www.elsevier.com/locate/yebeh

Epilepsy & Behavior 7 (2005) 37–50

Review

Psychiatric comorbidity in epilepsy

W.A.M. Swinkels a,*, J. Kuyk a, R. van Dyck b, Ph. Spinhoven c

a Stichting Epilepsie Instellingen Nederland (SEIN), Heemstede, The Netherlandsb Department of Psychiatry, Vrije Universiteit Amsterdam, Amsterdam, The Netherlands

c Division of Clinical and Health Psychology, Department of Psychiatry, Leiden University, Leiden, The Netherlands

Received 25 April 2005; accepted 26 April 2005Available online 21 June 2005

Abstract

Many studies on psychiatric comorbidity in epilepsy have been performed using many different patient groups and diagnosticinstruments. This methodological heterogeneity complicates comparison of the findings. In this article, psychiatric disorders in epi-lepsy are reviewed from the perspective of the DSM classification system. The empirical findings of axis I clinical disorders and axisII personality disorders are described separately. Furthermore, the existence and specificity of conditions such as interictal dysphoricdisorder, interictal behavior syndrome, and psychosis of epilepsy are discussed. From the many studies that have been performed onthis topic it can be learned that there is a need for well-controlled studies using representative patient groups and valid and stan-dardized diagnostic instruments. So far, the majority of the studies have concerned axis I disorders; relatively little research has beenperformed on axis II personality disorders. More research on personality disorders, as well as on the relative contributions of thedifferent (brain- and non-brain-related) factors to the relationship between epilepsy and psychiatric disorders, is recommended.� 2005 Elsevier Inc. All rights reserved.

Keywords: Epilepsy; Psychiatry; Comorbidity; Review

1. Introduction

Epilepsy is one of the most common chronic neurolog-ical disorders. The prevalence of epilepsy varies acrossstudies, but generally ranges from 4 to 10 per 1000 popu-lation [1–5]. This range is partly related to varying defini-tions, but also to true differences in prevalence rates acrossthe populations. The impact of epilepsy on quality oflife can be considerable. Epilepsy may, but does not inev-itably, affect emotional, behavioral, social, and cognitivefunctioning. Studies investigating these problemsfrequently use the term psychopathology. Because of thediverse and confusing ways this term has been used,psychopathology can refer to psychiatric problems ofvarious types, maladaptive emotional disorders, psycho-social adjustment difficulties, and behavioral and person-

1525-5050/$ - see front matter � 2005 Elsevier Inc. All rights reserved.

doi:10.1016/j.yebeh.2005.04.012

* Corresponding author. Fax +31 0 23 5588 279.E-mail address: mswinkels@sein.nl (W.A.M. Swinkels).

ality characteristics. To prevent further confusion, in thisarticle a distinction has been made between axis I clinicaldisorders and axis II personality disorders according totheDiagnostic and StatisticalManual ofMental Disorders

(DSM-IV). Currently, DSM-IV is the internationalstandard with respect to classification of mental disor-ders. The empirical findings regarding the relationshipbetween epilepsy and axis I and II disorders arereviewed by focusing on measures that somehow repro-duce the prevalence rates of these problems. For thatreason, we do not describe the extensive literature inwhich the Minnesota Multiphasic Personality Inventory(MMPI) [6] was used as a diagnostic instrument, as theMMPI is a dimensional instrument and does not lead toa psychiatric diagnosis. Apart from the empirical find-ings, the different concepts and theories about supposedepilepsy-specific disorders are also discussed. But first,we give a short overview of some historical facts aboutthe relationship between epilepsy and psychiatry.

38 W.A.M. Swinkels et al. / Epilepsy & Behavior 7 (2005) 37–50

2. Some historical facts

The term epilepsy is derived from the Greek word epi-

lepsia, which means ‘‘falling sickness.’’ The ancientGreeks considered epilepsy as a sacred disease. A personwith epilepsy was seen as seized by a god. Only a godcould throw a sane normal man to the ground, deprivehim of his senses, cause him to convulse, and then re-store him to normality. In his book The Sacred Disease

written in 400 BC, Hippocrates was the first to suggestthat epilepsy, like other diseases, has a natural and nota sacred cause. He argued that epilepsy was by no meansdivine or sacred and ascribed epilepsy to a disorder ofthe brain. During the Middle Ages, epileptic phenomenawere explained by various mystical, magical, and reli-gious concepts (see, e.g., [7]). Epilepsy patients were seenas being possessed, which worsened the already existingfears, anxieties, and feelings of shame. Many of the psy-chological and social problems experienced by epilepsypatients today have their origins in these unfortunateconcepts (that still persist in some cultures). Increasinginsight into the medical background of epilepsy beganin the Renaissance, but has particularly developed sincethe mid-19th century. The neurologist Hughlings Jack-son developed the first comprehensive understandingof seizure origin. In 1873, he gave the first accurate def-inition of epilepsy: ‘‘Epilepsy is the name given for occa-sional, sudden, excessive, rapid and local discharges ofgray matter’’ [8]. He is often considered the founder ofmodern epileptology. His contemporary, Gowers, fur-ther detailed the excessive variety of clinical epilepticsymptoms, and also was engaged in the separation ofepileptic convulsions from nonepileptic (hysterical) con-vulsions [9]. The differentiation between epilepsy and so-called ‘‘hystero-epilepsy’’ was also the work of theFrench neurologist Jean Martin Charcot and his group.Through such French workers as Briquet [10] and Morel[11] it became recognized that psychological distur-bances may occur as part of the seizure itself (ictal) oras an interictal disturbance involving various behavioraland cognitive functions. Understanding of epilepsyprogressed in the 1930s with the introduction of pheno-barbital and, more importantly, the use of the electroen-cephalogram (EEG), developed by Hans Berger, inpatients with epilepsy. It was realized that the localiza-tion of epileptic discharges in the brain and their associ-ation with lesions was more important than anythingelse in determining the character of the seizures. Never-theless, the clinical impression continued to exist thatpatients with epilepsy suffer from mental problems.Discovery of the temporal lobe focus (and, thus, tempo-ral lobe epilepsy) in 1949, together with the concept ofthe limbic system as important in the processing ofemotions, led to the idea of mental disorders beinglinked to epileptic disturbances in specific brain areas.People with epilepsy were considered mentally normal,

but it was assumed that brain dysfunctions would leadto seizures as well as psychological complaints.

Currently, it is believed that not only biological fac-tors (e.g., etiology, focus localization), but also medica-tion (e.g., number and types of medication) andpsychological and social factors (e.g., fear of seizures,perceived stigma) are important aspects in the develop-ment of psychiatric problems in epilepsy. It is thereforerecommended that all these factors be considered wheninvestigating psychopathology in epilepsy.

3. Comorbid psychiatric disorders in epilepsy

Psychiatric symptoms can be classified according totheir temporal relationship with seizure occurrence.They can be divided into peri-ictal (related to the seizureitself) and interictal (independent of the seizure) symp-toms. Peri-ictal symptoms are symptoms that precedethe seizure (preictal), clinical manifestations of the sei-zure itself (ictal), and symptoms that follow the seizuredirectly (postictal). Because this article concerns psychi-atric comorbidity in epilepsy, we focus on the interictaldisorders.

Although there are many studies on this topic, accu-rate estimates of psychiatric comorbidity are hard tofind. To a large extent, this is attributable to methodo-logical differences among the studies, which make com-parison of the results complicated. Most important arethe differences in patient groups and diagnostic instru-ments used. Patient groups are often small, and fre-quently, no (appropriate) control group is included.Furthermore, the epilepsy group is often not representa-tive of the total epilepsy population because, for exam-ple, only inpatients or outpatients, or sometimes avery specific subgroup of epilepsy patients (e.g., surgerypatients), are studied. Also important is the use of avariety of diagnostic instruments: only a few studiesuse standardized diagnostic instruments based on specif-ic criteria such as DSM-IV. Frequently, no distinction ismade between axis I clinical disorders and axis II per-sonality disorders. In the next two sections the empiricalfindings for DSM-IV axis I clinical disorders and axis IIpersonality disorders are reviewed.

