- Email: [email protected]
Seizure semiology in males with psychogenic nonepileptic seizures is associated with somatic complaints
Epilepsy Research 115 (2015) 153–157
Contents lists available at www.sciencedirect.com
Epilepsy Research journal homepage: www.elsevier.com/locate/epilepsyres
Seizure semiology in males with psychogenic nonepileptic seizures is associated with somatic complaints Shawn D. Gale a,∗ , Stacy W. Hill b , Caleb Pearson c a b c
Department of Psychology and Neuroscience Center, Brigham Young University, Provo, UT, USA Clearwater Neurosciences, Lewiston, ID, USA Center for Neuropsychology and Cognitive Neuroscience, University of Kansas Hospital, KS, USA
a r t i c l e
i n f o
Article history: Received 9 February 2015 Received in revised form 10 June 2015 Accepted 16 June 2015 Available online 21 June 2015 Keywords: Non-epileptic Somatization Psychogenic Gender differences
a b s t r a c t Aims: Psychopathology has been studied in patients with epileptic or psychogenic non-epileptic seizures in the context of diagnosis and treatment. Unfortunately, most PNES studies include few males and do not consider possible gender differences, making findings less generalizable to males with PNES. In this study we specifically compare males with PNES to females with PNES and to males with epilepsy. Methods: Males with PNES (n = 58), males with epilepsy (n = 86), females with PNES (n = 147), and females with ES (n = 142) were evaluated on an inpatient epilepsy monitoring unit. Self-reported objective measures of psychopathology, demographics, and PNES seizure semiology were compared. Results: Personality Assessment Inventory profiles revealed marked differences, particularly in somatic symptoms, between PNES and epilepsy. Females with PNES had higher levels of physiological depressive symptoms but lower antisocial features. Males with PNES who had clinically significant elevations on the somatic complaints scale were much more likely to have motor seizures while females with PNES classified similarly were equally likely to have either motor or non-motor events. Conclusion: Gender difference in PNES seizure semiology was associated with whether or not clinically significant somatic symptoms were present; males with elevated somatic symptoms were much more likely to have motor PNES. However, we did not find evidence of greater psychopathology in males with PNES compared to females with PNES. Gender differences in the behavioral manifestation of PNES in the context of presence or absence of somatization may have implications for diagnosis and treatment. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Psychogenic non-epileptic seizures (PNES) are paroxysmal events that mimic epileptic seizures (ES) but have psychological rather than neurological underpinnings and may be treatable with behavioral therapies (Dickinson and Looper, 2012). Although a careful clinical history and consideration of the behavioral components of the events (i.e. semiology) may increase confidence of a PNES diagnosis, video electroencephalography (vEEG) is the “gold standard” (Thomas et al., 2013). Given the complexity of this condition, psychological and neuropsychological testing may be coupled with vEEG (Dickinson and Looper, 2012; Walker, 2000). Given that the prevalence rate of PNES is three times higher in females than males, (Binder and Salinsky, 2007) most studies consist mainly of females, which could potentially affect diagnostic and treatment
∗ Corresponding author at: Department of Psychology and Neuroscience Center, Brigham Young University, 1060 SWKT, Provo, UT 84602, USA. E-mail address: shawn [email protected] (S.D. Gale). http://dx.doi.org/10.1016/j.eplepsyres.2015.06.011 0920-1211/© 2015 Elsevier B.V. All rights reserved.
decisions (Dworetzky et al., 2005; Oto et al., 2005). In addition, less is known regarding symptoms, including self-reported mood and personality characteristics, in males with PNES. The importance of specifically investigating males with PNES was recently suggested in a study which found it more difficult to identify probable PNES in males in the clinic and more difficult to confirm the diagnosis with vEEG (Noe et al., 2012). Moreover, males with PNES may have slightly different psychiatric symptoms than females including endorsement of different depressive symptoms (Asmussen et al., 2009) or poorer overall emotional adjustment (Holmes et al., 2001). In contrast, the semiology of the events does not seem to differ between males and females even though predictors of PNES and patient characteristics may be dissimilar (Dworetzky et al., 2005; Oto et al., 2005; Thomas et al., 2013). Although we previously demonstrated differences in selfreported depressive symptoms in patients with either ES or PNES (Asmussen et al., 2009) the present study was carried out to extend these findings with a much larger sample of males with PNES, as well as analyze additional mood and personality characteristics in the context of seizure semiology. Furthermore, in that prior
154
S.D. Gale et al. / Epilepsy Research 115 (2015) 153–157
work has suggested that somatic symptoms are particularly useful in identifying likely cases of PNES, (Hill and Gale, 2011b) we decided to specifically focus on how these self-reported symptoms might relate to gender differences in the behavioral manifestation of PNES. The aim of the current study was to (1) investigate selfreported psychiatric symptoms and personality traits in males with PNES in comparison to males with epilepsy and females with PNES in a larger sample, (2) determine if there are any differences in seizure semiology between the PNES males and females for those patients with and without clinical elevations on psychological testing and (3) compare the semiology of the events between males and females with PNES. 2. Material and methods 2.1. Subjects With approval of the institutional review board, all patients in this study were evaluated on the inpatient epilepsy monitoring unit at St. Joseph’s Hospital and Medical Center in Phoenix, Arizona. Psychological measures were typically completed on the day of admission prior to diagnosis, which was established by the attending neurologist via vEEG and clinical history. Those without clear vEEG diagnostic examinations, incomplete testing, or mixed etiology (e.g. both PNES and epilepsy) were excluded from the study. After these exclusions there were a total of 62 males with PNES, 158 females with PNES, 95 males with ES, and 165 females with ES. We then excluded patients whose scores on any of the Personality Assessment Inventory validity scales suggested an invalid profile (Morey, 1991, 2003). Specifically, we identified invalid profiles as those with elevations on the inconsistency scale (ICN T ≥ 73), the infrequency scale (INF T ≥ 75), the positive impression management scale (PIM T ≥ 68) and the negative impression management scale (NIM T ≥ 92). Thus, our final sample included 58 males with PNES, 147 females with PNES, 86 males with ES, and 142 females with ES. Demographic information is presented in Table 1. 