A critical and descriptive approach to interictal behavior with the Neurobehavior
Inventory (NBI)
Guilherme Nogueira M. de Oliveira
a,b,⁎
, Arthur Kummer
a, Renato Luiz Marchetti
c,
Gerardo Maria de Araújo Filho
d, João Vinícius Salgado
a, Anthony S. David
e, Antônio Lúcio Teixeira
a aNeuropsychiatric Unit, Neurology Division, School of Medicine, Universidade Federal de Minas Gerais, Belo Horizonte, BrazilbEpilepsy Treatment Advanced Centre (NATE), Felicio Rocho Hospital, Belo Horizonte, Brazil
cDepartment and Institute of Psychiatry, Faculdade de Medicina da Universidade de São Paulo (FMUSP), São Paulo, Brazil dDepartment of Neurology and Neurosurgery, Universidade Federal de São Paulo (UNIFESP), São Paulo, Brazil
eSection of Cognitive Neuropsychiatry, Institute of Psychiatry, King's College London, DeCrespigny Park, London SE5 8AF, UK
a b s t r a c t
a r t i c l e
i n f o
Article history:
Received 23 July 2012 Revised 21 August 2012 Accepted 23 August 2012 Available online 24 October 2012
Keywords:
Temporal lobe epilepsy Interictal behavior
Gastaut-Geschwind syndrome Neurobehavior Inventory Psychiatric disorders Depression
Purpose:The purpose of this study was to test the psychometric properties of the Neurobehavior Inventory (NBI) in a group of temporal lobe epilepsy (TLE) patients from a tertiary care center, correlating its scores with the presence of psychiatric symptoms.
Methods:Clinical and sociodemographic data from ninety-six TLE outpatients were collected, and a neuro-psychiatric evaluation was performed with the following instruments: Mini-Mental State Examination (MMSE), structured psychiatric interview (MINI-PLUS), Neurobehavior Inventory (NBI), and Hamilton Depression Rating Scale (HAM-D).
Results:Some traits evaluated by the NBI showed adequate internal consistency (mean inter-item correlation between 0.2 and 0.4) and were frequent, such as religiosity (74%) and repetitiveness (60.4%). Principal com-ponent analysis showed three factors, named here as emotions (Factor 1), hyposexuality (Factor 2), and un-usual ideas (Factor 3). Depressive symptoms on HAM-D showed a strong association with emotions and hyposexuality factors. When patients with left TLE and right TLE were compared, the former exhibited more sadness (p=0.017), and the latter, a greater tendency toward sense of personal destiny (p=0.028). Conclusion:Depression influences NBI scoring, mainly emotionality and hyposexuality traits. Neurobehavior Inventory subscales can be better interpreted with an appropriate evaluation of comorbid mood and anxiety disorders. Compromise in left temporal mesial structures is associated with increased tendency toward sad affect, whereas right temporal pathology is associated with increased beliefs in personal destiny.
© 2012 Elsevier Inc.
1. Introduction
Unusual behaviors have long been described in epilepsy, especially in temporal lobe epilepsy (TLE). More than 3000 years ago, the Babylonians reported mystical-religious states in people with epilepsy
[1]. In the nineteenth century, psychiatrists highlighted frequent irritability, anger, and rage in people with epilepsy. The concept of “larvate epilepsy”proposed by Morel, and supported by Falret, repre-sented periodic alternations in the mental state such as aggressive, manic, and depressive states that could indicate non-convulsive seizure activity[2]. Kraepelin called attention to the meticulousness, slowness, persistence, adhesion, perseveration, and mental stickiness present in patients with epilepsy that resulted in very circumstantial speech, a trait that came to be designated as viscosity [3]. The
alternating behavior with moments of excessive religious and ethical concern and episodes of irritability and temper tantrums were also pointed out by Kraepelin in the early twentieth century[4]. Curiously, Freud described the contradictory personality of Dostoevsky, who possibly had TLE[5], alternating periods of uncontrolled irritability, and violence with deepened guilt and attempts to follow religious and moral codes[3].
In the second half of the past century, many authors revisited the interictal behavioral changes in epilepsy. Gastaut noted that the emo-tional intensity,“stickiness”, and hyposexuality were exactly opposite to the placidity, unfocussed attention, and hypersexuality shown by primates following bilateral temporal lobectomy, a condition named Klüver-Bucy syndrome[6,7]. Later, Geschwind noticed that the tenden-cy of patients with TLE to write extensively (hypergraphia) was associ-ated with affective lability and preoccupation with ethical and religious issues. Thus, interictal behaviors characterized by changes in sexual behavior (usually hyposexuality), irritability, viscosity (slow thinking, circumstantial, and detailed speech), religiosity, hypergraphia, and fre-quent mood swings became known as Gastaut-Geschwind syndrome
–
⁎ Corresponding author at: Hospital Felício Rocho, Núcleo Avançado de Tratamento das Epilepsias (NATE), Av. do Contorno, 9530. Barro Preto, 30110‐934 Belo Horizonte MG, Brazil. Fax: +55 31 3514 7294.
E-mail address:[email protected](G.N.M. de Oliveira). 1525-5050 © 2012 Elsevier Inc.
http://dx.doi.org/10.1016/j.yebeh.2012.08.031
Contents lists available atSciVerse ScienceDirect
Epilepsy & Behavior
j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / y e b e h
Open access under the Elsevier OA license.
[8,9]. It is worth mentioning that some features can also be observed in the periictal period, such as hyperreligiosity and hypergraphia, mainly associated with postictal mania and psychosis[10,11].