4. Clinical disorders

It is commonly believed that epilepsy places a patientat increased risk of developing psychiatric problems.Most of the studies on this topic are based on epilepsypatients admitted to hospitals or specialized epilepsyclinics. There are some population-based studies: how-ever, they constitute a minority [12–15]. Table 1 is anoverview of studies on psychiatric comorbidity in adultepilepsy patients [12–28]. The percentage of psychiatric

Table 1Epidemiological studies on psychiatric disorders in adult epilepsy patientsa

Investigators (country, year) N Instruments Results Comments

Pond and Bidwell(UK, 1960) [12]

245 in- and outpatients Interview by psychiatricsocial worker

29% had significant psychologicaldisorders

No standardized instruments;no control group

Currie et al.(UK, 1971) [16]

666 hospital outpatients Interview and case note review Psychiatric aspects in 44% (especiallyanxious 19%, depressed 11%)

No control group; nostandardized instrument

Kogeorgos et al.(UK, 1982) [17]

66 epilepsy outpatients,50 controls

General Health Questionnaire(GHQ), Crown-Crisp ExperientialIndex (CCEI)

Psychiatric morbidity was found in45.5% of the epilepsy group, comparedwith 28% of the neurological controls(n.s.) and 21.6% of the communitysample (sign)

Control group consisted ofneurological outpatients;data were also comparedwith population norms

Schiffer and Babigian(USA, 1984) [18]

402 TLE, 368 MS, and124 ALS inpatients

Psychiatric contact rates (also clinicalpsychiatric diagnosis according toDSM-I or -II)

Prevalence rate of psychiatric contact was22.9% for TLE, 19.3% for MS, and4.8% for ALS

Retrospective study; inpatients;no standardized instruments

Edeh and Toone(UK, 1987) [13]

88 outpatients Clinical Interview Schedule (CIS) 48% had psychiatric problems Solid epilepsy diagnoses (both CTand EEG); valid instrument; nocontrol group

Gureje (Nigeria, 1991) [19] 204 outpatients Clinical Interview Schedule (CIS) 37% had psychiatric problems No control groupManchanda et al.(Canada, 1992) [20]

71 epilepsy inpatients General Health Questionnaire (GHQ) Psychiatric disorders in 45% of the patients Sample not representative; nostandardized instrument; nocontrol group

Fiordelli et al.(Italy, 1993) [21]

100 epilepsy outpatients,100 controls

Clinical Interview Schedule (CIS) Psychiatric disorders in 19% of epilepsypatients and 15% of controls (n.s.)

Control persons were healthyindividuals visiting the samehospital for minor reasons

Victoroff (USA, 1994) [22] 60 candidates for epilepsysurgery (all CPS)

Structured Clinical Interview forDSM-III-R patient version (SCID-P)

Psychiatric disorders (past or current) in 70%of patients; most prevalent are depressivedisorders (58.3%)

No control group; separatediagnoses on axis I and axis II

Silberman et al.(USA, 1994) [23]

21 epilepsy outpatients Schedule for Affective Disordersand Schizophrenia (SADS)

Lifetime psychiatric diagnosis in 71% of patients DSM-III-R version; small sample;no control group

Jalava and Sillanpaa(Finland, 1996) [14]

94 epilepsy patients Psychiatric diagnoses were takenfrom files or made by a psychiatristaccording to ICD8

Psychiatric disorders were found in 23% ofepilepsy patients, compared with 7% ofcontrols (significant)

Control groups were used;prospective cohort study (35 yearsfollow-up)

Manchanda et al.(Canada, 1996) [24]

300 candidates forepilepsy surgery

Present State Examination (PSE) 29.3% had a DSM-III-R axis I diagnosis No control group; separatediagnoses on axis I and axis II

Perini et al (Italy, 1996) [25] 20 TLE, 18 JME, 20 type Idiabetes, 20 healthy controls

Schedule for Affective Disorders andSchizophrenia (SADS), Beck DepressionInventory (BDI), state and trait anxietyscales (STAIX1 and STAIX2)

Psychiatric diagnosis in 80% of TLE, 22% ofJME, 10% of diabetes patients; mood disorderswere most prevalent (55% in TLE, 17% JME,10% diabetes)

Patient groups were matched(outpatients)

Arnold and Privitera(USA, 1996) [26]

27 epilepsy inpatients,14 PNES patients

Structured Clinical Interview forDSM-III-R Epilepsy Version (SCID)

Axis I current psychiatric disorders in 30% ofepilepsy and 43% of PNES patients (n.s.);lifetime diagnoses: 51% for epilepsy, 71%for PNES patients (n.s.)

Small sample size; no normalcontrol group

Stefansson et al.(Iceland, 1998) [15]

241 epilepsy patients,482 controls

Psychiatric classification madeby psychiatrist according to ICD-9

Psychiatric diagnosis in 35% of epilepsy patientsand 30% of controls (n.s.)

Control group was matched andconsisted of patients with othersomatic diseases; no standardizedinstrument

(continued on next page)

W.A.M

.Swinkels

etal./Epilep

sy&

Behavio

r7(2005)37–50

39

Table

1(continued)

Investigators

(country,

year)

NInstruments

Results

Comments

Ettingeret

al.

(USA,19

98)[27]

37epilepsy

outpatients,

38patients

withCLD

Brief

Sym

ptom

Inventory

(BSI)

andCenter

forEpidem

iologicStudies—

Depression

Scale

(CES-D

)

MeanBSIan

dCES-D

scoresmarkedly

elevated

inboth

epilepsy

andCLD

groupscompared

withnorm

alcontrols,butless

than

those

of

psychiatric

outpatients;nodifference

between

epilepsy

andCLD

fortotalscores

Resultswerecompared

with

dataofnorm

alcontrolsan

dpsychiatric

outpatients

Swinkelset

al.

(TheNetherlands,20

01)[28]

209epilepsy

inpatients

Composite

International

Diagn

ostic

Interview

(CID

I)Anxietydisordersin

24.9%

(12.4%

inpopulation)

andmooddisordersin

18.7%

ofpatients

(7.6%

inpopulation)

Inpatients;datawere

compared

withgeneral

population(n

=70

76)

aTLE,temporallobeepilepsy;CLD,chronic

Lym

edisease;MS,multiple

sclerosis;ALS,am

yotrophic

lateralsclerosis;CPS,complexpartial

seizures;PNES,psychogenic

nonepilepticseizures.

40 W.A.M. Swinkels et al. / Epilepsy & Behavior 7 (2005) 37–50

cases varies considerably (range between 19 and 80%).As can be observed in a substantial number of studies,no control group is included. The majority of studieswith a control group report increased psychiatric prob-lems in patients with epilepsy compared with normalcontrols.

An important question is: Is the epilepsy conditionitself predisposing these patients to psychiatric com-plaints or does the chronicity of epilepsy lead to psy-chiatric disorders? By chronicity of epilepsy, we meanthe psychosocial impact of (long-term) epilepsy, notthe duration of the condition. With this question inmind, Wells and colleagues [29] compared data onthe prevalence of psychiatric disorders in eight chronicmedical conditions in a community sample of 2554 per-sons. Psychiatric disorders were obtained with theDiagnostic Interview Schedule (DIS) for patients withchronic lung disease, diabetes mellitus, heart disease,hypertension, arthritis, physical handicap, cancer andneurological disorders (e.g., stroke). The results werecompared with those for a control group of personswith no medical condition. The prevalence of recentand lifetime psychiatric disorders was higher amongpersons with any chronic medical condition than inpersons without any of these conditions. These findingssuggest that it is reasonable to assume that chronic ill-ness is generally associated with an increased risk forpsychiatric disorders.

Several of the studies listed in Table 1 included, be-sides epilepsy patients, a control group of patients withanother chronic illness [15,17,18,25,27]. Most of thesestudies did not report increased psychiatric comorbidityin patients with epilepsy relative to patients with othermedical conditions and are in agreement with theassumption that the chronicity of the illness is, in greatpart, responsible for the increased risk of psychiatricdisturbances.

Besides being a chronic disease with its implied psy-chosocial difficulties, epilepsy is also a neurological dis-order, probably adding to the risk for developingpsychiatric disorders. Schiffer and Babigian [18] foundno difference in psychiatric disorders between TLE pa-tients and patients with multiple sclerosis (MS), exceptfor a significantly increased rate of depressed affect inMS patients. The prevalence rates for both groups werehigher than that for amyotrophic lateral sclerosis (ALS)patients. Both epilepsy and MS can be seen as unpre-dictable and chronic neurological conditions, so thisdoes not explain the higher rate of depression amongthe MS patients. Several earlier studies also found moredepressive symptoms in patients with MS comparedwith other neurological patients [30–33]. Possible expla-nations for this high rate of depressive diagnoses in MSare the structural involvement of the limbic system (bydemyelination), shared genetic vulnerabilities both todepression and to MS, and dysfunction of monoamine

W.A.M. Swinkels et al. / Epilepsy & Behavior 7 (2005) 37–50 41

metabolism within the central nervous system [18]. Cur-rently, it is presumed that MS-associated depression canalso be attributed to pathophysiological processes suchas cerebral inflammation (see e.g., [34]).

