with patients with
TLE without IED, although less marked than when compared with control subjects
(Z score3.49 at Talairach coordinates x66, y2, z26 mm). The SPM-based
Table 7.5. Neuropsychological and psychometric parameters
Variable TLE and IED TLE alone Group comparison
Mean FIQ (SD) 80.6 (8.5) 93.0 (13.9) ** (P0.000)

Mean VIQ (SD) 81.0 (8.6) 93.8 (14.1) ** (P0.000)
Mean PIQ (SD) 83.3 (11.8) 94.7 (15.1) ** (P0.005)
Mean BDI (SD) 8.8 (4.8) 4.2 (5.8) ** (P0.007)
Mean S-STAI 44.1 (14.9) 32.1 (10.9) ** (P0.005)
Mean T-STAI 50.1 (10.1) 34.7 (11.6) ** (P0.000)
Mean SDAS-9 (SD) 14.9 (7.5) 0.4 (0.7) ** (P0.000)
Mean SDAS-21 (SD) 30.9 (12) 3.1 (4.5) ** (P0.000)
Notes:
TLE, temporal lobe epilepsy; IED, intermittent explosive disorder; FIQ, full IQ; VIQ, verbal IQ;
PIQ, performance IQ; BDI, Beck Depression Inventory; STAI, State Trait Anxiety Inventory;
SDAS, Social Dysfunction and Aggression Scale; SD, standard deviation.
*P0.5, **P0.01 after Bonferroni correction.
Source: Tebartz van Elst et al. (2000a).
voxel-wise correlation of SDAS scores and automatically segmented grey matter in

all patients with TLE showed the left frontal grey matter area as negatively corre-
lated with these scores which expressed social consequences of interictal aective
aggression (Z score 3.65 at Talairach coordinates x66, y2, z26 mm). Age,
scores of depression and anxiety, IQ measures, or scores of verbal uency did not
signicantly correlate with specic decreases in grey matter in all patients with TLE.
Dual brain pathology in patients with affective aggression and epilepsy
In our studies we have demonstrated that amygdala-related brain pathology could

be recognized in about half of the patients with TLE and IED. Contrary to our
hypothesis however, there was no evidence for increased mesial temporal lobe
sclerosis or amygdala sclerosis in the aggressive patient group. Brain pathology in
patients with epilepsy and aggression was more diverse in nature and more diuse
in distribution. Interestingly, we found an increased prevalence of a history of
encephalitis in patients with epilepsy plus aggression. Encephalitis in the past might
have been a pathogenic element for the more diuse and widespread pathology in
the temporal lobes seen in our aggressive patients. Furthermore the increased prev-
alence of left-handedness in our aggressive patients may indicate early brain
pathology like encephalitis aecting the left hemisphere (i.e. lateralization of dom-
inance to the right hemisphere).
100
80
60
40
20
TLE with IED controls
Figure 7.2. Glass brain view of decreased grey matter in 24 patients with TLE with episodes of
affective aggression compared with 35 control subjects; displayed after correction for
multiple comparisons (Woermann et al., 2000).
Eleven of twelve patients with amygdala-related brain pathology displayed this

pathology on the left-hand side and the only patient with right-sided amygdala
atrophy alone was left-handed. Thus the dominant hemisphere seems to play a
more important role in the mediation of aective aggression than the nondomi-
nant hemisphere.
The nding of frontal cortical grey matter loss was clearly lateralized too.
Patients with TLE plus IED displayed highly signicant left frontal grey matter loss
that correlated with the SDAS scores. This, together with the left lateralized nding
of amygdala-related brain pathology, could support a theory of dual brain pathol-
ogy in patients with IED (see Figure 7.4): pathology within the amygdala or amyg-
daloid circuits might result in hyperarousal states where patients become angry and
Figure 7.3. Area of decreased grey matter (overlaid on normalized T1-weighted MRI) comparing 24
patients with TLE with episodes of affective aggression with 24 patients with TLE without
such episodes. At the location of maximum difference (Talairach coordinates: x66,
y2, z26 mm) the Z-score was 3.49 (Woermann et al., 2000).
aroused without a sucient external stimulus (hyperarousal syndrome). This dys-

functional arousal resulting in aggressive behavioural impulses normally can be
suppressed by learned behavioural rules. However, associated with additional
frontal lobe pathology, the capacity of the aected patients to suppress behavioural
impulses arising from the emotional brain is limited and this leaves the patients
vulnerable to the development of hyperarousaldyscontrol syndromes i.e. episodic
dyscontrol or intermittent explosive disorder.
This theory is in line with earlier functional imaging and MR spectroscopy studies
showing a reduced prefrontal glucose metabolism in murderers and signicantly
lower neuronal markers in the frontal lobes of repetitively violent patients with learn-
ing disabilities, although without clear lateralizing eects (Raine et al., 1997, 1998).
From the fact that in our patient sample frontal as well as limbic brain pathol-
ogy was clearly lateralized to the left i.e. the dominant hemisphere, one might con-
clude that only dual brain pathology of the dominant hemisphere results in
hyperarousaldyscontrol syndromes while a crossover dual brain pathology with
(Experience)
= hyperarousal Evaluation = dyscontrol
Cognitive
Primary Unimodal Polymodal Frontal

cortex cortex cortex cortex
Hypothalamus
Thalamus Hipp.
Sensory
input Amygdala Brainstem
Emotional
Input Output
(Perception) (Behaviour)
Figure 7.4. Dual brain pathology in episodic dyscontrol a theory of intermittent explosive disorder
as hyperarousaldyscontrol syndrome. Hipp, hippocampus.
pathology of the right frontal lobe or the right amygdala does not produce pheno-
menological states like this. However, this assumption is very speculative and
further investigations, possibly comparing patient groups with left- and right-sided
dual brain pathology may help clarify this question.
There is a controversy regarding the importance of hemispheric specialization
for aggressive behaviour (Bear, 1983; Nachson, 1991) with the majority of studies
pointing to a more important role of the left hemisphere (Saver et al., 1996). Our
MRI and neuropsychological ndings (i.e. particularly low verbal IQ) support this
assumption, which has been reported earlier (Herzberg and Fenwick, 1988).
It is important to note that in our sample there was a strong link between aggres-
sion and high levels of depression and anxiety, conrming other reports of such an
association in the nonpsychiatrically ill population (Bjork et al., 1997). It seems
plausible that high levels of anxiety result in states of hyperarousal that might be
facilitated by amygdala pathology as suggested by other authors (Cendes et al.,
1994). Regarding the relationship between depression and aggression there are only
few and nonconclusive reports (Braconnier and Jeanneau, 1997). Our ndings
point to a clear association between depression and IED in TLE.
Social and psychological aspects of aggression in epilepsy
Even though our studies concentrated on research into the neurobiological basis of
aggressive behaviour in patients with epilepsy, there is no question about the critical
role of social and psychological intervening variables for the development of aggres-
sive behaviour in general, be it in the context of epilepsy or not. Social disadvantage,
prejudice, poor housing, poverty, poor communication skills, all are factors that
make hyperarousal states and states of discontentment and anger more likely and
thus might increase the probability of aggressive behaviour. On the other hand a
precise and correct diagnosis always is the rst step in managing dicult behaviour,
and sociological, psychological and neurobiological views of the same problem are
not necessarily contradictory.
For example, we were able to demonstrate a close link between episodic dyscon-
trol, reduced IQ, depression and anxiety. Even though disentangling the complex
interaction between these dierent psychobiological elements is very dicult, it
nevertheless may suggest a possible way of treatment, irrespective of which of these
elements is the most important one from an aetiological point of view.
Treatment of aggression in epilepsy
If aggression is a problem in the clinical management of patients with epilepsy the

most important point is to establish a correct diagnosis (Figure 7.5). A careful
neurological, psychiatric and medical history and examination should be per-
formed to answer the following questions: 1. Is there any medical condition that
contributes to the aggressive behaviour such as endocrinological or immunologi-
cal diseases? Is there any medication that might contribute to the aggressive behav-
iour? 2. What is the correct neurological diagnosis? Are there any other cerebral
problems in addition to the epilepsy? 3. Are there any psychiatric diagnoses which
possibly are independent of the epilepsy, like bipolar disease or antisocial person-
ality disorder? If the epilepsy started early in life it is in fact often impossible to
establish if, for example, a clinical picture that fulls the criteria for an antisocial
personality disorder has to be judged as independent of the organic brain disease
indicated by the epilepsy or alternatively is an organic personality disorder. 4.
Finally, a careful behavioural analysis and possibly video-telemetry should clarify
if the aggressive behaviour is ictal, postictal or interictal and whether it occurs in
the context of altered states of consciousness or psychosis.
Following syndromatic and possibly nosological diagnosis, treatment should be
causal if possible i.e. intervening medical problems like endocrinological disorders
should be treated adequately. Neurological syndromes like the epilepsy itself should
be treated eectively but, as with all drugs that inuence cerebral functioning, the
question as to whether anticonvulsants might contribute to the behavioural
problem should be considered. For example benzodiazepines are well known to
have paradoxical eects in a minority of patients and may cause states of arousal
and aggression (Binder, 1987; Daderman, 1999; Marcus, 1995; Sheth, 1994).
Likewise phenobarbital is well known to cause behavioural problems with aggres-
Medical Neurological Psychiatric

history and exam history and exam history and exam
Neuropsychiatric
diagnosis
Behavioural analysis and diagnosis of
aggressive syndrome (ictal, postictal, interictal)
Treatment of medical, neurological and

psychiatric condition (look for depression and anxiety)
Possibly avoid Possibly avoid Causal

tricyclic antidepressants convulsant antipsychotics treatment
Consider paradox Consider psychotherapy,

side effects social support
Prophylactic Psychotherapy, social support,

interventions: lithium, SSRIs, antipsychotics
Symptomatic
treatment
Management of Benzodiazepines,
acute aggression: antipsychotics
Figure 7.5. Therapeutic guidelines for the treatment of aggression in patients with epilepsy.
sion in a considerable subgroup of patients with epilepsy often with learning dis-
ability (File, 1990). Besides, in individuals any given drug might have dierent
eects than those described in large groups and thus a careful behavioural analysis
of the sequence of events is the only way to establish any possible side eect, for
example of antiepileptic drugs.
Care should be taken to establish signs of depression or anxiety, since a close link
between these psychopathological states and aective aggression in epilepsy has been
established. Both should be treated medically and with psychotherapy at the same
time (Goldstein, 1997; Lorenzen, 1973). Behavioural therapy in particular in patients
with epilepsy and learning disability has been proven to be very eective (Davis,
1984; Holzapfel, 1998; Rapport, 1983). In the medical treatment of depression in
patients with epilepsy, SSRIs or other new antidepressants like venlafaxine should be
preferred to the old tricyclic antidepressants (TCA) since the latter are more likely to
provoke seizures (Blumer, 1997; Lambert, 1999). In fact, an anticonvulsant eect of
SSRIs is well documented in animal models of epilepsy (Browning, 1997; Lu, 1998;
Pasini, 1996; Wada, 1995) and is also described in humans (Favale, 1995).
Following treatment of all medical, neurological and psychiatric conditions that
may or may not contribute to the aggressive psychopathology a symptomatic
treatment of the aggression is mandatory. However, this depends on whether the

aggression is an ictal, postictal or interictal phenomenon.
Ictal aggression does not need symptomatic treatment but one might consider
interrupting a nonconvulsive status for example with benzodiazepines. Apart from
that, a patient who displays agitation and aggression during a seizure should not be
restricted since defensive violence is more common in such situations and the
aggressive behaviour is self-limited, as is the seizure.
The same is true for postictal confusional states, and even aggressive behaviour
in the context of postictal psychosis is self-limited. However, if the aggression is
severe and disturbing or self-harming, medical treatment with benodiazepines like
diazepam or clobazan and/or antipsychotics, for example risperidone, olanzapine
or quetiapine should be given. Good seizure control, avoiding severe complex
partial or secondary generalized seizures, is the best prophylactic intervention since
postictal confusional and psychotic states are more common after severe seizures.
In interictal aggression, prophylactic and acute symptomatic treatment of
aggression and agitation should be dierentiated. For the treatment of acute hyper-
arousaldyscontrol syndromes a combination of benzodiazepines like diazepam
and antipsychotics like haloperidol or sulpiride are still the most eective and safest
interventions. In cases of interictal psychosis however, the antipsychotic medica-
tion should eventually be switched to one of the atypical antipsychotic agents with
little proconvulsant potential like risperidone, olanzapine or quetiapine, since these
drugs are better tolerated. A good control of the psychosis is the best way to prevent
aggression if it is part of the psychosis.
In the case of interictal aggressive syndromes like IED there is no well-established
medical prophylactic therapy; however, there are many anecdotal reports of eec-
tive use of substances like lithium, valproate, carbamazepine, antipsychotics, beta-
blockers, clonidine and even psychostimulants (Fava, 1997; Grith, 1985;
Yudofsky, 1990). However, since there are hardly any well-conducted systematic
treatment studies at the moment the medical treatment still is very experimental
and single-case driven. Nevertheless, in the light of the very severe burden that is
put on patients and their relatives and caregivers by the sometimes devastating
behavioural episodes, a systematic trial of these agents seems to justied in special
cases. The use of cognitivebehaviour therapy for anger management should also
be considered, either as a rst-choice treatment, or combined with psychotropic
drugs.
Conclusion
Research into the neurobiological basis of aggression is still hampered by the di-
culty in dening phenomenological and nosological homogeneous study groups.
Nevertheless, it is important to develop a more precise understanding of the

complex interplay of social, psychological and neurobiological factors all contrib-
uting to the aggressive and violent behaviour. Aggression in epilepsy is rare.
However, if it occurs it imposes an immense burden on the patient, relatives and
caregivers. Clinically, it is crucial to rst of all establish a correct diagnosis.
Intervening medical, neurological and psychiatric disorders, in particular depres-
sion and anxiety, should be recognized and treated adequately. A correct syndro-
matic diagnosis of the aggressive syndrome and its relation to the seizures should
be made. Causal treatment should aim at any underlying medical, neurological or
psychiatric disorders. In symptomatic treatment of aggressive outbursts, treatment
of acute aggression and prophylactic treatment should be dierentiated. In acute
arousal states, if possible, patients should be left alone if aggression is a symptom
of a seizure since restrictive aggression is the most common form in this situation.
In cases of postictal or interictal aggression, it is important to establish a possible
relationship to postictal or interictal psychosis and treat adequately. For interictal
syndromes of aective aggression like IED (episodic dyscontrol), there are no gen-
erally established treatment protocols. However, tentative treatment with dierent
substances as mentioned above seems justied in single cases. Furthermore,
psychotherapeutic and social interventions may help prevent arousal states and
thus can be very eective.
R E F E R E N C ES
Aggleton, J.P. (1993). The contribution of the amygdala to normal and abnormal emotional
states. Trends Neurosci, 16, 32833.
Albert, D.J., Walsh, M.L. and Jonik, R.H. (1993). Aggression in humans: what is its biological
foundation? Neurosci Biobehav Rev 17, 40525.
Alheid, G.F., De Olmos, J.S. and Beltramino, C.A. (1995). Amygdala and extended amygdala. In
The Rat Nervous System, ed. G Paxinos, 495578. San Diego: Academic Press.
Amaral, D.G., Price, J.L., Pitknen, A. and Carmichael, S.T. (1992). Anatomical organisation of
the primate amygdaloid complex. In The Amygdala: Neurobiological Aspects of Emotion,
Memory, and Mental Dysfunction, ed. J.P. Aggleton, pp. 166. New York: Wiley-Liss.
(Fourth Revision) (DSMIV). Washington, DC: APA.
Andy, O.J. and Jurko, M.F. (1972). Hyperresponsive syndrome. In Psychosurgery, ed. E.
Hitchcock, L. Laitinen and K. Vaernet, pp. 11726. Springeld, ILL: Charles C. Thomas.
Bach-Y-Rita, G., Lion, J.R., Climent, C.E. and Ervin, F.R. (1971). Episodic dyscontrol: a study of
130 violent patients. Am J Psychiatry, 127, 14738.
Barratt, E.S., Stanford, M.S., Kent, T.A. and Felthous, A. (1997). Neuropsychological and cogni-
tive psychophysiological substrates of impulsive aggression. Biol Psychiatry, 41, 104561.
Bear, D. (1983). Hemispheric specialisation and the neurology of emotion. Arch Neurol, 40,
195202.
Behbehani, M.M. (1995). Functional characteristics of the midbrain periaquaeductal gray. Prog
Neurobiol, 46, 573605.
Bhatnagar, S. and Dallman, M. (1998). Neuroanatomical basis for facilitation of hypothalamic-
pituitary-adrenal responses to a novel stressor after chronic stress. Neuroscience, 84, 102539.
Binder, R.L. (1987). Three case reports of behavioral disinhibition with clonazepam. Gen Hosp
Bjork, J.M., Dougherty, D.M. and Moeller, F.G. (1997). A positive correlation between self-
ratings of depression and laboratory-measured aggression. Psychiatry Res, 69, 338.
Blumer, D. (1997). Antidepressant and double antidepressant treatment for the aective disor-
der of epilepsy. J Clin Psychiatry, 58, 311.
Braconnier, A. and Jeanneau, A. (1997). Anxiety, aggression, agitation and depression: psycho-
pathologic aspects. Encephale, 23, 437.
Brandao, M.L., Cardoso, S.H., Melo, L.L., Motta, V. and Coimbra, N.C. (1994). Neural substrate
of defensive behavior in the midbrain tectum. Neurosci Biobehav Rev, 18, 33946.
Browning, R.A. (1997). Enhancement of the anticonvulsant eect of uoxetine following block-
ade of 5-HT1A receptors. Eur J Pharmacol, 336, 16.
Cendes, F., Andermann, F., Gloor, P. et al. (1993). MRI volumetric measurement of amygdala and
hippocampus in temporal lobe epilepsy. Neurology, 43, 71925.
Cendes, F., Andermann, F., Gloor, P. et al. (1994). Relationship between atrophy of the amygdala
and ictal fear in temporal lobe epilepsy. Brain, 117, 73946.
Chapman, W.P., Schroeder, H.R., Geyer, G. et al. (1954). Physiological evidence concerning
importance of amygdala nuclear region in the integration of circulatory functions and
emotion in man. Science, 120, 94954.
Charney, D.S. and Deutch, A. (1996). A functional neuroanatomy of anxiety and fear: implications
for the pathophysiology and treatment of anxiety disorders. Crit Rev Neurobiol, 10, 41946.
Currie, S., Heatheld, W., Henson, R. and Scott, D. (1971). Clinical course and prognosis of tem-
poral lobe epilepsy: a survey of 666 patients. Brain, 94, 17390.
Daderman, A.M. (1999). Flunitrazepam (Rohypnol) abuse in combination with alcohol causes
premeditated, grievous violence in male juvenile oenders. J Am Acad Psychiatry Law, 27,
8399.
Damasio, A.R., Tranel, D. and Damasio, H. (1990). Individuals with sociopathic behavior caused
by frontal damage fail to respond autonomically to social stimuli. Behav Brain Res, 41, 8194.
Davis, G.R. (1984). Cognitive-behavioral treatment of depressed aect among epileptics: prelim-
inary ndings. J Clin Psychology, 40, 9305.
Davis, M., Rainnie, D. and Cassell, M. (1994). Neurotransmission in the rat amygdala related to
fear and anxiety. Trends Neurosci, 17, 20814.
Delgado-Escueta, A.V., Mattson, R.H., King, L. et al. (1981). Special report. The nature of aggres-
sion during epileptic seizures. N Engl J Med, 305, 71116.
Devinsky, O. and Bear, D. (1984). Varieties of aggressive behavior in temporal lobe epilepsy. Am
J Psychiatry, 141, 6516.
Dicks, P., Myers, R.E. and Kling, A. (1969). Uncus and amygdala lesions: eects on social behav-
ior in the free ranging rhesus monkey. Science, 165, 6971.
Duncan, J.S., Bartlett, P. and Barker, G.J. (1996). Technique for measuring hippocampal T2 relax-
ation time. Am J Neuroradiol, 17, 180510.
Elliot, F.A. (1982). Neurological ndings in adult minimal brain dysfunction and the dyscontrol
syndrome. J Nerv Ment Dis, 170, 6807.
Elliott, F.A. (1984). The episodic dyscontrol syndrome and aggression. Neurol Clin, 2, 11325.
Elliot, F.A. (1992). Violence the neurologic contribution: an overview. Arch Neurol, 49,
595603.
European Rating Aggression Group (1992). Social dysfunction and aggression scale (SDAS21)
in generalised aggression and in aggressive attacks: a validity and reliability study. Int J Methods
Psychiatric Res, 2, 1529.
Falconer, M.A. (1973). Reversibility by temporal-lobe resection of the behavioral abnormalities
of temporal-lobe epilepsy. N Engl J Med, 289, 4515.
Fava, M. (1997). Psychopharmacologic treatment of pathologic aggression. Psychiatr Clin N Am,
20, 42751.
Favale, E. (1995). Anticonvulsant eect of uoxetine in humans [see comments]. Neurology, 45,
19267.
Fenwick, P.B.C. (1986). Aggression and epilepsy. In Epilepsy and Psychiatry, ed. M.R. Trimble and
T. Bolwig, pp. 3160. Chichester: John Wiley & Sons.
Fenwick, P. (1989). The nature and management of aggression in epilepsy. J Neuropsychiatry Clin
Neurosci, 1, 41825.
File, S.E. (1990). Changes in seizure threshold and aggression during chronic treatment with
three anticonvulsants and on drug withdrawal. Psychopharmacology, 100, 23742.
Gallagher, M. and Chiba, A.A. (1996). The amygdala and emotion. Curr Opin Neurobiol, 6,
2217.
Gastaut, H., Morrin, G. and Lesevre, N. (1955). Etudes du compartment des epileptiques psycho-
moteur dans Linterval de leurs crises. Ann Med Psychol, 113, 129.
Gerard, M.E. (1998). Subacute postictal aggression [see comments]. Neurology, 50, 3848.
Geschwind, N. (1975). The clinical setting of aggression in temporal lobe epilepsy. In Neural Basis
of Violence and Aggression, ed. W.S. Fields and W.H. Sweet, pp. 27381. St. Louis, MI: Warren
H. Green.
Gloor, P. (1991). Mesial temporal sclerosis: historical background and an overview from a
modern perspective. In Epilepsy Surgery, ed. H.O. Luders, pp. 689703. New York: Raven Press.
Gloor, P., Olivier, A., Quesney, L.F., Andermann, F. and Horowitz, S. (1982). The role of the
limbic system in experiential phenomena of temporal lobe epilepsy. Ann Neurol, 12, 12944.
Goldstein, L.H. (1997). Eectiveness of psychological interventions for people with poorly con-
trolled epilepsy. J Neurol Neurosurg Psychiatry, 63, 13742.
Goldstein, M. (1974). Brain research and violent behavior: a summary and evaluation of the
status of biomedical research on brain and aggressive violent behavior. Arch Neurol, 30, 135.
Grith, J.L. (1985). Treatment of episodic behavioral disorders with rapidly absorbed benzodi-
azepines. J Nerv Ment Dis, 173, 31215.
Gunn, J. (1971). Epilepsy, automatism, and crime. Lancet, 1, 11736.

Gunn, J. (1977). Criminal behaviour and mental disorder. Br J Psychiatry, 130, 31729.
Halgren, E. (1992). Emotional neurophysiology of the amygdala within the context of human
cognition. In The Amygdala: Neurobiological Aspects of Emotion, Memory, and Mental
Dysfunction, ed. J.P. Aggleton, pp. 191228. New York: Wiley-Liss.
Herzberg, J.L. and Fenwick, P.B .(1988). The aetiology of aggression in temporal-lobe epilepsy.
Br J Psychiatry, 153, 505.
Holzapfel, S. (1998). Behavioral psychophysiological intervention in a mentally retarded epilep-
tic patient with brain lesion. Appl Psychophysiol Biofeedback, 23, 189202.
Hudson, L.P., Munoz, D.G., Miller, L., McLachlan, R.S., Girvin, J.P. and Blume, W.T. (1993).
Amygdala sclerosis in temporal lobe epilepsy. Ann Neurol, 33, 62231.
Jackson, G.D., Berkovic, S.F., Tress, B.M. et al. (1990). Hippocampal sclerosis can be reliably
detected by magnetic resonnance imaging. Neurology, 40, 186975.
Kalin, N.H. (1999). Primate models to understand human aggression. J Clin Psychiatry, 60
(Suppl. 15), 2932.
Kalviainen, R., Salmenpera, T., Partanen, K., Vainio, P., Riekkinen, P.S. and Pitkanen, A. (1997).
MRI volumetry and T2 relaxometry of the amygdala in newly diagnosed and chronic temporal
lobe epilepsy. Epilepsy Res, 28, 3950.
Kanemoto, K. (1996). Postictal psychosis: a comparison with acute interictal and chronic psycho-
ses. Epilepsia, 37, 5516.
Kanemoto, K. (1999). Violence and epilepsy: a close relation between violence and postictal
psychosis. Epilepsia, 40, 1079.
Kim, J.H., Tien, R.D., Felsberg, G.J. et al. (1995). Fast spin-echo MR in hippocampal sclerosis:
correlation with pathology and surgery. Am J Neuroradiol, 16, 62736.
Kligman, D. and Goldberg, D.A. (1975). Temporal lobe epilepsy and aggression. J Nerv Ment Dis,
160, 32441.
Kling, A.S. and Brothers, L.A. (1992). The amygdala and social behavior. In The Amygdala:
Neurobiological Aspects of Emotion, Memory, and Mental Dysfunction, ed. J.P. Aggleton, pp.
35377. New York: Wiley-Liss.
Krakowski, M. and Czobor, P. (1997). Violence in psychiatric patients: the role of psychosis,
frontal lobe impairment, and ward turmoil. Compr Psychiatry, 38, 2306.
LaBar, K.S., Ledoux, J.E., Spencer, D.D. and Phelps, E.A. (1995). Impaired fear conditioning fol-
lowing unilateral temporal lobectomy in humans. J Neurosci, 15, 684655.
Lambert, M.V. (1999). Depression in epilepsy: etiology, phenomenology, and treatment.
Lancman, M. (1999). Psychosis and peri-ictal confusional states. Neurology, 53, S338.
LeDoux, J.E. (1995). Emotions: clues from the brain. Annu Rev Psychol, 46, 20935.
Leicester, J. (1982). Temper tantrums, epilepsy and episodic dyscontrol. Br J Psychiatry, 141,
2626.
Lemieux, L., Wieshmann, U.C., Moran, N.F., Fish, D.R. and Shorvon, S.D. (1998). The detection
and signicance of subtle changes in mixed-signal brain lesions by serial MRI scan matching
and spatial normalisation. Med Image Analysis, 2, 22742.
Lishman, W.A. (1998). Organic Psychiatry The Psychological Consequences of Cerebral Disorder.
Oxford: Blackwell Science.
Logsdail, S.J. and Toone, B.K. (1988). Post-ictal psychoses. A clinical and phenomenological
Lorenzen, D. (1973). Initial results of behavior therapy in Psychiatric Regional Hospital
Weinsberg. [In German]. Nervenarzt, 44, 4237.
Lu, K.T. (1998). Endogenous serotonin inhibits epileptiform activity in rat hippocampal CA1
neurons via 5hydroxytryptamine 1A receptor activation. Neuroscience, 86, 72937.
Luo, B. (1998). Cholecystokinin B receptors in the periaqueductal gray potentiate defensive rage
behavior elicited from the medial hypothalamus of the cat. Brain Res, 796, 2737.
Mak, M. and De Konning, P. (1995). Clinical research in aggressive patients: pitfalls in study design
and measurements of aggression. Progr Neuropsychopharmacol Biol Psychiatry, 19, 9931017.
Maletzky, B.M. (1973). The episodic dyscontrol syndrome. Dis Nerv System, 34, 17885.
Marcus, A. (1995). Benzodiazepine administration induces exogenic psychosis: a case of child
abuse. Child Abuse Neglect, 19, 8336.
Margerison, J.H. and Corsellis, J. (1966). Epilepsy and the temporal lobe: a clinical, electroence-
phalographic and neuropathological study of the brain in epilepsy, with particular reference
to the temporal lobes. Brain, 89, 499530.
Miller, B.L., Darby, A., Benson, D.F., Cummings, J.L. and Miller, M.H. (1997). Aggressive, and
socially disruptive and antisocial behaviour associated with fronto-temporal dementia. Br J
Miller, L.A., McLachlan, R.S., Bouwer, M.S., Hudson, L.P. and Munoz, D.G. (1994). Amygdala
sclerosis: preoperative indicators and outcome after temporal lobectomy. J Neurol, Neurosurg
Moran, N.F., Lemieux, L., Maudgil, D.D., Kitchen, N.D., Fish, D.R. and Shorvon, S.D. (1999).
Analysis of temporal lobe resections in MR images. Epilepsia, 40, 107784.
Moyer, K.E. (1987). Violence and Aggression A Physiological Perspective. New York: Paragon
House Publishers.
Mungas, D. (1983). An empirical analysis of specic syndromes of violent behavior. J Nerv Ment
Dis, 171, 35461.
Nachson, I. (1991). Neuropsychology of violent behavior: controversial issues and new develop-
ments in the study of hemisphere function. In Neuropsychology of Aggression, ed. J.S. Miller,
pp. 93116. Boston: Kluwer Academic Publishers.
Pasini, A. (1996). The anticonvulsant action of uoxetine in substantia nigra is dependent upon
endogenous serotonin. Brain Res, 724, 848.
Petty, R.G., Bonner, D., Mouratoglou, V. and Silverman, M. (1996). Acute frontal lobe syndrome
and dyscontrol associated with bilateral caudate nucleus infarctions. Br J Psychiatry, 168,
23740.
Plummer, D.L. (1992). DispImage, a display and analysis tool for medical images. Revues
Neuroradiol, 5, 48995.
Raine, A., Buchsbaum, M. and Lacasse, L. (1997). Brain abnormalities in murderers indicated by
positron emission tomography. Biol Psychiatry, 42, 495508.
Raine, A., Stoddard, J., Bihrle, S. and Buchsbaum, M. (1998). Prefrontal glucose decits in mur-
derers lacking psychosocial deprivation. Neuropsychiatry, Neuropsychol Behav Neurol, 11, 17.
Ramamurthi, B. (1988). Stereotactic operation in behaviour disorders. Amygdalotomy and
hypothalamotomy. Acta Neurochirurgica Suppl, 44, 1527.
Rapport, M.D. (1983). Carbamazepine and behavior therapy for aggressive behavior. Treatment
of a mentally retarded, postencephalitic adolescent with seizure disorder. Behav Modication,
7, 25565.
Ratner, R.A. and Shapiro, D. (1979). The episodic dyscontrol syndrome and criminal responsibil-
ity. Bull Am Acad Psychiatry Law, 7, 42231.
Reis, D.J. (1969). Brain norepinephrine: evidence that neuronal release is essential for sham rage
behavior following brainstem transection in cat. Proc Nat Acad Sci USA, 64, 10812.
Rodin, E.A. (1973). Psychomotor epilepsy and aggressive behavior. Arch Gen Psychiatry, 28, 21013.
Rolls, E.T. (1992). Neurophysiology and functions of the primate amygdala. In The Amygdala:
Neurobiological Aspects of Emotion, Memory, and Mental Dysfunction, ed. J.P. Aggleton, pp.
14365. New York: Wiley-Liss.
Saver, J.L., Salloway, S.P., Devinsky, O. and Bear, D.M. (1996). Neuropsychiatry of Aggression. In
Neuropsychiatry, ed. B.S. Fogel, R.B. Schier and S.M. Rao, pp. 52348. Baltimore: Williams &
Wilkins.
Shaikh, M.B. (1997). Serotonin 5-HT1A and 5-HT2/1C receptors in the midbrain periaqueduc-
tal gray dierentially modulate defensive rage behavior elicited from the medial hypothalamus
of the cat. Brain Res, 765, 198207.
Sheth, R.D. (1994). Aggression in children treated with clobazam for epilepsy. Clin Neuro-
pharmacol, 17, 3327.
Soininen, H.S., Partanen, K., Pitkanen, A. et al. (1994). Volumetric MRI analysis of the amygdala
and the hippocampus in subjects with age-associated memory impairment: correlation to
visual and verbal memory. Neurology, 44, 16608.
Stein, D.J., Hollander, E., Cohen, L. et al. (1993). Neuropsychiatric impairment in impulsive per-
sonality disorders. Psychiatry Res, 48, 25766.
Stone, J.L., McDaniel, K.D., Hughes, J.R. and Hermann, B.P. (1986). Episodic dyscontrol disor-
der and paroxysmal EEG abnormalities: successful treatment with carbamazepine. Biol
Swartz, M.S., Swanson, J.W., Hiday, V.A., Borum, R., Wagner, H.R. and Burns, B.J. (1998).
Violence and severe mental illness: the eects of substance abuse and nonadherence to medi-
cation. Am J Psychiatry, 155, 22631.
Szabo, C.A. and Lancman, M.S.S. (1996). Postictal psychosis: a review. Neuropsychiatry,
Neuropsychol Behav Neurol, 9, 25864.
Tebartz Van Elst, L., Woermann, F., Lemieux, L. and Trimble, M.R. (1999). Amygdala enlarge-
ment in dysthymia A volumetric study of patients with temporal lobe epilepsy. Biol
Tebartz Van Elst, L., Woermann, F., Lemieux, L., Thompson, P.J. and Trimble, M.R. (2000a).
Aective aggression in patients with temporal lobe epilepsy a quantitative magnetic reso-
nance study of the amygdala. Brain, 123, 23443.
Tebartz Van Elst, L., Woermann, F., Lemieux, L. and Trimble, M.R. (2000b). Increased amygdala
volumes in female and depressed patients with temporal lobe epilepsy. Neurosci Lett, 281,
103106.
Thomson, C. (1989a). Aective disorders. In The Instruments of Psychiatric Research, ed. C.
Thomson, pp. 3446. New York: John Wiley & Sons.
Thomson, C. (1989b). Anxiety. In The Instruments of Psychiatric Research, ed. C. Thomson, pp.
4754. New York: John Wiley & Sons.
Treiman, D.M. (1991). Psychobiology of ictal aggression. In Advances in Neurology, ed. D. Smith,
D. Treiman and M.R. Trimble, Vol. 55, pp. 34156. New York: Raven Press.
Trimble, M.R. (1996) Biological Psychiatry. Chichester: John Wiley & Sons.
Trimble, M.R., Ring, H.A. and Schmitz, B. (1996). Neuropsychiatric aspects of epilepsy. In
Neuropsychiatry, ed. B.S. Fogel, R.B. Schier and S.M. Rao, pp. 771803. Baltimore: Williams
& Wilkins.
Valzelli, L. (1981). Psychobiology of Aggression and Violence. New York: Raven Press.
Van de Poll, N.E. and Van Goozen, S.H.M. (1992). Hypothalamic involvement in sexuality and
hostility: comparative psychological aspects. Progr Brain Res, 93, 34361.
Van Paesschen, W., Connelly, A., Johnson, C.L. and Duncan, J.S. (1996). The amygdala and
intractable temporal lobe epilepsy: a quantitative magnetic resonance imaging study [pub-
lished erratum: Neurology (1997) 48, 1751]. Neurology, 47, 102131.
Vitiello, B. and Sto, D.M. (1997). Subtypes of aggression and their relevance to child psychia-
try. J Am Acad Child Adolesc Psychiatry, 36, 30715.
Wada, Y. (1995). Prolonged but not acute uoxetine administration produces its inhibitory eect
on hippocampal seizures in rats. Psychopharmacology, 118, 3059.
Watson, C., Andermann, F., Gloor, P. et al. (1992). Anatomic basis of amygdaloid and hippocam-
pal volume measurement by magnetic resonance imaging. Neurology, 42, 174350.
Wieser, H.G. (1983). Electroclinical Features of the Psychomotor Seizures. Stuttgart: Fisher
Butterworths.
Woermann, F.G. (1999). Voxel-by-voxel comparison of automatically segmented cerebral gray
matter a rater-independent comparison of structural MRI in patients with epilepsy.
NeuroImage, 10, 37384.
Woermann, F.G., Barker, G.J., Birnie, K.D., Meencke, H.J. and Duncan, J.S. (1998). Regional
changes in hippocampal T2 relaxation and volume: a quantitative magnetic resonance
imaging study of hippocampal sclerosis. J Neurol Neurosurg Psychiatry, 65, 65664.
Woermann, F.G., Tebartz Van Elst, L., Free, S.L., Thompson, P.J., Trimble, M.R. and Duncan, J.S.
(1999). Reduction of frontal neocortical grey matter associated with aective aggression in
patients with temporal lobe epilepsy an objective voxel-by-voxel analysis of automatically
segmented MRI. Epilepsia, 40, 2056. (Abstract.)
Woermann, F., Tebartz Van Elst, L., Koepp, M.J. et al. (2000). Reduction of frontal neocortical
grey matter associated with aective aggression in patients with temporal lobe epilepsy. An
objective voxel-by-voxel analysis of automatically segmented MRI. J Neurol Neurosurg
Yudofsky, S.C. (1990). Pharmacologic management of aggression in the elderly. J Clin Psychiatry,
51 (Suppl.), 2228.
Zanchetti, A. (1968). Reex and brain stem inhibition of sham rage behaviour. Progr Brain Res,
22, 195205.
8
Epilepsy and suicide: a neuropsychiatric

analysis
Dietrich Blumer
University of Tennessee College of Medicine, Memphis, USA
Introduction
According to eight reports, death by suicide occurs in 5% of patients with epilepsy,

compared with 1.4% in the general population (Matthews and Barabas, 1981).
Based on four reports, the four- to vefold increase in suicides among patients with
epilepsy over the rate in the general population is magnied to approximately
25fold among patients with temporal lobe epilepsy (Barraclough, 1987). In a
Danish study covering 14 years (Henriksen et al., 1970), 164 of 2763 patients with
epilepsy died (an excess mortality rate of 273% compared to the number of deaths
expected in the general population). While epilepsy was the cause of death in 26%,
suicide was the cause of death in 20% (an excess mortality rate of 300%) at an
average age at death of 32 years. For studies of an association between epilepsy and
suicide to be statistically valid, the cohort must be standardized by age and sex,
according to Stenager and Stenager (1992); most studies have not met these stan-
dards. The same authors pointed out that the variety of forms of epilepsy necessi-
tates an examination of each syndrome singly for risk of suicide. Their second point
is very signicant, because the data available indicate that suicide is associated par-
ticularly with temporal lobe epilepsy.
Suicide and attempted suicide
Murphy (1994) points out that it is important to distinguish between suicide and
attempted suicide. Dierences between epileptic patients who attempt suicide and
those who complete the act are not established. In the general population, suicides
are three times as likely to be men as women, while attempted suicides are predom-
inantly women at a 2:1 ratio. Suicides are about equally frequent before and after
age 40, and the victims are most commonly suering from a major psychiatric
illness; attempted suicides are mostly under age 40 and are less likely to be suer-
ing from one of the psychiatric illnesses most commonly associated with suicide.
107
108 D. Blumer
Attempted suicide is about 10 times more frequent than suicide. The lifetime risk
of completed suicide after a suicide attempt is estimated to be slightly above 10%.
Suicide is carried out by eective means, while attempted suicides tend to use what
is at hand (8090% of the attempts are by overdose).
Several reports document the high frequency of suicide attempts by overdose
among epileptic patients. The incidence of self-poisoning in epilepsy has been
reported as sevenfold that of the general population (Mackay, 1979). Sixty-ve
percent of suicide attempts were carried out by ingesting antiepileptic drugs, 60%
of which were barbiturates (Hawton et al., 1980). The psychotoxic role of the bar-
biturates in suicide attempts is emphasized by two further studies. In a population
of 126 children and adolescents admitted to an emergency room for suicide
attempt, nine had epilepsy a 15-fold increase compared with the prevalence of
epilepsy in the same age group. Of those nine epileptic children and adolescents,
eight had been treated with phenobarbital (Brent, 1986). A much higher preva-
lence of suicidal ideation (47% vs. 4%) was noted among patients treated with
phenobarbital compared with those treated with carbamazepine (Brent et al.,
1987).
The generally dicult psychosocial circumstances of patients with chronic epi-
lepsy have been considered the leading factor responsible for their elevated suicide
rate, more important even than the presence of psychiatric illness or the availabil-
ity of drugs (Editorial, 1980). However, in general, psychiatric illness has been iden-
tied as the nearly universal antecedent of suicide (Murphy, 1994). Mendez et al.
(1989) studied the causative factors for suicide attempts by overdose in epilepsy and
concluded that interictal psychopathological factors were of primary importance.
A comparison of suicide attempts among patients with epilepsy and comparably
handicapped controls with other chronic disabilities found that 30% of those with
epilepsy had attempted suicide as compared with 7% of the controls (Mendez et al.,
1986). Psychosocial circumstances cannot be considered a sucient precipitant for
suicide attempts or suicide.
In his review of the topic, Diehl (1986) ranked the risk factors for suicide in epi-
lepsy as follows: (1) psychiatric disorders (psychotic episodes, dysphoric episodes,
twilight states, personality disorders); (2) relatively young males (ages 2549 years);
(3) generalized and temporal lobe seizures (with brain lesions); (4) prolonged
duration of the seizure disorder and inadequate therapy; (5) personal, social or
occupational diculties; and (6) availability of large amounts of antiepileptic
drugs.
While a good number of papers list statistical data on the topic, there is a regret-
table paucity of neuropsychiatric case reports that allow for a better understanding
of the psychopathology and pathogenesis specic to suicide in epilepsy and that
might point at methods to prevent the fatal outcome.
109 Epilepsy and suicide
Series of suicides providing neuropsychiatric data
Taylor and Marsh (1977) reported on the occurrence of suicide among 193 patients
who had undergone temporal lobectomy and who were followed for 5 to 24 years.
Of 37 deaths, nine were by suicide (six poisoned themselves). Including an addi-
tional six patients who died in unclear circumstances would have raised the suicide
rate observed to 50-fold of that expected (Taylor, 1987). Five of the nine who de-
nitely had committed suicide had been rendered seizure-free by the surgery. The
mental state of the victims was not described. The paper is of exceptional interest
for the statistics of suicide in epilepsy but is of limited value for our understanding
of the neuropsychiatric problem.
Mendez and Doss (1992) reported on the psychiatric aspects of four patients
who died by suicide out of 1611 patients with epilepsy followed in a neurology
clinic over a period of 8 years: two male patients with chronic psychosis and good
seizure control; one male patient with brief psychotic episodes associated with con-
fusion and increased bitemporal spikes and diuse slowing on EEG in the absence
of seizures; and one female patient with profound ictal and postictal depression
who committed suicide after three witnessed staring spells. The patient with brief
psychotic episodes and one of the patients with chronic psychosis experienced
voices commanding them to commit suicide. All four patients had suered from
complex partial seizures since childhood. All four patients committed suicide by
medication overdose.
Of our entire population treated for epilepsy at the Epi-Care Center in Memphis
over the past 12 years (10739 patients), ve patients have committed suicide. All
had a history of early onset (mean age 9.5 years) of longstanding complex partial
seizures (mean duration 29 years), with very high (often daily) seizure frequency
in all but one. Suicide occurred in all patients after a short interval (mean 13
months) of having obtained full control of seizures for the rst time by temporal
lobectomy (three patients), medication (one patient), or vagus nerve stimulation
(one patient). Three patients had a previous history of suicidal moods or suicide
attempts, but in three of the ve, the suicidal act was precipitate and not anticipated
at the time. Four were male and one was female. None of the deaths was caused by
overdose: one by hanging, one by drowning and three by gunshot.
Role of the interictal and peri-ictal psychopathology in suicide
The premodern psychiatrists who established the basis for our modern classica-
tion of mental disorders noted that specic mental changes were associated with
epilepsy. They had the advantage of observing institutionalized patients with
chronic epilepsy over prolonged periods and were familiar with the characteristic
110 D. Blumer
intermittent and pleomorphic changes that have eluded the modern cross-sectional
psychological assessment of patients with epilepsy. Modern assessments are usually
carried out with methods that have been validated for use in patients who do not
have epilepsy and that are insensitive to the task. Premodern psychiatrists had
arrived at the concept of the dysphoric disorder as the most common psychiatric
disorder of epilepsy, a distinct disorder that has only recently been rediscovered
(Blumer et al., 1995). One of its key symptoms is associated with the episodic sui-
cidal moods of patients with chronic epilepsy, i.e. mesial temporal lobe epilepsy
(Blumer, 2000; Blumer et al., 1995; Gastaut, 1956).
Kraepelin (1923) precisely described the intermittent dysphoric disorder of
patients with epilepsy. Dysphoric episodes present with depressive moods (very
frequently with utter disgust of life and suicidal bent), irritability, anxiety, head-
aches, insomnia or at times with euphoric moods. These polysymptomatic dys-
phoric episodes occur without external triggers with rapid onset and termination
and recur fairly regularly in a uniform manner in the absence of clouding of con-
sciousness. Dysphoric symptoms can be observed as prodromes of an attack or in
the aftermath of an attack, but they most commonly appear as phenomena inde-
pendent of the seizures, with a frequency varying from every few days to every few
months. A patient just awakens one day dysphoric, or the dysphoria develops insid-
iously through the course of a day. As a rule, the dysphoric state lasts from 1 to 2
days but might dissipate after just a few hours. Based on our own observations, we
have added anergia and phobic fears to Kraepelins six key symptoms of the dys-
phoric disorder and have dened it by the presence of at least three of the eight key
symptoms, each present to a troublesome degree. We have noted an average of ve
key symptoms among our patients with dysphoric disorder (Blumer et al., 1995).
The risk of suicide in patients with epilepsy is primarily associated with the epi-
sodes of intense depressive mood that occur during the interictal phase of patients
with a dysphoric disorder, and suicide in epilepsy tends to occur in a precipitate
manner. As already noted by Kraepelin, the dysphoric symptoms also tend to occur
peri-ictally, during the prodrome or in the aftermath of a seizure. The postictal
phase in particular may be associated with marked depressive mood (Blumer,
1992). A high suicidal risk has been observed in patients who experience ictal
depressive mood that extends into the postictal phase for a period of 1 hour to 3
days. Williams (1956) reported 21 such cases among his 100 patients with ictal
emotional experience, and 5 of the 21 patients made suicide attempts during their
postictal phase.
As noted by Kraepelin, interictal psychoses tend to develop among patients
with interictal dysphoric disorder (Blumer et al., 2000; Kraepelin, 1923). The dys-
phoric disorder persists during the psychotic state, and intense depressive moods
may occur in the course of an interictal psychosis. The presence of the hallucina-
tion of voices commanding patients to kill themselves represents a particular sui-

cidal risk.
The psychopathology of four patients (Mendez and Doss, 1992) who commit-
ted suicide is not reported in detail beyond the psychotic episodes in three of them
and the ictal and postictal depression in the fourth patient; the researchers state
only that two of those with psychosis also experienced depressive episodes. Four of
the ve patients from the Epi-Care series had longstanding dysphoric disorders.
The fth patient who had been found free of dysphoric symptoms earlier was not
examined during the 3-month interval from nally (after 20 years) becoming
seizure-free to his death by suicide; it was learned, however, that he had begun to
experience episodes of rage, symptomatic of a de novo dysphoric disorder.
Pathogenesis of the psychiatric disorders of epilepsy
Suicide in epilepsy appears to be caused by the interictal and sometimes by post-

ictal psychopathology. Since the early reports by Gibbs (1951), Landolt (1953), and
Hill (1953), evidence has been increasing that the interictal psychopathology tends
to emerge or to worsen upon improvement of the epilepsy as measured by seizure
frequency and EEG abnormalities. There is persuasive evidence that the psychiat-
ric disorders of epilepsy may result from the inhibitory activity that develops in
reaction to the excessive excitatory activity of the chronic seizure disorder, as pos-
tulated by Stevens (1975) and Engel (1989; Chapter 3). The precise nature of the
seizure-suppressing mechanisms is insuciently understood, and the phenome-
non is usually referred to as forced normalization (Landolt, 1955, 1958, 1963). The
evidence of a linkage of the psychiatric changes to inhibitory mechanisms can be
summarized as follows:
1. The development of interictal dysphoric and psychotic disorders is delayed fol-
lowing the onset of epilepsy (Gastaut et al., 1955; Slater and Beard, 1963) as
inhibitory mechanisms become increasingly established; this situation accords
with the particular linkage of the psychiatric disorders of epilepsy with its most
prominent chronic form mesial temporal lobe epilepsy.
2. Upon decrease and particularly upon full control of seizures, dysphoric symp-
toms and psychosis tend to be exacerbated or to emerge de novo.
3. Psychiatric changes emerge also at times when acute exacerbation of the seizure
activity engages an enhanced inhibitory response, commonly in the prodromal
and postictal phases, and rarely upon increased seizure activity as a seeming
opposite of forced normalization (Blumer et al., 2000).
4. There is a delayed phasing-out of the psychiatric changes after surgical elimina-
tion of the epileptogenic zone, presumably with only gradual fading of the inhib-
itory mechanism (Blumer et al., 1998).
112 D. Blumer
It is hypothesized that the dysphoric and psychotic symptoms, as well as the suici-
dality of epilepsy, are related, as are the seizures themselves, to the homeostatic ebb
and ow of excitatory and inhibitory inuences.
All patients of the two series of suicides providing neuropsychiatric details
(Mendez and Doss, 1992; Epi-Care patients noted above) had an onset of epilepsy
in early life, with a mean duration of the seizure disorder of 25 and 29 years, respec-
tively. This interval exceeds the mean interval from onset of epilepsy to the mani-
festation of interictal psychosis, reported as 14 years (Slater and Beard, 1963). Two
of the three patients who committed suicide in a psychotic state were under good
seizure control. The third patient showed the rare nding (Demers-Desrosiers et
al., 1978) of what seems the opposite of forced normalization: psychotic episodes
coinciding with the presence of increased electroencephalographic epileptiform
potentials in the absence of seizures, presumably resulting from the forceful engage-
ment of inhibitory mechanisms in response to enhanced subclinical seizure activ-
ity (Blumer et al., 2000). The increased suicide risk among patients with ictal and
persistent postictal depression (as in the fourth case of Mendez and Doss, 1992) was
previously documented by Williams (1956).
In our Epi-Care series, the striking nding of all ve suicides, which occurred
shortly after a longstanding seizure disorder had become controlled, points to
the risk of worsening (or de novo appearance) of a marked dysphoric disorder
once predominance of inhibitory mechanisms is established. It must be noted
again that ve of the nine patients from the British series who committed suicide
after temporal lobectomy had been rendered seizure-free (Taylor and Marsh,
1977).
Preventing suicide in epilepsy
Preventing suicide in epilepsy patients consists of eectively treating both the dys-
phoric disorder and the psychosis of the interictal phase (Blumer, 1997; Blumer et
al., 2000; Blumer and Zielinski, 1988). We now treat both the patients with suici-
dal dysphoric moods and those with interictal psychoses with double antidepress-
ant medication, enhanced if necessary with an atypical neuroleptic drug, e.g. with
the combination of 100150 mg imipramine, 2040 mg paroxetine and 24 mg ris-
peridone daily. The same treatment has been eective for patients with severe post-
ictal depressive mood, although we have not had the occasion to treat a patient with
ictal depression and suicidal intensity of the postictal phase. The dysphoric disor-
der is endogenous, and psychotherapy without pharmacotherapy leaves the patient
with suicidal moods at risk.
The bias against using antidepressants for the psychiatric disorders of epilepsy
on the grounds that they may lower the seizure threshold (McConnell and Duncan,
1998) is erroneous on both empirical and theoretical grounds. As our experience

over some 15 years has shown, the modest amounts of antidepressant medication
required do not increase seizure frequency in patients with chronic epilepsy whose
interictal dysphoric disorder or psychosis indicates the presence of marked inhibi-
tion (Blumer, 1997; Blumer et al., 2000; Blumer and Zielinski, 1988). Gastaut et al.
(1955) have pointed out that patients with temporal lobe epilepsy (in contrast to
those with primary generalized epilepsy) show a higher interictal seizure threshold
than do individuals without epilepsy. The proconvulsant eect of antidepressants
does not provoke seizures but may serve to mitigate the psychotoxic eect of exces-
sive inhibition. Avoiding antidepressants for a patient with suicidal dysphoric
moods may have fatal consequences.
The 10739 patients with epilepsy seen at the Epi-Care Center for the past 12 years
were evaluated and treated by a comprehensive team consisting of a neurologist,
neurosurgeon, psychiatrist and neuropsychologist. The ve suicides among this
population represent a much lower fatality rate if compared with the population
reported by Mendez and Doss (1992). This decreased rate may result from the ready
availability of a psychiatrist for treatment of every patient with dysphoric disorder
or psychosis.
There are limitations to suicide prevention, even if one is alerted to the risk. Of
our patients who committed suicide, two were noncompliant with our treatment
and two, for geographic reasons, were followed elsewhere prior to their suicide. In
retrospect, our fth patient, who had become dysphoric after control of his seizures
by medication, should have been brought in for treatment upon the emergence of
an episode of rage that preceded his suicidal act. In our experience, timely psycho-
pharmacological treatment of the dysphoric disorder or of an interictal psychosis
can usually prevent a suicidal outcome of the epileptic disorder.
Conclusion
Further neuropsychiatric studies of patients with epilepsy who commit suicide

need to be made available before one can draw more denite conclusions about all
the modes of suicide in epilepsy. The current review allows a number of conclu-
sions, similar to those reached earlier by Diehl (1986), but more specic in nature.
Suicide in epilepsy results from the psychiatric disorder of temporal lobe epi-
lepsy; that is, from a severe dysphoric disorder, from interictal psychosis (associated
with preceding and concomitant dysphoric disorder, and at times with command
hallucinations), or from a severe postictal depressive state. These psychiatric disor-
ders develop gradually as seizure-suppressing mechanisms become established or,
at times, upon acute engagement of the inhibition. Suicide in epilepsy has increased
with our improved ability to suppress seizures.
114 D. Blumer
Patients with early onset of temporal lobe epilepsy and prolonged duration of the
illness (more than 20 years) are at particular risk of suicide once their seizures are
suppressed. Males in the age range of 3050 are more at risk. Treatment with barbit-
urates, availability of drugs, loss of loved ones or of jobs, and other dicult psycho-
social predicaments are not aetiological factors but may contribute to the suicide risk.
Suicide in epilepsy tends to occur precipitately during a t of melancholy (as
van Gogh described the depressive mood of his dysphoric episodes) and is often
not anticipated. However, there are usually warnings that precede a suicide. Upon
the occurrence of episodes of suicidal moods, prompt intervention is required with
psychotropic medication, chiey of the antidepressant type, and with careful
follow-up that includes adjusting the psychotropic medication as needed.
Although transient suicidal moods among epileptic patients were observed fre-
quently by premodern psychiatrists, completed suicide was not often reported
(Delay et al., 1957). Deaths from seizures have been markedly decreased by our
progress in seizure control but may be surpassed by now in numbers by deaths from
suicide. Our ability to suppress seizures must become paired with our ability to
treat the psychiatric consequences of improved seizure control.
Acknowledgement
The study of the patients seen at the Epi-Care Center was made possible by the col-
laboration of Drs Keith Davies, Bruce Hermann, Georgia Montouris and Allen
Wyler.
R E F E R E N C ES
Barraclough, B.M. (1987). The suicide rate of epilepsy. Acta Psychiatr Scand, 76, 33945.
Blumer, D. (1992). Postictal depression: signicance for the treatment of the neurobehavioral
disorder of epilepsy. J Epilepsy, 5, 21419.
Blumer, D. (2000). Dysphoric disorders and paroxysmal aects: recognition and treatment of
epilepsy-related psychiatric disorders. Harvard Rev Psychiatry, 8, 817.
Blumer, D. and Zielinski, J. (1988). Pharmacologic treatment of psychiatric disorders associated
with epilepsy. J Epilepsy, 1, 13550.
Blumer, D., Montouris, G. and Hermann, B. (1995). Psychiatric morbidity in seizure patients on
a neurodiagnostic monitoring unit. J Neuropsychiatry Clin Neurosci, 7, 44556.
Blumer, D., Wakhlu, S., Davies, K. and Hermann, B. (1998). Psychiatric outcome of temporal
lobectomy for epilepsy: incidence and treatment of psychiatric complications. Epilepsia, 39,
47886.
Blumer, D., Wakhlu, S., Montouris, G. and Wyler, A. (2000). Treatment of the interictal psy-
choses. J Clin Psychiatry, 61, 11022.
Brent, D.A. (1986). Overrepresentation of epileptics in a consecutive series of suicide attempts
seen at a childrens hospital, 19781983. J Am Acad Child Psychiatry, 25, 2426.
Brent, D.A., Crumrine, P.K., Varma, R.R., Allan, M. and Allman, C. (1987). Phenobarbital treat-
ment and major depressive disorder in children with epilepsy. Pediatrics, 80, 90917.
Delay, J., Deniker, P. and Barande, R. (1957). Le suicide des pileptiques. Encphale, 46, 40136.
Demers-Desrosiers, L.A., Nestoros, J.N. and Vaillancourt, P. (1978). Acute psychosis precipitated
by withdrawal of anticonvulsant medication. Am J Psychiatry, 135, 9812.
Diehl, L.W. (1986). Epilepsie und Suizid. Psychiat Neurol Med Psychol, 38, 62533.
Editorial (1980). Suicide and epilepsy. Br Med J, 281, 530.
Engel, J. (1989). Seizures and epilepsy. Philadelphia: Davis.
Gastaut, H. (1956). tat actuel des connaissances sur lanatomie pathologique des pilepsies. Acta
Neurol Psychiatr Belg, 56, 520.
Gastaut, H., Morin, G. and Lesvre, N. (1955). tude du comportement des pileptiques psycho-
moteurs dans lintervalle de leurs crises: les troubles de lactivit globale et de la sociabilit. Ann
Med Psychol, 113, 127.
Ment Dis, 113, 5228.
Hawton, K., Fagg, J. and Marsack, P. (1980). Association between epilepsy and attempted suicide.
J Neurol Neurosurg Psychiatry, 43, 16870.
Henriksen, B., Juul-Jensen, P.P. and Lund, M. (1970). The mortality of epileptics. In Life
Assurance Medicine, ed. R.D.C. Brackenridge, pp. 13948. London: Pitman.
Hill, D. (1953). Psychiatric disorders of epilepsy. Med Press, 229, 4735.
Kraepelin, E. (1923). Psychiatrie, 8th edn. Leipzig: Barth.
Landolt, H. (1953). Some clinical electroencephalographical correlations in epileptic psychoses
(twilight states) [abstract]. Electronencephalogr Clin Neurophysiol, 5, 121.
Landolt, H. (1955). ber Verstimmungen, Dmmerzustnde und schizophrene Zustandsbilder
bei Epilepsie: Ergebnisse klinischer und electroenzephalographischer Untersuchungen.
Schweiz Arch Neurol Psychiatr, 76, 31321.
de Haas, pp. 91133. Amsterdam: Elsevier.
Landolt, H. (1963). Die Dmmer- und Verstimmungszustnde bei Epilepsie und ihre
Electroenzephalographie. Dtsch Z Nervenheilkunde, 185, 41130.
Mackay, A. (1979). Self-poisoning a complication of epilepsy. Br J Psychiatry, 134, 27782.
McConnell, H.W. and Duncan, D. (1998). Treatment of psychiatric comorbidity in epilepsy. In
Psychiatric Comorbidity in Epilepsy: Basic Mechanisms, Diagnosis, and Treatment, ed. H.W.
McConnell and P.J. Snyder, pp. 245361. Washington, DC: American Psychiatric Press.
Matthews, W.S. and Barabas, G. (1981). Suicide and epilepsy: a review of the literature.
Psychosomatics, 22, 51524.
Mendez, M.F. and Doss, R.C. (1992). Ictal and psychiatric aspects of suicide in epileptic patients.
Int J Psychiatry Med, 22, 2317.
116 D. Blumer
Mendez, M.F., Cummings, J.L. and Benson, D.F. (1986). Depression in epilepsy. Signicance and
phenomenology. Arch Neurol, 43, 76670.
Mendez, M.F., Lanska, D.J., Manon-Espaillat, R. and Burnstine, T.H. (1989). Causative factors
for suicide attempts by overdose in epileptics. Arch Neurol, 46, 10658.
Murphy, G.E. (1994). Suicide and attempted suicide. In The Medical Basis of Psychiatry, ed. G.
Winokur and P.J. Clayton, pp. 52944. Philadelphia: W.B. Saunders.
109, 95112.
Stenager, E.N. and Stenager, E. (1992). Suicide and patients with neurologic diseases.
Methodologic problems. Arch Neurol, 49, 1296303.
Stevens, J.R. (1975). Interictal clinical manifestations of complex partial seizures. In Advances in
Neurology, Vol. 11: Complex Partial Seizures and their Treatment, ed. J.K. Penry and D.D. Daly,
pp. 85112. New York: Raven.
Taylor, D.C. (1987). Psychiatric and social issues in measuring the input and outcome of epilepsy
surgery. In Surgical Treatment of the Epilepsies, ed. J. Engel Jr, pp. 485503. New York: Raven
Press.
Taylor, D.C. and Marsh, S.M. (1977). Implications of long-term follow-up studies in epilepsy:
with a note on the cause of death. In Epilepsy, the Eighth International Symposium, ed. J.K.
Penry, pp. 2735. New York: Raven Press.
Williams, D. (1956). The structure of emotions reected in epileptic experiences. Brain, 79,
2967.
9
Postictal psychoses, revisited

Kousuke Kanemoto
Aichi Medical University, Aichi, Japan
Historical background
Except for the immediate eects of a seizure on mental function, such as complex
partial status epilepticus and postictal confusion, modern epileptic psychoses can be
categorized into three main types: chronic, acute interictal and postictal psychoses. In
1953, Landolt stressed a seesaw relationship between epileptic seizures and psychoses,
and proposed the concept of forced normalization. In 1963, Slater made a rather com-
prehensive report on chronic psychoses in patients with epilepsy (Slater and Beard,
1963). In contrast, it was as late as 1988 before the concept of postictal psychoses was
revived by Logsdail and Toone. This delay in conceptual formation is all the more
peculiar, when considering the very old root of the concept of postictal psychosis.
In 1860, a French psychiatrist, Farlet, classied epileptic psychoses into three
categories: transient peri-ictal, chronic and true epileptic psychosis (Farlet,
1860/1961). As there was a lack of strict distinctions between preictal, intraictal and
postictal events at that time, it is not easy to compare Farlets classication to those
of the present. While transient peri-ictal psychosis overlaps postictal confusion and
the meaning of chronic psychosis is evident, Farlets true epileptic psychosis has no
clear counterpart in modern classications. Farlet assigned extreme psychomotor
agitation as well as extraordinarily aggressive and self-destructive behaviour to his
unique classication. As I will discuss later, these are salient psychopathological
traits of postictal psychoses. Indeed, John Hughlings Jackson (1875), a pioneer of
modern epileptology, stressed that the true epileptic psychosis described by Farlet
often followed clusters of seizures. However, authors of subsequent psychiatric lit-
erature have confused Farlets true epileptic psychosis with other types of epileptic
psychosis and even postictal confusion. While the literature has misunderstood the
true nature of the epileptic psychosis of Farlet for more than 100 years (Kolb and
Brodie, 1982; Taylor, 1972), that in the domain of epileptology has gradually for-
gotten it. As a result, and with the advent of antiepileptic drugs, the alternative
psychosis of Landolt (1963) became very popular, obscuring postictal psychosis
completely (Mendez et al., 1993; Onuma et al., 1995).
117
118 K. Kanemoto
This longstanding neglect, however, is being rapidly rectied, resulting in several

review articles (Lancman, 1999; Sachdev, 1998). Episodes of postictal psychosis can
no longer be ignored, as the intensive seizure monitoring under the diminution or
discontinuation of antiepileptic drugs currently practiced has dramatically
increased the chances for those who are involved in epileptic surgery to observe
postictal psychosis directly (Devinsky et al., 1995; Kanner et al., 1996). In this
chapter, we pay special attention to the natural history of postictal psychoses
outside of the seizure monitoring unit, in the belief that such a patient sampling
method would present in-vivo postictal psychoses, instead of the quasi-in-vitro
ones produced in the intensive care unit.
Following a description of two representative cases of postictal psychoses clini-
cal data concerning postictal psychosis are reviewed.
Case reports
CASE 1
A 28-year-old kimono shop manager had a 20-year history of paroxysmal fearful feelings of
being left alone. At age 11, complex partial seizures began to follow these moments of fear.
As his age advanced, his seizures increased in intensity as well as frequency, despite
maximum drug therapy. At age 26, the first manifest postictal mental derangement occurred,
after several bouts of complex partial seizures. One day after this cluster of seizures, he struck
his father, the owner of the kimono shop, who had asked whether he was all right, out of
uncontrollable rage. This peculiar dysphoric state lasted for a week, during which he was con-
tinuously prone to violent behaviour over minor matters. He reported that alien ideas had
invaded him and that opposing thoughts battled each other during this state. Afterwards, he
could recall perfectly the details of his violent behaviour, but could not understand why he
had acted in such a manner. Such postictal episodes and repeated violent behaviours tor-
mented those around him, and recurred every 2 months before admission. An MRI revealed
a marked asymmetry of the hippocampi (the left side was smaller than the right) with a lower
signal intensity from the left hippocampus on a reversed T2 condition. Although he was right-
handed, a 60 mg injection of Amytal into the right, but not the left, carotid artery caused the
patient to become aphasic. Ictal EEG recordings, including those with depth, plus subdural
electrodes, unanimously suggested that the left hippocampus was the origin of both the
simple and complex partial seizures. Subsequently, a left inferior lobectomy with a hippo-
campoamygdalotomy was performed. The resected tissue revealed Ammons horn sclerosis.
On awakening from anaesthesia, the patient showed a peculiar dysphoria, just as he had
during his postictal psychoses, which spontaneously disappeared within 3 days. However,
one month after the lobectomy, he became increasingly euphoric and elevated in mood. His
speech was loud, rapid and difficult to interrupt, while full of jokes, puns and other amusing
irreverence. He even exposed his genitalia in public during his elevated moods and courted
several copatients. Since he sang throughout the night, we needed strong sedatives to put
119 Postictal psychoses, revisited
him to sleep. This manic state, lasting for 4 weeks, gradually turned into a depressive one.
Two months after the operation, he felt quite upset and lost interest in all activities. He
became so agitated that he could not stand still for a moment, and walked around restlessly.
He complained of slowed thinking and difficulties in making decisions. After treatment with
tricyclic antidepressants, his mood improved steadily, but only gradually, over half a year. Six
months after the operation, he set out to do his previous work and his relations with col-
leagues improved dramatically. A year after the operation, his mood was stabilized without
the help of antidepressants and he was accepted once again as the manager of the kimono
shop. He has been completely seizure-free for 4 years postoperatively. Moreover, neither the
previous dysphoric episodes nor the mood disorders have recurred.
CASE 2
At age 3, a 38-year-old housewife suffered from an episode of prolonged febrile convulsions,

which lasted for more than 30 minutes and resulted in transient paralysis of the left upper
extremities. At age 12, the first episode of complex partial seizures occurred, in which she
unknowingly handed over an examination paper to a classmate who happened to be sitting
next to her. After that, every time during a seizure, she would unconsciously reiterate the
same phrase; Hes coming to collect my examination paper. What should I do?. At age 15,
paroxysmal feelings of a peculiar familiarity began to precede the complex partial seizures,
during which she felt that the atmosphere of her environment suddenly changed and it
seemed as if she had dissolved into the immediate surroundings. Her seizures remained
uncontrollable despite intensive medication. As the patients age advanced, she joined a
local religious sect as a devoted member and became increasingly eccentric. The first mani-
fest postictal psychosis occurred after bouts of complex partial seizures at age 29. After an
intervening 36-hour lucid interval, she would rapidly become more and more elevated in
mood, with loud rapid speech that was difficult to interrupt.
She would change subjects kaleidoscopically from one to another. She screamed to her
husband repeatedly, I love you, darling, and hugged and kissed him in public in a sensual
manner. Three days after the cluster of seizures, the euphoric state culminated in agitated
exaltation. She said, I am directly feeling all that is happening in every corner of the world
through the palpitating movement of my teeth. The circular movements of my teeth are syn-
chronized with the circular movements of the world. Through nerves in my teeth, I can sense
the future 2000 years from now. Because of extreme psychomotor agitation, a short stay
in the psychiatric ward as well as potent sedatives were required. This state disappeared
completely within 10 days. Throughout the episode, her orientation and memory remained
intact. After several episodes of such postictal psychotic states, she agreed to surgery. An
MRI revealed a marked asymmetry of the hippocampi (the left side was smaller than the
right) with a lower signal intensity from the left hippocampus on a reversed T2 condition.
Although she was right-handed, her dominant language side proved to be right. Ictal EEG
recordings unanimously suggested that the left hippocampus was the origin of both the
simple and complex partial seizures. During the course of intensive seizure monitoring,
postictal psychosis recurred once after a cluster of complex partial seizures. Subsequently,
120 K. Kanemoto
a left inferior lobectomy with a hippocampoamygdalotomy was performed and the resected
tissue revealed Ammons horn sclerosis. She recovered steadily without complications. A
year after the operation, the patient began to work as a manager of a Japanese restaurant.
She completely lost interest in the religious activity as if exorcised. She has been completely
seizure-free for 7 years postoperatively and no episodes of postictal psychosis have
recurred.
Clinical studies
In our work, we re-examined all of the outpatient cases from 1984 to 1999 at the
Kansai Regional Epilepsy Center who were known to have had epilepsy with psy-
chotic episodes (n177). Epilepsy and seizure classications were based on de-
nitions proposed by the International League against Epilepsy (Commission on
Classication and Terminology of the International League Against Epilepsy, 1981;
1989). In our study, psychosis was dened according to the following ICD10 cri-
teria: the presence of hallucinations, delusions, or a limited number of severe
abnormalities of behaviour, such as gross excitement and overactivity, and cata-
tonic behaviour (World Health Organization, 1992). However, we excluded
psychomotor retardation from the original denition. Ictal psychotic episodes
directly corresponding to ictal epileptiform discharge, such as nonconvulsive status
epilepticus, were also excluded from psychotic episodes. Postictal psychosis was
dened as one that occurred within 7 days after the last generalized tonic-clonic sei-
zures or clusters of complex partial seizures. We excluded those patients who exhib-
ited postictal psychosis only during or immediately after intensive seizure
monitoring. In this study, all episodes of psychosis, except for postictal psychosis,
were included in the category of interictal psychosis. Patients with continuous hal-
lucinations or delusions without remission were regarded as having chronic
psychosis. Patients with interictal psychosis but without chronic psychosis were
included with the acute interictal psychosis cases. We assessed the psychopatholog-
ical features of psychotic episodes according to a modied scale for assessing posi-
tive symptoms (SAPS; Andreasen, 1984), for which the test procedure details have
been described in a previous study (Kanemoto et al., 1996a). Statistical analyses
were made with Chi-square tests, with Yates modication for small numbers.
Incidence
Fifty-one (2%) out of 2905 patients with epilepsy treated at Kansai Regional
Epilepsy Center experienced postictal psychoses that were not articially induced
(Table 9.1). It is dicult to compare our data with previous studies, as they are
either multiple case reports (Lancman et al., 1994; Levin, 1952; Logsdail and Toone,
1988; Savard et al., 1991; Umbricht et al., 1995) or based on observations during
Table 9.1. Incidence of epileptic psychosis
Patients with psychotic disorder 177 (6%)

Postictal psychosis* 51 (2%)
Acute interictal psychosis* 75 (3%)
Chronic psychosis* 57 (2%)
Patients without psychotic disorder 2728 (94%)
Total 2905 (100%)
Note:
* Six patients experienced both postictal and interictal psychoses.
the seizure monitoring in preparation for epilepsy surgery (Devinsky et al., 1995;
Kanner et al., 1996). However, our nding seems to have a certain reliability,
because the prevalence of interictal psychosis in patients with epilepsy in the
current study (5%) agreed well with that seen in other specialized epilepsy clinics
(49%) (Edeh and Toone, 1987; Mendez et al., 1993; Onuma et al., 1995). The prev-
alence of postictal psychosis in patients with temporal lobe epilepsy (11%) proved
to be far higher than that in the general epilepsy population. Kanner et al. (1996)
reported that the annual incidence of postictal psychiatric events, including post-
ictal psychosis, at their monitoring unit was 7.8%. Seven out of the 13 patients in
their series had their rst-ever postictal psychiatric event during the monitoring
study, therefore, at most only 4% experienced a truly spontaneous postictal psychi-
atric event. Considering that their study was limited to patients with symptomatic
localization-related epilepsy, this gure is approximate to ours.
Lucid interval, recurrence and duration of postictal psychosis