2.2. Measures The Psychological Assessment Inventory (PAI) is a personality measure, which has been shown to be reliable and valid in identifying psychiatric symptoms (Morey, 1991). The PAI has also been shown to be reliable and valid in patients with neurologic conditions (Busse et al., 2014). It has been found to be useful in both PNES and ES samples (Hill and Gale, 2011b; Mason et al., 2000; Thompson et al., 2010). The PAI is a self-report measure consisting of 344 items that includes 11 clinical scales, four validity scales, five treatment scales, and two interpersonal scales, which only requires a 4th grade reading level (Morey, 1991). For the present study we limited our evaluation to the following clinical scales, and their subscales, Somatic Complaints (SOM), Anxiety (ANX), Depression
(DEP), Borderline Features (BOR), and Antisocial Features (ANT). We also included two of the treatment scales, Suicidal Ideation (SUI) and Stress (STR). PAI scales and subscales were scored and reported per the manual in standardized T-scores with a mean of 50 and a standard deviation of 10 with higher scores reflecting increased psychopathology. Separate normative data are used for males and females (Morey, 1991). The Beck Depression Inventory-second edition (BDI-II) is a 21-tem self-report measure of depression based on DSM-IV criteria for depressive disorders, which has been shown to be both reliable and valid for the use of identifying individuals with depressive symptoms (Beck et al., 1996). Finally, the observed seizures in those patients with PNES were categorized into semiological subgroups based on two models. First, we utilized the four-group classification model as described by Griffith et al. (2007) which includes the following categories: subjective, catatonic, minor motor and major motor. Patients with more than one seizure type were categorized according to the most frequent type (i.e. >60%). Second, we also used a simple two-group model (motor and non-motor) which we found useful when comparing neuropsychological function in patients with PNES (Hill and Gale, 2011a). 2.3. Statistical analysis Statistical analyses were carried out with IBM SPSS 21.0 (IBM Corp., Armonk, New York). Chi-square, t-tests, and MANOVA were used to make comparisons between groups on demographic and PAI variables. In addition, a logistic regression was carried out in the males with PNES to determine if the PAI SOM scale could predict seizure semiology. 3. Results 3.1. Demographics As indicated in Table 1, there were no group differences in years of education, handedness, or ethnicity. However, the males with epilepsy were younger than the males with PNES and both males and females with epilepsy had earlier age of seizure onset than males and females with PNES resulting in more years with a seizure disorder. 3.2. Psychiatric and personality differences A MANOVA comparing all patients with PNES to all patients with epilepsy on psychometric measures was statistically significant, F(23,409) = 9.854, p < 0.001 (Table 2). Although most of the PAI scales were different between groups, it should be pointed out that a clinical elevation on the PAI is defined as a score that is ≥70 which is two standard deviations above the mean for the control sample. Therefore, only the Somatic Complaints scale and
Table 1 Demographics for groups. Demographic
PNES males (n = 58)
PNES females (n = 147)
ES males (n = 86)
ES females (n = 142)
Age Education Handedness (right) Ethnicity Caucasian Hispanic African American Other Age at onset Years of seizures
42.9 (13.6) 12.8 (2.1) 52 (89.7%)
38.8 (12.3) 13.1 (2.5) 133 (90.5%)
36.6 (12.3) 13.2 (2.2) 70 (84.3%)
38.5 (14.0) 13.6 (2.4) 123 (87.2%)
51 (91.1%) 3 (5.4%) 1 (1.8%) 1 (1.8%) 33.8 (17.6) 9.2 (13.0)
116 (85.3%) 13 (9.6%) 6 (4.4%) 1 (0.7%) 30.6 (14.8) 8.4 (10.6)
68 (88.3%) 6 (7.8%) 2 (2.6%) 1 (1.3%) 18.3 (14.4) 18.2 (12.5)
113 (85.6%) 13 (9.8%) 6 (4.5%) 0 (0.0%) 19.7 (14.6) 19.6 (15.3)
Note. Sample size is smaller for reported ethnicity due to missing data.
F/2
p
2.73 2.00 3.38 4.63
.043 .113 .759 .865
24.98 23.49
.000 .000
S.D. Gale et al. / Epilepsy Research 115 (2015) 153–157 Table 2 Personality Assessment Inventory scales and subscales and Beck Depression Inventory by group PNES vs. ES. PAI scale/subscale
PNES (n = 205)
ES (n = 228)
p
SOM (somatic complaints) SOM-C (conversion) SOM-S (somatization) SOM-H (health concerns) ANX (anxiety) ANX-C (cognitive) ANX-A (affective) ANX-P (physiological) DEP (depression) DEP-C (cognitive) DEP-A (affective) DEP-P (physiological) BOR (borderline features) BOR-A (affective instability) BOR-I (identity problems) BOR-N (negative relationships) BOR-S (self-harm) ANT (antisocial features) ANT-A (antisocial behaviors) ANT-E (egocentricity) ANT-S (stimulus seeking) STR (stress) SUI (suicide ideation) Beck depression inventory-II
74.1 (13.3) 75.2 (15.9) 68.3 (12.5) 70.4 (12.9) 59.0 (12.8) 56.6 (12.8) 57.5 (12.3) 60.8 (13.0) 63.4 (13.2) 57.8 (13.8) 58.7 (14.0) 66.6 (11.6) 56.1 (12.2) 55.9 (12.7) 55.8 (11.9) 57.7 (12.8) 48.8 (10.0) 47.7 (8.4) 49.6 (10.1) 46.4 (7.5) 47.7 (9.0) 58.4 (11.8) 52.8 (11.7) 17.9 (11.1)
62.8 (11.4) 60.5 (13.3) 56.1 (11.6) 66.1 (11.1) 54.6 (10.7) 55.0 (10.5) 53.6 (11.3) 54.0 (10.9) 55.3 (11.6) 54.2 (11.8) 52.6 (10.9) 56.3 (11.8) 52.2 (10.7) 52.4 (11.1) 52.6 (10.8) 53.4 (11.3) 47.7 (9.1) 49.5 (9.0) 50.9 (11.3) 49.0 (9.0) 48.7 (8.8) 54.0 (10.9) 49.8 (10.7) 11.8 (9.5)
0.000 0.000 0.000 0.000 0.000 0.148 0.000 0.000 0.000 0.003 0.000 0.000 0.000 0.002 0.003 0.000 0.240 0.032 0.204 0.001 0.275 0.000 0.006 0.000
Note. Mean and standard deviation (SD) are provided for each variable. PAI scores are standardized T-scores (M = 50, SD = 10) and the Beck Depression Inventory-II is a raw score. PNES, psychogenic non-epileptic seizures; ES, epileptic seizures.