The role of brain dysfunction in the etiology of psy-chiatric disturbances is also supported by a study ofPerini et al. [25]. They reported a significantly higher fre-quency of psychiatric diagnoses in patients with TLE(80%), than in patients with juvenile myoclonic epilepsy(JME) (22%) and diabetic patients (10%). The authorsconclude that TLE is a neurological condition that isassociated with a high prevalence of interictal psychiat-ric disturbances, which more likely reflect limbic dys-function rather than psychological adjustment to achronic epileptic or medical condition.

It can be concluded that the chronicity of epilepsy isan important factor in the predisposition of these pa-tients to psychiatric disturbances, but that brain dys-function can pose an additional hazard, probablyrelated to the involvement of the limbic system.

4.1. Temporal lobe epilepsy

Whether patients with TLE are at increased risk ofdeveloping psychiatric disorders compared withpatients with other types of epilepsy remains one ofthe most controversial issues in the literature concern-ing epilepsy and psychiatry. This topic has beenrepeatedly investigated over the last decades. Severalinvestigators reported more psychiatric disturbancesin patients with TLE compared with extra-TLE and/or primary generalized epilepsy patients [12,19,35–37].Others, however, found no such differences [20,28,38–41]. Rodin et al. [37] made an important observation,namely, that patients with TLE often have more thanone seizure type, and they suggest that the number ofseizure types is more relevant to emotional and psychi-atric problems in epilepsy than the type of seizure perse. These findings were later confirmed by Hermannet al. [42] and Dodrill [43]. Apart from the numberof seizure types, it is likely that there are more riskfactors (e.g., age at onset, laterality of the temporalepileptiform focus) that predispose to or protect pa-tients with TLE from psychopathology [44,45]. Itmight appear that variables other than or in additionto TLE are important determinants of psychiatricdisorders in epilepsy. Because the limbic system isinvolved in the regulation of emotional behavior, itis likely that the exact localization and lateralizationof the epileptogenic zone are of importance. The lim-bic system is situated in the medial parts of the tempo-ral lobes, so more psychiatric disturbances (inparticular mood disorders) are expected to be foundin patients with an epileptic focus in these parts ofthe brain. Additionally, concomitant frontal lobe dys-function may also be of importance.

4.2. Model of Hermann and Whitman

Epilepsy should be considered a syndrome character-ized by different manifestations and etiologies. It is ar-gued that patients with severe, intractable epilepsy areat a higher risk of developing psychiatric problems thanpatients with a milder (less active) form of epilepsy. Formany patients, epileptic seizures are uncontrollable andunpredictable, which makes living with epilepsy difficult.Also, the patient�s fears and concerns regarding his orher seizures, perceived stigma and discrimination (par-ticularly in the area of employment), and lack of socialsupport are considered potential etiological (psychoso-cial) variables in the development of psychiatric distur-bances [46,47]. Degree of limbic system dysfunction,brain abnormalities, age at onset of seizures, numberof seizure types, etiology, etc., are also important riskfactors. Ideally, when psychiatric disorders in epilepsyare investigated, all these factors should be taken intoaccount. The inconsistency found among the studieson psychiatric comorbidity is, apart from the differentdiagnostic instruments used, in considerable part attrib-utable to this multitude of variables on which the patientgroups are dissimilar.

This brings us to the work of Hermann and Whitman[45]. They divided potential risk variables supposed tobe associated with psychiatric disorders in epilepsy intothree main categories: (1) brain-related factors, (2) non-brain-related factors, and (3) treatment-related factors.They proposed that these risk factors form a conceptualmodel of psychopathology in epilepsy. However, at thetime of its formulation it was premature to attempt toestimate the relative importance (or explanatory power)of the individual risk variables, and this so-called‘‘conceptual model’’ should rather be seen as a list ofvariables having some potential relationship to psycho-pathology in epilepsy (see Table 2).

An important benefit of such an overview is that ithelps to present the many known or suspected etiologi-cal variables in a manner that encourages new and moreempirical research. It should be kept in mind that thesevariables may not be independent of one another, butmay instead be highly intercorrelated and reflect a moregeneral factor, for example, severity of epilepsy. Theseverity of epilepsy can be deduced from such variablesas the presence of multiple seizure types, early age at on-set, poor seizure control, and symptomatic etiology. Thesame consideration most likely applies to non-brain-re-lated and treatment-related factors. The authors further-more hypothesized that the three factors probablyexplain different proportions of variance for differentdisorders: e.g., brain-related factors are supposed to ex-plain most of the variance in psychosis, whereas non-brain-related factors likely explain most of the variancein affective disorders. This assumption was recently elab-orated for depression in an article by Hermann et al.

Table 2Factors associated with psychopathology in epilepsy (from: Hermann & Whitman, 1984)

(1) Brain-relatedNeurological variables Results of history and neurological exam; structural brain abnormalities; neuropathological findings;

neurochemical/neurohormonal abnormalitiesEEG variables Topographic distribution of epileptiform activity; laterality of focal spike activity;

degree of limbic system dysfunction; nonepileptiform EEG abnormalitiesEpilepsy variables Etiology; seizure control; presence of multiple seizure types; age at onset; duration of disorder;

seizure-related experiences/clinical seizure features; seizure typeNeuropsychological variables Overall level of performance; pattern of performance

(2) Non-brain-relatedChronic Illness variables Presence of a chronic disorder; economic stress; limitations of activities and aspirations; otherEpilepsy variables Nature of epilepsy; medical ignorance and fear of seizures; stigma; social consequences and discriminationDevelopmental/epilepsy variables Familial considerations; effect of epilepsy on the family unit; altered expectations for a child with epilepsy;

age at onset; early experiencesDemographic/subject variables Socioeconomic status; age; sex; premorbid personality/psychological status; other

(3) Treatment-related(Reynolds, 1981) Folate deficiency; altered monoamine metabolism; effects on endocrine function;

induction of neuropathological changes in the CNS

Source. Reprinted, with permission, from Hermann and Whitman [45].

42 W.A.M. Swinkels et al. / Epilepsy & Behavior 7 (2005) 37–50

[48]. They reviewed 36 studies of depression in epilepsyto obtain more insight into the factors that are reliablepredictors of depression. In total, 60 different potentialpredictor variables were examined. Of all the variables,neurological and epilepsy variables (brain-related) weremost frequently studied in the case of interictal depres-sion, and psychological/social variables and medicationvariables were the least investigated. Nonetheless, thesebrain-related variables resulted in the fewest positivefindings (6%), whereas the psychological and social vari-ables were associated most frequently with depression(79% positive findings). Surprisingly, variables such asseizure frequency, age at onset, and duration of epilepsywere hardly associated with depression.

Since its introduction in the 1980s, the model of Her-mann and Whitman has been empirically evaluatedmany times, with generally the most attention beingpaid to the biological and psychosocial risk factors. Itis demonstrated that certain variables seem to be moreclosely related to depression than other variables; how-ever, little is understood about the causal relationshipsinvolved. Even now, many uncertainties and controver-sies remain with respect to the precise origin of the var-ious psychiatric disorders in epilepsy. Also, few attemptshave been made to integrate the three factors into acomprehensive model of psychopathology in epilepsy.

5. Personality disorders

Whereas the majority of the studies concentrate onpsychiatric disorders (in particular psychotic disorders,mood disorders, and anxiety disorders), relatively littleresearch is available on the comorbidity of personalitydisorders in epilepsy. Just as in studies concerning axisI clinical disorders, the MMPI is frequently used when

axis II personality disorders in epilepsy are studied[23,42,49,50]. For reasons mentioned earlier, we do notreport on these MMPI studies. Only a few investigatorsuse psychiatric diagnostic methods. Existing data onpersonality disorders in patients with epilepsy reveal aprevalence between 4 and 38% [21,22,24,26,51–54] (Ta-ble 3). In the majority of the studies no control groupis included, which makes it difficult to compare thesefindings with those for nonepileptic (normal) controlsubjects and control subjects with other chronic medicalconditions. Nevertheless, a number of studies have esti-mated the prevalence of personality disorders in thecommunity and reported prevalence rates between 5.9and 13.4% [55–58]. Compared with these figures, therate of prevalence of personality disorders in patientswith epilepsy seems to be slightly increased. In line withthese findings, Swinkels et al. [54] found that patientswith epilepsy exhibit more personality disorder traitscompared with a control group from the general popu-lation. Although these authors used dimensional scores(number of criteria met for each personality disorder) in-stead of rates of prevalence of personality disorders,they did use an instrument based on the criteria of axisII personality disorders according to DSM-IV. Whenthe results were compared with results for asthma out-patients [59], the epilepsy patients had higher scores.Previously, Schwartz and Cummings [51] also found sig-nificantly more personality disorders in patients withepilepsy, compared with a control group of neurologicalpatients.