Postictal psychosis has long been confused with the clouded consciousness
observed following complex partial seizures. Recent studies, including ours, have
proven that postictal psychosis cannot be reduced to a mere extension of postictal
confusions. The primary argument in support of this is twofold: (1) there is pre-
served orientation and memory seen during the episodes, and (2) lucid intervals
occur between the end of the seizures and the start of postictal psychoses
(Kanemoto et al., 1996a; Levin, 1952; Logsdail and Toone, 1988; Savard et al.,
1991). It is very dicult to explain the delayed manifestation of positive symptoms
after a short period of normality as a psychic equivalent of Todds paralysis, since a
mere exhaustion of the nervous system would have steadily recovered without
reversion, just as in postictal confusion. In 18 out of 51 patients, we conrmed the
presence of a lucid interval, which lasted from 1 to 3 days in 15 out of the 18 (83%)
(Table 9.2). This agrees well with previous studies (Kanemoto et al., 1996b; Kanner
122 K. Kanemoto
Table 9.2. Presence of a lucid interval in 18

of 51 patients observed
Less than 24 hours 1

1 day 4
2 days 6
3 days 5
47 days 2
Undetermined 33
Table 9.3. Duration of the postictal psychosis
1224 hours 4
17 days 16
1 week to 1 month 13
Longer than 1 month 3
Undetermined 15
et al., 1996; Logsdail and Toone, 1988; Savard et al., 1991). In comparison with
acute interictal psychosis, the duration of postictal psychosis is relatively short. In
more than half of the patients, psychotic episodes disappeared within a week, and
psychotic states lasting for more than one week but less than a month were seen in
another one third (Table 9.3). In a certain proportion of patients, postictal psycho-
sis has been shown to have a tendency to recur (Kanner et al., 1996; Lancman et al.,
1994). In our series, 49% of the patients with postictal psychosis experienced one
or more recurrences.
Patient background
In view of the common features that patients with interictal and postictal psycho-
sis share, such as a comparatively long latent period between epilepsy and psycho-
sis onset (longer than 10 years; Table 9.4) and a close association with temporal lobe
epilepsy, the prevailing view has been that interictal and postictal psychoses are
probably similar (Savard et al., 1991). However, we have found that postictal and
interictal psychoses dier in several fundamental demographic data. First, age at
epilepsy onset was signicantly younger in patients with interictal psychosis than
in those with postictal psychosis. Second, the latent period between psychosis and
epilepsy onset was still longer in postictal than interictal psychosis. Third, the pro-
portion of those patients with reduced intelligence quotient (IQ) was signicantly
Table 9.4. Patient background
Postictal psychosis Interictal psychosis

Demographic dataa (n45) (SD) (n126) (SD)
M/F 29/16 78/48

Age (year) 41.4 (12.5) 32.2 (9.8)
Age at epilepsy onset (year) 16.0 (11.3) 11.2 (6.2)
Duration of epilepsy (year) 25.6 (14.1) 20.5 (10.8)
Age at psychosis onset (year) 34.5 (10.3) 24.8 (7.3)
Latent period 18.1 (10.7) 13.2 (8.5)
Low intelligenceb 4/45 39/126
Notes:
a
There was a highly signicant dierence between the groups (P0.005) in all items except
for sex.
b
Full Scale IQ70.
higher in patients with interictal psychosis than in those with postictal psychosis.
These data agree well with the report of Umbricht et al. (1995), which conrmed
lower IQ and younger age at onset of epilepsy in interictal psychosis than in post-
ictal psychosis.
Association with temporal lobe epilepsy

Our data support the suggestion of a majority of previous studies that psychosis
in epilepsy might be preferentially associated with the temporal lobe (Bruens,
1971; Edeh and Toone, 1987; Gibbs, 1951; Perez and Trimble, 1980; Toone et al.,
1980). However, a direct comparison between postictal and interictal psychosis
revealed that the close link between psychotic episodes and temporal lobe epi-
lepsy was signicantly more remarkable in postictal than in interictal psychosis
(Table 9.5). Our data provided additional supportive evidence in this respect.
First, patients with postictal psychosis exhibited a temporal lobe pathology sig-
nicantly more often on an MRI than those with interictal psychosis. Second,
electroencephalographically, generalized spikes and waves were signicantly
more often recorded in interictal psychosis than in postictal psychosis patients
(Table 9.6). Furthermore, as for seizures, complex partial seizures were more
closely associated with postictal psychosis than interictal psychosis (Table 9.7).
These data support the view that the involvement of the limbic structure is almost
always indicated in cases of postictal psychosis and that postictal psychosis, in
contrast to the alternative psychosis of Landolt, almost never occurs in idiopathic
generalized epilepsy.
124 K. Kanemoto
Table 9.5. Epilepsy classification
Postictal Interictal No psychotic

psychosis psychosis episode
(n45a) (n126a) (n2728)
Idiopathic localization-related epilepsy 42

Symptomatic localization-related epilepsy
Temporal lobe epilepsyb 39 74 422
Extra-temporal lobe epilepsy 10 35 857
Idiopathic generalized epilepsy 5 308
Symptomatic/cryptogenic generalized epilepsy 7 369
Others 2 11 730
Notes:
a
Six patients with both postictal and interictal psychoses were excluded.
b
Chi-square5.14, statistically signicant dierence (p0.05).
Table 9.6. Laterality and localization of the laboratory findings
Postictal psychosis Interictal psychosis

(n45) (n126)
EEG ndings
Temporal foci 33 78
Extra-temporal foci 8 15
Sidedness (L/R) 11/18 43/41
Diuse SWC 1 21b
MRI localization
Temporala 16 25b
Extra-temporala 4 14
Sidedness (L/R) 12/9 34/20
Notes:
a
Patients with both temporal and extra-temporal pathology were excluded.
b
Statistically signicant dierence (P0.05).
SWC, spike and wave complex.
Table 9.7. Seizures in patients with postictal or interictal psychoses
Postictal psychosis (n45) Interictal psychosis (n126)
SPS 25 79
CPS 37 84a
GTC 36 97
Minor GS 0 9
Notes:
a
SPS, simple partial seizures; CPS, complex partial seizures; GTC, generalized tonic-clonic
seizures; minor GS, minor generalized seizures including absence, generalized myoclonic
seizure and generalized tonic seizure.
Table 9.8. Simple partial seizures in localization-related epilepsy
Autonomic 16 40
Dj vu 10 10a
Motor 4 15
Elementary visual 0 11
Note:
a
In a previous study based on MRI results (Kanemoto et al., 1996b), we suggested

that medial temporal lesions with additional neocortical involvement were especially
closely linked with postictal psychosis. We have also stressed the close association of
postictal psychosis with psychic auras, such as dj vu and ictal fear (Kanemoto et al.,
1996). In the present study as well, we conrmed a statistically signicant predomi-
nance of dj vu (Table 9.8). Considering the results of a cortical stimulation study
conducted by Gloor et al. (1982), which demonstrated involvement of the lateral as
well as medial temporal structure in the genesis of dj vu, this predominance of dj
vu aura agrees with the nding obtained from our MRI study. Although Savard et al.
(1991) suggested that the high incidence of ictal fear in patients with postictal psychic
aura was proof of the similarity between postictal and interictal psychoses, the
current study contradicts this, demonstrating that such psychic auras were a salient
feature of patients with postictal but not interictal psychosis. Furthermore, several
authors have pointed out that bilateral interictal epileptiform discharges predisposed
patients to develop postictal psychosis (Savard et al., 1991; Umbricht et al., 1995).
126 K. Kanemoto
Table 9.9. Psychopathological findings (Modified SAPS)a
Delusion of perception 0 37
Delusions of reference 3 113
Persecutory delusions 5 111
Verbal hallucinations 3 82
Visual hallucination 9 2
Grandiose delusions 12 1
Religious delusions 10 3
Pressure of speech 22 1
Illusion of familiarity 13 1
Mental diplopia 8 1
Notes:
a
Only features that demonstrated signicant dierence (P0.005) are listed.
SAPS, Scale for the Assessment of Positive Symptoms (Andreason, 1984).
Psychopathological features
The most striking dierences between interictal and postictal psychoses lay in the
domain of psychopathological phenomenology (Table 9.9). In a series presented by
Logsdail and Toone (1988), only one of 14 patients had primary delusions or thought
disorders (7%), whereas as many as nine exhibited a markedly abnormal mood
(64%). Our previous study, comparing the psychopathological features of postictal
psychoses with those of interictal psychoses, supported their data. The rst-rank
symptoms of Schneider, such as delusions of perception and voice commenting,
occurred signicantly less often in postictal psychoses than in acute interictal psy-
choses, whereas sexual indiscretions, religious delusions, and grandiosity, often in
the setting of an elevated mood, were observed in postictal psychosis ve times more
often than acute interictal psychosis. Illusions of familiarity, mental diplopia, and
feelings of impending death, which Jackson and Stewart (1899) described as hall-
marks of the dreamy state, occurred almost exclusively in postictal psychosis. The
frequent occurrence of grandiose delusions and religious delusions, in a setting of
markedly elevated moods and feelings of mystic fusion of the body with the universe,
characterizes the psychopathology of postictal psychosis. In the present extended
series of patients, we were able to amplify these ndings. In Table 9.7, psychopatho-
logical traits that were proved to show a highly signicant dierence are listed.
Violence and postictal psychosis

Violent behaviour elicited in the course of postictal psychosis deserves a special
comment. The argument against the view that epilepsy is closely related to a libera-
tion of aggressive impulses has marked modern epileptology, with the result that epi-
leptologists have almost succeeded in dismissing this old view. However, in the course
of our investigation of postictal psychosis, the sporadic episodes of abrupt violent
behaviour that we observed impressed us greatly. In a previous study (Kanemoto et
al., 1999), we compared violent attacks during episodes of postictal psychosis, acute
interictal psychosis, and postictal confusion immediately following complex partial
seizures in patients with temporal lobe epilepsy (TLE), and conrmed that severe
violent confrontational behaviour towards surrounding people with impending
danger occurred only rarely during the postictal confusions as previous studies have
also pointed out (Ashford et al., 1980; Rodin, 1973; Treiman, 1991). In contrast,
patients proved to be quite prone to violent behaviour during episodes of postictal
psychosis. Recently, Gerard et al. (1998) reported six cases who were identied as
having subacute postictal aggression, which supports our postulation that stresses a
close link between postictal psychosis and violent behaviour.
Postoperative de novo manic depressive illness
Hill et al. (1957) were one of the rst to recognize that depression could occur after
a temporal lobectomy. In a series evaluated by Taylor (1972), ve patients commit-
ted suicide. In another follow-up study, Taylor and Marsh (1977) reported that the
mortality rate during the rst 2 years postoperatively was twice as high as that in
any subsequent 2-year period. Further, in a Danish series investigated by Jensen and
Larsen (1979), all suicide attempts occurred within the rst postoperative month.
A literature search failed to nd any descriptions of postoperative hypomanic or
manic states, except for our own recent report (Kanemoto et al., 1998). However,
we were able to conrm the presence of a substantial number of cases with postop-
erative transient manic or hypomanic states, and a close relationship between post-
operative mood disorder and preoperative history of postictal psychoses.
Considering the intrinsic interrelatedness of postictal psychosis with dramatic
aective changes (Kanemoto et al., 1996a; Logsdail and Toone, 1988; Savard et al.,
1991), this preponderance of postoperative mood changes among patients with
preoperative postictal psychosis was all the more instructive. As Trimble (1991) has
warned, in view of the high incidence of suicide during the rst few years after a
temporal lobectomy alone, the need for continuing psychiatric observations of
patients who receive an operation, especially when they have a history of postictal
psychosis prior to surgery, is apparent.
Treatment
Treatment for postictal psychosis should be directed at two dierent stages. First,
once an episode of postictal psychosis appears, a direct shortening or alleviation of
128 K. Kanemoto
postictal psychosis should be attempted. While Kanner et al. (1996) recommend

dopamine blockers, Lancman et al. (1994) advise the use of benzodiazepines and
sedation with chloral hydrate.
However, this dierence in opinion is more apparent than real. In a typical case,
postictal psychosis begins with an initial hypomanic state, which develops rapidly
into a psychotic state with marked psychomotor agitation within 1248 hours. If
we succeed in making patients sleep amply during the initial hypomanic state,
appearance of frank psychosis could be nipped in the bud. In this way, a certain
proportion of the postictal psychosis could be prevented at the stage of the hypo-
manic state, especially in the seizure monitoring unit, where trained psychiatrists
could recognize the initial signs without delay. Indeed, postictal psychosis is self-
limited and lasts at most only for several days in a majority of cases. However, some
patients tend to commit violent behaviours or even suicide during the episodes.
Therefore, we strongly recommend prompt arrest of this state. Second, in contrast
to the alternative psychosis of Landolt, eorts to reduce complex partial and gen-
eralized tonic-clonic seizures could prevent recurrence of postictal psychosis in a
substantial number of patients. Because the direction of the treatment is opposite
at times as a function of alternative and postictal psychosis, the dierential diagno-
sis between these states is all the more important.
Conclusion
In conclusion, a concept of postictal psychosis is certainly contributing to the elu-

cidation of several longstanding neurobehavioural problems in patients with epi-
lepsy that have caused much controversy. Considering the conspicuous aective
nature of postictal psychosis as well as the preponderance of postoperative mood
disorders in patients with a preoperative history of postictal psychosis, further
investigations of this state could reveal underlying neurochemical changes which
could serve as a model for bipolar mood disorders. However, there remains much
controversy on the pathophysiological mechanism underlying postictal psychosis
(Wieser et al., 1985; Wolf, 1991). Although So et al. (1990) conrmed increased
spiking activity in the medial temporal region with depth-EEG recording during
an episode of postictal psychosis, they were sceptical about the causal relationship
between this nding and occurrence of frank psychosis, since increased spiking is
commonly observed as a postictal phenomenon in patients without psychosis. In a
similar study using depth EEG, Mathern et al. (1995) failed to conrm increased
postictal spike activity. Possibly, postictal psychosis should be further subdivided.
While some cases, in which peculiar feelings such as mental diplopia or dj vu are
prominent with the underlying tone of dysphoria, might be attributed to subclin-
ical limbic status epilepticus such that Wieser et al. (1985) conrmed with a depth-
EEG study, other cases in the setting of markedly elevated moods would not be
directly related to seizure activity but caused by some alterations in pathways of
neurotransmitters. In this regard, the serotoninergic mechanism should also be
considered in view of the marked aective nature of the postictal psychosis along
with dopaminergic hypersensitivity (Kanner et al., 1996; Logsdail and Toone, 1988;
Savard et al., 1991; So et al., 1990) as well as a GABA-mediated mechanism (Ring
et al., 1994; Szabo et al., 1996). Certainly, postictal psychosis deserves further atten-
tion.
R E F E R E N C ES
Andreasen, N.C. (1984). Scale for the Assessment of Positive Symptoms (SAPS). Iowa: The
University of Iowa.
Ashford, J.W., Schulz, S.C. and Walsh, G.O. (1980). Violent automatism in a partial complex
seizure. Arch Neurol, 37, 1202.
Bruens, J.H. (1971). Psychoses in epilepsy. Psychiatr Neurol Neurochir, 74, 17492.
(1981). Proposal for revised clinical and electroencephalographic classication of epileptic sei-
zures. Epilepsia, 11, 10213.
(1989). Proposal for revised classication of epilepsies and epileptic syndromes. Epilepsia, 30,
38999.
Devinsky, O., Abramson, H., Alper, K. et al. (1995). Postictal psychosis: a case control series of 20
patients and 150 controls. Epilepsy Res, 20, 24753.
Edeh, J. and Toone, B. (1987). Relationship between interictal psychopathology and the type of
epilepsy. Br J Psychiatry, 151, 95101.
Farlet, J. (1860/1961). De ltat des pileptiques. Arch Gn Md, 16, 66199; 17, 46191; 18,
42343.
Gerard, M.E., Spitz, M.C., Towbin, J.A. and Shantz, D. (1998). Subacute postictal aggression.
Ment Dis, 113, 5228.
Gloor, P., Olivier, A. and Quensney, L.F. (1982). The role of the limbic system in experiential phe-
nomena of temporal lobe epilepsy. Ann Neurol, 12, 12944.
Hill, D., Pond, D.W., Mitchell, W. and Falconer, M.A. (1957). Personality changes following tem-
poral lobectomy for epilepsy. J Ment Sci, 103, 1827.
Jackson, J.H. (1875). On temporary mental disorders after epileptic paroxysms. West Riding
Lunatic Asylum Med Rep, 5, 10529.
Jackson, J.H. and Stewart, P. (1899). Epileptic attack with a warning of a crude sensation of smell
and intellectual aura (dreamy state) in a patient who has symptoms pointing to gross organic
disease of right temporo-sphenoidal lobe. Brain, 22, 53449.
130 K. Kanemoto
Jensen, I. and Larsen, J.K. (1979). Mental aspects of temporal lobe epilepsy follow-up of 74
patients after resection of a temporal lobe. J Neurol Neurosurg Psychiatry, 42, 25665.
Kanemoto, K., Kawasaki, J. and Kawai, I. (1996a). Postictal psychoses: a comparison with acute
Kanemoto, K., Takeuchi, J., Kawasaki, J. and Kawai, I. (1996b). Characteristics of temporal lobe
epilepsy with mesial temporal sclerosis, with special reference to psychotic episodes.
Neurology, 47, 1199203.
Kanemoto, K., Kawasaki, J. and Mori, E. (1998). Postictal psychosis as a risk factor for mood dis-
orders after temporal lobe surgery. J Neurol Neurosurg Psychiatry, 65, 5879.
Kanemoto, K., Kawasaki, J. and Mori, E. (1999). Violence and epilepsy: a close relation between
violence and postictal psychosis. Epilepsia, 40, 1079.
Kanner, A.M., Stagno, S., Kotagal, P. and Morris, H.H. (1996). Postictal psychiatric events during
prolonged video-electroencephalographic monitoring studies. Arch Neurol, 53, 25863.
Kolb, L.C. and Brodie, H.K.H. (1982). Modern Clinical Psychiatry. Philadelphia: W.B. Saunders.
Lancman, M.E. (1999). Psychosis and peri-ictal confusional states. Neurology, 53 (Suppl. 2),
S338
Lancman, M.E., Craven, W.J., Asconap, J.J. and Penry, J.K. (1994). Clinical management of
recurrent postictal psychosis. J Epilepsy, 7, 4751.
Landolt, H. (1953). Einige klinisch-elektroencephalographische Korrelationen bei epileptischen
Dmmerzustnden. Nervenarzt, 24, 479.
Landolt, H. (1963). Die dammer- und verstimmungszustande bei epilepsie und ihre
Elektroencephalographie. Dtsch Z Nervenheilkunde, 185, 41130.
Levin, S. (1952). Epileptic clouded states. A review of 52 cases. J Nerv Ment Dis, 116, 21525.
Logsdail, S.J. and Toone, B.K. (1988). Postictal psychoses. A clinical and phenomenological
Mathern, G.W., Pretorius, J.K., Babb, T.L. and Quinn, B. (1995). Unilateral hippocampal mossy
ber sprouting and bilateral asymmetric neuron loss with episodic postictal psychosis. J
Neurosurg, 82, 22833.
Mendez, M.F., Grau, R., Doss, R.C. and Taylor, J.L. (1993). Schizophrenia in epilepsy: seizure and
psychosis variables. Neurology, 43, 10737.
Onuma, T., Adachi, N., Ishida, S., Katou, M. and Uesugi, S. (1995). Prevalence and annual inci-
dence of psychoses in patients with epilepsy. Epilepsia, 36 (Suppl. 3), S218
Perez, M.M. and Trimble, M.R. (1980). Epileptic psychosis-diagnostic comparison with process
schizophrenia. Br J Psychiatry, 141, 25661.
Ring, H.A., Trimble, M.R., Costa, D.C., Moriarty, J., Verhoe, N.P. and Ell, P.J. (1994). Striatal
dopamine receptor binding in epileptic psychosis. Biol Psychiatry, 35, 37580.
Rodin, E.A. (1973). Psychomotor epilepsy and aggressive behavior. Arch Gen Psychiatry, 28,
21013.
Sachdev, P. (1998). Schizophrenia-like psychosis and epilepsy: the status of the association. Am J
So, N.K., Savard, G., Andermann, A., Olivier, A. and Quensney, L.F. (1990). Acute postictal
psychosis: a stereo EEG study. Epilepsia, 31, 18893.
109, 95150.
Szabo, C.A., Lancman, M.L. and Stagno, S. (1996). Postictal psychosis: a review. Neuropsychiatry,
Neuropsychol Behav Neurol, 4, 25864.
Taylor, D.C. (1972). Mental state and temporal lobe epilepsy. A correlative account of 100 patients
treated surgically. Epilepsia, 13, 72765.
Taylor, D.C and Marsh, S.M. (1977). Implication of long-term-follow-up in epilepsy: with a note
on the cause of death. In Epilepsy: The 8th International Symposium, ed. J.K. Penry, pp. 2734.
New York: Raven Press.
Toone, B.K., Garralda, M.E. and Ron, M.A. (1980). The psychoses of epilepsy and the functional
psychoses. Br J Psychiatry, 137, 2459.
Treiman, D.M. (1991). Psychobiology of ictal aggression. In Advances in Neurology, Vol. 55, ed.
D. Smith, D. Treiman and M.R. Trimble, pp. 34156. New York: Raven Press.
Umbricht, D., Degreef, G., Barr, W.B., Lieberman, J.A., Pollack, S. and Schaul, N. (1995). Postictal
and chronic psychoses in patients with temporal lobe epilepsy. Am J Psychiatry, 152, 22431.
Wieser, H.G., Hailemariam, S. and Regard, M. (1985). Unilateral limbic epileptic status activity:
stereo-EEG, behavioral, and cognitive data. Epilepsia, 26, 1929.
Wolf, P. (1991). Acute behavioral symptomatology at disappearance of epileptiform EEG abnor-
mality: paradoxical or forced normalization. In Neurobehavioral Problems in Epilepsy, ed. D.
Smith, D. Treiman and M.R. Trimble, pp. 12742. New York: Raven Press.
World Health Organization (1992). The International Classication of Mental and Behavioural
Disorders: Clinical and Diagnostic Guidelines (Tenth Revision) (ICD10). Geneva: WHO.
Part III
Cognitive aspects
10
Dementia and epilepsy

Stephen W. Brown
Plymouth Postgraduate Medical School, Plymouth, UK
Definitions and context
As Lishman (1998) has pointed out, the term dementia has two potential mean-
ings in medical practice: rst, it may refer to a group of specic diseases, and
second, it is used to describe a clinical syndrome that can have many causes.
Although diseases in the former group are characterized by irreversible decline in
function, the latter includes conditions in which decline can be arrested, or in some
cases reversed.
Both ICD10 (World Health Organization, 1992) and DSMIV (American
Psychiatric Association, 1994) oer detailed diagnostic criteria for the syndrome,
with DSMIV dening additional principles for diagnosing dierent varieties of
dementia. Both of these, in attempting to ensure uniformity of populations that
might be used for research purposes, adopt a cookbook style of approach.
In this chapter, the term is used in the second of the meanings described by
Lishman. Dementia is regarded as an acquired global impairment of intellect,
memory and personality, which is independent of any impairment of conscious-
ness. The symptoms of dementia are typically of long duration, usually progressive
and often irreversible, but none of these latter features are essential to the concept.
How the concept of epilepsy came to be linked historically with that of demen-
tia, how these concepts became uncoupled, and how new connections came to be
made between them, are components of a story that contains much of the history
of psychiatry and of neurology.
Historical aspects
Aretaeus, in the second century AD (quoted by Temkin, 1971) described people in
whom epilepsy had become chronic as being languid, spiritless, stupid, inhuman,
unsociable, . . . not disposed to hold intercourse and slow to learn, from torpidity
of the understanding and of the senses. There was therefore a view that a continu-
ing propensity to seizures led to deterioration of those faculties that we would
probably refer to as intellect and personality. According to Berrios (1995) the term
135
136 S.W. Brown
dementia acquired a medical connotation in the second half of the eighteenth

century. He quotes the French encyclopaedia of 1765, which lists as causes:
1. damage to the brain caused by excessive usage, congenital causes or old age, 2. failure of the
spirit, 3. small volume of the brain, 4. violent blows to the head causing brain damage, 5. incur-
able diseases such as epilepsy, or exposure to venoms . . . Dementia is dicult to cure as it is
related to damage of brain bres and nervous uids; it becomes incurable in cases of congenital
defect or old age . . . [otherwise] treatment must follow the cause. (Diderot and dAlambert 1765)
A later concept, which was to carry much inuence, was the theory of degeneracy,
especially as that propounded by Morel (1857). He suggested that although mental
disorders might have an external or environmental cause in the rst instance, the
persons biological state is then modied, so that the disorder becomes hereditary.
Following this, each generation displays an increasing degree of pathology until the
line becomes extinguished in idiocy. Although degeneration was exemplied by the
deterioration across generations, it could also take place within the persons life-
time. Morel expressed a particular interest in epilepsy, and was responsible for sug-
gesting that there could exist a masked form (epilepsie larve), in which the main
features were not seizures but insanity. Thus a link was formalized between epilepsy
and deterioration. By the beginning of the twentieth century many European
medical writers took this as received fact. Thus:
The mental state of epileptics, as is well known, frequently presents deterioration, . . . mischie-
vous restlessness and irritability in childhood may develop to vicious and even criminal tenden-
cies in adult life. Every grade of intellectual defect may be met with, down to actual imbecility.
(Gowers, 1885)
And
The most numerous class of epileptics show, after the lapse of years, a slowly progressive dimming
of the active perceptions of the mind, a loss of memory, a blunting of the aections, a permanent
mental obtuseness which increases and grows, until, if the patient lives long enough, there is a
more or less absolute annihilation of all the faculties. (Berkley, 1901)
Most authorities seemed to believe that decline was consequent upon seizures. While
Gowers acknowledged that in a minority of cases deterioration is the expression of
a cerebral imperfection of which the epilepsy is another manifestation, he also stated
that in the majority of cases the failure must be regarded as a consequence of the
disease. Henry Maudsley also considered that seizures had a direct eect:
The mind is slowly weakened by the storms of fury through which it passes, and they sink nally
into the apathy of dementia a state of mere oblivion, in which they cease to hope or care more.
(Maudsley, 1874)
However, during the twentieth century the view that epilepsy alone might cause
intellectual deterioration has by and large been discredited, although replaced by
137 Dementia and epilepsy
observations regarding the eects of concomitant brain disease, recurrent head

injury due to seizures, and undesirable eects of antiepileptic treatment. It is easy
to dismiss nineteenth century writings as the consequence of prejudice-based,
evidence-free speculation, but the question really has to be considered as to
whether the patient population was dierent to that which we see today. Perhaps
aspects of the natural history of seizure disorders have changed with time. A
common cause for admission to an asylum in the late nineteenth century was
general paralysis of the insane. Neurosyphilis was very common then, but is now
rare. Seizures and dementia can both occur in this condition, which we now know
to be a late complication of syphilis, but the aetiology was not fully recognized until
the twentieth century. Indeed, speculation was rife about the possibility of psycho-
social causation (in some ways analogous to the debate in the 1960s and 1970s con-
cerning the cause of schizophrenia). Objectifying the cognitive consequences of
seizure disorders was a task that had to wait until the advent of psychometrics.
Origins of psychometry
Visitors and residents in nineteenth century London could attend an
Anthropometric Laboratory in South Kensington, and pay to have exact measure-
ments made of their height, weight, breathing power, strength of pull and squeeze,
colour sense and much else. This was the brainchild of Sir Francis Galton
(18221911), a gentleman-scholar, explorer, meteorologist and cousin of Charles
Darwin. Galton was interested in explaining the dierences between individuals.
He held views about the distribution of human abilities, and became convinced
that heredity was of prime importance in dening them. He was mentor to Karl
Pearson (18571936), whose name is well known to scientists for his signicant
contribution to the study of statistics, and who occupied the post of Galton
Professor of Eugenics at University College London after Galtons death. Both
Galtons and Pearsons views about the normal distribution of human abilities were
inuential in the subsequent design of intelligence quotient (IQ) tests. When Binet,
Wechsler and others were to develop standardized tests purporting to measure
intelligence, these tests were adjusted in order to provide a bell-shaped curve of dis-
tribution of results, and tests tended to be constructed to demonstrate previous
underlying assumptions rather than to discover new ones.
When the rst IQ tests were applied to people with epilepsy, there was sometimes
evidence of intellectual deterioration on retesting (Dawson and Conn, 1929)
although this was found to be part of a wider pattern of uctuation in testretest
performance in epilepsy that could occur in either direction (Fetterman and
Barnes, 1934; Patterson and Fonner, 1928; Sullivan and Gahagan, 1935). One study
found that improvement in seizure control was related to improvement in test per-
formance (Kugelmass et al., 1938) raising the possibility that observed uctuations
138 S.W. Brown
in IQ could be related to seizure frequency. Some studies did suggest a slight decline
in mean IQ with increasing duration of seizure disorder, but no very clear picture
emerged (Brown and Reynolds, 1981). Such observations, although supercially
interesting, were of limited value in developing a sophisticated model of the rela-
tionship between seizure disorders and cognitive function. This was to require
developments from an initially parallel, but later converging discipline, that of
neuropsychology.
Origins of neuropsychology
The origins of neuropsychology lie in direct clinical observation. Two nineteenth
century landmarks in the understanding of structural and functional relationships
in the brain were the observations of Broca and Wernicke. In 1861 Paul Broca iden-
tied the third frontal convolution of the left hemisphere as an area that if damaged,
would result in a specic impairment of expressive language. In 1874 Karl Wernicke
described specic impairment of receptive language associated with damage to an
area in the left hemisphere extending from the rst temporal convolution into the
parietal lobe. The recognition of clinically denable psychological syndromes
related to discrete brain pathology continued to develop along with a greater
understanding of cerebral localization. Modern neuropsychological testing
involves specic observation of memory, language, verbal and nonverbal uency
together with visuo-spatial and motor abilities. These assessments complement
tests of general cognitive functioning. Neuropsychological assessment is of course
used as part of the work-up for epilepsy surgery, but has been an essential tool in
studying the interaction between cognitive impairment and seizure-related vari-
ables (Goldstein 1997; Klviainen et al., 1992; Piccirilli et al., 1994; Rugland, 1990).
An issue in interpretation
In the process of scoring IQ tests in children, the raw test scores are subjected to
adjustment, which takes into account the age of the subject. This then gives a result
that is a comparison against the general population of the same age. As children
grow older, the mean raw scores for various items will rise as a result of learning
and normal development, but the adjustment will ensure that the mean population
IQ remains the same, and that IQ scores of the population of the same age t the
expected bell-shaped curve. This could lead to a situation where an individual child
improves test performance over a period of time and therefore improves raw test
scores, but does not improve as much as the rest of the population overall, and
therefore the IQ is seen to fall. This is of course an example of slow learning com-
pared to the rest of the population, but it does not represent deterioration. The
interpretation of serial IQ scores in children therefore requires attention to the raw
score changes. Clinical and educational psychologists will routinely take this into
account as part of their everyday practice, but the matter is perhaps worth drawing
to the attention of other clinicians who may not be so familiar with the details of
test construction, scoring and interpretation.
Possible relationships
As Gowers implied, dementia and epilepsy may both be consequences of the same
underlying disorder, rather than one being a consequence of the other. Some spe-
cic clinical epilepsy syndromes are associated with acquired disorders of intellect
and memory, although it is not clear whether seizure activity is responsible for the
cognitive changes or whether, like the previous group, there is a shared aetiology. It
is however known that individual seizures may result in a cognitive penalty, and
that interictal epileptiform EEG discharges can sometimes disrupt cognitive func-
tioning. Some antiepileptic drugs can also play a part. These potential relationships
are considered in turn.
Neurological disease in which both dementia and seizures occur

It is well recognized that symptomatic seizures occur in the context of dementia
syndromes in older people (Forsgren et al., 1996). Breteler et al. (1995) found that
people aged 5075 with a diagnosis of epilepsy had an overall relative risk of 1.5 of
subsequently developing a dementia, which they described as a moderately
increased risk over the expected rate. People with Downs syndrome are at partic-
ular risk of developing early Alzheimers disease, and this is frequently associated
with seizures (Collacott, 1993; Lott and Lai, 1982). In people with mental retarda-
tion (learning disabilities) who do not have Downs syndrome, the incidence and
prevalence of dementia is higher than expected, and is often associated with poorly
controlled epilepsy (Cooper, 1997). There are also isolated case reports showing an
apparent association between seizures, dementia and a pathological lesion. It may
well be that there are many such very rare or even one-o disorders (Yerby et al.,
1986). One example of a rare condition in which the relationship between cerebral
lesions, seizures and intellectual deterioration is clear is Sneddon syndrome. This
is of presumed autoimmune origin and is characterized by cerebral infarction,
livedo reticularis, hypertension, epilepsy and progressive dementia (ORiordan et
at., 1995; Stephens, 1992).
One specic pathology that is associated with epilepsy and cognitive deteriora-
tion is the hypothalamic hamartoma (Kuzniecky et al., 1997). This can give rise to
gelastic epilepsy in which brief, frequent and mechanical laughing seizures begin in
infancy against a background of otherwise normal development. Other seizure
types, together with cognitive deterioration, start to appear between the ages of
4 and 10 years (Berkovic et al., 1988).
140 S.W. Brown
Cognitive deterioration and worsening epilepsy also co-exist in Rasmussen syn-

drome (Rasmussen et al., 1958), a devastating progressive focal encephalitis that
has hitherto eluded aetiological explanation. Hart and her colleagues have recently
described some comorbidity with other cerebral pathologies (Hart et al., 1998).
Another hopeful lead has been the discovery by Rogers et al. (1994) of circulating
antibodies to the glutamate GLUR3 receptor in a patient with this condition who
showed some response to removal of the same antibodies by recurrent plasma
exchange.
There are also a number of specic genetic syndromes in which seizures and
dementia occur. Many show a Mendelian pattern, such as the autosomal dominant
hereditary prion disorders (Prusiner, 1994) or the progressive myoclonus of
Unverricht and Lundberg (Lehesjoki et al., 1992), which has a recessive mode of
inheritance. The number of these syndromes continues to expand, as does our
understanding of the pathological processes involved. Much more information will
become available as the map of the human genome continues to be elaborated.
Some specific epilepsy syndromes

Loss of previously acquired skills or abilities is a known adverse association with
certain epilepsy syndromes, most notably West syndrome, LennoxGastaut syn-
drome (LGS), the syndrome of continuous spike-wave discharges in slow wave
sleep (CSWS) and the LandauKlener syndrome (LKS). At rst sight, these con-
ditions all share the feature of frequent epileptiform EEG discharges, and it would
be tempting to suggest that such activity, by disrupting learning, interferes with
cognitive development. Although this may be partly true, it has been shown that
diuse electrical status can occur without apparent cognitive consequence
(Gokyigit and Caliskan, 1995). Besides, some authorities attribute at least part of
the deterioration in LGS to treatment eects, both pharmacological and psycho-
social (Genton and Dravet, 1997). LGS is also characterized in many people by fre-
quently disabling drop attacks (Oguni et al., 1996) which may have cumulative
adverse eects on cognitive function. Goldsmith et al. (2000) found that individu-
als with LGS with a normal cognitive outcome tended to have onset of LGS at a later
age than those who deteriorated.
EEG paroxysmal activity occurring during critical periods in development may
account for some of the specic cognitive and behavioural changes seen in LKS and
CSWS. Synaptic contacts that should have degenerated by apoptosis as part of
normal development may instead be strengthened. If this occurs in the temporop-
arietal cortex, this may lead to the acquired aphasia of LKS. If this is predominantly
frontal, higher cognitive and executive functions will be aected rst, leading to the
disintegrative psychotic-like presentation of CSWS (Roulet Perez et al., 1993;
Smith, 1997). This is consistent with the observation that the earlier the onset, the
worse the symptoms. It does not, however, explain the cause of the paroxysmal dis-
charges.
Cognitive effects of seizures

Seizures can cause disruption in cognitive functioning lasting for days after other-
wise apparent recovery (Dodrill, 1986), and a similar phenomenon lasting for
several hours has been described following other seizure types (Aldenkamp et al.,
1992). Related to this is the observation that seizures occurring during a nights
sleep may aect learning ability the next day (Aldenkamp, 1995).
Single complex partial or secondarily generalized seizures may be associated with
neuronal damage (Rabinowicz et al., 1996) and in humans, brain extracellular glu-
tamate builds up to potentially neurotoxic levels in partial seizures (During and
Spencer, 1993). Some other seizure-related variables that have been described as
adversely aecting cognitive function over time include: the presence of tonic-
clonic or complex partial seizures (Dodrill, 1986; Seidenberg et al., 1981), early
onset of seizures (OLeary et al., 1983, Strauss et al., 1995), the total number of sei-
zures experienced (Dodrill, 1986; Rodin et al., 1986), and repeated nonconvulsive
status epilepticus (Aldenkamp, 1997).
It has recently been reported that repetitive head injury in young adults is asso-
ciated with neocortical neurobrillary tangles (Geddes et al., 1999). Whether this
may be of clinical signicance in people with epilepsy who suer repeated head
trauma as a consequence of their seizures remains to be seen.
Critical periods in development may render cerebral functioning more suscep-
tible to interference (Holmes, 1997; Johnston, 1996). Not surprisingly, improve-
ment in seizure frequency is related to improvement in cognitive functioning
(Seidenberg et al., 1981). Jambaque et al. (2000) found that children with infantile
spasms due to tuberous sclerosis whose seizures became well controlled with vigab-
atrin showed signicant improvement in cognition and behaviour, in contrast to
the usual poor prognosis of this condition if seizures are not well controlled.
Cognitive effects of EEG discharges

Subclinical epileptiform EEG discharges can aect cognitive processes both gener-
ally and specically (Aarts et al., 1984). However, this may not be inevitable, and at
least one case has been reported in which diuse electrical status epilepticus
occurred without demonstrable learning problems (Gokyigit and Caliskan, 1995).
Nevertheless, other studies have suggested that EEG epileptiform discharges occur-
ring during IQ testing can aect test results (Siebelink et al., 1988) and such dis-
charges may also have a discreet eect on school performance (Kasteleijn-Nolst
Trenite et al., 1988). This phenomenon has been called transitory cognitive impair-
ment (TCI) and has been extensively studied. TCI can be demonstrated in about
142 S.W. Brown
half the cases in which testing is accompanied by discharges (Binnie, 1993).

Although the presence of generalized bursts of 3-Hz spike-wave lasting three
seconds or more is most likely to cause TCI, it can be demonstrated in some cases
with discharges of shorter duration. Focal spikes can also produce TCI, those on
the left being mainly associated with verbal errors and those on the right with non-
verbal task impairment (Binnie and Marston, 1992).
Bailet and Turk (2000) found that children with epilepsy and abnormal inter-
ictal EEGs scored lower than children with epilepsy with normal interictal EEGs
on reading and spelling measures. They concluded that overall long-term risks of
learning problems exist among children with epilepsy compared with controls. This
is true even for those children with normal IQs and whose seizures are well con-
trolled. This observation lends some support to the view previously expressed by
other authors that in some cases TCI can result in deterioration in psychological and
social functioning. It has been suggested that specic antiepileptic drug therapy
aimed at reducing the interictal EEG discharges, and not just clinical seizures, may
have a benecial eect (Besag, 1995; Marston et al., 1993). The issue of state-depen-
dent cognitive deterioration is discussed in more detail by Besag, Chapter 6.
Treatment effects
The eects of antiepileptic treatment on cognitive function remain ill-understood
despite much research activity, and the subject remains to some extent controver-
sial. A well-reasoned account of how this topic should be tackled has recently been
given by Aldenkamp and van Bronswijk (1999). In a previous review by Vermeulen
and Aldenkamp (1995) the authors concluded that the assembled evidence so far
is hardly a basis for denitive statements. The following is a synopsis of published
reports of varying scientic rigour, which nevertheless probably represents the
current consensus. It serves to complement the more extensive review of
Aldenkamp, Chapter 17.
Phenobarbitone can signicantly impair learning ability (Calandre et al., 1990),
and impairs attention and memory at even low therapeutic doses (Riva and Devoti,
1996).
A dementia-like picture sometimes seen with phenytoin has been long recog-
nized (Rosen, 1968); this may be related to the serum level of the drug (Matthews
and Harley, 1975) or to phenytoin-induced folate deciency (Neubauer, 1970). In
the latter case, folate supplementation can result in improved cognitive perfor-
mance (Froscher et al., 1995). Phenytoin has also been implicated in thiamine de-
ciency, resulting in discrete performance decits in visuo-spatial analysis,
visuo-motor speed and verbal abstracting ability. These improve with thiamine
supplementation (Botez et al., 1993). Phenytoin has also been reported as causing
a specic memory impairment (Butlin et al., 1984; Gillham et al., 1990), most
marked with visual memory (Pulliainen and Jokelainen, 1994), and it can also have
adverse eects on motor and mental speed (Aldenkamp et al., 1994). Phenytoin
also diminishes the practice eect expected to be seen on retesting in some cases
(Sabers et al., 1995).
Some authors have reported carbamazepine to aect adversely memory
(Forsythe et al., 1991) and psychomotor speed (Gillham et al., 1990), while others
report both carbamazepine and valproate as having negligible eects (Prevey et al.,
1996; Stores et al., 1992). Recently, reversible pseudodementia with cortical
pseudoatrophy induced by valproate has been described (Guerrini et al., 1998;
Papazian et al., 1995; Straussberg et al., 1998).
There are relatively few published studies in this area that relate to the newer
antiepileptic drugs. The evidence so far, such as it is, suggests that with standard-
ized testing procedures, vigabatrin, lamotrigine and tiagabine have no deleterious
eect on cognitive functioning, and in some cases may enhance performance, pre-
sumably by preventing adverse consequences of seizures (Banks and Beran, 1991;
Dodrill et al., 1997; Klviainen et al., 1996; Meador and Baker, 1997; Monaco, 1996;
Provinciali et al., 1996; Sveinbjornsdottir et al., 1994).
Psychosocial effects
It has been observed in a number of studies that children with epilepsy as a group
underachieve academically compared to that expected from objective measures of
their cognitive ability (Rutter et al., 1970). There are many possible reasons for this,
which will include loss of time at school due to illness and hospital appointments, and
denial of opportunity to take part in normal activities. It has also been shown (Long
and Moore, 1979) that signicant people in the lives of children with epilepsy (such
as their parents) have lower expectations of the childrens achievements than they do
of children without epilepsy. This may also have an eect on cognitive outcome.
Some observations in temporal lobe epilepsy
The author has for some years reported observations on some people with tem-
poral lobe epilepsy who show evidence of acquired cognitive problems in the
absence of the factors listed above (Brown, 1989, 1992, 1999; Brown and
Abeyasinghe, 1984; Brown and Vaughan, 1988, 1990). It seems that such problems
may occur more often in males than females, where there is a long history of poorly
controlled epilepsy of temporal lobe origin. The neuropsychological prole in
females is that of frontal and left temporal impairment; males may show impair-
ment of more cortical areas, especially the parietal lobes. In all cases the epilepsy
syndrome is temporal lobe epilepsy, and the EEG epileptic foci are temporal, not
frontal. A full account is given in Brown (1999).
144 S.W. Brown
A syndrome is described seen in people who develop temporal lobe epilepsy

before puberty, and who may come to the attention of psychologists or other
behavioural professionals after puberty because of behaviour problems. The di-
cult behaviour does not typically persist into adult life. Intellectual decline seems
to date from the late teenage years and is followed in most (but not all cases) by rel-
ative stability into adult life. It is not related to repeated minor head injuries caused
by seizures, although some people (mainly male) may have some dysarthria and
ataxia.
We identied two broad syndromes. A frontal subcortical picture was seen,
mainly in males with a prepubertal age of onset of epilepsy, where there is some dis-
turbance around or just after the time of puberty in which a stepwise decline of
intellectual functioning occurs. In addition, a cortical frontotemporal picture was
also seen, more commonly in females. These sometimes had a later age of onset of
epilepsy, and were more likely to acquire a psychiatric diagnosis. Other people with
localization-related epilepsy and discrete neuropsychological decits are also seen
with dierent histories and who show no intellectual deterioration.
Previous reports of metabolic changes in temporal lobe epilepsy have suggested
that frontal lobe functioning can also be aected. Ingvar (1984) comments:
. . . in temporal lobe epilepsy, the pathways from the deep temporal structures to the frontal lobes
do not function normally. The temporal gate to the frontal lobes may be closed. This may
explain symptoms of psychopathology in this common type of focal epilepsy.
More recently, Mackenzie and Miller (1994) examined senile plaques in temporal
lobectomy specimens and a control group. They found the age-related incidence of
senile plaques was signicantly greater in the temporal lobe epilepsy group, and
concluded that this suggested some aspects of TLE have a positive inuence on the
formation of senile plaques. The exact signicance of this nding is unclear.
Temporal Gate hypothesis

We have proposed a possible mechanism to account for the above observations,
which in deference to Ingvar (1984) we have called the Temporal Gate hypothesis
(Brown, 1999). This may be summarized as follows:
1. Certain neuropsychological functions of the frontal lobes come to maturity
during and after puberty, the process being inuenced by the neuroendocrine
axis.
2. This maturation of functioning is also dependent on frontal lobe input from an
intact temporo-limbic system.
3. If seizures of temporal lobe origin occur before and during puberty, the temporo-
limbic input to the frontal lobes during this critical period of development may
be disrupted. Such disruption might lead to a period of behavioural or adjust-
ment problems during and after puberty. Also, because frontal lobe function
does not develop in the normal way, this might lead to the observed neuro-
psychological eects. These eects may reach a plateau, while cases may exhibit
further decline.
4. In our studies, the neuropsychological disruption so caused is sucient to
impair some previously acquired skills, such as reading.
Some indirect, or collateral, support for the hypothesis comes from a study recently
reported by Metz-Lutz et al. (1999) who investigated EEG and neuropsychological
outcomes in children with an initial diagnosis of benign partial epilepsy. Their nd-
ings suggested to them that maturing cognitive functions subserved by a cortical
area distant from the epileptic focus are nevertheless susceptible to interference by
epileptic activity, and this may aect cognitive outcomes.
In other neurological conditions associated with deterioration, the phenomenon
of oxidative stress (the production of oxygen radicals beyond a threshold for proper
antioxidant neutralization) has been implicated. These include Alzheimers disease
(Sims, 1996), Parkinsons disease (Jenner and Olanow, 1994), amyotrophic lateral
sclerosis (Gorman et al., 1994), Picks disease (Castellani et al., 1995) and schizo-
phrenia (Ramchand et al., 1996). Various intracellular messenger systems involv-
ing glutamate are implicated in oxidative radical production. These systems are
involved in neuronal growth, dierentiation and apoptosis (Michaelis, 1998).
Glutamate is also known to play an important role in epilepsy. The author has
observed substantial improvement in cognitive functioning in two patients (one
male subcortical, one female temporal) after using lamotrigine, a glutamate
release inhibitor (Meldrum, 1994). There are various other reports, generally of an
anecdotal nature, of the positive eects of lamotrigine on cognitive functioning
and scholastic ability (Buoni et al., 1998; Meador and Baker, 1997). One could spec-
ulate that oxidative stress plays a part in the maintenance of the acquired frontal
syndrome decits described above.
Conclusions
True intellectual deterioration is now known to be a relatively uncommon associ-

ation of epilepsy. Where it occurs it is usually associated with specic causes. These
include concomitant degenerative brain disease, some specic epilepsy syndromes
of uncertain aetiology, the direct eects of seizures (especially tonic-clonic sei-
zures), transitory cognitive impairment due to interictal EEG epileptiform dis-
charges, and the consequences of some antiepileptic treatments. In some cases it
is possible that a seizure disorder with temporal lobe focus, if present during a crit-
ical developmental period, may adversely inuence maturation and maintenance
of normal frontal lobe function, giving rise to a clinical picture of cognitive
146 S.W. Brown
impairment or decline. The potential role of oxidative stress in this process

deserves further study.
Jokeit and Ebner (1999), as explored in Chapter 11, have described slow,
ongoing deterioration in some people with temporal lobe epilepsy, and state (with
an echo of Henry Maudsleys words from 125 years previously, about the mind
being weakened by the storms of fury through which it passes) It is assumed that
epilepsy-related noxious events and agents exhaust the compensatory capacity of
brain functions. Thus has our understanding come full circle, with still much left
to provide a good seam for investigators to mine.
R E F E R E N C ES
Aarts, J.H., Binnie, C.D., Smit, A.M. and Wilkins, A.J. (1984). Selective cognitive impairment
during focal and generalized epileptiform EEG activity. Brain, 107, 293308.
Aldenkamp, A.P. (1995). The impact of epilepsy on cognitive development and learning behav-
iour. In Epilepsy in Children and Adolescents, ed. A.P. Aldenkamp, W.O. Renier, F.E. Dreifuss
and T.P.B.M. Suurmeijer, pp. 22538. Boca Raton, FL: CRC Press.
Aldenkamp, A.P. (1997). Eect of seizures and epileptiform discharges on cognitive function.
Aldenkamp, A.P. and van Bronswijk (1999). Cognitive side-eects as outcome measures in anti-
epileptic drug treatment: the current debate. In Epilepsy and Mental Retardation, ed. M.
Sillanpaa et al., pp. 135146. Peterseld: Wrightson Biomedical Publishing.
Aldenkamp, A.P., Gutter, T. and Beun, A.M. (1992). The eect of seizure activity and paroxysmal
electroencephalographic discharge on cognition. Acta Neurol Scand, 86, 11122.
Aldenkamp, A.P., Alpherts, W.C., Diepman, L., vant Slot, B., Overweg, J. and Vermeulin, J.
(1994). Cognitive side-eects of phenytoin compared with carbamazepine in patients with
localization-related epilepsy. Epilepsy Res, 19, 3743.
(Fourth Edition) (DSMIV). Washington, DC: APA.
Bailet, L.L. and Turk, W.R. (2000). The impact of childhood epilepsy on neurocognitive and
behavioral performance: a prospective longitudinal study. Epilepsia, 41, 42631.
Banks, G.K. and Beran, R.G. (1991). Neuropsychological assessment in lamotrigine treated epi-
leptic patients. Clin and Exp Neurol, 28, 2307.
Berkley, H.J. (1901). A Treatise on Mental Disease. London: Henry Kimpton.
Berkovic, S.F., Andermann, F., Melanson, D., Ethier, R.E., Feindel, W. and Gloor, P. (1988).
Hypothalamic hamartomas and ictal laughter: evolution of a characteristic epileptic syndrome
and diagnostic value of magnetic resonance imaging. Ann Neurol, 23, 42939.
Berrios, G.E. (1995). Dementia. In A History of Clinical Psychiatry: The Origin and History of
Psychiatric Disorders, ed. G.E. Berrios and R. Porter, pp. 3451. London: The Athlone Press.
Besag, F.M. (1995). The therapeutic dilemma: treating subtle seizures or indulging in electroen-
cephalogram cosmetics? Semin Pediatr Neurol, 2, 2618.
Binnie, C.D. (1993). Signicance and management of transitory cognitive impairment due to
subclinical EEG discharges in children. Brain Develop, 15, 2330.
Binnie, C.D. and Marston, D. (1992). Cognitive correlates of interictal discharges. Epilepsia, 33
(Suppl. 6), S1117.
Botez, M.I., Botez, T., Ross-Chouinard, A. and Lalonde, R. (1993). Thiamine and folate treat-
ment of chronic epileptic patients: a controlled study with the Wechsler IQ scale. Epilepsy Res,
16, 15763.
Breteler, M.M., de Groot, R.R., van Romunde, L.K. and Hofman, A. (1995). Risk of dementia in
patients with Parkinsons disease, epilepsy, and severe head trauma: a register-based follow-up
study. Am J Epidemiol, 142, 13005.
Brown, S.W. (1989). Cognitive impairment in epilepsy. Educ Child Psychol, 6, 2532.
Brown, S.W. (1992). Organic brain syndromes in chronic epilepsy what is the truth about epi-
leptic dementia? Seizure, 1 (Suppl. A), S26/1.
Brown, S.W. (1999). Epilepsy dementia: intellectual deterioration as a consequence of epileptic
seizures. In Epilepsy and Mental Retardation, ed. M. Sillanpaa et al., pp. 11534. Peterseld:
Wrightson Biomedical Publishing.
Brown, S.W. and Abeyasinghe, R. (1984). Is there an epileptic dementia? Acta Neurol Scand Suppl,
70, 233.
Brown, S.W. and Reynolds, E.H. (1981). Cognitive impairment in epileptic patients. In Epilepsy
and Psychiatry, ed. E.H. Reynolds and M.R. Trimble. London: Churchill Livingstone.
Brown, S.W. and Vaughan, M. (1988). Dementia in epileptic patients. In Epilepsy, Behaviour and
Cognitive Function, ed. M.R. Trimble and E.H. Reynolds, pp. 17788. Chichester: John Wiley
& Sons.
Brown, S.W. and Vaughan, M. (1990). The nature of cognitive changes in people with chronic
epilepsy. Acta Neurol Scand Suppl, 82, 13.
Buoni, S., Grosso, S. and Fois, A. (1998). Lamotrigine treatment in childhood drug resistant epi-
lepsy. J Child Neurol, 13, 1637.
Butlin, A.T., Danta, G. and Cook, M.L. (1984). Anticonvulsant eects on the memory perfor-
mance of epileptics. Clin Exp Neurol, 20, 2735.
Calandre, E.P., Dominguez-Granados, R., Gomez-Rubio, M. and Molina-Font, J.A. (1990).
Cognitive eects of long-term treatment with phenobarbital and valproic acid in school chil-
dren. Acta Neurol Scand, 81, 5046.
Castellani, R., Smith, M.A., Richey, P.L., Kalaria, R., Gambetti, P. and Perry, G. (1995). Evidence
for oxidative stress in Pick disease and corticobasal degeneration. Brain Res, 696, 26871.
Collacott, R.A. (1993). Epilepsy, dementia and adaptive behaviour in Downs syndrome. J
Intellect Disabil Res, 37, 15360.
Cooper, S.A. (1997). High prevalence of dementia among people with learning disabilities not
attributable to Downs syndrome. Psychol Med, 27, 60916.
Dawson, S. and Conn, J.C.N. (1929). The intelligence of epileptic children. Arch Dis Child, 4,
14251.
Diderot & dAlambert (ed.) (1765). Encyclopdie ou Dictionnaire Raisonn des Sciences, des Arts
et des Mtieres, par une Societ de Gens de Lettres, Vol. 4, Paris: David Briasson; Le Breton:
Durand (quoted in Berrios, 1995).
148 S.W. Brown
Dodrill, C.B. (1986). Correlates of generalized tonic-clonic seizures with intellectual, neuro-
psychological, emotional and social function in patients with epilepsy. Epilepsia, 27, 399411.
Dodrill, C.B., Arnett, J.L., Sommerville, K.W. and Shu, V. (1997). Cognitive and quality of life
eects of diering dosages of tiagabine in epilepsy. Neurology, 48, 102531.
During, M.J. and Spencer, D.D. (1993). Extracellular hippocampal glutamate and spontaneous
seizure in the conscious human brain. Lancet, 341, 160710.
Fetterman, J. and Barnes, R.R. (1934). Serial studies of the intelligence of patients with epilepsy.
Neuropsychologia, 7, 287300.
Forsgren, L., Bucht, G., Eriksson, S. and Bergmark, L. (1996). Incidence and clinical characteriza-
tion of unprovoked seizures in adults: a prospective population-based study. Epilepsia, 37, 2249.
Forsythe, I., Butler, R., Berg, I. and McGuire, R. (1991). Cognitive impairment in new cases of
epilepsy randomly assigned to carbamazepine, phenytoin and sodium valproate. Dev Med
Child Neurol, 33, 52434.
Froscher, W., Maier, V., Laage, M. et al. (1995). Folate deciency, anticonvulsant drugs, and
psychiatric morbidity. Clin Neuropharmacol, 18, 16582.
Geddes, J.F., Vowles, G.H., Nicoll, J.A. and Revesz, T. (1999). Neuronal cytoskeletal changes are
an early consequence of repetitive head injury. Acta Neuropathol, 98, 1718.
Genton, P. and Dravet, C. (1997). LennoxGastaut syndrome and other childhood epileptic
encephalopathies. In Epilepsy: A Comprehensive Textbook, ed. J. Engel Jr and T.A. Pedley, pp.
235566. Philadelphia, PA: Lippincott-Raven.
Gillham, R.A., Williams, N., Wiedmann, K.D., Butler, E., Larkin, J.G. and Brodie, M.J. (1990).
Cognitive function in adult epileptic patients established on anticonvulsant monotherapy.
Epilepsy Res, 7, 21925.
Gokyigit, A. and Caliskan, A. (1995). Diuse spike-wave status of 9-year duration without behav-
ioral change or intellectual decline. Epilepsia, 36, 21013.
Goldsmith, I.L., Zupanc, M.L. and Buchhalter, J.R. (2000). Long-term seizure outcome in 74
patients with LennoxGastaut syndrome: eects of incorporating MRI head imaging in den-
ing the cryptogenic subgroup. Epilepsia, 41, 3959.
Goldstein, L.H. (1997). Neuropsychological assessment. In The Clinical Psychologists Handbook
of Epilepsy, ed. C. Cull and L.H. Goldstein, pp. 1834. London: Routledge.
Gorman, A.M., McGowan, A., ONeill, C. and Cotter, T. (1994). Oxidative stress and apoptosis
in neurodegeneration. J Neurol Sci, 139 (Suppl.), 4552.
Gowers, W.R. (1885). Epilepsy and Other Chronic Convulsive Diseases. New York: William Wood
and Company.
Guerrini, R., Belmonte, A., Canapicchi, R., Casalini, C. and Perucca, E. (1998). Reversible
pseudoatrophy of the brain and mental deterioration associated with valproate treatment.
Hart, Y.M., Andermann, F., Robitaille, Y., Laxer, K.D., Rasmussen, T. and Davis, R. (1998).
Double pathology in Rasmussens syndrome: a window on the etiology? Neurology, 50, 7315.
Holmes, G.L. (1997). Epilepsy in the developing brain: lessons from the laboratory and clinic.
Epilepsia. 38, 1230.
Ingvar, D.H. (1984). Epilepsy related to cerebral blood ow and metabolism. Acta Psychiatr Scand
Suppl, 313, 216.
Jambaque, I., Chiron, C., Dumas, C., Mumford, J. and Dulac, O. (2000). Mental and behavioural
outcome of infantile epilepsy treated by vigabatrin in tuberous sclerosis patients. Epilepsy Res,
38, 15160.
Jenner, P. and Olanow, C.W. (1994). Oxidative stress and the pathogenesis of Parkinsons disease.
Neurology, 47 (Suppl. 3), S16170.
Johnston, M.V. (1996). Developmental aspects of epileptogenesis. Epilepsia, 37 (Suppl. 1),
S29.
Jokeit, H. and Ebner, A. (1999). Long term eects of refractory temporal lobe epilepsy on cogni-
tive abilities: a cross sectional study. J Neurol Neurosurg Psychiatry, 67, 4450.
Klviainen, R., Aikia, M., Helkala, EL., Mervaala, E. and Riekkinen, P.J. (1992). Memory and
attention in newly diagnosed epileptic seizure disorder. Seizure, 1, 25562.
Klviainen, R., Aikia, M., Mervaala, E., Saukkonen, A.M., Pitknen, A. and Riekkinen, P.J. Sr
(1996). Long-term cognitive and EEG eects of tiagabine in drug-resistant partial epilepsy.
Kasteleijn-Nolst Trenite, D.G., Bakker, D.J., Binnie, C.D., Buerman, A. and Van Raaij, M. (1988).
Psychological eects of subclinical epileptiform EEG discharges. I. Scholastic skills. Epilepsy
Res, 2, 11116.
Kugelmass, I.N., Poull, L.E. and Rudnick, J. (1938). Mental growth of epileptic children. Am J Dis
Child, 55, 295303.
Kuzniecky, R., Guthrie, B., Mountz, J. et al. (1997). Intrinsic epileptogenesis of hypothalamic
hamartomas in gelastic epilepsy. Ann Neurol, 42, 607.
Lehesjoki, A.E., Koskiniemi, M., Pandolfo, M. et al. (1992). Linkage studies in progressive myo-
clonus epilepsy: UnverrichtLundborg and Laforas diseases. Neurology, 42, 154550.
Lishman, W.A. (1998). Organic Psychiatry: The Psychological Consequences of Cerebral Disorder,
Third Edition. Oxford: Blackwell Science.
Long, C.G. and Moore, J.R. (1979). Parental expectations for their epileptic children. J Child
Psychol Psychiatry, 24: 299312.
Lott, I.T. and Lai, F. (1982). Dementia in Downs syndrome: observations from a neurology clinic.
Appl Res Ment Retardation, 3, 2339.
Mackenzie, I.R. and Miller, L.A. (1994). Senile plaques in temporal lobe epilepsy. Acta
Neuropathol, 87, 50410.
Marston, D., Besag, F., Binnie, C.D. and Fowler, M. (1993). Eects of transitory cognitive impair-
ment on psychosocial functioning of children with epilepsy: a therapeutic trial. Dev Med Child
Neurol, 35, 57481.
Matthews, C.G. and Harley, J.P. (1975). Cognitive and motor-sensory performances in toxic and
nontoxic epileptic subjects. Neurology, 25, 1848.
Maudsley, H. (1874). Responsibility in Mental Disease. New York: Appleton.
Meador, K.J. and Baker, G.A. (1997). Behavioral and cognitive eects of lamotrigine. J Child
Neurol, 12 (Suppl. 1), S447.
Meldrum, B.S. (1994). The role of glutamate in epilepsy and other CNS disorders. Neurology, 44
(Suppl. 23).
Metz-Lutz, M.N., Kleitz, C., de Saint, M.A., Massa, R., Hirsch, E. and Marescaux, C. (1999).
Cognitive development in benign focal epilepsies of childhood. Dev Neurosci, 21, 18290.
150 S.W. Brown
Michaelis, E.K. (1998). Molecular biology of glutamate receptors in the central nervous system
and their role in excitotoxicity, oxidative stress and aging. Progr Neurobiol, 54, 369415.
Monaco, F. (1996). Cognitive eects of vigabatrin: a review. Neurology, 47 (Suppl. 1), S611.
Morel, B.A. (1857). Trait des Degnrscences Physiques, Intellectuelles et Morales de lEspce
Humaine et des Causes qui Produisent ces Variets Maladive. Paris: Baillire.
Neubauer, C. (1970). Mental deterioration in epilepsy due to folate deciency. Br Med J, 2,
75961.
Oguni, H., Hayashi, K. and Osawa, M. (1996). Long-term prognosis of LennoxGastaut syn-
drome. Epilepsia, 37 (Suppl. 3), 447.
OLeary, D.S., Lovell, M.R., Sackellares, J.C. et al. (1983). Eects of age of onset of partial and
generalized seizures on neuropsychological performance in children. J Nerv Ment Dis, 171,
6249.
ORiordan, J.I., Javed, M., Murphy, R. and Hutchinson, M. (1995). Sneddons syndrome: clinical
course and outcome. Irish Med J, 88, 667.
Papazian, O., Canizales, E., Alfonso, I., Archila, R., Duchowny, M. and Aicardi, J. (1995).
Reversible dementia and apparent brain atrophy during valproate therapy. Ann Neurol, 38,
68791.
Patterson, H.A. and Fonner, D. (1928). Some observations on the intelligence quotient in epilep-
tics. Psychiatr Quart, 31, 5428.
Piccirilli, M., DAlessandro, P., Sciarma, T. et al. (1994). Attention problems in epilepsy: possible
signicance of the epileptogenic focus. Epilepsia, 35, 10916.
Prevey, M.L., Delaney, R.C., Cramer, J.A., Cattanach, L., Collins, J.F. and Mattson, R.H. (1996).
Eect of valproate on cognitive functioning. Comparison with carbamazepine. The
Department of Veterans Aairs Epilepsy Cooperative Study 264 Group. Arch Neurol, 53,
100816.
Provinciali, L., Bartolini, M., Mari, F., Del Pesce, M. and Ceravolo, M.G. (1996). Inuence of
vigabatrin on cognitive performances and behaviour in patients with drug-resistant epilepsy.
Acta Neurol Scand, 94, 1218.
Prusiner, S.B. (1994). Biology and genetics of prion diseases. Annu Rev Microbiol, 48, 65586.
Pulliainen, V. and Jokelainen, M. (1994). Eects of phenytoin and carbamazepine on cognitive
functions in newly diagnosed epileptic patients. Acta Neurol Scand, 89, 816.
Rabinowicz, A.L., Correale, J., Boutros, R.B., Couldwell, W.T., Henderson, C.W. and DeGiorgio,
C.M. (1996). Neuron-specic enolase is increased after single seizures during inpatient
video/EEG monitoring. Epilepsia, 37, 1225.
Ramchand, C.N., Davies, J.I., Tresman, R.L., Griths, I.C. and Peet, M. (1996). Reduced suscep-
tibility to oxidative damage of erythrocyte membranes from medicated schizophrenic patients.
Prostaglandins Leukot Essent Fatty Acids, 55, 2731.
Rasmussen, T., Olszewski, J. and Lloyd-Smith, D. (1958). Focal seizures due to chronic localized
encephalitis. Neurology, 8, 43545.
Riva, D. and Devoti, M. (1996). Discontinuation of phenobarbital in children: eects on neuro-
cognitive behavior. Pediatr Neurol, 14, 3640.
Rodin, E.A., Schmaltz, S. and Twitty, G. (1986). Intellectual functions of patients with childhood-
onset epilepsy. Dev Med Child Neurol, 28, 2533.
Rogers, S.W., Andrews, P.I., Gahring, L.C. et al. (1994) Autoantibodies to glutamate receptor
GluR3 in Rasmussens encephalitis. Science, 265, 64851.
Rosen, J.A. (1968). Dilantin dementia. Trans Am Neurol Assoc, 93, 273.
Roulet Perez, E., Davido, V., Despland, P.A. and Deonna, T. (1993). Mental and behavioural
deterioration of children with epilepsy and CSWS: acquired epileptic frontal syndrome. Dev
Med Child Neurol, 35, 66174.
Rugland, A.L. (1990). Neuropsychological assessment of cognitive functioning in children with
epilepsy. Epilepsia, 31 (Suppl. 4), S414.
Rutter, M., Graham, P. and Yule, W. (1970). A neuropsychiatric study in childhood. Clinics in
Developmental Medicine Nos. 35/36. London: Spastics International and Heineman Medical.
Sabers, A., Moller, A., Dam, M. et al. (1995). Cognitive function and anticonvulsant therapy:
eect of monotherapy in epilepsy. Acta Neurol Scand, 92, 1927.
test-retest scores on the Wechsler Adult Intelligence Scale. Epilepsia, 22, 7583.
Siebelink, B.M., Bakker, D.J., Binnie, C.D. and Kasteleijn-Nolst Trenite, D.G. (1988).
Psychological eects of subclinical epileptiform EEG discharges in children. II. General intel-
ligence tests. Epilepsy Res, 2, 11721.
Sims, N.R. (1996). Energy metabolism, oxidative stress and neuronal degeneration in
Alzheimers disease. Neurodegeneration, 5, 43540.
Smith, M.C. (1997). LandauKlener syndrome and continuous spikes and waves during slow
sleep. In Epilepsy: A Comprehensive Textbook, ed. J. Engel Jr and T.A. Pedley, pp. 236777.
Philadelphia, PA: Lippincott-Raven.
Stephens, C.J. (1992). Sneddons syndrome. Clin Exp Rheumatol, 10, 48992.
Stores, G., Williams, P.L., Styles, E. and Zaiwalla, Z. (1992). Psychological eects of sodium val-
proate and carbamazepine in epilepsy. Arch Dis Child, 67, 13307.
Strauss, E., Loring, D., Chelune, G. et al. (1995) Predicting cognitive impairment in epilepsy:
ndings from the Bozeman epilepsy consortium. J Clin Exp Neuropsychol, 17, 90917.
Straussberg, R., Kivity, S., Weitz, R., Harel, L. and Gadoth, N. (1998). Reversible cortical atrophy
and cognitive decline induced by valproic acid. Eur J Paediatr Neurol, 2, 2138.
Sullivan, E.B. and Gahagan, L. (1935). On intelligence of epileptic children. Genet Soc Gen Psychol
Monogr, 17, 30975.
Sveinbjornsdottir, S., Sander, J.W., Patsalos, P.N., Upton, D., Thompson, P.J. and Duncan, J.S.
(1994). Neuropsychological eects of tiagabine, a potential new antiepileptic drug. Seizure, 3,
2935.
Temkin, O. (1971). The Falling Sickness, 2nd Edition revised. Baltimore: Johns Hopkins Press.
Vermeulen, J. and Aldenkamp, A.P. (1995). Cognitive side-eects of chronic antiepileptic drug
treatment: a review of 25 years of research. Epilepsy Res, 22, 6595.
Disorders: Diagnostic Criteria for Research (Tenth Revision) (ICD10). Geneva: WHO.
Yerby, M.S., Shaw, C.M. and Watson, J.M. (1986). Progressive dementia and epilepsy in a young
adult: unusual intraneuronal inclusions. Neurology, 36, 6871.
11
The risk of cognitive decline in patients with

refractory temporal lobe epilepsy
Hennric Jokeit1 and Alois Ebner2
1
Swiss Epilepsy Centre, Zurich, Switzerland
2
Epilepsy Centre Bethel, Bielefeld, Germany
Introduction
Patients with refractory temporal lobe epilepsy (TLE) are at higher risk for mental
and cognitive impairments than healthy controls (Hermann et al., 1987, 1997;
Trimble, 1988). Typically patients with right-sided TLE are frequently impaired in
visuo-spatial retention tasks; patients with left-sided TLE may exhibit decits of
verbal memory. Because of these frequent and prominent memory decits it is
sometimes neglected that many TLE patients perform below healthy control sub-
jects on a variety of neuropsychological tests including intelligence measures
(Hermann et al., 1997). The probable reason is that the temporal epileptogenic
zone is not only malfunctioning but also adversely inuences remote cerebral
structures resulting in additional cognitive decits (Engel et al., 1991; Lders and
Awad, 1991). One of our recent studies conrmed that assumption (Jokeit et al.,
1997).
We investigated 96 TLE patients by FDG-PET and neuropsychological assess-
ment who had a corresponding unilateral temporal hypometabolism, left hemi-
sphere speech dominance, full scale IQ of 70 and no extra-temporal lesion on
MRIs. The regional glucose metabolism was determined in each patient in homol-
ogous regions including prefrontal cortex. A multivariate analysis of variance
revealed that the observed prefrontal metabolic disturbances, that are remote from
the temporal epileptogenic zone, were associated with impaired intellectual abil-
ities. Patients who demonstrated prefrontal metabolic disturbances performed
worse on verbal as well as performance IQ measures than patients without prefron-
tal metabolic disturbances. Although patients who demonstrated prefrontal meta-
bolic disturbances had an earlier epilepsy onset, the revealed association with
cognitive impairment was unrelated to the age at onset. Nevertheless, age at epi-
lepsy onset is a well-documented risk factor for cognitive impairment (Bourgeois
et al., 1983; Glosser et al., 1997). It is assumed that an early epilepsy onset aects
152
153 Risk of cognitive decline in refractory TLE
considerably the maturation of brain functions and structures as well as the acqui-
sition of complex knowledge and abilities.
One of the most frequent questions asked by epilepsy patients and their relatives
is whether seizures are destructive and contribute to progressive decline of intellec-
tual abilities. Generally dementia is a very rare syndrome in TLE patients and does
not represent the typical course of refractory TLE. However, a growing number of
clinical and experimental studies suggest a slow but ongoing progression of symp-
toms with increasing duration of refractory TLE or total number of lifetime sei-
zures. A long duration of intractable epilepsy is related to a considerable number
of focal or generalized seizures, pathological interictal electric brain activity,
chronic and transient metabolic disturbances (Arnold et al., 1996; Savic et al., 1997;
Theodore et al., 1989) and chronic antiepileptic medication usually with high
serum levels (Hermanns et al., 1996). It is suggested that these noxious factors may
induce secondary neurophysiological and structural long-term changes (Beach et
al., 1995; Ben-Ari and Represa, 1990; Bengzon et al., 1997; Jokeit et al., 1997; Marsh
et al., 1997; Multani et al., 1994; Tasch et al., 1999; Theodore and Gaillard 1999;
Theodore et al., 1999).
A few neuropsychological studies have been aimed at elucidating this question,
whether the cognitive abilities of patients deteriorate with an increasing duration
of intractable epilepsy. But neither short-term longitudinal nor cross-sectional
studies demonstrated a convincing relationship between psychometric intelligence
and the duration of epilepsy in samples of adult patients (Bourgeois et al., 1983;
Brown, 1996; Brown and Vaughan, 1988; Dodrill and Wilensky, 1992; Rodin et al.,
1986; Seidenberg et al., 1981; Selwa et al., 1994; Strauss et al., 1995; Trimble, 1988).
On one hand, the absence of an evident duration eect suggests that probably no
dramatic cognitive changes occur within periods of some years in adult TLE
patients. On the other hand, methodological restrictions, for example, a limited
time range of longitudinal studies which rarely exceeds a decade, an undetected
cohort bias in cross-sectional studies, or confounded variables might cover possible
duration eects. Additionally, in studies with small sample sizes, the possibility of
a Type II error is frequently neglected. The results of densiometric and volumetric
cross-sectional studies in TLE patients demonstrate that dierences in neuro-
psychological measures became signicant only if patients diered in decades of
the duration of epilepsy (Barr et al., 1997; Breier et al., 1997; Jokeit et al., 1999;
Mathern et al., 1995a, b; Multani et al., 1994; Salmenper et al., 1998). Moreover,
these studies indicate that neuronal injury within and beyond the temporal lobes
continues to occur with ongoing seizure activity in TLE patients. However, it is well
known that the brain possesses a large degree of redundancy and plasticity, and has
compensatory mechanisms that may prevent or postpone a cognitive decline due
to small but ongoing brain damage (Calne et al., 1986; Lewin, 1980). Therefore, the
154 H. Jokeit and A. Ebner
individual brain reserve or spare capacity may also considerably inuence the
course of cognitive changes in patients with refractory TLE.
Cross-sectional studies
There are dierent approaches to infer cognitive changes along the time axis. From
a scientic as well as methodological point of view, prospective longitudinal studies
are generally superior to cross-sectional studies. Clinical observations and results
of longitudinal studies suggest that there is, if any, no rapid cognitive decline in
patients with refractory TLE. Consequently, longitudinal studies in TLE patients
are confronted with the problem of probably small eect sizes. Hence, to provide
statistical evidence, large samples of patients have to be observed over long periods
of time, probably exceeding decades. In contrast, cross-sectional studies allow the
recruitment of large sample sizes. However, the existence of a duration eect only
can be inferred from interindividual dierences in the duration of illness. Hence,
the interpretation of duration eects strictly presupposes that the patients were
recruited from the same population. Otherwise a cohort bias might considerably
aect the results. Although conclusions drawn from cross-sectional studies are
limited in some respects they may reveal trends that might be covered in longitu-
dinal studies restricted by time and sample size.
Also, dependent variables, usually intelligence measures, can be treated dieren-
tially as we demonstrate by the following studies that are based on independent
samples. In the rst study we considered dierences between an estimated measure
of former or premorbid intelligence and the current performance in an intelligence
test as a function of duration of epilepsy. This seems to be the only way to infer indi-
vidually a cognitive decline using psychometric instruments during a single neuro-
psychological investigation. In the second study we directly related the duration of
epilepsy with the current IQ test results. However, this approach is only appropri-
ate for sample studies.
Neuropsychological investigations of a patient frequently include so-called intel-
ligence trace tests to estimate the former intelligence in order to compare it with
current test results. Several studies have shown that most patients with cerebral
lesions or early dementia are unimpaired in intelligence trace tests, whereas the
same patients do worse in standard IQ tests. The dierence between the intelligence
trace test and standard IQ tests is considered to be a measure of cognitive decline
(Figure 11.1). If intellectual abilities deteriorate with increasing duration of epi-
lepsy the dierence between both measures should become larger. In our study
(Jokeit et al., 2000) we used the passive vocabulary-intelligence test (MWTB) as
an intelligence trace test (Lehrl, 1995). In this test, patients have to identify a real
word among four pseudo-words in rows of increasing diculty.
30
IQ Difference
20
10
10
20
30
40
50
0 10 20 30 40 50
Duration of epilepsy (years)