2 of its subscales, conversion and health concerns, were found to be clinically elevated in the PNES group compared to the ES group. However, we also report the proportion in each group with clinically elevated PAI scales. The proportion of patients in the PNES group with a clinically elevated Somatic Complaints scale was 63.4% vs. 25.9% in the ES group, 2 (1) = 61.83, p < 0.001). A clinically elevated Anxiety scale was found in 22.4% of the PNES group vs. 7.0% in the ES group, 2 (1) = 20.92, p < 0.001). The proportion of patients in the PNES group with a clinically elevated Depression scale was 34.1% vs. 14.0% in the ES group, 2 (1) = 24.24, p < 0.001). There was a higher proportion of patients in the PNES group with a clinically elevated Borderline Features scale, 18.5% compared to the ES group, 7.5%, 2 (1) = 11.95, p < 0.01). Finally, the proportion of patients in the PNES group with a clinically elevated Antisocial Features scale was 1.5% vs. 3.9% in the ES group, 2 (1) = 2.47, p > 0.05). The BDI-II was significantly higher in the PNES group with a score that would be considered to be in the mild range. We also carried out a MANOVA comparing males with PNES to females with PNES which was also statistically significant, F(23,181) = 2.059, p < 0.01 (Table 3). Consistent with our prior study, (Asmussen et al., 2009) females with PNES had higher scores on the physiological depression scale. No other depression scales were different between groups. Male with PNES had higher scores on 2 of the Borderline features subscales, affective instability, self-harm, as well as the Antisocial Features scale and all of its subscales. Comparing males with PNES to females with PNES (Table 4) suggested no significant differences in seizure semiology between groups for those that had this information available. This information was unavailable for 6 of the males with PNES and 7 of the females. History of reported head trauma was just outside the statistical threshold (p = .06). There was no difference in marital status between the PNES groups. 3.3. Seizure semiology differences by gender and PAI Fig. 1 depicts gender differences in the proportion of patients demonstrating motor or non-motor PNES events depending on
155
Table 3 Personality Assessment Inventory scales and subscales and Beck Depression Inventory in PNES group by gender. PAI scale/subscale
PNES males (n = 58)
PNES females (n = 147)
p
SOM (somatic complaints) SOM-C (conversion) SOM-S (somatization) SOM-H (health concerns) ANX (anxiety) ANX-C (cognitive) ANX-A (affective) ANX-P (physiological) DEP (depression) DEP-C (cognitive) DEP-A (affective) DEP-P (physiological) BOR (borderline features) BOR-A (affective instability) BOR-I (identity problems) BOR-N (negative relationships) BOR-S (self-harm) ANT (antisocial features) ANT-A (antisocial behaviors) ANT-E (egocentricity) ANT-S (stimulus seeking) STR (stress) SUI (suicide ideation) Beck depression inventory
74.4 (13.3) 76.6 (15.0) 67.1 (11.9) 70.9 (13.1) 59.0 (11.8) 55.6 (12.3) 58.4 (11.8) 61.1 (12.2) 63.6 (13.8) 59.6 (14.6) 60.4 (14.5) 63.8 (11.9) 58.6 (11.7) 58.7 (11.0) 57.0 (12.2) 58.6 (12.2) 52.2 (11.5) 51.5 (9.2) 53.8 (11.1) 48.7 (8.3) 50.2 (10.0) 60.9 (11.4) 54.5 (13.5) 19.1 (11.1)
74.0 (13.3) 74.6 (16.3) 68.7 (12.7) 70.1 (12.9) 59.0 (13.2) 57.0 (13.0) 57.2 (12.5) 60.7 (13.3) 63.3 (13.0) 57.1 (13.5) 58.0 (13.7) 67.6 (11.4) 55.1 (12.3) 54.8 (13.2) 55.4 (11.7) 57.3 (13.0) 47.5 (9.1) 46.2 (7.6) 47.9 (9.2) 45.4 (7.0) 46.8 (8.5) 57.4 (11.9) 52.1 (10.9) 17.4 (11.1)
0.856 0.425 0.400 0.702 0.988 0.499 0.527 0.865 0.910 0.237 0.253 0.036 0.066 0.048 0.364 0.500 0.002 0.000 0.000 0.004 0.015 0.057 0.192 0.340
Note. Mean and standard deviation (SD) are provided for each variable. PAI scores are standardized T-scores (M = 50, SD = 10) and the Beck Depression Inventory is a raw score. PNES, psychogenic non-epileptic seizures.
whether or not the Somatic Complaints scale was clinically elevated on the PAI. In all, 63.8% of the males with PNES and 62.8% of the females with PNES had a clinically elevated SOM scale. As a post hoc analysis, we found a much higher proportion of males with an elevation on this scale were found to have motor (78%) rather than non-motor (22%) seizures, 2 (1) = 5.967, p = .015. In contrast, within females, there was no interaction between an elevation on the Somatic Complaints scale and seizure semiology as 41% had nonmotor and 59% had motor seizures, 2 (1) = 0.148, p = 0.701. Within PNES males only however, binary logistic regression indicated that PNES males with an elevated SOM scale were 4.365 times (95% C.I. 1.29, 14.72) more likely to have motor seizure semiology than
Table 4 Seizure semiology and psychosocial characteristics in patients with PNES by gender. PNES males Seizure semiology (4 group model) Catatonic Major motor Minor motor Subjective Seizure semiology (2 group model) Motor Non-motor Marital status (S/D/M) Family Hx of seizures History of head trauma Currently taking an AED Currently taking psychiatric Number of AEDs Total AED mg Number of psychiatric
PNES females
n (percentage) 5 (9.6%) 13 (25.0%) 21 (40.4%) 13 (25.0%)
t/2
p
3.82
.281
.894
.344
26 (18.6%) 41 (29.3%) 40 (28.6%) 33 (23.6%)
34 (65.4%) 18 (34.6%) 26/5/26 13 (22.4%) 28 (48.3%) 43 (76.8%) 26 (44.8%)
81 (57.9%) 59 (42.1%) 48/30/65 29 (20.0%) 50 (34.2%) 111 (80.4%) 73 (49.7%)
5.06 .147 3.46 .324 .389
.079 .701 .063 .569 .533
Mean (SD) 1.1 (.90) 1197.4 (1279) 0.69 (.88)
1.3 (1.1) 1199.4 (1458.2) 0.76 (.93)
1.49 .009 .472
.136 .993 .637
PNES, psychogenic non-epileptic seizures; AED, antiepileptic drug, mg = milligrams; psychiatric, psychiatric medications. Note. n varies dues to missing data.
156
S.D. Gale et al. / Epilepsy Research 115 (2015) 153–157
Fig. 1. Percentage of PNES patients with seizure semiology (motor and non-motor) by gender and clinical elevation on PAI somatic complaints scale.