From these results, we can conclude with some reser-vation that the higher comorbidity of personality disor-ders in epilepsy dose not seem to be a consequence of thechronic medical condition per se, but is possibly relatedto the epilepsy condition itself. However, these findingsare based on a limited number of studies, and further

Table 3Epidemiological studies on personality disorders in adult epilepsy patientsa

Investigators (country, year) N Instruments Results Comments

Schwartz and Cummings(USA, 1988) [51]

21 epilepsy outpatients,24 neurological controls

Classification by psychiatristaccording to DSM-III

PDs in 38% of epilepsypatients and 4% of controls

No standardized instrument;small patient groups

Fiordelli et al. (Italy, 1993) [21] 100 epilepsy outpatients,100 controls

Clinical Interview Schedule(CIS) DSM-III-R criteria

PDs in 4% of patients and0% of controls

Control persons were healthyindividuals visiting the samehospital for minor reasons

Victoroff (USA, 1994) [22] 60 candidates for epilepsysurgery (all CPS)

Structured Clinical Interview forDSM-III-R patient version (SCID-P)

PDs in 18.33% of epilepsypatients

No control group; separatediagnoses on axis I and axis II

Manchanda et al.(Canada, 1996) [24]

300 candidates for epilepsysurgery

Present State Examination(PSE) DSM-III-R

PDs in 18% of epilepsy patients,especially cluster C disorders

No control group; separatediagnoses on axis I and axis II

Arnold and Privitera(USA, 1996) [26]

27 epilepsy inpatients,14 PNES patients

Structured Clinical Interview forDSM-III-R Epilepsy Version (SCID)

PDs in 18% of epilepsy patients(avoidant PD most prevalent) and36% of PNES patients (borderlineand avoidant PD most prevalent) (n.s.)

Small sample size; no normalcontrol group

Lopez-Rodriguez et al.(USA, 1999) [52]

52 epilepsy patients evaluated ascandidates for epilepsy surgery

Structured Clinical Interview forDSM-III-R PDs (SCID-II)

PDs in 21.15% of epilepsy patientsespecially cluster C disorders (15.8%)

No control group; medicallyrefractory epileptic patients

Krishnamoorthy et al.(UK, 2001) [53]

32 epilepsy patients,8 PNES patients

Standardized Assessment ofPersonality (SAP), DSM-III-R

PDs found for epilepsy group: 17.1%cluster A, 2.9% cluster B 17.1% cluster C

Small samples; half of the PNESgroup had comorbid epilepsy

Swinkels et al. (TheNetherlands, 2003) [54]

203 epilepsy inpatients,332 normal controls

Questionnaire on Personality Traits (VKP)according to DSM-IV criteria

Higher-dimensional scores forepilepsy patients on seven PDs

Large patient group; controlgroup; dimensional instead ofcategorical scores

a PD, personality disorder; CPS, complex partial seizures; PNES, psychogenic nonepileptic seizures.

W.A.M

.Swinkels

etal./Epilep

sy&

Behavio

r7(2005)37–50

43

44 W.A.M. Swinkels et al. / Epilepsy & Behavior 7 (2005) 37–50

investigation of the (relative) importance of the chronic-ity factor in the association between epilepsy andpersonality disorders is needed. Moreover, it is recom-mended that other chronic patient groups, in additionto the epilepsy group, be included to provide a moredirect comparison between different patient groups.

Furthermore, from Table 3 we see that several studiesreported a prevalence rate of personality disorders ofapproximately 18%, which is lower, on average, thanwhat was found for axis I clinical disorders. The sameapplies, however, to the general population, in whomthe prevalence of axis I disorders is also higher than thatof axis II disorders. In addition, the range of prevalencerates reported for personality disorders seems to besmaller. This may be explained by the fact that all stud-ies (with one exception) used standardized diagnosticinstruments based on DSM criteria. Another consistentfinding is the predominance of cluster C personality dis-orders in patients with epilepsy. However, this is not aunique finding because cluster C disorders are also mostprevalent in the general population.

It has been hypothesized that these personality disor-ders might be the result of the psychosocial consequenc-es of living with epilepsy, as a maladaptive reaction to achronic disorder, or the result of disrupted neuronalfunctioning, or a combination of both. However, Lo-pez-Rodriquez et al. [52] found that among all epilepsyvariables, only presence of an aura was positively corre-lated with the likelihood of having a personality disor-der. No difference in the rate of personality disordersbetween patients with TLE and patients with other typesof seizures was observed. Also, Swinkels et al. [54] didnot find any association between the localization ofthe epileptogenic zone and personality disorder traits.Moreover, they found that only a modest part of thevariance of particularly cluster C personality disordertraits can be explained by epilepsy-related variables(i.e., severity and duration of the epilepsy). So it canbe concluded that, at this time, the relative contributionof these co-called ‘‘brain-related’’ and ‘‘non-brain-relat-ed’’ factors to the association between epilepsy and per-sonality disorders remains uncertain. Only very fewstudies have investigated these risk factors associatedwith personality disorders.

6. Are there psychiatric disorders specific to epilepsy?

Several neuropsychiatrists with a special interest inepilepsy suggest that the existing systems of classifica-tion of psychiatric disorders and personality disordersare inadequate as far as epilepsy is concerned [60–62].They assume that there is an underestimation andunderrecognition of psychopathology due mainly tothe instruments used. These authors argue that theseinstruments are developed to detect mental disorders

in the general population rather than mental disordersspecific to epilepsy. They acknowledge that patients withepilepsy could have comorbid psychiatric disorders andpersonality disorders that fulfill criteria for the DSM-IVclassification system, but, in addition, they hypothesizethat there are also types of psychopathology that arerather specific to epilepsy and require another classifica-tion system. However, it remains to be seen if these sup-posedly epilepsy-specific disorders exist. In the nextsections, three of these syndromes are illustrated, name-ly, interictal dysphoric disorder, interictal behavior syn-drome, and the psychoses of epilepsy. We discusswhether the syndrome exists and how specific it is toepilepsy.

6.1. Interictal dysphoric disorder

Depressive disorders in epilepsy patients can beidentical to those of nonepileptic patients: they meetthe diagnostic criteria for major depression or dysthy-mic disorder. Depression is the most common comor-bid psychiatric disorder associated with epilepsy (see,e.g., [63,64]). The lifetime prevalence has been estimat-ed to be between 6 and 30% in population-based stud-ies and up to 50% among patients in tertiary centers.However, a certain number of patients have an atypi-cal clinical presentation of depressive symptoms thatfails to meet the criteria of any of the DSM affectivedisorders. Mendez et al. [65] observed that epilepsy pa-tients (although they met the criteria for a majordepressive disorder) had atypical peri-ictal featureswith more paranoia and psychotic symptoms. Thesepatients also had a more chronic dysthymic course be-tween the major depressive episodes, in which theyshowed more irritability and emotionality. Severalinvestigators believe that the various mood complaintsin epilepsy patients are different from those seen innonepileptic patients. Kraepelin [66] provided a clinicaldescription of such a disorder, and in recent years,Blumer et al. [67,68] introduced the term interictal dys-

phoric disorder (IDD) to refer to these depressivesymptoms observed specifically in patients with epilep-sy. The symptoms of IDD have an intermittent courseand can be categorized into depressive-somatoformsymptoms and affective symptoms. The depressive-somatoform symptoms include depressive mood,anergia, pain, and insomnia. The affective symptomsinclude irritability, euphoric mood, fear, and anxiety.Blumer et al. [67,68] reported that almost one-thirdto one-half of patients with epilepsy seeking medicalcare suffer from IDD of sufficient severity to requirepharmacological treatment. Although the overall sever-ity of the symptoms of IDD is milder than that of amajor depression, there is a significant impact on thepatient�s social relations, daily activities, and qualityof life. Because of the interrupted course of the symp-

W.A.M. Swinkels et al. / Epilepsy & Behavior 7 (2005) 37–50 45

toms, there is a failure to meet the DSM-IV criteria fordysthymic disorder [69].