Figure 11.1. Difference values of the estimated WAIS-R full scale IQ minus MWT-B-IQ (vocabulary
intelligence), as a function of the duration of epilepsy. Data of 37 patients with refractory
temporal lobe epilepsy are shown. The linear regression function confirms a negative
correlation between the difference value and the duration of epilepsy. (With permission
from Jokeit et al. 2000.)
Patients with a long duration of epilepsy more frequently demonstrated a con-

siderable dierence between both IQs compared with patients with a shorter dura-
tion of TLE.
In a larger study (Jokeit and Ebner, 1999), we examined the eects of duration
of epilepsy on the performance in a standard IQ-test as an indicator of global cog-
nitive abilities and integrity of higher brain functions. Furthermore, the inuence
of the variable education on psychometric intelligence and its interaction with
the duration of epilepsy was investigated. Epidemiological studies identied that
education as an indicator of brain reserve modies the clinical expression of

dementia and Alzheimers disease (Evans et al., 1997; Satz, 1993; Schmand et al.,
1997; Stern et al., 1994; Zhang et al., 1990).
We studied 209 consecutive patients of our epilepsy surgery programme with
TLE who fullled the following criteria: seizures of unilateral temporal origin as
demonstrated by continuous interictal and ictal video/EEG-monitoring, unilateral
lesions within the temporal lobes as demonstrated by MRI, and Full Scale
Intelligence Quotient (FSIQ) greater than 55 to exclude severely mentally impaired
patients. Detailed information on clinical and demographic variables are given in
the original study (Jokeit and Ebner, 1999).
Patients underwent a neuropsychological evaluation designed for patients with
TLE. To have a comprehensive psychometric measure of global cognitive abilities,
which was well normed to age-matched healthy controls, we used the Full Scale
Intelligence Quotient (FSIQ) from the German version of the Wechsler Adult
Intelligence Scale Revised (WAISR) (Tewes, 1991; Wechsler, 1981). The FSIQ
was estimated from the subtests Information, Comprehension, Similarities, Digit
Symbol, Picture Completion and Block Design.
First a data set of 16 independent variables for each patient was submitted to a
linear multiple regression analysis. The analyses on FSIQ revealed signicant con-
tributions by the variables education and duration of epilepsy. The equation
explained 34.6% of total variance. No further variable contributed signicantly.
To reveal possible factor interactions and to exclude linear eects of covariates in
a one-way ANOVA model, the continuous variable duration of epilepsy was
recoded into three values: 15 years0, 1530 years1, and 30 years of refrac-
tory epilepsy2. The side of seizure origin was submitted as the second factor (left,
right). The presence or absence of lesions beyond mesio-temporal structures in
MRI scans was submitted as the third factor (0, 1). Education, age at epilepsy onset,
ranked frequency of interictal epileptiform discharges, ranked frequency of habit-
ual seizures, presence of secondarily generalized seizures in a patients history,
ranked frequency of secondarily generalized seizures within the last year, serum
level for carbamazepine, phenytoin and phenobarbital, and AED mono- or poly-
therapy were controlled as covariates. Only the covariate education was signi-
cantly related to FSIQ. After removing linear eects of covariates the factor
duration of epilepsy was signicant. The factors side of seizure onset and presence
or absence of temporal lesions beyond mesio-temporal structures did not reach sig-
nicance. No interactions reached signicance. Post-hoc contrasts adjusted by
covariates revealed that patients with a duration of epilepsy for more than 30 years
performed worse than patients with less than 15 years of epilepsy and patients with
1530 years of epilepsy. Patients with less than 15 years of epilepsy did not dier
from patients with 1530 years of epilepsy in FSIQ values.
Since it is known that variable education explains a considerable amount of var-

iance of FSIQ values, we computed an ANOVA with factors duration of epilepsy (0,
1, 2) and education (low, high) to reveal possible interactions and to specify the
eect of variable education. The educational level was dichotomized into low
(patients who did not attend or did not nish at secondary school, Realschule in
Germany, n109) and high (patients who nished at least at secondary school,
n100). The side of seizure origin, presence or absence of lesions beyond mesio-
temporal structures, age at epilepsy onset, ranked frequency of interictal epilepti-
form discharges, ranked frequency of habitual seizures, presence of generalized
seizures in a patients history, ranked frequency of secondarily generalized seizures
within the last year, serum level for carbamazepine, phenytoin and phenobarbital,
and AED mono- or polytherapy were controlled as covariates. The ANOVA
revealed eects of the factors education and duration of epilepsy on the FSIQ. The
interaction between both factors did not reach signicance. No covariate was sig-
nicantly related to FSIQ. Figure 11.2 shows mean FSIQ values for both educa-
tional groups as a function of duration of TLE. Contrasts adjusted for covariates
revealed that the mean FSIQ values of groups with less than 15 years and 1530
years of TLE did not dier in patients with high educational attainment. However,
patients with more than 30 years of TLE performed worse than patients with less
than 15 years or 1530 years of epilepsy. Patients with low educational attainment
and less than 15 years of TLE performed better than patients with 1530 years of
TLE and patients with more than 30 years of TLE. But there was no signicant
dierence between patients with 1530 years and more than 30 years of TLE in the
low education group. Although not conrmed by an ANOVA interaction, these
contrasts suggest a duration-dependent dierence in mean FSIQ values of patients
with low and high educational attainment. The absence of a signicant ANOVA
interaction probably results from a similar decremental trend of the duration eect
in low- and high-educated patients.
Conclusions
Both studies demonstrate that patients with a long history of intractable TLE were
at higher risk of cognitive impairment than patients with a shorter duration of TLE.
Interestingly, in patients with higher educational attainment, the mean FSIQ was
stable for a longer duration of TLE than in less-educated patients.
We revealed that the educational level and the duration of epilepsy were the best
predictors for psychometric intelligence. Then we provided evidence that only a
long duration of TLE (more than 30 years) was related to impaired psychometric
intelligence in the total sample. The linear inuence of the variables age at epilepsy
onset, educational level of patients, patients serum level of rst-line antiepileptic
120
Mean FSIQ
110
100
90
80
lowhigh lowhigh lowhigh

70
15 1530 30
Duration of epilepsy (years)
Figure 11.2. Mean FSIQ values with 95% confidence intervals of patients with low (n109) and high
(n100) educational attainment for groups with a duration of epilepsy 15 years, 1530
years and 30 years. The factors education and duration of epilepsy were significant
(P0.01). Asterisks (* P0.05; ** P0.01; one tailed) indicate significant contrasts
between adjacent duration groups adjusted for covariates. (With permission from Jokeit
and Ebner, 1999.)
drugs, polypharmacy, frequency of habitual and secondarily generalised seizures

and frequency of interictal epileptiform discharges on psychometric intelligence
were statistically controlled. Age at testing controlled for duration of TLE was not
signicantly related to FSIQ. The factors side of TLE and presence or absence of
lesions beyond the mesio-temporal structures did not show main eects or inter-
actions. Therefore, the eect of duration of epilepsy cannot be attributed to those
covariates and factors. The patients educational level captured the most amount of
FSIQ variance. This could explain why in contrast to other studies (Strauss et al.,
1995) the remaining covariates (e.g. age at epilepsy onset) did not reach signi-
cance.
It is reasonable to assume that human brains develop a functional reserve or have
a spare capacity to cope with a stepwise neuronal loss by eciency, redundancy,
plasticity and reorganization (Lewin, 1980; Meier-Ruge et al., 1991; Stern et al.,
1996). Studies of dierent degenerative brain disorders (e.g. Parkinsons disease,
vascular and Alzheimers dementia) suggest that a functional decline becomes
apparent only if a certain amount of brain parenchyma is insulted (Boone et al.,
1992; Hornykiewicz, 1988; Pasquier and Leys, 1997; Small et al., 1995; Tomlinson et
al., 1970). A long duration of intractable TLE is related to a considerable number of
focal or secondarily generalized seizures, pathological interictal electric brain activ-
ity, chronic and transient metabolic disturbances due to morphological lesions
(Arnold et al., 1996), seizures (Savic et al., 1997) and antiepileptic medication
(Theodore et al., 1989), often the chronic antiepileptic medication with usually high
serum levels (Hermanns et al., 1996). It is suggested that each of these factors may
separately adversely aect cognitive functioning (Dreifuss, 1992; Lesser et al., 1986).
The presence of reactive microglia (Beach et al., 1995), reduced dendritic spine
density, dendritic swellings (Multani et al., 1994) and senile plaques (Mackenzie
and Miller, 1994) in adult temporal lobe specimens suggests that neuronal injury
continues to occur with ongoing seizure activity in TLE patients. Multani et al.
(1994) demonstrated a correlation between decreased dendritic spine density
remote from the epileptogenic zone and duration of seizure history. In patients
with mesial TLE densiometric techniques have revealed secondary declines in hip-
pocampal neuron densities associated with long histories of habitual seizures
(Mathern et al., 1995c, 1996). Recent studies have suggested a secondary decline of
hippocampal volume and temporal lobe metabolism in refractory TLE patients
(Barr et al., 1997; Breier et al., 1997; Jokeit et al., 1999; Salmenper et al., 1998).
Hermann et al. (1997) reported that patients with hippocampal sclerosis had more
generalized cognitive impairment, a signicantly longer history of intractable TLE
and a lower educational level than TLE patients without signicant hippocampal
sclerosis. We assume that a cumulation of small neuro-degenerative eects of
noxious neurochemical agents, abnormal brain electric events, and metabolic dis-
turbances over decades of epilepsy, accompanied by ageing, may increase the prob-
ability that the functional brain reserve or spare capacity is exhausted at a younger
age than expected (mean age of patients with TLE duration 30 years was 44
years), and deterioration of cognitive functions may begin (Meier-Ruge et al., 1991;
Mori et al., 1997; Stern et al., 1996).
The extent of functional reserve and therefore the vulnerability of brain func-
tions may vary considerably between people. It was suggested that higher educa-
tional attainment is related to a higher reserve against cognitive impairment due to
stepwise ongoing brain injury (Satz, 1993; Timiras, 1995). Our results of analyses
in patients with lower and higher educational attainment are in accordance with
ndings of epidemiological studies on dementia and Alzheimerss disease (Evans et
al., 1997; Satz, 1993; Schmand et al., 1997; Stern et al., 1994; Zhang et al., 1990). In
patients with higher educational attainment the mean FSIQ was stable for a longer
duration of TLE than in less-educated patients. Thus, higher educational attain-
ment as an indicator of higher cognitive reserve might delay the onset of cognitive
decline in patients with intractable TLE.
The fact that we found signicant eects in the whole sample analysis only in
patients with a history of intractable epilepsy lasting longer than three decades may
question conclusions drawn from negative ndings of studies on adverse eects of
duration of refractory TLE (Mackenzie et al., 1996; Selwa et al., 1994). Based on our
results, only prospective long-term studies which exceed three decades might reveal
the causes of a presumable decline in cognitive functioning of patients with intract-
able TLE. Such studies may solve the important question as to whether certain epi-
leptic syndromes are progressive disorders (Girvin, 1992; Gloor, 1991; Lesser et al.,
1986; Mathern et al., 1996; Sadzot 1997; Sutula et al., 1989). However, today well-
controlled cross-sectional studies comparing dierent well-dened epileptic syn-
dromes and including retest measures to compare dierent treatment strategies
may help to isolate adverse factors and to identify patients with increased risk of
cognitive decline and symptom progression.
R E F E R E N C ES
Arnold, S., Schlaug, G., Niemann, H. et al. (1996). Topography of interictal glucose hypo-
metabolism in unilateral mesiotemporal epilepsy. Neurology, 46, 142230.
Barr, W.B., Ashtari, M. and Schaul, N. (1997). Bilateral reductions in hippocampal volume in
adults with epilepsy and a history of febrile seizures. J Neurol Neurosurg Psychiatry, 63, 4617.
Beach, T.G., Woodhurst, W.B., MacDonald, D.B. and Jones, W.M. (1995). Reactive microglia in
hippocampal sclerosis associated with human temporal lobe epilepsy. Neurosci Lett, 191,
2730.
Ben-Ari, Y. and Represa, A. (1990). Brief seizure episodes induce long-term potentiation and
mossy bre sprouting in the hippocamus. Trends Neurosci, 13, 31218.
Bengzon, J., Kokaia, Z., Elmer, E., Nanobashvili, A., Kokaia, M. and Lindvall, O. (1997).
Apoptosis and proliferation of dentate gyrus neurons after single and intermittent limbic sei-
zures. Proc Natl Acad Sci USA, 94, 104327.
Boone, K.B., Miller, B.L., Lesser, I.M. et al. (1992). Neuropsychological correlates of white-matter
lesions in healthy elderly subjects. Arch Neurol, 49, 54954.
Bourgeois, B.F.D., Prensky, A.L., Palkes, H.S., Talent, B.K. and Busch, S.G. (1983). Intelligence in
epilepsy: a prospective study in children. Ann Neurol, 14, 43844.
Breier, J.L., Mullani, N.A., Thomas, A.B. et al. (1997). Eects of duration of epilepsy on the
uncoupling of metabolism and blood ow in complex partial seizures. Neurology, 48, 104753.
Brown, S.W. (1996). Epilepsy dementia: Intellectual deterioration as a consequence of epileptic
seizures. Epilepsia, 37 (Suppl. 4), S1223.
Brown, S.W. and Vaughan, M. (1988). Dementia in epileptic patients. In Epilepsy, Behaviour and
Cognitive Function, ed. M.R. Trimble and E.H. Reynolds, pp. 17788. Chichester: John Wiley
& Sons.
Calne, D.B., Eisen, A., McGeer, E. and Spencer, P. (1986). Alzheimers disease, Parkinsons disease,
and motoneurone disease: abiotrophic interaction between ageing and environment? Lancet,
2 (8515), 106770.
Dodrill, C.B. and Wilensky, A.J. (1992). Neuropsychological abilities before and after 5 years of
stable antiepileptic drug therapy. Epilepsia, 33, 32734.
Dreifuss, E.F. (1992). Cognitive function victim of disease or hostage to treatment? Epilepsia,
33 (Suppl. 1), S712.
Engel, J. Jr, Brandler, R., Grith, N.C. and Caldecott-Hazard, S. (1991). Neurobiological evi-
dence for epilepsy-induced interictal disturbances. Adv Neurol, 55, 97111.
Evans, D.A., Hebert, L.E., Becket, L.A. et al. (1997). Education and other measures of socioeco-
nomic status and risk of incident Alzheimer disease in a dened population of older persons.
Arch Neurol, 54, 1399405.
Girvin, J.P. (1992). Is epilepsy a progressive disorder? J Epilepsy, 5, 94104.
Gloor, P. (1991). Mesial temporal sclerosis: historical background and an overview from a
modern perspective. In Epilepsy Surgery, ed. H.O. Lders, pp. 689703. New York: Raven Press.
Glosser, G., Cole, L.C., French, J.A., Saykin, A.J. and Sperling, M.R. (1997). Predictors of intel-
lectual performance in adults with intractable temporal lobe epilepsy. J Int Neuropsychol Soc,
3, 2529.
Hermann, B.P., Wyler, A.R. and Richey, E.T. (1987). Epilepsy frontal lobe, and personality. Biol
Hermann, B.P., Seidenberg, M., Schoenfeld, J. and Davies, K. (1997). Neuropsychological char-
acteristics of the syndrome of mesial temporal lobe epilepsy. Arch Neurol, 54, 36976.
Hermanns, G., Noachtar, S., Tuxhorn, I., Holthausen, H., Ebner, A. and Wolf, P. (1996).
Systematic testing of medical intractability for carbamazepine, phenytoin, and phenobarbital
or primidone in monotherapy for patients considered for epilepsy surgery. Epilepsia, 37,
6759.
Hornykiewicz, O. (1988). Neurochemical pathology and the etiology of Parkinson disease. Mt
Sinai J Med, 55, 1120.
Jokeit, H., Seitz, R.J., Markowitsch, H.J., Neumann, N., Witte, O.W. and Ebner, A. (1997).
Prefrontal asymmetric interictal glucose hypometabolism and cognitive impairment in
patients with temporal lobe epilepsy. Brain, 120, 228394.
Jokeit, H., Ebner, A., Arnold, S. et al. (1999). Bilateral depressions of hippocampal volume,
glucose metabolism, and Wada hemispheric memory performance are related to the duration
of mesial temporal lobe epilepsy. J Neurol, 246, 92633.
Jokeit, H., Luerding, R. and Ebner, A. (2000). Cognitive impairment in temporal-lobe epilepsy.
Lancet, 355, 101819.
Lehrl, S. (1995). Mehrfachwahl-Wortschatz-Intelligenztest MWT-B [multiple-choice vocabulary
intelligence test]. Balingen: Perimed-spitta.
Lesser, R.P., Lders, H., Wyllie, E., Dinner, D.S. and Morris, III H.H. (1986). Mental deteriora-
tion in epilepsy. Epilepsia, 27 (Suppl. 2), S10523.
Lewin, R. (1980). Is your brain really necessary? Science, 210, 12324.
Lders, H.O. and Awad, I. (1991). Conceptual considerations. In Epilepsy Surgery, ed. H.O.
Lders, pp. 5162. New York: Raven Press.
Mackenzie, I.R.A. and Miller, L.A. (1994). Senile plaques in temporal lobe epilepsy. Acta
Neuropathol, 87, 50410.
Mackenzie, I.R.A., McLachlan, R.S., Kubu, C.S. and Miller, L.A. (1996). Prospective neuro-
psychological assessment of nondemented patients with biopsy proven senile plaques.
Marsh, L., Morrell, M.J., Shear, P.K. et al. (1997). Cortical and hippocampal volume decits in
temporal lobe epilepsy. Epilepsia, 38, 57687.
Mathern, G.W., Pretorius, J.K. and Babb, T.L. (1995a). Inuence of the type of initial precipitat-
ing injury and at what age it occurs on course and outcome in patients with temporal lobe sei-
zures. J Neurosurg, 82, 2207.
Mathern, G.W., Babb, T.L., Pretorius, J.K., Melendez, M. and Lvesque, M.F. (1995b). The patho-
physiologic relationships between lesion pathology, intracranial ictal EEG onsets, and hippo-
campal neuron losses in temporal lobe epilepsy. Epilepsy Res, 21, 13347.
Mathern, G.W., Babb, T.L., Vickrey, B.G., Melendez, M. and Pretorius, J.K. (1995c). The clini-
cal-pathogenic mechanism of hippocampal neuron loss and surgical outcomes in temporal
lobe epilepsy. Brain, 118, 10518.
Mathern, G.W., Babb, T.L., Leite, J.P., Pretorius, J.K., Yeoman, K.M. and Kuhlman, P.A. (1996).
The pathogenic and progressive features of chronic human hippocampal epilepsy. Epilepsy Res,
26, 15161.
Meier-Ruge, W., Hunziker, O. and Iwango, P. (1991). Senile dementia: a threshold phenome-
non of normal aging? A contribution to the functional reserve hypothesis of the brain. Ann NY
Acad Sci USA, 621, 10418.
Mori, E., Hirono, N., Yamashita, H. et al. (1997). Premorbid brain size as a determinant of
reserve capacity against intellectual decline in Alzheimers disease. Am J Psychiatry, 154,
1824.
Multani, P., Myers, R.H., Blume, H.W., Schomer, D.L. and Sotrel, A. (1994). Neocortical dendritic
pathology in human partial epilepsy: a quantitative Golgi study. Epilepsia, 35, 72836.
Pasquier, F. and Leys, D. (1997). Why are stroke patients prone to develop dementia? J Neurol,
244, 13542.
Rodin, E.A., Schmaltz, S. and Twitty, G. (1986). Intellectual functions of patients with child-
hood-onset epilepsy. Dev Med Child Neurol, 28, 2533.
Sadzot, B. (1997). Epilepsy: a progressive disease? Br Med J, 314, 3912.
Salmenper, T., Klviinen, R., Partanen, K. and Pitknen, A. (1998). Hippocampal damage
caused by seizures in temporal lobe epilepsy. Lancet, 351, 35.
Satz, P. (1993). Brain reserve capacity on symptom onset after brain injury: a formulation and
review of evidence for threshold theory. Neuropsychology, 7, 27395.
Savic, I., Altshuler, L., Baxter, L. and Engel, J. Jr (1997). Pattern of interictal hypometabolism in
PET scans with udeoxyglucose F 18 reects prior seizure types in patients with mesial tem-
poral lobe seizures. Arch Neurol, 54, 12936.
Schmand, B., Smit, J.H., Geerlings, M.I. and Lindeboom, J. (1997). The eects of intelligence and
education on the development of dementia. A test of the brain reserve hypothesis. Psychol Med,
27, 133744.
test-retest scores on the Wechsler Adult Intelligence Scale. Epilepsia, 22, 7583.
Selwa, L.M., Berent, S., Giordani, B., Henry, T.R., Buchtel, H.A. and Ross, D.A. (1994). Serial cog-
nitive testing in temporal lobe epilepsy: longitudinal changes with medical and surgical ther-
apies. Epilepsia, 35, 7439.
Small, G.W., Mazziota, J.C., Collins, M.T. et al. (1995). Apolipoprotein E type 4 allele and cere-
bral glucose metabolism in relatives at risk for familial Alzheimer disease. J Am Med Assoc, 273,
9427.
Stern, R.A., Silva, S.G., Chaisson, N. and Evans, D.L. (1996). Inuence of cognitive reserve on
neuropsychological functioning in asymptomatic human immunodeciency virus-1 infec-
tion. Arch Neurol, 53, 14853.
Stern, Y., Gurland, B., Tatemichi, T.K., Tang, M.X., Wilder, D. and Mayeux, R. (1994). Inuence of
education and occupation on the incidence of Alzheimers disease. J Am Med Assoc, 271, 100410.
Strauss, E., Loring, D., Chelune, G. et al. (1995). Predicting cognitive impairment in epilepsy:
ndings from the Bozeman Epilepsy Consortium. J Clin Exp Neuropsychol, 17, 90917.
Sutula, T., Cascino, G., Cavazos, J., Parada, I. and Ramirez, L. (1989). Mossy ber synaptic reor-
ganization in the epileptic human temporal lobe. Ann Neurol, 26, 32130.
Tasch, E., Cendes, F., Li, L.M., Dubeau, F., Andermann, F. and Arnold, D.L. (1999).
Neuroimaging evidence of progressive neuronal loss and dysfunction in temporal lobe epi-
lepsy. Ann Neurol, 45, 56876.
Tewes, U. (1991). Manual des Hamburg-Wechsler Intelligenztest fr Erwachsene Revision 1991.
Bern: Verlag Hans Huber.
Theodore, W.H. and Gaillard, W.D. (1999). Association between hippocampal volume and epi-
lepsy duration. Ann Neurol, 46, 800.
Theodore, W.H., Bromeld, E. and Onorati, L. (1989). The eect of carbamazepine on cerebral
glucose metabolism. Ann Neurol, 25, 51620.
Theodore, W.H., Bahita, S., Hatta, J. et al. (1999). Hippocampal atrophy, epilepsy duration, and
febrile seizures in patients with partial seizures. Neurology, 52, 1326.
Timiras, P.S. (1995). Education, homeostasis, and longevity. Exp Gerontol, 30, 18998.
Tomlinson, B.E., Blessed, G. and Roth, M. (1970). Observations on the brains of demented old
people. J Neurol Sci, 11, 20542.
Trimble, M.R. (1988). Cognitive hazards of seizure disorders. Epilepsia, 29 (Suppl. 1), S1924.
Wechsler, D. (1981). WAISR Manual. New York: The Psychological Corporation.
Zhang, M., Katzman, R., Salmon, D. et al. (1990). The prevalence of dementia and Alzheimers
disease in Shanghai, China: impact of age, gender, and education. Ann Neurol, 27, 42837.
12
Behavioural and neuropsychological aspects

of frontal lobe epilepsy
Christoph Helmstaedter
University Clinic of Epileptology Bonn, Bonn, Germany
Introduction
To date, neurobiological interest in behaviour and epilepsy has been concerned pri-
marily with temporal lobe epilepsies (TLE), and mesial temporal lobe epilepsies
(mTLE) in particular. This concentration on TLE is mostly due to the fact that this
type of epilepsy represents the majority of the focal epilepsies in the Bonn series
of patients with pharmacoresistant epilepsies this is about 80% and that mTLE
often forms an entity within the focal epilepsies regarding pathology (hippocam-
pal sclerosis), a frequent history of febrile convulsions, an early onset of epilepsy,
and memory problems as the prominent neuropsychological impairment. In TLE,
the aected cerebral structures and epileptogenic region are mostly circumscribed,
and structural pathology can be well quantied by quantitative MRI (T2 relax-
ometry and volumetry) or postoperative histopathological examinations of the
resected specimen. Frequency, homogeneity and quantiable pathology provide
ideal prerequisites for the study of the functional and behavioural correlates of
TLE. Great progress has been made during recent years, at least with respect to the
neuropsychological and cognitive aspects of TLE. Recent developments in the eld,
however, show that it is well recognized that temporal lobe functioning involves
more than memory and that its role in emotion and psychiatric symptoms is being
rediscovered.
The conditions we meet with frontal lobe epilepsy (FLE) are quite dierent.
Frontal lobe epilepsies, in spite of the size of the frontal lobes, are less frequent than
the temporal lobe epilepsies. In our own series, patients with FLE represent about
15% of the patients with pharmacoresistant epilepsies. Site and type of the under-
lying pathology are very heterogeneous. Furthermore, ictal and interictal clinico-
electric manifestations of FLE are infrequently localized, because multiple
connections to most other brain areas enable fast and widely distributed propaga-
tion of epileptic activity. The functional correlates of frontal pathology in epilepsy
are thus less well understood.
164
165 Behavioural and neuropsychological aspects of FLE
Frontal lobe and behaviour disorders
Correspondent to its cytoarchitectonic structure, the frontal lobe is traditionally

divided into two parts, which are important in two major areas. The posterior part
controls motor movement and is subdivided into a premotor and a motor area,
which control movement preparation and actual execution of movement respec-
tively. The anterior part of the frontal lobe, the prefrontal cortex, is especially
important in higher mental function, as in anticipation and planning, initiative,
judgement, and in aect control, will power, and the determination of personality
(Bechara et al., 1999; Raine et al., 2000). The prefrontal cortex can be further sub-
divided into the dorso-lateral cortex and the orbito-frontal cortex. This subdivision
of the prefrontal cortex is still simple and it should be noted that the orbito-frontal
cortex itself is a heterogeneous area connected with a wide range of other prefron-
tal, limbic, premotor, sensory areas and subcortical nuclei (Cavada et al., 2000). For
the purpose of this article it is important to know that there is evidence that damage
of the dorso-lateral part of the prefrontal cortex is more associated with impair-
ment of executive functions and functions of working memory, whereas damage of
the orbito-frontal cortex leads to impairment of the choice of behaviour, the estab-
lishment of emotional valences, and the evaluation and balancing of the past and
future consequences of a given behaviour (Bechara et al., 2000; Rolls, 2000; Sarazin
et al., 1998). Studies in common marmosets suggest a dissociation between the
lateral and the orbital-medial division of the prefrontal cortex according to which
the rst selects and controls actions on the basis of higher-order rules while the
latter controls dierent behaviour on the basis of lower-order rules (see Roberts
and Wallis, 2000).
The signicance of the orbito-frontal cortex for social and interpersonal behav-
iour recently came into focus when Anderson et al. (1999) reported two patients,
one with an accident at 15 months, the other with a frontal tumour resection at 16
months, both showing severe impairment of social and moral behaviour.
Traditionally, behavioural dyscontrol has been attributed to temporo-limbic
structures. Evidence for the involvement of the amygdala in aggression comes from
human and animal stimulation studies, from activating and inhibiting eects of anti-
epileptic drugs on aggression, and recently also from direct correlations of amygdala
volumes with aggression in patients with mesial epilepsies (Azouvi et al., 1999; Bearn
and Gibson, 1998; Trimble and Van Elst, 1999; Van Elst et al., 2000). However, aggres-
sion associated with the amygdala seems more defensive than oensive in nature
(Kalynchuk et al., 1999). Disinhibition phenomena, or a loss of impulse control as it
is observed with frontal lesions, may be an additional prerequisite for showing
impulsive aggressive behaviour. The orbito-frontal cortex as the border zone between
the frontal lobe and the limbic system is thus critical for the linkage of the frontal and
166 C. Helmstaedter
limbic aspects of behavioural dyscontrol disorders. Another important area in this

respect is the anterior-cingulate gyrus, which is also strongly connected to the amyg-
dala, and whose damage has also been associated with deviant social behaviour and
aect states (for overview see Devinsky et al., 1995).
The nding of antisocial and aggressive behaviour with frontal lobe damage is
not in itself new. A prominent and often cited example is the historic case of
Phineas Gage, who after an accident with a severe frontal brain injury changed from
a well-mannered man into an irresponsible and convention-neglecting person
(Damasio et al., 1994; Harlow, 1868). New in the study of Anderson et al. (1999) is
the nding that whether patients not only display severe behavioural disorders but
also fail to see the moral of the behaviour depends on the age at the lesion onset
(Dolan, 1999). Consequently, the orbito-frontal cortex seems not only important
for behaviour control but also for the acquisition of social and interpersonal rules.
It is important to note that irresponsible, aggressive and sociopathic behaviours
can occur independent of intellectual abilities, which are often well preserved in
frontal lesions. Other areas in which the orbital and medial prefrontal cortex are
believed to play a central role are addictive behaviour, attention-decit hyperactiv-
ity disorder, negative emotion and major depression (Drevets and Raichle, 1996;
London et al., 2000; Northo et al., 2000; Rubia et al., 2000). Davidson et al. (2000)
propose a key role of the prefrontal cortex in the regulation of emotion in violent
subjects and those predisposed to violence.
A theoretical basis of the role of the prefrontal cortex in the interplay of cogni-
tion and emotion has been provided with Damasios somatic marker hypothesis
(Damasio, 1996). This proposes that responses to external stimuli do not rely on
either conditioning processes or cognition alone, but on somatic marker signals
or autonomic response sets, which determine the conscious/unconscious connec-
tion between stimulus conditions, feelings and behaviour.
Epilepsy and behaviour disorders
Single case reports of behavioural and personality disorders in patients with severe
brain lesions often appear dramatic. However, with respect to focal epilepsies, these
reports nevertheless raise the question of whether there might be parallels in the
behaviour when epilepsy aects the same brain regions. With the exception of rare
cases of ictal aggression, postictal confusional states or psychosis (Marsh and
Krauss, 2000), behaviour and personality disorders observed in patients with FLE
appear less severe. Furthermore, as with TLE, one can hardly expect to nd the
prototypical frontal epileptic personality or Wesensnderung. Personality is by
denition more trait than state dependent and, particularly in epilepsy, it is quite
dicult to determine whether a given behaviour has trait characteristics or not.
Table 12.1. Factors affecting cognitive and mood states in epilepsy
States of epilepsy
preictal
ictal
postictal
interictal
(seizure free after successful surgery)
Seizures
frequency
generalization
nonconvulsive status epilepticus
Epileptic dysfunction
local versus distant eects
Lesion
e.g. alien tissues vs. migration and developmental disorders (confounded with dierent ages
at lesion/epilepsy onset)
extent, location, lateralization
Antiepileptic drugs
positive vs. negative psychotropic eects
individual incompatibility
drug-induced encephalopathy
intoxication
Conclusions about persistence and continuity require follow-up observations with

longer time intervals. In epilepsy several factors can be discerned, which can lead
to more or less reversible changes in patients cognitive abilities and mood states
(see Table 12.1). Furthermore, although we can now look back on a long history of
successful epilepsy surgery, it is still not clear to what degree the fact of having sei-
zures is a prerequisite of behaviour and mood disorders in epilepsy.
The patient with epilepsy must always be seen in his state relative to seizures, e.g.
whether he is ictal, postictal or interictal. According to recent ndings with regard
to seizure prediction by nonlinear measures of complexity loss as recorded by intra-
cranial EEG, signicant seizure-precipitating drops in complexity, i.e. synchroniza-
tion, can be recorded long before the seizure starts (Elger and Lehnertz, 1998).
Accordingly, one must assume also preictal states, which would t well with
patients reports of increased dysphoric mood and cognitive problems before the
seizure starts. Finally, since many patients can now become seizure-free on a long-
term basis by epilepsy surgery, one can suggest an additional state of well-
controlled epilepsy after successful epilepsy surgery.
168 C. Helmstaedter
Epileptic activity can aect distant brain areas and cause cognitive and behav-
iour problems not related to the primary lesion or epileptogenic zone (Shulman,
1984). Notwithstanding seizures and epileptic activity, one must also dierentiate
the underlying pathologies, which can be more or less systemic, have dierent ages
of onset in life, and thus have dierent eects on brain maturation and the devel-
opment of cognition and personality. We must nally consider inuences of often
longstanding antiepileptic medication in these patients. Antiepileptic drugs may
have positive or negative psychotropic side eects, and can show incompatibilities
in the individual patient (Schmitz, 1999). Interactive eects of pathology, epilepsy
and treatment must be considered. Apart from idiosyncratic actions, drugs can
have dierent eects in lesion and nonlesion patients, and they may act dierently
dependent on seizure control.
Taking this into consideration, it will be shown in the next sections that there is
nevertheless evidence of specic behavioural abnormalities in patients with FLE,
which can be interpreted within a theoretical framework of frontal lobe dysfunc-
tion. This will be outlined with the example of interictal behaviour as assessed by
neuropsychological examination and self-report measures concerning quality of
life, everyday activities, personality and psychiatric symptoms. In addition, seizure
semiology and impairment during frontal lobe seizures and frontal nonconvulsive
status epilepticus will be considered, to convey an idea of what the behavioural con-
sequences of impaired frontal lobe functions in FLE might be.
The neuropsychology of frontal lobe epilepsy
The development of neuropsychology in FLE is probably best reected by Brenda

Milners (1995) description of her evaluation of Penelds patient K.M., the frontal
counterpart of the temporal patient H.M. This patient had a penetrating head
injury in 1928, developed seizures, and underwent surgery of the anterior parts of
both frontal lobes. Surgery successfully controlled the seizures and led to improved
behaviour as well as improved IQ. However, when re-evaluated with the newly
developed Wisconsin Card Sorting test in 1962 he showed severe impairment in
exible categorical thinking and concept formation whilst his IQ still was average.
This case exemplies how much outcome interpretation depends on both the test
sensitivity and test selection.
Since then there have been surprisingly few attempts to apprehend the cognitive
characteristics of patients with FLE with group studies (Shulman, 1984).
Unfortunately, most of the data from Milners era stem from patients after surgery
and thus tell us more about frontal lobe lesions than about FLE. Furthermore, most
earlier studies focused on single functions, more or less following a rather monis-
tic view of a frontal central executive (Baddeley and Hitch, 1974). Major impair-
ments indicated by these studies are problems in concept formation, response inhi-
bition (Milner, 1964), estimations (Smith and Milner, 1984), conditional associa-
tive learning (Petrides, 1985; Petrides and Milner, 1982), and prot from
information provided in advance in choice reaction tasks (Alivisatos and Milner,
1989). Focusing on memory, Delaney et al. (1980) found no dierences in meas-
ures of memory when nonoperated patients with unilateral frontal lobe foci were
compared to healthy controls. The rst of our own studies found that decits in
attention are the most signicant problem in patients with FLE (Kemper et al.,
1992).
Later systematic group studies in nonresected patients with FLE followed the
theoretical suggestion of dierent frontal subfunctions (Stuss and Benson, 1986)
and met the requirements of the manifold frontal lobe pathology by the use of a
broader range of tests. These addressed dierent aspects of attention, motor coor-
dination, psychomotor speed, uency, response inhibition, conceptual formation
and shift, as well as planning, guessing or estimating.
Between 1996 and 1999 Upton and Thompson published a series of ve articles
reporting dierent ndings on neuropsychology in their sample of 74 subjects with
FLE. Using a test battery with dierent measures of executive functions and motor
skills, they come to the conclusion that patients with epilepsy show a decit pattern
which is similar to that found in frontal lobe dysfunction in general (Upton and
Thompson, 1996a, b). As compared with patients with TLE, frontal patients show
poorer motor coordination, guessing, estimation and response inhibition.
Similarly, we found in 23 patients with FLE that cognitive problems could be
diagnosed with a broad range of 10 frontal tasks with about double as many test
parameters. The great number of test parameters, however, turned out to be highly
redundant and could be statistically reduced to four relatively independent func-
tional areas, namely psychomotor speed/attention, motor coordination, working
memory, and response inhibition. These four factors explained 70% of the total
variance. When compared with patients with TLE, those with FLE were character-
ized by prominent impairment in motor skills and response inhibition
(Helmstaedter et al., 1996). Problems in speed/attention and working memory
were frequent but they appeared rather nonspecic since they were also observed
in the temporal lobe group. This, however, does not necessarily contradict the
assumption that these are frontal functions. An imaging study of Jokeit et al. (1997)
showed in this respect that in patients with TLE, prefrontal metabolic asymmetries
are evident which are associated with frontal lobe measures and intelligence.
In another of our own studies we addressed the cognitive consequences of frontal
lobe surgery. We evaluated 33 patients pre- and postoperatively and were able to
conrm the impairment pattern of impaired motor skills and response inhibition.
We also showed that frontal surgery does not cause considerable additional damage
170 C. Helmstaedter
if eloquent cortex (SMA; motor and language area) is spared. However, when
surgery included resection of the SMA the most prominent neuropsychological
symptom besides neurological decits directly after surgery was a SMA deciency
syndrome (impairment of initiation) with aphasia (speech arrest and transcortical
aphasia) (Helmstaedter et al., 1998). Additional psychomotor slowing was
observed in lobectomies as compared to lesionectomies.
Looking closer at clinical variables which might explain the impairment pattern
in nonresected patients, no consistent picture emerges. According to Upton and
Thompson (1997a), seizure frequency and the duration of epilepsy have an eect
on performance, but this appears to be a nonspecic eect rather than a consistent
nding over dierent tests. With the exception of motor skills, which were spared
in early right-sided FLE, no systematic eect of the assumed inuence of the age at
the onset of epilepsy on cognitive development could be concluded from their data
(Upton and Thompson, 1997b).
The impact of having epileptic seizures on cognition can be demonstrated by our
postoperative ndings, indicating that in seizure-free patients adjacent functions
recovered after surgery. Comparable release eects have been also reported after
temporal lobe surgery (Hermann et al., 1988). However, one should not go so far
as to conclude that all decits are due to epileptic dysfunction and are thus rever-
sible, as has been suggested by Boone et al. (1988) in a single case report.
In conclusion, with regards to the neuropsychological ndings in FLE, it appears
that indeed dierent frontal subfunctions can be dierentiated. Nevertheless, the
measures which characterize FLE have in common the demand of adequate
response selection and initiation and response inhibition. This holds for tests which
explicitly assess interferences and response inhibition, but also for testing of motor
skills or working memory. Ending up again with a unique central executive func-
tion, one may hypothesize that the particular problem in FLE is the impairment of
response selection/initiation/inhibition with varying emphasis depending on
dierent functional areas. Which area is aected then depends on the type and the
localization of the underlying lesion, including the possibility that symptoms are
overshadowed by spreading epileptic dysfunction.
Besides this, it is important to mention that the development of appropriate test
instruments for the assessment of frontal lobe dysfunction is not yet complete and
still represents a challenge for neuropsychologists. Most psychometric tests which
allow quantication of test behaviour provide patients with a clear structure for
behaviour, i.e. with test instructions, rules, time constraints etc. This enables the
patient to behave in an ordered way and real problems with behaviour organiza-
tion arising from frontal pathology are easily overlooked. If provided with greater
freedom and demands on spontaneous interactions with complex situations, the
same patient would otherwise reveal decits. A possible solution to this dilemma
Table 12.2. Ictal frontal seizure semiology (n 15)
Localization Positive symptoms
Motor area Nearly 1:1 manifestation of seizure activity in

myoclonic and tonic or clonic motor activity
SMA (speech and motor area) Tonic posturing
Premotor Contraversive head and eye movements
Prefrontal (including gyrus cinguli) Explosive and complex motor automatisms
(including vocalizations); bizarre and hysterical
behaviour; mood change
Negative symptoms
Mesial propagation and secondary Loss of consciousness

generalization
Impaired executive control

Pathological excitation and disinhibition
could be to design tasks which evoke spontaneous behaviour and decisions which
have to be made by the patient, as has been done by Bechara et al. (1998) with the
gambling task, by Goldberg and Podell (2000) with their Cognitive Bias Task, or by
Upton and Thompson (1999) with their Twenty Questions Task.
Ictal behaviour in frontal lobe seizures: positive and negative phenomena
Like others before us, we recently analysed seizure phenomena in patients with FLE
by video-EEG monitoring. The main purpose was to get hints from seizures for
dierential diagnosis. On the other hand, seizures can be studied in terms of tran-
sient dysfunctions which are more or less localized and seizure semiology; pre-
served functions, as well as impaired functions, can tell us something about the
cerebral functional organization of cognition and consciousness.
We studied positive phenomena in terms of seizure semiology, and negative
phenomena in terms of impairment when patients were neuropsychologically
tested during their seizures (Helmstaedter et al., unpublished data; Lux et al., 2000;
Scherrmann and Elger, 1999).
Ictal phenomena in frontal seizures are mostly positive phenomena (see Table
12.2). On the one hand, this means a nearly 1: 1 relationship between discharges and
motor excitation when direct access to motor neurons is possible in primary motor
area seizures, for example. On the other hand, this means release and disinhibition of
172 C. Helmstaedter
Table 12.3. Negative ictal symptoms in focal epilepsy (n 116)
Location of seizure activity
Percentage impaired when Frontal Right temporal Left temporal Bitemporal

tested ictally n29 n21 n38 n28
Orientation reex 62 10 18 57
Receptive speech (commands) 48 15 59 93
Expressive speech 77 11 47 76
Memory 31 0 46 100
Consciousness 33 12 39 100
complex behaviours and behaviour chains when precentral areas are involved.
Examples are posturing and contraversive movements in SMA and premotor seizures,
and explosive, bizarre, and emotional unstable behaviours in prefrontal seizures
including its mesial parts. Negative phenomena like loss of consciousness are com-
monly observed in seizures with mesial propagation and secondary generalization.
For frontal seizures one can thus conclude that the prominent feature is impairment
of executive control in terms of a pathological hyper-excitation or disinhibition.
Neuropsychological examination of the cognitive impairment during seizures
can provide additional insight into the ictal event. We performed ictal testing in 116
patients, most of them being candidates for epilepsy surgery. These patients under-
went ictal examinations which included examination of orientation reexes
(verbal, nonverbal, tactile), expressive/receptive language (commands, naming
repetition), nonverbal reception/expression (commands and imitation) and nally
awareness and memory (interview after the seizure). Testing was performed by the
video-EEG monitoring sta and started as soon as possible after seizure onset.
Functions were tested hierarchically according to their complexity and testing was
continued until the seizure ended. About half of the patients had implanted strip
or depth electrodes for invasive EEG recordings.
Table 12.3 shows the impairment pattern which results when the distribution of
ictal EEG activity at the time of testing is considered. In comparison to lateralized
and bilateral temporal lobe seizures frontal lobe seizures are characterized by
prominent impairment of orientation reexes and expressive speech, which are
typical frontal functions. Receptive speech is often preserved. Patients can try, for
example, to follow body commands even when they appear involved in excessive
motor activity. In contrast to left and bitemporal seizures, consciousness (aware-
ness of any kind) and memory for the test situation during the seizures is mostly
preserved.
Table 12.4. Ictal symptoms in frontal nonconvulsive status epilepticus (n 5)
Performance Impairment
Motor functions (including speech) Generally reduced activity

Rarely, automatisms (fumbling etc. . . .)
Orientation Fluctuating
Executive functions (including language) No self-initiated directed actions
Increased perseverations
Intrusions
Apractic signs in object use and imitation
Reasoning Problems with concept formation and shift
(colour/form . . .)
Working memory Only impaired when complex mental
information processing is required
Emotion Emotional instability (dysphoric)
Impaired executive control

Pathological inhibition
A very interesting behaviour and neuropsychological pattern of impairment can

be observed in patients with frontal nonconvulsive status epilepticus. It is impor-
tant to note that in contrast to a generalized tonic-clonic status, which is the repe-
tition of the same seizure, the nature of frontal nonconvulsive status and frontal
seizures is completely dierent. In contrast to frontal lobe seizures, seizure semiol-
ogy of nonconvulsive status is dominated by negative seizure phenomena. Without
EEG recording the epileptic nature of this state is easily overlooked, and patients
appear rather strange, since they are slowed, dysphoric, morose and dicult. When
neuropsychologically examined during the seizure we found in ve cases consis-
tently generally reduced activity, uctuating orientation, reexive and no self-ini-
tiated behaviour, perseverations, intrusions, apractic signs, problems to shift
between tasks, impaired working memory on a higher cognitive level and emo-
tional instability (see Table 12.4). In 1997 we described a single patient with a non-
convulsive status epilepticus who showed a generalized EEG pattern but focal
cognitive decits when neuropsychologically tested during this state. Today, with
better diagnostic tools, we would probably be able to reinterpret this case as frontal
nonconvulsive status (Bauer et al., 1997).
Contrasting frontal seizures to frontal nonconvulsive status, one could interpret
the latter in terms of an impaired executive control by pathological hyper-
inhibition. Impressive recovery to normal behaviour can be observed in these
174 C. Helmstaedter
1.6
1.4
1.2
0.8
0.6
0.4
0.2
Figure 12.1. Preictal baseline measures and postictal course of verbal memory and decision times in
patients with frontal and left or right temporal lobe epilepsy. The bars indicate
performance of healthy subjects when tested repeatedly in the same intervals.
patients when the status is successfully cessated by injection of diazepam. This is

one form of state-dependent cognitive impairment, as discussed in more detail in
Chapter 6 by Besag.
Postictally, patients, since they often do not lose consciousness during frontal
lobe seizures, are quickly reoriented. Figure 12.1 shows the course of verbal
memory and decision times in pre- and postictal memory testing after frontal lobe
seizures as compared with left/ right temporal seizures and repeated testing in
healthy controls. After lateralized temporal lobe seizures, material-specic memory
impairment can be observed for at least one hour after complete reorientation.
What is shown for left temporal patients in verbal memory in Figure 12.1 has its
counterpart for right temporal patients in gural memory. As for frontal lobe sei-
zures, it is remarkable that there is no postictal deterioration in memory nor sig-
nicant slowing down of reaction times. However, when seizures secondarily
generalize, lasting memory impairment can be observed following frontal seizures
(Helmstaedter et al., 1994).
We can conclude so far that from frontal lobe seizures a dysexecutive syndrome
results, with mostly preserved awareness and consciousness, only reexive but no
self-initiated behaviour, and a seizure semiology, which is dominated by a state of
hyperexcitation and disinhibition or hyperinhibition. This would conrm the
impression from neuropsychological ndings that the major problem in FLE is
appropriate response selection/initiation and inhibition of behaviour. A further

dierentiation according to lesions or foci within particular sites of the frontal
lobes can be suggested but has not yet been proven. From a neuropsychological
point of view it is still dicult to decide whether one central executive function or
dierent executive functions should be assumed. As already mentioned a compro-
mise is favoured at the moment, which suggests that the frontal subfunctions are
constituted by similar processes of response selection/initiation and inhibition in
dierent domains and modalities of behaviour.
Behavioural correlates of frontal lobe epilepsy
If we propose problems with behaviour selection/initiation and inhibition as a

functional complex which is mainly aected in frontal lobe epilepsy, the obvious
question is whether or not this dysfunction has a correlate in personality and
behaviour.
With respect to this question we applied several self-rating scales to a group of
95 patients with either frontal (n18) or mesial temporal lobe epilepsy (n77).
Epilepsy groups were matched regarding sex, the age at the onset of epilepsy (mean
11 years) and the duration of epilepsy (mean 24 years). The tests in use were the
BPSE activity subscale assessing frequencies of activities (Helmstaedter and Elger,
1994); depression and anxiety were assessed by the Beck Depression Inventory
(BDI; Beck et al., 1981) and the Zung Self Rating Anxiety Scale/SAS (Zung, 1971);
personality was assessed by the Neo Five Factor Inventory, a German version of the
NEO personality inventory (Costa and McCrea, 1992). Quality of life in epilepsy
was assessed by a German modied QOLIE-10 (English version: Cramer et al.,
1996); and nally we evaluated education and employment in order to add some
objective data.
Group comparisons considering localization and lateralization of epilepsy
revealed only slight dierences (Table 12.5). Patients with mTLE as a trend showed
poorer mood and signicantly increased anxiety scores, they described themselves
as more active at home, less active with respect to outdoor cultural activities, and
less open for experiences than patients with FLE. It is important to note that, when
compared to normative data of healthy control subjects, the result regarding
outdoor cultural activities must be interpreted in a way that patients with FLE are
more active than the controls and patients with mTLE. Furthermore, when com-
pared to data of a healthy control group, the neuroticism score of patients with
mTLE and the conscientiousness score of patients with FLE appeared elevated.
As regards quality of life (QOL), patients were categorized as having poor QOL
when they showed scores below the 25th percentile. As is shown in Figure 12.2
patients with TLE generally tended to report poorer QOL than patients with FLE.
176 C. Helmstaedter
Table 12.5. Group comparisons considering localization and lateralization of epilepsy
Scale Group Mean SD Signicance
Mood (BDI SAS)

Depression FLE 7.6 7.3 n.s.
mTLE 11.1 9.0
a
Anxiety FLE 29.7 7.9
mTLE 35.9 7.4
Activities (BPSE: activity subscale)

a
Home activities FLE 25.3 4.9
mTLE 27.8 5.7
Social activities FLE 20.2 5.3 n.s.
mTLE 18.7 6.1
a
Cultural activities FLE 16.3 6.5
mTLE 12.8 5.4
Personality (NEO FFI)

Neuroticism FLE 21.4 5.4 n.s.
mTLE 24.7 7.4
Extraversion FLE 26.2 4.3 n.s.
mTLE 26.0 6.2
a
Open to experiences FLE 28.6 6.6
mTLE 25.1 5.3
Agreeableness FLE 31.6 4.6 n.s.
mTLE 30.2 4.2
Conscientiousness FLE 34.9 5.6 n.s.
mTLE 33.2 5.3
Notes:
FLE, frontal lobe epilepsy; mTLE, mesial temporal lobe epilepsy.
For decriptions of scales used, see text.
a
Signicantly dierent; n.s., not signicantly dierent.
Impaired mood, memory problems and social limitations correspond well to the
features of TLE found with the other instruments in this evaluation.
Our current approach to behavioural problems and personality in patients with
focal epilepsies is less led by classication systems, which may be useful in idio-
pathic psychiatric disorders. As already mentioned in the introduction, there is a
long history of personality research in epilepsy and up to now no consistent fea-
tures have been discerned. So far this has been explained by the multifactorial
determination of psychiatric problems in patients with symptomatic epilepsies. As
Figure 12.2. Quality of Life (QOL) in frontal lobe epilepsy (FLE) as compared with mesial temporal
lobe epilepsy (mTLE). Values 25th percentile were considered as reflecting perception
of impaired QOL. Asterisks indicate significant group differences in Chi-square testing.
AED, antiepileptic drug.
far as psychometric approaches are concerned previous studies of temporal lobe

epilepsy mostly used the MMPI (Rose et al., 1996) or more specically the
BearFedio Inventory (Bear et al., 1982; Devinsky and Najjar, 1999). It is our daily
experience that commonly used psychiatric scales or psychological personality
inventories largely fail to reect objectively what seems to the examiner clearly to
be an epilepsy-related change in personality or a behaviour disorder.
At the moment we are evaluating our own clinical personality inventory, which
was empirically designed according to a collection of behavioural problems per-
ceived by the clinical psychological sta in the University Hospital of Epileptology
in Bonn, Germany (Helmstaedter et al., 2000a). For preliminary analysis the ques-
tionnaire was consecutively applied to 59 patients with TLE, 17 patients with FLE,
9 patients with parieto-occipital epilepsy and 44 healthy controls. It consists of 82
questions concerning 15 dierent behavioural domains. The answer style is a six-
stepped frequency of occurrence rating with 1occurs not at all and 6occurs
very frequently. Second order factor analysis resulted in six factors, which were
interpreted as follows: 1. organic personality change with patients reporting com-
munication problems, emotional lability, being indecisive, susceptible to interfer-
ence, perseverative and hypoactive; 2. depressed mood including depressive
mood, reduced vitality, anxiety, and insensitivity; 3. addiction and obsession
including addiction to legal and illegal substances, compulsion, and obsession;
178 C. Helmstaedter
Impairment Lateralization Localization Sex Pathology

Scale
< 1 SD 37 right/48 left 59 T/17 F/ 9 P 59/70 28 with AHS
1. Organic personality change 32%** LEFT* TLE/FLE* W>M /
2. Addiction/obsession 20%* RIGHT* FLE* / /
3. Depressed mood 28% / / / AHS (right)*
4. Extraversion 32%** LEFT/RIGHT* PLE* / /
5. Aggression 31%** LEFT* / / /
6. Adaptivity/hyperactivity 29%* / FLE* / /
*P < 0.05, **P < 0.01 (significantly different from control subjects)
Figure 12.3. Results obtained with the clinical personality questionnaire. TLE, temporal lobe epilepsy;
FLE, frontal lobe epilepsy; PLE, parieto-occipital epilepsy.
4. extraversion comprised sociability, curiosity and self-determined behaviour;

5. aggression comprised aggression, sensation-seeking, nonadaptive behaviour
and violence; 6. hyperactivity and adaptivity. When taking clinical data as well as
sex as independent variables some interesting and comprehensible results could be
obtained (Figure 12.3).
The data rst of all indicate that problems in the respective areas are evident in
2030 % of patients. Organic personality changes are preferentially seen in left epi-
lepsies of either origin, in women more than in men. Addiction and obsession are
more frequent in right epilepsies and in frontal epilepsies in particular. Depressed
mood is preferentially seen in patients with hippocampal sclerosis, a nding which
is in line with one of our recent publications (Quiske et al., 2000). All patients and
patients with parietal epilepsies in particular show reduced extraversion. Aggressive
behaviour seems more frequent in left epilepsies, and patients with FLE show
increased hyperactivity and adaptivity, which may parallel the nding of increased
outdoor/cultural activities and openness for experiences. It is important to note
that these results are preliminary and that larger control groups and validation
studies are still required. However, the data indicate that the often-cited depressive
mood is not the only behavioural problem in patients with focal epilepsy, and that
apart from this there are specic behavioural aspects which appear related to local-
ized and lateralized lesions or epileptic dysfunctions.
Although no dierences between patients with FLE and TLE were obtained, it is
worth reporting the results with regard to the organic personality change scale in
more detail. As shown in Figure 12.4 for selected items, about 20% of the patients
report that they oend others; between 20 and 35% of the patients report problems
Organic personality change
Offends others 18%
Receptive problems 23%
Perseverative 26%
Misunderstandings 33%
Circumstantial 34%
Irritates others 51%
0% 10% 20% 30%
% deviant ratings > 75% percentile as compared

with healthy controls
Figure 12.4. Items extracted out of the organic personality change scale of the clinical personality
inventory. Bars represent the per cent of patients with focal epilepsies reporting increased
problems in communication and interpersonal contact.
with reception, misunderstandings, that they were perceived as perseverative or

circumstantial; and 50% report that their behaviour irritates others. This is similar
to the epileptic personality, and taken together with the depressed mood, one
might well think of the dysphoric and paroxysmal mood disorder as it has been
proposed from a more psychiatric point of view (Blumer, 2000).
Academic achievement and employment in frontal lobe epilepsy
From patients with frontal lesions, it is well known that they may show unimpaired
cognitive functions but nevertheless fail on everyday demands of job and career
because of behavioural problems, unsteadiness, concentration problems, increased
susceptibility to interference and problems with timing and planning. Subjective
data may not detect behavioural problems because patients with frontal lobe
lesions have been reported to underestimate their impairments. With school
achievement and employment, however, we have indirect markers, which allow us
to infer how far patients are adapted to everyday life. As indicated in Table 12.6 it
is not the group with FLE but the one with mTLE which is less educated, and the
job situation is comparable in both groups.
180 C. Helmstaedter
Table 12.6. Academic achievement, employment and epilepsy
Academic achievement level %
No regular Lowa Medium Higha Percentage

school (Hauptschule) (Realschule) (Gymnasium) employed
FLE (n18) 17 22 22 39 68
mTLE (n83) 10 54 21 15 59
Note:
a
Chi-square: signicant dierence.
Trait or state
As shown above, patients with FLE have behavioural disorders which appear very
mild when compared with those reported in patients with frontal mass lesions. With
respect to mood disorders, they appear less aected than patients with TLE, and they
also show better academic achievement. The nding that hyperactivity, addiction and
obsession might be a behavioural feature of FLE is of great interest, and can be dis-
cussed as reecting frontal dysfunction in general and to be in line with the behavi-
our observed in neuropsychological examination and during seizures. The question
which remains open, is how consistent the behaviour in focal epilepsies is over time.
We can not yet give a conclusive answer to this question on the basis of long-term
follow-up observations. The impact of epilepsy and seizures on behaviour,
however, can be estimated by comparisons of patients who after surgery still have
seizures with those who became completely seizure-free. We therefore analysed
data from surgically and nonsurgically treated patients who participated in a long-
term follow-up study, which was originally designed to show the cognitive devel-
opment of these patients over time (Helmstaedter et al., 2000b). However, at the
time of the long-term follow-up visit we also assessed depression by use of the BDI
and quality of life by use of a German modied QOLIE10. For the present purpose
we extracted from the total database only the data of the patients with temporo-
mesial epilepsy and hippocampal sclerosis as compared with those with FLE. Fifty-
seven patients had mTLE with hippocampal sclerosis (27 had surgery, 20 were
treated conservatively) and 30 patients had FLE (16 had surgery and 14 were treated
conservatively). Taking depression and quality of life measures as the dependent
variables in a multivariate analysis with consideration of surgery, localization and
lateralization of epilepsy as independent variables, and age and the follow-up inter-
val (mean 56 months; 210 years) as covariates, seizure outcome turned out to be
the only signicant predictor. Only 14% of the seizure-free patients in contrast to
51% of those who still had seizures showed elevated depression scores greater than
the cut-o score of 12 points. It should be noted that 14% is much less than the
normally reported 30% of patients with focal epilepsy and depressive mood, and
that 51% clearly exceed this number. Comparably, 45% of the seizure-free patients
reported good quality of life with QOLIE10, as compared to only 11% of the
patients who continued to have seizures. Although these data are not follow-up
data and although depression and quality of life represent only two facets of the
whole range of behaviour, these data show quite impressively what a dierence the
presence or absence of seizures can make. The nding parallels recent ndings in
children who after successful epilepsy surgery show marked improvement in
behaviour disorders (Lendt et al., 2000). Long-term follow-up studies on person-
ality and behaviour disorders are thus badly needed to complete our understand-
ing of the interaction between brain damage, epilepsy and behaviour.
Conclusion
We can conclude that in FLE, frontal dysfunctions can be suggested which char-
acteristically become evident in cognition, seizures and behaviour. The main
common feature of the behavioural problems met in FLE concerns behaviour
control in terms of response selection/initiation and inhibition. The domains in
which these problems become apparent may vary with clinical features. Following
our own ndings, hyperactivity, conscientiousness, obsession and addiction can be
seen as reecting frontal lobe dysfunction in frontal lobe epilepsy. Depression,
anxiety, neuroticism, cognitive (memory) impairment and social limitations in
contrast seem more a feature of mTLE. However, methodological diculties
regarding the adequacy of the clinical measures in use as well as confounding eects
of lesions, epileptic dysfunction, antiepileptic drugs and psychosocial status do not
yet allow further distinctions such as can be made, for example, in neurobiological
models about the frontal lobes and behaviour. Full-blown personality disorders are
very rare in FLE and symptoms appear rather mild when compared with patients
with mass lesions. As regards the state/trait discussion in epilepsy, the eects of
seizure control indicate that a great part of the observed behavioural problems is
indeed state-dependent. However, follow-up evaluations are badly needed to
partial out the contribution of lesions and epileptic dysfunctions to behaviour dis-
orders and to demonstrate how far these are reversible or not.
R E F E R E N C ES
Alivisatos, B. and Milner, B. (1989). Eects of frontal or temporal lobectomy on the use of
advance in information in a choice reaction time task. Neuropsychologia, 27, 495503.
182 C. Helmstaedter
Anderson, S.W., Bechara, A., Damasio, H. et al. (1999). Impairment of social and moral behav-
ior related to early damage in human prefrontal cortex. Nat Neurosci, 2, 10327.
Azouvi, P., Jokic, C., Attal, N., Denys, P., Markabi, S. and Bussel, B. (1999). Carbamazepine in
agitation and aggressive behavior following severe closed head injury: results of an open trial.
Brain Injury, 13, 797804.
Baddeley, A.D. and Hitch, G. (1974). Working memory. In Recent Advances in Motivation and
Learning, Vol. 8, ed. G.A. Bower, pp. 4790. New York: Academic Press.
Bauer, J., Helmstaedter, C. and Elger, C.E. (1997). Nonconvulsive status epilepticus with gener-
alized fast activity. Seizure, 6, 6770.
Bear, D., Levin, K., Blumer, D., Chetham, D. and Ryder, J. (1982). Interictal behaviour in hospi-
talized temporal lobe epileptics: relationship to idiopathic psychiatric syndromes. J Neurol
Neurosurg Psychiatry, 45, 4818.
Bearn, R.G. and Gibson, R.J. (1998). Aggressive behavior in intellectually challenged patients
with epilepsy treated with lamotrigine. Epilepsia, 39, 2802.
Bechara, A., Damasio, H., Damasio, A.R. and Lee, G.P. (1999). Dierent contribution of the
human amygdala and ventromedial prefrontal cortex to decision making. J Neurosci, 19,
547381.
Bechara, A., Damasio, H. and Damasio, A.R. (2000). Emotion, decision making and the orbito-
frontal cortex. Cereb Cortex, 10, 295307.
Bechara, A., Damasio, H., Tranel, D. and Anderson, S.W. (1998). Dissociation of working
memory from decision making within the human prefrontal cortex. J Neurosci, 18, 42837.
Beck, A.T., Rush, A.J., Shaw, B.F. and Emery, G. (1981). Kognitive Therapie der Depression.
Mnchen: Urban und Schwarzenberg.
Blumer, D. (2000). Dysphoric disorders and paroxysmal aects: recognition and treatment of
epilepsy related disorders. Harv Rev Psychiatry, 8, 817.
Boone, K.B., Miller, B.L., Rosenberg, L., Durazo, A., McIntyre, H. and Weil, M. (1988).
Neuropsychological and behavioural abnormalities in an adolescent with frontal lobe seizures.
Cavada, C., Compay, T., Tejedor, J., Cruz-Rizzolo, R.J. and Reinoso-Surez, F. (2000). The ana-
tomical connections of the macaque monkey orbitofrontal cortex. A review. Cereb Cortex, 10,
22042.
Costa, P.T. and McCrea, R.R. (1992). Revised NEO Personality Inventory (NEO PI-R) and the NEO
Five Factor Inventory. Professional Manual. Odessa, FL: Psychological Assessment Resources.
Cramer, J.A., Perrine, K., Devinsky, O. and Meador, K. (1996). A brief questionnaire to screen
quality of life in epilepsy the QOLIE10. Epilepsia, 37, 57782.
Damasio, A.R. (1996). The somatic marker hypothesis and the possible functions of the prefron-
tal cortex. Philos Trans R Soc Lond B Biol Sci, 351, 141320.
Damasio, H., Grabowski, T., Frank, R., Galaburda, A.M. and Damasio, A.R. (1994). The return
of Phineas Gage: clues about the brain from the skull of a famous patient. Science, 20/264,
11025.
Davidson, R.J., Putnam, K.M. and Larson, C.L. (2000). Dysfunction in the neural circuitry of
emotion regulation a possible prelude to violence. Science, 289, 5914.
Delaney, R.C., Rosen, A.J., Mattson, R.H. and Novelly, R.A. (1980). Memory function in focal
epilepsy: a comparison of non-surgical, unilateral temporal lobe and frontal lobe samples.
Cortex, 16, 10317.
Devinsky, O. and Najjar, S. (1999). Evidence against the existence of a temporal lobe syndrome.
Neurology, 53, S1325.
Devinsky, O., Morrell, M.J. and Vogt, B.A. (1995). Contributions of anterior cingulate cortex to
behavior. Brain, 118, 279306.
Dolan, R. (1999). On the neurology of morals. Nature Neurosci, 2, 9279.
Drevets, W.C. and Raichle, E. (1996). Positron emission tomographic imaging studies in human
emotional disorders. In The Cognitive Neurosciences, ed. M.S. Gazzaniga, pp. 115362.
Massachusetts: MIT Press.
Elger, C.E. and Lehnertz, K. (1998). Seizure prediction by non-linear time series analysis of brain
electrical activity. Eur J Neurosci, 10, 7869.
Goldberg, E. and Podell, K. (2000). Adaptive decision making, ecological validity, and the frontal
lobes. J Clin Exp Neuropsychol, 22, 5668.
Harlow, J.M. (1868). Recovery from passage of an iron bar through the head. Mass Med Soc Publ,
2, 32746.
Helmstaedter, C. and Elger, C.E. (1994). Cognitive-behavioral aspects of quality of life in presur-
gical patients with epilepsy. J Epilepsy, 7, 22031.
Helmstaedter, C., Elger, C.E. and Lendt, M. (1994). Postictal courses of cognitive decits in focal
epilepsies. Epilepsia, 35, 10738.
Helmstaedter, C., Kemper, B. and Elger, C.E. (1996). Neuropsychological aspects of frontal lobe
epilepsy. Neuropsychologia, 34, 399406.
Helmstaedter, C., Gleiner, U., Zentner, J. et al. (1998). Neuropsychological consequences of epi-
lepsy surgery in frontal lobe epilepsy. Neuropsychologia, 36, 33341.
Helmstaedter, C., Gleiner, U. and Elger, C.E. (2000a). Clinical Personality Scales (CPS) in focal
epilepsy: preliminary results. Epilepsia, 42 (Suppl.), 236.
Helmstaedter, C., Kurthen, M., Lux, S. et al. (2000b). Long-term clinical, neuropsychological, and
psychosocial follow-up in surgically and nonsurgically treated patients with drug-resistant
temporal lobe epilepsy. Nervenarzt, 71, 62942.
Hermann, B.P., Wyler, A.R. and Richey, E.T. (1988). Wisconsin card sorting test performance in
patients with complex partial seizures of temporal lobe origin. J Clin Exp Neuropsychol, 10,
46776.
Jokeit, H., Seitz, R.J., Markowitsch, H.J., Neumann, N., Witte, O.W. and Ebner, A. (1997).
Prefrontal asymmetric interictal glucose hypometabolism and cognitive impairment in
patients with temporal lobe epilepsy. Brain, 120, 228394.
Kalynchuk, L.E., Pinel, J.P. and Treit, D. (1999). Characterization of the defensive nature of -
kindling-induced emotionality. Behav Neurosci, 113, 76675.
Kemper, B., Helmstaedter, C. and Elger, C.E. (1992). Kognitive Prole von prchirurgischen
Patienten mit Frontal- und Temporallappenepilepsie. In Epilepsie 91, ed. D. Schener, pp.
34550. Reinbeck: Einhorn Presse Verlag.
Lendt, M., Helmstaedter, C., Kuczaty, S., Schramm, J. and Elger, C.E. (2000). Behavioural disor-
ders in children with epilepsy: early improvement after surgery. J Neurol Neurosurg Psychiatry,
69, 73944.
184 C. Helmstaedter
London, E.D., Ernst, M., Grant, S., Bonson, K. and Weinstein, A. (2000). Orbitofrontal cortex
and human drug abuse: functional imaging. Cereb Cortex, 10, 33442.
Lux, S., Helmstaedter, C., Kurthen, M., Hartje, W. and Elger, C.E. (2000). Localizing value of ictal
neuropsychological decits in focal epilepsies. Epilepsia, 41 (Suppl. 7), 153.
Marsh, L. and Krauss, G.L. (2000). Aggression and violence in patients with epilepsy. Epilepsy
Behav, 1, 1608.
Milner, B. (1964). Some eects of frontal lobectomy in man. In The Frontal Granular Cortex, ed.
J.M. Warren and K. Akert, pp. 31334. New York: McGraw-Hill.
Milner, B. (1995). Aspects of human frontal lobe function. In Advances in Neurology, Vol. 66, ed.
H.H. Jasper, S. Riggio and P.S. Goldman-Rakic, pp. 6784. New York: Raven Press.
Northo, G., Richter, A., Gessner, M. et al. (2000). Functional dissociation between medial and
lateral prefrontal cortical spatiotemporal activation in negative and positive emotions: a com-
bined fMRI/MEG study. Cereb Cortex, 10, 93107.
Petrides, M. (1985). Decits on conditional associative learning tasks after frontal- and temporal
lobe lesions in man. Neuropsychologia, 5, 60114.
Petrides, M. and Milner, B. (1982). Decits on subject-ordered tasks after frontal and temporal
lobe lesions in man. Neuropsychologia, 3, 24962.
Quiske, A., Helmstaedter, C., Lux, S. and Elger, C.E. (2000). Depression in patients with temporal
lobe epilepsy is related to mesial temporal sclerosis. Epilepsy Res, 39, 1215.
Raine, A., Lencz, T., Bihrle, S., LaCasse, L. and Coletti, P. (2000). Reduced prefrontal gray matter
volume and reduced autonomic activity in antisocial personality disorder. Arch Gen Psychiatry,
67, 11927.
Roberts, A.C. and Wallis, J.D. (2000). Inhibitory control and aective processing in the prefron-
tal cortex: neuropsychological studies in the common marmoset. Cereb Cortex, 10, 25262.
Rolls, E.T. (2000). The orbitofrontal cortex and reward. Cereb Cortex, 10, 28494.
Rose, K.J., Derry, P.A. and McLachlan, R.S. (1996). Neuroticism in temporal lobe epilepsy: assess-
ment and implications for pre- and postoperative psychosocial adjustment. Epilepsia, 37,
48491.
Rubia, K., Overmeyer, S., Taylor, E. et al. (2000). Functional frontalization with age: mapping
neurodevelopmental trajectories with fMRI. Neurosci Biobehav Rev, 24, 1319.
Sarazin, M., Pillon, B., Giannakopoulos, P., Rancurel, G., Samson, V. and Dubois, B. (1998).
Clinicometabolic dissociation of cognitive functions and social behavior in frontal lobe
lesions. Neurology, 51, 1428.
Scherrmann, J.M. and Elger, C.E. (1999). Generation of seizure phenomena in the frontal and
temporal lobe. Epilepsia, 40 (Suppl. 7), 102.
Schmitz, B. (1999). Psychiatric syndromes related to antiepileptic drugs. Epilepsia, 40 (Suppl. 10),
6570.
Shulman, M.B. (1984). The frontal lobes, epilepsy, and behavior. Epilepsy Behav, 1, 38495.
Smith, M.L. and Milner, B. (1984). Dierential eects of frontal-lobe lesions on cognitive esti-
mation and spatial memory. Neuropsychologia, 22, 697705.
Stuss, T.D. and Benson, T. (1986). The Frontal Lobes. New York: Raven Press.
Trimble, M.R. and Van Elst, L.T. (1999). On some clinical implications of the ventral striatum
and the extended amygdala. Investigations of aggression. Ann NY Acad Sci USA, 29, 63844.
Upton, D. and Thompson, P.J. (1996a). General neuropsychological characteristics of frontal

lobe epilepsy. Epilepsy Res, 23, 16977.
Upton, D. and Thompson, P.J. (1996b). Epilepsy and the frontal lobes: neuropsychological char-
acteristics. J Epilepsy, 9, 21522.
Upton, D. and Thompson, P.J. (1997a). Neuropsychological test performance in frontal-lobe epi-
lepsy: the inuence of aetiology, seizure type, seizure frequency and duration of disorder.
Seizure, 6, 4437.
Upton, D. and Thompson, P.J. (1997b). Age at the onset and neuropsychological function in
frontal lobe epilepsy. Epilepsia, 38, 110313.
Upton, D. and Thompson, P.J. (1999). Twenty question tests and frontal lobe dysfunction. Arch
Clin Neuropsychol, 14, 20316.
Van Elst, L.T., Woermann, F.G., Lemieux, L., Thompson, P.J. and Trimble, M.R. (2000). Aective
aggression in patients with temporal lobe epilepsy: a quantitative MRI study of the amygdala.
Brain, 123, 23443.
Zung, W.W.K. (1971). A rating instrument for anxiety disorders. Psychosomatics, 12, 3719.
Part IV
Nonepileptic attacks
13
Epilepsy, dissociation and nonepileptic

seizures
Richard J. Brown
Introduction
Within the context of psychiatry and neurology, the term dissociation is extremely
dicult to dene satisfactorily. Since its introduction in the late nineteenth century,
the term has been applied to a wide range of neurological, psychiatric and psycho-
logical phenomena. As a result, there is considerable confusion over what actually
constitutes dissociation and the concept is frequently misapplied. This is particularly
true within the eld of epilepsy. Many epileptic phenomena have been labelled as dis-
sociative, including the sensory, aective and cognitive features of partial seizures,
behavioural automatisms, postictal amnesia and fugue (Thomas and Trimble, 1997).
Similarly, certain psychiatric phenomena that mimic epileptic events, so-called non-
epileptic seizures, have been identied as primarily dissociative in nature. Indeed,
many authorities have argued that the dissociative nature of nonepileptic seizures
could provide the basis for their conceptual and practical dierentiation from
genuine epileptic events (Kuyk et al., 1999). If, however, dissociation is experienced
both by individuals with epilepsy and by those with pseudo-epileptic seizures, how
can its occurrence aid in the dierential diagnosis of these conditions?
In this chapter, I will explore what we mean by dissociation and how it relates to
the phenomena of epilepsy and nonepileptic seizures. In this way, I aim to demon-
strate that (i) many epileptic phenomena previously labelled as dissociative should
not be regarded as dissociative at all; (ii) dissociation is a fundamental aspect of
nonepileptic seizures; (iii) those epileptic phenomena that are genuinely dissocia-
tive involve a dierent type of dissociation to that involved in nonepileptic seizures;
and (iv) the concept of dissociation needs to be used far more precisely if it is to
help distinguish between epileptic and nonepileptic seizures.
What is dissociation?
Cardea (1994) has described a useful taxonomy that captures the dierent ways in
which the concept of dissociation has been used. According to this scheme, there
189
190 R.J. Brown
Table 13.1. Classification of dissociative disorders in ICD10 and DSMIV
ICD10 dissociative (conversion) disorders DSMIV dissociative disorders
Dissociative amnesia Dissociative amnesia