those PNES males without an elevated SOM (Wald statistic = 5.642, p = 0.018). 4. Discussion The current study provides additional information regarding males with PNES. First, results did not suggest differences between male and females with PNES in self-reported psychiatric symptoms and personality characteristics in the areas of somatization, anxiety, or depression as measured by the PAI with the exception of the physiological symptoms depression subscale which was higher in females consistent with our prior study (Asmussen et al., 2009). While there were gender differences in the Antisocial Features scale and its subscales, and the self-harm subscale of the Borderline Features scale, the mean scores for both PNES groups were not in the clinically elevated range. Second, there were clear differences between the group with PNES and the group with ES on the psychometric tests. Most notably, PAI scores related to somatic complaints were much higher in the PNES group and were disproportionally elevated in comparison to other evaluated features of psychopathology. Third, consideration of group differences may also include evaluating the proportion of patients in each group with clinical elevations on the other major PAI scales analyzed here. For example, the proportion of PNES patients with a clinically elevated Borderline Features scale was higher compared to the ES group (18.5% vs. 7.5%). Similarly, the proportion of PNES patients with elevations on the Depression scale was higher than in the ES group (34.1% vs. 14.0%) as was the proportion with elevations on the Anxiety scale (22.4% vs. 7.0%). Fourth, when taking into account elevations on the Somatic Complaints scale, there tended to be a higher proportion of males with PNES presenting with motor seizures whereas females were more evenly distributed across semiological presentation. Consistent with prior studies (Dworetzky et al., 2005; Thomas et al., 2013) the proportion of patients presenting with a given seizure semiology was similar between males and females with PNES when we utilized a four-group classification model (Griffith et al., 2007). However, when we utilized a two-group classification model (motor and non-motor), which we found to be equally useful in a previous study (Hill and Gale, 2011a), and considered whether or not patients had evidence of somatization on the PAI, we found a significantly higher proportion of PNES males with motor compared to non-motor seizures. However, this association was not present in the females with PNES. van Merode et al. (1997) reported that a much higher proportion of males with PNES demonstrated
tonic-clonic (major motor) seizures compared to females (80% vs. only 45%). In contrast, Oto et al. (2005), Thomas et al. (2013) and Asadi-Pooya et al. (2013) did not observe the gender differences in semiology reported by van Merode. Our results do not necessarily conflict with these prior studies. First, our data suggests that the gender differences may only be apparent in the context of clinically elevated somatic complaints, which were not measured in the other studies. Second, we categorized patients by most frequent overall seizure semiology rather than specific seizure characteristics and we also utilized a two-group semiology classification (motor vs. non-motor). In addition, we would also point out that our study has a larger sample size than Oto et al. (2005) and Thomas et al. (2013) and is equivalent to Asadi-Pooya et al. (2013). Still, the underlying reason for this finding is unclear though our percentage of PNES males with motor seizures was 77.4%, which is very close to the proportion of 80% reported by van Merode et al. (1997). The cases used by van Merode were all from individual or small sample published case studies and thus, may have been more likely to include patients with more significant psychopathology. Therefore, since we separated males with PNES who had clinical significant somatization in order to investigate this, our sample may have been similar. In terms of the etiology, van Merode et al. (1997) speculated that males with PNES may be more likely to have motor rather than nonmotor seizures because of an increased likelihood that the attacks would be perceived as representing a “real” (medical) condition or perhaps because a motor manifestation (e.g. tonic-clonic attack) may be more representative of a “male form of acting-out” (p. 314). Others have suggested that males with PNES may have more difficulties with emotional adjustment than females with PNES or in comparison to either gender with epilepsy (Holmes et al., 2001). We did not find evidence that the males in our study had more psychopathology than the females. However, we utilized the PAI whereas the study that found gender differences in psychopathology in PNES used the MMPI and it may be that these two tests are measuring slightly different things. In fact, in a prior study which administered both the PAI and MMPI to patients with PNES it was a combination of the measures that best predicted diagnosis (i.e. PNES vs. ES) suggesting they may be capturing slightly different aspects of the associated somatization (Gale and Hill, 2012). Alternatively, the reason for this observation could possibly represent a true difference between males and females with PNES or may represent a sampling effect in that only the more maladjusted males with PNES were likely to seek medical treatment and thus present to the EMU (Holmes et al., 2001). Still, if it is true that males with PNES tend to have more maladjustment than females with PNES,
S.D. Gale et al. / Epilepsy Research 115 (2015) 153–157
then it is possible that the expression of this emotional maladjustment may differ based on gender. Furthermore, Kaplan et al. (2013) found an association between alexithymia and conversion disorder in PNES and suggested that the emotion that has not been properly exhibited because of alexithymia may present physically, as in resulting in PNES. There is prior work suggesting higher levels of alexithymia in males compared to females (Levant et al., 2009) though whether this might be related to male differences in semiology when somatization is present is uncertain. Future studies may wish to analyze seizure semiology in the context of psychological symptoms. Limitations of the current study are those associated with crosssectional studies. Thus, though there was an association between seizure semiology and presence or absence of somatic complaints in males, the direction of this association cannot be determined. Other limitations include the fact that the patients in this study may be affected by selection bias in that we could only include those that completed the psychological testing and had their stereotypic events captured while on the EMU. Noe et al. (2012) found that in those patients predicted to demonstrate PNES with vEEG on an EMU, failure to capture events was 39% in probable PNES males but only 12% in probable PNES females. Furthermore, there may be a difference between those patients, particularly males with PNES, which are evaluated on an EMU compared to those that are not. Another limitation is the fact that the finding of differences in seizure type between males and females with PNES based on SOM score was a post hoc analysis and increases the possibility that it was a spurious or false positive finding. Finally, though our sample was larger than most in terms of the number of males with PNES, it was still much smaller compared to our sample of females with PNES. Thus, the findings from this study will need to be replicated. In conclusion, we describe a sample of males with PNES evaluated on an EMU that suggested an increased likelihood of motor seizure semiology when the males with PNES had abnormal scores on the somatic complaints subscale of the PAI. Thus, while we confirmed that males with PNES compared to males with ES have elevated somatization on psychological testing as shown previously, we suggest that the presence of self-reported somatization is associated with a more dramatic seizure presentation which may have implications for diagnosis and treatment. Further, although the proportion of clinically elevated psychopathology was greater in PNES compared to the epilepsy group, there was no significant difference in the rate of psychopathology between PNES gender groups, which is in contrast to prior evidence suggesting greater psychopathology in PNES males compared to PNES females. Future studies may wish to determine if potential gender differences in the interaction between self-reported symptoms of psychopathology and their relationship to the behavioral manifestation of PNES, specifically seizure semiology, may influence diagnostic or treatment decisions.