Though personality study concepts are relevant in the investiga-tion of long‐standing behavioral changes, the overall term“epileptic personality”is nowadays regarded as obsolete, stigmatizing, and im-precise[4,15], and we preferred the term“interictal behavior”when referring to the characteristics of Gastaut-Geschwind syndrome. The issue is surrounded by controversy as some studies were unable to discriminate a specific syndrome in TLE when compared with other neurological and psychiatric control groups[12–14]. Though classi-cally described, the diagnostic hallmarks of the syndrome are not clearly defined and, therefore, subject to much criticism [15]. The Neuro-Behavioral Inventory (NBI)[4], an adaptation of the earlier Bear–Fedio Inventory (BFI)[16], was introduced to evaluate such as-pects of interictal epileptic behavior (viscosity, emotions, moralism, spirituality, and hyposexuality) as well as episodic states that may have clinical relevance (mood instability, irritability, paranoia, and anxiety) and has been translated and adapted to the Brazilian popula-tion[17]. The purpose of this study was to evaluate NBI performance in a group of TLE patients and to correlate the behavioral characteris-tics with psychopathology and clinical variables.
2. Subjects and statistical methods
A cross-sectional study was conducted with 96 TLE outpatients. All attended the Epilepsy Clinic of the Neurology Service, Hospital das Clínicas, Universidade Federal de Minas Gerais (UFMG), which is a regional referral service for refractory epilepsy. This service encom-passes around 1000 patients who are referred from general practi-tioners, general physicians, or general neurologists from the whole state of Minas Gerais, Brazil. This study was approved by the local ethics committee in accordance with the Declaration of Helsinki. The inclusion criteria were as follows: diagnosis of TLE according to the ILAE criteria[18], age of more than 18 years, and capacity to pro-vide written informed consent for participation in the study. Patients were excluded if they had severe medical or neurologic diseases other than epilepsy, history of previous neurosurgery, and severe cognitive impairment according to MMSE performance[19].
Sociodemographic (age, gender, ethnicity, marital status, employ-ment situation, and educational level) and clinical information (age at onset and duration of the epilepsy, seizure type, seizure frequency, antiepileptic drug regime, MRI, and EEGfindings) were collected. Data of epileptiform activity on interictal EEG were registered for de-scriptive purposes, as this information has limited lateralizing proper-ties since it can represent propagation from other sites or can be normal in patients with epilepsy. As only 51 patients (53.1%) were assessed with VEEG telemetry, we decided to use the mesial temporal sclerosis (MTS) diagnosis on MRI as a lateralizing reference.
Psychiatric diagnosis was made according to the Mini Internation-al Neuropsychiatric Interview (MINI) Plus version 5.0.0.[20–22], a structured clinical interview that provides diagnosis according to the DSM-IV criteria. Hamilton Depression Rating Scale (HAM-D)
[23–25]was used to measure depressive symptoms. We used the 24-item HAM-D (total score range: 0–75), which includes helpless-ness, hopelesshelpless-ness, and low self-esteem items[25], in order to reduce the weight of the somatic symptoms, increasing its specificity.
The Brazilian-Portuguese version of the NBI[17]was used to as-sess the behavioral traits of the Gastaut-Geschwind syndrome. The inventory is composed of 100 true or false items (total score range: 0–100), grouped in 20 subscales (5 items in each subscale). For subscale scores (total score range: 0–5), 3 or more positive answers were considered to be indicative of caseness, and a total scale score of 20 or more determined overall caseness, as suggested by the inven-tory developer[4].
Descriptive analysis of categorical variables and proportions were calculated and presented. Cronbach's coefficient alpha and mean inter-item correlations were computed to ascertain NBI total and subscale internal consistency. Non-parametric correlations (Spearman's rho) were calculated between the NBI subscale scores and the HAM-D. Principal component analysis was performed in order to identify correlation within the set of traits. Varimax rotation was performed to assist interpretation of extracted factors. An adjusted score was at-tributed to every factor. Each subscale had a weighted score (subscale total score multiplied by the factor correlation value). The factor ad-justed score was then calculated: sum of factor subscale weighted scores divided by the total number of subscales in each factor. For comparison of categorical variables between groups, Fisher's exact test was performed, and continuous variables were evaluated using the Kruskal–Wallis and Mann–WhitneyUtests.
All tests were performed using SPSS version 15.0 for Windows. A two-sidedpvalue lower than 0.05 was considered significant for all tests.
3. Results
The sociodemographic and clinical characteristics of the ninety-six outpatients with TLE included in the study are displayed inTable 1.
Most of the patients (88.5%) were on more than one antiepileptic drug (AED). Carbamazepine (62.5%), clobazam (47.9%), lamotrigine (22.9%), clonazepam (21.9%), and topiramate (17.7%), in monotherapy or in combination, were the most prescribed AEDs. Other AEDs used were valproate (14.6%), phenytoin (14.6%), phenobarbital (12.5%), and oxcarbazepine (6.3%).