Dissociative fugue Dissociative fugue
Dissociative motor disorders Dissociative identity disorder
Dissociative convulsions Depersonalization disorder
Dissociative anaesthesia and sensory loss Dissociative disorder not otherwise specied
Dissociative stupor
Trance and possession disorders
Mixed dissociative (conversion) disorders
Other dissociative (conversion) disorders
Dissociative (conversion) disorder, unspecied
are three major facets to the dissociation construct: (i) dissociation as noncon-
scious or nonintegrated mental modules or systems; (ii) dissociation as an altera-
tion in consciousness; and (iii) dissociation as a defence mechanism.
Dissociation as nonintegrated mental modules or systems

This is probably the most widely used denition of dissociation and is reected in
the so-called dissociative disorders categories of current psychiatric taxonomies.
According to DSMIV, the essential feature of the dissociative disorders is a dis-
ruption in the usually integrated functions of consciousness, memory, identity, or
perception of the environment (p. 477; American Psychiatric Association, 1994).
Dissociation in this sense reects the original meaning of the term dsagrgation,
oered by Janet as a label for the putative process whereby previously integrated
memories can become inaccessible to consciousness and control behaviour in a way
that the individual is unaware of (Janet, 1889, 1924). According to Janet, such a
process is the basic psychopathological mechanism underlying so-called hysteri-
cal symptoms, phenomena that have now been reclassied within the dissociative
and somatoform disorders categories in DSMIV and ICD10. The DSMIV dis-
sociative disorders category encompasses dissociative amnesia, dissociative fugue,
dissociative identity disorder (formerly multiple personality disorder), deperson-
alization disorder and dissociative disorder not otherwise specied (see Table 13.1).
A slightly broader denition is oered in ICD10, which identies the loss of
control over bodily movements as an additional dissociative phenomenon (World
Health Organization, 1992). As such, the ICD10 dissociative (conversion) disor-
ders category encompasses dissociative amnesia, dissociative fugue, dissociative
motor disorders, dissociative convulsions, dissociative anaesthesia and sensory
191 Seizures, epilepsy and dissociation
loss, dissociative stupor, mixed dissociative (conversion) disorders, other dissocia-

tive (conversion) disorders and dissociative (conversion) disorders, unspecied
(see Table 13.1). The somatoform and dissociative disorders are related in that these
phenomena are characterized by symptoms that, on the face of it, resemble those
that occur in certain physical conditions, but which are presumed to be psycholog-
ical in origin. Dissociative and somatoform phenomena dier in that the symptoms
of the former resemble those of neurological illness, while the symptoms of the
latter are more akin to those encountered in internal medicine. As Kihlstrom
(1994) has cogently argued, all of the phenomena identied as dissociative disor-
ders within DSMIV and ICD10 are linked by the fact that each has a temporary
disruption in consciousness or volition as its primary dening feature.
The dierences between DSMIV and ICD10 in their classication of the dis-
sociative and somatoform disorders are readily apparent. First, unlike ICD10,
DSMIV places nonepileptic attacks in the somatoform rather than the dissocia-
tive disorders category, along with other so-called conversion phenomena, such as
unexplained motor and sensory symptoms, that are identied as dissociative in
ICD10. This dierence is more practical than conceptual, with DSMIV placing
greater emphasis on the importance of excluding physical illness in the dierential
diagnosis of these phenomena (American Psychiatric Association, 1994). Second,
unlike DSMIV, ICD10 does not identify depersonalization as a dissociative phe-
nomenon, due to the lack of any signicant loss of control over sensation, memory
or movement in this condition, and its limited eect on personal identity. Third,
DSMIV identies a distinct category for multiple personality disorder, relabelled
dissociative identity disorder in the latest edition of this scheme. In contrast,
ICD10 places multiple personality disorder in the other dissociative (conversion)
disorders category, reecting controversy over whether this syndrome is iatrogenic
or culturally bound to North America. Inconsistencies aside, both DSMIV and
ICD10 explicitly state that physical conditions such as epilepsy should be excluded
in the dierential diagnosis of the dissociative and somatoform disorders.
In addition to the dissociative and somatoform disorders, Cardea (1994) iden-
ties other pathological phenomena characterized by a lack of integration between
mental modules or systems that are caused by neurological rather than psychiatric
events. Blindsight, a rare condition in which the suerer displays above-chance
visual discrimination despite reporting a lack of visual experience, provides one
example of how normally integrated functions can become dissociated through
neurological damage. Many of the unusual behaviours often displayed by patients
following commissurotomy also fall within this category, as do those exhibited by
individuals suering from hemi-neglect. In each of these cases, the dissociation is
between the individuals ongoing behaviour and their introspective verbal report.
Neurological phenomena of this sort are important because they demonstrate the
192 R.J. Brown
distributed, semimodular (Fodor, 1983; Kosslyn, 1994) organization of brain,

mind, consciousness and behaviour.
Neurological dissociations such as blindsight are supercially analogous to those
observed in psychiatric instances of dissociation, such as the preservation of
implicit perception1 in the context of dissociative blindness (Kihlstrom, 1994).
Neurological and psychiatric dissociation dier, however, in that the former is often
permanent, reecting irreversible damage to the underlying neurological sub-
systems in question (Kihlstrom, 1994). Psychiatric instances of dissociation, in
contrast, are thought to be the product of an alteration in the parameters govern-
ing otherwise intact psychological functions; they are, therefore, reversible by de-
nition. Relatedly, neurological and psychiatric dissociations dier in that, unlike
the former, the latter involves symptoms (e.g. glove anaesthesia) that need not, and
typically do not, relate to the actual organization of the nervous system and its
many distributed components. On these grounds, it is apparent the dissociation
in these cases is an entirely dierent phenomenon and the two must not be con-
fused.
The idea that normally integrated psychological processes can become tempor-
arily dissociated and exist in isolation of one another has also been cited as the basis
for other, less pathological, phenomena (Cardea, 1994; Hilgard, 1977; Woody and
Bowers, 1994). Many apparently hypnotic phenomena fall within this category,
including profound amnesia, the loss of perceptual experience and complex beha-
viours characterized by a sense of involuntariness, all of which can be temporarily
produced by appropriate suggestions in certain individuals. The extent to which
similar processes are involved in these phenomena and those displayed by individ-
uals with dissociative psychopathology has been a matter of debate since the time
of Janet. Conceptually, there are good grounds to assume a common mechanism
in hypnotic and dissociative phenomena (Oakley, 1999) and recent functional
imaging evidence provides some support for a link between the two (Halligan et al.,
2000).
According to Cardea (1994), this particular denition of dissociation has also
been inappropriately applied to a number of other normal psychological phenom-
ena. Following Hilgard (1977), the execution of complex behaviours with only
minimal conscious awareness, such as the action of driving a car whilst holding a
conversation, has often been identied as a dissociative phenomenon. As Cardea
has pointed out, however, the dissociation label should not simply be applied to any
behaviour or psychological process that, for whatever reason, occurs without full
1
Implicit perception is evidenced when an external stimulus produces psychological eects despite not being
perceived consciously (Kihlstrom, 1994). The phenomenon of blindsight provides one such example.
Implicit perception is akin to the concept of implicit memory, in which behaviour is inuenced by learnt
information that the individual cannot consciously recall.
awareness. Such a practice ignores the fact that, in many such cases, the individual
can bring the apparently dissociated process into awareness by an act of selective
attention. Other such cases involve dissociation between systems or processes that
one would not normally expect to operate in an integrated fashion. According to
Cardea, mental modules or systems should only be regarded as truly dissociated
from one another if their dissociation is (i) in contrast to a normal state of integra-
tion; and (ii) cannot be overcome by an act of will.
Dissociation as an alteration in consciousness

A second usage of the dissociation concept refers to an altered state of conscious-
ness characterized specically by a disengagement from the self or the environment
(Cardea, 1994). As Cardea has pointed out, this sense of the dissociation concept
should not be applied to everyday phenomena, such as daydreaming and other
states of distraction, where engagement with the environment is less than complete.
Rather, it should be reserved specically for states that are regarded by the experi-
encing individual as qualitatively dierent to their normal state of awareness.
Although a number of dierent phenomena fall within the bounds of this deni-
tion (e.g. trance and possession states), probably the most commonly reported
are depersonalization and derealization. In depersonalization, the individual
experiences a profound feeling of detachment from their thoughts, perceptions,
actions and emotions, often characterized by a sense of numbness or disembodi-
ment. In derealization, the individual experiences an intact sense of self coupled
with a feeling of detachment from the external environment, which often feels
unreal or at a distance. Such feelings are extremely common, frequently occurring
in the context of psychiatric illnesses such as depression and anxiety; they also
occur as a circumscribed problem in their own right, such as in depersonalization
disorder. Although DSMIV identies depersonalization disorder as a dissociative
phenomenon, this condition clearly relates to a dierent sense of dissociation than
that which applies to the other members of this category; this dierence further jus-
ties the separation of depersonalization disorder from the dissociative disorders
category in ICD10. Depersonalization and derealization are also found in certain
drug states (e.g. those produced by marijuana, LSD and ketamine), neurological
conditions such as temporal lobe epilepsy and can occur spontaneously in the
context of stress or fatigue.
Dissociation as a defence mechanism

Finally, dissociation has been described as a defence mechanism that protects the
individual from potentially overwhelming pain or anxiety. In many respects, this
account of dissociation is indistinguishable from the Freudian concept of repres-
sion (Erdelyi, 1985). This sense of the dissociation concept is typically used to
194 R.J. Brown
describe the psychological function served by the creation of a dissociated state of

consciousness, or the dissociation of mental modules or systems from one another
(Cardea, 1994). By this view, exposure to a traumatic situation may trigger the
compartmentalization of memories, which preserves psychological integrity by
preventing the distressing material from entering consciousness after the event.
Alternatively, such traumatic exposure may spontaneously elicit a depersonalized
state that prevents extreme emotion from inhibiting an appropriate behavioural
response. As such, this denition of dissociation may relate to either of the deni-
tions described previously. In both cases, dissociation of this sort could either be
an acute response to an isolated traumatic event or a trait-like characteristic
acquired as a result of repeated exposure to trauma.
Epilepsy and dissociation
Although the dierential diagnosis of ICD10 and DSMIV dissociative disorders

explicitly requires the exclusion of symptoms with an identiable neurological
basis, many of the phenomena associated with epilepsy, particularly temporal lobe
epilepsy, have been regarded as dissociative in nature (Devinsky et al., 1989; Good,
1993). Indeed, ICD10 includes a specic category for dissociative disorders due to
a general medical condition, which encompasses many of the symptoms exhibited
by individuals with epilepsy. The absence of such a category from DSMIV,
however, reects doubt concerning the value of attaching the dissociative label to
these phenomena. In my view, such doubt is well justied in many cases.
The notion that many epileptic phenomena can be regarded as dissociative is
based, to a considerable extent, on the frequent occurrence of amnesia in epilepsy
(Good, 1993). Individuals with complex partial or generalized seizures typically
display profound amnesia for events occurring during the ictus. Moreover, certain
people experience a postictal fugue state characterized by apparently purposeful
behaviour for which they are subsequently amnesic, much like dissociative fugue.
Despite their prima facie resemblance to dissociative phenomena, however, these
events should not be regarded as episodes of dissociation (Good, 1993).
Dissociative amnesia is characterized by an inability to retrieve information that
has been learnt and is present in memory despite its inaccessibility (American
Psychiatric Association, 1994). Amnesia for ictal events, in contrast, reects a dis-
ruption in normal information processing, resulting from uncontrolled neural
activity, that prevents the encoding of new material during the ictus. The ictal
amnesia is not a product of a retrieval failure, therefore, but simply the absence of
memories to retrieve. It is for this reason that this form of amnesia is irreversible,
unlike most cases of dissociative amnesia (American Psychiatric Association,
1994).
Postictal fugue should not be regarded as a dissociative episode for similar

reasons. Unlike dissociative fugue, postictal fugue is characterized by a disruption
in consciousness associated with signicant confusion and an abnormal EEG
(Thomas and Trimble, 1997). The apparently purposeful behaviour displayed in
postictal fugue is not dissociated from ongoing cognitive activity as it is in dissoci-
ative fugue; rather, it occurs in the relative absence of such activity. The inability to
reverse the amnesia associated with postictal fugue serves as an illustration of this
fact. The behavioural automatisms often observed in the context of complex partial
seizures and regarded as a dissociative phenomenon by some (Good, 1993) are
amenable to a similar interpretation. Like the behaviours exhibited during postictal
fugue, ictal automatisms only occur in the context of a disruption in consciousness
and disturbed behavioural control; genuinely dissociated behaviours are note-
worthy because they occur despite an otherwise intact ability to control action2. In
both cases, it is likely that these behaviours result from the uncontrolled activation
of circumscribed motor programs by epileptic discharges in neural sites associated
with behavioural control. The fact that such automatisms are particularly charac-
teristic of seizures originating in the frontal lobes lends support to this view.
Many of the phenomena associated with partial seizures originating in the tem-
poral lobes, such as hallucinations, sensory and cognitive auras, dj vu and dj
veu should also be distinguished from true episodes of dissociation. Although hal-
lucinations and auras have a phenomenology that departs from external reality,
these phenomena involve the paradoxical integration of information within con-
scious awareness. As such, they may be more appropriately regarded as phenom-
ena of association rather than dissociation (Kihlstrom, 1994). The phenomenology
of epileptic hallucinations and auras probably originates in the activation of repre-
sentational structures in the temporal lobes, either directly by seizure activity in
representational networks, or indirectly through seizure-related stimulation of
limbic structures such as the amygdala and anterior cingulate (Bancaud et al.,
1994). Experiences of epileptic dj vu and dj veu are also more associative than
dissociative and may involve a similar neurophysiological process. For example,
stimulation of the amygdala by seizure-related discharges could imbue current per-
ceptions and cognitions with an unwarranted emotional colouring that may be
experienced as a sense of having encountered the situation before (Bancaud et al.,
1994; Sierra and Berrios, 1998).
Certain phenomena associated with temporal lobe epilepsy can, however, be
regarded as genuine examples of dissociation according to the scheme described by
Cardea (1994). Depersonalization and derealization commonly occur in the
2
Dissociated behaviours should be distinguished from normal automatic behaviours because the latter are
generally in accord with system goals and can therefore be considered voluntary (Brown, 1999; see also
Cardea, 1994).
196 R.J. Brown
context of temporal lobe epilepsy and involve an alteration in consciousness char-

acterized by dissociation from the self and/or the environment. According to Sierra
and Berrios (1998), depersonalization and derealization are the products of a ves-
tigial defence mechanism evolved to provide the optimum processing conditions
for adaptive behaviour in the face of threat. By this view, extreme anxiety triggers
an inhibitory response from the left prefrontal cortex that dampens output from
the sympathetic nervous system, through inhibition of the amygdala and anterior
cingulate. In turn, the right prefrontal cortex is activated by ascending arousal
systems controlled by uninhibited amygdala circuits, generating further inhibition
of the cingulate. As a result, the individual experiences a sense of vigilant alertness
devoid of any emotional or cognitive content, a state that is ideally adapted for the
control of action in the face of extreme and potentially debilitating danger. If this
response is triggered in the absence of threat, however, the resulting sense of deper-
sonalization and derealization can itself be highly unpleasant and incapacitating.
Given the validity of this account, depersonalization and derealization in the
context of temporal lobe epilepsy may be the result of seizure activity in the amyg-
dala that prevents the emotional tagging of perceptual and cognitive information
prior to its entry into conscious awareness3. It may be that this process is the
product of transient disconnection between the amygdala and sensory areas result-
ing directly from seizure activity. Alternatively, it may reect an indirect defensive
response in the face of anxiety elicited by seizure-based stimulation of the amyg-
dala. Intuitively, one suspects that the former is the more plausible possibility,
although the latter cannot be ruled out a priori. Following this account of deper-
sonalization and derealization, these phenomena can be regarded as dissociative in
sense (ii) of the term; whether they should, in the context of epilepsy, be regarded
as the result of a dissociative defence mechanism remains an empirical issue.
The fact that few epileptic phenomena can be regarded as dissociative in any
strict sense reects widespread confusion over what actually constitutes dissocia-
tion. Although widely endorsed, the idea that any breakdown in memory, con-
sciousness, identity, perception or behavioural control is dissociative overextends
the term and diminishes its descriptive validity (Cardea, 1994). Amnesia cannot
be considered genuinely dissociative unless it involves an inability to retrieve intact
information that should, under normal circumstances, be available for recall
(Kihlstrom, 1994). Amnesia resulting from a failure to encode information, includ-
ing that which occurs in the context of epilepsy, does not fall within this category.
Loss of behavioural control can only be considered dissociative if it is within the
context of an otherwise intact ability to control action. Seizure-related motor phe-
3
Such a process could also be responsible for the paradoxical sense of unfamiliarity that characterizes jamais
vu, a phenomenon commonly observed in epilepsy.
nomena, including complex automatisms, are not dissociative because they occur
only in the context of reduced behavioural control in general. Current psychiatric
taxonomies do not make these distinctions clearly enough, relying instead on a
purely descriptive approach that precludes precise classication based on the
mechanisms underlying dierent phenomena.
Nonepileptic seizures and dissociation
Epileptologists frequently encounter patients who present with paroxysmal events

that, despite resembling epileptic episodes, are actually nonepileptic. Indeed, as
many as 50% of patients referred to specialist epilepsy centres may turn out not to
have epilepsy (Francis and Baker, 1999). While some nonepileptic seizures may be
attributable to physical causes other than epilepsy (see Gates and Erdahl, 1993), a
demonstrable organic basis is absent in many such cases. Of these, some are attrib-
utable to an identiable psychiatric illness, such as psychosis, that can produce
seizure-like symptoms. In other cases, however, nonepileptic seizures4 occur as an
isolated psychiatric problem in their own right. Identifying such cases represents a
considerable challenge to neurologists working within this domain. At present,
there are few, if any, reliable criteria for an inclusive diagnosis of nonepileptic attack
disorder. As a result, current diagnostic practice is essentially based on the exclu-
sion of epilepsy and other physical disorders (see Brown and Trimble, 2000). The
concept of dissociation is particularly important in relation to nonepileptic attacks
as it sheds light on both the mechanisms and, potentially, the dierential diagnosis
of these phenomena.
Although nonepileptic seizures have a presentation as diverse as that associated
with epilepsy itself, it is possible to impose some order on their general semiology.
In a sample of 110 patients with well-documented nonepileptic attacks, Meierkord
et al. (1991) report that approximately one-third of all cases involved a collapse
with limpness, while two-thirds involved prominent motor activity such as limb
thrashing. In this study, decreased responsivity to verbal stimulation was evident in
three-quarters of all cases, while purposeful or semi-purposeful motor behaviour
4
Much has been written about the relative merits of the various terms used to describe these phenomena.
Many have argued that terms such as pseudoseizures are pejorative because they imply that the events in
question are not subjectively compelling (Betts, 1990). In contrast, the neutral alternative term nonepilep-
tic seizures has been criticized for being imprecise, as it fails to distinguish between the many dierent types
of paroxysmal events that are nonepileptic (Kuyk et al., 1999). My own view is that the term somatoform
seizures would be more appropriate than either of the above, as it is nonpejorative, has descriptive preci-
sion and emphasizes the importance of excluding physical illness in the dierential diagnosis of these
events. For the sake of descriptive continuity, however, the terms nonepileptic seizures and nonepileptic
attacks will be adopted here; the term nonepileptic attack disorder will be used to describe the condition
characterized by these events. In the present context, it should be assumed that the term refers specically
to those events that cannot be attributed to either an organic cause or an identiable psychiatric illness.
198 R.J. Brown
was observed in 44%. Just over two-thirds of all cases presented with stereotyped
seizures. These observations are broadly consistent with those reported by Betts
and Boden (1992). In this study, three types of emotional nonepileptic attacks
were identied in addition to those deliberately simulated for primary or secondary
gain and those attributable to a recognizable psychiatric attack disorder (e.g. panic
disorder). Approximately 21% of these individuals experienced so-called swoon
attacks, in which the individual characteristically sinks to the oor and lies inert
and unresponsive for the duration of the attack. Roughly 33% displayed so-called
tantrum attacks, involving a sudden drop to the oor followed by screaming and
thrashing of the limbs. Finally, 46% displayed so-called abreactive attacks, involv-
ing gasping, limb-thrashing, pelvic thrusting and stiening of the body with back
arching.
In all cases, nonepileptic attacks involve a temporary loss of behavioural, sensory
or cognitive control that occurs in the context of intact neuropsychological func-
tioning, as evidenced by a normal EEG during the nonepileptic ictus. The absence
of paroxysmal brain discharges serves as the principal feature that distinguishes
nonepileptic from genuine epileptic events. By itself, however, the EEG cannot
provide a completely reliable basis for the identication of epileptic and nonepilep-
tic seizures (Brown and Trimble, 2000), underlining the potential value of dissoci-
ation as a criterion for an inclusive diagnosis of nonepileptic attack disorder.
Several converging lines of evidence indicate that these events involve a dissoci-
ative psychological mechanism (see Kuyk et al., 1997). In the rst instance, nonepi-
leptic seizures are commonly found in the context of other forms of dissociative
psychopathology. Bowman (1993) and Bowman and Markand (1996) found that
the vast majority of individuals with nonepileptic attacks meet criteria for DSMIV
dissociative disorders such as dissociative amnesia, identity disturbance and deper-
sonalization. Post-traumatic stress disorder, commonly assumed to involve a dis-
sociative mechanism, was also particularly common in this group of patients
(Bowman, 1993; Bowman and Markand, 1996). Other studies have found that
nonepileptic seizures frequently occur alongside other unexplained physical symp-
toms (Krishnamoorthy et al., 2001; Meierkord et al., 1991), suggesting that they
may be one aspect of a broader tendency to express psychological distress somati-
cally, so-called somatization (Lipowski, 1968). A number of authorities have sug-
gested that dissociation is an important aspect of this phenomenon also (Nemiah,
1991). Eating disorder symptoms, which have been linked to a dissociative process
(Pettinati et al., 1985), also appear to be particularly common in patients with non-
epileptic seizures (Krishnamoorthy et al., 2001).
The frequent co-occurrence of dissociative psychopathology in patients with
nonepileptic attacks appears to indicate a general propensity for dissociative exper-
iences in these individuals. Consistent with this notion is a recent study by Kuyk et
al. (1999), showing that individuals with nonepileptic attacks display elevated levels
of hypnotic susceptibility. In a related vein, nonepileptic attacks can, in many such
individuals, be provoked using suggestion, placebo or hypnosis (Dericioglu et al.,
1999). High hypnotic susceptibility is commonly found in patients with dissocia-
tive psychopathology (Frischholz et al., 1992; Pettinati et al., 1985; Spiegel et al.,
1988), and a dissociative interpretation of hypnosis has been oered by a number
of authorities (Hilgard, 1977; Woody and Bowers, 1994). Bowman (1993) also
found that individuals with nonepileptic seizures yield elevated scores on the
Dissociative Experiences Scale (DES; Bernstein and Putnam, 1986), a self-report
measure assessing everyday occurrences of dissociation, compared with nonclini-
cal controls. However, in a more recent study, Alper et al. (1997) found that DES
scores are also elevated in patients with complex partial seizures (see also Devinsky
et al., 1989); indeed, there was no signicant dierence in overall DES scores
between these patients and a group with nonepileptic seizures. Nevertheless, both
epileptic and nonepileptic groups scored higher on the DES than typically observed
in nonclinical populations. In my view, this particular nding demonstrates the
danger of conating the various denitions of dissociation within a single measure
such as the DES. As the DES treats dissociation as a unitary concept, it cannot
dierentiate between conditions that are characterized by dierent forms of disso-
ciative phenomena, such as epilepsy and nonepileptic attack disorder.
It is widely thought that traumatic experiences precipitate the development of
dissociative symptoms, which serve a defensive function that protects the individ-
ual from extreme anxiety and psychological disintegration. Indeed, both DSMIV
and ICD10 make an explicit link between traumatic events and the onset of dis-
sociative symptoms. Moreover, a number of studies have found disproportionately
high rates of physical, sexual and emotional abuse in patients with dissociative dis-
orders (Chu and Dill, 1990; Irwin, 1994; Pribor et al., 1993). As such, evidence indi-
cating an increased prevalence of traumatic experiences in individuals with
nonepileptic seizures could be viewed as additional support for a dissociative inter-
pretation of this phenomenon. To this end, Bowman (1993) found that 70% and
77% of her sample of 27 nonepileptic seizure patients had experienced physical or
sexual abuse respectively. Similarly, Betts and Boden (1992) obtained positive
sexual abuse histories from 54% of 96 patients with nonepileptic seizures.
Evidence implicating high dissociative comorbidity, hypnotic susceptibility
and exposure to trauma in individuals with nonepileptic seizures provides only
indirect evidence for a dissociative interpretation of this phenomenon. Although
such evidence suggests that a tendency to dissociate may be a common feature of
these individuals, it does not constitute conclusive proof that nonepileptic attacks
200 R.J. Brown
are themselves dissociative. A recent study by Kuyk et al. (1999) places such an
interpretation on a rmer footing. Like genuine epileptic seizures, nonepileptic
attacks are often associated with a dense amnesia for events occurring during the
ictus. As I have hopefully demonstrated, genuine epileptic amnesia should not be
considered a dissociative phenomenon because it arises from a seizure-related
disruption in memory encoding rather than an inability to retrieve intact
memory traces. However, Kuyk et al. (1999) have shown that the amnesia asso-
ciated with nonepileptic attacks may actually be the product of such a retrieval
decit. Kuyk et al. (1999) compared a group of individuals with amnesia for
events occurring during well-documented nonepileptic attacks with a group dis-
playing amnesia following complex partial and generalized epileptic seizures. All
subjects were hypnotized and given suggestions designed to facilitate the recov-
ery of ictal events; the experimenter remained blind to group status at all times.
Using a free-recall paradigm, 17 out of 20 patients with nonepileptic seizures
recovered signicant information concerning the designated attack; this informa-
tion was veried by video recordings or third-party reports. In contrast, not one
of the 17 patients with epilepsy retrieved information concerning their attack
during hypnosis. Such a nding appears to demonstrate that, unlike that found
in epilepsy, nonepileptic amnesia results from a process that prevents the individ-
ual from accessing memories successfully encoded during the attack. This appar-
ent separation of intact memorial information from conscious awareness
following a nonepileptic attack, coupled with the phenomenological character of
these events, clearly identies these phenomena as dissociative in sense (i) of the
term.
Dissociative mechanisms of nonepileptic seizures

If one assumes that nonepileptic seizures are a dissociative phenomenon, what
mechanisms might be involved in the generation and maintenance of these events?
Current theorizing in this domain is still very much dominated by nineteenth
century ideas concerning the mechanisms underlying so-called hysterical5 phe-
nomena. Indeed, the term dissociation originates in the work of Pierre Janet who
proposed one of the earliest systematic accounts of the psychological mechanisms
underlying hysteria. According to Janet (1889, 1924), a fundamental weakness in
the hysterical individuals mental character makes them susceptible to a breakdown
in the normally integrated functions of consciousness when faced by environmen-
tal stress or trauma. As a result, organized sets of knowledge pertaining to the
5
In recent years, the term hysteria has fallen from favour due to its unwarranted and pejorative connota-
tions. The multiple unexplained symptoms captured by this concept, of which nonepileptic seizures are
one, are now subsumed within the somatoform and dissociative disorder categories of DSMIV and
ICD10.
trauma may become dissociated from the main body of consciousness, and may
serve to take control of behaviour and experience if activated by environmental
events. The automatic activation of these dissociated memories results in a hyster-
ical reaction (or somnambulism) that, in some instances, takes the form of a non-
epileptic attack. According to this view, two aspects of the hysterical individuals
psychophysiological make-up are responsible for the processes of dissociation and
somnambulism. First, the hysterical individual possesses an abnormally high
degree of suggestibility that allows ideas from the external environment to develop
within them in the absence of their own eort or awareness. Second, the hysterical
individual suers from an attentional dysfunction or retraction of the eld of con-
sciousness (Janet, 1924, p. 314) which prevents them from entertaining alternative
states of mind, thereby accentuating their responsivity to external suggestion.
Although a century old, the theoretical analysis of hysteria oered by Janet con-
tinues to inuence theory and research concerning dissociation and the dissociative
disorders. This inuence is particularly evident in Hilgards (1977) neodissociation
theory and its conceptual derivatives (Woody and Bowers, 1994). According to
Hilgard, behaviour is controlled by the operation of a large set of functionally auton-
omous low-level cognitive control systems specialized for the execution of particular
behavioural acts. Each control system comprises an organized set of well-learned
behavioural and cognitive routines or schema6 developed following extensive expe-
rience with the environment; these control systems are hierarchically organized
beneath an executive ego, a higher-level cognitive structure associated with volition
and consciousness. The executive ego is responsible for selecting the appropriate cog-
nitive control systems for any given task; however, in order to conserve the limited
resources of the executive, once control systems are selected they are able to function
with a considerable degree of autonomy from the ego. As such, they have become dis-
sociated from the executive, with their continued operation occurring largely outside
of awareness; in this way, well-learned behaviours can be performed eortlessly and
concurrently. According to Hilgard, such processing dissociations are fundamental
to human cognition, with the conscious representation of information being largely
unnecessary for the execution of extremely complex behaviours.
Hilgards theory provides an elegant account of the psychological mechanisms
involved in dissociation. Although it was originally developed as a general account
of the mechanisms involved in behavioural control, neodissociation theory has
been most inuential as an account of hypnotic behaviour. According to neodisso-
ciation theory, the hypnotic induction brings about a functional inhibition of the
executive ego, causing it to fractionate into two separate elements (Kirsch and
6
In contemporary cognitive theory, schemas are organized knowledge structures that represent the sequence
of actions and/or processing operations involved in a given behaviour e.g. the movements involved in con-
trolling a car.
202 R.J. Brown
Figure 13.1. The neodissociation account of dissociative behaviours. (Adapted from Kirsch and Lynn,
1995.)
Lynn, 1995). Although one part of the fractionated executive continues to function
as normal during hypnosis, the second part is concealed from awareness by the for-
mation of an amnesic barrier. This part of the ego can exert behavioural control in
the usual fashion but such control is prevented from representing itself in con-
sciousness by the amnesia (see Figure 13.1). Hypnotic behaviours result from the
selection of cognitive control systems via the part of the executive ego concealed by
the amnesia; as the hypnotized individual is aware of only the resulting behaviour
and not the process by which it was selected (due to the amnesia), they experience
the execution of hypnotic suggestions as occurring involuntarily.
The neodissociative model of hypnotic behaviours also provides the basis for an
account of more pathological phenomena such as dissociative amnesia, fugue and
multiple personality disorder (Kihlstrom, 1994; Spiegel and Cardea, 1991).
According to such a view, the formation of amnesic barriers within the executive
ego is a common defensive response in the face of trauma. These barriers serve an
adaptive function in that they protect the individual from experiencing potentially
overwhelming negative aect associated with the traumatic event. However, patho-
logical dissociative amnesia can arise if the barrier within the executive ego endures
to the point where the memory loss itself becomes distressing or debilitating. In the
case of dissociative fugue, the amnesic barrier conceals large tracts of autobio-
graphical memory as well as the traumatic events themselves. Without access to this
autobiographical information, the fugue suerer not only reports amnesia but also
a profound loss of personal identity (Kihlstrom, 1994). Neodissociation theory also

postulates that dierent parts of the executive ego, separated from each other by
amnesic barriers, can be activated at dierent times (e.g. during traumatic events).
If a nonprimary aspect of the ego is activated often enough or for a sucient length
of time, it may become associated with enough autobiographical information to
constitute a separate identity in its own right. Such a process could allow for the
development of multiple identities within a single individual, a phenomenon that
characterizes dissociative identity disorder. Neodissociation theory can also
account for dissociative phenomena characterized by a loss of behavioural control,
such as nonepileptic attacks. By this view, dissociative behaviours are controlled by
that part of the executive ego concealed beneath the amnesic barrier (Kirsch and
Lynn, 1995). As such, neodissociation theory would identify nonepileptic attacks
as essentially voluntary behaviours; however, they are perceived as involuntary
because the individual is amnesic for the executive control involved in their execu-
tion (Kihlstrom, 1994).
Hilgards neodissociation theory represents an important step forward in our
understanding of the dissociation concept. Perhaps most importantly, it recasts
Janets original ideas within a modern cognitive psychological framework that
bridges the gap between pathological (e.g. dissociative amnesia) and nonpatholog-
ical (e.g. hypnotic amnesia) instances of dissociation. While Janet regarded disso-
ciation as an abnormal process provoked by stress in the constitutionally weak
mind of the hysterical individual, neodissociation theory identies dissociation as
a normal psychological process that is utilized by the organism as a defence mech-
anism in the face of trauma. Pathological dissociation arises when this normally
adaptive process becomes generalized to situations where such a response is inap-
propriate. Similarly, hypnotic suggestions capitalize on the dissociative nature of
brain and mind to produce unusual and entertaining phenomena.
Despite its popularity within psychiatry and psychology, there are a number of
problems with neodissociation theory as an account of the mechanisms underly-
ing the dissociative disorders. In the rst instance, neodissociation theory assumes
that pathological dissociation results from a defensive response to trauma that has
become overgeneralized. However, not all individuals who display dissociative
symptomatology have been exposed to trauma. Second, neodissociation theory
assumes that pathological dissociation arises from the generation of an amnesic
barrier. However, amnesia is not apparent in all cases of dissociative symptomatol-
ogy. Many individuals with nonepileptic seizures, for example, are not amnesic for
the loss of behavioural control that they experience during the ictus. Third, neo-
dissociation theory assumes that similar mechanisms underlie both pathological
and nonpathological dissociations, such as those observed during hypnosis.
204 R.J. Brown
However, as Woody and Bowers (1994) have pointed out, the amnesic-barrier
concept that forms the basis of neodissociation theory does not provide a good
explanation of hypnotic behaviours and experiences. If such a notion were correct,
it would imply that all hypnotic phenomena involve some degree of spontaneous
amnesia. This is, in fact, extremely uncommon. One would also expect hypnotic
(i.e. suggested) amnesia to be far more common than is actually the case. Moreover,
many individuals are unable to experience hypnotic amnesia despite being respon-
sive to other hypnotic suggestions.
More recently, Woody and Bowers (1994) have oered an alternative account of
the dissociative mechanisms underlying hypnotic behaviours and experiences.
Intrinsic to neodissociation theory is the idea that dissociative behaviours are con-
trolled by part of an executive system that is concealed beneath an amnesic barrier.
In this sense, neodissociation theory is based on the idea of dissociated experience;
that is, the individual experiences a loss of behavioural control that is, in fact, illu-
sory7 (Woody and Bowers, 1994). Woody and Bowers (1994) have proposed an
alternative account based on the concept of dissociated control. By this view, the dis-
sociation is between the executive ego and the lower-level systems controlling beha-
viour, rather than within the executive itself. According to Woody and Bowers
(1994), this dissociation between higher- and lower-level cognitive control results
from an inhibition of the executive system due to the induction of hypnosis. In this
model, dissociated behaviours are generated by the automatic activation of lower-
level behavioural routines by environmental cues (see Figure 13.2). As such, disso-
ciated behaviours are genuinely involuntary, because they are not controlled by the
executive ego.
Although originally developed to account for the dissociative phenomena
observed in hypnosis, dissociated control theory could also be applied to more
pathological instances of dissociation, such as nonepileptic attacks. By this view,
nonepileptic attacks would be the product of a learned behavioural routine trig-
gered automatically (i.e. without input from the executive control system) by cues
from the environment. The apparently involuntary and stereotyped nature of non-
epileptic attacks (see Meierkord et al., 1991) certainly appears to indicate that this
phenomenon is controlled by the activation of a low-level behavioural routine or
schema. If this were the case, the semiology of any given nonepileptic attack would
correspond to the nature of the behavioural representation underlying it. In addi-
tion to behavioural features (e.g. those representing the movements involved in the
seizure), this schema could also involve a cognitive control component that serves
to inhibit the operation of the executive system and the information that is passed
7
This is based on the notion that any behaviour controlled by the executive ego is volitional, regardless of
whether the executive is concealed beneath an amnesic barrier. See Brown (1999) for a fuller discussion.
Figure 13.2. The dissociated control theory account of dissociative behaviours. (Adapted from Kirsch
and Lynn, 1995.)
to it by lower-level systems. This could account not only for the behavioural aspect
of nonepileptic seizures but also for the loss of behavioural control and disruption
in subjective awareness that characterizes many such events.
Such an account of the mechanisms involved in nonepileptic seizures raises two
important questions. First, what are the origins of the behavioural routines under-
lying these events? The fact that individuals with nonepileptic events have often
been exposed to epilepsy (either in themselves or others) suggests that some sort of
behavioural modelling is involved in this process. Alternatively, it may be that non-
epileptic attacks are a learned behavioural analogue of a biological reaction such as
the sham death reex (Kretschmer, 1926), perhaps triggered previously as an acute
response to threat.
Second, what are the environmental cues that trigger the nonepileptic event? If
one assumes that nonepileptic seizures serve something of a defensive function, it
is likely that a common precipitant of nonepileptic attacks will be feelings of
anxiety, or thoughts, images or memories associated with such feelings. In princi-
ple, however, triggers could be anything that is associated in memory with the
behavioural representation underlying the attack; these are likely to vary from indi-
vidual to individual depending largely on the circumstances surrounding the initial
occurrence of the attacks.
A dissociated control account of nonepileptic seizures has all the advantages of
neodissociation theory and fewer problems. Like neodissociation theory, it can
account for the basic features of nonepileptic seizures, but does so without recourse
206 R.J. Brown
to the much-criticized amnesic barrier notion (Brown, 1999; Woody and Bowers,
1994). Moreover, unlike neodissociation theory, a dissociated control account of
nonepileptic seizures would identify them as genuinely involuntary behaviours.
Such an idea is more intuitively appealing than the neodissociation theory view that
dissociative behaviours are volitional but erroneously perceived as involuntary.
Indeed, the idea that an individual can execute a behaviour deliberately but not be
aware of it (even though they may be paying it full attention) seems something of
a contradiction in terms.
One of the problems with dissociated control theory in its current form is that
the direct activation of low-level cognitive systems by environmental cues is actu-
ally a fundamental aspect of routine behavioural control (Brown, 1999; Kirsch and
Lynn, 1997). This is well illustrated by everyday action-slips (Reason, 1979), such
as dialling a familiar but out-of-date telephone number despite being aware that it
is no longer valid. As such, one need not assume that executive inhibition is neces-
sary for the occurrence of dissociative behaviours. This is entirely consistent with
the present account of nonepileptic attacks, which posits that executive inhibition
occurs after the activation of a low-level behavioural schema (i.e. as part of the
schema itself), not before it.
Summary
Dissociation is a complex and multifaceted concept that is frequently misapplied

within the eld of epilepsy. In this chapter, I have explored the various components
of the dissociation concept and how they relate to the phenomena of epilepsy and
nonepileptic seizures. I have demonstrated why many epileptic phenomena often
thought to be instances of dissociation, such as amnesia, postictal fugue, behav-
ioural automatisms, auras and hallucinations, should not be regarded as dissocia-
tive at all. I have also argued, however, that depersonalization and derealization
occurring in the context of epilepsy can be regarded as genuinely dissociative, in
the sense that they involve an altered state of consciousness characterized by disen-
gagement from the self or environment. I have also presented evidence indicating
that nonepileptic attacks should be considered a dissociative phenomenon, in this
case involving a temporary disruption in behavioural control and subjective aware-
ness despite intact neuropsychological functioning. Although certain aspects of
epilepsy are dissociative, therefore, they do not involve the same type of dissocia-
tion as that underlying nonepileptic attacks.
If the dissociation concept is to prove useful in this area, much greater precision,
both conceptual and methodological, is required. Researchers and clinicians
should be explicit about which denition of dissociation they are referring to and
eorts should be made to construct measures of dissociation that are, unlike the
DES, phenomenon-pure. Assessing the reversibility of nonepileptic amnesia,

which may prove invaluable as an aid to dierential diagnosis in this area, provides
one illustration of the potential utility of such an endeavour.
R E F E R E N C ES
Alper, K., Devinsky, O., Perrine, K. et al. (1997). Dissociation in epilepsy and conversion nonepi-
leptic seizures. Epilepsia, 38, 9917.
(Fourth Edition) (DSMIV). Washington, DC: APA.
Bancaud, J., Brunet-Bourgin, F., Chavel, P. and Halgren, E. (1994). Anatomical origin of dj vu
and vivid memories in human temporal lobe epilepsy. Brain, 127, 7190.
Bernstein, E. and Putnam, F.W. (1986). Development, reliability and validity of a dissociation
scale. J Nerv Ment Dis, 174, 72735.
Betts, T. (1990). Pseudoseizures: seizures that are not epilepsy. Lancet, 336, 1634.
Betts, T. and Boden, S. (1992). Diagnosis, management and prognosis of 128 patients with non-
epileptic attack disorder. Part I. Seizure, 1, 1926.
Bowman, E.S. (1993). Etiology and clinical course of pseudoseizures: relationship to trauma,
depression and dissociation. Psychosomatics, 4, 33342.
Bowman, E.S. and Markand, O.N. (1996). Psychodynamics and psychiatric diagnoses of pseudo-
seizure patients. Am J Psychiatry, 153, 5763.
Brown, R.J. (1999). An Integrative Cognitive Theory of Suggestion and Hypnosis. PhD thesis,
University College London.
Brown, R.J. and Trimble, M.R. (2000). Dissociative psychopathology, non-epileptic seizures and
neurology. J Neurol Neurosurg Psychiatry, 69, 28591.
Cardea, E. (1994). The domain of dissociation. In Dissociation. Clinical and Theoretical
Perspectives, ed. S.J. Lynn and J.W. Rhue, pp. 1531. New York: Guilford Press.
Chu, J.A. and Dill, D.L. (1990). Dissociative symptoms in relation to childhood physical and
sexual abuse. Am J Psychiatry, 147, 88792.
Dericioglu, N., Saygi, S. and Ciger, A. (1999). The value of provocation methods in patients sus-
pected of having non-epileptic seizures. Seizure, 8, 1526.
Devinsky, O., Putnam, F., Grafman, J., Bromeld, E. and Theodore, W.H. (1989). Dissociative
states and epilepsy. Neurology, 39, 83540.
Erdelyi, M.H. (1985). Psychoanalysis: Freuds Cognitive Psychology. New York: W.H. Freeman.
Fodor, J.A. (1983). The Modularity of Mind. Cambridge, MA: MIT Press.
Francis, P. and Baker, G.A. (1999). Non-epileptic attack disorder (NEAD): a comprehensive
review. Seizure, 8, 5361.
Frischholz, E.J., Lipman, L.S., Braun, B.G. and Sachs, R. (1992). Psychopathology, hypnotizabil-
ity, and dissociation. Am J Psychiatry, 149, 15215.
Gates, J.R. and Erdahl, P. (1993). Classication of non-epileptic events. In Non-Epileptic Seizures,
ed. J.R. Gates and A.J. Rowan, pp. 2130. Oxford: ButterworthHeinemann.
208 R.J. Brown
Good, M.I. (1993). The concept of an organic dissociative syndrome: what is the evidence? Harv
Rev Psychiatry, 1, 14557.
Halligan, P.W., Athwal, B.S., Oakley, D.A. and Frackowiak, R.S.J. (2000). Imaging hypnotic par-
alysis: implications for conversion hysteria. Lancet, 355, 9867.
Hilgard, E.R. (1977). Divided Consciousness: Multiple Controls in Human Thought and Action.
New York: John Wiley & Sons.
Irwin, H.J. (1994). Proneness to dissociation and traumatic childhood events. J Nerv Ment Dis,
8, 45660.
Janet, P. (1889). Lautomatisme Psychologique. Paris: Felix Alcan.
Janet, P. (1924). The Major Symptoms of Hysteria (2nd edition). New York: Macmillan.
Kihlstrom, J.F. (1994). One hundred years of hysteria. In Dissociation: Clinical and Theoretical
Perspectives, ed. S.J. Lynn and J.W. Rhue, pp. 36594. New York: Guilford Press.
Kirsch, I. and Lynn, S.J. (1995). The altered state of hypnosis. Am Psychologist, 50, 84658.
Kirsch, I. and Lynn, S.J. (1997). Hypnotic involuntariness and the automaticity of everyday life.
Am J Clin Hypnosis, 40, 32948.
Kosslyn, S.M. (1994). Image and Brain. Cambridge, MA: MIT Press.
Kretschmer, E. (1926). Hysteria. New York: Nervous and Mental Disease Publishing Co.
Krishnamoorthy, E.S., Brown, R.J. and Trimble, M.R. (2001). Personality and psychopathology
in non-epileptic attack disorder (NEAD): a prospective study, Epilepsy Behav, 2, 41822.
Kuyk, J., van Dyck, R. and Spinhoven, P. (1997). The case for a dissociative interpretation of
pseudoepileptic seizures. J Nerv Ment Dis, 184, 46874.
Kuyk, J., Spinhoven, P. and van Dyck, R. (1999). Hypnotic recall: a positive criterion in the
dierential diagnosis between epileptic and pseudoepileptic seizures. Epilepsia, 40, 48591.
Lipowski, Z.J. (1968). Review of consultation psychiatry and psychosomatic medicine. III.
Theoretical issues. Psychosom Med, 30, 395422.
Meierkord, H., Will, B., Fish, D. and Shorvon, S. (1991). The clinical features and prognosis of
pseudoseizures diagnosed using video-EEG telemetry. Neurology, 41, 16436.
Nemiah, J.C. (1991). Dissociation, conversion and somatization. In American Psychiatric Press
Annual Review of Psychiatry, ed. A. Tasman and S. Goldnger, Vol. 10, pp. 24860.
Washington, DC: APA Press.
Oakley, D.A. (1999). Hypnosis and conversion hysteria: A unifying model. Cognitive
Neuropsychiatry, 4, 24365.
Pettinati, H.M., Horne, R.L. and Staats, J.M. (1985). Hypnotizability in patients with anorexia
nervosa and bulimia. Arch Gen Psychiatry, 42, 101416.
Pribor, E.E., Yutzi, S.H., Dean, T.J. and Wetzel, R.D. (1993). Briquets syndrome, dissociation, and
abuse. Am J Psychiatry, 150, 150711.
Reason, J.T. (1979). Actions not as planned. In Aspects of Consciousness, Vol. 1: Psychological Issues,
ed. G. Underwood and R. Stevens, pp. 6789. London: Academic Press.
Sierra, M. and Berrios, G.E. (1998). Depersonalization: neurobiological perspectives. Biol
Spiegel, D. and Cardea, E. (1991). Disintegrated experience: the dissociative disorders revisited.
J Abnorm Psychol, 100, 36678.
Spiegel, D., Hunt, T. and Dondershine, H.E. (1988). Dissociation and hypnotizability in post-
traumatic stress disorder. Am J Psychiatry, 145, 3015.
Thomas, L. and Trimble, M.R. (1997). Dissociative disorders. In Epilepsy: A Comprehensive
Textbook, ed. J. Engel and T.A. Pedley, pp. 277584. Philadelphia: LippincottRaven
Publishers.
Woody, E.Z. and Bowers, K.S. (1994). A frontal assault on dissociated control. In Dissociation:
Clinical and Theoretical Perspectives, ed. S.J. Lynn and J.W. Rhue, pp. 5279. New York:
Guilford Press.
Disorders: Clinical Descriptions and Diagnostic Guidelines (Tenth Revision) (ICD10). Geneva:
WHO.
14
Psychobiology of psychogenic
pseudoseizures
J. Chris Sackellares and Dalma Kalogjera-Sackellares
Malcolm Randall VA Medical Center and University of Florida, Florida, USA
Introduction
Psychogenic pseudoseizures are paroxysmal events which mimic epileptic seizures.

While patients suering from these symptoms are referred to neurologists because
they are mistakenly believed to have epilepsy, neurologists consider the underlying
disorder to be of psychological aetiology. Our observations on the nature of
pseudoseizures are based on a carefully studied sample of 100 patients with pseudo-
seizures evaluated over a period of 5 years at the University of Michigan Medical
Center. The patients were evaluated by intensive neurological, clinical psychologi-
cal and/or neuropsychological investigations. Psychodynamic psychotherapy was
performed in a portion of this group. The results of these eorts have provided us
with insights into the psychopathology of the disorder (Kalogjera-Sackellares,
1995; Kalogjera-Sackellares and Sackellares, 1997a, b, 1999). In addition, we dis-
covered evidence to suggest that neurological disturbances may play a role in the
pathophysiology of the disorder (Kalogjera-Sackellares and Sackellares, 1999). A
recurrent theme among our patients was trauma. This included physical as well as
emotional trauma (Kalogjera-Sackellares, 1995; Kalogjera-Sackellares and
Sackellares, 1999). The prevalence of trauma, as well as the post-traumatic charac-
ter of many of the symptoms, do not t with the current diagnostic framework of
the somatoform disorders (Guggenheim and Smith, 1995) under which pseudosei-
zures are typically considered. The concept of somatoform disorders obscures the
operative impact of both the emotional as well as physical trauma which are cor-
nerstones of our approach to this disorder. Our own observations regarding the
psychopathology as well as the possibility of neurological disturbance led us to an
extensive review of the literature on pseudoseizures and hysteria.
Among the richest sources of information on the subject are the case descrip-
tions by the nineteenth century clinical investigators, Paul Briquet (a psychiatrist)
and Jean-Martin Charcot (a neurologist). Briquet wrote extensively on the subject
of hysterical seizures (a historic term for pseudoseizures) in his classic book on hys-
210
211 Psychobiology of psychogenic pseudoseizures
teria (Briquet, 1859). Charcots case descriptions as well as his comments on the
subjects have been captured for posterity in the form of published lectures
(Charcot, 1879; Goetz, 1987; Harris, 1991). These clinicians contributed much to
the foundations of both neurology and psychiatry. As we came to learn from the
review of their work, both believed that hysteria resulted from a neurological dis-
order, but failed to nd neuroanatomical abnormalities to support this view. As a
result, neurological investigations into this disorder all but ceased. The emphasis
shifted to investigations into the psychopathology of the disorder. These psychiat-
ric investigations occurred under the historic impact of the pioneering insights of
Janet, Freud and others. Our analysis of the literature was aided considerably by the
translation of Charcot by G. Siegerson (Charcot, 1879), as well as the more recent
translations (Goetz, 1987; Harris, 1991). Analysis of Briquet, on the other hand,
was largely based on direct translations (by DKS) of his text (Briquet, 1859) because
no English translation of his work was available.
In this chapter, we will review the historical events that led to the commonly held
view that pseudoseizures are solely grounded in psychopathology, in the absence of
detectable neurological disorders. In reviewing the works of Charcot and Briquet,
it is apparent that the importance of trauma in the development of hysteria was rec-
ognized by both Briquet (1859) and Charcot (1879; Harris, 1991). However, focus-
ing on a series of well-described cases of Charcot, we nd that it was emotional
trauma, not physical trauma that he emphasized. Yet, we will provide ample evi-
dence that, in Charcots cases, physical trauma (particularly head injury) may have
played an equally important role. In addition to our review of historical cases, we
will discuss the results of recent studies supporting the concept that neurobiologi-
cal disturbances interact with traumatic environmental stresses, leading to the
development of psychogenic pseudoseizures.
Historically, psychogenic pseudoseizures have been considered to be a classic
manifestation of hysteria. Until the emergence of the modern view of hysteria in
the nineteenth century, hysteria was generally believed to result from disturbances
in the reproductive system, specically the uterus. Nineteenth century investiga-
tors, particularly Pierre Briquet, and Jean-Martin Charcot, rejected the aetiological
role of the uterus. In addition to making this historic contribution, they further-
more hypothesized that hysteria arose from disturbances in the central nervous
system (Briquet, 1859; Goetz, 1987; Harris, 1991). Thus, both of these clinical
investigators believed hysteria to be a neurological disorder (although Briquets
contribution to the conceptualization of hysteria as a disorder of the central
nervous system is rarely acknowledged and has only recently been discussed at
some length in the context of the forgotten avantgarde of neuroscience (Mai, 1982;
Mai and Merskey, 1981). First Briquet (Mai, 1982; Mai and Merskey, 1981), and
later Charcot, attempted to identify lesions in the nervous system to account for the
212 J.C. Sackellares and D. Kalogjera-Sackellares
symptoms of hysteria. They were unable to nd relevant anatomical lesions in their

patients. However, their quest was limited by the histological methods of the times.
In response to the absence of identiable lesions, both men postulated physiologi-
cal or dynamical neurological disturbances to account for the functional distur-
bances and physical ndings they observed in their patients.
As noted earlier, Charcot emphasized the importance of trauma in patients with
hysteria. However, he focused upon the emotional trauma, or shock rather than
physical injury to the nervous system. Review of his cases fully substantiates the
importance of shock and emotional trauma in these unfortunate individuals
(Harris, 1991). However, his case reports also describe physical trauma, and in par-
ticular, head injury as well as debilitating illnesses in the vast majority of his cases.
While Charcot described these physical insults, he did not postulate a causal rela-
tionship between head injury and hysteria, presumably because of the absence of
neurological lesions in postmortem tissue.
Given the obvious presence of emotional trauma and psychological symptoms
among hysterical patients and the failure to nd neurological lesions, it was natural
that subsequent clinical investigators, led by Janet and Freud, would focus upon
psychodynamics. The success of psychological investigations and failure of biolog-
ical investigations led neurologists to abandon the subject of hysteria. The primary
interest of neurologists became limited to the dierentiation of hysteria from true
neurological disorders.
Apart from recognizing the importance of emotional trauma in his patients,
Charcot was unable to provide further insight into the aetiology of hysteria.
However, he did provide extraordinarily useful descriptions of ndings on neuro-
logical examination as well as vivid description of hysterical seizures. His extraor-
dinarily detailed studies of the phenomenology of hysterical patients can be found
in the lectures chronicled by his students (Charcot, 1879; Harris, 1991). Based on
these observations, Charcot described four phases of the archetypal hysterical
seizure: (1) a prodromal phase, (2) an epileptoid phase, (3) a phase characterized
by bizarre postures and (4) an emotional phase. Less severe or elaborate seizures
could occur. These modied attacks consisted of only parts of the full-blown arche-
typal seizures. The study of these phases was aided by drawings that captured the
important details of hysterical seizures (see Figures 14.1 and 14.2). A large sample
of these drawings was published in a book by Richer (1885). Richers drawings cap-
tured the motor manifestations of pseudoseizures (for example, see Figure 14.1).
Even more impressive was his success in rendering facial expressions which capture
the intense emotional and experiential aspects of certain phases of the hysterical
seizure. An example is illustrated in Figure 14.2.
The question arises as to whether todays patient with psychogenic pseudosei-
zures suers from the same malady as the hysteria cases of Charcots time. Given the
Figure 14.1. Reproduction of a drawing by Richer depicting a patient experiencing the epileptoid phase
of a hysterical seizure (from Richer, 1885).
high incidence of nonepileptic seizures (hystero-epilepsy) in hysterical patients of

the nineteenth century, it seems reasonable to assume that Charcot was describing
the same clinical entity as that underlying todays psychogenic pseudoseizure. Citing
Briquet, Charcot stated that hysterical seizures occurred in all but 25% of patients
with hysteria (Harris, 1991). Thus, nonepileptic seizures have been considered a
typical manifestation of hysteria from the beginning of modern neurology. In the
nineteenth century, hysteria was grouped along with other disorders which had no
known anatomical lesion. Examples of such disorders included epilepsy and chorea,
neither of which were associated with neuropathological lesions demonstrable with
the histological techniques of the time. Disorders considered to be neurological, yet
without a demonstrable lesion, were classied as neuroses. Paradoxically, while the
term neurosis has persisted, its modern meaning, in contrast to that of nineteenth
century clinical investigators, indicates psychological causation.
The modern vestige of this early attempt at classication is a tendency among
many clinicians to consider nonepileptic seizures as functional rather than
Figure 14.2. Reproduction of the period of passionate attitude (the emotional phase), the fourth
stage of a classical hysterical seizure.
neurological or organic in aetiology. The implication is that nonepileptic pseudo-

seizures do not have a biological substrate. However, careful reading of Charcots lec-
tures suggests that he used the term functional to mean a disorder of the nervous
system that is not associated with demonstrable lesions of the nervous system. Today,
with knowledge gained from more powerful investigative tools, we have extended
the concept of organic to include a number of toxic, metabolic, degenerative and
genetic disorders that are not associated with physical lesions. In clinical practice
today, the term functional, applied to a symptom or nding on examination, is
taken to mean that the cause is psychological as opposed to biological in nature. It
is generally accepted that many disorders once considered to be functional are due
to biological causes (e.g. Gilles de la Tourette syndrome, schizophrenia, bipolar
aective disorder). In an analogous manner, it is possible that a biological substrate
for pseudoseizures will be identied, using todays sophisticated investigative tools.
Charcot meticulously applied the prevailing clinicalanatomical method, a
method which he helped perfect, to the investigation of patients with hysteria. The
clinicalanatomical method involved the correlation of clinical signs and symp-
toms to anatomical lesions in the nervous system. His application of this method
to the study of the progressive muscular atrophies, amyotrophic lateral sclerosis,
and multiple sclerosis resulted in major contributions to the understanding of
those disorders. Such cases were classied as organic because they were associated
with localized lesions of the nervous system. Such lesions were detectable by visual
inspection of autopsy material. More detailed localization and description of these
organic lesions was made possible by skilful use of the microscope, a tool that had
been recently introduced into medical science.
In contrast to his success with the organic neurological disorders, Charcot was
unable to unravel the mystery of hysteria through the application of the clini-
calanatomical method. Hysterical patients presented a plethora of observable
physical ndings on neurological examination. In addition to seizures, these nd-
ings included blindness, diplopia, limb weakness, sensory loss, tremors and invol-
untary movements. However, these ndings occurred in patterns that diered from
those found in patients with observable anatomical lesions of the nervous system.
In some cases, the pattern of visual disturbance, weakness, or sensory loss con-
icted with the rules of functional neuroanatomy known at the time. More impor-
tantly, the profound clinical abnormalities were not associated with detectable
lesions in the nervous system. Charcot was convinced that the clinical ndings of
hysterical patients were due to appropriately located anatomical abnormalities.
However, in the absence of observable structural lesions, he referred to these lesions
as dynamical lesions (Harris, 1991).
Although the clinical ndings in hysterical patients diered from those of the
organic neurological disorders, Charcot was able to demonstrate classical clinical
ndings that have become the hallmark of hysteria. These ndings include con-
stricted visual elds, monocular diplopia, monoplegia, strange contractures and
hemianaesthesia. The motor weakness observed in hysterics was not associated
with the reex abnormalities which were seen in patients with demonstrable lesions
of the nervous system. Even today, such ndings are taken as evidence for the pres-
ence of hysteria, although modern psychiatrists prefer to apply more modern
terms, such as conversion disorder.
Charcot preferred not to oer theories to explain neurological disorders, generally
preferring to conne his discussions to empirical ndings. However, his observations
of the eects of hypnosis and suggestion in hysterics served as a stimulus for the sub-
sequent development of the psychological theories pioneered by Janet and Freud
(Goetz, 1987). The clinical histories he provided during his lectures frequently
described neurological as well as psychiatric illness in family members. His com-
ments suggest that he considered heredity an important factor in the development of
hysteria. Although he did not speculate about this issue, his discussions indicate that
various neurological symptoms appearing in the same family represented dierent
clinical manifestations of an underlying hereditary predisposition (Harris, 1991).
Charcots lectures included descriptions of traumatic events in his female
patients with hysterical seizures (Charcot, 1879); however, we noted that it was in
his male hysterics that he emphasized the history of antecedent trauma and previ-
ous illness (Harris, 1991). In his lectures on seven cases of male hysteria (in all of
whom pseudoseizures gured prominently), he emphasized the history of antece-
dent trauma and its relationship to the onset of symptoms. During his lectures, his
comments focused upon psychological trauma. However, our careful review of his
lectures on seven male patients with hysteria reveals numerous references to signi-
cant closed head injuries. Indeed, he describes closed head injuries in four of the
seven cases described. A fth patient lost consciousness as a result of an accident
that caused severe blood loss. Three were either physically abused or attacked.
Antecedent debilitating illnesses occurred in ve cases. Emotional trauma is
described in all seven cases. Paternal alcohol abuse was described in four cases, but
there is only one patient with a history of past alcohol abuse. Charcots emphasis
on the role of emotional trauma and not physical trauma is understandable, given
the intolerable psychological and social stresses patients experienced. What is less
clear is that he did not consider the impact of the physical trauma which was equally
apparent in his patients.
Consistent with the historic heritage discussed in this chapter, a number of con-
temporary authors have rediscovered serious traumatic experiences in patients
with psychogenic pseudoseizures (Alper et. al., 1993; Arnold and Privitera, 1996;
Bowman, 1993; Cartmill and Betts, 1992; Gross, 1982; Harden, 1997; Kalogjera-
Sackellares, 1995). This recurrent theme of trauma gured so prominently in the
histories of pseudoseizure patients that it played a key role in the classication of
pseudoseizure syndromes proposed by Kalogjera-Sackellares (1995). The continu-
ity of historical and contemporary reference to trauma makes the study of trauma
a natural starting point for investigations into the nature of pseudoseizures
(Kalogjera-Sackellares and Sackellares, 1999). It must be considered relevant to the
understanding of its causes and to the development of approaches to clinical man-
agement (Kalogjera-Sackellares, 1995).
Although psychic trauma is a dominant factor in most cases of psychogenic
pseudoseizures, physical trauma, often involving closed head injuries also is a
Table 14.1. Demographics of 100 patients with psychogenic pseudoseizures referred to

the neuropsychology programme at the University of Michigan
Males 20%
Females 80%
Mean age 33.7%
Pure pseudoseizures 71%
Mixed epilepsy and pseudoseizures 29%
Abnormal neurological examination 4%
Structural brain abnormality on neuroimaging 8%
History of closed head injury 52%
History of substance abuse 13%
Source: Adapted with permission.
prominent part of the history in many patients. Several recent authors have com-
mented on head injuries in their patients with psychogenic pseudoseizures. A high
incidence of head trauma occurring during physical abuse or during accidents was
reported by Bowman (1993). Similarly, Lancman et al. (1993) found many patients
with head trauma associated with loss of consciousness. In addition, Westbrook et
al. (1998) found that 32% of their patients with nonepileptic seizures had experi-
enced an antecedent head trauma which was classied as minor in 91% of the cases.
Upon reviewing the histories of 100 patients with psychogenic pseudoseizures
evaluated at the University of Michigan (Kalogjera-Sackellares and Sackellares,
1999), we found a documented history of signicant closed head injury in 52% (see
Table 14.1). In most instances, the trauma would have been categorized as mild
head injury. These patients sustained head injury as a result of motor vehicle acci-
dents, accidents on the job, falls or physical assault. Often, the physical assault was
part of a chronic pattern of physical abuse. In many instances, these injuries
occurred during childhood. However, there are some cases in which there was no
history of signicant head trauma until adulthood. As in many of Charcots cases,
in those cases with head trauma during adulthood, the injury preceded seizure
onset by days to months and the patients usually ascribed a causal role to the head
injury.
The high incidence of head trauma in patients with psychogenic pseudoseizures
raises the question as to whether the head trauma may have played some role in the
pathogenesis of the disorder. Is it possible that mild brain injury can render a
patient more vulnerable to psychogenic pseudoseizures? In our sample, other
potential causes of brain injury were reported. A history of drug or alcohol abuse
was found in 13% of patients (Table 14.1). Approximately 29% (Table 14.1) were
found to have a prior or concurrent history of epilepsy. In another sample, we

found evidence for active epilepsy in approximately 25% of patients with psycho-
genic pseudoseizures (Kalogjera-Sackellares and Sackellares, 1997a, b). In their case
series, Wilkus et al. (1984) reported a high incidence of events that may have con-
tributed to neurological impairment. Such events occurred in fully 80% of their
sample. These events included head injury with loss of consciousness or other
sequelae in 7 of their 25 cases; an infectious disorder with sequelae in 2 cases; prior
brain surgery had occurred in 2 cases. Other events such as birth trauma, sun
stroke, severe heat exhaustion, partial drowning or gas exposure also occurred in
their patients.
Further evidence of impaired brain function in patients with psychogenic
pseudoseizures comes from studies of cognitive and intellectual function in these
patients. Neuropsychological test batteries are employed in clinical practice to detect
evidence of organic brain dysfunction. Therefore, it is natural to expect that patients
with a purely functional disorder would perform normally on these tests.
Alternatively, one might expect the pattern of test performance to dier from
patients with organic brain disorders. However, several investigators have reported
impaired neuropsychological performances in patients with pseudoseizures. In fact,
several studies, reported by dierent investigators, have produced very similar
results (Kalogjera-Sackellares and Sackellares, 1999). Wilkus et al. (1984; Wilkus and
Dodrill, 1989) reported results from administration of an intensive clinical neuro-
psychological test battery in 25 patients with psychogenic pseudoseizures evaluated
at the University of Washington. These investigations revealed that their sample of
patients performed in the impaired range on 51.2% of the tests. Sackellares et al.
(1985) analysed a sample of patients evaluated at the University of Virginia and
University of Michigan. The authors found that the cognitive performance of that
group was less than would be expected for normal individuals of similar intelli-
gence. Binder et al. (1998) compared a group of patients with nonepileptic seizures
with a group of patients with epilepsy and with normal control subjects on a battery
of neuropsychological tests. They found that the two seizure groups performed sig-
nicantly more poorly than normal controls. However, there were no signicant
dierences between the two seizure groups on any items of the extensive test battery.
They also found that performance on neuropsychological tests was more strongly
correlated with measures on the MMPI/MMPI-2 in the nonepileptic seizure group.
Based on these observations, the investigators concluded that neuropsychological
tests do not discriminate between patients with epilepsy and those with nonepilep-
tic seizures. These investigators concluded that impairment of neuropsychological
performance was strongly related to emotional and psychosocial factors.
In a more recent study, we examined intellectual and neuropsychological test
performance in 53 patients with psychogenic pseudoseizures (Kalogjera-
Sackellares and Sackellares, 1999). The group comprised 44 patients with pseudo-
seizures who had no evidence of concomitant epilepsy, and no denitive historical
data that would support the diagnosis of antecedent epilepsy. This subgroup was
designated as the pure pseudoseizure subgroup. In addition, there were nine
patients with documented pseudoseizures and concomitant epilepsy, or a well-
documented history of epileptic seizures. This subgroup was designated as the
mixed pseudoseizures and epilepsy subgroup. In both subgroups, there was a pre-
ponderance of women (77% in the pure subgroup and 78% in the mixed sub-
group). Intellectual functioning had been measured, using the WAISR in each of
these patients. The HalsteadReitan Neuropsychological Test Battery (Boll, 1981)
was administered to all but one. These tests were performed as a part of standard
clinical evaluation which included medical and neurological history, physical and
neurological examinations, routine EEG, neuroimaging, neuropsychological eval-
uations, and in most cases, long-term EEG-video monitoring. Interestingly, there
were no statistically signicant dierences in the two subgroups with respect to
WAISR scores. The group as a whole (pure and mixed subgroups combined)
revealed an interesting pattern of scores (see Figure 14.3). The mean verbal IQ, per-
formance IQ, and full scale IQ all fell in the average range. Scores were highly var-
iable and ranged from the mentally decient to the very superior. However, the
overall distribution was skewed toward the low end of the average range. Scores on
the HalsteadReitan Neuropsychological Test Battery were not signicantly dier-
ent for the pure and mixed subgroups. Of particular importance is the nding that
more than 50% of the group as a whole performed in the impaired range on more
than half of individual subtests of the battery (based on published cuto scores of
Jarvis and Barth (1984). The percentage of patients scoring in the impaired range
on each of the HalsteadReitan tests for which cuto scores are available is shown
in Table 14.2. The Halstead Impairment Index (HII) is a summary score for the
HalsteadReitan Neuropsychological Test Battery. The mean HII for the entire
sample and for each subgroup was in the impaired range. However, individual HII
scores varied substantially (Figure 14.4). Scores ranged from 0.1 (normal) to 1.0
(impaired performance on all the constituent tests used to calculate the HII). The
HII fell within the moderately to severely impaired range (0.7 and higher, based on
the criteria of Jarvis and Barth (1984)) in nearly half (48.8%) of the pure pseudo-
seizure group. This nding was particularly important given that this group had
not been diagnosed with any dened neurological disorder and their symptoms
had been assumed to result from purely psychological causes.
This study (Kalogjera-Sackellares and Sackellares, 1999) involved a large sample
of patients with well-documented psychogenic pseudoseizures. The intellectual
and neuropsychological ndings of this study are consistent with those of previ-
ous studies. These studies together underscore the prevalence of measurable
16
14
Number of subjects
12
10
0
6069 7079 8089 9099 100109 110119 120129 130139
IQ range
Verbal IQ
Performance IQ
Full scale IQ
Figure 14.3. Distribution of scores on the Wechsler Adult Intelligence Scale Revised (WAISR) in a
sample of 53 patients with psychogenic pseudoseizures. Scores range from the mentally
deficient to the very superior range. However, the overall distribution was skewed toward
the low end of the average range. (Adapted with permission.)
neuropsychological and cognitive decits in patients with psychogenic pseudo-

seizures. Although some investigators have attributed these decits to emotional
factors (Binder et al., 1998), an equally plausible explanation is that many patients
with pseudoseizures, with or without concomitant epilepsy, have undiagnosed
neurological abnormalities. In the vast majority of cases, structural abnormalities
are not reported on MRI and CT scans. In the vast majority of cases, neurological
examinations do not reveal focal or lateralized abnormalities that t patterns that
are seen with lesions of the nervous system. In fact, focal or lateralized neurologi-
cal ndings do not t characteristic patterns associated with anatomical lesions of
the nervous system. However, this does not exclude the possibility of biological
disturbances aecting brain function. Not all such disturbances cause gross
anatomical changes detectable by routine CT or MRI scans.
A surprising observation in our patient sample provides additional evidence that
biological factors may play an important role in the pathogenesis of psychogenic
Table 14.2. Percentage of subjects scoring in the impaired range on HalsteadReitan

Neuropsychological Test Battery
Groups Pure Mixed pseudoseizures

combined pseudoseizures and epilepsy
Test
Halstead Index 63% 61% 75%
Categories 67% 73% 33%
TPT-Total 46% 43% 63%
TPT-Memory 18% 14% 38%
TPT-Location 66% 62% 88%
Speech Sounds Perception Test 60% 59% 63%
Seashore Rhythm Test 62% 63% 57%
Trails A Time 24% 21% 38%
Trails B Time 42% 38% 63%
Finger tapping dominant hand 98% 98% 100%
Source: Reprinted with permission.
pseudoseizures. We recently discovered the very interesting fact that approximately