157
References Asadi-Pooya, A.A., Emami, M., Emami, Y., 2013. Gender differences in manifestations of psychogenic non-epileptic seizures in Iran. J. Neurol. Sci. 332, 66–68. Asmussen, S.B., Kirlin, K.A., Gale, S.D., Chung, S.S., 2009. Differences in self-reported depressive symptoms between patients with epileptic and psychogenic nonepileptic seizures. Seizure 18, 564–566. Beck, A.T., Steer, R.A., Brown, G.K., 1996. BDI-II Manual. The Psychological Corporation, San Antonio. Binder, L.M., Salinsky, M.C., 2007. Psychogenic nonepileptic seizures. Neuropsychol. Rev. 17, 405–412. Busse, M., Whiteside, D., Waters, D., Hellings, J., Ji, P., 2014. Exploring the reliability and component structure of the personality assessment inventory in a neuropsychological sample. Clin. Neuropsychol. 28, 237–251. Dickinson, P., Looper, K.J., 2012. Psychogenic nonepileptic seizures: a current overview. Epilepsia 53, 1679–1689. Dworetzky, B.A., Strahonja-Packard, A., Shanahan, C.W., Paz, J., Schauble, B., Bromfield, E.B., 2005. Characteristics of male veterans with psychogenic nonepileptic seizures. Epilepsia 46, 1418–1422. Gale, S.D., Hill, S.W., 2012. Concurrent administration of the MMPI-2 and PAI in a sample of patients with epileptic or non-epileptic seizures: implications for an inpatient epilepsy monitoring unit. Epilepsy Behav. 25, 181–184. Griffith, N.M., Szaflarski, J.P., Schefft, B.K., Isaradisaikul, D., Meckler, J.M., McNally, K.A., Privitera, M.D., 2007. Relationship between semiology of psychogenic nonepileptic seizures and Minnesota Multiphasic Personality Inventory profile. Epilepsy Behav. 11, 105–111. Hill, S.W., Gale, S.D., 2011a. Neuropsychological characteristics of nonepileptic seizure semiological subgroups. Epilepsy Behav. 22, 255–260. Hill, S.W., Gale, S.D., 2011b. Predicting psychogenic nonepileptic seizures with the Personality Assessment Inventory and seizure variables. Epilepsy Behav. 22, 505–510. Holmes, M.D., Dodrill, C.B., Bachtler, S., Wilensky, A.J., Ojemann, L.M., Miller, J.W., 2001. Evidence that emotional maladjustment is worse in men than in women with psychogenic nonepileptic seizures. Epilepsy Behav. 2, 568–573. Kaplan, M.J., Dwivedi, A.K., Privitera, M.D., Isaacs, K., Hughes, C., Bowman, M., 2013. Comparisons of childhood trauma, alexithymia, and defensive styles in patients with psychogenic non-epileptic seizures vs. epilepsy: implications for the etiology of conversion disorder. J. Psychosom. Res. 75, 142–146. Levant, R.F., Hall, R.J., Williams, C.M., Hasan, N.T., 2009. Gender differences in alexithymia. Psychol. Men Masc. 10, 190–203. Mason, S.L., Doss, R.C., Gates, J.R., 2000. Clinical utility of the Personality Assessment Inventory in the diagnosis of psychogenic non-epileptic seizures. Epilepsia 41, 156. Morey, L., 1991. The Personality Assessment Inventory. Psychological Assessment Resource, Odessa, FL. Morey, L., 2003. Essentials of PAI Assessment. John Wiley & Sons, Hoboken, NJ. Noe, K.H., Grade, M., Stonnington, C.M., Driver-Dunckley, E., Locke, D.E., 2012. Confirming psychogenic nonepileptic seizures with video-EEG: sex matters. Epilepsy Behav. 23, 220–223. Oto, M., Conway, P., McGonigal, A., Russell, A.J., Duncan, R., 2005. Gender differences in psychogenic non-epileptic seizures. Seizure 14, 33–39. Thomas, A.A., Preston, J., Scott, R.C., Bujarski, K.A., 2013. Diagnosis of probable psychogenic nonepileptic seizures in the outpatient clinic: does gender matter? Epilepsy Behav. 29, 295–297. Thompson, A.W., Hantke, N., Phatak, V., Chaytor, N., 2010. The Personality Assessment Inventory as a tool for diagnosing psychogenic nonepileptic seizures. Epilepsia 51, 161–164. van Merode, T., de Krom, M.C., Knottnerus, J.A., 1997. Gender-related differences in non-epileptic attacks: a study of patients’ cases in the literature. Seizure 6, 311–316. Walker, J.A., 2000. Use of neuropsychological testing to differentiate neurologic from non-neurologic disorders. In: Gates, J.R., Rowan, A.J. (Eds.), Non-epileptic Seizures. , second ed. Butterworth-Heinemann, Boston, pp. 115–122.
Contents lists available at www.sciencedirect.com
Epilepsy Research journal homepage: www.elsevier.com/locate/epilepsyres
Seizure semiology in males with psychogenic nonepileptic seizures is associated with somatic complaints Shawn D. Gale a,∗ , Stacy W. Hill b , Caleb Pearson c a b c
Department of Psychology and Neuroscience Center, Brigham Young University, Provo, UT, USA Clearwater Neurosciences, Lewiston, ID, USA Center for Neuropsychology and Cognitive Neuroscience, University of Kansas Hospital, KS, USA
a r t i c l e
i n f o
Article history: Received 9 February 2015 Received in revised form 10 June 2015 Accepted 16 June 2015 Available online 21 June 2015 Keywords: Non-epileptic Somatization Psychogenic Gender differences
a b s t r a c t Aims: Psychopathology has been studied in patients with epileptic or psychogenic non-epileptic seizures in the context of diagnosis and treatment. Unfortunately, most PNES studies include few males and do not consider possible gender differences, making findings less generalizable to males with PNES. In this study we specifically compare males with PNES to females with PNES and to males with epilepsy. Methods: Males with PNES (n = 58), males with epilepsy (n = 86), females with PNES (n = 147), and females with ES (n = 142) were evaluated on an inpatient epilepsy monitoring unit. Self-reported objective measures of psychopathology, demographics, and PNES seizure semiology were compared. Results: Personality Assessment Inventory profiles revealed marked differences, particularly in somatic symptoms, between PNES and epilepsy. Females with PNES had higher levels of physiological depressive symptoms but lower antisocial features. Males with PNES who had clinically significant elevations on the somatic complaints scale were much more likely to have motor seizures while females with PNES classified similarly were equally likely to have either motor or non-motor events. Conclusion: Gender difference in PNES seizure semiology was associated with whether or not clinically significant somatic symptoms were present; males with elevated somatic symptoms were much more likely to have motor PNES. However, we did not find evidence of greater psychopathology in males with PNES compared to females with PNES. Gender differences in the behavioral manifestation of PNES in the context of presence or absence of somatization may have implications for diagnosis and treatment. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Psychogenic non-epileptic seizures (PNES) are paroxysmal events that mimic epileptic seizures (ES) but have psychological rather than neurological underpinnings and may be treatable with behavioral therapies (Dickinson and Looper, 2012). Although a careful clinical history and consideration of the behavioral components of the events (i.e. semiology) may increase confidence of a PNES diagnosis, video electroencephalography (vEEG) is the “gold standard” (Thomas et al., 2013). Given the complexity of this condition, psychological and neuropsychological testing may be coupled with vEEG (Dickinson and Looper, 2012; Walker, 2000). Given that the prevalence rate of PNES is three times higher in females than males, (Binder and Salinsky, 2007) most studies consist mainly of females, which could potentially affect diagnostic and treatment
∗ Corresponding author at: Department of Psychology and Neuroscience Center, Brigham Young University, 1060 SWKT, Provo, UT 84602, USA. E-mail address: shawn [email protected] (S.D. Gale). http://dx.doi.org/10.1016/j.eplepsyres.2015.06.011 0920-1211/© 2015 Elsevier B.V. All rights reserved.