Nevertheless, the existence and specificity of a charac-teristic depressive syndrome such as IDD amongpatients with epilepsy is controversial. As far as the exis-tence of IDD is concerned, whether it represents a differ-ent syndrome unique to patients with epilepsy or issimply a syndrome not fully meeting the DSM-IV crite-ria for a depressive disorder (i.e., subthreshold) remainsuncertain. This issue is part of a more general probleminherent to the use of a categorical classification systemsuch as the DSM. In many cases the number of symp-toms experienced is just below the threshold to confirma psychiatric (DSM) diagnosis. This does not automati-cally mean that the problems the patient experiences arenot severe enough to pay attention to. Unfortunately,studies comparing depression with IDD are notablyabsent (with the exception of the study by Kanner[64]). It is possible that there exists a spectrum with achronic dysthymic state characterized by the featuresof IDD that may intermittently exacerbate and at thattime meet the criteria for major depressive disorder.Another question is how specific such a syndrome is toepilepsy. It is unknown whether patients with otherchronic medical conditions (with a paroxysmal course)also experience the same combination of symptoms.Clearly, more systematic studies are needed to clarifythese issues.

6.2. Interictal behavior (or personality) syndrome

Specific interictal personality characteristics in epilep-sy have been recognized for centuries and describedextensively [70–73]. Initially, the characteristics werederived largely from case studies and anecdotal observa-tions. Later, personality traits were more systematicallyinvestigated. In the course of time, different names wereused to describe the cluster of interictal characteristics(viz. ‘‘interictal behavior syndrome,’’ ‘‘interictal person-ality syndrome,’’ ‘‘epileptic personality’’). All of thesenames refer to the same subset of features. It shouldbe realized that these interictal personality traits occurin some patients with epilepsy, whereas the majority ofpatients have a normal personality structure. Thesetraits are difficult to define, quantify, and study. It is fre-quently not clear which traits are attributable to the epi-lepsy and which result from other factors. For example,structural brain lesions such as mesial temporal sclerosisand tumors can cause personality changes independentof the epilepsy itself. Likewise, (antiepileptic) medicationand psychosocial problems can also have significantbehavioral consequences. Furthermore, it is often diffi-cult to distinguish postictal states from interictal peri-ods. Again, the specificity of the personality featuresto epilepsy is an open question. None of the interictaltraits in epilepsy have ever proven to be unique to epi-

lepsy; i.e., they also occur among psychiatric and otherneurological patients and normal subjects [74,75].

Many epilepsy patients do not meet the DSM-IV cri-teria for a specific personality disorder, yet they showsome typical personality oddities. Waxman andGeschwind [71] described a subset of personality char-acteristics, including deepened emotionality, circum-stantiality, disrupted religious and sexual concerns,and hypergraphia, among patients with TLE, andattributed these to an ‘‘interictal behavioral syndrome.’’These features are not necessarily maladaptive (or indic-ative of psychopathology), and the syndrome should beseen as one of behavioral change rather than of abehavioral disorder. A few years later, Bear and Fedio[72] developed an inventory specifically to assess theseand other reported personality traits. This so-calledBear–Fedio Inventory (BFI) consists of 18 traits, whichreflect characteristics in behavior, thought, and affectthat previously have been associated in the literaturewith TLE. These traits are listed in Table 4. Each traitwas assessed with five true-false items, so the scores ran-ged from 0 to 5.

Frequencies of all 18 traits (self-reports) were in-creased in patients with TLE, compared with those ofnormal and neurological controls. The most significantdifferences were humorlessness, circumstantiality,dependence, and sense of personal destiny. Raters (i.e.,family or friends) identified TLE patients as significantlydifferent from controls on 14 traits, most strongly forcircumstantiality, obsessionalism, and dependence. Bear[73] and Bear and Fedio [72] suggest that these person-ality changes result from a sensory–limbic hyperconnec-tion syndrome, in which the epileptic focus leads toenhanced associations between affects and stimuli (e.g.,irritability, deepened emotion, hyperreligiosity, hyper-graphia). To some extent, this is the opposite of the Klu-ver–Bucy syndrome [76] in which limbic dysfunctionleads to failure to attribute the appropriate emotionalsignificance to stimuli (reflects sensory–limbic discon-nection). It should be kept in mind that the majorityof epilepsy patients do not spontaneously offer descrip-tions of these traits; therefore, these could be easilymissed unless specifically asked for.

Since then, many studies have been performed usingthe BFI with mixed results (for an overview, see, e.g.,[75]). In their overview, Devinsky and Najjar came totwo general conclusions. First, the BFI consistently dis-criminates epilepsy patients from normals and nonpsy-chiatric (somatic) control patients, but fails todistinguish epilepsy patients from psychiatric patients.Second, the comparisons of TLE and primary general-ized epilepsy patients are inconclusive.

From these findings, it can be concluded that the BFIdoes not seem to define a personality syndrome specificto TLE. However, some traits seem to occur more oftenor more intensely in epileptic patients than in the general

Table 4Interictal personality traits ascribed to temporal lobe epilepsy

Trait Clinical observation

Viscosity Stickiness, tendency to repetitionHumorlessness, sobriety Overgeneralized ponderous concern, humor lackingSadness/depression Discouragement, tearfulness, self-depreciationAnger Increased temper, irritabilityAggression Overt hostility, rage attacks, violent crimes, murderAltered sexual interest Loss of libido, hyposexualismCircumstantiality Loquacious, pedantic, overly detailedParanoia Suspicious, overinterpretative of motives and eventsGuilt Tendency toward self-scrutiny and self-recriminationHyperreligiosity Holding deep religious beliefsFeeling of personal destiny Egocentricity, personal events highly chargedHypergraphia Keeping extensive diaries, detailed notesPhilosophical interest Nascent metaphysical or moral speculations, cosmological theoriesElation, euphoria Grandiosity, exhilarated moodHeightened emotionality Deepening of all emotions, sustained intense affectDependence, passivity Cosmic helplessness, ‘‘at hands of fate’’Obsessiveness Ritualism, orderliness, compulsive attention to detailHypermoralism Attention to rules with inability to distinguish significant from minor infraction

46 W.A.M. Swinkels et al. / Epilepsy & Behavior 7 (2005) 37–50

population. Most of these patients have partial seizuresof temporal lobe origin. Several of these traits (e.g., vis-cosity, hypergraphia, circumstantiality, hyperreligiosity,and dependence) resemble characteristics of cluster Cpersonality disorders, for example, obsessive–compul-sive disorder. From studies of axis II personality disor-ders we know that cluster C personality disorders arefrequently observed in patients with epilepsy. Thus,some overlap seems to exist, and it may be suggestedthat the traits described by the BFI are features associ-ated with axis II personality disorders (particularly clus-ter C disorders). Nevertheless, it is still uncertainwhether a cluster of personality traits exists that is spe-cific to patients with (temporal lobe) epilepsy as theBFI fails to distinguish epilepsy patients from psychiat-ric patients.

6.3. Psychosis of epilepsy

Psychotic syndromes in epilepsy have traditionallybeen classified according to their temporal relationshipwith the seizures (ictal, postictal, interictal). Because thisarticle concerns interictal psychopathology, we focus onthe interictal psychosis presumed to be associated withepilepsy. A distinction is made between brief interictalpsychosis and more chronic psychotic states.

Itmay often be difficult to distinguish postictal psycho-sis from brief interictal psychotic episodes also called ‘‘al-ternating psychoses [77].’’ In these brief interictalpsychotic states, an antagonistic relationship between epi-lepsy and psychosis is presumed in which periods of in-creased seizure activity can alternate with seizure-freeintervals during which a patient becomes psychotic. Thisphenomenon, called forced normalization [78], refers tostabilization of the EEG and not necessarily to seizurecontrol. The episodes are brief (last from several days to

weeks), may end in a convulsion, the consciousness is usu-ally clear, and often an associationwithTLE is found [78–82]. The phenomenology is further characterized by para-noid delusions and auditory hallucinations, and affectivesymptoms may also occur [77,83–85]. The evidence is,however, limited, somore research on this topic is needed.