30% of our patients with pure pseudoseizures are left-handed (Kalogjera-
Sackellares and Sackellares, 2001). This is markedly increased in comparison with
the 10% prevalence of left-handers reported in the general population (Hardyck
and Petrinovich, 1977). This observation can be explained in one of three ways: (1)
it is a chance observation, (2) left-handers are prone to developing psychogenic
pseudoseizures, or (3) our sample contains pathological left handers. The probabil-
ity of nding this high incidence of left-handers by chance alone is extremely low
(P0.000381) (see Kalogjera-Sackellares and Sackellares, 2001). That normal left-
handers are predisposed to developing some sort of functional disorder character-
ized by seizure-like activity seems quite unlikely. Given the neuropsychological and
cognitive diculties observed in patients with pseudoseizures, it is more likely that
the over-representation of left-handers in our sample is due to the presence of some
pathological left-handers (see Kalogjera-Sackellares and Sackellares, 2001).
Pathological left-handers are individuals who were destined to be right-handers,
but developed left-hand dominance due to an insult to the left frontal lobe pre-
natally, during infancy or early in childhood.
In summary, we traditionally consider psychogenic pseudoseizures to be a mani-
festation of a purely psychological disorder. This view stems largely from the
absence of gross neuroanatomical lesions that can explain the seizures or associated
neurological symptoms and signs. Further, these patients manifest a plethora of
222
Number of subjects J.C. Sackellares and D. Kalogjera-Sackellares
Impairment index
Figure 14.4 Distribution of the Halstead Impairment Index (HII), the summary score for the
HalsteadReitan Neuropsychological Test Battery, administered to 52 of the same patients
shown in Figure 14.3. A score of 0.1 is normal. Scores of 0.7 and above are in the
moderately to severely impaired range. A score of 1.0 indicates performance in the
pathological range on all constituent tests used to compute the HII. Note that individual
scores ranged from normal to 1.0. However, 46.9% of patients in the entire group (pure
pseudoseizures and mixed pseudoseizures and epilepsy subgroups combined) scored in
the moderately to severely impaired range; and 48.8% of the pure group (n44) scored
in the moderately to severely impaired range. (Adapted with permission.)
emotional and personality disturbances (Bowman, 1993; Cartmill and Betts, 1992;
Kalogjera-Sackellares, 1995; Kalogjera-Sackellares and Sackellares, 1997a, b; Roy,
1979; Vanderzant et al, 1986) which direct the clinician toward psychiatric issues.
The importance of psychological factors is further underscored by the prevalence
of traumatic emotional experiences in these patients. Yet, a persistent question is
why these traumatic experiences trigger pseudoseizures and other pseudoneuro-
logical symptoms in some individuals, but not others. At least one possibility is that
some individuals are biologically susceptible to the development of psychogenic
pseudoseizures in response to traumatic emotional experiences.
We suggest that, at least in some individuals, the biological susceptibility may

result from neurological impairment. As our previous discussion indicates, a
number of investigators have reported a high incidence of closed head injury and
other neurological insults in their patients. Several investigators have reported
impaired performance on tests of neuropsychological and cognitive function.
Finally, we have found an over representation of left-handers in our pseudo-
seizure patient sample (Kalogjera-Sackellares and Sackellares, 2001). This over-
representation of left-handers suggests one of two possibilities. First, it is possible
that left-handers subjected to certain neurological insults or severe emotional
trauma may be predisposed to psychogenic pseudoseizures. A second, and more
plausible explanation is the presence of pathological left-handers in the sample. If
the additional left-handers are pathological left-handers, it suggests that some
patients may have experienced neurological insults to the left hemisphere prior to
the establishment of hand dominance (Kalogjera-Sackellares and Sackellares,
2001).
Taken together, all of these observations suggest that biological factors play an
important aetiological role in the development of psychogenic pseudoseizures.
This possibility in no way negates the role of emotionally traumatic experiences or
other environmental factors in the development of psychiatric disorders character-
ized by pseudoseizures. Nonetheless, it may help to explain why all individuals
experiencing traumatic life events do not develop pseudoseizures.
We feel that it is extremely important to consider the strong possibility that
psychobiological factors may play an important role in the pathophysiology of epi-
lepsy. Because pseudoseizures have long been assumed to stem from purely func-
tional causes, there have been essentially no investigations into the neurobiology
of this disorder.
This disorder causes as much disability as medically intractable epilepsy. Yet,
little progress has been made in the exploration of new therapeutic interventions.
Psychotherapy can be benecial to many of these patients (Kalogjera-Sackellares,
1995). However, not all respond, and many do not have access to psychotherapy.
To our knowledge, there have been no organized scientic investigations into the
use of psychopharmacology in the treatment of this disorder.
R E F E R E N C ES
Alper, K., Devinsky, O., Perrine, K., Vazquez, B. and Luciano, D. (1993). Nonepileptic seizures
and childhood sexual and physical abuse. Neurology, 43, 19503.
Arnold, L.M. and Privitera, M.D. (1996). Psychopathology and trauma in epileptic and psycho-
genic seizure patients. Psychosomatics, 37, 43843.
Binder, L.M., Kindermann, S.S., Heaton, R.K. and Salinsky, M.C. (1998). Neuropsychologic
impairment in patients with nonepileptic seizures. Arch Clin Neuropsychology, 13, 51322.
Boll, T.J. (1981). The HalsteadReitan Neuropsychology Test Battery. In Handbook of Clinical
Neuropsychology, ed. S.B. Filskov and T.J. Boll, pp. 577607. New York: John Wiley & Sons.
Bowman, S. (1993). Etiology and clinical course of pseudoseizures relationship to trauma,
depression and dissociation. Psychosomatics, 34, 33342.
Briquet, P. (1859). Trait Clinique et Thrapeutique de Lhystrie. Paris: J.B. Bailliere.
Cartmill, A. and Betts, T. (1992). Seizure behavior in a patient with post-traumatic stress disor-
der following rape. Notes on the etiology of pseudoseizures. Seizure, 1, 336.
Charcot, J.M. (1879). Lectures on the Diseases of the Nervous System Delivered at La Salpetrire.
Translated by G. Siegerson. Philadelphia: Henry O. Lea.
Goetz, C.G. (1987). Charcot the Clinician: The Tuesday Lessons. New York: Raven Press.
Gross, M. (1982). Incest and hysterical seizures. Med Hypnoanalysis, 3, 14652.
Guggenheim, F.G. and Smith, G.R. (1995). Somatoform disorders. In Comprehensive Textbook of
Psychiatry, Sixth Edition, Vol. 1, ed. H.I. Kaplan and B.J. Sadock, pp. 125170. Baltimore:
Williams & Wilkins.
Harden, C.L. (1997). Pseudoseizures and dissociative disorders: a common mechanism involv-
ing traumatic experiences. Seizure, 6, 1515.
Hardyck, C. and Petrinovich, L.F. (1977). Left-handedness. Psychol Bull, 84, 385404.
Harris, R. (ed.) (1991) Charcot, J.M.: Clinical Lectures on Diseases of the Nervous System. London:
Tavistock/Routledge.
Jarvis, P.E. and Barth, J.T. (1984). The HalsteadReitan Test Battery: an Interpretive Guide. Odessa,
FL: Psychological Assessment Resources Inc.
Kalogjera-Sackellares, D. (1995). Psychological disturbances in patients with pseudoseizures. In
Psychological Disturbances in Epilepsy, ed. J.C. Sackellares and S. Berent, pp. 191217. Boston:
ButterworthHeinemann.
Kalogjera-Sackellares, D. and Sackellares, J.C. (1997a). Personality proles of patients with
pseudoseizures. Seizure, 6, 17.
Kalogjera-Sackellares, D. and Sackellares, J.C. (1997b). Analysis of MMPI patterns in patients
with psychogenic pseudoseizures. Seizure, 6, 41927.
Kalogjera-Sackellares, D. and Sackellares, J.C. (1999). Intellectual and neuropsychological fea-
tures in patients with psychogenic pseudoseizures. Psychiatry Res, 86, 7384.
Kalogjera-Sackellares, D. and Sackellares, J.C. (2001). Impaired motor function in patients with
psychogenic pseudoseizures. Epilepsia, 42, 16006.
Lancman, M.E., Brotherton, T.A., Asconape, J.J. and Penry, J.K. (1993). Psychogenic seizures in
adults: a longitudinal analysis. Seizure, 2, 2816.
Mai, F.M. (1982). The forgotten avant-garde. Trends Neurosci, 5, 678.
Mai, F.M. and Merskey, H. (1980). Briquets treatise on hysteria. Arch Gen Psychiatry, 37, 14015.
Mai, F.M. and Merskey, H. (1981). Briquets concept of hysteria: an historical perspective. Can J
Richer, P. (1885). Etudes Cliniques sur la Grande Hystrie ou lHystropilpsie. Paris: Delahaye et
LeCrosnier.
Roy, A. (1979). Hysterical seizures. Arch Neurol, 36, 447.
Sackellares, J.C., Giordani, B., Berent, S. et al. (1985). Patients with pseudoseizures: intellectual
and cognitive performance. Neurology, 35, 11619.
Vanderzant, C.W., Giordani, B., Berent, S. Dreifuss, F.E. and Sackellares, J.C. (1986). Personality
of patients with pseudoseizures. Neurology, 36, 6648.
Westbrook, L.E., Devinsky, O. and Geocadin, R. (1998). Nonepileptic seizures after head injury.
Wilkus, R.J., Dodrill, C.B. and Thompson, P.M. (1984). Intensive EEG monitoring and psycho-
logical studies of patients with pseudoepileptic seizures. Epilepsia, 25, 1007.
Wilkus, R.J. and Dodrill, C.B. (1989). Factors aecting the outcome of MMPI and neuropsycho-
logical assessments of psychogenic and epileptic seizure patients. Epilepsia, 30, 33947.
15
Epilepsy and panic disorder

Howard A. Ring and Nuri Gene-Cos
St Bartholomews and the Royal London School of Medicine, London, UK
Introduction
Epilepsy is a neurological condition associated with well-dened abnormalities of

brain electroencephalographic (EEG) activity. It is generally treated with antiepi-
leptic agents and occasionally with resective neurosurgery. Panic disorder (PD) is a
psychiatric condition mostly treated with antidepressant medication and cogni-
tivebehaviour therapy (CBT). Why then should the conditions in any way be con-
sidered together in the same chapter? This paper will initially point out clinical
similarities and dierences and will then discuss the question of whether there are
any pathophysiological similarities between the two conditions.
Epilepsy
Epilepsy is a complex condition characterized by recurrent episodes of paroxysmal
disturbance of normal brain functioning. For a secure diagnosis of epilepsy to be
made there needs to be evidence that these paroxysms of disturbance involve both
disruptions of ongoing behaviour and abnormalities of brain electroencephalo-
graphic (EEG) activity recognized as epileptiform.
The clinical manifestations of epilepsy depend on the site of origin, or focus of
the seizure, the pattern, manner and extent of propagation of epileptiform electri-
cal activity through the brain and the aetiology of the epilepsy. From the point of
view of the person with epilepsy, the subjective experience of repeated seizures is
generally relatively similar, or stereotyped, although the actual nature of the experi-
ence is highly variable between individuals. Hence some may experience hardly any
awareness of a seizure, even if they collapse to the ground unconscious and proceed
to suer a sustained tonic-clonic convulsion. Other people with epilepsy may expe-
rience emotional experiences such as fear, unreality or dj vu which, although
perhaps not associated with any behaviours that an external observer would recog-
nize as abnormal, nevertheless may be very disruptive and unpleasant.
The current International League against Epilepsy (ILAE) classication of sei-
zures is based on EEG ndings and seizure phenomenology. The two largest groups
226
227 Epilepsy and panic disorder
of seizure types are the partial seizures, which have a focal onset, and the primary
generalized seizures, which do not have any identiable focal origin. When consid-
ering the relationship between epilepsy and PD, whether with respect to dieren-
tial diagnosis or putative commonalities in pathophysiology, it is the partial
seizures which are of most relevance.
Panic disorder
In both the ICD10 and DSMIV diagnostic classications PD is considered as an
anxiety disorder. Although these classicatory systems do not represent the last
word in mechanistic understanding of behavioural disorders, it is clear from the
inclusion of PD within the anxiety disorder/neurotic disorder grouping that the
general view is that PD has a psychological rather than a biological aetiology.
However, whilst it is clear that the core subjective experience of PD is one of
extreme fear, this does not in itself prove that the disorder is simply an extreme end
of a continuum that starts with mild anxiety.
Panic attacks may (in common with epileptic seizures) be described as paroxys-
mal events. They are discrete periods of intense fear or emotional discomfort,
accompanied by a range of somatic symptoms including palpitations, trembling, a
feeling of shortness of breath (which may be associated with hyperventilation),
sweating, feelings of choking and psychological symptoms including depersonal-
ization, fear of losing control and fear of dying. The attacks occur spontaneously,
without warning, and although they may occur in situations in which they have
previously occurred, when the patient is concerned that an event may happen, they
may also occur unexpectedly. Individual panic attacks are self-limiting although
estimates of duration vary. Retrospective estimates by suerers suggest an average
duration of between 10 and 20 minutes. However, a prospective study reported
considerably longer attacks, with a mean of between 15 and 50 minutes (Taylor et
al., 1986). Whilst some people experience attacks accompanied by most if not all of
the associated symptoms described above, in others there may be very few experi-
enced apart from paroxysmal fear and anxiety. The frequency of attacks varies, both
between individuals, and over time within individuals, from several attacks in a day
to only occasional attacks over a whole year.
The clinical diagnosis of PD is characterized by panic attacks, avoidance of sit-
uations in which previous panic attacks have occurred and ongoing worry regard-
ing the possibility of future attacks. However, these recurrent attacks of extreme
fear and a feeling of impending death or disaster are not restricted to any partic-
ular environmental setting or set of circumstances. In addition, it is important to
note that although patients with PD worry about having further panic attacks,
this worry is of a lower magnitude than the emotions experienced during an
attack.
228 H.A. Ring and N. Gene-Cos
There is good evidence, based on clinical accounts, that PD is not a homogene-

ous disorder. In some people pure PD exists with panic attacks in the absence of any
other psychopathology. However, sizeable proportions of those with PD are comor-
bid for agoraphobia or depression or both. It has been reported that women with
PD are more likely to report depression, anxiety or agoraphobic avoidance than men
with this diagnosis (Chambless and Mason, 1986). Increased anxiety and depression
in aected women was also noted by Oie et al. (1990). Whilst women may show
greater agoraphobic avoidance, men may increase their alcohol intake to cope with
their symptoms. Although these dierences in the associated symptoms and beha-
viours between the manifestations of PD in men and women exist, the core features
of panic attacks appear to be relatively similar across the sexes. Hence, although in
therapeutic terms it is clearly important to consider the whole syndrome with which
the patients present, in mechanistic terms it may be that both sexes share a common
aetiology of panic attacks, but that cultural and behavioural responses to these
attacks then determine the sex dierences in associated symptoms.
In a large American epidemiological study, the National Comorbidity Survey, it
was found that the prevalence of PD was greatest in the age range 1525 (Eaton et
al., 1994). However, separating data from males and females revealed that whilst
peak age of onset was within this range for males, for females it was older. This
study also noted that people with less than 12 years of education were up to ten
times more likely to suer from PD than those with more than 16 years of educa-
tion (i.e. including a college education).
Perhaps surprisingly, in recent years it has become clear that panic attacks are
associated with an increased risk of attempted and completed suicide (Markowitz
et al., 1989). The rate of suicide attempts is reported as 20% in those with PD and
12% in those with panic attacks but without the avoidance of panic-inducing situ-
ations and anxiety regarding future attacks that contribute to the diagnosis of PD.
In this context it is also noted that there is an increased risk of suicide and attempted
suicide in people with epilepsy, as discussed by Blumer (Chapter 8).
Potential pathophysiological commonalities
Biochemical pathophysiology
Although there are various models of neurochemical disturbance underlying
seizure genesis and also various models proposed to underlie PD, in both condi-
tions aetiological disruptions of the GABA system have been proposed.
It is well established that GABA exerts an inhibitory control on neuronal excit-
ability by its rapid action on Cl channels. The mechanisms of several antiepilep-
tic drugs, such as phenobarbitone, vigabatrin and benzodiazepines, involve
enhancing GABAergic activity in various ways.
As yet, no specic defect in GABA functioning has been identied in patients

with PD. However, GABA is indirectly implicated in PD through benzodiazepines.
Benzodiazepine receptor agonists produce neuronal inhibition via benzodiazepine
receptor modulation of GABAa receptor mechanisms, leading to, amongst other
eects, anxiolysis, muscle relaxation and sedation. Benzodiazepine antagonists
occupy the benzodiazepine receptor site without producing pharmacological
eects, while inverse agonists, such as beta-carboline, are anxiogenic and procon-
vulsant (Katz et al., 1993). Clinical data show that high potency benzodiazepine
agonists, such as alprazolam and clonazepam, have marked antipanic eects.
Moreover, other studies have suggested subsensitivity of benzodiazepine receptors
in these patients (Eison, 1990). Possibly in relation to this, umazenil, a benzodi-
azepine antagonist, has been reported to be panicogenic in patients with PD but
not in healthy controls.
Electrophysiology
A variety of abnormal electrophysiological ndings have been reported in groups
of patients with PD in comparison with healthy control subjects.
EEG studies
Patients with panic attacks have been reported to have an increased amount of par-
oxysmal EEG activity (Hughes, 1996), with this occurring up to four times more
often than is the case in patients with a depressive illness. Temporal lobe abnormal-
ities have been highlighted in brain electrical activity mapping (BEAM) studies in
patients with panic attacks.
However, other studies have failed to detect any EEG changes resembling epilep-
tiform activity in people with PD. Stein and Uhde (1989) evaluated a group of 35
medication-free patients with PD (Research Diagnostic Criteria). The EEGs were
performed over 45 to 75 minutes by using a 21-channel scalp EEG. In addition, 31
patients had an EEG performed with additional use of nasopharyngeal or anterior
temporal leads. Twenty-two patients had been sleep-deprived for 24 hours before the
EEG, and recordings were performed during drowsiness or light sleep whenever pos-
sible. In all patients, EEGs were obtained during a 2-minute period of hyperventila-
tion and in response to photic stimulation. Patients were divided into two groups: 15
with psychosensory symptoms and 20 without psychosensory symptoms.
Their results showed that EEG abnormalities of any type were infrequent, occur-
ring in a total of 5 (14%) of the 35 patients. None of these abnormalities suggested
the presence of an epileptiform disturbance but were nonspecic in nature. One
patient experienced a severe panic attack during his EEG, yet his EEG recording was
normal. Moreover, the authors found no signicant association between the pres-
ence or absence of EEG abnormalities and the presence or absence of prominent
psychosensory symptoms. However, they concluded that given the technical limi-
tations of surface EEG recordings, their ndings cannot exclude the possibility that
PD and complex partial seizures share common pathophysiological mechanisms or
sites of dysfunction. Their ndings suggest that although it is not likely that PD is
an epileptiform disorder, temporal lobe and limbic structures may play a major role
in the pathophysiology of panic.
In agreement with Stein and Uhdes work, Lepola et al. (1990) reported normal
EEG ndings in the majority of a group of 54 patients with PD who were investi-
gated using extensive EEG recordings and computerized tomography (CT) scan.
Fifteen (28%) had previously been treated for temporal lobe epilepsy or another
neurological condition. The vast majority of patients exhibited normal EEG and
CT ndings. Only in 13 (24%) patients did the EEG show increased slow-wave
activity, whilst CT scans revealed incidental abnormalities in 6 (20%) of the 30
patients so investigated. The authors commented that, although they did not use a
control group to compare the ndings with, neither the EEG nor CT showed any
focal abnormalities related to PD itself.
The situation is slightly dierent when patients with what have been described
as atypical panic attacks are studied. Edlund et al. (1987) described a series of six
patients who presented with atypical PA involving hostility, irritability, severe dere-
alization, and social withdrawal. All the patients underwent standard EEG record-
ings. None of the patients had clear temporal lobe epilepsy but most had minor and
nonspecic temporal EEG abnormalities.
Weilburg et al. (1995) studied 15 subjects who met DSM criteria for panic attack
but who also had atypical features including at least one of the following: sensory
distortions, change in level of consciousness, aphasia, focal paraesthesia, altered
sense of body position, hallucinations, sudden shifts in mood, headache or auto-
nomic changes. These subjects underwent prolonged ambulatory EEG monitoring
which included sphenoidal recordings. Eleven of their subjects were thus recorded
during the course of at least one, and in three subjects multiple, panic attacks. In
45% (5/11, including the three recorded during multiple attacks) the clinical symp-
toms were associated with focal paroxysmal EEG changes. However, even in those
who on some occasions had abnormal EEG changes associated with a panic attack,
this only occurred in a proportion (with an average of 35%) of their recorded
attacks. However, as the authors acknowledge, rst, their results may not be appli-
cable to those with panic attacks without atypical features and second, that at least
a proportion of their patients, although not meeting diagnostic criteria for epilepsy,
may nevertheless have been manifesting atypical ictal activity accounting for the
atypical (or possibly the typical) features of panic.
Overall, if these studies demonstrate anything, it is that there may be a grey area
in which some symptoms associated with panic attacks are associated with abnor-
mal EEG activity. This does not mean to say that the whole episode is driven by
electrophysiological disturbances, but it does raise the possibility that there is, in at
least a proportion of people with the symptoms of panic attack, some detectable
pathophysiological change in brain activity.
Feelings of derealization and depersonalization occur relatively frequently in
people with PD and are also accepted to occur from time to time in people with tem-
poral lobe epilepsy. Although less common in those with epilepsy, when these symp-
toms do develop they tend to be experienced as more robust phenomena.
Interestingly, there is some evidence that there are electrophysiological dierences
between those with PD whose symptoms include derealization or depersonalization
and those who do not experience these phenomena. Locatelli et al. (1993) investi-
gated computerized EEG activity derived from the temporal lobes (F7, T3, T5, F8,
T4, T6) in 30 healthy subjects and 37 patients with PD (DSMIIIR; American
Psychiatric Association, 1987) (with or without agoraphobia), in a resting condition
and also in an odour stimulation condition designed to activate temporal lobe struc-
tures. The patients with PD were divided into two groups: 17 with depersonaliza-
tion and/or derealization during their panic attacks and 20 without
depersonalization and/or derealization. Patients with PD without depersonalization
or derealization and healthy controls showed an increase of fast activity (beta 2:
1830 Hz) and a decrease of slow activities (delta 2: 24 Hz; theta 1: 46 Hz) inde-
pendent of odour stimulation. PD patients with depersonalization and/or dereal-
ization showed an increase of slow activity (delta 1: 12 Hz; delta 2: 24 Hz) and
bilateral lack of responsiveness in the fast alpha (alpha 2: 1012 Hz) frequency band
during odour stimulation. The authors suggested that the EEG changes during
odour stimulation (a temporal lobe activation task) could be interpreted as the
orienting reaction to the activating procedure; and that this appears to be dierent
depending on whether or not patients have temporal lobe symptomatology (deper-
sonalization/derealization). These ndings indicate that PD patients with temporal
symptoms respond to procedures activating their temporal regions with hypersyn-
chronization of electrical activity. The increase of delta activity can be seen as a low-
ering of the sensitivity threshold of the deep temporal regions, supporting the
hypothesis of temporal lobe involvement in patients with PD and temporal lobe
symptomatology. These ndings also demonstrate that this subset of patients with
PD have an abnormality of temporal lobe electrophysiology that in certain circum-
stances produces clinical symptoms and also, though not necessarily at the same
time, show a tendency towards abnormal synchronization of electrical activity. In
this context it is noted that hypersynchronous EEG activity may be a feature of ictal
EEG discharges. Indeed, in human temporal lobe epilepsy hippocampal hypersyn-
chronous discharges are present and may evolve into a recruiting rhythm leading to
propagation of ictal activity beyond the site of onset (Engel, 1998).
Event-related potentials
Event-related potentials (ERP) are changes in electrical brain activity that provide
a neurophysiological reection of information processing. They are derived from
averaged EEG recordings made whilst subjects undergo repeated presentations of
various stimuli in a variety of experimental paradigms. The study of ERP compo-
nents recorded from subjects whilst they perform cognitive tasks enables the assess-
ment of cerebral information processing with millisecond resolution (Pfeerbaum
et al., 1995). The P3a, occurring about 300 milliseconds after the stimulus, is asso-
ciated with the orienting of attention. It is elicited by irrelevant novel sounds in a
sequence of repetitive standard tones. It is generated by centres in the frontal lobes
and the hippocampi (Alho et al., 1998; Knight and Nakada, 1998).
It has been reported (Clark et al., 1996) that patients with PD, compared with
normal controls, have increased peak amplitude fronto-central P3a responses to all
tones (not just to novel sounds). The authors suggest that the presence of a large
P3a in PD patients might indicate an abnormal cognitive response to processes that
otherwise would have been dealt with automatically. PD patients appeared to apply
unnecessary attention to the processing of stimuli that should have been ltered out
at an earlier processing stage, engaging conscious attention unnecessarily. As P3a is
normally not seen in active attention tasks, as it is swamped by the task-related P3b,
its presence in PD would indicate, as well as an excess of stimuli processing, a failure
to reduce their response to these stimuli after repeated presentation. The P3a nor-
mally habituates with repeated stimulus presentation and it may be that there is
reduced habituation in PD. If this was to be conrmed it may explain why PD
patients became excessively aroused in environments such as crowds or supermar-
kets, where there is a high level of irrelevant stimuli. The enlarged P3a to task-
relevant stimuli is characteristic of activity that would be expected in reaction to
novel, task-irrelevant events and is consistent with specic, functional pathology
involving the prefrontallimbic pathways.
The mismatch negativity (MMN), is a relatively early ERP that is considered to
reect the earliest cortical event in cognitive processing of auditory stimuli
(Tiitinen et al., 1994), reecting the preconscious processing of unexpected audi-
tory stimuli. The main sites of MMN generation are in the superior temporal cor-
tices. It is elicited in the laboratory as a response to infrequent stimuli in sequences
of frequent homogeneous stimuli. This potential characteristically occurs about
150 milliseconds after the stimulus and can be elicited by changes in simple tones,
complex stimuli or components of speech such as phonemes (Naatanen, 1992). In
recent studies we have investigated MMN in age and sex-matched groups consist-
ing of 10 patients with panic attacks and PD, 9 patients with epilepsy and 10 normal
controls. The results are displayed in Table 15.1. It is noted that whilst MMN
parameters dier signicantly from controls in a number of sites, in general the
Table 15.1. Mismatch negativity (MMN) results in patients with epilepsy (n9), PD
(n10) and in a normal control group (n10)
MMN parameters Epilepsy group Panic group Control group
Onset latency (ms) 66 767 772

Duration (ms) 88.4 775.2b 110b
Fz (V/ms) 54a 136 180a
F4 (V/ms) 46a, c 134c 179a
Cz (V/ms) 26a 112 177a
Pz (V/ms) 32 763 110
Notes:
Group dierences (P0.05): a epilepsy vs. controls; b panic vs. controls; c epilepsy vs. panic.
The MMN was recorded from four electrode sites (Fz, F4, Cz, Pz) dened using the standard
1020 system. The data in the table demonstrate that the patients with panic disorder showed
signicantly shorter duration of the MMN potential than the control group. Considering the
MMN amplitude (measured as V/ms at Fz, F4, Cz and Pz), it is noted that the patients with
epilepsy had signicantly smaller amplitude MMNs than controls at the three more anterior
electrodes and a signicantly smaller amplitude MMN than the PD patients at the electrode
nearest the right anterior temporal region.
results for the patients with panic disorder lie between those observed in the epi-
lepsy and the control groups. These results suggest that whilst epilepsy and panic
disorder do not share the same electrophysiological abnormalities, nevertheless
there are disturbances in temporal lobe electrophysiology in patients with PD.
Structural and functional imaging

Fontaine et al. (1990) carried out MRI scans in a group of 30 patients (age between
20 and 40 years) with PD (DSMIII criteria) and 20 matched controls. All patients
had been treated with clonazepam for up to 3 months and the MRIs were done
when the patients had signicantly improved from their anxiety symptoms; more-
over, all patients took an additional 2 mg of clonazepam in the hours before the
MRI took place. In contrast, none of the controls were on clonazepam. The main
nding of this study was the increased incidence of focal abnormalities in the right
mesiotemporal area in the PD group. There were a variety of circumscribed high-
signal lesions in the white matter which were detected by the MRI as well as asym-
metric atrophy of the temporal lobes. The authors emphasized that their ndings
may be relevant to panic and phobic disorders as not only were limbic structures
involved but both the parahippocampal gyrus and the hippocampal formation play
a major role in receiving input from the association areas for all sensory modalities.
Moreover, these structures could initiate a marked defensive response through the
septo-amygdalar complex and brain stem structures. Fontaine et al. (1990) con-
cluded that although they observed an increased incidence of focal neuroanatom-
ical changes in the temporal lobes, it was unclear whether these abnormalities were
related to any genetic predisposition to PD.
Lucey et al. (1997) compared regional cerebral blood ow (rCBF), using single
photon emission tomography, in three groups of patients, 15 patients with PD and
agoraphobia, 16 patients with post-traumatic stress disorder (PTSD) and 15 patients
with obsessive compulsive disorder (OCD). Their main nding was a reduction in
caudate and superior frontal cortical perfusion in both OCD and PTSD groups com-
pared with PD and healthy controls. The caudate reduction correlated negatively
with depression (Beck Depression Inventory) and with the PTSD syndrome sever-
ity (Impact of Events scale). No dierences were found in temporal lobes.
Experimental models of anxiety

In a fMRI activation study in normal volunteers it has been demonstrated that the
amygdala is involved in conditioning and extinction of fear responses in a fashion
similar to that previously observed in experimental animals (LaBar et al., 1998).
There is evidence from studies in normal volunteers that abnormal patterns of limbic
activity may result in symptoms resembling both features of temporal lobe complex
partial seizures and features of panic attacks. In one study, intravenous injections of
procaine resulted in a range of subjective experiences including emotional, somatic
and visceral experiences often similar to those experienced in the auras of temporal
lobe epilepsy as well as resulting in the development of panic attacks in four out of
ten subjects. These experiences also included; euphoria, anxiety, depression, fear and
derealization. Positron emission tomography (PET) scanning of the subjects during
this experiment revealed that all these experiences, described as powerful and over-
whelming, were associated with increased activity in anterior limbic and paralimbic
regions (Servan-Schreiber et al., 1998). Both these authors and Ketter et al. (1996),
using a similar paradigm, noted that procaine-related activation of the left amygdala
was positively correlated with symptoms of fear and negatively correlated with feel-
ings of calmness or euphoria. In this context it is interesting to note that it is left tem-
poral more than right temporal lobe epilepsy that is more associated with the
development of negative aective states in people with epilepsy. Amygdala activation
(in this case bilaterally) has also been observed following cholecystokinin tetrapep-
tide-induced anxiety in normal volunteers (Benkelfat et al., 1995). Hence there is
good evidence that neurochemical activation of the amygdala and surrounding
structures is associated with brief but extreme experiences of panic and anxiety.
Similarly, it is well established that seizure activity may originate in the amygdala
in some patients with epilepsy. It has further been reported that seizures originat-
ing in the region of the amygdala are particularly associated with subjective emo-
tional experiences that resemble aspects of PD.
The biological significance of phenomenological similarities between PD and

epilepsy
It has been reported that ictal fear occurs in between 5 and 15% of those with tem-
poral lobe epilepsy. The experience may vary from mild uneasiness to intense feel-
ings of panic and impending doom (Devinsky and Vazquez, 1993). Fear is a
relatively common aura experience in patients with temporal lobe epilepsy, repre-
senting 10% of the experiences reported by Taylor and Lochery (1987). Williams
(1956) investigated emotional phenomena in 2000 patients with epilepsy and
found that 100 of them reported an emotion as part of the epileptic experience.
As in the study by Taylor and Lochery (1987), the most commonly reported
emotion was fear, occurring in Williams sample in 61% of the 100 patients with
emotional phenomena. On some occasions this fear was quite pervasive, with
psychic and somatic features. More recently, it has been reported that patients who
experience ictal fear as part of their temporal lobe epilepsy are more likely to have
previously suered an anxiety disorder than those who have not experienced ictal
fear (Devinsky and Vazquez, 1993).
It is well established that fear may be provoked by activity in medial temporal
structures. It is the most common experiential phenomenon produced by depth
electrode brain stimulation in antero-medial temporal regions (Halgren et al.,
1978). Hence there is evidence that abnormal electrical activity, both epileptiform
and experimental, leads to a subjective experience of intense fear of sudden onset:
a cardinal feature of panic.
In experiments with rats, LeDoux et al. (1990) found that the lateral nucleus of
the amygdala receives direct input from the sensory thalamus. By this pathway, the
amygdala is able to detect aversive input and fear-conditioned stimuli even if
sensory neocortical areas are disconnected, lesioned or ablated. As a consequence,
it can quickly and automatically elicit autonomic, endocrine and motor fear
responses even before the neocortex is able to build up a coherent representation
of the triggering threat stimulus. Moreover, it transmits an alarm signal to the neo-
cortex, which causes it to allocate its attentional resources to the current sensory
input. This can be achieved, technically, by activating ascending neuromodulatory
transmitter systems such as serotonin, acetylcholine and noradrenaline (Grae et
al., 1993). These neurotransmitters are presumed to aect the signal-to-noise ratio
of neocortical processing (Robbins and Everitt, 1995), correlated with focal atten-
tion and conscious perception (Crick, 1994). LeDoux (1995) found that neocorti-
cal processing is necessary for discriminative conditioning as well as for the
extinction of conditioned fear responses and speculated that in the absence of input
from primary sensory areas, potential fear information cannot be relayed to higher
regions such as the prefrontal cortex. Thus when the activity of the amygdala is not
suciently controlled and inhibited by these more discriminative modules, it may
tend to give rise to false threat alarms.
In summary, it seems that preattentive processing of potential threat tends to
elicit false threat alarms that automatically activate both physiological fear
responses as well as attention directed to the current sensory input (leading to con-
scious perception and analysis). However, in the intact brain, both of these eects
can be modied and extinguished by more discriminative and more elaborated
modes of processing (Windmann, 1998).
Integrating the previous models into a theory, Windmann (1998) proposes what
he calls the false-alarm-theory of PD. Symptoms of panic and pathological states
of anxiety arise from the hyperfunctioning of a preattentive alarm system whose
structural basis is closely related to the amygdala and its connections to ascending
neuromodulatory transmitter systems. The hyperfunction results in a tendency to
signal potential threat to the neocortex which is not adequately modied by more
elaborated processing in patients with PD.
Conclusions
Some studies have found temporo-limbic abnormalities in subsets of people with

PD or panic attacks. In general, there is no evidence of a specic relationship
between epilepsy, more specically complex partial seizures, and PD; indeed, in PD
most of the EEG, structural and functional ndings are nonspecic. However, there
clearly is an overlap of phenomenology between the two conditions. A possible
explanation is that PD is associated with disturbances of temporo-limbic function,
but that these disturbances are not epileptiform or epileptogenic in nature. The
clinical similarities between aspects of PD and complex partial seizures, of tem-
poral lobe origin in particular, arise by virtue of the similarity in aected brain
structures and the functions of these structures. If there is an overrepresentation of
people with both temporal lobe epilepsy and PD, then it may be that both arise,
though independently, from a common pathophysiological precursor.
R E F E R E N C ES
Alho, K., Winkler, I., Escera, C. et al. (1998). Processing of novel sounds and frequency changes
in the human auditory cortex: magnetoencephalographic recordings. Psychophysiology, 35,
21124.
and Related Health Problems (Third Edition, Revised) (DSMIIIR). Washington, DC: APA.
Barraclough, B. (1981). Suicide and epilepsy. In Epilepsy and Psychiatry, ed. E.H. Reynolds and
M.R. Trimble, pp. 726. Edinburgh: Churchill Livingstone.
Benkelfat, C., Bradwejn, J., Meyer, E. et al. (1995). Functional neuroanatomy of CCK4-induced
anxiety in normal healthy volunteers. Am J Psychiatry, 152, 11804.
Chambless, D.L. and Mason, J. (1986). Sex, sex-role stereotyping and agoraphobia. Behav Res
Ther, 24, 2315.
Clark, C.R., McFarlane, A.C., Weber, D.L. and Battersby, M. (1996). Enlarged frontal P300 to
stimulus change in panic disorder. Biol Psychiatry, 39, 84556.
Crick, F. (1994). The Astonishing Hypothesis: The Scientic Search for the Soul. New York: Simon
& Schuster.
Devinsky, O. and Vazquez, B. (1993). Behavioural changes associated with epilepsy. Neurol Clin,
11, 12749.
Eaton, W.W., Kessler, R.C., Wittchen, H.U. and Magee, W.J. (1994). Panic and panic disorder in
the United States. Am J Psychiatry, 151, 41320.
Edlund, M.J., Swann, A.C. and Clothier, J. (1987). Patients with panic attacks and abormal EEG
results. Am J Psychiatry, 144, 5089.
Eison, M.S. (1990). Serotonin: a common neurobiologic substrate in anxiety and depression. J
Clin Psychopharmacol, 10 (Suppl.), 2630.
Engel, J. Jr (1998). Research on the human brain in an epilepsy surgery setting. Epilepsy Res, 32,
111.
Fontaine, R., Breton, G., Dry, R., Fontaine, S. and Elie, R. (1990). Temporal lobe abnormalities
in panic disorder: an MRI study. Soc Biol Psychiatry, 27, 30410.
Grae, F.G., Silveira, M.C., Nogueira, R.L., Audi, E.A. and Oliveira, R.M. (1993). Role of the
amygdala and periaqueductal gray in anxiety and panic. Behav Brain Res, 58, 12331.
Halgren, E., Walter, R.D., Cherlow, D.G. and Crandall, P.H. (1978). Mental phenomena evoked by
electrical stimulation of the human hippocampal formation and amygdala. Brain, 101, 83117.
Hughes, R.N. (1996). Drugs which induce anxiety: caeine. New Zealand J Psychol, 25, 3642.
Katz, R.J., Landau, M., Lott, A. et al. (1993). Serotonergic (5HT), medication of anxiety: thera-
peutic eects of serazepine in generalised anxiety disorder. Biol Psychiatry, 34, 414.
Ketter, T.A., Andreason, P.J., George, M.S. et al. (1996). Anterior paralimbic mediation of
procaine-induced emotional and psychosensory experiences. Arch Gen Psychiatry, 53, 5969.
Knight, R.T. and Nakada, T. (1998). Cortico-limbic circuits and novelty: a review of EEG and
blood ow data. Rev Neurosci, 9, 5770.
LaBar, K.S., Gatenby, J.C., Gore, J.C., LeDoux, J.E. and Phelps, E.A. (1998). Human amygdala
activation during conditioned fear acquisition and extinction: a mixed-trial fMRI study.
Neuron, 20, 93745.
LeDoux, J.E. (1995). Emotion: clues from the brain. Ann Rev Psychol, 46, 20935.
LeDoux, J.E., Cicchetti, P., Xagoraris, A. and Romanski, L.M. (1990). The lateral amygdaloid
nucleus: sensory interface of the amygdala in fear conditioning. J Neurosci, 10, 10629.
Lepola, U., Nousiainen, U., Puranen, M., Riekkinen, R. and Rimn, R. (1990). EEG and CT nd-
ings in patients with panic disorder. Biol Psychiatry, 28, 7217.
Locatelli, M., Bellodi, L., Perna, G. and Scarone, S. (1993). EEG power modications in panic dis-
order during a temporolimbic activation task: relationships with temporal lobe clinical symp-
tomatology. J Neuropsychiatry Clin Neurosci, 5, 40914.
Lucey, J.V., Costa, D.C., Adshead, G. et al. (1997). Brain blood ow in anxiety disorders. Br J
Markowitz, J., Weissman, M.M., Ouellette, R. et al. (1989). Panic disorder and quality of life. Arch
Gen Psychiatry, 46, 98492.
Mathews, W.S. and Barabas, G. (1981). Suicide and epilepsy: a review of the literature.
Psychosomatics, 22, 51524.
Naatanen, R. (1992). Attention and Brain Function. Hillsdale, NJ: Lawrence Erlbaum Associates.
Oei, T.P.S., Wanstall, K. and Evans, L. (1990). Sex dierences in panic disorder with agorapho-
bia. J Anxiety Disord, 4, 31724.
Pfeerbaum, A., Roth, W.T. and Ford, J.M. (1995). Event-related potentials in the study of psy-
chiatric disorders. Arch Gen Psychiatry, 52, 55963.
Robbins, T.W. and Everitt, B.J. (1995). Arousal systems and attention. In The Cognitive
Neurosciences, ed. M. Gazzaniga, pp. 70320. Cambridge, MA: MIT Press.
Servan-Schreiber, D., Perlstein, W.M., Cohen, J.D. and Mintun, M. (1998). Selective pharmaco-
logical activation of limbic structures in human volunteers: a positron emission tomography
study. J Neuropsychiatry Clin Neurosci, 10, 14859.
Stein, M.B. and Uhde, T.W. (1989). Infrequent occurrence of EEG abnormalities in panic disor-
der. Am J Psychiatry, 146, 51720.
Taylor, C.B., Sheikh, J., Agras, W.S. et al. (1986). Ambulatory heart rate changes in patients with
panic attacks. Am J Psychiatry, 143, 47882.
Taylor, D.C. and Lochery, M. (1987). Temporal lobe epilepsy; origin and signicance of simple
and complex auras. J Neurol Neurosurg Psychiatry, 50, 67381.
Tiitinen, H., May, P., Reinikainen, K. and Naatanen, R. (1994). Attentive novelty detection in
humans is governed by pre-attentive sensory memory. Nature, 372, 9092.
Weilburg, J.B., Schachter, S., Worth, J. et al. (1995). EEG Abnormalities in patients with atypical
panic attacks. J Clin Psychiatry, 56, 35862.
Williams, D. (1956). The structure of emotions reected in epileptic experiences. Brain, 79,
2967.
Windmann, S. (1998). Panic disorder from a monistic perspective: integrating neurobiological
and psychological approaches. J Anxiety Disord, 12, 485507.
Part V
Treatment complications
16
The effects of antiepileptic drugs on

behaviour
Bettina Schmitz
Humboldt University, Berlin, Germany
Introduction
All antiepileptic drugs (AEDs) may have eects on thinking, mood and behaviour
in individual patients. These psychotropic eects are not simply idiosyncratic but
depend on the drugs anticonvulsive strength and the persons biological and
psychological predisposition.
If a psychiatric disorder occurs in a patient with epilepsy this always has a
multifactorial aetiology, anticonvulsant pharmacotherapy being only one of
many risk factors. Because of the complex pathogenesis of psychiatric complica-
tions in epilepsy, aetiology-related epidemiological data are dicult to obtain,
particularly since chronic eects of anticonvulsants are almost impossible to
identify. Among a series of consecutive patients who developed either a schizo-
phreniform psychosis or a major depression, peri-ictal syndromes predominated
in psychotic patients, and interictal syndromes in depressed patients (Schmitz et
al., 1999). Twenty-eight per cent of depressive episodes and 15% of psychotic epi-
sodes were attributed to drug treatment, including alternative syndromes, intox-
ication and withdrawal syndromes. However, these gures may not be
representative for todays praxis because the cases were largely collected prior to
the introduction of new AEDs.
In a more recent series from Japan, the percentage of AED-related psychoses was
signicantly higher; 40% of unselected cases had psychoses following a change of
their AED regime (half of which occurred with the new anticonvulsant zonisamide,
which is not as yet licensed in Europe (Matsuura, 1999)).
The psychotropic eects of AEDs can be broken down into those which are nega-
tive and those which are positive or prophylactic. The major psychotropic eect of
AEDs is via their antiepileptic properties. Therefore, the best prophylaxis of psychiat-
ric complications is early and complete control of seizures. Unfortunately, even with
the expanding repertoire of eective AEDs there are many patients in whom complete
and lasting seizure control cannot be achieved immediately after rst manifestation
241
242 B. Schmitz
of seizures. Patients whose seizures are dicult to treat are particularly at risk to
develop psychiatric complications secondary to AED treatment.
There are (at least) six questions in the context of severe behavioural adverse
events of AEDs, most of which cannot be satisfactorily answered:
1. Drug-related incidence rates: How often and when in the course of treatment are
complications to be expected when a specic drug is given?
2. Medication-related issues: What are the dierences between monotherapy and
polytherapy; and is there a relationship with titration rates and maximum dosages?
3. Psychopathology and outcome of the psychiatric reaction: Are there specic
psychiatric syndromes; what is the prognosis, which therapeutic actions ought
to be taken?
4. Neurological, epileptological and psychiatric risk factors: How can we identify
vulnerable patients?
5. Is there a relationship with cognitive side eects; are there predictable positive
psychotropic eects?
6. What are the underlying mechanisms?
The aim of this chapter is an overview on the current knowledge on psychiatric
eects of anticonvulsants. One of the major methodological problems of studies
looking at the relationship between AED and mental state in epilepsy relate to clas-
sication and terminology. Unfortunately, diagnostic criteria for psychiatric side
eects are neither dened nor standardized. The four major categories used in the
literature are psychoses, aective syndromes, behavioural or personality disorders
and encephalopathies. From a nosological point of view these categories are
not distinct and not specic. The label psychosis may stand for a chronic and
schizophrenia-like illness, or a short lasting, transient and delirious episode, or a
psychosis in the context of a severe, depressive illness.
Data from drug trials are usually restricted to isolated psychopathological symp-
toms, such as nervousness, anxiety, depressed mood or abnormal thinking. The
clinical signicance and the broader psychiatric context of these remain unclear. An
extrapolation towards a specic syndrome like depression would be inappropriate,
but is nevertheless often done. Anxiety, for example, is a nonspecic aective
symptom occurring with almost all psychiatric syndromes.
Classical antiepileptic drugs
With respect to the classical antiepileptic drugs there are almost no systematic data
on their psychiatric side eects. Our knowledge is largely empirical, based on small
case series or anecdotal reports (for a synopsis see Table 16.1).
A number of studies have suggested a link between depression and treatment
with primidone or phenobarbital both in adults and in children (Brent, 1986; Brent
243 Effects of antiepileptic drugs on behaviour
Table 16.1. Psychotropic effects of antiepileptic drugs
Positive psychotropic Psychoses and other

eects Negative aective eects complications
Barbiturates, Aggression, depression, ADHD in children

Primidone withdrawal syndromes
Benzodiazepines Anxiolytic, sedative Withdrawal syndromes Disinhibition
Ethosuximide Insomnia Alternative psychoses
Phenytoin Toxic schizophreniform
psychoses, encephalopathy
Carbamazepine Mood stabilizing, Rarely mania and
impulse control depression
Valproate Mood stabilizing, Acute and chronic
antimanic encephalopathy
Vigabatrin Aggression, depression, ADHD, encephalopathy,
psychosis withdrawal alternative psychoses
syndromes
Lamotrigine Mood stabilizing, Insomnia Rarely psychoses
antidepressive
Felbamate Stimulating? Agitation Psychoses possible
Gabapentin Anxiolytic, Rarely aggression in
antidepressive? children
Tiagabine Depression Nonconvulsive status
epilepticus
Topiramate Mood stabilizing ?? Depression Psychoses
Levetiracetam
Notes:
? represents minimal data. , not applicable; ADHD, attention-decit hyperactivity disorder.
et al., 1987; Robertson et al., 1987). In children a conduct disorder resembling the
attention decit hyperactivity disorder may be provoked by many antiepileptic
drugs, the most frequently implicated drugs being, however, the barbiturates.
Irritability and aggressive behaviour are side eects particularly often seen in men-
tally handicapped patients.
Phenytoin may provoke schizophrenia-like psychoses at high serum levels
(McDanal and Bolman, 1975). These psychoses are dose related, thus toxic syn-
dromes, but they are not associated with cerebellar symptoms which are the most
244 B. Schmitz
common central nervous side eects of phenytoin. A chronic encephalopathy has

also been described with phenytoin and has been referred to as Dilantin dementia
(Trimble and Reynolds, 1976).
Psychoses typically following cessation of seizures and associated with a normal-
ization of the EEG occur in 2% of children treated with ethosuximide. The risk for
forced normalization is higher (8%) in adolescents and adults treated with etho-
suximide for persisting absence seizures (Wolf et al., 1984).
Aective problems are rare complications of treatment with carbamazepine
(Dalby, 1975). These are either depressive disorders or mania, the latter being
explained as a paradoxical eect due to the antidepressant properties of carbamaz-
epine which is chemically related to tricyclic antidepressants (Drake and Peruzzi,
1986).
Rarely, valproate is associated with acute or chronic encephalopathies
(Sackellares et al., 1979; Schndienst and Wolf, 1992; Zaret and Cohen, 1986).
These encephalopathies are related to dose and perhaps polytherapy and are rever-
sible with dose reduction.
New antiepileptic drugs
In Europe, eight new antiepileptic drugs have been introduced in the last decade.
With respect to the psychiatric side eects of these new AEDs, data exist particu-
larly from premarketing studies. However, drug trials are designed to test antiepi-
leptic ecacy and psychiatric adverse events are not systematically reported. Thus
severity and psychopathological nature of behavioural problems remain obscure.
Further, dierences in patients included in trials do not allow comparisons of
psychiatric risks of specic drugs, particularly since following the early vigabatrin
experience patients with a psychiatric history were often excluded from trials.
Finally, these data are not always entirely transparent to the interested epileptolo-
gist, at least not as soon and to an extent as one might wish.
There are three aspects which need to be dierentiated in the analysis of the
psychotropic proles of the new AEDs in epilepsy: the rst relates to psychiatric
side eects of new AEDs when used in epilepsy patients, the second to the evidence
for positive psychotropic eects when these drugs are used in psychiatric patients,
and the third relates to data on positive psychotropic eects of new AEDs when
used in epilepsy patients.
Psychiatric side effects of new antiepileptic drugs

Vigabatrin
Of the newer agents vigabatrin has been the most studied, perhaps on account of
its being the rst to be tested and licensed. Shortly after the introduction of viga-
batrin, a London group had published an incidence of signicant psychiatric com-

plications in 7% of treated patients (Sander et al., 1991). Thomas et al. (1996) have
analysed case records of psychiatric complications, episodes of psychoses or major
depression, reported to the manufacturer of vigabatrin. With respect to psychoses
the authors identied three patterns: of a total of 28 psychotic patients, eleven had
become seizure free with vigabatrin, six had a postictal psychosis following a cluster
of seizures after initial seizure control, possibly related to tolerance, and two psy-
choses occurred after withdrawal of vigabatrin.
In children, particularly in association with learning disabilities, the most common
psychiatric side eects are agitation and excitation, hyperkinesia and aggression, a
behaviour syndrome similar to that seen with barbiturates. In an early French study
the incidence of behavioural disturbances was as high as 26% (Dulac et al., 1991).
Since these early reports, the clinical signicance of vigabatrin-associated behav-
ioural problems has been a matter of controversy, prompting Ferrie to perform a
meta-analysis of psychoses and severe behavioural reactions leading to drug dis-
continuation, in seven placebo-controlled European studies (Ferrie et al., 1996).
The overall incidence of these complications was 3.4% in the vigabatrin group and
0.6% in the placebo group. Although this is a signicant dierence in statistical
terms, the authors conclude that the risk for psychiatric complications is not
increased with vigabatrin referring to published incidence rates of psychiatric dis-
orders in epilepsy patients in general. Remarkably, the incidence rates seem to be
rather dierent in dierent studies, ranging from 1% to 12% suggesting that either
the risk is not the same for all patient groups or that the threshold to report psychi-
atric side eects is not the same among dierent investigators.
Another meta-analysis on the risk for psychosis has only recently been published
(Levinson and Devinsky, 1999). These authors used a uniform dictionary and
translated psychopathological symptoms described in the investigator forms into
standardized psychiatric terminology, which was then summarized into a syndro-
matic diagnosis. This analysis of American and non-American double-blind
studies demonstrated that there is a signicantly increased risk for psychosis and
particularly for depression. Psychoses occurred in 2.5% of patients treated with
vigabatrin compared to an incidence of 0.3% in the placebo group (P < 0.05), and
depression occurred in 12.1% of patients treated with vigabatrin in contrast to only
3.5% in the placebo group (P < 0.001). Less than 2% of the patients were taken o
vigabatrin suggesting that these were relatively mild manifestations of depression
in the majority of patients. It is however important to recognize that the duration
of the analysed studies was 34 months, later complications are therefore not
included. It is a pity that this useful analysis only came out in 1999, since most epi-
leptologists have almost stopped using vigabatrin because of the risk for visual eld
defects since 1998.
246 B. Schmitz
Psychiatric side eects are not restricted to patients with complicated epilepsies
who receive vigabatrin as an add-on treatment. The monotherapy trials showed a
signicantly increased incidence of depressive disorders in 5% of vigabatrin-
treated patients as compared to only 1% in a carbamazepine-treated group
(Chadwick, 1999).
Lamotrigine
In contrast to vigabatrin, lamotrigine has early on gained a reputation of having
positive psychotropic properties, improving both mood and cognitive functions.
Severe psychiatric complications seem to be uncommon with lamotrigine, and
psychosis and depression occurred only in very few cases in the trials (Fitton and
Goa, 1995).
Insomnia, which may be associated with irritability, anxiety or even hypomania,
is the only signicant psychiatric side eect, occurring in 6% of patients treated
with lamotrigine in monotherapy, compared to 2% in patients treated with carba-
mazepine and 3% in patients treated with phenytoin (Brodie et al., 1995).
When there were reports that carers complained that handicapped patients
became more alert and demanding, this was interpreted as reecting inadequate
rehabilitation facilities rather than being a negative side eect (Binnie, 1997). Besag
refers to this as a release phenomenon (Besag, 2001). There are now, however, a
number of reports that children with learning diculties and adults with mental
handicap develop behavioural problems such as aggression with lamotrigine
(Beran and Gibson, 1998; Ettinger et al., 1998). More recently there have been
reports on the induction of a reversible Gilles de la Tourette syndrome, which in
some cases was accompanied by obsessivecompulsive symptoms (Lombroso,
1999).
Felbamate
Felbamate is at present only used in a minority of patients, particularly with
LennoxGastaut syndrome, due to its haematologic and hepatic toxicity. According
to the manufacturer psychoses are rare, reported as serious adverse events in 0.02%
of all patients treated in 1996 (Essex Pharma, personal communication). Felbamate
may lead to increased alertness, inducing sleep problems and behavioural problems
related to agitation in some patients, again, particularly in children with learning
disabilities (McConnell et al., 1996).
Ketter et al. (1999) have specically investigated psychotropic eects of felba-
mate. They concluded from their study of 30 refractory epilepsy patients that the
stimulating eects of felbamate may be benecial or negative depending on preex-
isting psychopathology. Patients with baseline insomnia or anxiety experienced a
deterioration in their psychic state, while other children improved.
Gabapentin
Beyond somnolence, negative psychotropic eects have not been demonstrated in
the controlled studies of gabapentin, which is a generally well-tolerated but also a
relatively weak anticonvulsant. However, there are a number of studies suggesting
that gabapentin may induce behavioural problems such as aggression in children
with learning disabilities and adults with mental handicap (Lee et al., 1996; Tallian
et al., 1996; Wolf et al., 1995). It is not clear whether this could be related to rapid
titration.
Tiagabine
A specic problem with tiagabine is the paradoxical provocation of de novo non-
convulsive status epilepticus due to a relatively narrow therapeutic window
(Schapel and Chadwick, 1996). Therefore, EEG investigations are necessary when
behavioural problems arise. Unfortunately, this complication was discovered fol-
lowing the initial trials and it is therefore not possible to know whether psychiatric
complications in the trials were related to underlying epileptic activity or not.
In the placebo-controlled add-on studies nervousness and depressed mood were
both increased in the tiagabine group (Leppik, 1995) (12% vs. 3%, 5% vs. 1%). The
incidence for serious adverse events presenting as psychosis was not signicantly
increased in the tiagabine group (2% vs. 1%). A total of 84 psychoses had been
reported to the manufacturer in 1996. In 30 patients tiagabine was withdrawn, in
38 patients tiagabine was reduced, and in 16 cases tiagabine was continued at an
unchanged dosage.
The records of 19 patients with psychoses classied as serious adverse events have
been further analysed (Schmitz, unpublished data). Psychoses occurred after a var-
iable duration of treatment with tiagabine, with a mean of 267 days (range 13606
days). The mean dosage was 48 mg/day (range 880 mg). Seven psychoses occurred
postictally. Only one patient had an alternative psychosis following seizure control
and one patient became psychotic following tiagabine withdrawal. There was no
systematic EEG monitoring in these cases, so nonconvulsive status cannot be
excluded and may have gone unrecognized in some cases. The psychoses were of a
paranoid-hallucinatory type in 12 cases and duration was less than a week in 10,
and less than a month in 4 cases. In summary, for psychoses with tiagabine there
are no distinguishable patterns, as described for vigabatrin.
Topiramate
Topiramate, promising polytherapy with a single drug because of three dierent
modes of action, is a highly eective anticonvulsant, working against a broad spec-
trum of seizure types. Topiramate also has a high rate of reported side eects. These
might in part relate to rapid titration schemes in the earlier studies. However, there
248 B. Schmitz
is no evidence so far that severe psychiatric complications can be avoided by slow

titration. An unusual idiosyncratic side eect of topiramate is amnesic or motor
aphasia, and in the controlled trials 1728 % of patients developed symptoms clas-
sied as abnormal thinking (Janssen-Cilag, 1996).
The rate of aective symptoms is clearly dose-dependent with an incidence of
9% and 19% with a daily dose of 200 mg and 1000 mg respectively in one clinical
study (Janssen-Cilag, 1996). The high incidence of depressive syndromes with
topiramate may be related to cognitive side eects, which are particularly common
with this drug. But this relationship has not been studied.
Trimble et al. (2000) have looked at the clinical data from a series of patients with
psychoses and depression related to topiramate. In this analysis there was no spe-
cic pattern for the precipitation of psychoses. Interestingly, depressive syndromes
were related to a complete control of seizures in ve cases. They also suggested that,
in contrast to vigabatrin, at least some of these psychoses were associated with high
serum levels, and possible intoxication.
Levetiracetam
Levetiracetam is the latest drug, which only recently has been launched in Europe.
Data on psychiatric eects of levetiracetam are limited but a preliminary analysis
suggests that there are no signicant psychiatric risks associated with this drug
(Trimble, 2000). Aective disorders were reported in 0.02% and psychosis in
0.007% of patients treated with levetiracetam in clinical trials.
Mechanisms
There are a number of theoretical mechanisms linking antiepileptic drugs and
psychiatric disorders. These are: (1) dose-related drug toxicity, (2) dose-unrelated
or idiosyncratic eects in vulnerable individuals, (3) drug withdrawal and (4)
eects related to antiepileptic ecacy (forced normalization). The most impor-
tant mechanisms both from an epidemiological and a theoretical point of view are
idiosyncratic side eects and alternative syndromes associated with the phenome-
non of forced normalization.
Idiosyncratic, dose-unrelated effects

The mode of action is more or less known for the new substances, if this can be reliably
concluded from animal models. In a simplistic scheme, the major antiepileptic mech-
anisms are either via the sodium channel or GABAergic or antiglutamatergic. It has
been pointed out by Trimble that psychiatric problems and in particular depressive
disorders are signicantly increased with those AEDs which have strong GABAergic
properties: vigabatrin, tiagabine and topiramate (Trimble, 1997); all three drugs had
an increased rate of depressive symptomatology in placebo-controlled trials.
This observation is not easy to explain but has been used as further evidence for
the GABA hypothesis of depression. A number of clinical observations and experi-
mental studies have shown that GABAergic mechanisms are involved in the patho-
genesis of depression (Petty, 1995).
Trimbles hypothesis of a link between psychiatric complications and GABAergic
mechanisms of AEDs has been extended by Ketter et al. (1994). These authors dis-
tinguish two categories of AED, the rst being GABAergic with sedating, anxiolytic
and antimanic properties. This category comprises barbiturates, benzodiazepines,
valproate, vigabatrin, tiagabine and gabapentin. The second category comprises
antiglutaminergic drugs which are claimed to have activating, anxiogenic and anti-
depressive eects: felbamate and lamotrigine. In this scheme topiramate holds an
intermediate position because of its multiple mechanisms. The authors suggest that
anticonvulsant drugs have dierent psychiatric eects depending on the preexist-
ing mental status of patients. They predict that patients who are primarily activated
may benet from drugs which belong to the sedating category and become worse
with activating drugs. On the other hand, patients who are primarily sedated
would benet from a drug from the activating category, while the same patients
would worsen with a sedating anticonvulsant. Although this scheme might repre-
sent an oversimplication, taking the primary psychopathological status of
patients into account explains the sometimes unexpected and seemingly paradox-
ical eects of some AED in individual patients. Therefore, the consideration of the
patients preexisting mental state beyond the epileptic syndrome when choosing an
AED is a useful approach which deserves further study.
Forced normalization
The concept of forced normalization goes back to the publications of Heinrich
Landolt, head of the Swiss epilepsy centre in Zurich from 1955 until 1971 (Landolt,
1958). Cases of forced normalization or alternative psychoses have been reported
with all of the novel drugs but seem to be particularly common with vigabatrin.
There are a number of reports with topiramate, very few with tigabine and lamot-
rigine, and only one case with gabapentin (Blumer, personal communication).
There seems to be a link between the incidence of these alternative syndromes with
a number of drugs which happen to be more ecacious than others, when accept-
ing the results of the meta-analysis by Elferink and Van Zwieten Boot (1997). These
authors analysed drug trials and compared AEDs by looking at the number of
patients which are needed to be treated in order to nd one responder. According
to this analysis, vigabatrin and topiramate hold relatively good positions while gab-
apentin and lamotrigine appear to be less eective.
The phenomenon of forced normalization is not restricted to drug-induced
seizure control. It is likely that in patients who develop de novo psychosis following
250 B. Schmitz
epilepsy surgery, forced normalization plays a role, and recently a rst case of an
alternative psychosis secondary to vagus nerve stimulation has been published
(Gatzonis et al., 2000).
Positive psychotropic effects of antiepileptic drugs in psychiatric patients

The positive psychotropic properties of carbamazepine and valproate are well
established. Both anticonvulsants are frequently used in psychiatric patients.
Carbamazepine is indicated for the prophylaxis of bipolar disorder and the man-
agement of episodic dyscontrol, and valproate is particularly useful in the treat-
ment of acute mania (Walden et al., 1998).
The antidepressive eects of lamotrigine, which were suspected shortly after its
introduction, have been conrmed in controlled studies of patients with bipolar
and rapid cycling aective disorder (Calabrese et al., 1999; Kasumakar and Yatham,
1997). However, in its present formulation, lamotrigine is not useful in acute
psychiatric disorders because of the slow titration rate.
Gabapentin is increasingly being prescribed for an almost unlimited spectrum
of psychiatric disorders (Letterman and Markowitz, 1999). This is largely based on
positive case reports or small open studies. Controlled studies have established e-
cacy as yet only in subforms of anxiety disorders (social phobia) (Pande et al.,
1999).
Tiagabine has been used in only few patients with psychiatric problems, result-
ing in mixed, but mostly negative, results (Grunze et al., 1999). More promising are
the preliminary experiences regarding topiramate. This drug is currently being
intensively tested for various psychiatric indications (Marcotte, 1998). Data from
open studies in mania and depression have shown positive results (Normann et al.,
1999). As yet, there are no data on eects of levetiracetam in psychiatric disorders.
Positive psychotropic effects of AEDs in epilepsy patients