3.1. Psychiatric and behavioral features
Temporal lobe epilepsy patients had a high frequency of lifetime psychiatric disorders (68.8%) (Table 2). Depression (35.4%) was the most common psychiatric disorder, while anxiety disorders, as a group, (40.6%) were also prevalent. Panic disorder was diagnosed in 10 patients (10.4%), agoraphobia in 12 (12.5%), social phobia in 13 (13.5%), simple phobia in 12 (12.5%), generalized anxiety disorder in 22 (22.9%), and mixed anxiety-depression disorder in 1 (1%) pa-tient. Of note, in some patients, more than one anxiety disorder diag-nosis was observed. Suicide risk was present in almost a third (31.3%) of TLE patients. One patient had epileptic (right temporal seizures) and psychogenic non-epileptic seizures recorded during VEEG telem-etry. This patient showed no major psychiatric disorder according to the MINI-PLUS and exhibited only minimum depressive symptoms (HAM-D=3 points). Curiously, he had a high NBI total score (49 points) and maximum scores (5 points) on persistence and repeti-tiveness, interest in details, religious convictions, cosmic interests, and feelings about sex subscales. He also had high scores (4 points) on orderliness and sense of personal destiny subscales.
NBI total score showed a high Cronbach's alpha (0.9), but a low mean inter-item correlation (0.107). Some subscales showed mean inter-item correlation lower than 0.2, such as fearfulness, sense of personal destiny, dependency, and happiness. Seriousness showed mean item correlation lower than 0.1. Total and subscale inter-nal consistencies and frequency of traits in the sample are shown in
Table 3.
3.2. Factor analysis
Neurobehavior Inventory's 20 subscales were subjected to princi-pal component analysis (PCA). Correlation values lower than 0.1 were suppressed. Factorability of the correlation matrix was supported by Bartlett's Test of sphericity below 0.001 and Kaiser-Meyer-Olkin Mea-sure of Sampling Adequacy of 0.869.
Principal component analysis revealed the presence of 3 main components with eigenvalues over 1, explaining 35%, 10%, and 6.8% of the variance, respectively. Varimax rotation solution is presented
inTable 4.
Subscales were grouped corresponding to some features of the Gastaut-Geschwind syndrome, described here as the following: emo-tions (Factor 1), hyposexuality (Factor 2), and unusual ideas (Factor 3). Emotions factor (mood, guilt, dependence, sadness, and anger) accounted for the larger percentage (35%) of the total variance. Inter-estingly, the so-called viscosity items (orderliness, interest in details, persistence, and repetitiveness) were also grouped in the emotions factor. Viscosity seems to be a prevalent subtrait in TLE as 70% of the patients had at least 2 elevated scores in the subscales pertaining to viscosity, and only 15% of the patients were negative for all 3
subscales related to viscosity. The hyposexuality factor not only was represented by hyposexuality, anxiety symptoms (fear), seriousness, and somatic complaints, but also showed some association with emo-tions factor (sadness and anger) subscales. By the same token, sub-scales of the hyposexuality factor (fearfulness and somatization) were associated with the emotions factor. The unusual ideas factor was diametrically opposed to emotions and hyposexuality factors and was represented mainly by subscales related to spirituality. When spirituality components (religious convictions, cosmic inter-ests, and sense of personal destiny) were analyzed together, only 22.9% of the patients did not show any of these traits. Hypergraphia showed a weak correlation (b0.3) with all factors, but was more
closely related to the unusual ideas factor.
3.3. Impact of depression and other psychiatric disorders on behavioral aspects of the NBI
As depression was the most frequent disorder, we analyzed the in-fluence of depressive symptoms (HAM-D) on NBI traits, as depicted in
Fig. 1.
Fig. 1 stresses the strong correlation between depression
(HAM-D) and NBI subscales. Of note, HAM-D was associated with the emotions (Factor 1) and hyposexuality (Factor 2) factor subscales, but not with unusual ideas factor (Factor 3) subscales.
3.4. Psychopathology regarding laterality
Patients were divided in four groups regarding laterality of MTS: 40 patients with right MTS (RMTS), 35 with left MTS (LMTS), 10 with bitemporal MTS (BMTS), and 11 patients without MTS (WMTS). Groups were statistically different according to secondary generalization of partial seizures (p=0.012), underrepresented in the RMTS group, present in 5 patients (12.5%), compared to 14 (40%) patients with LMTS, 5 (50%) with BMTS, and 4 (36.4%) patients with WMTS. Also, there were differences across groups regarding du-ration of epilepsy (p=0.011), as the WMTS group had a lower mean time since the onset of uncontrolled seizures, mean (SD) of 19.6 (15.1) years, in comparison with 29.1 (9.4) years (SD) in RMTS, 33.9 (11.5) years (SD) in LMTS, and 32 (11) years (SD) in the BMTS group. There were no other significant differences on the sociodemographic and clinical variables.
When NBI behavioral traits were compared, there were no signif-icant differences across all four groups. However, when patients with LMTS and RMTS were compared separately, the LMTS group scored significantly higher on sadness (p =0.017) and lower on sense of per-sonal destiny (p=0.028). No other significant differences were ob-served in this analysis.