decisions (Dworetzky et al., 2005; Oto et al., 2005). In addition, less is known regarding symptoms, including self-reported mood and personality characteristics, in males with PNES. The importance of specifically investigating males with PNES was recently suggested in a study which found it more difficult to identify probable PNES in males in the clinic and more difficult to confirm the diagnosis with vEEG (Noe et al., 2012). Moreover, males with PNES may have slightly different psychiatric symptoms than females including endorsement of different depressive symptoms (Asmussen et al., 2009) or poorer overall emotional adjustment (Holmes et al., 2001). In contrast, the semiology of the events does not seem to differ between males and females even though predictors of PNES and patient characteristics may be dissimilar (Dworetzky et al., 2005; Oto et al., 2005; Thomas et al., 2013). Although we previously demonstrated differences in selfreported depressive symptoms in patients with either ES or PNES (Asmussen et al., 2009) the present study was carried out to extend these findings with a much larger sample of males with PNES, as well as analyze additional mood and personality characteristics in the context of seizure semiology. Furthermore, in that prior
154
S.D. Gale et al. / Epilepsy Research 115 (2015) 153–157
work has suggested that somatic symptoms are particularly useful in identifying likely cases of PNES, (Hill and Gale, 2011b) we decided to specifically focus on how these self-reported symptoms might relate to gender differences in the behavioral manifestation of PNES. The aim of the current study was to (1) investigate selfreported psychiatric symptoms and personality traits in males with PNES in comparison to males with epilepsy and females with PNES in a larger sample, (2) determine if there are any differences in seizure semiology between the PNES males and females for those patients with and without clinical elevations on psychological testing and (3) compare the semiology of the events between males and females with PNES. 2. Material and methods 2.1. Subjects With approval of the institutional review board, all patients in this study were evaluated on the inpatient epilepsy monitoring unit at St. Joseph’s Hospital and Medical Center in Phoenix, Arizona. Psychological measures were typically completed on the day of admission prior to diagnosis, which was established by the attending neurologist via vEEG and clinical history. Those without clear vEEG diagnostic examinations, incomplete testing, or mixed etiology (e.g. both PNES and epilepsy) were excluded from the study. After these exclusions there were a total of 62 males with PNES, 158 females with PNES, 95 males with ES, and 165 females with ES. We then excluded patients whose scores on any of the Personality Assessment Inventory validity scales suggested an invalid profile (Morey, 1991, 2003). Specifically, we identified invalid profiles as those with elevations on the inconsistency scale (ICN T ≥ 73), the infrequency scale (INF T ≥ 75), the positive impression management scale (PIM T ≥ 68) and the negative impression management scale (NIM T ≥ 92). Thus, our final sample included 58 males with PNES, 147 females with PNES, 86 males with ES, and 142 females with ES. Demographic information is presented in Table 1. 2.2. Measures The Psychological Assessment Inventory (PAI) is a personality measure, which has been shown to be reliable and valid in identifying psychiatric symptoms (Morey, 1991). The PAI has also been shown to be reliable and valid in patients with neurologic conditions (Busse et al., 2014). It has been found to be useful in both PNES and ES samples (Hill and Gale, 2011b; Mason et al., 2000; Thompson et al., 2010). The PAI is a self-report measure consisting of 344 items that includes 11 clinical scales, four validity scales, five treatment scales, and two interpersonal scales, which only requires a 4th grade reading level (Morey, 1991). For the present study we limited our evaluation to the following clinical scales, and their subscales, Somatic Complaints (SOM), Anxiety (ANX), Depression
(DEP), Borderline Features (BOR), and Antisocial Features (ANT). We also included two of the treatment scales, Suicidal Ideation (SUI) and Stress (STR). PAI scales and subscales were scored and reported per the manual in standardized T-scores with a mean of 50 and a standard deviation of 10 with higher scores reflecting increased psychopathology. Separate normative data are used for males and females (Morey, 1991). The Beck Depression Inventory-second edition (BDI-II) is a 21-tem self-report measure of depression based on DSM-IV criteria for depressive disorders, which has been shown to be both reliable and valid for the use of identifying individuals with depressive symptoms (Beck et al., 1996). Finally, the observed seizures in those patients with PNES were categorized into semiological subgroups based on two models. First, we utilized the four-group classification model as described by Griffith et al. (2007) which includes the following categories: subjective, catatonic, minor motor and major motor. Patients with more than one seizure type were categorized according to the most frequent type (i.e. >60%). Second, we also used a simple two-group model (motor and non-motor) which we found useful when comparing neuropsychological function in patients with PNES (Hill and Gale, 2011a). 2.3. Statistical analysis Statistical analyses were carried out with IBM SPSS 21.0 (IBM Corp., Armonk, New York). Chi-square, t-tests, and MANOVA were used to make comparisons between groups on demographic and PAI variables. In addition, a logistic regression was carried out in the males with PNES to determine if the PAI SOM scale could predict seizure semiology. 3. Results 3.1. Demographics As indicated in Table 1, there were no group differences in years of education, handedness, or ethnicity. However, the males with epilepsy were younger than the males with PNES and both males and females with epilepsy had earlier age of seizure onset than males and females with PNES resulting in more years with a seizure disorder. 3.2. Psychiatric and personality differences A MANOVA comparing all patients with PNES to all patients with epilepsy on psychometric measures was statistically significant, F(23,409) = 9.854, p < 0.001 (Table 2). Although most of the PAI scales were different between groups, it should be pointed out that a clinical elevation on the PAI is defined as a score that is ≥70 which is two standard deviations above the mean for the control sample. Therefore, only the Somatic Complaints scale and
Table 1 Demographics for groups. Demographic
PNES males (n = 58)
PNES females (n = 147)
ES males (n = 86)
ES females (n = 142)
Age Education Handedness (right) Ethnicity Caucasian Hispanic African American Other Age at onset Years of seizures
42.9 (13.6) 12.8 (2.1) 52 (89.7%)
38.8 (12.3) 13.1 (2.5) 133 (90.5%)
36.6 (12.3) 13.2 (2.2) 70 (84.3%)
38.5 (14.0) 13.6 (2.4) 123 (87.2%)
51 (91.1%) 3 (5.4%) 1 (1.8%) 1 (1.8%) 33.8 (17.6) 9.2 (13.0)
116 (85.3%) 13 (9.6%) 6 (4.4%) 1 (0.7%) 30.6 (14.8) 8.4 (10.6)
68 (88.3%) 6 (7.8%) 2 (2.6%) 1 (1.3%) 18.3 (14.4) 18.2 (12.5)
113 (85.6%) 13 (9.8%) 6 (4.5%) 0 (0.0%) 19.7 (14.6) 19.6 (15.3)
Note. Sample size is smaller for reported ethnicity due to missing data.