Antiepileptic drugs are increasingly being marketed for a broad spectrum of
psychiatric disorders. Unfortunately we cannot extrapolate from studies per-
formed with primary psychiatric patients that there are positive psychotropic
eects when these drugs are prescribed for epilepsy patients. Most studies in psy-
chiatry are done with patients suering from bipolar disorder, and this is a rela-
tively uncommon diagnosis in patients with epilepsy. As an example, the
well-established mood-stabilizing eects of carbamazepine have never been dem-
onstrated in epilepsy.
Given the high incidence of psychiatric comorbidity in epilepsy it would be very
useful to know whether antiepileptic drugs have a positive inuence on the psychic
status of epilepsy patients beyond their inuence on seizure activity. However, there
is hardly any scientic evidence for this. The methodological problems of four
studies which have looked at psychotropic eects of AED in epilepsy will be dis-
cussed below.
Ketter et al. (1994) prospectively investigated psychopathology in 32 patients
who were openly withdrawn from all antiepileptic drugs. Thirty-eight per cent
developed moderate to severe psychopathology, the most prominent symptoms
being anxiety and depression. According to the authors, and the complex statistics,
this withdrawal emergent psychopathology was not fully explained by an increase
in seizures, demographic factors or psychiatric history suggesting pharmacody-
namic eects following drug discontinuation. This study is often quoted as evi-
dence for positive psychotropic eects of AEDs. However, the patients were
withdrawn from all AEDs within a short period of time in order to start a felba-
mate monotherapy trial, and were in a very specic anxiety-provoking situation.
Therefore, these data should not be used to claim positive psychotropic eects of
anticonvulsants.
As mentioned before, the improvements in mood and alertness related to lamot-
rigine treatment are claimed to be independent from seizure control and have been
described in many studies. Most of these studies are however uncontrolled, and the
controlled data are dicult to interpret. For example, in the early quality of life
study by Smith et al. (1993), patients improved only on two psychiatric scales (hap-
piness and alertness), but not on the other four scales which measured self-esteem,
mood, anxiety and depression.
In another study, the eects of lamotrigine were compared to carbamazepine
after 4 weeks of treatment (Gilham et al., 1996) resulting in signicant dierences
on a self-report questionnaire with respect to psychological parameters such as dys-
phoria and worries. Patients treated with lamotrigine showed signicantly better
results on these scores than patients in the carbamazepine group. However, given
the slow titration rate of lamotrigine in the rst 4 weeks of treatment, this result
probably reects the adverse eect prole of carbamazepine rather than positive
psychotropic eects of lamotrigine.
A study by Harden et al. (1999) claims to demonstrate that gabapentin has a ben-
ecial eect on mood in epilepsy. However, the evidence for this is very weak; the
design of this study is open, and patients only improved on one scale, a clinician-
rated questionnaire on dysthymia (Cornell), while patients did not improve on
self-rating mood scales such as the Beck Depression Inventory. (At least, this weak
positive psychotropic eect of gabapentin is likely to be independent from antiepi-
leptic eects, since there were no signicant dierences with respect to seizure
control compared to placebo.)
In summary, there is clearly a need for further, well-designed studies (prospec-
tive, double-blind and placebo-controlled) looking specically at the psychotropic
eects of anticonvulsants in epilepsy patients.
252 B. Schmitz
Summary and clinical recommendations
The risk for psychiatric complications of AEDs is likely to be linked to severity of

epilepsy, polytherapy, rapid titration and high dosages of drugs. Patients with pre-
vious psychiatric problems or a familial predisposition seem to be specially prone
to behavioural side eects. It is important to recognize patients at risk in order to
inform them and their families about the possibility of psychiatric side eects, in
order to apply a careful titration scheme and to make sure that patients are seen fre-
quently. Psychiatric complications are mild and reversible in most cases, when rec-
ognised at an early stage. Risk factors for psychiatric complications are not a strict
contraindication for any particular drug, and it is not always necessary to com-
pletely withdraw the responsible drug. Depending on the pathophysiology and the
severity of the syndrome, a dose reduction or a comedication with a neuroleptic or
antidepressive drug may be a good compromise.
Psychiatric disorders in epilepsy have a multifactorial aetiology, pharmacotherapy
being only one of many risk factors which are both biological and psychosocial.
Among psychiatric adverse events of anticonvulsants, a variety of nonpsychotic
behavioural problems are reported most commonly, followed by aective disorders,
and psychosis being a relatively rare though severe complication. Psychotropic eects
of anticonvulsants warrant further research because many relevant parameters
related to pathomechanism, frequency, psychopathology and prognosis are not
known.
Behavioural side-eect proles, both negative and positive psychotropic eects,
ought to be considered in the choice of the optimal antiepileptic drug for an indi-
vidual patient. Despite the extensive existing literature covering this topic, there is
a need for more studies specically devoted to the psychiatric eects of AEDs in
patients with epilepsy. We need these studies in order to better identify patients at
risk for severe behavioural reactions with specic drugs and also in order to iden-
tify patients who have a good chance to benet from potentially positive psycho-
tropic eects of AEDs.
R E F E R E N C ES
Beran, R.G. and Gibson, R.J. (1998). Aggressive behaviour in intellectually challenged patients
with epilepsy treated with lamotrigine. Epilepsia, 39, 2802.
Besag, F.M.C. (2001). Behavioural eects of the new anticonvulsants. Drug Safety, 24, 51336.
Binnie, D.B. (1997). Lamotrigine. In Epilepsy. A Comprehensive Textbook, ed. J. Engel and T.A.
Pedley, pp. 153140. Philadelphia, New York: LippincottRaven.
Brent, D.A. (1986). Overrepresentation of epileptics in a consecutive series of suicide attempters

seen at a childrens hospital, 19781983. J Am Acad Child Psychiatry, 25, 2426.
Brent, D.A., Crumrine, P.K., Varma, R.R., Allan, M. amd Allman, C. (1987). Phenobarbital treat-
ment and major depressive disorder in children with epilepsy. Pediatrics, 80, 90917.
Brodie, M.J., Richens, A. and Yuen, A.W. (1995). Double-blind comparison of lamotrigine and
carbamazepine in newly diagnosed epilepsy. UK Lamotrigine/Carbamazepine Monotherapy
Trial Group. Lancet, 345, 4769.
Calabrese, J., Bowden, C., Sachs, G., Ascher, J., Monaghan, E. and Rudd, D. (1999). A double
blind placebo controlled study of lamotrigine monotherapy in outpatients with bipolar I
depression. J Clin Psychiatry, 60, 7988.
Chadwick, D. (1999). Safety and ecacy of vigabatrin and carbamazepine in newly diagnosed
epilepsy: a multicentre randomised double-blind study. Vigabatrin European Monotherapy
Study Group. Lancet, 354, 1319.
Dalby, M.A. (1975). Behavioral eects of carbamazepine. In Complex Partial Seizures and their
Treatment. Advances in Neurology, Vol. 11, ed. J.K. Penry and D.D. Daly, pp. 33143. New York:
Raven Press.
Drake, M.E. and Peruzzi, W.T. (1986). Manic state with carbamazepine therapy of seizures. J Natl
Med Assoc, 78, 11057.
Dulac, O., Chiron, D., Cusmai, R., Pajot, N., Beaumont, D. and Mondragon, S. (1991). Vigabatrin
in childhood epilepsy. J Child Neurol, 6 (Suppl. 2), 307.
Elferink, J.A. and Van Zwieten Boot, B.J. (1997). New antiepileptic drugs. Analysis based on
number needed to treat shows dierences between drugs studied. Br Med J, 314, 603.
Ettinger, A.B., Weisbrot, D.M., Saracco, J., Dhoon, A., Kanner, A. and Devinsky, O. (1998).
Positive and negative psychotropic eects of lamotrigine in patients with epilepsy and mental
retardation. Epilepsia, 39, 8747.
Ferrie, C.D., Robinson, R.O. and Paniotopoulos, C.P. (1996). Psychotic and severe behavioural
reactions with vigabatrin: a review. Acta Neurol Scand, 93, 18.
Fitton, A. and Goa, K.L. (1995). Lamotrigine. Drugs, 50, 691713.
Gatzonis, S.D., Stamboulis, E. and Siafakas, A. (2000). Acute psychosis and EEG normalisation
after vagus nerve stimulation. J Neurol Neurosurg Psychiatry, 69, 2789.
Gillham, R., Baker, G., Thompson, P. et al. (1996). Standardisation of a self-report questionnaire
for use in evaluating cognitive, aective and behavioural side-eects of anti-epileptic drug
treatments. Epilepsy Res, 24, 4755.
Grunze, H., Erfurth, A., Marcuse, A., Amann, B. and Walden, J. (1999). Tiagabine appears not to
be ecacious in the treatment of acute mania. J Clin Psychiatry, 13, 1949.
Harden, C.L., Lazar, L.M., Pick, L.H. et al. (1999). A benecial eect on mood in partial epilepsy
patients treated with gabapentin. Epilepsia, 40, 112934.
Janssen-Cilag (1996). Topiramate. Product monograph.
Kasumakar, V. and Yatham, L.N. (1997). An open study of lamotrigine in refractory bipolar
depression. Psychiatry Res, 19, 1458.
Ketter, T.A., Nalow, B.A., Flamini, R., White, S.R., Post, R.M. and Theodore, W.H. (1994).
Anticonvulsant withdrawal-emergent psychopathology. Neurology, 44, 5561.
254 B. Schmitz
Ketter, T.A., Post, R.M. and Theodore, W.H. (1999). Positive and negative psychiatric eects of
antiepileptic drugs in patients with seizure disorders. Neurology, 53 (Suppl. 2), 5367.
de Haas, pp. 91131. Amsterdam: Elsevier.
Lee, D.O., Steingard, R.J., Cesena, M., Helmers, S.L., Riviello, J.J. and Mikati, M.A. (1996).
Behavioral side eects of gabapentin in children, Epilepsia, 37, 8790.
Leppik, E. (1995). Tiagabine: the safety landscape. Epilepsia, 36 (Suppl. 6), 1013.
Letterman, L. and Markowitz, J.S. (1999). Gabapentin: a review of published experience in the
treatment of bipolar disorder and other psychiatric conditions. Pharmacotherapy, 19, 56572.
Levinson, D.F. and Devinsky, O. (1999). Psychiatric adverse events during vigabatrin therapy.
Lombroso, C.T. (1999). Lamotrigine-induced tourettism. Neurology, 52, 11914.
Marcotte, D. (1998). Use of topiramate, a new anti-epileptic as a mood stabilizer. J Aect Disord,
50, 24551.
Matsuura, M. (1999). Epileptic psychoses and anticonvulsant drug treatment. J Neurol Neurosurg
McConnell, H., Snyder, P.J., Duy, J.D. et al. (1996). Neuropsychiatric side eects related to treat-
ment with felbamate. J Neuropsychiatry Clin Neurosci, 8, 3416.
McDanal, C.E. and Bolman, W.M. (1975). Delayed idiosyncratic psychosis with diphenylhydan-
toin. J Am Med Assoc, 231, 1063.
Normann, C., Langosch, J., Schaerer, L., Grunze, H. and Walden, J. (1999). Treatment of acute
mania with topiramate. Am J Psychiatry, 156, 201415.
Pande, A.C., Davidson, J.R., Jeerson, J.W. et al. (1999). Treatment of social phobia with gaba-
pentin: a placebo-controlled study. J Clin Psychopharmacol, 19, 3418.
Petty, F. (1995). GABA and mood disorders: a brief review and hypothesis. J Aect Disord, 34,
27581.
Robertson, M.M., Trimble, M.R. and Townsend, H.R.A. (1987). Phenomenology of depression
in epilepsy. Epilepsia, 28, 36472.
Sackellares, J.C., Lee, S.I. and Dreifuss, F.E. (1979). Stupor following administration of valproic
acid to patients receiving other convulsant drugs. Epilepsia, 20, 697703.
Sander, J.W.A.S., Hart, E.M., Trimble, M.R. and Shorvon, S.D. (1991). Vigabatrin and psychosis.
J Neurol Neurosurg Psychiatry, 54, 4359.
Schapel, G. and Chadwick, D. (1996). Tiagabine and non-convulsive status epilepticus. Seizure,
5, 1536.
Schmitz, B., Robertson, M. and Trimble, M.R. (1999). Depression and schizophrenia in epilepsy:
social and biological risk factors. Epilepsy Res, 35, 5968.
Schndienst, M. and Wolf, P. (1992). Zur mglichkeit neurotoxischer Sptwirkungen von val-
proinsure. In Valproinsure, ed. G. Krmer and M. Laub, pp. 25965. Berlin: Springer.
Smith, D., Baker, G., Davies, G., Dewey, M. and Chadwick, D. (1993). Outcomes of add-on treat-
ment with lamotrigine in partial epilepsy. Epilepsia, 34, 31222.
Tallian, K.B., Nahata, M.C., Lo, W. and Tsao, C.Z. (1996). Gabapentin associated with aggressive
behavior in pediatric patients with seizures. Epilepsia, 37, 5012.
Thomas, L., Trimble, M.R., Schmitz, B. and Ring, H.A. (1996). Vigabatrin and behaviour disor-
ders: a retrospective study. Epilepsy Res, 25, 217.
Trimble, M.R. and Reynolds, E.H. (1976). Anticonvulsant drugs and mental symptoms: a review.
Psychol Med, 6, 16978.
Trimble, M.R. (1997). Neuropsychiatric consequences of pharmacotherapy. In Epilepsy. A
Comprehensive Textbook, ed. J. Engel and T.A. Pedley, pp. 216170. Philadelphia, New York:
LippincottRaven.
Trimble, M.R. (2000). Anticonvulsants and behaviour: The prole of new drugs with respect to
psychosis and depression. Epilepsia, 41 (Suppl. Florence), 114.
Trimble, M.R., Rusch, N., Betts, T. and Crawford, P.M. (2000). Psychiatric symptoms after
therapy with new antiepileptic drugs: psychopathological and seizure related variables.
Seizure, 9, 24954.
Walden, J., Normann, C., Langosch, J., Berger, M. and Grunze, H. (1998). Dierential treatment
of bipolar disorders with old and new antiepileptic drugs. Neuropsychobiology, 38, 1814.
Wolf, P., Inoue, Z., Rder-Wanner, U.U. and Tsai, J.J. (1984). Psychiatric complications of absence
therapy and their relation to alteration of sleep. Epilepsia, 25, 569.
Wolf, S.M., Shinnar, S., Kang, H., Balaban gil, K. and Mosh, S.L. (1995). Gabapentin toxicity in
children manifesting as behavioral changes. Epilepsia, 36, 12035.
Zaret, B.S. and Cohen, R.A. (1986). Reversible valproic acid-induced dementia: a case report.
17
Antiepileptic drug treatment and epileptic

seizures effects on cognitive function
Albert P. Aldenkamp
University of Amsterdam, Amsterdam, the Netherlands
Introduction
One of the consequences of epilepsy is impairment of cognitive function and

memory impairments, mental slowing and attentional decits are the most fre-
quent disorders (Aldenkamp et al., 1995a; Dodson and Trimble, 1994). Sometimes
such cognitive consequences are more debilitating for the individual patients than
the seizures.
The exact causes of cognitive impairment in epilepsy have not been explored
fully, but clearly three factors are involved: aetiology, the seizures and the central
side eects of treatment (Dodson and Pellock, 1993). Aetiology-related factors are
often clear, are separate from epilepsy or concern a small group; especially in the
age-dependent encephalopathies, such as the West syndrome (Berg and Shinnar,
1997).
We concentrate here on the cognitive eects of seizures and antiepileptic drug
treatment. When evaluating these factors it is imperative to realize that in practice
most cognitive problems have a multifactorial origin and the three aforementioned
factors combined are responsible for most of the make-up of a cognitive problem
in an individual patient. Moreover the factors are related, which causes therapeu-
tic dilemmas when seizure control can only be achieved with treatments that are
associated with cognitive side eects. Nonetheless, these factors are discussed here
separately and sequentially. In the last paragraph the overall impact of the factors
combined is discussed.
Cognitive side effects of antiepileptic drugs
Treament of seizures requires antiepileptic drug (AED) treatment for the large
majority of patients and may be accompanied by unwanted eects on cognitive
function. Although the magnitude of such cognitive side eects is generally con-
sidered to be mild to moderate for most of the AEDs, their impact may be substan-
256
257 Drug treatment, seizures and cognitive function
tial in some patients when critical functions are involved such as learning in chil-
dren (Aldenkamp et al., 1995a) or driving capacities in adults (often requiring
milliseconds precision); or when functions are impaired that are already vulner-
able, such as memory function in the elderly (Trimble, 1987). Moreover, as the cog-
nitive side eects represent the long-term outcome of AEDs, the eects may
increase with prolonged therapy, which contributes to the impact on daily life func-
tioning in refractory epilepsies (Committee on Drugs, 1985).
The following topics are relevant to clinical practice:
The combined eects of seizures and AEDs on cognitive function
Undoubtedly many controversies concern the relative contribution of AED
therapy, compared to the eect of seizure activity on cognition. Improved
seizure control (when for example a new AED is added into the existing drug
regime) may cause cognitive improvement that itself may camouage genuine
cognitive side eects of the new drug. In many situations it will thus be impos-
sible to separate seizure eects from genuine AED eects. Subjective patient
complaints may enlarge this problem, as patients often believe that their cogni-
tive problems are caused by external factors, such as the drugs they have to take
instead of by internal factors, such as their own seizures.
Habituation
In most drugs early side eects may occur, fortunately only for a short period,
i.e. during the rst few days or weeks of drug exposure. After this period nor-
malization occurs, possibly due to the development of so-called positive toler-
ance or habituation (Kulig and Meinardi, 1977). Although little is known about
how tolerance to the cognitive eects of AEDs develops, a failure to take this
factor into account may lead to overestimation of the negative eects of drugs
on cognition. In clinical assessment we should only conclude cognitive side
eects if these persist during long-term treatment.
Subjective patient complaints
Most of the studies use formal neuropsychological tests to assess the cognitive
side eects of AEDs, although in clinical practice the opportunities for such
assessment are usually very limited. Patient complaints, suggestive of problems
in cognitive behaviour often represent the only available evidence of possible
cognitive dysfunction. Nevertheless, it would be unwise to take all patient com-
plaints that suggest cognitive dysfunction at face value. While some patients may
have a clear insight into their own failures, others may have a poor understand-
ing of their performance. Thus, although subjective cognitive complaints are an
important factor to be considered, their use may generate discrepancies between
patient complaints and the outcome of cognitive tests, a situation that, in our
experience, occurs fairly regularly in clinical practice and is a frequent cause for
controversies. When changing to another drug, patients may be doing better but
258 A.P. Aldenkamp
feeling worse, and vice versa, a topic that has been discussed since it was raised
in 1890. A possible solution is to use standardized scales to assess the subjective
impressions of patients. Several of these scales with proven validity now exist
(e.g. the ABNAS neurotoxicity scale; Aldenkamp and Baker, 1997; Aldenkamp
et al., 1995b).
The relationship with serum levels
The specicity of cognitive side eects of AEDs came back into debate recently
when several studies failed to nd earlier reported cognitive side eects of AEDs
when serum concentrations were suciently controlled (Meador et al., 1990).
When Dodrill and Troupin initiated their research on cognitive side eects of
AEDs in 1977 (Dodrill and Troupin, 1977), they reported signicant drug-
induced impairment in patients using phenytoin. However, no impairment
remained, when they removed all patients with phenytoin serum levels exceed-
ing the upper limit of the therapeutic reference interval (30 g/ml). This was done
in a reanalysis of their original work and published about 15 years later (Dodrill
and Troupin, 1991).
For other types of AEDs the eects of higher serum levels seem to be milder,
but were nonetheless also found for carbamazepine and for valproate. The
implication of these ndings is that some of the reported drug-induced cogni-
tive impairments and dierences between drugs actually may have been due to
dierences in drug concentrations during the study, rather than representing
specic eects on cognition. Although the value of routine serum level moni-
toring in clinical treatment is criticized it may be helpful to control the serum
levels when using high dosing if patients have cognitive complaints.
In addition to the eect of dose, the pharmacokinetic properties of antiepi-
leptic drugs may have an eect on cognitive functioning. In carbamazepine,
transient cognitive decits have been detected in relation to high peak serum
levels (Aldenkamp et al., 1987). The pharmacokinetic prole of this drug, char-
acterized by rapid and marked uctuations in serum levels, may dierentially
aect test performance across short periods during the day. In drugs with large
dierences between trough and peak levels, a part of the cognitive assessment
procedure should be performed repeatedly during the day, allowing a compari-
son of performance at peak levels with other periods.
Generics and branded formulations of one drug
It is sometimes suggested that dierent formulations of the same drug may have
dierent eects on cognitive function (Crawford et al., 1996). This may espe-
cially be true for carbamazepine, due to dierences in absorption rate and phar-
macokinetics among the dierent formulations. In a recent study, we did not
obtain signicant dierences in cognitive performance when patients were
switched from the conventional branded form of carbamazepine to several
generic formulations (Aldenkamp et al., 1998). Nonetheless, patients perfor-

mance should be monitored when frequent swiches are made between dierent
formulations of an antiepileptic drug.
The specic eects of the AEDs on cognitive function have been assessed in a
meta-analysis that evaluated 25 years of studies (Vermeulen and Aldenkamp,
1995):
Polypharmacy shows most impact on cognitive function as opposed to mono-
therapy, irrespective of the type of AEDs included. Two drugs with mild cogni-
tive eects may show potentiation of tolerability problems and thus yield serious
cognitive impairment, when used in combination (Trimble, 1987).
All established AEDs have absolute cognitive side eects, i.e. all the investi-
gated drugs have eects when compared to no treatment. These eects are larger
for phenobarbitone (PHB) than for phenytoin (PHT), carbamazepine (CBZ) or
valproate (VPA). But even the latter drugs, that are generally considered to be
drugs with a safe cognitive prole, have cognitive eects, mostly resulting in a
mild general psychomotor slowing (Aldenkamp et al., 1993).
The respective dierences between the four most investigated AEDs: PHB,
PHT, CBZ and VPA, can be considered as small, when studied within a normal
therapeutic dose, with the exception of the cognitive eects of PHB that has a
less favourable cognitive prole when compared to PHT, VPA and CBZ. Possibly
the most marked ndings were that all AEDs have some cognitive eects and
that the eects of PHT are more moderate than has been suggested previously
(Meador et al., 1990).
The cognitive side eects of each antiepileptic drug. Phenobarbitone is the
only AED with specic absolute eects: when compared with no treatment
memory functions are aected. In long-term treatment even impairment
or delay of psychological development is reported, leading to intellectual
deterioration. No information is available on dose-eects.
Phenytoin has a much milder impact on cognitive function than had hitherto
been suspected. When compared with no treatment PHT-induced attentional
decit and mental slowing is observed but these are not markedly dierent from
similar eects of CBZ and VPA. No correlation with dose was found.
Carbamazepine has been compared with several drugs and some dierences
have been found in respect of VPA and PHT. There is some evidence that dose-
eects may occur and that controlled-release formulations have a more favour-
able cognitive prole than conventional forms. No dierences were found
between dierent generic forms of CBZ.
Valproate has mild psychomotor slowing as an absolute eect and has a
similar cognitive prole when compared with CBZ and PHT. In contrast with
CBZ, dose-relationships have not been obtained and there is no eect when
260 A.P. Aldenkamp
patients are switched from conventional to controlled-release formulations.

For the newer antiepileptic drugs there is a suggestion of equivalence with
CBZ and VPA for oxcarbazepine, gabapentin and vigabatrin (although mood
eects have been reported for vigabatrin). For the other newer drugs, lamotri-
gine and tiagabine, no data from controlled studies have been published yet.
Clinical anecdotal information suggests mood eects during titration of tiaga-
bine and attentional-decits in relation to treatment with lamotrigine in chil-
dren. A recent study (Aldenkamp et al., 2000) showed cognitive impairment in
topiramate related to dose escalation speed.
Cognitive effects of seizures
When evaluating the impact of seizures on cognition we should consider the time-
scale of such impairments; they may be short-term reactions (direct or acute eects
of seizures) or they may persist over time. Somewhere in this time-factor there is
the crucial issue of some aspects of reversibility versus irreversibility of the cogni-
tive impairment. Some aspects of this are discussed by Besag (Chapter 6).
Short-term or acute effects

Short-term or direct (acute) eects of single seizures are reported on several cog-
nitive functions especially on alertness and short-term learning (Bornstein et al.,
1988) even when the period with unconsciousness is short or when there is no or
only partially disturbed consciousness. Most impact of the seizures, however, is
through postictal eects, with impaired alertness sometimes over considerable
periods (Dodrill, 1986). These postictal eects are more dicult to detect and
therefore to acknowledge in the patient. Nevertheless, post-ictal eects may be
detectable on attentional tests during days after a single tonic-clonic seizure and
during several hours after other seizure types (Aldenkamp et al., 1996). In addition
to the cognitive eects related to postictal confusion, some studies point to the dis-
ruptive eects of epileptic discharges on long-term potentiation (Moore et al.,
1993), the physiological process important for learning and especially for stabiliz-
ing information in memory. It is thus not just the epileptic seizure (i.e. the actual
ictal period) that causes cognitive impairment that explains the sometimes
extended periods with cognitive impairment as an aftermath of a seizure, even in
patients with short nonconvulsive seizures.
This is probably also the reason that repeated nocturnal seizures may have cog-
nitive eects. These are found to have detrimental eects on language functions, on
memory and alertness, possibly also through the eects of disturbed sleep patterns
(Snead and Saito, 1993).
Seizure type must be taken into account, with most eects on cognitive function
being noted with secondarily generalized tonic-clonic and complex partial seizures
(Prevey et al., 1998).
One might argue that if a single seizure causes cognitive impairment then more
seizures will cause more severe cognitive decits, an issue that is associated with
seizure frequency. Indeed, seizure frequency has been successfully correlated to
cognitive impairment in a number of studies (Aldenkamp and Alpherts, 1999). A
particularly convincing case is the follow-up of identical twins concordant for inci-
dence of epilepsy, revealing seizure frequency as the only factor associated with cog-
nitive and educational problems (Dodrill and Troupin, 1976).
This factor may also explain the sometimes remarkable improvements of cogni-
tive function that are found after starting treatment in children with nonconvulsive
seizures. Nonconvulsive seizures are by denition more dicult to detect than con-
vulsive attacks, and there is an eect of having seizures for a considerable period of
time before being clinically detected (Mandelbaum and Burack, 1997). Such
periods with uncontrolled seizures may even cause deterioration of intelligence
(Aldenkamp et al., 1996), although this may be reversible after initiation of treat-
ment (Mandelbaum and Burack, 1997).
The conclusion is therefore that single seizures may have an aftermath on cogni-
tive dysfunction over sometimes extended periods (even with short seizures) and
ongoing seizure activity may result in impressive cognitive eects due to the accu-
mulating eects of the seizures. This is especially relevant in patients with dicult-
to-detect seizures as these seizures may accumulate to serious cognitive
impairment, without recognition of the source of deterioration. Conversely,
seizure control may cause cognitive improvements.
Long-term (irreversible?) effects

The direct eects of seizures will thus dissolve in the majority of the patients after
the seizures are adequately controlled. Although some experimental studies, using
several animal models, found damage to the hippocampus and related limbic struc-
tures after a single prolonged generalized or limbic seizure (Meldrum, 1997), there
is no convincing evidence for similar eects in humans (Holmes, 1991). Even with
high seizure frequency most cognitive eects will therefore dissolve after seizure
control is achieved (Hoch et al., 1994). Berg and Shinnar (1997) conclude in their
review that epilepsy in humans is usually not a progressive disorder.
The discussion about irreversibility of seizure-related cognitive eects thus con-
centrates on cases with ongoing seizure activity and consequently on factors such
as seizure duration, number of seizures before treatment (Cameld et al., 1996),
number of seizures during lifetime (Dodrill, 1986) or number of years with seizures
(Jokeit and Ebner, 1999). This is very reminiscent of Gowers suggested mental
decay (epileptic dementia) as a consequence of the pathological long-term
262 A.P. Aldenkamp
sequelae of the seizures. In general it can be concluded that the severity of cogni-
tive impairment is associated with the number of years in which seizures actually
occurred (Bourgeois et al., 1983; Rodin et al., 1986). For secondarily generalized
seizures, Dodrill (1986) obtained a clear relationship between cognitive deteriora-
tion and number of secondarily generalized tonic-clonic seizures during lifetime,
with 100 seizures as the crucial cut-o point. For complex partial seizures recent
studies do not point to number of seizures but to a time-window that allows for
normalization of cognitive eects. In the study of Kotagal et al. (1987) irreversible
eects were found after 5 years of continuing seizures. Most controlled studies,
however, showed a longer time-window and point to irreversible cognitive eects
after periods of about 20 years with continuing complex partial seizures (Jokeit and
Ebner, 1999; Chapter 11). Additionally, dierent results were reported for patients
when reaching seizure remission after a short period with seizures (Seidenberg et
al., 1981), revealing cognitive improvement, versus patients reaching remission
after an extended period with seizures (Rodin et al., 1986). This latter group had
lower outcomes on IQ tests and showed no improvement after remission.
The conclusion is therefore that irreversible cognitive eects of seizures occur
only as an eect of continuing seizures over longer periods. Some studies suggest
the existence of a time-window: when seizure remission is achieved within this
time-window, cognitive function will normalize to premorbid levels (when aetio-
logy does not interfere). Outside this time-window, cognitive impairments may
become irreversible. This time-window has been set for secondarily generalized
seizures to be within 100 seizures and for complex partial seizures at 5 years with
continuing seizures. This illustrates the need for achieving seizure control before
such a time-window for irreversibility is exceeded.
Conclusion; seizures and drugs
Single seizures have an aftermath in regard to cognitive impairment. Although this

sometimes aects the individual over extended periods, these eects are generally
reversible and dissolve, but may be substantial in patients with high seizure fre-
quency. This may be especially important in patients with dicult-to-detect sei-
zures, as the eect of undetected seizures may accumulate to serious cognitive
impairment and even (temporary) deterioration of intelligence.
Irreversible cognitive eects have been demonstrated in patients with recurrent
seizures that persist over longer periods.
On the other hand, all antiepileptic drugs also aect cognitive function to some
extent, although this eect is much more limited compared to the eects of the sei-
zures. However, when high dosing or adjunctive polytherapy is needed the balance
between benets and disadvantages may be negatively biased against drug treat-
ment. Drug treatment therefore requires careful balancing in the attempt to reach
maximal seizure control and avoidance of neurotoxic side eects.
The relevance of cognitive impairment in this balance is illustrated in studies
using quality of life as an outcome measure. The integrity of cognitive function is
highly correlated with the possibility of achieving important goals in life, such as
satisfactory occupational opportunities and social relationships. Probably this rela-
tionship is mediated through an eect of cognitive impairment on education.
Whenever cognitive impairments occur, even if these are temporary eects, they
may aect educational progress and lead to restricted occupational opportunities
in later life (Austin et al., 1999). Sillanp et al. (1998) studied the long-term prog-
nosis of seizures, using a 30-year follow-up of 220 patients. The study showed a
clear correlation between socio-economic status at endpoint and seizure remission.
Whenever seizures continued, adverse social eects occurred and these tended to
persist for extended periods of their life even when seizure remission was achieved
in a later phase of life. The longer the period with seizures, the more impact on daily
life was obtained, but the most crucial period appeared to be the early period
immediately after onset of epilepsy, if the epilepsy had its onset in childhood. This
conrms the aforementioned relationships and illustrates that early achievement of
seizure remission is thus a crucial factor preventing the development of cognitive
impairments and consequently adverse educational and social eects of epilepsy.
R E F E R E N C ES
Aldenkamp, A.P. and Alpherts, W.C.J. (1999). Psychological assessment. In Handbook of Clinical
Neurology, ed. H. Meinardi and G. Bruyn, pp. 387408.
Aldenkamp, A.P. and Baker, G. (1997). The Neurotoxicity Scale-II; Results of a patient-based
scale assessing neurotoxicity in patients with epilepsy. Epilepsy Res, 27, 16573.
Aldenkamp, A.P., Alpherts, W.C.J., Moerland, M.C., Ottevanger, N. and van Parijs, J.A.P. (1987).
Controlled release carbamazepine: cognitive side eects in patients with epilepsy. Epilepsia, 28,
50714.
Aldenkamp, A.P., Alpherts, W.C.J., Blennow, G. et al. (1993). Withdrawal of antiepileptic medi-
cation; eects on cognitive function in children the results of the multicentre Holmfrid
study. Neurology, 43, 4151.
Aldenkamp, A.P., Dreifuss, F.E. and Renier, W.O. (1995a). Epilepsy in Children and Adolescents.
New York: CRC-Press Publishers.
Aldenkamp, A.P., Baker, G., Pieters, M.S.M., Schoemaker, H.C., Cohen, A.F. and Schwabe, S.
(1995b). The Neurotoxicity Scale; the validity of a patient-based scale, assessing neurotoxicity.
264 A.P. Aldenkamp
Aldenkamp, A.P., Overweg, J., Gutter, Th., Beun, A.M., Diepman, L. and Mulder, O.G. (1996).
Eect of epilepsy, seizures and epileptiform EEG discharges on cognitive function. Acta Neurol
Scand, 93, 2539.
Aldenkamp, A.P., Rentmeester, Th., Hulsman, J. et al. (1998). Pharmacokinetics and cognitive
eects of carbamazepine formulations with dierent dissolution rates. Eur J Clin Pharmacol,
54, 18592.
Aldenkamp, A.P., Baker, G., Mulder, O.G. et al. (2000). A multicentre randomized clinical study
to evaluate the eect on cognitive function of topiramate compared with valproate as add-on
therapy to carbamazepine in patients with partial-onset seizures. Epilepsia, 41, 116778.
Austin, J.K., Huberty, T.J., Huster, G.A. and Dunn, D.W. (1999). Does academic achievement in
children with epilepsy change over time? Dev Med Child Neurol, 41, 4739.
Berg, A.T. and Shinnar, S. (1997). Do seizures beget seizures? An assessment of the clinical evi-
dence in humans. J Clin Neurophysiol, 14, 102110.
Bornstein, R.A., Drake M.E. Jr and Pakalnis, A. (1988). WAIS-R factor structure in epileptic
patients. Epilepsia, 29, 1418.
Bourgeois, B.F.D., Presky, A.L. and Palkes, H.S. (1983). Intelligence in epilepsy: a prospective
study in children. Ann Neurol, 14, 43844.
Cameld. C., Cameld, P., Gordon, K. and Dooley, J. (1996). Does the number of seizures before
treatment inuence ease of control or remission of childhood epilepsy? Not if the number is
10 or less. Neurology, 46, 414.
Committee on Drugs (1985). Behavioral and cognitive eects of anticonvulsant therapy.
Pediatrics, 76, 6447.
Crawford, P., Hall, W.W., Chappell, B., Collings, J. and Stewart, A. (1996). Generic prescribing
for epilepsy is it safe? Seizure, 5, 15.
Dodrill, C.B. (1986). Correlates of generalized tonic-clonic seizures with intellectual, neuro-
psychological, emotional, and social function in patients with epilepsy. Epilepsia, 27,
399411.
Dodrill, C.B. and Troupin, A.S. (1976). Seizures and adaptive abilities. A case of identical twins.
Arch Neurol, 33, 6047.
Dodrill, C.B. and Troupin, A.S. (1977). Psychotropic eects of carbamazepine in epilepsy: a
double-blind comparison with phenytoin. Neurology, 27, 10238.
Dodrill, C.B. and Troupin, A.S. (1991). Neuropsychological eects of carbamazepine and phen-
ytoin; a reanalysis. Neurology, 41, 1413.
Dodson, W.E. and Pellock, J.M. (1993). Pediatric Epilepsy: Diagnosis and Treatment. New York:
Demos Publications.
Dodson, W.E. and Trimble, M.R. (1994). Epilepsy and Quality of Life in Epilepsy. New York: Raven
Press.
Hoch, D.B., Hill, R.A. and Oas, K.H. (1994). Epilepsy and mental decline. Neurol Clin, 12, 1013.
Holmes, G.L. (1991). Do seizures cause brain damage? Epilepsia, 32 (Suppl. 5), S1418.
Kotagal, P., Rothner, A.D., Erenberg, G., Cruse, R.P. and Wyllie, E. (1987). Complex partial sei-
zures of childhood onset. A ve year follow-up study. Arch Neurol, 44, 117780.
Kulig, B. and Meinardi, H. (1977). Eects of antiepileptic drugs on motor activity and learned
behavior in the rat. In Advances in Epileptology, ed. H. Meinardi and A.J. Rowan, pp. 98104.
Amsterdam: Swets & Zeitlinger.
Mandelbaum, D.E. and Burack, G.D. (1997). The eect of seizure type and medication on cog-
nitive and behavioral functioning in children with idiopathic epilepsy. Dev Med Child Neurol,
39, 7315.
Meador, K.J.M., Loring, D.W., Huh, K. et al. (1990). Comparative cognitive eects of anticonvul-
sants. Neurology, 40, 3914.
Meldrum, B.S. (1997). First Alfred Meyer Memorial Lecture. Epileptic brain damage: a conse-
quence and a cause of seizures. Neuropathol Appl Neurobiol, 23, 185201.
Moore, S.D., Barr, S.D. and Wilson, W.A. (1993). Seizure-like activity disrupts LTP in vivo.
Prevey, M.L., Delaney, R.C., Cramer, J.A. and Mattson, R.H. (1998). Complex partial and secon-
darily seizure patients: cognitive functioning prior to treatment with antiepileptic medication.
VA Epilepsy Comparative Study 264 Group. Epilepsy Res, 30, 19.
Rodin, E.A., Schmaltz, S. and Twitty, G. (1986). Intellectual functions of patients with childhood-
onset epilepsy. Dev Med Child Neurol, 28, 2533.
Seidenberg, M., OLeary, D.S. and Giordani, B. (1981). Test-retest changes of epilepsy patients:
assessing the inuence of practice eects. J Clin Neuropsychol, 3, 23755.
Sillanp, M., Jalava, M., Kaleva, O. and Shinnar, S. (1998). Long-term prognosis of seizures with
onset in childhood. N Engl J Med, 338, 171522.
Snead, O.S. and Saito, M. (1993). Encephalopatic epilepsy after infancy. In Pediatric Epilepsy:
Diagnosis and Therapy, ed. W.E. Dodson and J.M. Pellock. New York: Demos Publications.
Trimble, M.R. (1987). Anticonvulsant drugs and cognitive function: a review of the literature.
Vermeulen, J. and Aldenkamp, A.P. (1995). Cognitive side-eects of chronic antiepileptic drug
treatment: a review of 25 years of research. Epilepsy Res, 22, 6595.
18
Psychiatric effects of surgery for temporal

lobe epilepsy
Steffi Koch-Stoecker
Klinik Mara I, Bielefeld, Germany
Introduction
The cessation of epilepsy by surgical intervention although it may be an end in

itself primarily aims at a reduction of injuries and embarrassment, in order to
enable patients to do what they formerly were precluded from doing. It oers pos-
sibilities to live a more fullled and self-condent life. But if the disasters and
injuries are somehow self-imposed, if the embarrassment arises mainly from pecu-
liarities of character, if the sense of stigmatization is a deeply held conviction . . .
removing the epilepsy will not necessarily alter these conditions (Taylor et al.,
1997). Those patients although seizure-free may stay in their disabled situation.
The experience that the results of epilepsy surgery (ES) even if successful from
a neurological point of view may be far from satisfying for special groups of
patients leads to the following discussion.
On the one hand, a growing interest in psychiatric comorbidity has led to better
prognostic knowledge about negative outcomes. This led to a tendency to set up
contraindications for surgery for those patients who suered from the most severe
psychiatric disorders, namely chronic psychoses, because this patient group con-
tinued to exhibit their psychotic features postoperatively, and the positive eects of
the surgical intervention seemed doubtful. In principle, this strategy to exclude
patients with predictable barriers to surgical success is reasonable, but it is still
unclear which group of psychotic patients to preclude from surgery (see below).
On the other hand, epileptologists started to consider more than just postoper-
ative seizure frequency for the evaluation of surgical outcome. Evidence for this
tendency is to be found in the widely accepted system of outcome classication by
Engel et al. (1993). They stated that some degree of consideration for the impact
of residual seizures on quality of life is necessary, which led to a dierentiation
between the two outcome categories, Class III: Worthwhile improvement and
Class IV: No worthwhile improvement. With that dierentiation, Engel et al.
explicitly went beyond the level of pure seizure-counting and introduced aspects of
266
267 Psychiatric effects of surgery for TLE
quality and subjective value: the same postoperative seizure frequency may be a
worthwhile success for one patient but a negligible seizure reduction for another
one. With that introduction of aspects of subjective value, Engel et al. have intro-
duced a new direction into the outcome discussion which forces evaluators to take
into account the individual experiential conditions and personal assessments of
patients postoperatively. It therefore makes sense to go one step further and employ
such a worthwhile-category also with respect to Engels outcome Class I, namely
the seizure-free patients.
Engel et al. (1993) emphasize the present lack of quantitative measures to distin-
guish between Class III and Class IV. Of course, quality of life research has devel-
oped some approaches to comprehend quality with quantitative methods, but solid
answers to questions of individual success still need single case reconstructions,
including, among others, an understanding of patients thoughts and emotions, of
sorrows and fears, and of lucid and blind spots in social perception.
It is a genuine psychiatric-psychotherapeutic task to conduct such reconstruc-
tions, and it requires time, expertise and costs. However, it can make a crucial
dierence to the indications for and the evaluation of surgery. With the help of the
psychiatric concept of personality disorders, some basic predictors for subjective
surgical success beyond the number of postoperative seizures can be found.
In short, there are at least two aspects to the psychiatric evaluation of surgical
candidates: assessment is not only useful for the prognosis of psychiatric compli-
cations, but also with respect to the estimation of the subjective value of surgery for
individual patients.
In the following I shall present results and ask questions concerning both aspects:
what do we know about the occurrence of psychiatric disorders in the context of epi-
lepsy surgery, and how can the analysis of individual personality traits be used to
explain peculiarities and set up personal aims of patients around surgery? I want to
propose how to integrate both aspects into psychiatric assessment in epilepsy centres.
Presented data refer to surgical candidates with temporal lobe epilepsies only.
Whenever mention is made of our own data (Bethel), I refer to a psychiatric
outcome study of the rst 100 adult patients who had a temporal lobe resection in
our epilepsy centre, Mara I, in Bethel, Bielefeld, whom I have followed up for at
least 2 years (Koch-Stoecker, 2001).
Psychiatric disorders in the context of temporal lobe resections
In order to achieve relevant information about psychiatric diagnoses in the context

of epilepsy surgery it is imperative to collect psychiatric data preoperatively and set
them into relation with postoperative changes. Moreover, in order to ensure the
comparability of results, it would be sensible for dierent epilepsy centres to
268 S. Koch-Stoecker
Table 18.1. Total psychiatric comorbidity in surgical candidates
Jensen and Larsen 1979: 80%

Polkey 1983: 43%
Naylor et al., 1994: 43%
Manchanda et al., 1996: 47%
Ring et al., 1998: 52%
Blumer et al., 1998: 57%
Glosser et al., 2000: 51% (psychiatric syndromes, Axis I DSMIIIR)
Bethel: 43% (psychiatric syndromes, Axis I DSMIIIR);
72% (psychiatric syndromes plus personality disorders)
Note:
DSM-III-R; American Psychiatric Association (1987).
attempt an agreement about the diagnostic methods to be used. Those two basic
demands were already formulated in the 1960s (Ferguson and Rayport, 1965), but
are still unfullled today. What is lacking is the integration of psychiatrists into the
epilepsy surgery units, who can routinely conduct a psychiatric assessment for all
surgical candidates.
Instead, in the literature, there are many retrospective analyses. These studies
attempt to reconstruct psychiatric disorders only using patients notes, in which the
reports of psychiatric case histories are often incomplete or even missing. This leads
to unreliable results, which are especially susceptible to underestimations and false
classications of psychiatric disorders.
Further, there are several studies using psychological questionnaires. Yet,
although well constructed and standardized for specic disorders, these question-
naires often are not validadted for problems of patients with epilepsy, and in any
case cannot substitute for clinical psychiatric diagnostics.
In the following these limitations of some of the presented data are to be kept in
mind.
Total psychiatric morbidity in patients with temporal lobe surgery
Results on total psychiatric comorbidity dier depending on the ways patients are
referred to the centres, on the strategies of evaluation, and on the absence or pres-
ence of a psychiatric assessment, the latter leading to higher frequencies of psychi-
atric diagnoses. There are two notable trends.
First, comorbidity in surgical candidates is surprisingly high, ranging between
43 and more than 80% (Table 18.1). The most plausible explanation of these nd-
ings lies in the fact that mesio-temporal structures which are disturbed in tempo-
ral lobe epilepsy (TLE) are central to experiencing and processing emotions and are
susceptible to the development of dysfunctional cellular connections, with an eect
on behaviour.
Second, the total amount of psychopathology diminishes only slightly after
surgery: stable patients usually stay stable, some deteriorate, some especially if
seizure-free improve (see below for more details). As for the prognosis of poor
psychiatric outcome, in Bethel we have noted the existence of severe personality
disorders as a potent indicator. Anhoury et al. (2000) found the presence of preop-
erative psychiatric disorders, bilateral independent spike discharges, and the size of
surgical resections as predictors.
Psychoses
The term psychosis designates globally spoken severe psychiatric syndromes,

characterized by thoughts, feelings and actions that are incomprehensible for a
neutral observer. Diagnoses are often based on spectacular symptoms such as delu-
sions and hallucinations without any further diagnostic dierentiation. Given this
weak assessment basis, Savard (1991) found in a meta-analysis of diverse studies,
preoperative rates of psychoses between 7 and 16%, and postoperatively between
10 and 28%.
The importance of an exact classication of psychoses in the context of epilepsy
is emphasized by Trimble and Schmitz (1997). They distinguish between ictal,
postictal, peri-ictal, interictal and alternative psychoses. Except for interictal psy-
choses, which require neuroleptic treatment, all other psychoses in epilepsy require
a regulation of the antiepileptic drugs as the rst therapeutic intervention. Thus,
clear diagnoses can save patients from referrals to psychiatric hospitals.
Postictal psychoses
Poor diagnostic dierentiation between the psychoses, especially between postictal
and interictal ones, can have severe consequences for surgical candidates. Thus,
without exact psychiatric classication, the already-mentioned tendency to exclude
psychotic patients from surgery could mislead surgeons into regarding postictal
psychoses as a contraindication for ES. Mislead because patients with postictal
psychoses can prot from surgery in two ways. If the resection is successful, they
lose their seizures. However, in addition they will lose their directly seizure-related
psychosis. For these reasons Fenwick (1994) has even suggested that postictal psy-
choses should be regarded as a psychiatric indication for ES.
The neurological condition of these patients is, however, complicated and this
may result in a less favourable seizure prognosis. For example, bitemporal (Savard
et al., 1991) and extratemporal EEG discharges, clusters of seizures (Umbricht et
Table 18.2. Postictal psychoses
Incidence in TLE
4% (Kanemoto et al., 1996)
Incidence in surgical candidates

18% (Umbricht et al., 1995);
13% (Kanemoto et al., 1998);
6% (Bethel)
Psychiatric outcome
Psychoses
none (Bethel)
Temporary mood disorders
60% (Kanemoto et al., 1998);
50% (Bethel)
Seizure outcome (class I)
33%; further 33% after second resection (Bethel)
al., 1995), and nocturnal GTCS (Kanemoto et al., 1996) are all reported. This
emphasizes the need for a comprehensive neuropsychiatric evaluation of such
cases.
The occurrence of postictal psychoses was quoted at 4% in a large study group
of more than 800 patients with TLE (Kanemoto et al., 1996), but the incidence in
surgical candidates is higher (between 6 and 18%). Whether seizure outcome is less
favourable than in the total group needs further evaluation. Results on incidence
and postsurgical course are shown in Table 18.2.
Chronic interictal psychoses

Because of the widespread reservation about operating on chronic psychotic
patients, due to the argument that the psychoses would continue to persist anyhow,
and surgery would therefore not be protable, the number of severe psychotic
patients evaluated has become small in epilepsy surgery centres. However, the argu-
ment of the positive eects of a seizure reduction in psychotic patients is relevant,
and Fenwick (1988) has argued that psychotic patients without seizures could be
much better o than patients with psychosis and seizures. Moreover there are
reports of permanent remittance of psychoses after surgery (Jensen and Larsen,
1979), and enduring deteriorations on the other hand have not been found (Taylor,
1972).
Nevertheless, surgical interventions in chronic psychotic patients are compli-
cated. First, the comorbidity of psychosis and TLE may be evidence for an extended
limbic dysfunction. Second we know that these patients are vulnerable and tend to
experience acute exacerbations of their psychosis during stressful life events, like
surgery. It is therefore imperative to provide special perioperative care and tailored
postoperative rehabilitation settings for such patients (Krahn et al., 1996; Taylor,
1987) in order to prevent acute crises. It is recommended that discussion with such
patients includes information about the indications and the aims of surgery but to
dierentiate between their epilepsy and their psychosis, including the information
that the psychosis will most probably continue.
With these special preparatory conditions, we have achieved worthwhile results
in Bethel. All three chronic psychotic patients we operated on experienced acute
psychotic exacerbations after surgery, but in the long run their psychoses became
milder in all cases, going along with the seizure reduction. However, only one of the
three patients has become seizure-free.
Postoperative psychoses
One question is, whether or not there are typical de novo psychoses induced by ES.
According to one position, postoperative psychoses primarily occur as so-called de
novo postictal psychoses (Savard et al., 1998) in patients with persistent seizures,
and thus are only indirectly connected with the surgical event. Another position is
that surgery only has the function of a trigger that releases a manifest psychosis,
which was already latent and might have found its preoperative expression in par-
anoid personality traits (Ferguson et al., 1993). However, there are still good argu-
ments for the diagnostic entity de novo psychosis as aetiologically linked to the
surgical intervention. Mace and Trimble (1991) consider them to be an eect
related to a nondominant hemisphere hypofunction, because they predominantly
occur after right/nondominant resections. They further argue that the sudden inhi-
bition of seizure activity through surgery may induce mechanisms parallel to those
of forced normalization.
Altogether there is no doubt about the occurrence of postoperative psychoses,
but again most data concerning aetiology and predictors as well as clinical features
have to be interpreted with caution, because of well-known methodological
reasons (retrospective analyses, missing reliable preoperative information, prob-
lems of classication).
Savard (1991) found a spectrum of frequencies between 0.5 and 21%, and
Trimble (1992) between 3.8 and 35.7%, with a mean of 7.6%. Concerning
morphology, several studies have suggested that gangliogliomas predispose to post-
operative psychoses (Andermann et al., 1999; Bruton, 1988), but we did not nd
such a correlation in our centre. Nondominant temporal foci are frequent (Mace
and Trimble, 1991; Bethel) in contrast to an excess of left-sided, dominant lesions
noted for chronic psychoses in epilepsy.
Table 18.3. Postoperative de novo psychoses
Frequency
0.521% (meta-analysis, Savard, 1991); 3.835.7% (meta-analysis, Trimble, 1992);
11% (Bethel)
Morphology
Gangliogliomas preferred (Bruton, 1988; Andermann et al., 1999)
Laterality
Nondominant epileptic focus (Mace and Trimble, 1991); 80% nondominant (Bethel)
Preoperative psychopathology
100% personality disorders (Bethel)
Symptoms
Starting with depressive symptoms, sleep disorders, going on with delusions (frequently after
rst seizure-relapse)
Psychotic contents
Coping with surgery/new psychosocial demands (Ferguson and Rayport, 1965)
Long-term development
Variable: some chronic, some free of psychosis after second resection, some remitting with
neuroleptic treatment (Bethel)
Concerning preoperative psychopathology, we found that all patients had per-

sonality disorders before surgery (Koch-Stoecker, 1997), which indicates that
surgery may be the critical event overwhelming the psychotic threshold in patients
with already preoperatively weakened personality structures.
Psychoses often start with symptoms of mood and sleep disturbance and then
continue with delusions, which are frequently initiated by a seizure-relapse. The
psychotic contents mainly relate to two themes (Ferguson and Rayport, 1965): (1)
a psychotic structure of impressions of the surgical context (such as suspecting
microchips or laser inuence in the brain), (2) the apprehension of new psycho-
social demands (such as the paranoid ideation that neighbours control the patients
actions).
The long-term course in some cases depends on patients compliance to take pre-
scribed medications, in others on further treatment of epilepsy (e.g. reoperation).
Many patients respond to neuroleptic medication, but in others there is a necessity
for repeated treatment in psychiatric hospitals. These ndings are summarized in
Table 18.3.
Affective disorders and anxiety
As is the case for psychoses, it is also true for the aective disorders that the usual
psychiatric diagnostic categories do not oer adequate classication for epilepsy
patients. Typical constellations of symptoms allowing the diagnosis of a major

depressive episode are rare, dysthymic or organic depressive states rather frequent.
Blumer and Altshuler (1997) have attempted to introduce a useful classication
category of aective disorders in patients with TLE, which they call interictal dys-
phoric disorder. Predominant features are, among others, depressive mood, par-
oxysmal irritability leading to outbursts of verbal aggressivity with consecutive
feelings of shame, a sudden onset and a brief duration of only days.
Blumer et al. (1998) found that those interictal dysphoric mood disorders
present in 57% of their patients faded away after surgery in 20% of patients, 36%
stayed stable with their dysphoric disorder, and 44% worsened after surgery, some
with a remission after antidepressant treatment. There were also about 40% of the
psychiatrically intact group, predominantly those who continued to have seizures,
who developed their dysphoria after surgery.
The enduring remittance of depressive symptoms depends on complete seizure
relief (Blumer et al., 1998; Hermann and Wyler, 1989). This nding is supported
by our results. Moreover we found that our preoperatively depressed patients
showed dierences in psychopathology after surgery related to laterality: dominant
resections led to somatoform symptoms as surrogates of depression (headache,
backache, etc.), while nondominant resected patients frequently had postoperative
depressions.
Emotional irritation and lability with sudden mood changes, uncertainty con-
cerning the future, reduced stress tolerance etc. during the rst months after ES are
very typical (41% of all resected patients, Fraser 1988; 45% during the rst 6 weeks,
Ring et al., 1998).
Additionally, circumscribed episodes of depression occur after ES. As early as in
1957 Hill et al. described their occurrence, being independent of seizure outcome,
with a remission within the rst 18 months after surgery. Because of their tempor-
ary character, Trimble (1992) designates them as complications of surgery. Their
frequency is about 810% of surgically treated patients (Naylor et al., 1994). They
occure more with nonlesional resections or mesio-temporal scleroses (Bruton,
1988), in nondominant resected patients (Fenwick et al., 1993; Bethel) and in pre-
operatively aggressive patients, who lose their aggressiveness after surgery and tend
to develop depressions (Taylor, 1987). There are hints from one research group
(Kanemoto et al., 1998) of correlations with dominant resections and with post-
ictal psychoses before surgery. The occurrence of postoperative depression was
found to be independent of seizure outcome (Hill et al., 1957), except for psychi-
atrically preoperatively intact patients in whom depression seems to be linked to
seizure recurrence (Blumer et al., 1998). For an overview see Table 18.4.
Mania seldom occurs in patients with TLE (Wolf, 1982). However, with
respect to postsurgical outcome there are some hints for the occurrence of manic
syndromes. Krahn et al. (1996) describe hypomanic states immediately after
Table 18.4. Episodes of postoperative depression
Duration
Remission within 18 months (Hill et al., 1957)
Aetiological hypothesis
Process of scarring (Trimble, 1992)
Frequency
810 % of resected patients (Bruton, 1988; Naylor et al., 1994; Bethel)
Morphology
Mesio-temporal sclerosis or nonlesional resections (Bruton, 1988)
Laterality
Nondominant resections (Fenwick et al., 1993; Bethel); dominant resections (Kanemoto
et al., 1998)
Psychiatric predictors
Aggressivity leads to postoperative depression (Taylor, 1987); postictal psychoses leads to
postoperative depression (Kanemoto et al., 1998)
Seizure outcome
Independent occurrence (Hill et al., 1957; Bethel)
surgery and Kanemoto et al. (1998) reported about 10% of the resected patients
showing (hypo)manic episodes directly after surgery. It may well be that the inci-
dence of manic disorders is usually underestimated because of two dierent
reasons: (1) the dierentiation between optimistic gladness after successful resec-
tion and symptomatic euphoria is dicult in some cases; (2) manic symptoms
may have already vanished and may not be remembered at the time of the rst
postoperative evaluation, which in many centres takes place at 3 or 6 months
after surgery.
Symptoms of anxiety are very common in epilepsy patients, but their classica-
tion covers many problems of dierentiation between fear of seizures, fear as a
symptom of seizures, avoidant behaviours due to stigmatization, fear as a symptom
of depression, and others.
Accordingly, preoperative estimations of anxiety disorders in candidates for
surgery vary between 10% (Manchanda et al., 1996) and 44% (Bladin, 1992). More
than 2 years after surgery Koch-Weser et al. (1988) even found higher rates of
anxiety than before surgery.
Ring et al. (1998) reported a frequency of 42% of early postoperative symptoms
of anxiety which had already diminished after 3 months. In the early weeks after
surgery, an exact dierentiation between the woven symptoms of irritability,
anxiety, and mood uctuation is not easy and might even be impossible. This time
period deserves to be better evaluated from the psychopathological perspective.
Table 18.5. Postoperative nonepileptic attacks
Frequency
10% (Glosser et al., 1999);
5% (Ney et al., 1998);
4% (Bethel)
Preferred incidence
Gender
Women (Glosser et al., 1999; Bethel)
Laterality
Right (Glosser et al., 1999; Bethel); Left (Ney et al., 1998)
Onset-time
After adolescence (Glosser et al., 1999)
Preoperative psychopathology
High (Ney et al., 1998);
Borderline personality disorders (Bethel)
IQ
Low (Ney et al., 1998)
Operative complication rate
High (Ney et al., 1998)
Nonepileptic attacks (nonepileptic seizures)
Most centres are reluctant to operate on patients with epileptic seizures which
occur in association with nonepileptic attacks. Even if the epilepsy is cured by
surgery, there is a high probability of the dissociative attacks continuing. Therefore,
only after nonepileptic attacks are well treated by psychotherapeutic interventions
should surgery be considered (Henry and Drury, 1997).
Concerning postoperative nonepileptic attack disorders (NEADs), there are
some reports in early surveys on the psychiatric eects of epilepsy surgery
(Ferguson and Rayport, 1965; Taylor, 1972). After a long period of scientic
neglect, they recently are attracting attention again.
Glosser et al. (1999) found NEADs in just under 10% of cases within a timeframe
of 10 years after epilepsy surgery. They started within the rst months after surgery,
aected predominantly women, lateralization of resection was right temporal, and
the seizure-onset was frequently after adolescence. Ney et al. (1998) found 5% post-
operative NEADs, with left lateralization, a high rate of preoperative psychopathol-
ogy, low IQ and high frequency of perioperative complications. We had 4%
postoperative dissociative attacks, all of them right temporal resections, all of them
with preoperative borderline personality disorders (Table 18.5).
While Glosser et al., interpret the incidence of nonepileptic attacks as somato-

form disorders, we would regard them as phenomena of dissociation, which enable
the patient to shut out conscious experience temporarily, in distinct situations of
unbearable, overwhelming emotions by using mechanisms comparable to epilep-
tic seizure activity.
Personality disorders a gateway to an individual understanding of patients
After a period of 20 years of neglect, a growing scientic interest has again been
directed to personality disorders during the last decade. One reason is our increas-
ing knowledge about the neurobiological basis of behaviour. Questions about the
demarcation of personality disorders from manifest psychiatric syndromes at one
end and from normal variants of behaviour patterns at the other end have been dis-
cussed, as well as the aetiological components and the predictive value of person-
ality disorders.
According to current psychiatric theory, personality disorders represent endur-
ing patterns of thoughts, emotions, and actions which dier considerably from
expectations of sociocultural surroundings and lead to impairment and suering.
They usually become manifest during childhood and adolescence. Constitutional,
biographic and experience-related conditions are discussed as aetiological factors.
With respect to TLE, there has been an ongoing debate about the epileptic per-
sonality, which has been shown to be more harmful than helpful. Yet, there is much
evidence that epilepsy patients, especially those with a mesiotemporal seizure
focus, show behaviour disturbances, which could partly be seizure-related due to
limbic system hyperactivity and interictal inhibitory mechanisms, partly linked to
the brain lesion itself, and partly be due to the eects of antiepileptic drugs, etc.
(Engel et al., 1991; Chapter 3).
To recapitulate our main results on personality disorders in the surgical context:
First, 60% of our patients with temporal lobe resections had personality disorders;
second, about one-third of all patients with severe personality disorders suered
from postoperative psychiatric deteriorations; third, we had no new psychoses after
surgery without preexisting personality disorder (paranoid features in most cases)
and nally, we had no new dissociative attacks after surgery without preexisting
personality disorders (all borderline type).
These results have implications for our preoperative information to the patients.
However, we can do more than inform patients about their disorders and warn
them about postoperative complications, in order to full the criteria for informed
consent to operation. An analysis of the individual development of a personality
disorder in each single case permits us to gain an insight into the complex struc-
ture of the internal aairs and subjective values of patients and provides hooks for
psychotherapeutic interventions.
How could the development of personality disorders and their neuronal basis be
explained?
Many of our neuronal networks are constructed from birth by the repetitive use of
cognition and emotion. They dier in complexity depending on the variety of
dierent pathways in use. Optimal stimulation enhances the spectrum of possible
reactions, and mechanisms like kindling facilitate the reactive choices. If there are
severe limitations of the capacity or function of neuronal connectivity, the person
will not always be able to respond appropriately to dierentiated situational
demands. Instead, they will repeatedly rely on use of available standard reaction
types. Such stereotype reactions, provoked due to limitations in limbic connectiv-
ity could be the organic basis of personality disorders.
Various factors could lead to such limitations. One may be a temporal lobe epi-
lepsy itself, which provokes intermittent overexcitations within limbic structures,
with the result of a disturbance in processing emotional reactions. It may perhaps
lead to sudden unexplained experiences of fear. The seizure-induced kindling
process of fear may then, as a generalizing reaction, facilitate avoidant behaviour
and lead to an avoidant personality. The same behaviour strategy could develop as
a reaction to punishment-induced fear, or it could be a consequence of feelings of
inferiority in social communication due to severe memory decits, etc. In any of
these cases, each single behaviour of the avoidant strategy however strange it may
seem is selected as the most adaptive of the available alternative reactions, which
are reduced due to functional or structural limbic decits. For that reason the
persons themselves will not understand that their behaviour is judged as inconven-
ient or even as a psychiatric disorder.
Beyond the problem of maladaptive behaviour itself, personality disorders
involve a reduced stress tolerance and a heightened psychic vulnerability, as an
additional result of the limitations due to dysfunctional neuronal connections.
Thus it becomes evident that in so-called stressful life events processing capacities
are easily overwhelmed and the mental system breaks down, which often results in
psychotic decompensations.
For epilepsy patients with personality disorders, the context of surgery itself is
a stressful event. This may facilitate neuronal excitation in unusual directions. In
addition, and supporting the escalating process, the surgical disconnection of
temporal structures forces other parts of the brain to take over functions during
the time of scarring and healing. Thus the postoperative period is a double deli-
cate time-span, involving changes in the cerebral mechanisms of excitation and
inhibition.
Such a model of interaction of psychosocial and neurobiological factors could
be paradigmatic for the development of all psychoses: maladaptive schemata of
action and behaviour, acquired by constitutional and/or experiential faults, are pre-
conditions, which emerge as personality disorders. Under special emotional stress
conditions they easily run into overstimulation, become dysfunctional and end up
in psychotic confusion.
According to this model personality disorders change their status from a cate-
gory of psychiatric diagnosis to meaningful developmental tracks. This may lead to
a better understanding of the very special views and values of these patients.
CASE REPORT
A 40-year-old woman, whose husband has a going concern with the sale of cars, and who
has two children aged 20 and 8, exhibits friendly manners without obvious problems in the
neurological examination. She seems a bit worried especially when in contact with the nurses.
After two seizures at the age of two, epilepsy started at 16, and worsened after her first
pregnancy. She has about 68 seizures per month in two clusters about the time of ovula-
tion and menstruation. The aura contains massive fear of dying. A right mesio-temporal
sclerosis was diagnosed and she was operated on with an optimal prognosis. In fact she
became seizure-free, except for some auras (Engel-classification: class I, category B).
Now to the postoperative psychiatric situation: she had a severe major depression start-
ing shortly after surgery. The use of antidepressants was limited because she refused to take
them after only a few days. She had massive feelings of disgust concerning her husband.
About half a year after surgery she recovered from her depression, but then started to throw
all conventions overboard and showed manic symptoms. Within the following months she
developed the delusional idea of having a love-affair with a neurologist at our centre. She
went from manic symptoms to paranoid-hallucinatory experiences of being influenced
through the internet in her thoughts and emotions. She left her home and was recently hos-
pitalized against her will and put on neuroleptic drugs.
Where are the hints from her biography?
She was raised by her mother under poor social conditions with the message: we are poor
but proud. Her mother then married again and the girl suddenly was confronted with an
aggressive alcoholic stepfather, who had no appreciation for her self-confidence and broke
her will. This was when she started to have seizures, which were accompanied by auras of
massive fear. She left her home with an unstable conception of the world and immediately
became pregnant and married. At that time she had a nervous breakdown and became a
psychiatric inpatient. From then on, shortly after the start of her married life, she tried to hide
her inner world and to fulfil the demands of a good housewife, showing a well-functioning
smiling face to her husband. She also tried to hide her seizures and told people that she suf-
fered from circulatory lability. She suspected neighbours envied her success and invested
much energy in fulfilling their imagined expectations. She blamed herself for being uncon-
trolled and aggressive towards her children.
Her expectations about the time after surgery were those of definitely getting the chance
to be what she always believed she really was: strong, beautiful, self-confident, the way she
had felt before the traumatizing experience with her stepfather.
Wasnt it predictable that the adjustment of her self-confidence would be difficult and
that an overestimation of her own personal capacities could result, when the combined
seizure-induced and trauma-induced neuronal pathways of fear and caution were discon-
nected? That the surgical-induced imbalance of excitation and inhibition could lead to
uncontrolled emotional discharges in her case, which had been bundled in the seizure
activity before?
Couldnt we have foreseen that she would get into trouble with her husband, when she
was seizure-free? That she would search for a dreamlike man who takes care of her and that
this picture could easily be projected to the neurologist who set her free of seizures, which
symbolized the negative time of her life?
Wasnt it predictable that she would try to free herself from barriers and flee from home,
and that she had no real chance to escape except into psychiatry? All this happened.
Maybe we could have saved her from at least some of these traps, if we had invested
more time analysing the case-history before surgery and insisted on more transparency of
her developmental needs, such as the narcissistic feeding by her mother, the unexpected
traumatization by the stepfather with the consequence of starting epilepsy and a global dis-
trust and suspicion against men, and later against everybody. Instead we only diagnosed her
combined narcissistic and partly paranoid personality disorder and recommended psycho-
therapy, which did not happen.
A proposal for psychiatric assessment strategies
It is neither necessary, nor economic, nor possible to carry out such an extensive
analysis for every patient. It would make more sense to have a diagnostic screening
for everybody and then decide who needs a detailed examination of organic, bio-
graphic and situational aspects of their personality in order to discover the individ-
ual traps and try to remove them before surgery, or at least to inform the patient
about possible dangers.
If therapeutic interventions take place, they must include three steps. First, to
nd out and accept the dierent limitations within the individual social and neuro-
nal network. Second, to instruct the patient about the possibility of using dierent,
more adaptive behaviour strategies. Finally, to train the patient to create new path-
ways and extinguish unsuitable old ones.
Unfortunately these possibilities are still seldom attainable.
R E F E R E N C ES
(Third edition, revised) (DSMIIIR). Washington, DC: APA.
Andermann, L.F., Savard, G., Meencke, H.J., McLachlan, R., Mosh, S. and Andermann, F.
(1999). Psychosis after resection of ganglioglioma or DNET: evidence for an association.
Epilepsia, 40, 837.
Anhoury, S., Brown, R.J., Krishnamoorthy, E.S. and Trimble, M.R. (2000). Psychiatric outcome
following temporal lobectomy: a predictive study. Epilepsia, 41, 160815.
Bladin, P.F. (1992). Psychosocial diculties and outcome after temporal lobectomy. Epilepsia, 33,
898907.
Blumer, D. and Altshuler, L.L. (1997). Aective disorders. In Epilepsy: A Comprehensive Textbook,
ed. J. Engel Jr and T.A. Pedley, pp. 28399. Philadelphia, New York: LippincottRaven.
Blumer, D., Wakhlu, S., Davies, K. and Hermann, B.P. (1998). Psychiatric outcome of temporal
lobectomy for epilepsy: incidence and treatment of psychiatric complications. Epilepsia, 39,
47886.
Bruton, C.J. (1988). The Neuropathology of Temporal Lobe Epilepsy. Maudsley Monographs 31.
Oxford: Oxford University Press.
Engel, J. Jr, Bandler, R., Grith, N.C. and Caldecott-Hazard, S. (1991). Neurobiological evidence
for epilepsy-induced interictal disturbances. In Advances in Neurology, Vol. 55, ed. D. Smith,
D. Treiman and M.R. Trimble, pp. 97111. New York: Raven Press.
Engel J. Jr, Van Ness, P.C., Rasmussen, T.B. and Ojemann, L.M. (1993). Outcome with respect to
epileptic seizures. In Surgical Treatment of the Epilepsies, Second edition, ed. J. Engel Jr,
pp. 60921. New York: Raven Press.
Fenwick, P. (1988). Psychiatric assessment and temporal lobectomy. Acta Neurol Scand, 78
(Suppl. 117), 96101.
Fenwick, P. (1994). Psychiatric assessment and temporal lobectomy. In The Surgical Management
of Epilepsy, ed. A.R. Wyler and B.P. Hermann, pp. 21733. London: ButterworthHeinemann.
Fenwick, P., Blumer, D.P., Caplan, R., Savard, G. and Victoro, J.I. (1993). Presurgical psychiat-
ric assessment. In Surgical Treatment of the Epilepsies, Second edition, ed. J. Engel Jr, pp.
27390. New York: Raven Press.
Ferguson, S.M. and Rayport, M. (1965). The adjustment to living without epilepsy. J Nerv Ment
Disease, 140, 2637.
Ferguson, S.M., Rayport, M., Blumer, D.P., Fenwick, P. and Taylor, D.C. (1993). Postoperative
psychiatric changes. In Surgical Treatment of the Epilepsies, Second edition, ed. J. Engel Jr, pp.
64961. New York: Raven Press.
Fraser, R.T. (1988). Improving functional rehabilitation outcome following epilepsy surgery. Acta
Neurol Scand, 78 (Suppl. 117), 1228.
Glosser, G., Roberts, D. and Glosser, D.S. (1999). Nonepileptic seizures after resective epilepsy
surgery. Epilepsia, 40, 17504.
Glosser, G., Zwil, A., Glosser, D.S., OConnor, M.J. and Sperling, M.R. (2000). Psychiatric aspects
of temporal lobe epilepsy before and after anterior temporal lobectomy. J Neurol Neurosurg
Henry, T.R. and Drury, I. (1997). Non-epileptic seizures in temporal lobectomy candidates with
medically refractory seizures. Neurology, 48, 137482.
Hermann, B.P. and Wyler, A.R. (1989). Depression, locus of control, and the eects of epilepsy
surgery. Epilepsia, 30, 3328.
Hill, D., Pond, D.A., Mitchell. W. and Falconer, M.A. (1957). Personality changes following tem-
poral lobectomy for epilepsy. J Ment Sci, 103, 1827.
Jensen, I. and Larsen, J.K. (1979). Mental aspects of temporal lobe epilepsy. J Neurol Neurosurg
Kanemoto, K., Kawasaki, J. and Kawai, I. (1996). Postictal psychosis: a comparison with acute
Kanemoto, K., Kawasaki, J. and Mori, E. (1998). Postictal psychosis as a risk factor for mood dis-
orders after temporal lobe surgery. J Neurol Neurosurg Psychiatry, 65, 5879.
Koch-Stoecker, S. (1997). Psychotische phnomene bei patienten mit operativ behandelten tem-
porallappenepilepsien. Epilepsie-Bltter, 10, 325.
Koch-Stoecker, S. (2001). Psychiatric outcome. In Epilepsy Surgery, Second edition, ed. H. Lders
and Y. Comair, pp. 83744. Philadelphia: Lippincott Williams & Wilkins.
Koch-Weser, M., Garron, D.C., Gilley, D.W. et al. (1988). Prevalence of psychologic disorders
after surgical treatment of seizures. Arch Neurol, 45, 130811.
Krahn, L.E., Rummans, T.A. and Peterson, G.C. (1996). Psychiatric implications of surgical treat-
ment of epilepsy. Mayo Clin Proc, 71, 12014.
Mace, C.J. and Trimble, M.R. (1991). Psychosis following temporal lobe surgery: a report of six
cases. J Neurol Neurosurg Psychiatry, 54, 63944.
Manchanda, R., Schaefer, B., McLachlan, R.S. et al. (1996). Psychiatric disorders in candidates for
surgery for epilepsy. J Neurol Neurosurg Psychiatry, 61, 829.
Naylor, A.S., Rogvi-Hansen, B., Kessing, L. and Kruse-Larsen, C. (1994). Psychiatric morbidity
after surgery for epilepsy. J Neurol Neurosurg Psychiatry, 57, 137581.
Ney, G., Barr, W.B., Napolitano, C., Decker, R. and Schaul, N. (1998). New-onset psychogenic
seizures after surgery for epilepsy. Arch Neurol, 55, 72630.
Polkey, C.E. (1983). Eects of anterior lobectomy apart from relief of seizures: a study of 40
patients. J R Soc Med, 76, 3548.
Ring, H.A., Moriarty, J. and Trimble, M.R. (1998). A prospective study of the early postsurgical
psychiatric associations of epilepsy surgery. J Neurol Neurosurg Psychiatry, 64, 6014.
Savard, G. (1991). Psychosis and surgery of epilepsy. In Epilepsy Surgery, ed. H. Lders, pp. 4615.
New York: Raven Press.
Savard, G., Andermann, L.F., Reutens, D. and Andermann, F. (1998). Epilepsy, surgical treatment
and postoperative psychiatric complications: a re-evaluation of the evidence. In Forced
Normalization and Alternative Psychoses of Epilepsy, ed. M.R. Trimble and B. Schmitz, pp.
17992. Peterseld and Bristol: Wrightson Biomedical Publishing.
Taylor, D.C. (1972). Mental state and temporal lobe epilepsy. A correlative account of 100 patients
treated surgically. Epilepsia, 13, 72765.
Taylor, D.C. (1987). Psychiatric and social issues in measuring the input and outcome of epilepsy
surgery. In Surgical Treatment of the Epilepsies, ed. J. Engel Jr, pp. 485503. New York: Raven Press.
Taylor, D.C., Neville, B.G.R. and Cross, J.H. (1997). New measures of outcome needed for the
surgical treatment of epilepsy. Epilepsia, 38, 62530.
Trimble, M.R. (1992). Behavior changes following temporal lobectomy, with special reference to
psychosis. J Neurol Neurosurg Psychiatry, 55, 8991.
Trimble, M.R. and Schmitz, B. (1997). The psychoses of epilepsy/schizophrenia. In Epilepsy: A

Comprehensive Textbook, ed. J. Engel and T.A. Pedley, pp. 207181. Philadelphia, New York:
Lippincott-Raven.
Umbricht, D., Degreef, G., Barr, W.B., Lieberman, J.A., Pollack, S. and Schaul, N. (1995). Postictal
and chronic psychoses in patients with temporal lobe epilepsy. Am J Psychiatry, 152, 22431.
Wolf, P. (1982). Manic episodes in epilepsy. In Advances in Epileptology: 8th Epilepsy International
Symposium, ed. H. Akimoto, H. Kazamatsuri, M. Seino and A.A. Ward Jr, pp. 23740. New
York: Raven Press.
19
Vagus nerve stimulation and mood

Christian E. Elger and Christian Hoppe
Department of Epileptology, University of Bonn, Germany
Introduction
When asked for their connotations with the vagus nerve, most medically trained
people will think of the parasympathetic eerent tasks of cranial nerve X: heart rate
modulation, regulation of ingestion/digestion, eects on lung functioning and so
on. And these connotations are what gave this nerve its name: vagus, latin the wan-
derer. However, the vagus nerve is more a sensory than a motor nerve, since it has
about 80% aerent but only 20% eerent bres (Foley and DuBois, 1937). First
reports on the cerebral eects of an electrical stimulation of the vagus were pub-
lished by Bailey and Bremer in 1938. In the following decades, researchers revealed
that vagus nerve stimulation (VNS) may inuence surface EEG, suppress epileptic
electrical activity in the brain, and even terminate seizures in animal models of epi-
lepsy (Zabara, 1985, 1992). The rst single-patient trial on VNS for treatment of
epilepsy was set up in 1988 in the USA (Penry and Dean, 1990).
The electrical stimulation of the left vagus nerve trunk by a totally implanted
stimulation device (NCPTM-system, Cyberonics Inc.) was approved for treatment
of drug-resistant epileptic seizures by the FDA in 1997 and by the European
Community in 1994 (Schachter and Saper, 1998). The pulse generator is implanted
into the chest wall, in a similar fashion to cardiac pacemakers, while the spiral plat-
inum electrodes are attached to the left vagus nerve trunk below the cardiac branch.
The pulse generator may be programmed telemetrically by a programming wand
that is held over the generator and connected to a portable computer. At standard
settings, the stimulator would deliver electrical pulses to the vagus every 5 minutes
for about 30 seconds (pulse frequency: 2030 Hz, pulse width: 250500 s).
Alternatively, rapid cycles with 12 seconds o-stimulation time and 7 seconds on-
stimulation time may be programmed.
The treatment usually gets started with an output currency of 0.25 mA after
implantation which can be stepwise increased during the next weeks depending on
the seizure outcome and adverse side eects (maximum: 3.5 mA). Several studies
have shown eectiveness (Amar et al., 1999; Ben-Menachem et al., 1994; DeGiorgio
283
284 C.E. Elger and C. Hoppe
et al., 2000; Handforth et al., 1998), safety (Annegers et al., 1998; Fisher and
Handforth, 1999; Ramsay et al., 1994), and sucient costbenet ratios (Boon et
al., 1999) of VNS for patients with intractable seizures.
While the antiseizure eect of VNS is usually attributed to the aerent, that is,
direct cerebral eects, the most common adverse eects such as hoarseness, cough,
or throat paraesthesia, are supposed to result from the eerent portion of stimula-
tion which is unavoidable since the entire nerve is stimulated. Regarding the issues
reviewed here, one should keep in mind that VNS may achieve its eects by an eer-
ent peripheral mechanism as well.
Anxiety and depressive disorders are common psychiatric conditions in patients
with epilepsy (Jacoby et al., 1996; Kohler et al., 1999). About one-third to one-half
of patients score high on anxiety and depression self-report scales, but only one-
third of the aected patients are recognized by general practitioners to have psychi-
atric problems (ODonoghue et al., 1999). Depressive mood states and poor quality
of life are part of a complex interplay of clinical measures (e.g. seizure frequency,
seizure severity, epilepsy duration, age at onset) and psychosocial parameters
(employment, marital status) (Jacoby et al., 1996; Roth et al., 1994; Smith et al.,
1991). However, depression in epilepsy patients may not be fully accounted for by
either clinical or psychosocial factors since biological mechanisms involved in epi-
leptogenesis may also contribute to depression (Hermann et al., 1996; Schmitz et
al., 1999). Therefore, seizure outcome is only one even though leading outcome
measure of epilepsy treatment. Psychiatric aspects of epilepsy have to be considered
and new drugs and methods, as for example VNS, have to be evaluated for their
benets regarding mood and quality of life as well as any eect on seizures.
Mood improvements by VNS in epilepsy patients
Reports from the early randomized controlled trials on VNS for epilepsy treatment
(EO3, EO5) suggested improved quality of life in a majority of patients (Ben-
Menachem et al., 1994; Handforth et al., 1998): At the 14-weeks follow-up, about
5060% of the patients stated that their quality of life has improved since implanta-
tion. From a psychometric point of view, reliability and validity of these data were
questionable and these reports had to be considered as preliminary. But they initiated
some studies of this phenomenon which will be reviewed in the following section.
Harden et al. (2000) recently published a study on VNS and mood in which 20
epilepsy patients under VNS and 20 control patients were enrolled (a nonrandom-
ized, nonblinded, controlled clinical trial). Mood outcome measures were scores
from standard psychiatric rating scales (Cornell Dysthymia Rating Scale; Mason et
al., 1993; Hamilton Depression Rating Scale/Hamilton Anxiety Rating Scale;
Hamilton, 1960) and from the Beck Depression Inventory (Beck, 1967), an estab-
285 Vagus nerve stimulation and mood
lished self-report questionnaire. Signicant improvements of depression scores

from baseline to follow-up were observed only in the VNS treatment group. A slight
decrement in the score of the Hamilton Anxiety Rating Scale was not signicant.
No signicant mood changes were observed within the control group. No between-
group dierences were obtained at baseline or follow-up and the authors failed
to demonstrate a signicant interaction eect (group time) to conrm their
hypothesis; only for the BDI score did the interaction eect approach signicance
(P0.07). From the within-group changes which were limited to the VNS group,
Harden et al. (2000) concluded that their ndings, for the rst time, showed
improvements of depressed mood in epilepsy patients during VNS treatment by
means of psychometrically evaluated measures.
Elger et al. (2000) completed the EO3 study on seizure outcome of VNS, in which
their unit participated with some patients in 1993, with a comprehensive psychiat-
ric evaluation. This international multisite outcome study included 14 weeks of a
randomized control trial (RCT). Patients were randomly assigned to a low or high
stimulation condition (dose-eect study). In order to give patients from both
groups the feeling of participating in an optimal treatment condition, patients
from the low stimulation group were told that VNS has optimal eects when sub-
jects are just able to recognize the signal. In contrast, patients from the high stim-
ulation group were informed that VNS should be maximized according to
subjective tolerability of adverse eects (with 1.75 mA as the predened
maximum). Seizure and psychiatric data were recorded 4 weeks before implanta-
tion (baseline) and at the 3- and 6-month follow-ups. Medication was unchanged
during the entire duration of the study. From our unit, 11 patients with severe
drug-resistant epileptic seizures enrolled in the EO3 study and agreed to partici-
pate in the psychiatric study as well. An experienced psychiatrist, to whom the stim-
ulation conditions were masked, completed several standard rating scales, such as
the Montgomerysberg Depression Rating Scale (MADRS; Montgomery and
sberg, 1979) and the Scale for the Assessment of Negative Symptoms (SANS;
Andreasen, 1981). Figure 19.1 depicts improvements of depressive mood states
during VNS treatment as revealed by the MADRS.
Despite the small sample size, and consequently small statistical power, MADRS
changes from baseline to the 6-month follow-up were signicant in the total
sample (nonparametrical ANOVA, Friedman tests: P0.05; post hoc pairwise anal-
ysis by Wilcoxon test: P0.05). Interestingly, patients from the high and low stim-
ulation group tended to experience dierent courses of mood improvements: in the
high stimulation group mood improvements appeared earlier and were already
seen at the 3-month follow-up and sustained until the 6-month follow-up. Patients
from the low stimulation group experienced signicant mood improvements, par-
ticularly after nishing the RCT, that is, after output currencies were increased and
ns ns
Low
High
Total
P 0.10; *P 0.05
pre VNS 3 months post 6 months post