Whereas brief interictal psychoses are relativelyuncommon (about 10% of all psychoses in epilepsy),chronic interictal (schizophrenia-like) psychoses are pre-sumed to occur more often (in 20% of the psychoses inepilepsy) [86]. It is suggested that schizophrenia-like psy-chosis is 6 to 12 times more likely to occur in epilepticpatients than in the general population [87]. The riskhas probably been overestimated because many studiesused highly selected patient populations. For example,the landmark study by Slater et al. [88] received consid-erable criticism principally because of its selection bias(the epilepsy patients with psychosis were specificallysent to these hospitals because they were difficult cases).Nevertheless, this study was of major importance be-cause these investigators were the first to perform a sys-tematic study of psychosis in a sample of 69 hospitalizedepilepsy patients and they introduced the term schizo-

phrenia-like psychosis to indicate the similarities betweenthese two disorders. According to the present-day classi-fication of the DSM-IV, the diagnosis of schizophreniarequires the presence of at least two of the following fivesymptom categories with a minimal duration of 1month: delusions, hallucinations, disorganized speech,grossly disorganized or catatonic behavior, and negativesymptoms. Also, patients must be dysfunctional in so-cial and occupational domains or in self-care for a min-imum of 6 months, during which they may present withnegative symptoms or with at least two of the symptomsoutlined above, but of lesser severity. Slater and co-workers proposed that, if the patients in their study

W.A.M. Swinkels et al. / Epilepsy & Behavior 7 (2005) 37–50 47

had not had epilepsy, the psychoses probably wouldhave been diagnosed as schizophrenia. Yet, togetherwith several other investigators, they stated that theschizophrenia-like psychosis of epilepsy has some atypi-cal features that are different from what is seen in classicschizophrenia. Notable and frequently mentioneddifferences are the lack of negative symptoms (i.e.,preservation of warm affect) and adequate or evenwell-preserved personalities and interpersonal relationsin schizophrenia-like psychoses [88–96]. Slater et al.[88] also mentioned a high frequency of delusions andreligious, mystical experiences. A number of investiga-tors described the symptoms as being largely paranoid-hallucinatory [88–91,95,97]. Furthermore, the onsetof schizophrenia-like psychosis usually occurs after a10- to 20-year duration of epilepsy, and an associationwith TLE is repeatedly suggested [88,91,93–95].

It has to be mentioned that several of these investiga-tors are from the same institute and so their findings arebased on the same patient material. Also, the lack of a(psychiatric) control group and phenomenologicaldescriptions that are based on impression onlymake thesefindings not very convincing. Perez and colleagues [98]tried to overcome these weaknesses by comparing 11 pa-tients with epilepsy and psychosis with 9 nonepilepticschizophrenic controls using the Present State Examina-tion (PSE). They concluded that the psychotic episodesin patients with epilepsy are indistinguishable from thoseof classic schizophrenics. More of these studies areurgently needed to investigate similarities and dissimilar-ities between schizophrenia and schizophrenia-like psy-chosis systematically with standardized psychiatricmeasures.

There is also much controversy about the risk factors.For instance, the relation between temporal lobe seizurefoci and psychosis remains to be proven (e.g.,[39,44,99]). These authors argued that the apparent ex-cess of TLE in patients with both epilepsy and psychosisreflects nothing more than that TLE is the commonestform of epilepsy. There seems to be, however, consensusthat psychosis of epilepsy is less severe than schizophre-nia and responds better to therapy [100].

The position that a specific interictal psychotic disor-der exists in patients with epilepsy has some support. Itseems that patients with epilepsy can experience psy-chotic symptoms rather similar to those of schizophre-nia. Several authors noted the relative lack of negativesymptoms and a more benign course for epileptic schizo-phrenia but controlled studies on this topic are lacking.Apparently, some differences, albeit small ones, exist be-tween interictal psychoses in epilepsy and psychosis inclassic schizophrenia. Apart from that, the boundary be-tween postictal and brief interictal psychosis is poorlydefined, as is the distinction of brief interictal psychosisfrom more chronic psychotic episodes. Clearly, moreinvestigation is needed.

7. Summary and conclusions

It is commonly believed that epilepsy carries with itan increased risk of developing interictal psychiatric dis-turbances. Research in the field of epilepsy and psychia-try has concentrated on epilepsy mainly as a biologicalcondition. Currently, it is being recognized that the med-ical and psychosocial dimensions of epilepsy are just as(or even more) important.

A multiplicity of studies on psychiatric comorbidityin epilepsy have been performed, and many different pa-tient groups and diagnostic instruments used. Themajority of these studies concern clinical disordersaccording to axis I of the DSM. Prevalence rates of axisI psychiatric problems vary between 19 and 80%. Thislarge variation in percentages is probably attributableto the different patient groups investigated and the evengreater variety of diagnostic methods (i.e., some studiesuse no diagnostic instrument at all). A considerablenumber of studies did not use a control group of‘‘healthy’’ subjects or other patient groups with whichto compare results. From the studies that did include acontrol group of healthy (nonmedical) subjects, it wasfound that the prevalence of axis I disorders is higherfor patients with epilepsy. Studies that (also) includedanother chronic medical patient group assume that thechronicity of the medical condition is an important fac-tor for an increased risk among epilepsy patients fordeveloping axis I psychiatric disorders, but brain dys-function seems to be just as important. The precise con-tribution of the different epilepsy-related variables to theincreased risk of psychiatric problems in patients withepilepsy remains unclear. It seems that both brain-relat-ed and non-brain-related factors are important.

Whereas many investigations concentrate on axis Idisturbances, relatively little research is performed onthe prevalence of axis II personality disorders. The prev-alence rate of personality disorders in epilepsy varies be-tween 4 and 38%, which is more consistent than what isfound for axis I disorders. Probably this smaller vari-ability is the result of the fact that almost all studies usedstandardized diagnostic instruments based on the DSMcriteria for personality disorders. In a comparison of theresults for epilepsy patients with findings from the gen-eral population, it seems that the prevalence of person-ality disorders in patients with epilepsy is slightlyincreased. From the little research that is available, italso seems that more personality disorders occur in epi-lepsy patients compared with patients with other chronicmedical conditions. In accordance with what is found inthe general population is the proneness of cluster C per-sonality disorders in epilepsy. So far, little is knownabout the relative contribution of biological and psycho-social factors to personality disorders.

Despite the high prevalence of especially axis I disor-ders, but also axis II disorders, observed with traditional

48 W.A.M. Swinkels et al. / Epilepsy & Behavior 7 (2005) 37–50

diagnostic methods, some neuropsychiatrists suggestthat the prevalence of psychiatric disorders in epilepsyis underestimated. They assume the existence of psychi-atric disorders specific to patients with epilepsy and thatare not detected with traditional instruments. Commonexamples of such syndromes are interictal dysphoric dis-order, interictal behavior or personality syndrome, andthe psychoses of epilepsy. Notwithstanding the largenumber of articles on these syndromes, the existenceand specificity of such syndromes solely in patients withepilepsy remain open questions. Possibly, these syn-dromes can be diagnosed as psychiatric disordersaccording to DSM or are manifestations of axis I or axisII disorders on a subthreshold level. The latter possibil-ity is a problem that might be seen as inherent to the useof classification systems (i.e., categorical, instead ofdimensional) and does not imply the existence of realepilepsy-specific disorders. However, this issue is stillcontroversial.

From the many studies that have been performed onpsychiatric comorbidity in epilepsy, it can be learned thatthere is a need for well-controlled studies using represen-tative patient groups and valid and standardized diagnos-tic instruments.More research on personality disorders inparticular is recommended. Finally, future studies shouldfocus more on the relative contributions of the differentbrain- and non-brain-related variables to the relation-ships between epilepsy and psychiatric disorders.

References

[1] Cowan LD, Leviton A, Bodensteiner JB, Doherty L. Problemsin estimating the prevalence of epilepsy in children: the yieldfrom different sources of information. Peadiatr Perinat Epidem-iol 1989;3:386–401.

[2] Hauser WA, Annegers JF, Kurland LT. Prevalence of epilepsy inRochester, Minnesota: 1940–1980. Epilepsia 1991;32:429–45.

[3] Rwiza HT, Kilonzo GP, Haule J, et al. Prevalence and incidenceof epilepsy in Ulanga, a rural Tanzanian district: a community-based study. Epilepsia 1992;33:1051–6.

[4] Placencia M, Sander JW, Roman M, et al. The characteristics ofepilepsy in a largely untreated population in rural Ecuador. JNeurol Neurosurg Psychiatry 1994;57:320–50.

[5] Shackleton DP, Westendorp RGJ, Kasteleijn-Nolst TreniteDGA, De Boer A, Herings RMC. Epilepsy medication: a roadto determing the number of individuals with seizures. J ClinEpidemiol 1997;50:1061–8.