Table 1
Sociodemographic and clinical characteristics of the study sample. Patients (n=96) n or mean (SD)
Proportion (%) or median (range) Gender
Male 42 43.8%
Female 54 56.3%
Age, years 40.7 (10.1) 41 (21–65) Ethnic group
White 47 49%
Black 10 10.4%
“Pardo”a 39 40.6%
Educational level, years of study 8.0 (3.7) 8 (0–15) Marital statusb
Single 39 40.6%
Married/stable union 41 42.7% Divorced/separate 12 12.5%
Widowed 4 4.2%
Employment situation
Employed 41 42.7%
Unemployed 18 18.8%
Retired (age or time of service purposes) 6 6.3% Retired (due to illness) 31 32.3% Seizure type
Simple partial 45 46.9% Complex partial 91 94.8% Partial evolving to secondarily generalized 28 29.2% Localization of epileptiform discharges
Right temporal lobe 36 37.5% Left temporal lobe 30 31.3% Bilateral temporal lobe 20 20.8% No epileptiform discharges 10 10.4% Religion
Catholic 54 56.3
Protestant 33 34.4
Spiritism 4 4.2
Other 1 1.0
Without religion 4 4.2 Age at onset of epilepsy, years 10.7 (9.6) 9.5 (0–49) Duration of epilepsy, years 30.0 (11.7) 29 (3–54) Seizure frequency, seizures per month 6.3 (9.6) 3 (0–60) Seizure free for 6 months or longer 8 8.3% MRI
WMTS 11 11.5%
RMTS 40 41.7%
LMTS 35 36.5%
BMTS 10 10.4%
AEDs therapy regime (%)
Monotherapy 11 11.5%
Dualtherapy 58 60.4%
3 AEDs 23 24.0%
4 AEDs 4 4.1%
SD = Standard deviation; MRI = Magnetic resonance imaging; WMTS = Without MTS; RMTS = Right mesial temporal sclerosis; LMTS = Left mesial temporal sclerosis; BMTS = Bilateral MTS; AEDs = Antiepileptic drugs.
a“Pardo”: mixed race or color (mulato, mestizo). b One patient did not disclose marital status.
Table 2
Psychiatric profile in patients with temporal lobe epilepsy (note patients may have more than one diagnosis).
Psychiatric disorder/symptom and suicidality Patients (n=96) n or mean (SD)
Proportion (%) or median (range) Any disorder (current) 55 57.3% Any disorder (lifetime) 66 68.8% Bipolar Disorder (BD) 8 8.3% Major depressive episode (lifetime) 34 35.4% Major depressive episode (current) 27 28.1% Anxiety disorders 39 40.6% Obsessive–Compulsive Disorder (OCD) 8 8.3%
Psychosis 5 5.2%
Somatoform disorders 13 13.5% Presence of current suicide risk 30 31.3% HAM-D 15.2 (11.4) 12 (0–48) HAM-D = Hamilton Depression Rating Scale.
In an attempt to estimate the syndrome frequency, factor adjusted scores extracted from PCA were categorized with a split on the medi-an (caseness was defined if factor score was over the median). In this sense, twenty-one patients (21.9%) were negative for all factors, whereas eighteen (18.8%) patients were positive for all 3 factors, thirty-five (36.5%) for 2 factors, and twenty-two (22.9%) patients for 1 factor. None of the patients fulfilled the 20 positive traits, median (range) of 11 (0–19) elevated subscales per patient.
4. Discussion
In our sample, the majority of patients had high scores on the NBI (97% had scores over 20 points). The presentfindings suggest that the syndrome with all 20 traits evaluated by the NBI is rare, as none of the patients were positive for all subscales. When the traits were grouped in PCA, the presence of all three factors in the same patient was ob-served in a minority of the patients, around 20%. An importantfinding is that the frequency of many traits was similar to the ones reported
on the translation of the NBI into Brazilian-Portuguese that was ap-plied to 15 patients[17], showing the stability of the scale across this population.
Nowadays, personality experts seem to be more interested on sin-gle psychological constructs of limited domain rather than on the de-velopment of global measures of personality. Personality studies usually focus on the assessment of the variation across individuals. Therefore, measures of relationship (correlation coefficients) are used instead of mean differences (analysis of variance). When a large number of characteristics are measured, it is important to specify accurately each trait separately in groups of highly inter-correlated items[26]. According to Briggs and Cheek[26], an optimal level of homogeneity is reached when the mean inter-item correla-tion is within the 0.2 to 0.4 range. If inter-item correlacorrela-tion is lower than 0.1, it is unlikely that the total score represents the items' com-plexity, and if it is higher than 0.5, the items on the scale tend to be redundant. From this point of view, it is appropriate to suggest that future research may revise some items with low internal consistency, such as seriousness.
Based on the assumption that a“single scale ought to measure a single construct”, it is important to take into account that“any test that measures more than one common factor to a substantial degree yields scores that are psychologically ambiguous and very difficult to interpret”[26]. Thus, the mean pairwise correlation among the 100 items of 0.107 (range from−0.369 to 0.626) leads to the conclu-sion that there is a weak correlation among items, as many presented a negative correlation. An alpha of 0.93 indicates that this sample of items probably correlates around 93% with another sample of items drawn from the same population. However, a mean inter-item corre-lation of 0.107 indicates that the items are very disparate. The Neurobehavior Inventory probably measures a construct that is broad and polymorphous. One argument supporting this hypothesis is that there are 20 subscales within the NBI. This would imply that the NBI total score has limited value.
It is suggested that the characteristics of the Gastaut-Geschwind syndrome are manifested primarily in patients with TLE. Bear et al. reported higher sadness, emotionality (emotional intensity), and reli-gious and philosophical interests in 10 patients with TLE compared to 10 patients with generalized and focal non-temporal epilepsy[27]. Recently, a study using the BFI stressed that the only different trait between TLE patients subjects with psychogenic non-epileptic sei-zures (PNES) was hypergraphia, which was more frequent in TLE Table 3
Frequency of personality traits and internal consistency of the Neurobehavior Inventory.