Figure 19.1. Improvement of depressed mood during VNS treatment: high (n6) vs. low stimulation
(n5).
adapted to the other patients during the second ramp-up period. At the 3-month
follow-up, group dierences between the MADRS scores approached signicance
(MannWhitney test: P0.10). This nding is consistent with the notion of a
(partial) VNS doseresponse relationship which may indicate specicity of the
observed eects. In addition, data revealed a (partial) independence of mood
improvements and the antiseizure eect of VNS: 9 of 11 patients presented mood
responses whereas only 3 of 11 patients had about 50% reductions in seizure fre-
quency, that is, mood improvements in 6 of 9 patients could not be attributed to
improved seizure control.
Both studies have to be regarded as preliminary and the most critical methodo-
logical aspects, such as a possible rater bias due to missing attempts of masking the
experimental conditions in the Harden et al. (2000) study and the missing control
group in the Elger et al. (2000) study, are carefully discussed in both papers. From
a rigorous methodological point of view, both studies fail to denitely prove the
specicity of the observed eect and to exclude a mere placebo eect. However, it
is questionable whether this issue can be taken further: blinding is dicult in sur-
gical treatments such as VNS; arranging an appropriate group of control patients
is very questionable; masking VNS patients to their stimulation condition appears
almost impossible; and it is unclear whether a placebo condition needs to include
surgery as well. Furthermore, even a controlled doseresponse study may be di-
cult to perform since todays patients are probably well informed about the eects
of low and high stimulation, due to comprehensive patient information.
In an on-going study we are trying to circumvent these problems by using self-

report questionnaires: here patients are asked to evaluate their present mood states
(e.g. last 4 weeks). At the follow-up, 6 months after implantation, they are pre-
sumed to be unable to remember their former answers to single items. Mood score
changes can be compared to normative data as recorded during construction of the
questionnaire, e.g. to exclude a mere regression to the mean eect.
VNS as an antidepressant treatment in nonepileptic patients
From the evidence reported above it is a small step to the evaluation of VNS for
treatment of clinical depressions in nonepileptic patients. Recently, Rush et al.
(2000) published their ndings from a rst single-arm study on this issue. On the
basis of ethical considerations, only the most aected patients from dierent clinics
in Dallas were included and provided with a vagus nerve stimulator (n30).
Inclusion criteria were a DSMIV diagnosis of major depression disorder (MDD),
bipolar I or bipolar II disorder (American Psychiatric Association, 1994). Patients
had to be in a major depressive episode (MDE) which either was lasting 2 years
or which was one of at least four MDEs during life. Finally, they had to have failed
on at least two antidepressant medication treatments from dierent medication
groups during the current MDE.
The study design was as follows: the baseline period was up to 4 weeks before
implantation. Implantation was followed by a 2-weeks recovery period with the
stimulation device turned o. During these weeks patients were left unclear about
whether the device was already turned on or not. During the next 2 weeks output
currency was increased stepwise to the maximum tolerated level (ramp-up period).
In the following 8 weeks stimulation was left unchanged (xed-stimulation
period). After implantation patients were evaluated for mood states every week.
The main outcome score was the Hamilton Depression Rating Scale (28-item
version, HDRS-28; Hamilton, 1960, 1967). A subsequent open clinical observation
period lasted at least 9 months in all patients reported on.
No eects were obtained during the rst 2 weeks after implantation (recovery
period with inactive stimulator). Using a 50% reduction in the HDRS-28 total
score to dene response, a 40% response rate was found at the end of the 8-weeks
xed stimulation period. Complete response (HDRS-2810) was observed in 17%
of the patients. Follow-up data from the open clinical observation period showed
sustained mood improvement in all responders. In some cases response was further
increased so that at the end of the reported observation period, 7 of 10 responders
from the acute study demonstrated complete response (23% of the total sample).
Since the depressions were very severe and chronic, a mere placebo eect was
very unlikely to occur; even with the lack of a control group and despite the small
sample size these ndings are intriguing and justify further investigations. From a
clinical point of view, these data raise hope for a totally new approach in the treat-
ment of severe clinical depressions, and a multisite double-blind randomized
control trial is underway in the USA in which more than 200 severely depressed
patients will enrol.
Cerebral mechanisms of action
A further development of the indications and the eectiveness of VNS depends on

a comprehensive understanding of the mechanisms of action underlying VNS.
Theoretical considerations have to rely on the anatomical conditions of the vagus
(George et al., 2000; Rutecki, 1990), functional imaging studies (Henry et al., 1998,
1999; Vonck et al., 2000), studies on the neurochemical eects of VNS in epilepsy
patients (Ben-Menachem et al., 1995; Hammond et al., 1992b; Naritoku et al., 1995;
Walker et al., 1999), and behavioural studies on VNS (Clark et al., 1998, 1999).
Most researchers assume direct cerebral mechanisms while possible peripheral
parasympathetic eects are widely ignored.
Several studies with dierent methodological approaches have demonstrated
that VNS may bilaterally activate a wide range of brain regions, as to be expected
from the anatomical relations of the vagus: the nucleus of the solitary tract (NTS)
as the primary target of aerent vagal input; brain stem regions in the direct vicin-
ity of NTS; thalamus, hypothalamus, amygdala, hippocampus and isocortex. These
brain regions are particularly relevant in the context of both epileptogenesis and
neuropsychiatric disorders. The details of activation remain somewhat unclear
since dierent studies report slightly dierent eects, e.g. regarding a possible lat-
eralization of the obtained activations. The behavioural studies of Clark et al.
(1998, 1999) have demonstrated dose-dependent enhancement eects of VNS on
retention and recognition performance in animals and epilepsy patients, and thus
shown the functional impact of these activations.
From a clinical point of view, there is the striking convergence of anticonvulsant
and antidepressant treatments (Post et al., 1992; Trimble, 1998): several antiepilep-
tic drugs (AEDs), for example carbamazepine (Okuma et al., 1973), valproate
(Swann et al., 1997), gabapentin (Harden et al., 1999; Letterman and Markowitz,
1999), and lamotrigine (Calabrese et al., 1999; Suppes et al., 1999), have been
shown to be eective also in depression. Vice versa, electroconvulsive therapy
(ECT), as one of the most ecacious therapies for drug-resistant depressions
(Olfson et al., 1998) is also known for its antiseizure eect (Regenold et al., 1998;
Sackheim, 1999; Sackheim et al., 1983). Thus, one may assume biochemical mech-
anisms which underlie both epileptogenesis and mood disorders, and which are
aected and tuned by AEDs and ECT and potentially VNS.
Several mechanisms which could explain the anticonvulsant and/or the antide-
pressant eect of VNS have been proposed (George et al., 2000; Harden et al., 2000).
Noradrenergic system
Naritoku et al. (1995) investigated regional c-Fos immunoreactivity and reported
activation of the locus coeruleus (LC) by therapeutic VNS. By means of lesion
studies in animal models of epilepsy, Krahl and coworkers have demonstrated that
the antiseizure eect of injected norepinephrine depends on the vagus (Krahl et al.,
2000) and that the seizure-attenuating eect of VNS is mediated by and totally
depends on LC activity (Krahl et al., 1998). The locus coeruleus is the major origin
of the noradrenergic system in the brain and has projections to brain regions which
are involved in both mood regulation and epileptogenesis (e.g. thalamus, hippo-
campus, amygdala and isocortex). Norepinephrine has an inhibitory inuence on
postsynaptic neurons which may explain the antiseizure eect of activating the LC
by VNS. At the same time, the noradrenergic system is involved in neuropsychiat-
ric disorders like depression: one major eect of tricyclic antidepressant medica-
tion is to increment the noradrenergic tone of the brain (Schatzberg and
Schildkraut, 1996). Thus, roughly speaking, VNS is supposed to stimulate LC
neurons, which in turn increase the delivery of norepinephrine which is supposed
to be an endogenous antidepressant and anticonvulsant.
Serotonergic system
Another hypothesis, with some evidence from neurochemical studies on VNS in
epilepsy patients (Ben-Menachem et al., 1995), relies on the role of the serotoner-
gic system in mood regulation. From the NTS there are direct connections to the
raphe nuclei which are the main and nearly the only origin of the cerebral seroto-
nergic system. Changes in the serotonergic tone of the brain are clearly associated
with mood changes.
Cortical inhibition
From clinical experience with ECT, a third hypothesis may be derived: inhibition of
cortical activity, as the major mechanism of ECT (Sackheim et al., 1996), may
enhance depressive mood states and may improve seizure control. Since VNS seems
to produce rather similar eects as ECT, one may speculate also that VNS increases
the inhibitory inuences on cortical tone. However, evidence conrming this specu-
lation and replication of the ndings from ECT is missing (Hammond et al., 1992a).
GABA and glutamate

The possible role of GABA and glutamate transmission from the NTS was high-
lighted by an animal study of Walker et al. (1999). These authors tested four
substances for their potential to attenuate seizures in standard animal models of

epilepsy (bicuculline methiodide, pentylenetetrazol) when they were microin-
jected into the mediocaudal NTS (mNTS): a GABA-A receptor agonist (musci-
mol), a GABA-a receptor antagonist (bicuculline methiodide), a glutamate
receptor antagonist (kynurenate) or a local anaesthetic (lidocaine). The ndings
revealed that increased GABA or decreased glutamate transmission from the NTS
reduces susceptibility for limbic seizures. This may provide a possible mechanism
for seizure attenuation due to VNS as well as for other secondary limbic alterations
which may aect mood regulation.
Peripheral mechanisms
Since the vagus is a mixed nerve, VNS always comprises a portion of eerent stim-
ulation which may alter peripheral functions. For instance, it may cause hoarseness
as the most common adverse eect of VNS. As a matter of course, any changes con-
cerning mood or epileptic seizures must be due to cerebral changes. However, from
a theoretical point of view one has good reasons to expect that peripheral changes
induced by eerent VNS may in turn result in cerebral changes relevant for the
issues discussed here.
In mammalians the eerent branch of the vagus plays a decisive role in emotion
regulation and expression (Porges 1997; Porges et al., 1994). Furthermore, the
vagus is supposed to coordinate and protect the organisms metabolic resources,
e.g. by retarding heart rate more or less (vagal brake) (Porges, 1995). This more
theoretical view is conrmed by clinical observations in neuropsychiatric disorders
such as depression and anxiety which reveal clear associations between mood and
parasympathetic functions (Glassman, 1998; Lehofer et al., 1997, 1999). Diurnal
mood variations in some depressed patients may be associated with parasympa-
thetic activity (Rechlin et al., 1995). Regarding cardiac measures, it is noteworthy
that there is no evidence for altered vagal tone in unmedicated clinical depressions
but for increased sympathetic tone (heart rate) which may be due to increased
anxiety in depressed patients (Lehofer et al., 1997; Yeragani et al., 1991).
Interestingly, experimentally induced panic attacks (hyperventilation, sodium
lactate administration) are accompanied by an attenuated vagal tone (George et al.,
1989) suggesting that anxiety disorders may be even more susceptible for VNS
treatment than depressions. Other authors have also suggested a linkage between
vagal functions and anxiety disorders (Watkins et al., 1998).
In animal experiments, one can transiently block eerent neural transmission by
a lidocaine injection below the point of electrical stimulation (Brodin, 1985). In
such an experiment, Clark et al. (1998) could show that eects on retention and
recognition exclusively resulted from the aerent portion of VNS. Investigating the
role of eerent vagal transmission in patients is dicult. One would have to record
peripheral physiological measures and consider them as covariates during data
analysis. Even if there is no evidence for general alterations of cardiac or gastroin-
testinal functions due to VNS in the sense of adverse eects (Ramsay et al., 1994),
peripheral changes induced by VNS may be small and more dicult to register.
In the Elger et al. (2000) study, mood improvements were particularly expressed
in a reduction of negative symptoms as recorded by the Scale for the Assessment of
Negative Symptoms (Andreasen, 1981) or by the anergia scale of the Brief
Psychiatric Rating Scale (Overall and Gorham, 1962). We propose that negative
symptoms and particularly anergia may be interpreted as a lack of energy in which
the autonomic nervous system and particularly the vagus may be involved.
Preliminary data of our on-going self-report questionnaire study suggest that
VNS improves anxiety and unpleasant exertion as recorded by the Self-Rating
Anxiety Scale (Zung, 1971) or the BendlichkeitsSkala (Zerssen et al., 1970). In
contrast, improvement of depressed mood, which was measured by the Beck
Depression Inventory, appears to be a smaller eect. One has to consider that this
self-report questionnaire particularly accounts for higher cognitive and emotional
aspects of depression. Therefore, we assume that eerent VNS may contribute to
mood improvements, rst and more unspecically, by tuning the basic autonomic
balance and the vagal management of metabolic resources, or second and more
specically, by attenuating sympathetic tone and peripheral symptoms of anxiety.
Finally, we would like to allude to some theoretical diculties associated with the
fact that the vagus is more part of a system than two one-way routes: VNS has an
impact on the entire vagal brainperiphery feedback loop and electrical stimula-
tion aects signalling in both directions. A vagus under VNS may make the brain
think that peripheral functions have changed even if they actually have not, that
is, even if no objective changes can be revealed by psychophysiological measure-
ments. Such a mechanism could be described as virtually peripheral. Conversely,
VNS may distort or suggest commands coming from the brain which are to be
transmitted to the periphery by the vagus. This virtually cerebral mechanism
results in peripheral eects, as for example hoarseness. Studies on the alterations of
neural transmission within the vagus as induced by VNS would be required. So far,
the articial stimulation of the vagal system by VNS with its unphysiological duty
cycles, output currents and pulse frequencies has to be regarded as very coarse. In
fact, some authors assume that this is the true reason why more serious cardiac side
eects do not occur in patients under VNS (George et al., 2000). A better under-
standing of vagal neurotransmission will provide the basis for more subtle, more
adaptive and hopefully even more eective brain stimulation techniques in the
future. VNS is probably the promising beginning of this intriguing development
and an important scientic tool for human research on these issues.
R E F E R E N C ES
Amar, A.P., DeGiorgio, C.M., Tarver, W.B. and Apuzzo, M.L. (1999). Long-term multicenter
experience with vagus nerve stimulation for intractable partial seizures: results of the XE5 trial.
Stereotact Funct Neurosurg, 73, 1048.
(Fourth edition) (DSMIV). Washington, DC: APA.
Andreasen, N.C. (1981). Scale for the Assessment of Negative Symptoms (SANS). Iowa City:
University of Iowa.
Annegers, J.F., Coan, S.P., Hauser, W.A., Leetsma, J., Duell, W. and Tarver, B. (1998). Epilepsy,
vagal nerve stimulation by the NCP system, mortality, and sudden, unexpected, unexplained
death. Epilepsia, 39, 20612.
Bailey, P. and Bremer, F. (1938). A sensory cortical representation of the vagus nerve. J
Neurophysiology, 1, 40512.
Beck, A.T. (1967). Depression: Clinical, Experimental, and Theoretical Aspects. New York: Hoeber.
Ben-Menachem, E., Manon Espaillat, R., Ristanovic, R. et al. (1994). Vagus nerve stimulation for
treatment of partial seizures, 1. A controlled study of eect on seizures (First International
Vagus Nerve Stimulation Study Group). Epilepsia, 35, 61626.
Ben-Menachem, E., Hamberger, A., Hedner, T. et al. (1995). Eects of vagus nerve stimulation
on amino acids and other metabolites in the CSF of patients with partial seizures. Epilepsy Res,
20, 2217.
Boon, P., Vonck, K., DHave, M., OConnor, S., Vandekerckhove, T. and De Reuck, J. (1999). Cost-
benet of vagus nerve stimulation for refractory epilepsy. Acta Neurol Belg, 99, 27580.
Brodin, P. (1985). Dierential inhibition of A, B and C bres in the rat vagus nerve by lidocaine,
eugenol and formaldehyde. Arch Oral Biol, 30, 47780.
Calabrese, J.R., Bowden, C.L., Sachs, G.S., Ascher, J.A., Monaghan, E. and Rudd, G.D. (1999). A
double-blind placebo-controlled study of lamotrigine monotherapy in outpatients with
bipolar I depression. Lamictal 602 Study Group. J Clin Psychiatr, 60, 7988.
Clark, K.B., Smith, D.C., Hassert, D.L., Browning, R.A., Naritoku, D.K. and Jensen, R.A. (1998).
Posttraining electrical stimulation of vagal aerents with concomitant vagal eerent inactiva-
tion enhances memory storage processes in the rat. Neurobiol Learning Memory, 70, 36473.
Clark, K.B., Naritoku, D.K., Smith, D.C., Browning, R.A. and Jensen, R.A. (1999). Enhanced rec-
ognition memory following vagus nerve stimulation in human subjects. Nature Neurosci, 2,
948.
DeGiorgio, C.M., Schachter, S.C., Handforth, A. et al. (2000). Prospective long-term study of
vagus nerve stimulation for the treatment of refractory seizures. Epilepsia, 41, 1195200.
Elger, G., Hoppe, C., Falkai, P., Rush, A.J. and Elger, C.E. (2000). Vagus nerve stimulation is asso-
ciated with mood improvements in epilepsy patients. Epilepsy Res, 42, 20310.
Fisher, R.S. and Handforth, A. (1999). Reassessment: Vagus nerve stimulation for epilepsy: A
report of the Therapeutics and Technology Assessment Subcommittee for the American
Academy of Neurology. Neurology, 53, 6669.
Foley, J.O. and DuBois, F. (1937). Quantitative studies of the vagus nerve in the cat. I. The ratio
of sensory and motor studies. J Comp Neurol, 67, 4967.
George, D.T., Nutt, D.J., Walker, W.V., Porges, S.W., Adino, B. and Linnoila, M. (1989). Lactate
and hyperventilation substantially attenuate vagal tone in normal volunteers. Arch Gen
George, M.S., Sackheim, H.A., Rush, J.A. et al. (2000). Vagus nerve stimulation: a new tool for
brain research and therapy. J Biol Psychiatry, 47, 28795.
Glassman, A.H. (1998). Depression, cardiac death, and the central nervous system.
Neuropsychobiology, 37, 803.
Hamilton, M. (1960). A rating scale for depression. J Neurol Neurosurg Psychiatry, 12, 3719.
Hamilton, M. (1967). Development of a rating scale for primary depressive illness. Br J Soc Clin
Psychol, 6, 27896.
Hammond, E.J., Uthman, B.M., Reid, S.A. and Wilder, B.J. (1992a). Electrophysiological studies
of cervical vagus nerve stimulation in humans: I. EEG eects. Epilepsia, 33, 101320.
Hammond, E.J., Uthman, B.M., Wilder, B.J. et al. (1992b). Neurochemical eects of vagus nerve
stimulation in humans. Brain Res, 583, 3003.
Handforth, A., DeGiorgio, C.M., Schachter, S.C. et al. (1998). Vagus nerve stimulation therapy
for partial-onset seizures: a randomized active-control trial. Neurology, 51, 4855.
Harden, C.L., Lazar, L.M., Pick, L.H. et al. (1999). A benecial eect on mood in partial epilepsy
patients treated with gabapentin. Epilepsia, 40, 112934.
Harden, C.L., Pulver, M.C., Ravdin, L.D., Nikolov, B., Halper, J.P. and Labar, D.R. (2000). A pilot
study of mood in epilepsy patients treated with vagus nerve stimulation. Epilepsy Behav, 1,
939.
Henry, T.R., Bakay, R.A.E., Votaw, J.R. et al. (1998). Brain blood ow alterations induced by ther-
apeutic vagus nerve stimulation in partial epilepsy: acute eects at high and low levels of stim-
ulation. Epilepsia, 39, 98390.
Henry, T.R., Votaw, J.R., Pennell, P.B. et al. (1999). Acute blood ow changes and ecacy of vagus
nerve stimulation in partial epilepsy. Neurology, 52, 116673.
Hermann, B.P., Trenerry, M.R. and Colligan, R.C. (1996). Learned helplessness attributional style
and depression in epilepsy (Bozeman Epilepsy Surgery Consortium). Epilepsia, 37, 6806.
Jacoby, A., Baker, G.A., Stehen, N., Potts, P. and Chadwick, D.W. (1996). The clinical course of
epilepsy and its psychosocial correlates: ndings from a UK Community study. Epilepsia, 37,
14861.
Kohler, C., Norstrand, J.A., Baltuch, G. et al. (1999). Depression in temporal lobe epilepsy before
epilepsy surgery. Epilepsia, 40, 33640.
Krahl, S.E., Clark, K.B., Smith, D.C. and Browning, R.A. (1998). Locus coeruleus lesions suppress
the seizure attenuating eects of vagus nerve stimulation. Epilepsia, 39, 70914.
Krahl, S.E., Senanayake, S.S. and Handforth, A. (2000). Seizure suppression by systemic epineph-
rine is mediated by the vagus nerve. Epilepsy Res, 38, 1715.
Lehofer, M., Moser, M., Hoehn-Saric, R. et al. (1997). Major depression and cardiac autonomic
control. Biol Psychiatry, 42, 91419.
Lehofer, M., Moser, M., Hoehn-Saric, R. et al. (1999). Inuence of age on the parasympatholytic
property of tricyclic antidepressants. Psychiatry Res, 85, 199207.
Letterman, L. and Markowitz, J.S. (1999). Gabapentin: a review of published experience in the
treatment of bipolar disorder and other psychiatric conditions. Pharmacotherapy, 19, 56572.
Mason, B., Kocsis, J., Leon, A. et al. (1993). Measurement of severity and treatment response in
dysthymia. Psychiatr Ann, 23, 62531.
Montgomery, S.A. and sberg, M. (1979). A new depression scale designed to be sensitive to
change. Br J Psychiatry, 134, 3829.
Naritoku, D.K., Terry, W.J. and Helfert, R.H. (1995). Regional induction of Fos immunoreactiv-
ity in the brain by anticonvulsant stimulation of the vagus nerve. Epilepsy Res, 22, 5362.
ODonoghue, M.F., Goodridge, D.M., Redhead, K., Sander, J.W. and Duncan, J.S. (1999).
Assessing the psychosocial consequences of epilepsy. A community-based study. Br J Gen
Pract, 49, 21114.
Okuma, T., Kishimoto, A., Inoue, K., Matsumoto, H. and Ogura, A. (1973). Anti-manic and pro-
phylactic eects of carbamazepine (Tegretol) on manic depressive psychosis. A preliminary
report. Folia Psychiatr Neurol Jpn, 27, 28397.
Olfson, M., Marcus, S., Sackheim, H.A., Thophson, J. and Pincus, H.A. (1998). Use of ECT for
the inpatient treatment of recurrent major depression. Am J Psychiatry, 155, 229.
Overall, J.E. and Gorham, D.R. (1962). The Brief Psychiatric Rating Scale. Psychol Rep, 10,
799812.
Penry, J.K. and Dean, J.C. (1990). Prevention of intractable partial seizures by intermittent vagal
stimulation in humans: preliminary results (Abstract). Epilepsy, 31 (Suppl.), S403.
Porges, S.W. (1995). Orienting in a defensive world: mammalian modications of our evolution-
ary heritage. A polyvagal theory. Psychophysiology, 32, 30118.
Porges, S.W. (1997 ). Emotion: an evolutionary by-product of the neural regulation of the auto-
nomic nervous system. Ann New York Acad Sci, USA, 807, 6277.
Porges, S.W., Doussard-Roosevelt, J.A. and Maiti, A.K. (1994). Vagal tone and the physiological
regulation of emotion. Monogr Soc Res Child Dev, 59, 16786.
Post, R.M., Weiss, S.R.B. and Chuang, D.M. (1992). Mechanisms of action of anticonvulsant in
aective disorders: comparisons with lithium. J Clin Psychopharmacol, 12 (Suppl. 1), S2335.
Ramsay, R.E., Uthman, B.M., Augustinsson, L.E. et al. (1994). Vagus nerve stimulation for treat-
ment of partial seizures, 2. Safety side eects and tolerability (First International Vagus Nerve
Stimulation Study Group). Epilepsia, 35, 62736.
Rechlin, T., Weis, M. and Kaschka, W.P. (1995). Is diurnal variation of mood associated with
parasympathetic activity? J Aect Disord, 34, 24955.
Regenold, W.T., Weintraub, D. and Taller, A. (1998). Electroconvulsive therapy for epilepsy and
major depression. Am J Geriatr Psychiatry, 6, 1803.
Roth, D.L., Goode, K.T., Williams, V.L. and Faught, E. (1994). Physical exercise, stressful life
experience and depression in adults with epilepsy. Epilepsia, 35, 124855.
Rush, A.J., George, M.S., Sackheim, H.A. et al. (2000). Vagus nerve stimulation (VNS) for treat-
ment-resistant depression, a multicenter study. J Biol Psychiatry, 47, 27686.
Rutecki, P. (1990). Anatomical, physiological, and theoretical basis for the antiepileptic eect of
vagus nerve stimulation. Epilepsia, 31, S16.
Sackheim, H.A. (1999). The anticonvulsant hypothesis of the mechanism of action of ECT,
current status. J ECT, 15, 526.
Sackheim, H.A., Decina, P., Malitz, S., Reesor, S.R. and Prohovnik, I. (1983). Anticonvulsant and
antidepressant properties of electroconvulsive therapy, a proposed mechanism of action. Biol

Sackheim, H.A., Luber, B., Katzman, G.P. et al. (1996). The eects of electroconvulsive therapy
on quantitative electroencephalograms, relationship to clinical outcome. Arch Gen Psychiatry,
53, 81424.
Schachter, S.C. and Saper, C.B. (1998). Vagus nerve stimulation. Epilepsia, 39, 67786.
Schatzberg, A.F. and Schildkraut, J.J. (1996). Recent studies on norepinephrine systems in mood
disorders. In Psychopharmacology, 4th Generation of Progress, ed. F. Bloom and D. Kupfer, pp.
pp. 91120. New York: Raven Press.
Schmitz, E.B., Robertson, M.M. and Trimble, M.R. (1999). Depression and schizophrenia in epi-
lepsy, social and biological risk factors. Epilepsy Res, 35, 5968.
Smith, D.F., Baker, G.A., Dewey, M., Jacoby, A. and Chadwick, D.W. (1991). Seizure frequency,
patient-perceived seizure severity and the psychosocial consequences of intractable epilepsy.
Suppes, T., Brown, E.S., McElroy, S.L. et al. (1999). Lamotrigine for the treatment of bipolar dis-
order, a clinical case series. J Aect Disord, 53, 958.
Swann, A.C., Bowden, C.L., Morris, D. et al. (1997). Depression during mania. Treatment
response to lithium or divalproex. Arch Gen Psychiatry, 54, 3742.
Trimble, M.R. (1998). New antiepileptic drugs and psychopathology. Neuropsychobiology, 38,
14951.
Vonck, K., Boon, P., Van Laere, K. et al. (2000). Acute single photon emission computed tomo-
graphic study of vagus nerve stimulation in refractory epilepsy. Epilepsia, 41, 6019.
Walker, B.R., Easton, A. and Gale, K. (1999). Regulation of limbic motor seizures by GABA and
glutamate transmission in nucleus tractus solitarius. Epilepsia, 40, 10517.
Watkins, L.L., Grossman, P., Krishnan, R. and Sherwood, A. (1998). Anxiety and vagal control of
heart rate. Psychosom Med, 60, 498502.
Yeragani, V.K., Pohl, R., Balon, R. et al. (1991). Heart rate variability in patients with major
depression. Psychiatry Res, 37, 3546.
Zabara, J. (1985). Peripheral control of hypersynchronous discharge in epilepsy.
Electroencephalogr Clin Neurophysiol, 61s, S162.
Zabara, J. (1992). Inhibition of experimental seizures in canines by repetitive vagal stimulation.
Zerssen, D.v., Koeller, D.M. and Rey, E.R. (1970). Die bendlichkeits-skala Bf-S ein einfaches
instrument zur objektivierung von bendlichkeits-strungen insbesondere im rahmen von
lngsschnitt-untersuchungen. Arzneimittelforschung, 20, 91518.
Zung, W.W.K. (1971). A rating instrument for anxiety disorders. Psychosomatics, 12, 3719.
Part VI
Treatment
20
On the use of psychotropic drugs in patients

with seizure disorder
M.R. Trimble1 and Anke Hensiek2
1
2
Addenbrookes Hospital, Cambridge, UK
Introduction
It is now accepted that many patients with epilepsy have psychiatric problems.
Recent epidemiological evidence from selected clinics suggests that over 50% of
patients may have a recognizable psychiatric disorder (Krishnamoorthy and
Trimble, unpublished data). It is also known that many patients with epilepsy
receive psychotropic drugs, sometimes, but not always, on account of their psychi-
atric symptoms. Thus, it has to be acknowledged that there is an overlap between
anticonvulsant drugs and psychotropic agents, such that many of the former are
known to have mood-regulating properties, while a number of the latter (for
example, benzodiazepines) have anticonvulsant properties.
A classication of psychotropic drugs currently in use is given in Table 20.1.
The main part of this text relates to the prescription of antidepressant and anti-
psychotic drugs in patients with epilepsy, and then a brief comment will be made
about some of the other agents.
Antidepressant drugs seizures and epilepsy
Ever since the introduction of tricyclic drugs into clinical practice, seizures have
been recognized as a side eect. This has been reviewed on several occasions
(Trimble, 1980, 1987).
The position has changed in the last few years because of the introduction of a
number of new antidepressant drugs, and following a brief review of the older lit-
erature, these will be discussed.
A review of some early studies

Animal data
A number of laboratory and clinical investigations were carried out to assess the
eect of tricyclic drugs on seizures and the seizure threshold. The early laboratory
299
300 M.R. Trimble and A. Hensiek
Table 20.1. Classification of psychotic drugs
Antidepressants
Antipsychotics
Minor tranquillizers
Mood stabilizers
Psychostimulants
Others (beta blockers etc.)
investigations essentially revealed the proconvulsant eect of these agents, and it

emerged from these studies that clomipramine, amitriptyline and maprotiline were
probably the most proconvulsant compounds (Trimble, 1987).
Luchins et al. (1984) used spike activity in perfused guineapig hippocampal slices
as an indication of epileptogenicity and reported the eect of a variety of antide-
pressants. Imipramine, amitriptyline, nortriptyline, desipramine and maprotiline
generally increased spike activity, while viloxazine, protriptyline and trimipramine
appeared to decrease neuronal excitability. Nomifensine, a drug no longer avail-
able, had a biphasic eect, increasing excitability initially, and then producing ces-
sation of spikes. Similarly, doxepin produced a signicant increase in excitability,
and then signicant decreases. In this model mianserin had no eect.
Using the photosensitive baboon, Papio papio, Trimble et al. (1977) compared
two tricyclic drugs, namely clomipramine and imipramine, and the quadricyclic
maprotiline with the nontricyclic nomifensine. The rst three all lowered the
seizure threshold, while nomifensine had little eect and in some animals was anti-
convulsant.
One interesting drug examined in these early studies was viloxazine. In two
animal models (Luchins et al., 1984; Meldrum et al., 1982) it was observed, if any-
thing, to have a seizure-protective eect.
These early animal studies therefore suggested the potential for both anticonvul-
sant and proconvulsant eects of these drugs; in some compounds the potential
was a dose-related eect.
Clinical data
The clinical data at that time came mainly from reports of government agencies
such as the Committee for the Safety of Medicines (CSM), and from clinical trial
data. Several reviews emphasized the poor quality of the available information
(Edwards, 1985). However, from the clinical studies the highest reporting of sei-
zures was with maprotiline and clomipramine, and the lowest reporting with pro-
triptyline.
301 The use of psychotropic drugs in seizure disorder
The estimated risk of seizures with tricyclic drugs was around 0.06 to 0.1%
(Burley, 1977; Jick et al., 1983) The incidence of seizures with imipramine was
0.7%, and with clomipramine 3.0%.
Garvey and Tollefson (1987) drew attention to myoclonic seizures that could
occur in relationship to tricyclic drug prescribing, and suggested that as many as
40% of patients reported some kind of myoclonic event after starting them. They
calculated the frequency of the eect with the dierent drugs in descending order
from maprotiline, trazodone, nortriptyline, desipramine, amitriptyline to imipra-
mine. Doxepin was not associated with this eect.
Other information that emerged from the early clinical studies was the low
reporting of seizures with viloxazine (Edwards and Glen-Bott, 1984), and for pro-
convulsant drugs a relationship of seizure reporting to the therapeutic dose, with
higher doses of the drug having a higher frequency of seizures, a clear relationship
to overdose, and the relationship of seizures to the number of psychotropic drugs
prescribed.
The time relationship between commencing the drug and the seizures varies
considerably between studies. The attack may occur from 24 hours to several weeks
after starting an antidepressant, although early seizures (less than a week later)
would seem to be associated with lower dosing schedules and possibly more
patient-related factors that lower the seizure threshold (Trimble, 1980) than later-
onset seizures.
It was generally concluded that patients were more likely to have seizures (if they
did not have epilepsy) if they had a family history of seizures, or a past history of
relevant medical conditions such as a head injury or a cerebrovascular accident.
Few of the above antidepressant drugs were tried in patients with epilepsy.
Viloxazine was studied, but it was shown to easily lead to anticonvulsant drug tox-
icity, and was therefore not recommended in epilepsy (Pisani et al., 1984).
Paradoxically, there were some clinical reports of tricyclic antidepressants being
anticonvulsant. Ojemann et al. (1983) reported retrospective data, which suggested
that doxepin improved seizure frequency in 15 of 19 patients who were prescribed
the drug. This included a diminution of both partial and generalized tonic-clonic
seizures.
Conclusions from these earlier clinical studies were that, in general, antidepress-
ant drugs were proconvulsant, although they were not all proconvulsant to the
same degree, and some may not alter the seizure threshold or have a biphasic eect,
sometimes revealing some anticonvulsant eects. Table 20.2 shows factors which
were thought to be interlinked with lowering of the seizure threshold, emphasizing
that this is a problem of the use of these drugs which is not conned only to epi-
lepsy. There are many patients without epilepsy, who have a lowered seizure thresh-
old, who are susceptible to psychotropic-induced seizures.
Table 20.2. Factors that lower the seizure threshold
A family history of epilepsy

Head injury, especially with a prolonged posttraumatic amnesia, or intracranial injury
Neurological illness
Table 20.3. Newer antidepressants
Noradrenergic uptake inhibitors reboxetine

Serotoninnoradrenaline uptake inhibitors venlafaxine
Serotonin antagonist/reuptake inhibitor nefazodone
Noradrenaline selective serotonin uptake inhibitors mirtazapine
Newer antidepressant drugs

There have been several developments of antidepressants since the tricyclic era.
Some drugs have briey been mentioned above, which were nontricyclic, such as
mianserin, maprotiline and viloxazine. However, the major development in the last
few years has been of agents that selectively inhibit reuptake and either noradren-
aline, or serotonin, or both.
Table 20.3 shows a list of the newer drugs. Table 20.4 gives a receptor prole and
the epileptogenic potential of these compounds.
In brief, the selective serotonin reuptake inhibitors (SSRIs) are represented by
citalopram, uoxetine, uvoxamine, sertraline and paroxetine. Of these, citalo-
pram is the most selective on serotonergic reuptake, inhibiting serotonin reuptake
3000 times more than noradrenaline uptake, and 22000 times more than dopamine
(Noble and Beneld, 1997). In general, the SSRIs are better tolerated and safer in
overdose compared to tricyclic drugs.
The latest generation of antidepressants has been developed to derive their ther-
apeutic benets from tailor-made action at specic monoamine receptors and
reuptake sites, in theory providing better ecacy and better tolerability (Feighner,
1999).
Reboxetine is a selective noradrenergic reuptake inhibitor (NARI) with low an-
ity for histaminergic, cholinergic, dopaminergic and alpha-1 adrenergic receptors. It
appears to be equally eective as the tricyclics in treating depression, and there is a
suggestion that it may be more eective than uoxetine (Montgomery, 1997).
Venlafaxine is a serotoninnoradrenergic reuptake inhibitor (SNRI), which is
similar to the earlier generation of antidepressants, but it does not interact with
histaminergic or cholinergic receptors, thus diminishing side eects due to those
receptor systems. Several studies have indicated equi-potentiability or superior
Table 20.4. Receptor profile and epileptogenic potential of antidepressants
Action on receptor
Drug H1 M1 NA 5HT1 5HT2 5HT3 Seizures/epilepsy
Imipramine 0.14%
In overdose:
3.88%
Paroxetine Prolonged seizures during ECT
In overdose:
No seizures in 15 patients with maximum
dose 850 mg
Sertraline Rare reports of seizures secondary to SIADH
In overdose:
No seizures in 40 patients up to 8000 mg
Fluoxetine 1/1000a
Citalopram No worsening of epilepsy in 16 patients
In overdose:
100 mg 1.9 g: 18% seizures; 1.9 g: 49%
Reboxetine / 0.13%a
Venlafaxine 0.18%a
In overdose:
Seizures in dosages over 1000 mg
Nefazodone No seizures in premarketing trials, since
then rare reports of convulsionsa
Mirtazapine 0.1%a
Notes:
, no/negligible eect; , stimulation; , blockade; a information from premarketing trials and product
monograph. Receptors: H1, histamine; M1, muscarinergic; NA, noradrenaline; 5HT1, 5HT2, 5HT3, serotonin.
eectiveness with this compound compared with tricyclics (Burnett and Dinan,
1994).
Nefazodone is a noradrenalineserotonin reuptake inhibitor whose most potent
action is blockade of 5HT2 postsynaptic receptors, leading to a dual mechanism of
action on the serotonin system. Noradrenaline reuptake inhibition is only minimal,
and there is no interaction with histamine or cholinergic receptors.
Mirtazapine (a noradrenaline-specic serotoninergic antidepressant or NASSA)
has a selective action at alpha-2 adrenoreceptors, and only at some serotonin recep-
tor subtypes. Its actions are to increase noradrenergic and serotoninergic transmis-
sion by blocking the alpha-2 autoreceptors. However, because it also blocks 5HT2
and 5HT3 receptors, the increased serotonin turnover only stimulates the 5HT1
receptors. Thus it enhances noradrenergic and 5HT1A-mediated serotonergic
neurotransmission. It is free of muscarinic, alpha-1 adrenergic and 5HT2- and
5HT3-related side eects, but its eect on histamine receptors can cause sedation
and increased appetite. Several studies have shown equal or superior eciency of
this compound compared with other antidepressants (Bremner, 1995).
New antidepressant drugs and seizures

All of the SSRIs have been associated with seizures in clinical practice, although
there is some evidence which suggests that they may have less seizure potential than
the earlier agents.
Krijzer et al. (1984), used freely moving rats implanted with subcortical elec-
trodes. Almost all of the antidepressants tested caused epileptogenic EEG changes;
mianserin was the most potent. However, uvoxamine caused only minimal eects.
None of the newer agents have been tested in such models and clinical informa-
tion is largely derived from the clinical trials and postmarketing surveys. Generally
the gures for seizure incidence given for all of these new compounds are less than
for the tricyclics, the lowest gures being recorded so far for mirtazapine and nefad-
azone.
Citalopram has been used in depression in patients with epilepsy; no change of
seizure frequency was noted in 16 patients in an open study (Specchio et al., 1999).
In another study, Hovorka et al. (2000) gave citalopram to 43 patients with epilepsy
and comorbid depression, and assessed them over an 8-week period. Sixty-ve per
cent were judged to be responders to the antidepressant eect. No change of seizure
frequency was noted, and no de novo generalized tonic-clonic seizures were
observed.
The SSRI most used in patients with epilepsy is paroxetine. Blumer (1997;
Chapter 8) has reported on the eective use of paroxetine in the management of
patients with what he refers to as the interictal dysphoric disorder of epilepsy. In his
studies, paroxetine is often given in combination with a tricyclic antidepressant. He
reports this combination to be safe and ecacious in this population. Exacerbation
of seizures was not reported, and dysphoric symptoms resolved in the majority of
his cases.
We have observed three patients with persistent epilepsy who have become
seizure-free on paroxetine. In one, the eect was short-lived, but in the other two
the eect has been sustained. This in spite of no changes to the anticonvulsant pre-
scriptions.
The other SSRI recently evaluated in epilepsy is sertraline (Kanner et al., 2000).
They prospectively evaluated the eect of this antidepressant in 100 consecutive
patients with epilepsy and depression (n97) or obsessivecompulsive disorder
(n3). They noted an increase in seizures following start of therapy in 6% of
patients, patients being assessed from 0.2 to 38 months. Interestingly, the mean
dose of sertraline in these six was lower than in the other patients. They reported
that depressive symptoms resolved in 54% of patients, but also described (Blumer,
1997) the pleomorphic clinical picture of these patients, and the symptom dier-
ences from typical major aective disorder.
Pharmacokinetic interactions
It has been emphasized for a long time now that serum anticonvulsant-level mon-
itoring can be of value in obtaining good seizure control, and checking compliance.
The administration of additional drugs can cause metabolic interactions, leading
to either a fall or a rise in the anticonvulsant serum levels. This may lead to a recru-
descence or a worsening of seizure frequency, or precipitate anticonvulsant toxicity.
There are occasional but nevertheless important reports of interactions between
tricyclic drugs and both phenytoin and carbamazepine, leading to toxicity. The case
of viloxazine has also been noted above.
The new generation of antidepressant drugs diers considerably in their ability
to induce liver enzymes of the P450 system. Most psychotropic drugs are metabo-
lized by four isoenzymes (CYP1A2, CYP2C, CYP2D6 and CYP3A4) (Monaco and
Cicolin, 1999). The anticonvulsants mainly aect CYP3A4, there thus being some
potential for pharmacokinetic interactions.
In general, drugs which induce liver enzymes may lower the levels of antidepress-
ant drugs, and this may have therapeutic consequences. Information on the eect of
SSRIs on plasma levels of anticonvulsants are limited, although there are reports of
carbamazepine toxicity in patients given uoxetine (Dursan et al., 1993). Keller et
al. (1997) looked for interactions between uoxetine and carbamazepine in patients
with epilepsy, but did not nd any change of plasma levels over an observation
period of 20 days. Fluoxetine has been associated with case histories of increased
phenytoin (Jalil, 1992) and sodium valproate levels (Cruz-Flores et al., 1995).
Sertraline has less of an inuence on concomitant anticonvulsant levels, prob-
ably because it has little or no eect on the cytochrome P450 3A4 system. However,
possible interactions between sertraline and lamotrigine have been suggested
(Kaufman and Gerner, 1998).
Paroxetine also does not inhibit the CYP3A4 system, and may not therefore
provoke any interactions. The limited clinical data on 20 patients with epilepsy
given this drug did not reveal any signicant alteration of anticonvulsant levels
(unpublished data).
With regard to the newer non-SSRI agents, no information is available.
Conclusions
The data on the use of antidepressants in epilepsy suggest the following. Nearly all
of the tricyclic drugs are proconvulsant, but there are clinical reports in which, at
Table 20.5. Antipsychotic drugs
Phenothiazines
Butyrophenones
Atypical
Others (sulpiride; tetrabenazine)
least with some of them, an anticonvulsant eect has been noted. The reason for
this paradoxical eect is unclear. In psychiatric practice, in recent years, there has
been a move away from the use of tricyclic drugs, mainly on account of their other
side eects and risk of death with overdose; in patients with a reduced seizure
threshold, not necessarily having epilepsy, they should be avoided. Other drugs that
are proconvulsant include maprotiline and mianserin.
Of the newer generation of drugs, the SSRIs appear to provoke less in the way of
seizures than the tricyclic drugs. It is a possibility that the even newer, more selec-
tive drugs provoke less in the way of seizures than the SSRIs, but more data on these
compounds are needed.
Metabolic interactions occur with some of these compounds, which may lead to
anticonvulsant toxicity. Any change in patient-reported symptoms, or a deteriora-
tion of the aective disorder, that may suggest a toxic eect, need to be watched out
for. Whichever drug is used it is advisable to start, if clinical needs will allow, at
smaller doses, and increase the dose relatively slowly in order to avoid precipitation
of potential seizures.
Antipsychotic drugs
A classication of the antipsychotic drugs is given in Table 20.5; as with antide-

pressant drugs, in recent years there have been several newer agents introduced into
clinical practice. These essentially, with some exceptions, fall into the class of atyp-
ical antipsychotics. Table 20.6 gives a review of the receptor-binding proles of a
number of these agents.
The classical neuroleptic drugs, such as chlorpromazine and haloperidol, antag-
onize dopamine D2 receptors. Essentially their clinical eciency has been shown
to correlate with inhibitory activity at these receptor subtypes. However, these
drugs block dopamine receptors in the striatum leading to catalepsy in animal
models, and unwanted extrapyramidal side eects in clinical practice.
The new generation of antipsychotic drugs essentially fall into two categories;
those that are clozapine related, which includes olanzapine and quetiapine, and
others such as risperidone.
Table 20.6. Receptor binding profiles of antipsychotics
Anity for receptor (Ki in nmol/l)
Drug D1 D2 D4 1 2 H1 5HT2a 5hHT2c M
Haloperidol 25 1 5 46 360 1000 78 1000 570

Clozapine 85 126 9 7 8 6 12 8 1.0
Risperidone 75 3 7 2 3 155 0.6 26 1000
Olanzapine 31 112 27 19 228 7 4 11 2.1
Quetiapine 455 160 NA 7 87 11 220 615 56
Notes:
Receptors: D1, D2, D4, dopamine; 1, 2, adrenergic; M, muscarinergic; 5HT2a, 5HT2c,
serotonine; H, histamine; NA, not available.
Source: Information from premarketing trials and product monographs.
Although clozapine has been available for many years, it was initially removed
from clinical practice (except in some selected countries) on account of its poten-
tial to produce agranulocytosis. However, it was reintroduced as a model of an
atypical antipsychotic. The term relates to the low potential of these compounds to
cause extrapyramidal problems, and they also have minimal eects on serum pro-
lactin levels. The mechanism of atypicality seems to relate to dierent receptor pro-
les.
In general, the atypical antipsychotics occupy lower levels of D2 receptors than
the classical antipsychotics (2060% as opposed to 8090%) (Kapur et al., 1999).
One reason for their prole may be due to the rapid displacement of these agents
from receptors by endogenous dopamine, on account of their being more loosely
bound.
The newer antipsychotic agents also have a lower relative anity for striatal D2
receptors as opposed to limbic D2 receptors (dorsal vs. ventral striatum). Further,
of all the newer agents, clozapine is the one that seems not to bind to the core of
the nucleus accumbens.
Since their introduction, antipsychotic drugs have been shown to be proconvul-
sant. Early animal models, using the photosensitive baboon Papio papio, suggested
that there may be dierences between the phenothiazine-derived agents, such as
chlorpromazine, and the butyrophenones, represented for example by haloperidol
and pimozide. Pimozide in particular seemed to have less of an eect on the seizure
threshold. In clinical practice it has recently been problematic to prescribe because
of the need to carry out ECG investigations before prescription. It is one of a
growing number of drugs associated with prolonging the Q-T interval, with the
possibility of being associated with cardiac complications.
In general, the use of intramuscular preparations, such as uphenazine deca-
noate, was not associated with any change in the frequency of reporting of seizures
in patients with epilepsy who also had psychosis.
As with the newer antidepressants, there is much less information about the
eect of the atypical neuroleptics on the seizure threshold, with the single excep-
tion of clozapine. The latter was known to be proconvulsant from early studies, the
seizures being a dose-related eect. The incidence of seizures rises to about 5% at
doses of 600 mg, although EEG changes may be recorded at lower doses. The sei-
zures are often myoclonic, but can be generalized tonic-clonic, or partial, depend-
ing on the individual patient.
It is perhaps no coincidence that the drug which appears to be the most eective
antipsychotic, namely clozapine, is also associated with a high frequency of sei-
zures. The relationship of convulsive seizures to the relief of psychopathology is an
integral part of psychiatric therapy, through ECT. It is often forgotten that the latter
was introduced for the treatment of dementia praecox, and has clinically and theo-
retically important antipsychotic eects.
There are some patients with epilepsy who are nonresponsive to neuroleptic
drugs, and need clozapine. In particular there is a group of patients whose seizure
frequency decreases or who become seizure-free, whose psychosis deteriorates in
this setting. For them clozapine may be the drug of choice.
CASE REPORT
A 30-year-old patient, diagnosed as having leucine-sensitive hypoglycaemia at 10 months

of age developed epilepsy at the age of four. This typically presented with clusters of several
episodes daily, lasting 3 or 4 days, recurring at monthly intervals. During her seizure she
would have a typical aura with a feeling of fear and butterflies in her stomach lasting about
30 seconds. This was followed by a scream, and a generalized seizure, which would last
about a minute. Prior to these seizures she would often have a prodrome of 23 days with
a build-up of verbal and physical aggression.
She had been treated with various medications, but at the age of 26 she was changed to
sodium valproate, and her seizure frequency improved, indeed she became seizure-free.
She had gradually developed a psychotic illness, but this dramatically deteriorated with res-
olution of the seizures and she was admitted twice to psychiatric hospitals under the Mental
Health Act, aged 26 and 27.
An EEG had revealed left anterior temporal abnormalities, and an MRI showed prominent
ventricles.
She had been prescribed several antipsychotic drugs, including the atypical antipsychot-
ics risperidone and olanzapine. None of these were of any help in resolving her psychosis.
She was therefore started on clozapine, in gradually increasing doses. Her EEG was initially
monitored.
On clozapine, her EEG revealed more frequent sharp waves over the left temporal region,
and she began to develop auras again, although had no complex partial or generalized sei-
zures. The auras were simple partial attacks, and were of little concern to her.
A dramatic improvement in her psychosis was noted, such that she is once again living
independently in the community, with stable mood, infrequent auditory hallucinations,
and with more insight into her paranoia. She remains on sodium valproate and clozapine
(400 mg a day).
Figures for the incidence of seizures with the other atypical antipsychotics vary
from a reporting of 0.1% of seizures in double-blind clinical trials of rispiridone,
to 0.20.9% for olanzapine, and 9 out of 1710 cases for quetiapine. These latter
gures come from the reporting of seizures in clinical trials, and do not necessar-
ily reect a direct causeeect relationship between prescription of the drug and
the seizure event.
Pharmacokinetic interactions
The interactions between antipsychotic drugs and antiepileptic drugs have been
even less studied than the antidepressants. Some psychotropics, such as haloperi-
dol, mainly metabolize using the P450 system, others such as chlorpromazine use
dierent liver mechanisms. However, decreases in the levels of some neuroleptics
can occur in patients prescribed anticonvulsant drugs, and several studies have
been carried out in patients with schizophrenia who have received both carbamaz-
epine and a neuroleptic. Haloperidol levels can drop by up to 50% following coad-
ministration of the antiepileptic (Arana et al., 1986). Clozapine and olanzapine
primarily use the CYP1A2 isoenzyme, which may lead to interactions with some of
the tricyclic antidepressants, and carbamazepine.
Conclusions
As with antidepressants, further work needs to be done in the important area of man-
aging patients with neurological disease, particularly epilepsy, with antipsychotic
agents. At present, particularly in epilepsy, the tendency is away from using the more
traditional neuroleptics, to using the atypical neuroleptics, for several reasons. The
main one relates to the potential danger of the long-term development of extrapyra-
midal motor disorders, which are much less likely to occur with the atypical neuro-
leptics. The latter are mainly well tolerated by patients with epilepsy, and seizures are
not usually a problem clinically. As emphasized, clozapine can be used in patients
with epilepsy, particularly if the psychosis is proving intractable to treatment. The
drug is introduced slowly, and the EEG monitored. Patients are warned that their
seizure frequency may rise; however, at doses below about 600 mg/day, clinical
problems have not been encountered. One particular caution relates to the develop-
ment of agranulocytosis, and patients (in the UK at least) need to be placed on a
special register, and also have regular haematological assessments. Further, it is a
contraindication to prescribe clozapine at the same time as carbamazepine.
Other psychotropic agents
Patients with epilepsy are prescribed a variety of other psychotropics, the main
ones being benzodiazepines, either as hypnotics or anticonvulsants, and lithium, a
mood stabilizer.
Benzodiazepines should be used in caution in patients with epilepsy, the main
problem being the potential for a paradoxical increase in seizures, or withdrawal
seizures on stopping the prescription. Further, some of these drugs have a poten-
tial for the development of dependency.
There appear to be dierences between the 1,5 and 1,4 benzodiazepines, the
former being represented by clobazam. This drug was introduced initially as an
anxiolytic, but was shown to have eective and sustained anticonvulsant proper-
ties. It is recommended as an adjunct treatment for the management of patients
with intractible epilepsy, and may be particularly of value in patients with epilepsy
with a high level of anxiety, who may also present with panic attacks. It is less cereb-
rotoxic than the 1,4 equivalents such as clonazepam, and is recognized to have
inherent psychotropic properties. Clobazam is of particular value in patients with
intermittent clusters of seizures (such as catamenial episodes), and for the supres-
sion of clusters of seizures. The latter are associated in some patients with postictal
psychosis, and prevention of the cluster may well abort a potential psychosis. Ten
milligrams given 46 hourly for 2448 hours may be all that is required. Clobazam
can also be given after the cluster, if any psychiatric symptoms seem to be develop-
ing, using a similar schedule.
Lithium, which is also proconvulsant, can be used as a mood stabilizer in patients
who have recurrent cyclical mood disorders, or recurrent outbursts of aective
aggressive behaviours. Caution should be exercised when combining lithium with
carbamazepine, as patients occasionally develop a cerebrotoxic syndrome.
Monitoring of serum levels of lithium is mandatory, as is observing patients over
time for the development of secondary complications of lithium therapy such as
hypothyroidism, or diabetes insipidus.
Conclusions
Psychotic drugs are used with considerable frequency in patients with epilepsy, and
used appropriately and cautiously they add considerably to management. However,
like all CNS drugs, they have a variety of side eects, and the exacerbation or pre-
cipitation of seizures is important in this patient group. It is particularly relevant
in patients who may have been seizure-free for a period of time, and who then go
on to develop psychiatric disorders.
Recent years have seen an expansion in the number of psychotropic drugs avail-
able, particularly with regard to the antidepressants and the antipsychotics. The
newer developed agents generally seem to have a more favourable prole than older
agents for use in patients with epilepsy.
R E F E R E N C ES
Arana, G.W., Go, D.C., Freidman, H. et al. (1986). Does carbamazepine-induced reduction in
haloperidol plasma levels worsen psychotic symptoms? Am J Psychiatry, 143, 6589.
Bremner, J.D. (1995). A double blind comparison of ORG 3770, amitriptyline and placebo in
major depression. J Clin Psychiatry, 56, 51926.
Burley, D.M. (1977). A brief note on the problem of epilepsy and antidepressant treatment. In
Depression: The Biochemical and Physiological Role of Ludiomil, ed. A. Jewkes, pp. 2023.
Horsham: Ceiba.
Burnett, F.E. and Dinan, T.G. (1994). The clinical eectiveness of venlafaxine in the treatment of
depression. Rev Contemp Pharmacother, 9, 30320.
Cruz-Flores, S., Ghazala, R., Hyat, R. and Mirza, W. (1995). Valproaic toxicity with uoxetine
therapy. Missouri Med, 92, 2967.
Dursan, S.M., Natthew, V.W. and Reveley, M.A. (1993). Toxic serotonin syndrome after uoxe-
tine plus carbamazepine. Lancet, 342, 4423.
Edwards, J.D. and Glen-Bott, M. (1984). Does viloxazine have epileptogenic properties. J Neurol
Neurosurg Psychiatry, 47, 9604.
Edwards, J.G. (1985). Antidepressants and seizures: epidemiological and clinical aspects. In The
Psychopharmacology of Epilepsy, ed. M.R. Trimble, pp. 119139. Chichester: John Wiley &
Sons.
Feighner, G.P. (1999). Mechanism of action of antidepressant medication. J Clin Psychiatry, 60,
411.
Garvey, M.J. and Tollefson, G.D. (1987). Occurrence of myoclonus in patients treated with cyclic
antidepressants. Arch Gen Psychiatry, 44, 26972.
Hovorka, J., Herman, E. and Nemcova, I. (2000). Treatment of interictal depression with citalo-
pram in patients with epilepsy. Epilepsy Behav, 6, 4448.
Jalil, P. (1992). Toxic reaction following the combined administration of uoxetine and pheny-
toin. J Neurol Neurosurg Psychiatry, 55, 41213.
Jick, H., Dinan, B.J., Hunter, J.R. et al. (1983). Tricyclic antidepressants and convulsions. J Clin
Psychopharmacol, 3, 1825.
Kanner, A.M., Kozak, A.M. and Frey, M. (2000). The use of sertraline in patients with epilepsy,
is it safe? Epilepsy Behav, 1, 1005.
Kapur, S., Zipursky, R.B. and Remmington, G. (1999). Clinical and theoretical implications of
5HT2 and D2 receptor occupancy of clozapine, risperidone and olanzapine. Am J Psychiatry,
156, 28693.
Kaufman, K.R. and Gerner, R. (1998). Lamotrigine toxicity secondary to sertraline. Seizure, 7,
1635.
Keller, R.. Tortar Prolo, P., Ravizza, L. and Monaco, F. (1997). Interazioni farmacocinetiche tra
uoxetina e farmaci anti epilettica. Bolletino Lega Italiana Contro Lepilepsia, 99, 1836.
Krijzer, F., Snelder, M. and Bradford, D. (1984). Comparison of the proconvulsant properties of
uvoxamine and clovoxamine with other antidepressants in an animal model.
Neuropsychobiology, 12, 24954.
Luchins, D.J., Oliver, A.P. and Wyatt, R.J. (1984). Seizures with antidepressant: an in vitro tech-
nique to assess relative risk. Epilepsia, 25, 2532.
Meldrum, B.S., Anlezark, G., Adam, H.K. and Greenwood, D.T. (1982). Anticonvulsant and pro-
convulsant properties and viloxazine hydrochloride. Psychopharmacology, 76, 21217.
Monaco, F. and Cicolin, A. (1999). Interactions between anticonvulsants and psychoactive drugs.
Epilepsia, 40 (Suppl.), S716.
Montgomery, S.A. (1997). Riboxetine: additional benets to depressed patients. J
Psychopharmacol, 11 (Suppl.), S915.
Noble, S. and Beneld, P. (1997). Citalopram: a review of its pharmacology, clinical eciency and
tolerability in the treatment of depression. CNS Drugs, 8, 41031.
Ojemann, L.M., Friel, P.N., Trejow, J. and Dudley, D.L. (1983). Eect of doxepin on seizure fre-
quency and depressed epileptic patients. Neurology, 33, 6468.
Pisani, F., Narvone, M.C., Fazio, A. et al. (1984). Increased serum carbamazepine levels by vilox-
azine in epileptic patients. Epilepsia, 25, 4825.
Specchio, L.M., La Neve, A., Spinelli, A. et al. (1999). Il trattamento antidepressevo con citalo-
pram in pazienti con epilepssia. Bollettino Lega Italiana Crontro Lepilepsia, 99, 1878.
Trimble, M.R. (1980). New antidepressant drugs and the seizure threshold. Neuropharmacology,
19, 12278.
Trimble, M.R. (1987). Antidepressant drugs seizures and epilepsy. In Quo Vadis. Montpelier:
Sano Group.
Trimble, M.R., Anlezark, G. and Meldrum, B. (1977). Seizure activity in photosensitive baboons
following antidepressant drugs, and the role of serotoninergic mechanisms. Psychopharma-
cology, 51, 15964.
21
The role of psychotherapy in the treatment

of epilepsies
Martin Schndienst
Epilepsie-Zentrum Bethel, Bielefeld, Germany
Introduction
The use of psychotherapy in epilepsy is a complicated topic, because it always has

to take into account the neurological dimension of the underlying epileptic disor-
der as well as the specic psychiatric vulnerability epilepsy engenders. As a result,
the following cannot provide a simple set of treatment rules. Instead, it presents:
1. Introductory comments on the complex relation between (neurological) epilep-
tological and psychotherapeutic approaches.
2. This is followed by four case reports to give some idea of how the general rules
of diagnosis have to be applied in an individual way to produce a unique treat-
ment in every single case.
3. Then, some ideas are presented on indications and contraindications.
4. Finally, ndings from a posttreatment comparison of conventional inpatient
treatment of epilepsy vs. treatment supplemented by psychotherapy are reported
to help decide whether psychotherapy is in any way eective in treating epilepsy
or is only well-intended.
Despite the above-mentioned complexity, it is easy to gain an overview on
psychotherapy in the treatment of epilepsy because there seems to be practically no
recent literature on the topic. The indices of major modern textbooks on epileptol-
ogy refer to psychotherapies exclusively in connection with pseudoseizures (Engel
and Pedley, 1997) or give, at best, the terse reference that this modality is largely
neglected in the literature on epilepsy for several reasons (Stagno, 1993, p. 1154).
Except for pseudoseizures, psychotherapeutic procedures have not been addressed
in any relevant epileptological journals during the last 10 years.
It is quite remarkable, and almost worth studying in its own right, that psycho-
therapy in the treatment of epilepsy has hardly ever been a subject of scientic
research, even though there is no denying the frequency of psychological distur-
bances in such patients. In line with the situation in the literature, medical care
reveals a disproportionately high number of secondary psychological disturbances
313
314 M. Schndienst
in epilepsy patients but a disproportionately low level of psychotherapeutic assis-

tance. This has less to do with the psychotherapists themselves, despite the frequent
truism that psychotherapy could aggravate an epilepsy and is, therefore, contrain-
dicated. It is far more the case that epileptology itself has remained conspicuously
silent as far as demands for the development of psychotherapeutic concepts for its
clients are concerned.
To some extent, this may be due to the way in which epileptologists often view
psychotherapists: namely, as little more than those who seem to have all the time at
their disposal that one would wish for ones own work. They also make compara-
tively little use of the available psychotherapeutic treatments compared to what are
considered to be potent drug options. Finally, clinicians often view psychotherapists
as persons who always only conrm what they know and anticipate already: partic-
ularly overprotective mothers, destructive fathers, strangulated aect in the
patients or some kind of spectacular narratives from a patients prior life that only
get in the way of the abstraction necessary in clinical work.
Finally, there are a few major dierences in the ways of thinking themselves that
create major handicaps to the integration of epileptological and psychotherapeu-
tic perspectives: whereas the epileptologist strives continuously to test whether
external treatments such as new drugs or, if necessary, surgery should be applied,
the psychotherapist focuses specically on the action potentials of the patients
themselves, be they in developing an individual technique for stopping seizures,
improving insight or correcting problematic patterns of interpersonal interactions.
Instead of relying on the positive, visible ndings of neuro-imaging techniques,
neurophysiology and neuropsychology, psychotherapy focuses on the gaps and
contradictions that rst make it possible to grasp defensive mechanisms such as
projections, denials and isolations with all their intra- and interindividual conse-
quences.
In contrast to the usual and necessary practice in epileptology of advancing
knowledge by tracing back the disorder increasingly more precisely to a circum-
scribed cause that is localized as accurately as possible, psychotherapy is concerned
with how symptoms are embedded in a persons individual life context. Hence, a
psychotherapeutic diagnosis is based on the idea that by creating a suitable setting,
cautious interventions and on-line observations of the interplay in the counselling
session, the decisive problem areas for the patients will be revealed in interaction
and also become accessible to modication.
Furthermore, there is also a justiable apprehension that focusing on, for
example, defensive processes or (counter) transference might lead to the neglect of
physical aspects that are just as essential for treatment. Each recognized dimension
added to the relevant data eld, in this case, the usual discourse on epileptology (i.e.
case history, EEG and, not least, neuro-imaging techniques) multiplies the number
315 Psychotherapy in the treatment of epilepsies
of interactions that need to be taken into account, and produces not only a factual
increase in knowledge but also an equally factual increase in potentially confusing
complexity. A reluctance to tackle such a lack of transparency and a preference to
narrow the eld to neurological treatments has a certain cognitive rationality and
is indubitably better than forcing oneself to take more demanding perspectives,
particularly when they are not encouraged by the hospital or surgery framework.
Nonetheless, epileptologically dicult treatment situations often emerge in
which it is popular to talk about a noncompliance or pharmacoresistance that
apparently cannot be explained further. These are situations in which the integra-
tion of a psychotherapeutic approach can bring about a decisive change in the
course of treatment. However, it is harder to recognize such a need for psychother-
apy in patients with epilepsy than in those without such an organic disease: many
treatment problems are attributed too hastily to the epilepsy or to the seizures as
such, to the medication or also to accompanying neuropsychological decits rather
than being conceived as problems accessible to psychotherapy. This neglects the
problems due to either more or less unconscious conicts (e.g. of identity, self-
esteem or dependency) or so-called ego-structural decits (e.g. a highly reduced
perception of self and/or other; inadequate self-control; unstable aect; or imma-
ture defence mechanisms such as, in particular, denial or dissociation, neurotic or
even psychotic projection and unstable attachment behaviour).
In the following, four short case reports will be used to illustrate how epilepto-
logical, psychiatric, and, in the stricter sense, psychotherapeutic dimensions inter-
relate.
F O U R C A S E R E P O R TS
The first patient has an epilepsy with focal and generalized seizures plus perimenstrually
peaking diffuse tonic-clonic seizures with onset at 15 years. Although her left-cephalic aura
indicated a right-temporal focus, photosensitivity and corresponding spike-waves as well as
a hereditary factor pointed to a generalized disorder.
The patient was resistant to phenytoin and carbamazepine and was referred to us with a
relatively high phenobarbital level.
The admission interview was characterized by the recurring and piercingly expressed
theme that her last physician had said he was referring her to us after telling her that he had
nothing left up his sleeve.
Yes, and then, for a while, I had a doctor who was on television, and everybody gave him
a lot of praise, and he almost cost me my life and treated me with valproic acid until I was
in a coma.
Another recurring phrase was, I cant fall into the open arms of a doctor.
Hence, these and similar communications linked together major therapeutic, erotic and
destructive ideas of reference in an indiscriminable, confusing and entangled manner.
316 M. Schndienst
We admitted her for inpatient treatment with the diagnosis of an agitated depression plus
occasionally severe suicidal intent.
Initially after the experiences reported above she must almost have thought we
wanted to murder her when we proposed changing her medication to the combination of
valproic acid and lamotrigine that is particularly promising for such mixed epilepsies.
Against the background of such breakdown fantasies, the planned changeover was even
more difficult because it also included an inpatient phenobarbital detoxification with all the
risks of seizures in withdrawal. None of those involved were spared in any way.
The patient was good at eliciting a number of strong reactions whose impact was bound
to remain destructive as long as it was not understood as an actualization and externaliza-
tion of her self- and self-esteem conflict and associated fears of doom and destruction.
Although the wish for a less exhausting patient is understandable in such phases of treat-
ment, it would only make the therapeutic relationship superficial or lead to a cessation of
treatment.
A few themes recurred during treatment:
1. Poisoning by the prior therapist along with the peevish reproach that, nowadays, it
seemed that such poisonings were simply taken for granted.
2. The fathers working in the garden.
3. Buying expensive shoes.
4. Her mothers own occasional seizure 20 years before.
Although this list may seem absurd, its contents provided opportunities for symbolic
understanding.
In her accounts of weekend visits to her parents that always featured her father working
in the garden, the patient found a way to leave behind all her bitterness and was at times
so warm-heartedly humorous as to not only disclose the both decisive and psychosexually
fixated relationship with her father but also enable her to become aware of this through a
reflective self-distancing.
The patients rather strange deliberations over whether to buy a wonderful pair of very
expensive winter boots not only reflected her self-conflict ranging from her emerging iden-
tity plans and oppressive material restrictions, but it was also capable of being named as
such. This was joined by dreams that made it possible for her to see the possibility of taking
an intermediate position somewhere between delusions of grandeur and depression.
At the same time, the mother started to recollect her single occasional seizure 30 years
before that, as now threatening the daughter, had led her to give up a career as a school
teacher while also suggesting a biological flaw in the family on which the daughter had
fixated in anxious anticipation.
One may consider that all this has little to do with the epilepsy but perhaps with a com-
pletely independent narcissistic neurosis. However, it is precisely the specific constellation of
connections in each single case between (a) directly seizure-related breakdown experiences,
(b) the narcissistic crises associated with every seizure for many patients, (c) the cumulative
growth of resignation, (d) the fragility of self-esteem and (e) the ambivalence towards med-
icines that, like a disappointing object, are perceived simultaneously as an indispensable pro-
tection but also repeatedly as a failure and disappointment that makes it necessary for the
therapist to consider not only the epilepsy and its best possible pharmacotherapy, but also
the ego-structural sequelae of both the seizures themselves and the medicines.
Naturally, it is impossible to reach a final decision on whether the long-term reduction in
anxiety and calming of this patient is due to her now being free from seizures for a com-
plete year, to the withdrawal from phenobarbital, to the combined treatment with valproic
acid and lamotrigine or to the processing of the above-mentioned (in this case, self- and
oedipal) conflicts with the resulting increase in internal latitude. Whatever the case, the
psychotherapeutic setting was certainly necessary to generate a tolerance for the difficult
changeover without which it would have been impossible to contain a patient in such a pre-
carius condition. After freedom from seizures had been achieved, I found it more than ironic
when the patient told me how her mother had asked her impatiently during her menstru-
ation whether she had had another seizure. After the second seizure-free menstruation, the
mother surprised the patient one morning by telling her that she had had her first seizure
for 30 years the night before, or at least she had woken up after biting her tongue.
The next patient, a 28-year-old male with a right-hemisphere epilepsy and cerebral hemi-
hypotrophy and somatosensory auras, frontal hypermotor and generalized tonic-clonic sei-
zures had not, at the time of admission, left his parents home for more than 10 years
because of his fear of having a seizure on the street.
The drug changeover was accompanied by two half-hour psychotherapy sessions per
week. A number of these sessions were characterized by the patients complaints over the
way seizures prevented him from getting anywhere in life along with my personal tiredness
in reaction to this that was almost impossible to control. After I had turned up late for several
of our sessions, I realized how far the patient seemed to accept my tardiness with complete
indifference. I mentioned this to him, and this transformed my role from a sacrosanct phy-
sician into that of an assailable other, leading to an incredible change: unexpectedly, I
became the focus of very excessive demands. He simply thought that I or the hospital should
help him to find not only a flat and a job but also, when possible, a mate. Hence, a regres-
sion to the level of grandiose infantile wishes had occurred. Several sessions pursued a kind
of reality test of the patients wishes that led inevitably to disappointed anger and the nec-
essary emotional counter-control of the therapist known in the literature as containing. After
a hefty but brief depressive reaction, the patient found his way out of his regressive arrest
and began to structure his future in small practical steps rather than getting caught up in
grandiose desires. He worked out his own desensitization programme to overcome his fear
of the streets that was soon a success. In addition, he managed to learn the necessary DC
potential shifts in a biofeedback treatment so that although auras continue to occur daily,
these have not turned into major seizures for more than 18 months.
The next 30-year-old patient with a focal epilepsy and right-temporal lobe hypotrophy was
admitted with such prior diagnoses as an abnormal personality development with depres-
sions and anxiety states and even a tentative diagnosis of onset of psychosis with auditory
hallucinations. A diagnostic phase within a preoperative institute had taken a relatively tur-
bulent course in interactive terms.
318 M. Schndienst
Alongside improving antiepileptic drug therapy, the 3.5-month treatment consisted ini-
tially in a very cautious approach to her strange and prolonged disturbances to perception
and experience. For a long time, these led to a threatening atmosphere that was hard to
understand. In the past, her so-called auditory (pseudo) hallucinations had led to the ten-
tative diagnosis of a psychosis, and these were currently also the reason for major disinte-
gration anxieties expressed in, for example, the sensation then, I feel like a dictionary in
which more and more pages are simply blank or everything I want to think gets sucked out
of my brain as if it was a dry sponge.
During psychotherapy, it took some time to verbalize the various episodic sensations to
which she was exposed because of the difficulty of putting the quality of these experiences
into words. It was also necessary to distinguish very slowly and carefully her auras from other
episodes that had the character of anticipatory anxiety or mental disintegration anxiety. This
also made it possible to calm down the patients own apprehension that she was facing a
creeping psychotic disintegration, an apprehension that she would suddenly start to project
on her communication partner with corresponding surges in anger when treatment first
commenced.
Hence, in this case, it was the verbalization of what was experienced in the auras that
helped to make the repeated sudden affects and intensively conflictual relationship patterns
accessible to a more relaxed self-observation. The patient developed more precise differen-
tiations of affect, and the initially threatening quality of the treatment atmosphere gradually
disappeared.
The final 19-year-old patient was referred to us after spending 11 months in a psychiatric
hospital for adolescents. For several years, she had been treated by an epileptic outpatient
department for the prior diagnosis of a focal frontal epilepsy. Our own diagnosis, in contrast,
was very clearly an idiopathic generalized epilepsy with myoclonic and generalized tonic-
clonic seizures.
Correcting the diagnosis was important, because the administration of lamotrigine and
valproic acid led immediately to a remarkable improvement in her seizures. Only isolated
relapses occurred either when the patient was woken abruptly or when medication had not
been taken. Nonetheless, such irregularities were very frequent because of her very irregu-
lar bedtimes as well as a failure to take medication. Encouraged by the impressive initial
effect of changing her drugs, we thought that the disruptive gaps in compliance could be
closed through simple advice. However, contact with the patient then became reduced: she
appeared to draw back from her previous attentiveness into a dour, seemingly almost
unreachable silence. After an exhausting processing that confronted major familial scotom-
ization tendencies in her life history, it emerged that at the age of 1014 years, that is, at the
onset of her epilepsy, her mother had blocked any adequate pharmacotherapy due to her
own fears of poisoning engendered by her personal psychotic development. Her father, in
contrast, had advised her emphatically to take her medicine, but had been unable to assert
himself. The reconstruction of this history in conjunction with the impression that the patient
drifted off precisely when attempting to discuss the background of her clearly self-injurious
noncompliance revealed the intrapsychological conflict behind the biographical drama: by
switching apparently meaninglessly between adequate and inadequate health behaviour,

the patient subconsciously maintained the inner relationship to both her father and her
mother. Compliance would have meant the loss of the internal mother. It is precisely when
such a meaning of noncompliance becomes recognized that compliance becomes conceiv-
able and, indeed, unproblematic.
Discussion
I would like to present some remarks on the setting for a psychotherapeutic

approach to epilepsies. My rst point addresses the apparently marginal element of
a xed timeframe, that is, the dened amount of time specied at the beginning of
treatment on the basis of the initial psychostructural situation, for example, four
times 20 minutes per week or two times 25 minutes per week. This changes time
from something that the physician uses to attain certain goals into something that
is primarily at the patients disposal, to be used or abused as he or she sees t.
The second point is embedded in the rst, and concerns the patients right and
need to start the discussion without being belittled by any kind of guiding ques-
tions on the choice of topic. The therapist is then no longer the knowledgeable one
who already knows what the patient should think about, what should be recom-
mended and what should be discouraged. The therapist becomes a companion in
the patients process of self-discovery.
A further point is the need to pay attention to methods for recognizing ego-
structural problems behind the more obvious surface symptoms.
Hence, behind the rst patients irritable depression lies the instability of her self-
esteem regulation that she externalized in fantasies of being poisoned with medi-
cines. In the second patient, the agoraphobic avoidance covers the tendency to
regress to the level of passive infantile need. In the third patient, schizoid rejection
of contacts hides her fear of psychotic disintegration as a consequence of her auras
that she had previously been unable to identify. In the fourth patient, depressive
withdrawal resulted from the unconscious conict arising from her identication
with both very hostile parents.
Interventions have to be selected to t the specic disorder, and they require a
willingness in the therapist to participate emotionally in what is perceived.
In the rst patient, for example, this meant involvement in the anxiety associated
with her seizures. In the second patient, the therapist had to counter his regression
in a demanding way and risk his anger. In the third patient, it was necessary to rec-
ognize the fearful atmosphere of vague threat. In the fourth patient, this meant
coming to terms with the seemingly (at rst glance) inappropriate switches of aect.
When collating the numerous case reports in which supplementing epilepsy
treatment with psychotherapy has proved to be of lasting benet, we have found
320 M. Schndienst
again and again that the decisive element in these treatments is to reconstruct the
remarkably blurred selfobject borders in these patients. One could say that it is
important for the therapist to take over various functions of a patients self (in the
sense of taking on a helping ego function) and reconstruct them successively.
This requires a correction of defence mechanisms, particularly of projection and
regression, that is never without some emotional strain.
The therapist also has to cope repeatedly with a certain form of balance. He or she
has to know the specic strains on self-regulation from having epilepsy that develop
not least from the loss of the availability of the self that is particular to the disease
and through which seizures can also take on the character of narcissistic crises.
In the rst patient, this meant commencing practicable confrontations and
demarcations instead of being trapped by both negative and positive idealizations
of her images of her parents. In the second patient, this meant letting the archaic
grandiose self be replaced by a more mature, disappointed self that was capable of
formulating and implementing phase-appropriate limited goals. In the third
patient, the concern was to overcome threatening projections and calm the prepsy-
chotic disintegration anxiety and develop the self-reective ability to discriminate
between auras and depersonalizations.
Questions of indication and contraindication
Alongside obvious indications such as a marked anxiety, obsessivecompulsive or

borderline disorder, we consider that the indication for a supplementary psycho-
therapy should always be examined in epilepsy patients when (a) an epilepsy treat-
ment is dicult in itself, and particularly when (b) noncompliance or (c)
pharmacoresistance are present.
One particular impediment to recognizing the indication for psychotherapy is
found in those underdiagnosed cases in which an epilepsy is accompanied by a
posttraumatic stress disorder (PTSD) (Rosenberg et al., 2000). This does not just
mean cases involving exceptionally violent biographical experiences. Recurrent
severe seizures themselves may represent a directly traumatic experience in a
stricter sense, particularly when as in many frontal epilepsies the patient retains
a considerable degree of consciousness. Several of the characteristics of PTSD men-
tioned in DSM-IV (American Psychiatric Association, 1994) such as the numbing
of general responsivity, constriction of aect, re-experiencing the traumatic event
or an exaggerated startle response can be considered characteristic of many epilep-
sies as well. This makes it all the more surprising that so little research has focused
on the potentially traumatizing character of seizures. One reason for this may be
the well-known fact in psychotraumatology that the more destructive stressful
events have been, the more dicult they are to express in words. Contraindications
Table 21.1. Follow-up of inpatient epilepsy treatments
Without psychotherapeutic With psychotherapeutic

add-on add-on
6 months after discharge 12 months after discharge

n43 n34
Completely seizure-free 5 11
Note: Exact Fisher test: P0.025
are given particularly when professional and institutional preconditions are inad-
equate. Commencing treatment without the right setting (founded psychotherapy
training, dened session times, discussions regarding the duration, form, and goals
of treatment, supervision) borders on malpractice. Likewise, one should not com-
mence psychotherapy with an epilepsy patient without being aware of the multi-
tude of psychopathological phenomena that may occur in epilepsies. Finally,
supplementary psychotherapy cannot be recommended in an inpatient treatment
framework when this is not supported by the institution, and sound framing con-
ditions are not provided.
Effective or only well-intended?