F/2
p
2.73 2.00 3.38 4.63
.043 .113 .759 .865
24.98 23.49
.000 .000
S.D. Gale et al. / Epilepsy Research 115 (2015) 153–157 Table 2 Personality Assessment Inventory scales and subscales and Beck Depression Inventory by group PNES vs. ES. PAI scale/subscale
PNES (n = 205)
ES (n = 228)
p
SOM (somatic complaints) SOM-C (conversion) SOM-S (somatization) SOM-H (health concerns) ANX (anxiety) ANX-C (cognitive) ANX-A (affective) ANX-P (physiological) DEP (depression) DEP-C (cognitive) DEP-A (affective) DEP-P (physiological) BOR (borderline features) BOR-A (affective instability) BOR-I (identity problems) BOR-N (negative relationships) BOR-S (self-harm) ANT (antisocial features) ANT-A (antisocial behaviors) ANT-E (egocentricity) ANT-S (stimulus seeking) STR (stress) SUI (suicide ideation) Beck depression inventory-II
74.1 (13.3) 75.2 (15.9) 68.3 (12.5) 70.4 (12.9) 59.0 (12.8) 56.6 (12.8) 57.5 (12.3) 60.8 (13.0) 63.4 (13.2) 57.8 (13.8) 58.7 (14.0) 66.6 (11.6) 56.1 (12.2) 55.9 (12.7) 55.8 (11.9) 57.7 (12.8) 48.8 (10.0) 47.7 (8.4) 49.6 (10.1) 46.4 (7.5) 47.7 (9.0) 58.4 (11.8) 52.8 (11.7) 17.9 (11.1)
62.8 (11.4) 60.5 (13.3) 56.1 (11.6) 66.1 (11.1) 54.6 (10.7) 55.0 (10.5) 53.6 (11.3) 54.0 (10.9) 55.3 (11.6) 54.2 (11.8) 52.6 (10.9) 56.3 (11.8) 52.2 (10.7) 52.4 (11.1) 52.6 (10.8) 53.4 (11.3) 47.7 (9.1) 49.5 (9.0) 50.9 (11.3) 49.0 (9.0) 48.7 (8.8) 54.0 (10.9) 49.8 (10.7) 11.8 (9.5)
0.000 0.000 0.000 0.000 0.000 0.148 0.000 0.000 0.000 0.003 0.000 0.000 0.000 0.002 0.003 0.000 0.240 0.032 0.204 0.001 0.275 0.000 0.006 0.000
Note. Mean and standard deviation (SD) are provided for each variable. PAI scores are standardized T-scores (M = 50, SD = 10) and the Beck Depression Inventory-II is a raw score. PNES, psychogenic non-epileptic seizures; ES, epileptic seizures.
2 of its subscales, conversion and health concerns, were found to be clinically elevated in the PNES group compared to the ES group. However, we also report the proportion in each group with clinically elevated PAI scales. The proportion of patients in the PNES group with a clinically elevated Somatic Complaints scale was 63.4% vs. 25.9% in the ES group, 2 (1) = 61.83, p < 0.001). A clinically elevated Anxiety scale was found in 22.4% of the PNES group vs. 7.0% in the ES group, 2 (1) = 20.92, p < 0.001). The proportion of patients in the PNES group with a clinically elevated Depression scale was 34.1% vs. 14.0% in the ES group, 2 (1) = 24.24, p < 0.001). There was a higher proportion of patients in the PNES group with a clinically elevated Borderline Features scale, 18.5% compared to the ES group, 7.5%, 2 (1) = 11.95, p < 0.01). Finally, the proportion of patients in the PNES group with a clinically elevated Antisocial Features scale was 1.5% vs. 3.9% in the ES group, 2 (1) = 2.47, p > 0.05). The BDI-II was significantly higher in the PNES group with a score that would be considered to be in the mild range. We also carried out a MANOVA comparing males with PNES to females with PNES which was also statistically significant, F(23,181) = 2.059, p < 0.01 (Table 3). Consistent with our prior study, (Asmussen et al., 2009) females with PNES had higher scores on the physiological depression scale. No other depression scales were different between groups. Male with PNES had higher scores on 2 of the Borderline features subscales, affective instability, self-harm, as well as the Antisocial Features scale and all of its subscales. Comparing males with PNES to females with PNES (Table 4) suggested no significant differences in seizure semiology between groups for those that had this information available. This information was unavailable for 6 of the males with PNES and 7 of the females. History of reported head trauma was just outside the statistical threshold (p = .06). There was no difference in marital status between the PNES groups. 3.3. Seizure semiology differences by gender and PAI Fig. 1 depicts gender differences in the proportion of patients demonstrating motor or non-motor PNES events depending on
155
Table 3 Personality Assessment Inventory scales and subscales and Beck Depression Inventory in PNES group by gender. PAI scale/subscale
PNES males (n = 58)
PNES females (n = 147)
p
SOM (somatic complaints) SOM-C (conversion) SOM-S (somatization) SOM-H (health concerns) ANX (anxiety) ANX-C (cognitive) ANX-A (affective) ANX-P (physiological) DEP (depression) DEP-C (cognitive) DEP-A (affective) DEP-P (physiological) BOR (borderline features) BOR-A (affective instability) BOR-I (identity problems) BOR-N (negative relationships) BOR-S (self-harm) ANT (antisocial features) ANT-A (antisocial behaviors) ANT-E (egocentricity) ANT-S (stimulus seeking) STR (stress) SUI (suicide ideation) Beck depression inventory
74.4 (13.3) 76.6 (15.0) 67.1 (11.9) 70.9 (13.1) 59.0 (11.8) 55.6 (12.3) 58.4 (11.8) 61.1 (12.2) 63.6 (13.8) 59.6 (14.6) 60.4 (14.5) 63.8 (11.9) 58.6 (11.7) 58.7 (11.0) 57.0 (12.2) 58.6 (12.2) 52.2 (11.5) 51.5 (9.2) 53.8 (11.1) 48.7 (8.3) 50.2 (10.0) 60.9 (11.4) 54.5 (13.5) 19.1 (11.1)
74.0 (13.3) 74.6 (16.3) 68.7 (12.7) 70.1 (12.9) 59.0 (13.2) 57.0 (13.0) 57.2 (12.5) 60.7 (13.3) 63.3 (13.0) 57.1 (13.5) 58.0 (13.7) 67.6 (11.4) 55.1 (12.3) 54.8 (13.2) 55.4 (11.7) 57.3 (13.0) 47.5 (9.1) 46.2 (7.6) 47.9 (9.2) 45.4 (7.0) 46.8 (8.5) 57.4 (11.9) 52.1 (10.9) 17.4 (11.1)
0.856 0.425 0.400 0.702 0.988 0.499 0.527 0.865 0.910 0.237 0.253 0.036 0.066 0.048 0.364 0.500 0.002 0.000 0.000 0.004 0.015 0.057 0.192 0.340
Note. Mean and standard deviation (SD) are provided for each variable. PAI scores are standardized T-scores (M = 50, SD = 10) and the Beck Depression Inventory is a raw score. PNES, psychogenic non-epileptic seizures.