Domains (items) Cronbach's alpha Caseness (%) Mean (SD) Median (range)
Mean inter-item correlation Total score 0.925 97.9 52.29 (16.228) 54 (16–85) 0.107 Writing tendency (6,18,53,62,98) 0.527 38.5 2.25 (1.392) 2 (0–5) 0.181 Sense of law and order (19,41,49,75,83) 0.494 74 3.24 (1.288) 3.5 (0–5) 0.170 Religious convictions (13,24,48,94,99) 0.595 74 3.35 (1.361) 4 (0–5) 0.238 Anger and temper (20,25,36,43,91) 0.741 44.8 2.22 (1.736) 2 (0–5) 0.364 Orderliness (5,14,21,42,71) 0.481 63.5 2.78 (1.363) 3 (0–5) 0.161 Feelings about sex (7,10,52,77,84) 0.672 49 2.67 (1.639) 2 (0–5) 0.290 Fearfulness (16,40,57,64,86) 0.465 49 2.59 (1.342) 2 (0–5) 0.150 Feelings of guilt (4,55,61,70,80) 0.537 60.4 2.74 (1.416) 3 (0–5) 0.193 Seriousness (29,45,66,88,100) 0.323 76 3.45 (1.123) 3 (1–5) 0.072 Sadness (33,38,73,85,92) 0.739 38.5 2.01 (1.645) 2 (0–5) 0.363 Emotions (9,23,54,69,89) 0.598 55.2 2.79 (1.514) 3 (0–5) 0.225 Suspicion (12,17,22,34,63) 0.689 41.7 2.09 (1.610) 2 (0–5) 0.307 Interest in details (26,50,58,67,81) 0.532 75 3.43 (1.351) 4 (1–5) 0.169 Cosmic interests (37,46,51,72,76) 0.555 52.1 2.58 (1.491) 3 (0–5) 0.200 Sense of personal destiny (1,11,27,32,97) 0.438 63.5 2.91 (1.290) 3 (0–5) 0.131 Persistence and repetitiveness (35,44,60,78,90) 0.698 60.4 2.90 (1.651) 3 (0–5) 0.316 Hatred and revenge (56,68,82,87,96) 0.668 24 1.49 (1.436) 1 (0–5) 0.291 Dependency (3,15,28,39,93) 0.514 40.6 2.31 (1.300) 2 (0–5) 0.145 Happiness (2,31,59,65,79) 0.475 29.2 1.89 (1.280) 2 (0–5) 0.154 Physical well-being (8,30,47,74,95) 0.601 55.2 2.60 (1.483) 3 (0–5) 0.225
Table 4
Principal component analysis of the 20 subscales from the Neurobehavior Inventory. Subscale Component
1 2 3
Suspicion .760 .103
Emotions .756 .112 .149
Feelings of guilt .733 .181
Dependency .726 .186 .107
Sadness .716 .426 −.140
Persistence and repetitiveness .710 .125 Anger and temper .698 .392 .188 Orderliness .665 .107
Sense of law and order .646 .173 Hatred and revenge .626 .156 Interest in details .521 .193 .253 Feelings about sex .107 .712
Seriousness .230 .621 .255 Fearfulness .452 .563
Physical well-being .487 .523
Religious convictions −.172 .189 .815
Cosmic interests .302 .703 Sense of personal destiny .144 −.129 .656
Happiness .171 −.443 .492
Writing tendency .153 .267
[12]. Mungas argues that a large percentage of the variation of traits evaluated by the BFI could be explained by the presence of mental dis-orders and general psychopathology. Hence, the inventory would be nonspecific and unable to distinguish TLE patients with the proposed syndrome from psychiatric patients[13]. In view of this, the present study highlights the major impact of depression on various features of the Gastaut-Geschwind syndrome, especially the“episodic traits”, such as anger and sadness. For example, sexual interests were influenced by depression, and the analysis of hyposexuality must consider the high frequency of depression among these patients. Also, more stable traits, such as viscosity subscales, were associated with depression. Circumstantiality and obsessionalism are characteristics observed in some patients without epilepsy in a mood disorders clinic
[28], and we could hypothesize that bidirectional mechanisms increase the vulnerability to mood disorders and viscosity, especially in TLE. Some of these symptoms may be better described by standard psychiatric evaluation, including assessment with instruments as the BDI, Brief Psychiatric Rating Scale, or even specific mood instruments such as the Interictal Dysphoric Disorder Inventory[29–31]or the Neu-rological Disorders Depression Inventory for Epilepsy[32,33]. Thus, an alternative explanation for so many traits in BFI and NBI is that some of these traits are‘artifacts’of underecognized and undertreated de-pression and anxiety in TLE. One message for clinicians is that treatment for mood/anxiety disorders should be optimal before behaviors are con-sidered to be due to enduring personality changes.
A possible explanation for the dichotomy between the unusual ideas factor (linked with spirituality) and emotions factor (and also hyposexuality factor) may lie on the secondary association with de-pression in the latter group. Consequently, the lack of association of the unusual ideas factor with depressive symptoms can be explained by the fact that strong spiritual and religious beliefs may be protective against depression[34]. In addition, hypergraphia was associated with the unusual ideas factor, pointing to a possible correlation with spiri-tuality. Curiously, many reports point out the religious themes written by many hypergraphic people with epilepsy [5,10,35]. It is also plausible that hypergraphia is an isolated component of the syn-drome, observed only in a few patients. Gastaut-Geschwind syndrome characteristics are difficult to measure, and some traits are really dif-ferent from pure depression/anxiety and could be better explained by impairment in language function, theory of mind, self-awareness, and environmental monitoring related to damage and decreased func-tional connectivity of mesial temporal structures[36].