In my opinion, the above considerations justify conceiving psychotherapy in an epi-
lepsy centre within an interdisciplinary framework and practising it as a supplement
to conventional treatment hence, fashionably speaking, in the sense of an add-on.
However, to examine the methodologically dicult question as to whether its eec-
tiveness can be conrmed, we (the head of the nursing service, H. Welteke, and the
author) carried out a follow-up study. Seizure-related outcomes were surveyed in 34
patients at least 12 months after discharge from inpatient treatment. Findings were
classied with the adapted version of the so-called Engel Scale (Engel, 1987) for drug
treatment histories (categories are: Class I, completely free of disabling seizures;
Class II, rare disabling seizures (almost seizure-free); Class III, quantiable worth-
while improvement; and Class IV, no measurable improvement in life quality).
Results showed that 11 (32.4%) former patients were completely free of seizures,
a further 4 (11.8%) were almost seizure-free, 6 (17.6%) had experienced a worth-
while improvement, whereas the remaining 13 (38.2%) had to be classied to
Engels Class IV.
It is interesting to compare our ndings with a parallel 6-month follow-up in the
rest of the Bethel Epilepsy Centre (Pfin and May, in press). This study reports a
complete freedom from seizures in only 12.4% of patients. Table 21.1 compares the
data.
322 M. Schndienst
This suggests that the number of optimal treatment outcomes was signicantly
higher in the group with supplementary psychotherapy at an alpha level of 0.05. It
can be concluded that there now is less need to justify provision of a psychothera-
peutic add-on for epilepsy patients, but, instead, more need to justify why the
potentials of such treatments often remain unexploited.
R E F E R E N C ES
(fourth revision) (DSMIV). Washington, DC: APA.
Engel, J. Jr (1987). Outcome with respect to epileptic seizures. In Surgical Treatment of the Epi-
lepsies, ed. J. Engel Jr. New York: Raven Press.
Engel, J. and Pedley, T. (1997). Epilepsy A Comprehensive Textbook, Vol. IIII. Philadelphia, New
York: Lippincott-Raven.
Pfin, M. and May, T. (in press). Comprehensive care in an epilepsy clinic. In Comprehensive
Care for People with Epilepsy, ed. R.T. Fraser, M. Pfin, R. Thorbecke, U. Specht and P. Wolf.
London: John Libbey.
Rosenberg, H.J., Rosenberg, S., Williamson, P. and Wolford, G. (2000). A comparative study of
trauma and posttraumatic stress disorder prevalence in epilepsy patients and psychogenic
nonepileptic seizure patients. Epilepsia, 41, 44752.
Stagno, S.J. (1993). Psychiatric aspects of epilepsy. In The Treatment of Epilepsy: Principles and
Practice, ed. E. Wyllie. Philadelphia, New York: Lea and Febiger.
22
Choosing measures to assess quality of life

(QOL) in epilepsy
Caroline E. Selai, Katja Elstner and M.R. Trimble
Why assess quality of life?
Whilst quality of life (QOL) measures have been developed for a number of reasons
(Fitzpatrick et al., 1992), two basic aspects of health care underlie most of the ques-
tions that QOL appraisals set out to answer: outcome of treatment and cost. With
increasingly sophisticated life-saving and life-prolonging medical interventions,
and a range of options between alternative treatments, quality of life has emerged
as an important outcome. Also, it is argued that no country in the world can aord
to do all that it is technically possible to do to improve the health of its citizens and
so the need has arisen for some system of setting priorities. Quality of life and other
outcome data are informing health economic decisions and debate about the allo-
cation of scarce resources. The eld of QOL research is thriving, and much progress
has been made in the last 10 years.
Types of QOL measure
There is no gold standard for measuring QOL and the range of instruments avail-
able, or still undergoing development, is remarkable in terms of both quantity and
heterogeneity. The range of categories of QOL/health status measures has been
comprehensively reviewed elsewhere (Brooks, 1995). In brief, generic instruments
cover a broad range of QOL domains in a single instrument. Their chief advantage
is in facilitating comparisons among dierent disease groups. Disease-specic
instruments reduce patient burden by including only relevant items for a particu-
lar illness but their main disadvantage is the lack of comparability of results with
those from other disease groups.
Health proles provide separate scores for each of the dimensions of QOL,
whereas a health index, a type of generic instrument, gives a single summary score,
usually from 0 (death) to 1 (perfect health).
323
324 C.E. Selai, K. Elstner and M.R. Trimble
Another type of measure, developed within the economic tradition, is the utility
measure. Whilst some decisions are taken for individual patients, others, such as
those made by health policy makers, concern groups of patients. Here the focus is
on society as a whole and the societal allocation of scarce resources. For this
purpose, preference-weighted measures are required.
Finally, some researchers have questioned the appropriateness of the xed ques-
tionnaire to assess QOL. They argue that, since quality of life is a uniquely personal
perception, most standardized measurements of QOL in the medical literature
seem to aim at the wrong target (Gill, 1995). Their argument, that any technique
to assess QOL should be tailored to the individual respondent, is discussed further
below.
Individualized, patient-driven QOL assessment techniques
There is a fundamental tension in the measurement of QOL. Since what is deemed

important for QOL is acknowledged to be subjective and idiosyncratic, dierences
being inuenced by a variety of personal and cultural factors, an appraisal of QOL
should strive to capture the individuals subjectively appraised phenomenological
experience. On the other hand, the hallmark of scientic measurement is reliable,
objective, empirical data collection. Researchers have devised QOL measurement
techniques at various stages of this subjective/objective continuum and there now
exist over 1000 instruments that have been developed taking a variety of approaches
to measurement (Hedrick et al., 1996). The two poles of the qualitativequantitative
continuum have dierent strengths. It has been suggested that qualitative methods
are more valid whilst quantitative methods are more reliable (Mays and Pope, 1996).
The QOL literature advocates a robust and rigorous programme of instrument
development and testing, and most QOL measures are developed within the
psychometric tradition (Juniper et al., 1996; McDowell and Newell, 1987).
However, since quality of life is uniquely personal, most standardized measure-
ments of QOL are inappropriate (Gill, 1995). It is argued that quality of life can be
suitably measured only by determining the opinions of patients and by supple-
menting (or replacing) the instruments developed by experts (Gill and Feinstein,
1994). Scales developed within the psychometric tradition often omit items impor-
tant to the beliefs and values of individual patients (Gill, 1995; Hunt, 1999) and the
psychometric aim of internal reliability is in conict with the goals of achieving
comprehensiveness and content validity (Brazier and Deverill, 1999). In response
to this, a number of individualized, patient-driven techniques have been devel-
oped whereby the patient can nominate items of importance to him/herself (Fraser
et al., 1993; Geddes et al., 1990; Guyatt et al., 1987a, b; OBoyle et al., 1993; Ruta et
325 Measures to assess quality of life in epilepsy
al., 1994; Tugwell et al., 1990). The literature on individual QOL assessment tech-
niques has recently been reviewed (Joyce et al., 1999).
The choice of measure will depend upon the goal of the study; a common rec-
ommendation is to include both disease-specic and generic measures in an inves-
tigation.
QOL measures for use in epilepsy

The history, development and current status of measures to assess health-related
quality of life (HRQOL) in epilepsy have been comprehensively reviewed (Cramer,
1996; Hermann, 1995; Trimble and Dodson, 1994). It is well documented that the
occurrence of seizures and stigma of epilepsy lead to a variety of physical, psycho-
logical and social problems which have an impact on many areas such as self-
esteem, work, career prospects, family, leisure, (in)ability to drive and relationships
(Cramer, 1996).
Therapeutic outcome: change in seizure frequency
Whilst approximately 70% of patients are well controlled on monotherapy, with

standard antiepileptic drugs (AEDs), for the remaining 30% of patients polyther-
apy is considered. Surgery (mainly temporal lobectomy) is another option for some
patients with intractable epilepsy.
Drug trials
In clinical trials, a change in seizure frequency has traditionally been the main
measure of ecacy. Those patients who experience a 50% reduction in seizure
frequency are described as responders whilst all other participants are nonres-
ponders (Smith et al., 1995). QOL has rarely been an outcome measure in clinical
trials of AEDs.
Surgery
More stringent criteria have been adopted for the seizure-dened outcome after
epilepsy surgery. In addition to seizure freedom, some researchers have chosen a
75% (Bladin, 1992; Hermann et al., 1992; Malgrem et al., 1997) and some a 90%
reduction in seizures (McLachlan et al., 1997; Rose et al., 1996).
The two studies reported here were designed to look at the relationship between
changes in seizure frequency (clinically dened end-point) and changes in QOL.
The data were collected from patients attending clinics at the National Hospital for
Neurology, Queen Square. The main thrust of this chapter is to compare the sensi-
tivity of the Quality of Life Assessment Schedule (QOLAS), an individualized,
patient-driven technique, to the EuroQol EQ-5D, a generic, single index which

yields a single, summary score.
Study 1: patients starting adjunctive AED therapy
Drug study: subjects and methods

Patients attending a follow-up appointment at the outpatient epilepsy clinics at
Queen Square were approached after their medical consultation. The patients
recruited were about to start on one of the ve antiepileptic drugs (vigabatrin, clob-
azam, lamotrigine, gabapentin and topiramate) as add-on therapy. Those willing
to take part, and who gave informed consent, were oered the choice of a telephone
interview at home as an alternative to a face-to-face interview and most patients
chose this option. The timing of the interviews was: (i) baseline, (ii) 3 months from
baseline and (iii) 6 months from baseline. Only the baseline and 6-months follow-
up data are reported here. Seizure frequency and seizure severity were assessed
using the National Hospital Seizure Severity Scale (ODonoghue et al., 1996).
Quality of Life (QOL)

QOL was measured by the QOLAS (Kendrick and Trimble, 1994) and the EuroQol
instrument, the EQ-5D (Brooks, 1996; EuroQol Group, 1990). These are briey
described below.
The QOLAS
The Quality of Life Assessment Schedule (QOLAS) is an individualized QOL
assessment technique which is tailored to each individual patient, and is a revised
version of the QoLASCA, a method originally based on the Repertory Grid
Technique (Kendrick, 1997; Kendrick and Trimble, 1994; McGuire, 1991). The full
QoLASCA technique was somewhat burdensome and the revised method
(QOLAS) has been considerably streamlined. Two main aspects of the original
theoretical work have been maintained: (i) the original emphasis (in order to assess
therapeutic outcome) on a careful and comprehensive interview, recording items
of importance to the patient in the patients own words; (ii) the idea that QOL is a
function of the conceptual distance between how I am now and how I would like
to be, the gap between actuality and expectation. This is known in the medical lit-
erature as Calmans gap, since Calman suggested that a key aim of medical care
should be to narrow the gap between a patients hopes and expectations and what
actually happens (Calman, 1994).
The QOLAS has been used in studies of patients with epilepsy (Selai and
Trimble, 1998; Selai et al., 2000), dementia (Selai et al., 2001) and Gilles de la
Tourette syndrome (Elstner et al., 2001).
QOLAS interview
The QOLAS interview used in this study is as follows:
1. Introduction and rapport-building.
2. The respondent is invited to recount what is important for his/her QOL and ways
in which their current health condition is aecting their QOL. Key constructs are
extracted from this narrative. Prompting is sometimes required.
3. In total, ten constructs are elicited, two for each of the following domains of
QOL: physical, psychological, social, daily activities and cognitive functioning
(or well-being).
4. The patient is asked to rate how much of a problem each of these is now on a 05
scale where 0no problem; 1very slight problem; 2mild problem; 3
moderate problem; 4big problem and 5it could not be worse.
5. The patient is asked to rate how much of a problem they would like each of these
to be ideally on a 05 scale as above.
6. At follow-up interview, the respondents individual constructs are read out to
them and they are invited to re-rate each on the 05 scale for how much of a
problem there is with each now.
QOLAS scoring
(i) For each construct, the like score is subtracted from the now, giving a score
for the distance between expectation and reality.
(ii) The scores, calculated in (i) above, for the two constructs per domain are
summed to give a domain score out of ten. The total for each of the ve domains
is summed to give an overall QOLAS score out of 50.
The EuroQol EQ-5D

The EQ-5D is a generic instrument for describing and evaluating health-related
quality of life, developed to complement other forms of quality of life measure, and
to generate a cardinal index of health, thus giving it considerable potential for use
in economic evaluation (Brooks, 1996; EuroQol Group, 1990). It has three compo-
nents, each providing separate data. In the rst part, which yields a simple descrip-
tive prole, the respondent rates his own health today on ve questions, one for each
of the dimensions: mobility, self-care, usual activities, pain/discomfort and
anxiety/depression. Each question has three response options: no problems, some
problem and extreme problems. This descriptive classication thus denes 243 pos-
sible health states. The respondent next rates his own health, today, on a visual ana-
logue scale (VAS), calibrated from 0100 with the end-points 100best imaginable
health state and 0worst possible health state. Finally, valuations for each of the 243
health states have been obtained and so, according to how the respondent has rated
himself on the descriptive prole, the corresponding utility value can be ascertained.
The EQ-5D has been used in a number of clinical studies, but in only one study
in epilepsy, to our knowledge. In this study, the EQ-5D VAS was used in a compar-
ison of four preference measures (Stavem, 1998) but data on the HRQOL of
patients was not presented.
Side eects, adverse events, dened as any epilepsy-related health event requir-
ing urgent medical attention, and the reason for stopping medication were also
recorded. The outcome criterion, chosen after review of the current literature, was
50% reduction in seizures. We do not report the relative performance of each
individual drug.
Drug study: statistics

The data were analysed by the chi-square statistic or paired t-tests, two-tailed, as
appropriate. The QOLAS data were normally distributed, as were the VAS scores
(although the latter were slightly skewed). The EQ-5D utility data were markedly
skewed and so the nonparametric Wilcoxon signed-ranks test was performed.
Drug study: results

A total of 146 patients who were about to start on one of the following ve AEDs
were recruited: vigabatrin, clobazam, lamotrigine, gabapentin and topiramate.
These included 125 patients on whom complete data for 6 months were collected
and 21 patients who failed to attend follow-up. The mean age of the patients was
37.2. Thirty-ve per cent of patients were working or studying (at least part-time).
Of the 125 patients, 15 started on vigabatrin (of which 10 were male), 20 on clob-
azam (8 male), 26 on lamotrigine (14 male), 17 started on gabapentin (8 male) and
47 on topiramate (28 male). Of the other 21 patients, 3 had started on vigabatrin,
4 on clobazam, 6 on gabapentin, 6 on lamotrigine and 2 on topiramate.
Drug study: seizures at baseline

Almost half of the patients (49%) had experienced more than one type of seizure in
the last month and 8% had experienced more than two seizure types. At baseline,
most of our patients were classied as having severe epilepsy (Vickrey et al., 1995).
A total of 115 patients (92%) reported having more than 12 seizures in the previous
year, 6 patients (5%) were having from 212 seizures, no patients were seizure-free
or having only auras. The remaining patients (3%) reported having seizures in their
sleep and were unable to report seizure frequency. The percentage of patients in each
outcome group experiencing convulsions at baseline was identical (both 46%).
Drug study: seizure reduction

Table 22.1 summarizes the status of the patients at the 6-month follow-up. Forty-six
patients (37%) had achieved 50% reduction in seizures (27 male; 19 female). There
Table 22.1. Drug study: clinical status at 6-months follow-up
Status at 6 months Patients (n125) %
Still on drug 75 60
Experiencing side eectsa 49 39
Experienced serious adverse eventsb 15 12
50% or more reduction in seizures 46 37
Did not attend follow-up interview 21
Notes:
a
Side eects as reported by patients and attributed by them to the add-on therapy.
b
Serious adverse events are epilepsy-related events requiring urgent medical intervention.
Table 22.2. Drug study: summary of outcome measures scores
50% reduction 50% reduction 50% reduction 50% reduction

baseline follow-up baseline follow-up
Outcome measure mean (SD) mean (SD) mean (SD) mean (SD)
QOLAS 31.85 (10) 23a (11) 32.85 (8) 30.85 (11)

EQ-VAS 67.85 (23) 75b (17) 63.85 (19) 64.85 (20)
EQ-utility 0.86 (0.20) 0.89 (0.16) 0.85 (0.16) 0.85 (0.19)
EQ-utility median 1 1 0.85 0.85
Notes:
a
P0.001, b P0.02.
was no signicant dierence in age between those patients who achieved 50% seizure
reduction (mean age36 years) and those who did not (mean age38 years).
Drug study: quality of life measures

The three issues most frequently raised in each of the ve QOLAS domains are
reported in Appendix 22.1 (a full report on these qualitative data is available from
the authors).
Table 22.2 summarizes the quantitative QOLAS and EQ-5D results for this study.
The group whose seizures were reduced by 50%, showed signicantly lower
QOLAS scores i.e. a signicant improvement in HRQOL, at follow-up compared
to baseline (t6.18, P0.001).
Drug study: EQ-5D profile scores

Tables 22.3, 22.4 and 22.5 report descriptive EQ-5D data. Table 22.3 shows the
baseline prole scores, for each of the ve domains, for the whole group (n125)
Table 22.3. Drug study: comparison of baseline EQ-5D profile data with UK norms.
Percentage reporting no problems
Our study (n125) UK survey (n3395)

EQ domain % %
Mobility 88 82
Self-care 93 96
Usual activities 78 84
Pain/discomfort 82 67
Anxiety/depression 65 79
Table 22.4. Drug study: EQ-5D descriptive health profile data. 50% seizure reduction
group: baseline and follow-up (n46)
No problems Some problems Severe/extreme
EQ domain t1 t2 t1 t2 t1 t2
Mobility 40 (87)a 39 (85) 6 (13) 7 (15) 0 (0) 0 (0)

Self-care 43 (93) 44 (96) 3 (07) 2 (04) 0 (0) 0 (0)
Usual activities 35 (76) 38 (83) 9 (20) 8 (17) 2 (4) 0 (0)
Pain/discomfort 36 (78) 39 (85) 10 (22) 7 (15) 0 (0) 0 (0)
Anxiety/depression 36 (78) 29 (63) 8 (17) 16 (35) 2 (5) 1 (2)
Note:
a
Figures in parentheses are percentage of patients.
Table 22.5. Drug study: EQ-5D descriptive health profile data. < 50% seizure reduction
group: baseline and follow-up (n79)
No problems Some problems Severe/extreme
EQ domain t1 t2 t1 t2 t1 t2
Mobility 70 (89)a 70 (89) 9 (11) 9 (11) 0 (0) 0 (0)

Self-care 73 (92) 75 (95) 6 (08) 4 (5) 0 (0) 0 (0)
Usual activities 63 (80) 68 (86) 15 (19) 11 (14) 1 (1) 0 (0)
Pain/discomfort 66 (84) 61 (77) 13 (16) 16 (20) 0 (0) 2 (3)
Anxiety/depression 45 (57) 45 (57) 30 (38) 31 (39) 4 (5) 3 (4)
Note:
a
Figures in parentheses are percentage of patients.
in comparison to UK normative data (Kind et al., 1998). This table shows the
number of patients reporting no problem for each of the EQ-5D domains. In our
study, fewer patients reported no problem in the domain anxiety/depression
compared with the UK survey. On the other hand, more patients reported no
problems with pain/discomfort than the UK survey. Table 22.4 shows descriptive
data comparing baseline to follow-up for the group who achieved 50% seizure
reduction (n46). Table 22.5 shows descriptive data comparing baseline to follow-
up for the group who did not achieve 50% reduction in seizures. At follow-up, there
was no signicant dierence between the two outcome groups on the health prole.
Drug study: EQ-5D VAS scores

Most patients queried the VAS. Forty-seven per cent of patients said they thought
that health did not include their epilepsy. These patients said that if the VAS was to
include epilepsy the score would be up to 70 VAS points lower. Table 22.2 sum-
marizes the EQ-VAS scores for the two groups of patients at baseline and follow-up.
The group whose seizures were reduced by 50%, showed signicantly higher VAS
scores i.e. a signicant improvement in HRQOL/health status (t2.48, P0.02).
Drug study: EQ-5D utility scores

Table 22.2 summarizes the EQ-utility scores for the two groups of patients at base-
line and follow-up. The group whose seizures were reduced by 50% or more did
not show a statistically signicant improvement on the utility score (Wilcoxon
Z0.470; P0.64).
Drug study: summary

The results of this study suggest that HRQOL improves in patients with severe epi-
lepsy on adjunctive treatment who experience a 50% or greater seizure reduction.
This is the rst study to report health status using the EQ-5D in patients with epi-
lepsy. The QOLAS and the EQ-VAS were sensitive to change but the EQ-5D prole
and EQ-5D utility were not responsive. Compared with a large UK survey, a similar
percentage of patients in our study reported any problem in each EQ-5D domain.
Fewer patients in our study reported any pain/discomfort compared with the UK
survey but more of our patients reported any problem with anxiety/depression.
Study 2: epilepsy surgery
Surgery: subjects and methods

A total of 145 patients undergoing evaluation for denitive treatment for intract-
able epilepsy were interviewed during their stay on the telemetry unit of the
National Hospital, Queen Square. Quality of life was assessed using the Quality of
Life Assessment Schedule (QOLAS); the Epilepsy Surgery Inventory (ESI-55) and
the EuroQol EQ-5D.
QOL measures
The QOLAS and the EQ-5D have been described above.
The Epilepsy Surgery Inventory (ESI-55)

The ESI-55 consists of the generic SF-36 plus a number of epilepsy-specic ques-
tions (Vickrey et al., 1992). The scoring produces 11 subscales (health, energy,
QOL, social functioning, emotional functioning, cognitive functioning, role-
emotional, role-memory, role-physical, physical function and pain) and three com-
posite scores for physical health, mental health and role functioning.
Surgery study: statistics

The data were analysed by the chi-square statistic or paired t-tests, two-tailed, as
appropriate. The EQ-5D utility data were markedly skewed and so the nonpara-
metric Wilcoxon signed ranks test, was performed. Criterion validity and internal
reliability were assessed using Spearmans rank correlations and the coecient
alpha. The dierences between the QOLAS scores at baseline and follow-up were
tested using the Wilcoxon signed ranks test.
Epilepsy surgery: results

Seizure reduction
A subgroup of 40 patients was interviewed at follow-up (mean time to follow-
up1 year). Of these 40 patients, 15 had not had surgery at follow-up interview
and 25 were postsurgery. Sixteen (64%) of the surgical patients had left temporal
lobe resection, four patients (16%) had right temporal lobe resection and 5 (20%)
had extra temporal lobe resection. Of the 25 patients who had undergone surgery,
22 had 75% reduction in seizures and 3 patients did not. Of the 15 patients who
had not gone on to have surgery, no patient achieved a 75% seizure reduction.
These data are summarized in Table 22.6.
The mean duration of epilepsy in years was 23.1 (SD 9.9) and the median was 24
years. The mean age was 32.8 years (SD 8.6) and the median 31.0 years. The mean
follow-up period was 14.3 months (SD 8.7) and median 12.5 months. The mean
age of onset of epilepsy was 9.6 years (SD 9.4) and the median was 8 years. Table
22.7 summarizes the patient outcome data.
QOLAS scores
The results for the QOLAS at baseline and at follow-up are presented in Table 22.7.
There was signicant improvement in all QOLAS domains at follow-up in the
operated group who achieved 75% or greater reduction in seizure frequency.
Table 22.6. Epilepsy surgery: summary of patient data
75% seizure reduction 75% seizure reduction
Number of patients 18 22
Male 8 7
Female 10 15
No surgery 15 0
Surgery 3 22
Table 22.7. Epilepsy surgery: summary of outcome measures scores
75% reduction 75% reduction No surgery No surgery

Outcome mean (SD) mean (SD) mean (SD) mean (SD)
measure t1 t2 t1 t2
QOLAS 32.3 (8.0) 17.1a (8.8) 31.3 (6.7) 29.3 (8.3)

EQ-VAS 61.6 (20.3) 76.6a (15.6) 64.6 (17.4) 69.1 (13.6)
ESI-CMH 62.2 (14.3) 74.8b (12.1) 59.9 (14.9) 63.4 (14.5)
ESI-CPH 73.2 (14.0) 82.9b (11.6) 65.6 (26.8) 71.1 (18.3)
ESI-CRF 69.6 (22.9) 78.5b (20.8) 56.5 (27.1) 67.4 (27.9)
Notes:
ESI-CMH, ESI-55 composite mental health score; ESI-CPH, ESI-55 composite physical health
score; ESI-CRFESI-55 composite role functional score.
a
P0.05, b P0.01.
ESI-55 composite scores

Table 22.7 summarizes the ESI-55 composite scores. At follow-up, there was a sta-
tistically signicant improvement in QOL in comparison with baseline scores, on
two of three ESI-55 composite scores: Composite Mental Health (CMH) t4.3;
df21, P0.0001, 95% CI (18.7; 6.5); Composite Physical Health (CPH)
t4.4, df20, P0.0001, 95% CI (14.8; 5.3). Although Composite Role
Functioning (CRF) scores showed improvement at follow-up, they did not reach
statistical signicance.
EQ-5D
Tables 22.8 and 22.9 show descriptive EQ-5D prole data at baseline and at follow-
up.
As with the drug study, most patients queried the EQ-5D visual analogue scale
(VAS). Forty-two per cent of patients said they thought that health did not
include their epilepsy. Again, if the VAS was to include their epilepsy, they would
Table 22.8. Baseline EQ-5D profile data (n22)
EQ domain No problema Some problems Severe problems
Mobility 86 9 5
Self-care 86 14 0
Usual activities 72 18 9
Pain/discomfort 82 18 0
Anxiety/depression 59 32 9
Note:
a
Figures are percentage of patients.
Table 22.9. EQ-5D profile data at follow-up
EQ domain No problemsa Some problems Severe problems
Mobility 90 10 0
Self-care 100 0 0
Usual activities 89 11 0
Pain/discomfort 85 15 0
Anxiety/depression 80 20 0
Note:
a
Figures are percentage of patients.
have given a score up to 70 points lower on the VAS scale. For example, one patient
said that his health was excellent in general and he felt well on the day of the inter-
view so he scored himself as 80. On reection, he added that if the score was sup-
posed to include his epilepsy and seizures, then he would have adjusted it to 30.
Although we noted the qualitative data, we took the score each patient originally
gave for their health since this is what the EQ-5D asks. Table 22.7 summarizes the
EQ-5D VAS scores for the two groups of patients at baseline and follow-up. There
was signicant improvement at follow-up (t2.6, df20, P0.02, 95% CI
(26.0; 2.8)).
Table 22.10 shows the baseline prole scores for the whole group (n145) in
comparison with UK normative data (Kind et al., 1998). Fewer patients in our
surgery study reported no problem on most of the EQ-5D domains compared
with the UK survey. A marked dierence was observed for the anxiety/depression
domain with 49% of the surgery patients reporting no problem compared to 79%
of the UK population.
Table 22.10. Surgery study: comparison of baseline EQ-5D profile data with UK norms.
Percentage reporting no problems
EQ domain Surgery study (n145) UK survey (n3395)

% %
Mobility 84 82
Self-care 86 96
Usual activities 70 84
Pain/discomfort 81 67
Anxiety/depression 49 79
Table 22.11. Epilepsy surgery: summary of outcome measures scores
EQ-Utility 75% reduction No operation
Median Mean (SD) Median Mean (SD)
t1 0.85 0.81 (0.31) 0.85 0.77 (0.24)

t2 1.00 0.91 (0.11) 0.85 0.90 (0.11)
EQ-5D utility scores

The EQ-5D utility scores at baseline and at follow-up were: mean 0.81 (median
0.85) and mean 0.91 (median 1.0) respectively (Table 22.11). The dierence was
not signicant.
Summary
The gures for the epilepsy surgery study are similar to the gures for the anti-
epileptic drug audit. The QOLAS, the EQ-VAS and 2/3 ESI-55 Composite Scores
were sensitive to change as shown by statistically signicant changes in scores. For
the two patient groups there were dierences in the EQ-5D prole at baseline but
the percentages of problems experienced by the two groups at follow-up were
similar. The EQ-5D utility scores showed improvement but the changes were not
signicantly dierent and this nding requires further discussion.
Discussion
The results of this study suggest that HRQOL improves both in patients with severe
epilepsy on adjunctive treatment who experience a 50% seizure reduction and in
patients who have undergone epilepsy surgery who achieved a 75% seizure
reduction. The QOLAS, the EQ-VAS and two out of three subscales of the ESI-55
were sensitive to change but the EQ-5D prole and EQ-5D utility were not respon-
sive. There are two main possible explanations for our ndings.
Choice of clinical outcome

There is no simple relationship between seizure severity, seizure frequency and the
consequences of epilepsy, and there is debate whether a reduction in (but not elim-
ination of) seizures does lead to an improvement in HRQOL (Smith et al., 1995).
Even if a new agent reduces the seizure frequency by 50%, patients will continue
to experience seizures, and these may be unpredictable with severe ictal or post-
ictal phenomena (Baker, 1995). It has been suggested that the ultimate goal of new
antiepileptic drugs, therefore, should be seizure freedom (Walker and Sander,
1996).
Patients about to undergo surgical treatment for epilepsy often have high, but
not necessarily realistic, expectations of signicant positive changes postopera-
tively (Baxendale and Thompson, 1996). The EQ-5D might not be sensitive to the
relatively modest changes perceived after reduction in (but not elimination of) sei-
zures.
Ability of the QOLAS and the EQ-5D to detect change in HRQOL

The QOLAS asks patients to nominate the HRQOL topics of concern to them. They
are able to both choose the items and, importantly, discuss them in their own lan-
guage or idiom. The QOLAS therefore taps each patients perceived change in their
own HRQOL rather than objectively veriable changes in status e.g. role/social
functioning.
The EQ-5D, a generic instrument, might not have basic face validity for partic-
ular patient populations, and epilepsy in particular. Comparing our audit data with
the UK norms, the only domain where the epilepsy patients reported signicantly
more problems was the anxiety/depression domain. Numerous studies, however,
have shown that seizures and the stigma of epilepsy considerably impair HRQOL,
and HRQOL is certainly poor in patients with intractable epilepsy, who are taking
a cocktail of antiepileptic drugs, and who have been referred to a centre of tertiary
referral such as Queen Square. In the surgery study, where the patients were inter-
viewed in their hospital beds attached to video-telemetry monitoring equipment,
more problems (compared to UK norms) were reported in most of the EQ-5D
domains with a marked dierence in percentages reported for the anxiety/depres-
sion domain. Nevertheless, the utility scores for the patients who achieved 75%
seizure reduction were not signicantly dierent from baseline.
For patients with intractable epilepsy, the occurrence of seizures and stigma of
epilepsy lead to a variety of physical, psychological and social problems which have
impact on many areas such as self-esteem, career prospects, family, leisure,

(in)ability to drive and relationships (Cramer, 1996). All of our patients mentioned
some problems due to epilepsy with at least one of: work; studies; career; house-
work; child-care; leisure; sports; relationships, family and social life, yet 78% of the
audit patients (and 70% of the surgery patients being interviewed in their hospital
beds) ticked no problem with the EQ-5D Usual Activities domain at baseline. An
explanation could be that the EQ-5D might be more useful for acute rather than
chronic illness since it does not capture chronic problems to which the patient has
adapted. This phenomenon of coping/adjustment resulting in the underreporting
of problems on HRQOL instruments by patients with epilepsy has been previously
discussed (Devinsky et al., 1995). Because the EQ-5D utility score is dependent
upon the EQ-5D prole, and given the problems outlined above, it is not surpris-
ing that the utility scores were not sensitive to changes in seizure outcome.
Most of our patients with epilepsy queried the VAS. The most common
comment (47% of patients) was that health does not include epilepsy. Most of
these patients said that, if the VAS was to include epilepsy, the score would be up to
70 points lower. Even though the VAS was sensitive to change, these qualitative data
raise questions about the interpretation of the numerical data obtained from the
VAS in this patient group. It has previously been reported that patients with epi-
lepsy have diculties completing visual analogue scales (Falloweld, 1994).
Conclusions
The QOLAS, an individualized, patient-centred measure was demonstrated to be

sensitive to change.
The EQ-5D VAS, which is measuring some aspect of health-related HRQOL, is
sensitive to change. Our data suggest, however, that the EQ-5D does not have face
validity for patients with severe epilepsy. The EQ-5D prole, and, therefore, the
EQ-5D utility scores, might not have validity for this patient group. We recommend
that further research is required into the suitability of generic measures to assess
QOL in patients with epilepsy. Meanwhile, we would urge all researchers to be cau-
tious in their choice of QOL measure.
Acknowledgements
For assistance with recruitment of patients we gratefully acknowledge the assis-

tance of: Prof. S. Shorvon, Prof. J. Duncan, Dr J.W.A. Sander, Dr D. Fish and Dr S.
Smith. Thanks also to Dr M. ODonoghue who provided training in the use of the
National Hospital Seizure Severity Scale. An earlier version of this paper was pre-
sented at the EuroQol Plenary meeting in Hannover, 13 October 1998.
R E F E R E N C ES
Baker, G. (1995). Health-related quality of life issues: optimizing patient outcomes. Neurology,
45 (Suppl. 2), S2934.
Baxendale, S.A. and Thompson, P.J. (1996). If I didnt have epilepsy . . .: patient expectations of
epilepsy surgery. J Epilepsy, 9, 27481.
Bladin, P.F. (1992). Psychosocial diculties and outcome after temporal lobectomy. Epilepsia, 33,
898907.
Brazier, J. and Deverill, M. (1999). A checklist for judging preference-based measures of health
related quality of life: learning from psychometrics. Health Econ, 8, 4151.
Brooks, R.G. (1995). Health Status Measurement: A Perspective on Change. Basingstoke:
Macmillan.
Brooks, R. (1996). EuroQol: The current state of play. Health Policy, 37, 5372.
Calman, K.C. (1994). Quality of life in cancer patients: an hypothesis. J Med Ethics, 10, 1247.
Cramer, J. (1996). Quality of life assessment for people with epilepsy. In Quality of Life and
Pharmacoeconomics in Clinical Trials, Second edition, ed. B. Spilker, pp. 90918. Philadelphia,
New York: Lippincott-Raven.
Devinsky, O., Vickrey, B.G., Cramer, J. et al. (1995). Development of the quality of life in epilepsy
inventory. Epilepsia, 36, 1080104.
Elstner, K., Selai, C.E., Trimble, M.R. and Robertson, M.M. (2001). Quality of life of patients with
Gilles de la Tourette syndrome. Acta Psychiatr Scand, 103, 529.
EuroQol Group (1990). EuroQol a new facility for the measurement of health-related quality
of life. Health Policy, 16, 199208.
Falloweld, L. (1994). An overview of quality of life measurements. In Epilepsy and the Quality
of Life, ed. M.R. Trimble and W.E. Dodson, pp. 8598. New York: Raven Press.
Fitzpatrick, R., Fletcher, A.E., Gore, S.M., Jones, D.R., Spiegelhalter, D.J. and Cox, D.R. (1992).
Quality of life measures in health care. I: applications and issues in assessment. Br Med J, 305,
10747.
Fraser, S.C.A., Ramirez, A.J., Ebbs, S.R. et al. (1993). A daily diary card for quality of life meas-
urement in advanced breast cancer trials. Br J Cancer, 67, 3416.
Geddes, D.M., Dones, L., Hill, E. et al. (1990). Quality of life during chemotherapy for small cell
lung cancer: assessment and use of a daily diary card in a randomised trial. Eur J Cancer, 26,
48492.
Gill, T.M. (1995). Quality of life assessment: values and pitfalls. J R Soc Med, 88, 6802.
Gill, T.M. and Feinstein, A.R. (1994). A critical appraisal of the quality of quality of life measure-
ments. J Am Med Assoc, 272, 61926.
Guyatt, G.H., Berman, L.B., Townsend, M., Pugsley, S.O. and Chambers, L.W. (1987a). A
measure of quality of life for clinical trials in chronic lung disease. Thorax, 42, 7738.
Guyatt, G.H., Townsend, M., Pugsley, S.O. et al. (1987b). Bronchodilators in chronic air-ow lim-
itation. Am Rev Resp Disorders, 135, 106974.
Hedrick, S.C., Taeuber, R.C. and Erickson, P. (1996). On learning and understanding quality of
life: a guide to information sources. In Quality of Life and Pharmacoeconomics in Clinical Trials,
Second edition, ed. B. Spilker, pp. 5964. Philadelphia: Lippincott-Raven.
Hermann, B.P. (1995). The evolution of health-related quality of life assessment in epilepsy. Qual
Life Res, 4, 87100.
Hermann, B.P., Wyler, A.R and Somes, G. (1992). Preoperative psychological adjustment and
surgical outcome are determinants of psychosocial status after anterior temporal lobectomy. J
Neurol Neurosurg Psychiatry, 55, 4916.
Hunt, S. (1999). The researchers tale: a story of virtue lost and regained. In Individual Quality of
Life: Approaches to Conceptualisation and Assessment, ed. C.R.B. Joyce, C.A. OBoyle and H.
McGee, pp. 22532. Amsterdam: Harwood Academic Publishers.
Joyce, C.R.B., OBoyle, C.A. and McGee, H. (eds.) (1999). Individual Quality of Life: Approaches
to Conceptualisation and Assessment. Amsterdam: Harwood Academic Publishers.
Juniper, E.F., Guyatt, G.H. and Jaeschke, R. (1996). How to develop and validate a new health-
related quality of life instrument. In Quality of Life and Pharmacoeconomics in Clinical Trials,
Second edition, ed. B. Spilker, pp. 4956. Philadelphia, New York: Lippincott-Raven.
Kendrick, A. (1997). Quality of life. In The Clinical Psychologists Handbook of Epilepsy, ed. C. Cull
and L.H. Goldstein, pp. 13048. London: Routledge.
Kendrick, A.M. and Trimble. M.R. (1994). Repertory grid in the assessment of quality of life in
patients with epilepsy: the Quality of Life Assessment Schedule. In Epilepsy and the Quality of
Life, ed. M.R. Trimble and W.E. Dodson, pp. 15163, New York: Raven Press.
Kind, P., Dolan P., Gudex, C. and Williams, A. (1998). Variations in population health status:
results from a United Kingdom national questionnaire survey. B Med J, 316, 73641.
Malgrem, K., Sullivan, M., Ekstedt, G., Kullberg, G. and Kumlien, E. (1997). Health-related
quality of life after epilepsy surgery: a Swedish multicentre study. Epilepsia, 38, 8308.
Mays, N. and Pope, C. (1996). Rigour and qualitative research. In Qualitative Research in Health
Care, ed. N. Mays and C. Pope, pp. 1019. London: BMJ Publishing Group.
McDowell, I. and Newell, C. (1987). Measuring Health: A Guide to Rating Scales and
Questionnaires. Oxford: Oxford University Press.
McGuire, A.M. (1991). Quality of life in women with epilepsy. In Women and Epilepsy, ed. M.R.
Trimble, pp. 1330. Chichester: John Wiley & Sons.
McLachlan, R.S., Rose, K.J., Derry, P.A., Bonnar, C., Blume, W.T. and Girvin, J.P. (1997). Health-
related quality of life and seizure control in temporal lobe epilepsy. Ann Neurol, 41, 4829.
OBoyle, C.A., McGee, H.M., Hickey, A. et al. (1993). The Schedule for the Evaluation of Individual
Quality of Life (SEIQoL). Administration Manual. Dublin: Royal College of Surgeons in Ireland.
ODonoghue, M. F., Duncan, J.S. and Sander, J.W.A.S. (1996). The National Hospital Seizure
Severity Scale: A further development of the Chalfont Seizure Severity Scale. Epilepsia, 37,
56371.
Rose, K.J., Derry, P.A. and McLachlan, R.S. (1996). Neuroticism in temporal lobe epilepsy: assess-
ment and implications for pre- and postoperative psychosocial adjustment and health-related
quality of life. Epilepsia, 37, 48491.
Ruta, D.A., Garratt, A.M., Leng, M., Russell, I.T. and MacDonald, L.M. (1994). A new approach
to the measurement of quality of life: the patient-generated index (PGI). Med Care, 32,
110926.
Selai, C.E. and Trimble, M.R. (1998). Adjunctive therapy in epilepsy with new antiepileptic drugs:
is it of any value? Seizure, 7, 41718.
Selai, C.E., Elstner, K. and Trimble, M.R. (2000). Quality of life pre- and post epilepsy surgery.
Selai, C.E., Trimble, M.R., Rossor, M. and Harvey, R.J. (2001). Assessing quality of life (QOL) in
dementia: the feasibility and validity of the Quality of Life Assessment Schedule (QOLAS). J
Neuropsychol Rehab, 11, 21943.
Smith, D., Baker, G.A., Jacoby, A. and Chadwick, D.W. (1995). The contribution of the measure-
ment of seizure severity to quality of life research. Qual Life Res, 4, 14358.
Stavem, K. (1998). Quality of life in epilepsy: comparison of four preference measures. Epilepsy
Res, 29, 2019.
Trimble, M.R. and Dodson, W.E. (ed.) (1994). Epilepsy and Quality of Life. New York: Raven
Press.
Tugwell, P., Bombardier, C., Buchanan, W.W. et al. (1990). Methotrexate in rheumatoid arthri-
tis: impact on quality of life assessed by traditional standard item and individualized patient
preference health status questionnaires. Arch Intern Med, 150, 5982.
Vickrey, B.G., Hays, R.D., Graber, J., Rausch, R., Engel, J. and Brook, R.H. (1992). A health-
related quality of life instrument for patients evaluated for epilepsy surgery. Med Care, 30,
299319.
Vickrey, B.G., Hays, R.D., Engel, J. et al. (1995). Outcome assessment for epilepsy surgery: the
impact of measuring health-related quality of life. Ann Neurology, 37, 15866.
Walker, M.C. and Sander, J.W.A.S. (1996). The impact of new anti-epileptic drugs on the prog-
nosis of epilepsy: seizure freedom should be the ultimate goal. Neurology, 46, 91214.
Appendix 22.1. QOLAS qualitative data. The
three most frequently nominated items per
QOLAS domain
QOLAS domain Number of times nominating
Physical
Seizure-related (injury, incontinence) 153
Tiredness 34
Drug side eects 21
Psychological
Depression 65
Anger (why me..?) 57
Anxiety 42
Social
Inability to drive 59
Restricted social life/leisure 66
Family 38
Work/economic
Unable to work/keep job 49
Interference with career/promotion 63
Discrimination (job application) 28
Cognitive
Memory 91
Concentration impaired 45
Thinking (diculties with) 33
341
Index
Numbers in italics indicate tables and gures. place in the limbic system 1920
seizures originating in 2345
abreactive attacks 198 amygdalohippocampectomy 23
absence seizures amygdalotomy, bilateral 84
frequent 724 animals, aggression subtypes 82
risk of psychosis 42, 43 anterior commissure 27
see also childhood absence epilepsy (CAE) Anthropometric Laboratory 137
absence status 44 anticonvulsant drugs see antiepileptic drugs
action-slips 206 (AEDs)
aective disorders in epilepsy antidepressant drugs 97
anatomic substrates 289 interactions with AEDs 305
and temporal lobe surgery 2724 newer drugs 3024
see also depression in epilepsy; dysphoric and seizures
disorders; mania early animal data 299300
agoraphobia 228, 317 early clinical data 3001
aggression 81 newer drugs 3045
classication 812 suicide prevention 11213
clinical relevance 82, 83 see also vagus nerve stimulation (VNS):
in epilepsy antidepressant eects
anatomic substrates 301 antiepileptic drugs (AEDs)
diagnosis 96, 97 antiglutaminergic 249
postictal psychosis 1267 cognitive side eects 67, 1423, 2567
prevalence 845 combined eects with seizures 257
social and psychological aspects 956 dierent drug formulations 2589
treatment 968, 99 habituation 257
types of behaviours 856 meta-analysis of studies 25960
neurobiology 834 patient complaints 2578
see also intermittent explosive disorder (IED) related to serum levels 258
alcohol consumption GABAergic 2489
men with panic disorder 228 interactions with psychotropic drugs 305,
precipitation of JME 48 309
alprazolam 229 positive psychotropic eects 243
alternative psychoses 45, 47, 24950, 269 epilepsy patients 2501
amitriptyline 300 psychiatric patients 250
amnesia psychiatric side eects 7, 241, 243
dissociative 194, 200, 202 classical drugs 2424
ictal 194, 200 classication/terminology 242
amnesic barriers 202, 204 clinical recommendations 252
amygdala 234 mechanisms 24850
in anxiety 31, 33, 234 new drugs 2448
direct thalamic input 235 and preexisting mental state 249
false threat alarms 2356 questions regarding 242
in fear-induced aggression 84 quality of life study 32631
organization 24, 25 in suicide attempts 108
pathology in TLE with IED 8790, 91 see also specic drugs
343
344 Index
antipsychotic drugs pharmacokinetic interactions 305, 309

atypical 307 psychotropic eects 243, 244, 250, 251
case report (clozapine in epilepsy) 3089 cats, defensive rage 301, 32
classical 306 character neurosis in JME 54
classication 306 Charcot, Jean-Martin see hysteria: Charcots work
interactions with other drugs 309 childhood absence epilepsy (CAE) 42
newer drugs 3067, 308 attention decits 523
receptor binding proles 307 frequency of psychosis 42
and seizures 3078 personality traits 52
antisocial personality disorder (APD) 82 see also absence seizures: frequent
anxiety chlorpromazine 306
in epilepsy 10 cingulate gyrus 27, 166
aggression 95, 97 citalopram 302, 303, 304
temporal lobe surgery 274 classication of psychiatric disorders in epilepsy
experimental models 234 11, 1315
role of the amygdala 31, 33, 234 clobazam 310
susceptibility to VNS 290, 291 clomipramine 300, 301
see also antiepileptic drugs (AEDs): psychiatric clonazepam 229
side eects; panic disorder (PD) clozapine 307, 308, 30910
area tempesta 19 cognitive decline in refractory TLE 152, 153
attention decit hyperactivity disorder 243 duration eects 1534
attention decits in FLE 169 cross-sectional studies 1547
auras 125, 195 exhaustion of brain functional reserve 159
autistic features 78 education and 156, 157, 158, 15960
awakening epilepsy see idiopathic generalized measurement 154, 155
epilepsy (IGE) prefrontal metabolic disturbances 152
proles in males and females 143, 144
senile plaque formation 144
BearFedio scale 11, 13 temporal gate hypothesis 1445
behavioural disturbances in epilepsy 1668 cognitive eects of epilepsy
anatomic substrates 1819 eects of seizures see seizures: cognitive eects
aective disorders 289 factors for consideration 1678
aggression 301 frequent localized discharges 756
anxiety 31, 33 LandauKlener syndrome 778, 140
schizophrenia 2930 postictal state-dependent impairment 767
assessment framework 701 prevention 789
epidemiological studies 70 psychosocial 143
frequent absence seizures 724 and quality of life 263
and learning disorders 634 subtle, dened 71
prevention 789 TLE see cognitive decline in refractory TLE
subtle 71 transitory cognitive impairment 745, 1412
see also aggression; antiepileptic drugs (AEDs): see also antiepileptic drugs (AEDs): cognitive
psychiatric side eects; specic types of side eects; dementia and epilepsy; learning
epilepsy disorders in epilepsy
benzodiazepines 229 cognitive function 62
and aggression 96, 98 complex partial seizures
for postictal psychoses 128 attention tests on patients 53
psychotropic eects 243 and depression 29
and seizures 310 continuous spike-wave discharges in slow wave
beta-carboline 229 sleep (CSWS) 140
blindsight 191 corpus callosum 27
brain functional reserve 159 cortical inhibition and VNS 289
Briquet, Paul 210, 21112
Broca, Paul 138 Damasios somatic marker hypothesis 166
defensive aggression 81, 82, 84
CAE see childhood absence epilepsy defensive rage (cats) 301, 32
Calmans gap 326 degeneracy theory 136
carbamazepine dj vu 125, 195
and clozapine 310 dementia, denitions 135
cognitive eects 143, 258, 259 dementia and epilepsy
and lithium 310 historical links 1357
345 Index
possible relationships 139 population-based studies 69, 12

cognitive eects of EEG discharges 1412 prevalence 1011
cognitive eects of seizures 141 psychiatric disorders 56
neurological disease with dementia and epilepsy on awakening see idiopathic generalized
seizures 13940 epilepsy (IGE)
specic epilepsy syndromes 1401 epilepsy surgery inventory (ESI-55) 332
treatment eects see antiepileptic drugs epileptiform discharges
(AEDs): cognitive side eects cognitive eects 745, 1412
see also cognitive eects of epilepsy early treatment 789
dentate gyrus 21, 22 frequent
depersonalization 193, 1956, 231 frontal 71
depersonalization disorder 193 hemispheric 76
depression left temporal 71
in epilepsy 10 localized 756
aggression 95, 97 episodic dyscontrol see intermittent explosive
anatomic substrates 289 disorder (IED)
postoperative 127, 2724 EQ-5D 3278, 3367
suicide risk 110 ESES see electrical status epilepticus of slow wave
TLE and FLE 176, 178, 1801 sleep
GABA hypothesis 2489 ESI-55 (epilepsy surgery inventory) 332
in panic disorder 228 ethosuximide 46, 243, 244
see also antiepileptic drugs (AEDs): psychiatric EuroQuol EQ-5D 3278, 3367
side-eects; dysphoric disorders event-related potentials 2323
derealization 193, 1956, 231
desipramine 300 fear see anxiety; panic disorder
discharges see epileptiform discharges felbamate 243, 246
dissociated control theory 2046 FLE see frontal lobe epilepsy (FLE)
dissociation umazenil 229
classication of disorders 1901, 190 uoxetine 302, 303, 305
confusion over concept 189, 196 uphenazine 308
and epilepsy 1947 uvoxamine 304
facets 18990 folate supplementation 142
alteration in consciousness 193 forced normalization 457, 111, 24950
defence mechanism 1934 frontal lobe
nonintegrated mental modules 1903 and aggression 83, 84, 166
and nonepileptic seizures see nonepileptic assessment test design 170, 171
seizures: dissociative mechanisms and behaviour disorders 1656
dissociative amnesia 194, 200, 202 cognitive consequences of surgery 16970
dissociative fugue 2023 structure 165
dissociative identity disorder 203 frontal lobe epilepsy (FLE) 164
dopaminergic hypothesis of schizophrenia behaviour and personality 175, 176
2930 clinical personality inventory 1779
dorso-lateral cortex 165 education and employment 17980
Downs syndrome 139 quality of life 175, 177, 1801
doxepin 300, 301 state/trait 1801
dysphoric disorders 13, 11011, 112, 273 neuropsychology 16870
postictal reorientation 174
educational performance seizure analysis
childhood absence epilepsy 52 negative phenomena 171, 1723
epidemiological studies 70 nonconvulsive status epilepticus 173, 174
FLE and mTLE 17980 positive phenomena 171, 172
electrical status epilepticus of slow wave sleep fugue
(ESES) 778 dissociative 2023
electroconvulsive therapy (ECT) 28, 288, 289 postictal 194, 195
entorhinal cortex 20, 212, 23 Full Scale Intelligence Quotient (FSIQ) 156
epidemiology
epilepsy 5 gabapentin 243, 247, 250, 251
learning diculties 70 Galton, Sir Francis 137
psychiatric comorbidity in epilepsy 6 gamma aminobutyric acid (GABA)
AED-related 241 in depression 2489
hospital-based studies 910 in panic disorder 2289
346 Index
gamma aminobutyric acid (GABA) (cont.) implicit perception 192

regulation of seizures 28990 inhibitory interneurons 24, 26
generalized tonic-clonic seizures (GTCS; grand intelligence 62
mal) 42, 434 intelligence quotient see IQ
glutamate 145, 28990 intelligence trace tests 154
interictal aggression 856, 98
hallucinations 45, 11011, 195 interictal dysphoric disorder see dysphoric
haloperidol 306, 307, 309 disorders
head trauma 141, 216, 217 interictal psychoses 13, 45
health-related quality of life (HRQOL) see quality associated seizures 125
of life (QOL) chronic, and temporal lobe surgery 2701
hemispherectomy 76 demographic data 1223
hemispheric specialization 28, 95 in dierent types of epilepsy 124
Hilgards neodissociation theory 2014 in dysphoric disorder 11011
hippocampal commissure 278 pathogenesis 11112
hippocampal formation 21, 25 psychopathological features 126
hippocampus treatment 98
place in the limbic system 19, 20 intermittent explosive disorder (IED) 83, 86
in schizophrenia 30 studies in TLE
sclerosis 234, 25, 26, 867, 159 aims and rationale 867
structure 213, 25 amygdala pathology 8790, 91
trisynaptic circuit 21, 223 cortical abnormalities 902, 93, 94
hyperarousal syndrome 934 demographic data 889
hypersynchronization 24, 26 dual brain pathology 92, 93, 945
hypnotic phenomena 192 neuropsychological and psychometric
dissociated control theory 204 parameters 92
neodissociation theory 2012, 204 treatment 98
hypothalamic hamartomas 139 IQ
hypothalamus 83 and aggression 86, 90
hysteria early studies in epilepsy 1378
Briquets work 210, 21112 eects of duration of TLE 1548
Charcots work 211 interictal vs. postictal psychoses 1223
classical clinical ndings 215 interpretation of serial scores 1389
clinicalanatomical method 215
descriptions of seizures 212 Janet, Pierre 190, 2001
importance of trauma 211, 212, 216 juvenile myoclonic epilepsy (JME) 42
male hysteria 216 frequency of psychoses 42, 43
Janets work 2001 frontal impairment 556
nonepileptic seizures 21214 personality traits 545
Richers drawings 212, 213, 214 precipitating factors 48
role of the uterus 211 sleep deprivation 48, 4950
see also nonepileptic seizures; psychogenic
pseudoseizures kainic acid 31
ictal aggression 85, 98 lamotrigine

ictal amnesia 194, 200 cognitive eects 143, 145
ictal fear 125, 235 interaction with sertraline 305
ictal psychoses 44 psychiatric side eects 7, 246
idiopathic generalized epilepsy (IGE) 412 psychotropic eects 243, 246, 250, 251
classication of syndromes 42 LandauKlener syndrome 65, 778, 140
precipitating factors 478 learning disorders in epilepsy
lifestyle 489 causes 645
sleep deprivation 48, 4950 eects of subtle seizures 66
psychoses 47 links with behavioural disorders 634
frequency 42, 43 permanent 65, 66
relation to seizure type 42, 434 prevalence 64, 70
types 447 state-dependent 656
see also childhood absence epilepsy (CAE); terminology 623
juvenile myoclonic epilepsy (JME) treatment 67
IED see intermittent explosive disorder (IED) see also cognitive eects of epilepsy
imipramine 300, 301, 303 left-handedness 92, 221, 223
347 Index
LennoxGastaut syndrome (LGS) 140 norepinephrine 289

levetiracetam 248 nortryptiline 300
limbic system 1920
amygdala see amygdala olanzapine 98, 307, 309
hippocampus see hippocampus opioids 22, 28
interhemispheric connections 278 orbito-frontal cortex 165, 166
lithium 310 oxidative stress 145
locus coeruleus activation by VNS 289
P3a and P3b 232
magnetic resonance imaging (MRI) studies 86, 87, panic disorder (PD) 226
88, 2334 aetiology 227
mania cerebral blood ow 234
postoperative 127, 273, 274 diagnosis 227
valproate treatment 250 EEG studies 22930
maprotiline 300 atypical panic attacks 230
mental retardation derealization/depersonalization symptoms
aggression 82 231
dened 63 event-related potentials 2323
dementia 139 epidemiology 228
in epilepsy 63, 64, 70 experimental models 234
mesial temporal lobe epilepsies (mTLE) 164 false-alarm theory 2356
mesial temporal sclerosis 867 GABA system implicated 2289
mianserin 304 MRI 2334
mirtazapine 3034 panic attacks 227, 230, 290
mismatch negativity (MMN) 232, 233 suicide risk 228
Monolog 723 Papez circuit 19, 20
mossy bres 21, 22, 24, 25 paroxetine 303, 304, 305
MRI see magnetic resonance imaging (MRI) PD see panic disorder
studies Pearson, Karl 137
multiple subpial transection 78 periaqueductal grey 30, 834
permanent cognitive impairment 712
narcolepsy (sleep epilepsy) 48, 49 personality disorders 276
NBI (Neurobehavioural Inventory) 13 case report 2789
NEAD see nonepileptic attack disorder development 2778
nefazodone 303, 304 and temporal lobe surgery 276, 277
neodissociation theory 2014 personality traits 7, 10
Neurobehavioural Inventory (NBI) 13 clinical inventory 1779
neuropsychology, origins of 138 FLE 175, 176, 1779
neurosyphilis 137 IGE 501
nomifensine 300 childhood absence epilepsy 523
nonconvulsive status epilepticus 44, 65, 72, 173, juvenile myoclonic epilepsy 546
174 TLE 11, 13, 175, 176, 1789
nonepileptic attack disorder (NEAD) 14, 197 see also behavioural disturbances in epilepsy
postoperative 275 petit mal see absence seizures
nonepileptic seizures 189, 197 phenobarbital
dissociative mechanism (evidence) behavioural eects 96, 97
amnesia 200 cognitive eects 142, 259
dissociative comorbidity 198 and depression 242
hypnotic susceptibility 199 and suicide 108
traumatic experiences 199 phenytoin
dissociative mechanisms cognitive eects 1423, 258, 259
dissociated control theory 2046 interactions with tricyclic drugs 305
Janets analysis of hysteria 2001 psychiatric complications 2434
neodissociation theory 2014 pimozide 3078
presentations 1978 positron emission tomography (PET) 55, 234
and temporal lobe surgery 2756 postictal aggression 85, 98
terminology 197 postictal fugue 194, 195
see also psychogenic pseudoseizures postictal psychoses 445
nontemporal lobe epilepsies (non-TLE), frequency aggression 85, 1267
of psychosis 42 case reports 11820
noradrenergic system and VNS 289 clinical studies 120
348 Index
postictal psychoses (cont.) psychometry, origins of 1378

clinical studies (cont.) psychoses in epilepsy
association with TLE 1234 AED-related 241, 243
demographic data 1223 see also specic antiepileptic drugs
duration 122 epidemiology 8, 910, 11, 1201
incidence 1201, 270 forced normalization 457, 111, 24950
lucid intervals 1212 ictal 44
psychic auras 125 interictal see interictal psychoses
psychopathological features 126 in men 9
recurrences 122 postictal see postictal psychoses
seizures 125 postoperative 127, 2712
historical background 11718 relation to seizure type 42, 434
pathophysiological mechanism 128 see also psychiatric disorders and temporal lobe
prevention 128, 310 surgery: psychoses; schizophrenia
and temporal lobe surgery 127, 26970 psychotherapy in epilepsy
treatment 98, 1278 case reports
postoperative psychoses 127, 2712 agoraphobic avoidance 317
posttraumatic stress disorder (PTSD) 320 agitated depression 31517
predatory aggression 81, 82 depressive withdrawal 31819
prefrontal cortex 29, 1656 fear of psychotic disintegration 31718
prepsychotic dysphoria 45 clinicians views of psychotherapists 314
primidone 242, 243 contraindications 3201
procaine 234 follow-up study 3212
protriptyline 300 handicaps to integration 31415
pseudoseizures see psychogenic pseudoseizures indications 320
psychiatric comorbidity in epilepsy a neglected area 31314
classication of psychiatric disorders 11, 1315 therapists roles 31920
epidemiology see epidemiology: psychiatric time frames 319
comorbidity in epilepsy psychotropic drugs
pathogenesis 11112 antidepressant see antidepressant drugs
temporal lobe surgery candidates 2689 antipsychotic see antipsychotic drugs
see also specic disorders classication 300
psychiatric disorders and temporal lobe surgery other agents 310
2678 see also specic drugs
aective disorders 127, 2724 pyknolepsy see childhood absence epilepsy (CAE)
anxiety 274
case report 2789 Quality of Life Assessment Schedule (QOLAS)
evaluation of surgical outcome 2667 3267, 336
limitations of presented data 268 quality of life (QOL)
nonepileptic attacks 2756 assessment
personality disorders 276, 277 epilepsy surgery inventory (ESI-55) 332
psychiatric assessment strategies 279 EuroQuol EQ-5D instrument 3278, 3367
psychoses 266, 269 individualized techniques 3245
chronic interictal 2701 QOLAS 3267, 336
postictal 26970 qualitative vs. quantitative methods 324
postoperative 127, 2712 reasons for 323
total psychiatric comorbidity 2689 types of measure 3234
psychiatric eects of anticonvulsants see and cognitive function in epilepsy 263
antiepileptic drugs (AEDs): psychiatric side FLE 175, 177, 1801
eects and seizure frequency 325, 336
psychic auras 125, 195 drug study 32631, 341
psychogenic pseudoseizures 210 surgery study 3315
emotional trauma 222 TLE 175, 177, 1801
evidence for biological factors 220, 221, 223 quetiapine 98, 307, 309
historical review see hysteria
intellectual and neuropsychological test ndings Rasmussen syndrome 76, 140
21819, 2202 reboxetine 302, 303
left-handedness 221, 223 receptor proles
patient demographics 217 antidepressants 303
physical trauma 216, 217 antipsychotics 307
see also nonepileptic seizures risperidone 98, 307, 309
349 Index
Schaer collaterals 21, 22 prevention 11213

schema 201 risk factors 108, 114
schizophrenia in epilepsy seizure control and 11112, 114
anatomic substrates 2930 suicide and panic attacks 228
epidemiology 8, 910 sulpiride 98
in men 9 surgery
school performance see educational performance bilateral amygdalotomy 84
seizures eects on learning 67
absence see absence seizures evaluation of outcome 2667
and antipsychotic drugs 3078 frontal lobe, cognitive consequences 16970
and benzodiazepines 310 hemispherectomy 76
cognitive eects 141 multiple subpial transection 78
acute/short-term 2601 quality of life study 3315
frequent absence seizures 724 see also psychiatric disorders and temporal lobe
impairment in FLE 172, 173 surgery
long-term 2612 swoon attacks 198
neuronal damage 65
nocturnal seizures 77 tantrum attacks 198
complex partial 29 TCI see transitory cognitive impairment (TCI)
FLE see frontal lobe epilepsy (FLE): seizure temporal lobe epilepsy (TLE) 18
analysis behaviour and personality
nonepileptic see hysteria; nonepileptic seizures; compared with FLE 17581
psychogenic pseudoseizures traits 11, 13
prediction 167 cognitive decline see cognitive decline in
subjective experience of 226 refractory TLE
subtle 71 ictal fear 235
type related to risk of psychosis 42, 434 memory decits 152
see also antidepressant drugs: and seizures; mesial (mTLE) 164
quality of life (QOL): and seizure frequency postictal memory impairment 174
selective serotonin reuptake inhibitors (SSRIs) psychoses
302 forced normalization 456
and seizures 3045 frequency 43, 45
see also specic drugs interictal 45, 124
senile plaques 144 postictal 121, 1234, 127
serotonergic system and VNS 289 quality of life 175, 177
sertraline 302, 303, 3045 senile plaques 144
sleep deprivation and IGE 4750 see also intermittent explosive disorder (IED):
sleep epilepsy 48, 49 studies in TLE; suicide in epilepsy
Sneddon syndrome 139 temporal lobe surgery, psychiatric disorders see
sodium valproate 78 psychiatric disorders and temporal lobe
somatic marker hypothesis 166 surgery
somatoform seizures see nonepileptic seizures; thiamine deciency 142
psychogenic pseudoseizures tiagabine
somnambulism 201 cognitive eects 143
spike-wave monitoring 723 psychiatric side eects 247
SSRIs see selective serotonin reuptake inhibitors psychotropic eects 243, 250
(SSRIs) TLE see temporal lobe epilepsy (TLE)
state-dependent cognitive impairment 712, 767 topiramate
state-dependent learning disorders 656 cognitive impairment 260
statistical parametric mapping 90 psychiatric side eects 2478
status epilepticus 64, 72 psychotropic eects 243, 250
ESES 778 transitory cognitive impairment (TCI) 656, 745,
nonconvulsive 44, 65, 72, 173, 174 1412
subtle seizures 71 trauma
suicide in epilepsy 107 dissociation response 1934, 199
attempted 1078 emotional 212, 222
dysphoric disorder 11011, 112 head 141, 216, 217
interictal and peri-ictal psychopathology 10911 and pseudoseizures/hysteria 21012, 21617,
neuropsychiatric data 109 222
overdose 108 trimipramine 300
postoperative 127 twin studies, schizophrenia 30
350 Index
vagus nerve stimulation (VNS) 2834 venlafaxine 302, 303

antidepressant eects vigabatrin
epilepsy patients 2846 cognitive eects 67, 143
nonepileptic patients 2878 psychiatric side eects 7, 243, 2446
anxiety disorders 290, 291 viloxazine 300, 301
epilepsy treatment 2834 violence
mechanisms of action dened 81
cerebral 28890 in postictal psychosis 1267
peripheral 2901 see also aggression
procedure 283 VNS see vagus nerve stimulation (VNS)
valproate vocabulary intelligence test (MWT-B) 154
cognitive eects 143, 25960
encephalopathy 244 Wernicke, Karl 138
psychotropic eects 243, 250

The Neuropsychiatry of Epilepsy

Hochgeladen von

Dokumentinformationen

Originalbeschreibung:

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

The Neuropsychiatry of Epilepsy

Hochgeladen von

Copyright:

Verfügbare Formate

The Neuropsychiatry of Epilepsy

Michael Trimble is Professor of Behavioural Neurology at the Institute of Neurology in

Cambridge University Press

Cambridge University Press 2002

This book is in copyright. Subject to statutory exception and to the provision of

First published in print format 2002

Cambridge University Press has no responsibility for the persistence or accuracy of

List of contributors page viii

2 Neuropsychiatric disorders in epilepsy epidemiology and classication 5

3 Limbic connectivity: anatomical substrates of behavioural disturbances

Part II Clinical aspects

4 The psychiatry of idiopathic generalized epilepsy 41

5 Epilepsy and learning disorders 62

6 Subtle cognitive and behavioural eects of epilepsy 70

7 Aggression and epilepsy 81

8 Epilepsy and suicide: a neuropsychiatric analysis 107

9 Postictal psychoses, revisited 117

Part III Cognitive aspects

10 Dementia and epilepsy 135

11 The risk of cognitive decline in patients with refractory temporal

12 Behavioural and neuropsychological aspects of frontal lobe epilepsy 164

Part IV Nonepileptic attacks

13 Epilepsy, dissociation and nonepileptic seizures 189

14 Psychobiology of psychogenic pseudoseizures 210

15 Epilepsy and panic disorder 226

Part V Treatment complications

16 The eects of antiepileptic drugs on behaviour 241

17 Antiepileptic drug treatment and epileptic seizures eects on

18 Psychiatric eects of surgery for temporal lobe epilepsy 266

19 Vagus nerve stimulation and mood 283

20 On the use of psychotropic drugs in patients with seizure disorder 299

21 The role of psychotherapy in the treatment of epilepsies 313

22 Choosing measures to assess quality of life (QOL) in epilepsy 323

Albert P. Aldenkamp Richard J. Brown

Jerome Engel Dieter Janz

F. Lopez-Rodriquez Caroline E. Selai

In 1997, Dr Pete Engel, President of the International League Against Epilepsy,

Neuropsychiatric disorders in epilepsy

Psychiatric Association, 1994), and the mental disorders component of the

Population-based studies of psychiatric comorbidity in epilepsy

health professional, and a lack of standardized techniques to assess patients with

Jalava and Sillanpaa (1996) examined a prospective population-based cohort

Some large hospital-based studies

data supported a distinct association of schizophrenic disorders with epilepsy, par-

Are psychiatric disorders commoner in epilepsy?

Of all the dierent psychiatric disorders in epilepsy, it is psychosis for which

The classification of psychiatric disorders in epilepsy

The classication of psychiatric disorders in epilepsy has always been controversial.

Table 2.1. Important epidemiological studies of neuropsychiatric comorbidity in epilepsy

Psychiatric disorders are common in epilepsy, and encompass the spectrum of

Limbic connectivity: anatomical substrates of

Behavioural disturbances associated with epilepsy are due, in part, to a number of

ring homeostatic seizure-suppressing mechanisms which maintain the interictal

The limbic system

Mid-brain tegm. CC VHC DHC AC

Regio Medial entorhinal

The amygdala has been implicated in a number of important primal functions,

Mossy fibre reinnervation

Interhemispheric limbic connections

autopsy specimens has revealed the presence of a dorsal hippocampal commissure,

Possible anatomic substrates of some behavioural disturbances associated

The behavioural disturbances most associated with epilepsy include depression