[6] Hathaway SR, McKinley JC. Minnesota multiphasic personalityinventory. Minneapolis: University of Minnesota; 1943.

[7] Tempkin O. The falling sickness. second ed. Baltimore: JohnsHopkins Univ. Press; 1971.

[8] Jackson JH. Selected writings. In: Taylor J, editor. On epilepsyand epileptiform convulsions, Vol. 1. London: Hodder &Stoughton; 1931.

[9] Gowers WR. Epilepsy and other chronic convulsive diseas-es. London: J&A Churchill; 1881.

[10] Briquet P. Traite clinique et therapeutique de l�hysterie. Paris.1859.

[11] Morel BA. Traite des malades mentales. Paris: Masson et Cie;1860.

[12] Pond DA, Bidwell BH. A survey of epilepsy in fourteen generalpractices: II. Social and psychological aspects. Epilepsia 1959/1960;1:285–99.

[13] Edeh J, Toone B. Relationship between interictal psychopathol-ogy and the type of epilepsy. Br J Psychiatry 1987;151:95–101.

[14] Jalava M, Sillanpaa M. Concurrent illnesses in adults withchildhood-onset epilepsy: a population-based 35-year follow-upstudy. Epilepsia 1996;37:1155–63.

[15] Stefansson SB, Olafsson E, Hauser WA. Psychiatric morbidity inepilepsy: a case controlled study of adults receiving disabilitybenefits. J Neurol Neurosurg Psychiatry 1998;64:238–41.

[16] Currie S,HeathfieldKWG,HensonRA, ScottDF.Clinical courseand prognosis of temporal lobe epilepsy. Brain 1971;94:173–90.

[17] Kogeorgos J, Fonagy P, Scott DF. Psychiatric symptom patternsof chronic epileptics attending a neurological clinic: a controlledinvestigation. Br J Psychiatry 1982;140:236–43.

[18] Schiffer RB, Babigian HM. Behavioral disorders in multiplesclerosis, temporal lobe epilepsy, and amyotrophic lateralsclerosis: an epidemiologic study. Arch Neurol 1984;41:1067–9.

[19] Gureje O. Interictal psychopathology in epilepsy: prevalence andpattern in a Nigerian clinic. Br J Psychiatry 1991;158:700–5.

[20] Manchanda R, Schaefer B, McLachlan R, Blume WT. Interictalpsychiatric morbidity and focus of epilepsy in treatment-refrac-tory patients admitted to an epilepsy unit. Am J Psychiatry1992;149:1096–8.

[21] Fiordelli E, Beghi E, Bogliun G, Crespi V. Epilepsy andpsychiatric disturbance. Br J Psychiatry 1993;163:446–50.

[22] Victoroff J. DSM-III-R psychiatric diagnoses in candidates forepilepsy surgery: lifetime prevalence. Neuropsychiatry Neuro-psychol Behav Neurol 1994;7:87–97.

[23] Silberman EK, Sussamn N, Skillings G, Callanan M. Auraphenomena and psychopathology: a pilot investigation. Epilep-sia 1994;35:778–84.

[24] Manchanda R, Schaefer B, McLachlan RS, et al. Psychiatricdisorders in candidates for surgery for epilepsy. J NeurolNeurosurg Psychiatry 1996;61:82–9.

[25] Perini GI, Tosin C, Carraro C, et al. Interictal mood andpersonality disorders in temporal lobe epilepsy and juvenilemyoclonic epilepsy. J Neurol Neurosurg Psychiatry1996;61:601–5.

[26] Arnold LM, Privitera MD. Psychopathology and trauma inepileptic and psychogenic seizure patients. Psychosomatics1996;37:438–43.

[27] Etttinger AB, Weisbrot DM, Krupp LB, Jandorf L, Gaudino E,Cramer J. Symptoms of psychiatric disturbances in epilepsy. JEpilepsy 1998;11:10–4.

[28] Swinkels WAM, Kuyk J, De Graaf EH, Van Dyck R, SpinhovenPh. Prevalence of psychopathology in Dutch epilepsy inpatients:a comparative study. Epilepsy Behav 2001;2:441–7.

[29] Wells KB, Golding JM, Burnam MA. Psychiatric disorder in asample of the general population with and without chronicmedical conditions. Am J Psychiatry 1988;145:976–81.

[30] Baldwin MV. A Clinico-experimental investigation into thepsychologic aspects of multiple sclerosis. J Nerv Ment Dis1952;115:299–342.

[31] Surridge D. An investigation into some psychiatric aspects ofmultiple sclerosis. Br J Psychiatry 1969;115:749–64.

[32] Whitlock FA, Siskind MM. Depression as a major symptom ofmultiple sclerosis. J Neurol Neurosurg Psychiatry 1980;43:861–5.

[33] Dalos NP, Rabins PV, Brooks BR. Disease activity andemotional state in multiple sclerosis. Ann Neurol 1983;13:573–7.

[34] Pollak Y, Orion E, Goshen I, Ovadia H, Yirmiya R. Experi-mental autoimmume encephalomyelitis-associated behavioralsyndrome as a model of �depression due to multiple sclerosis�.Brain Behav Immun 2002;16:533–43.

[35] Gibbs FA, Gibbs EL, Furster B. Psychomotor epilepsy. ArchNeurol Psychiatry 1948;60:331–9.

W.A.M. Swinkels et al. / Epilepsy & Behavior 7 (2005) 37–50 49

[36] Gudmundsson G. Epilepsy in Iceland. Acta Neurol Scand1966;43(Suppl.25):1–124.

[37] Rodin EA, Katz M, Lennox K. Differences between patientswith temporal lobe seizures and those with other forms ofepileptic attacks. Epilepsia 1976;17:313–20.

[38] Small JG, Small IF, Hayden MP. Further psychiatric investiga-tions of patients with temporal and nontemporal lobe epilepsy.Am J Psychiatry 1966;123:303–10.

[39] Stevens JR. Psychiatric implications of psychomotor epilepsy.Arch Gen Psychiatry 1966;14:461–71.

[40] Mignone RJ, Donnelly EF, Sadowsky D. Psychological andneurological comparisons of psychomotor and non-psychomo-tor epileptic patients. Epilepsia 1970;11:345–59.

[41] Standage KF, Fenton GW. Psychiatric symptom profiles ofpatients with epilepsy: a controlled investigation. Psychol Med1975;5:152–60.

[42] Hermann BP, Dikmen S, Wilensky AJ. Increased psychopathol-ogy associated with multiple seizure types: fact or artifact?.Epilepsia 1982;23:587–96.

[43] Dodrill CB. Number of seizure types in relation to emotionaland psychosocial adjustment in epilepsy. In: Porter RJ, Ward JrAA, Mattson RH, Dam M, editors. Advances in epileptology:XVth Epilepsy International Symposium. New York: RavenPress; 1984. p. 541–4.

[44] Stevens JR, Hermann BP. Temporal lobe epilepsy, psychopathol-ogy and violence: the state of evidence. Neurology1981;31:1127–32.

[45] Hermann BP, Whitman S. Behavioral and personality correlatesof epilepsy: a review, methodological critique, and conceptualmodel. Psychol Bull 1984;95:451–97.

[46] Hauser WA, Hesdorffer DC. Epilepsy: Frequency, causes andconsequences. New York: Demos; 1990.

[47] Hermann BP, Whitman S. Psychopathology in epilepsy: the roleof psychology in altering paradigms of research, treatment, andprevention. Am Psychologist 1992;47:1134–8.

[48] Hermann BP, Seidenberg M, Bell B. Psychiatric comorbidity inchronic epilepsy: identification, consequences, and treatment ofmajor depression. Epilepsia 2000;41(Suppl. 2):S31–41.

[49] Hermann BP, Dikmen S, Schwartz MS, Karnes WE. Interictalpsychopathology in patients with ictal fear: a quantitativeinvestigation. Neurology 1982;32:7–11.

[50] Provinciali L, Franciolini B, Del PesceM, SignorinoM, CeravoloG. Influence of neurological factors on the personality profile ofpatients with temporal lobe epilepsy. J Epilepsy 1989;2:239–44.

[51] Schwartz J, Cummings JL. Psychopathology and epilepsy: anoutpatient consultation–liaison experience. Psychosomatics1988;29:295–300.

[52] Lopez-Rodriguez F, Altshuler L, Kay J, Delarhim S, Mendez M,Engel J. Personality disorders among medically refractory epilep-tic patients. J Neuropsychiatry Clin Neurosci 1999;11:464–9.