Lesions in the non-dominant temporal lobe were associated with depression initially by Flor-Henry[37]. Subsequently, other studies have been devoted to this issue, and linked depression to an epileptic focus on the right hemisphere, while others do not show this associa-tion (30–32). For instance, Victoroff et al. found that left temporal hypometabolism was associated with depression in patients with par-tial seizures. However, patients with hypometabolism in the right temporal lobe also had a history of depression. These authors state that the onset of seizures in the left temporal lobe and the degree of hypometabolism in the temporal lobes were independent risk factors of depression[38]. Inactivation of the left hemisphere by injection of amobarbital to the left carotid artery can induce crying, pessimistic statements, guilt, complaints, and worries about the future[39]. Bear and Fedio[16]also pointed out that patients with right TLE minimized depressive complaints on self-reports, while patients with left TLE tended to exaggerate them. Altshuler et al. described that patients with left TLE scored higher on self-ratings for depression[40]. Van Elst et al. investigated the characteristics of the Geschwind syndrome with NBI in a group of nine patients with TLE and bilateral hippocam-pal atrophy in comparison to a group of 14 patients with TLE but with normal hippocampal volumes[41]. The group with bilateral hippo-campal atrophy showed significant reduction in the left amygdala (pb0.01), which may explain the tendency of this group to show
emo-tionality, fear, guilt, and sadness. Hence, thefindings of the present study are in line with the studies that link damage to the left temporal lobe to negative thoughts and feelings such as sadness, an important symptom of depressive disorder.
Bear and Fedio described 27 outpatients with TLE (15 with a right focus and 12 with left focus) that showed differences on behavioral profile according to the laterality of epileptic focus and also when compared with patients with neurological disease and asymptomatic controls[42]. These authors described a double dissociation at the principal component factor analysis in which thefirst factor was char-acterized by traits of emotionality (anger, aggression, and sadness), most frequently found in patients with right TLE, and an ideational/ ruminative factor (religion, seriousness, and personal destiny), more common in left TLE[42]. Our data go on the opposite direction of the Bear and Fedio study regarding sense of personal destiny, which was associated with RMTS in the present study [16]. It must be highlighted that our study was based on self-rating profiles, whereas in the Bear and Fedio study, this difference was observed on rater re-ports. Other explanations for this difference may be based on meth-odological issues; we used MTS on MRI to determine laterality, while Bear and Fedio used interictal EEG. Another difference is the number of patients used in each study [16]. Cultural differences across countries may have also played a role, and the fact that sense of personal destiny subscale had a low internal consistency (mean inter-item correlation of 0.131) can indicate that this is a fragile con-struct. Regardless of these methodological distinctions, other studies showed the involvement of the right temporal lobe in religious expe-riences. Smaller right hippocampus volume was observed in patients with epilepsy with high religiosity on the NBI[43]. Accordingly, a re-cent study showed a possible lateralizing value of a manual automa-tism usually made by a group of Christians, the Sign of the Cross, which was present in the ictal period of four patients, all with unilat-eral right mesial TLE[44]. Some authors also report ictal religious ex-periences linked to epileptic focus in the non-dominant hemisphere
[45–47]. As personal destiny was associated with spirituality factor and more frequent in RMTS in the present study, we can hypothesize that this supports the involvement of right temporal structures in subjective spiritual perception.
One important limitation of this study is the absence of a caregiver rating evaluation. Many patients came from the countryside without a close relative (or the relative had a low level of education) which impaired the rating evaluation. The next of kin impressions are need-ed as individuals with cognitive impairment or even some degree of
C o rr e la ti o n w it h H A M -D (S pe a rm a n' s rho ) NB I Tot al Score* * Susp ici on* * Em o tio n s ** Gu il t** D e pendency ** Sadness ** R e petitivene ss* * Anger* * O rderliness ** Law an d O rd e r* Ha tr e d ** Details* Hyposexuality* * Ser iousness ** Fearfulne ss* * So m a tiz atio n ** R e li gio sit y Cosmic Destiny H a ppi ness Hypergraphia -0.2 0.0 0.2 0.4 0.6 0.8 Factor 1 Factor 2 Factor 3
Fig. 1.Correlation of behavioral traits with depressive symptoms according to the Hamilton Depression Rating Scale (HAM-D). Subscales are displayed in the sequence presented in factor analysis (*pb0.05; **pb0.001).
psychopathology have impairment in insight[4]. However, self-rating instruments have benefits as many caregivers are unaware of the spiritual and religious beliefs of the patients[4]. The conclusion that depression influences NBI scoring is limited by the transversal design of the present study, which limits causal relationships. It is possible that some patients with behavioral changes may emphasize negative aspects of behavior, increasing self‐reports of depressive symptoms. However, considering the high frequency of depression in TLE, as well as the application in the present study of an interviewer rating instrument to quantify depression (HAM-D), we would be more in-clined to propose the main relevance of depression as a confounding factor on NBI scoring. Another limitation is the absence of VEEG te-lemetry to the whole sample, which could have improved lateralizing correlations and identification of PNES. We were able to diagnose PNES in only one patient, who showed no major psychiatric disorder according to conventional psychiatric evaluation. However, this pa-tient exhibited high scores on the NBI, mainly on religiosity, viscosity, and hyposexuality subscales. As it is not possible to generalize from a single case, we suggest the necessity of further studies designed to address the relevance of interictal behavior in PNES and vice versa. The present study is strengthened by a representative sample of TLE patients evaluated by a multidimensional structured psychiatric evaluation.
The fact that instruments such as the BFI and NBI do not support a global syndrome does not invalidate the observations made in these patients over decades pointing to some common behavioral charac-teristics. The present study demonstrated that NBI has valid factors to evaluate specific traits described in epilepsy, but many subscales are still unsatisfactory for research purposes. Also, new adaptations of the instrument must consider excluding or modifying redundant items of NBI subscales with inappropriate mean inter-item correla-tions, as well as subscales that overlapped with depressive symptoms, and this is one perspective of the present study. Future studies shall also compare TLE to other epilepsy syndromes and control groups without epilepsy to clarify the role of seizures and temporal lobe in-volvement in different aspects of interictal behavior. The instrument has a clear clinical value to assess some features of the behavior that the patient may not report spontaneously. Though sometimes maladaptive, some behavioral traits must not be mistaken with psy-chopathological symptoms and, therefore, must be distinguished from personality disorders. It is relevant to address this issue as it can orientate specific treatment programs, directed not only to psy-chiatric comorbidities, such as depression, but also cognitive rehabil-itation of dysfunctional behaviors.
5. Conclusion
Behavioral traits of the Gastaut-Geschwind syndrome are frequent in TLE. However, the presence of all traits in the same patient is uncommon, and possibly the syndrome as a whole is a rare event. Depression has a major impact on NBI scoring, principally on dimensions connected to emotions (anger and viscosity) and hyposexuality (fearfulness and somatization) factors. The high frequen-cy of underdiagnosed depression and anxiety in these patients may explain the description of a large number of subtraits seen in the emo-tive and hyposexual factors. Left temporal pathology may lead to nega-tive affect characterized by sadness, whereas right temporal pathology can prompt increased beliefs in personal destiny. Future adaptations of the NBI can bring new insights to the behavioral aspects of TLE.
Funding
This work was partly funded by the Brazilian funding agencies CAPES and Fapemig.
Acknowledgments
We are thankful to Mery Natali Silva Abreu for her statistical support.
References
[1] Reynolds EH, Wilson JVK. Psychoses of epilepsy in Babylon: the oldest account of the disorder. Epilepsia 2008;49(9):1488-90.
[2] Berrios GE. Epilepsy and insanity during the early 19th century: a conceptual history. Arch Neurol 1984;41(9):978-81.
[3] Blumer D. Evidence supporting the temporal lobe epilepsy personality syndrome. Neurology 1999;53(5 Suppl. 2):S9–S12.
[4] Blumer D. Personality disorders in epilepsy. Neuropsychiatry of personality disor-ders. Boston: Blackwell Science; 1995. p. 230-63.
[5] Seneviratne U. Fyodor Dostoevsky and his falling sickness: a critical analysis of seizure semiology. Epilepsy Behav 2010;18(4):424-30.
[6] Gastaut H. Interpretation of the symptoms of psychomotor epilepsy in relation to physiologic data on rhinencephalic function. Epilepsia 1954;3:84-8.
[7] Kluver H, Bucy PC. Preliminary analysis of functions of the temporal lobes in monkeys. Arch Neurol Psychiatry 1939;42(6):979–1000.
[8] Waxman SG, Geschwind N. The interictal behavior syndrome of temporal lobe epilepsy. Arch Gen Psychiatry 1975;32(12):1580-6.
[9] Trimble M, Freeman A. An investigation of religiosity and the Gastaut-Geschwind syn-drome in patients with temporal lobe epilepsy. Epilepsy Behav 2006;9(3):407-14. [10] Trimble MR. The soul in the brain: the cerebral basis of language, art and belief.
Baltimore: The Johns Hopkins University Press; 2007.
[11] Nishida T, Kudo T, Inoue Y, et al. Postictal mania versus postictal psychosis: differ-ences in clinical features, epileptogenic zone, and brain functional changes during postictal period. Epilepsia 2006;47(12):2104-14.
[12] Tremont G, Smith MM, Bauer L, et al. Comparison of personality characteristics on the Bear–Fedio Inventory between patients with epilepsy and those with non-epileptic seizures. J Neuropsychiatry Clin Neurosci 2012;24(1):47-52. [13] Mungas D. Interictal behavior abnormality in temporal lobe epilepsy. A specific
syn-drome or nonspecific psychopathology? Arch Gen Psychiatry 1982;39(1):108-11. [14] Locke DEC, Fakhoury TA, Berry DTR, Locke TR, Schmitt FA. Objective evaluation of
personality and psychopathology in temporal lobe versus extratemporal lobe epilepsy. Epilepsy Behav 2010;17(2):172-7.
[15] Devinsky O, Najjar S. Evidence against the existence of a temporal lobe epilepsy personality syndrome. Neurology 1999;53(5 Suppl. 2):S13-25.
[16] Bear DM, Fedio P. Quantitative analysis of interictal behavior in temporal lobe epilepsy. Arch Neurol 1977;34(8):454-67.
[17] de Oliveira GNM, Kummer A, Salgado JV, et al. Crosscultural adaptation of the neurobehavior inventory (NBI) for Brazil. J Epilepsy Clin Neurophysiol 2009;15:123-9. [18] Commission on Classification and Terminology of the International League Against Epilepsy. Proposal for revised classification of epilepsies and epileptic syndromes. Epilepsia 1989;30(4):389-99.
[19] Brucki SMD, Nitrini R, Caramelli P, Bertolucci PHF, Okamoto IH. Suggestions for utilization of the mini-mental state examination in Brazil. Arq Neuropsiquiatr 2003;61(3B):777-81.
[20] Amorim P. Mini International Neuropsychiatric Interview (MINI): validation of a short structured diagnostic psychiatric interview. Rev Bras Psiquiatr 2000; 22(3):106-15.