[53] Krishnamoorthy EA, Brown RJ, Trimble MR. Personality andpsychopathology in nonepileptic attack disorder and epilepsy.Epilepsy Behav 2001;2:418–22.

[54] Swinkels WAM, Duijsens IJ, Spinhoven Ph. Personality disordertraits in patients with epilepsy. Seizure 2003;12:587–94.

[55] Samuels JF, Nestadt G, Romanoski AJ, Folstein MF, McHughPR. DSM-III personality disorders in the community. Am JPsychiatry 1994;151:1055–62.

[56] Samuels JF, Eaton WW, Bienvenu III OJ, Brown CH, Costa JrPT, Nestadt G. Prevalence and correlates of personality disor-ders in a community sample. Br J Psychiatry 2002;180:536–42.

[57] JacksonHJ, Burgess PM. Personality disorders in the community:a report from the Australian National Survey of Mental HealthandWellbeing. Soc Psychiatry Psychiatr Epidemiol 2000;35:531–8.

[58] Torgerson S, Kringlen E, Cramer V. The prevalence of person-ality disorders in a community sample. Arch Gen Psychiatry2001;58:590–6.

[59] Bauer H, Duijsens IJ. Personality disorders in pulmonarypatients. Br J Med Psychol 1998;71:165–73.

[60] Krishnamoorthy ES. An approach to classifying neuropsychiat-ric disorders in epilepsy. Epilepsy Behav 2000;1:373–7.

[61] Krishnamoorthy ES. Neuropsychiatric disorders in epilepsy:epidemiology and classification. In: Trimble M, Schmitz B,editors. The neuropsychiatry of epilepsy. Cambridge: Cam-bridge Univ. Press; 2002. p. 5–17.

[62] Barry JJ, Lembke A, Huynh N. Affective disorders in epilepsy.In: Ettinger AB, Kanner AM, editors. Psychiatric issues inepilepsy. Philadelphia: Lippincott Williams & Wilkins; 2001. p.45–71.

[63] Altshuler L. Depression and epilepsy. In: Devinsky O, TheodoreWH, editors. Epilepsy and behavior. New York: Wiley–Liss;1991. p. 47–65.

[64] Kanner AM. Depression in epilepsy: prevalence, clinical semi-ology, pathogenic mechanisms, and treatment. Biol Psychiatry2003;54:388–98.

[65] Mendez MF, Cummings JL, Benson DF. Depression in epilepsy:significance and phenomenology. Arch Neurol 1986;43:766–70.

[66] Kraepelin E. Psychiatrie, vol. 3. Leipzig: Johann AmbrosiusBarth; 1923.

[67] Blumer D. Epilepsy and disorders of mood. In: Smith D,Treiman D, Trimble M, editors. Neurobehavioral problems inepilepsy. New York: Raven Press; 1991. p. 85–196.

[68] Blumer D, Montouris G, Hermann B. Psychiatric morbidity inseizure patients on a neurodiagnostic monitoring unit. J Neuro-psychiatry Clin Neurosci 1995;7:445–56.

[69] Kanner AM, Kozak AM, Frey M. The use of sertraline inpatients with epilepsy: is it safe?. Epilepsy Behav 2000;1:100–5.

[70] Gastaut H, Morin G, Leserve N. Etude du comportement desepileptiques psycho-moteurs dans l�intervalle de leurs crises. AnnMedico-Psychol 1955;113:1–27.

[71] Waxman SG, Geschwind N. The interictal behaviour syndromeof temporal lobe epilepsy. Arch Gen Psychiatry 1975;32:1580–6.

[72] Bear DM, Fedio P. Quantitative analysis of interictal behaviourin temporal lobe epilepsy. Arch Neurol 1977;34:454–67.

[73] Bear D. Temporal lobe epilepsy: a syndrome of sensory-limbichyperconnection. Cortex 1979;15:357–84.

[74] Devinsky O, Vazquez B. Behavioral changes associated withepilepsy. Neurol Clin 1993;11:127–49.

[75] DevinskyO,Najjar S. Evidence against the existence of a temporallobe personality syndrome. Neurology 1999;53(Suppl. 2):S13–25.

[76] Kluver H, Bucy P. Preliminary analysis of functions of thetemporal lobe in man. Arch Neurol Psychiatry 1939;42:979–1000.

[77] Tellenbach H. Epilepsie als anfallsleiden und als psychose. Uberalternatieve psychosen paranoider pragung bei �forcierter nor-malisierung� (Landolt) des electroencephalogramms epilepti-scher. Nervenartz 1965;36:190–202.

[78] Landolt H. Serial electroencephalographic investigations duringpsychotic episodes in epileptic patients and during schizophrenicattacks. In: Lorentz de Haas AM, editor. Lectures on epilep-sy. Amsterdam: Elsevier; 1958.

[79] Landolt H. Some clinical electroencephalographic correlations inepileptic psychoses (twilight states). Electroencephologr ClinNeurophysiol 1953;5:121.

[80] Landolt H. Uber Verstimmungen, Dammerzustande und schizo-phrene Zustandsbilder bei Epilepsie. Schweiz Arch NeurolPsychiatry 1955;76:313–21.

[81] Glaser GH, Newman RJ, Schafer R. Interictal psychosis inpsychomotor-temporal lobe epilepsy: an EEG psychologicalstudy. In: Glaser GH, editor. EEG and behavior. New York: -Basic Books; 1963.

[82] Dongier S. Statistical study of clinical and electroencephalograph-ic manifestations of 536 psychotic episodes occurring in 516epileptics between clinical seizures. Epilepsia 1959/1960;1:117–42.

50 W.A.M. Swinkels et al. / Epilepsy & Behavior 7 (2005) 37–50

[83] Ramani V, Gumnit RJ. Intensive monitoring of interictalpsychosis in epilepsy. Ann Neurol 1982;11:613–22.

[84] Wolf P. The clinical syndromes of forced normalization. Jpn JPsychiatry Neurol 1984;38:187–92.

[85] Pakalnis A, Drake ME, John K, Kellum JB. Forced normali-zation: acute psychosis after seizure control in 7 patients. ArchNeurol 1987;44:289–92.

[86] Schmitz B. Psychiatric issues. In: Schmidt D, Schachter SC,editors. Epilepsy: Problem solving in clinical practice. Lon-don: Martin Dunitz Ltd; 2000. p. 169–89.

[87] Sachdev P. Schizophrenia-like psychosis and epilepsy: the statusof the association. Am J Psychiatry 1998;155:325–36.

[88] Slater E, Beard AW, Glithero E. The schizophrenia-likepsychosis of epilepsy. V. Discussion and conclusions. Br JPsychiatry 1963;109:95–150.

[89] Hill D. Psychiatric disorders of epilepsy. Med Press1953;229:473–5.

[90] Pond DA. Psychiatric aspects of epilepsy. J Indian Med Prof1957;3:142–5.

[91] Bruens JH. Psychoses in epilepsy. Psychiatr Neurol Neurochir1971;74:17–192.

[92] Stoudemire A, Nelson A, Houpt JL. Interictal schizophrenia-likepsychosis in temporal lobe epilepsy. Psychosomatics 1983;24:331–9.

[93] Kristensen O, Sindrup EH. Psychomotor epilepsy and psychosis:I. physical aspects. Acta Neurol Scand 1978;57:361–9.

[94] Oyebode F, Davison K. Epileptic schizophrenia: clinical featuresand outcome. Acta Psychiatr Scand 1989;79:327–31.

[95] PerezMM, Trimble MR. Epileptic psychosis: diagnostic compar-ison with process schizophrenia. Br J Psychiatry 1980;137:245–9.

[96] Toone BK, Garralda ME, Ron MA. The psychosis of epilepsyand the functional psychosis: a clinical and phenomenologicalcomparison. Br J Psychiatry 1982;141:256–61.

[97] Parnas J, Korsgaard S, Krautwald O, Jensen PS. Chronicpsychosis in epilepsy: a clinical investigation of 29 patients. ActaPsychiatr Scand 1982;66:282–93.

[98] PerezMM,TrimbleMR,MurrayNMF,ReiderI.Epilepticpsychosis:an evaluation of PSE profiles. Br J Psychiatry 1985;146:155–63.

[99] McKenna PJ, Kane JM, Parrish K. Psychotic syndromes inepilepsy. Am J Psychiatry 1985;142:895–904.

[100] Kanner AM. Psychosis of epilepsy: a neurologist�s perspective.Epilepsy Behav 2000;1:219–27.