[21] Jones JE. Are anxiety and depression two sides of the same coin? In: Kanner A, Schachter S, editors. Psychiatric controversies in epilepsy. Amsterdam: Elsevier Inc.; 2008. p. 89–109.
[22] Mula M, Schmitz B, Jauch R, et al. On the prevalence of bipolar disorder in epilep-sy. Epilepsy Behav 2008;13(4):658-61.
[23] Krishnamoorthy ES. The evaluation of behavioral disturbances in epilepsy. Epilepsia 2006;47(s2):3-8.
[24] Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry 1960;23(1):56-62.
[25] Moreno RA, Moreno DH. Hamilton (HAM-D) and Montgomery & Asberg (MADRS) rating scales. Rev Psiquiatr Clín 1998;25(5):262-72.
[26] Briggs SR, Cheek JM. The role of factor analysis in the development and evaluation of personality scales. J Pers 1986;54(1):106-48.
[27] Bear D, Levin K, Blumer D, Chetham D, Ryder J. Interictal behaviour in hospitalised temporal lobe epileptics: relationship to idiopathic psychiatric syndromes. J Neurol Neurosurg Psychiatry 1982;45(6):481-8.
[28] Akiskal HS, Hirschfeld RM, Yerevanian BI. The relationship of personality to affec-tive disorders. Arch Gen Psychiatry 1983;40(7):801-10.
[29] Araújo Filho GM A, de Oliveira GNM, Oliva CH, et al. Translation and cross-cultural adaptation of the Interictal Dysphoric Disorder Inventory (IDDI). J Epilepsy Clin Neurophysiol 2010;16(4):155-61.
[30] Mula M, Jauch R, Cavanna A, et al. Interictal dysphoric disorder and periictal dys-phoric symptoms in patients with epilepsy. Epilepsia 2010;51(7):1139-45. [31] Mula M, Jauch R, Cavanna A, et al. Clinical and psychopathological definition of
the interictal dysphoric disorder of epilepsy. Epilepsia 2008;49(4):650-6. [32] Gilliam FG, Barry JJ, Hermann BP, Meador KJ, Vahle V, Kanner AM. Rapid detection
of major depression in epilepsy: a multicentre study. Lancet Neurol 2006;5(5): 399-405.
[33] de Oliveira GNM, Kummer A, Salgado JV, et al. Brazilian version of the Neurological Disorders Depression Inventory for Epilepsy (NDDI-E). Epilepsy Behav 2010;19(3): 328-31.
[34] Miller L, Wickramaratne P, Gameroff MJ, Sage M, Tenke CE, Weissman MM. Religiosity and major depression in adults at high risk: a ten-year prospective study. Am J Psychiatry 2012;169(1):89-94.
[35] Roberts J, Robertson M, Trimble M. The lateralising significance of hypergraphia in temporal lobe epilepsy. J Neurol Neurosurg Psychiatry 1982;45(2):131-8. [36] Pittau F, Grova C, Moeller F, Dubeau F, Gotman J. Patterns of altered functional
connectivity in mesial temporal lobe epilepsy. Epilepsia 2012;53(6):1013-23. [37] Flor-Henry P. Schizophrenic-like reactions and affective psychoses associated
with temporal lobe epilepsy: etiological factors. Am J Psychiatry 1969;126(3): 400-4.
[38] Victoroff JI, Benson DF, Grafton ST, Engel J, Mazziotta JC. Depression in complex partial seizures electroencephalography and cerebral metabolic correlates. Arch Neurol Fevereiro 1994;51(2):155-63.
[39] Hecht D. Depression and the hyperactive right-hemisphere. Neurosci Res 2010;68(2):77-87.
[40] Altshuler LL, Devinsky O, Post RM, Theodore W. Depression, anxiety, and tempo-ral lobe epilepsy. Latetempo-rality of focus and symptoms. Arch Neurol 1990;47(3): 284-8.
[41] van Elst LT, Krishnamoorthy ES, Bäumer D, et al. Psychopathological profile in patients with severe bilateral hippocampal atrophy and temporal lobe epilepsy: evidence in support of the Geschwind syndrome? Epilepsy Behav 2003;4(3):291-7. [42] Reynolds EH. Interictal behaviour in temporal lobe epilepsy. Br Med J (Clin Res
Ed) 1983;286(6369):918-9.
[43] Wuerfel J, Krishnamoorthy E, Brown R, et al. Religiosity is associated with hippo-campal but not amygdala volumes in patients with refractory epilepsy. J Neurol Neurosurg Psychiatry 2004;75(4):640-2.
[44] Lin K, Marx C, Caboclo LOSF, Centeno RS, Sakamoto AC, Yacubian EMT. Sign of the Cross (Signum Crucis): observation of an uncommon ictal manifestation of mesial temporal lobe epilepsy. Epilepsy Behav 2009;14(2):400-3.
[45] Åsheim Hansen B, Brodtkorb E. Partial epilepsy with“ecstatic”seizures. Epilepsy Behav 2003;4(6):667-73.
[46] Ozkara C, Sarý H, Hanoğlu L, Yeni N, Aydoğdu I, Ozyurt E. Ictal kissing and reli-gious speech in a patient with right temporal lobe epilepsy. Epileptic Disord 2004;6(4):241-5.
[47] Carrazana E, Cheng J. St Theresa's dart and a case of religious ecstatic epilepsy. Cogn Behav Neurol 2011;24(3):152-5.