whether or not the Somatic Complaints scale was clinically elevated on the PAI. In all, 63.8% of the males with PNES and 62.8% of the females with PNES had a clinically elevated SOM scale. As a post hoc analysis, we found a much higher proportion of males with an elevation on this scale were found to have motor (78%) rather than non-motor (22%) seizures, 2 (1) = 5.967, p = .015. In contrast, within females, there was no interaction between an elevation on the Somatic Complaints scale and seizure semiology as 41% had nonmotor and 59% had motor seizures, 2 (1) = 0.148, p = 0.701. Within PNES males only however, binary logistic regression indicated that PNES males with an elevated SOM scale were 4.365 times (95% C.I. 1.29, 14.72) more likely to have motor seizure semiology than
Table 4 Seizure semiology and psychosocial characteristics in patients with PNES by gender. PNES males Seizure semiology (4 group model) Catatonic Major motor Minor motor Subjective Seizure semiology (2 group model) Motor Non-motor Marital status (S/D/M) Family Hx of seizures History of head trauma Currently taking an AED Currently taking psychiatric Number of AEDs Total AED mg Number of psychiatric
PNES females
n (percentage) 5 (9.6%) 13 (25.0%) 21 (40.4%) 13 (25.0%)
t/2
p
3.82
.281
.894
.344
26 (18.6%) 41 (29.3%) 40 (28.6%) 33 (23.6%)
34 (65.4%) 18 (34.6%) 26/5/26 13 (22.4%) 28 (48.3%) 43 (76.8%) 26 (44.8%)
81 (57.9%) 59 (42.1%) 48/30/65 29 (20.0%) 50 (34.2%) 111 (80.4%) 73 (49.7%)
5.06 .147 3.46 .324 .389
.079 .701 .063 .569 .533
Mean (SD) 1.1 (.90) 1197.4 (1279) 0.69 (.88)
1.3 (1.1) 1199.4 (1458.2) 0.76 (.93)
1.49 .009 .472
.136 .993 .637
PNES, psychogenic non-epileptic seizures; AED, antiepileptic drug, mg = milligrams; psychiatric, psychiatric medications. Note. n varies dues to missing data.
156
S.D. Gale et al. / Epilepsy Research 115 (2015) 153–157
Fig. 1. Percentage of PNES patients with seizure semiology (motor and non-motor) by gender and clinical elevation on PAI somatic complaints scale.
those PNES males without an elevated SOM (Wald statistic = 5.642, p = 0.018). 4. Discussion The current study provides additional information regarding males with PNES. First, results did not suggest differences between male and females with PNES in self-reported psychiatric symptoms and personality characteristics in the areas of somatization, anxiety, or depression as measured by the PAI with the exception of the physiological symptoms depression subscale which was higher in females consistent with our prior study (Asmussen et al., 2009). While there were gender differences in the Antisocial Features scale and its subscales, and the self-harm subscale of the Borderline Features scale, the mean scores for both PNES groups were not in the clinically elevated range. Second, there were clear differences between the group with PNES and the group with ES on the psychometric tests. Most notably, PAI scores related to somatic complaints were much higher in the PNES group and were disproportionally elevated in comparison to other evaluated features of psychopathology. Third, consideration of group differences may also include evaluating the proportion of patients in each group with clinical elevations on the other major PAI scales analyzed here. For example, the proportion of PNES patients with a clinically elevated Borderline Features scale was higher compared to the ES group (18.5% vs. 7.5%). Similarly, the proportion of PNES patients with elevations on the Depression scale was higher than in the ES group (34.1% vs. 14.0%) as was the proportion with elevations on the Anxiety scale (22.4% vs. 7.0%). Fourth, when taking into account elevations on the Somatic Complaints scale, there tended to be a higher proportion of males with PNES presenting with motor seizures whereas females were more evenly distributed across semiological presentation. Consistent with prior studies (Dworetzky et al., 2005; Thomas et al., 2013) the proportion of patients presenting with a given seizure semiology was similar between males and females with PNES when we utilized a four-group classification model (Griffith et al., 2007). However, when we utilized a two-group classification model (motor and non-motor), which we found to be equally useful in a previous study (Hill and Gale, 2011a), and considered whether or not patients had evidence of somatization on the PAI, we found a significantly higher proportion of PNES males with motor compared to non-motor seizures. However, this association was not present in the females with PNES. van Merode et al. (1997) reported that a much higher proportion of males with PNES demonstrated
tonic-clonic (major motor) seizures compared to females (80% vs. only 45%). In contrast, Oto et al. (2005), Thomas et al. (2013) and Asadi-Pooya et al. (2013) did not observe the gender differences in semiology reported by van Merode. Our results do not necessarily conflict with these prior studies. First, our data suggests that the gender differences may only be apparent in the context of clinically elevated somatic complaints, which were not measured in the other studies. Second, we categorized patients by most frequent overall seizure semiology rather than specific seizure characteristics and we also utilized a two-group semiology classification (motor vs. non-motor). In addition, we would also point out that our study has a larger sample size than Oto et al. (2005) and Thomas et al. (2013) and is equivalent to Asadi-Pooya et al. (2013). Still, the underlying reason for this finding is unclear though our percentage of PNES males with motor seizures was 77.4%, which is very close to the proportion of 80% reported by van Merode et al. (1997). The cases used by van Merode were all from individual or small sample published case studies and thus, may have been more likely to include patients with more significant psychopathology. Therefore, since we separated males with PNES who had clinical significant somatization in order to investigate this, our sample may have been similar. In terms of the etiology, van Merode et al. (1997) speculated that males with PNES may be more likely to have motor rather than nonmotor seizures because of an increased likelihood that the attacks would be perceived as representing a “real” (medical) condition or perhaps because a motor manifestation (e.g. tonic-clonic attack) may be more representative of a “male form of acting-out” (p. 314). Others have suggested that males with PNES may have more difficulties with emotional adjustment than females with PNES or in comparison to either gender with epilepsy (Holmes et al., 2001). We did not find evidence that the males in our study had more psychopathology than the females. However, we utilized the PAI whereas the study that found gender differences in psychopathology in PNES used the MMPI and it may be that these two tests are measuring slightly different things. In fact, in a prior study which administered both the PAI and MMPI to patients with PNES it was a combination of the measures that best predicted diagnosis (i.e. PNES vs. ES) suggesting they may be capturing slightly different aspects of the associated somatization (Gale and Hill, 2012). Alternatively, the reason for this observation could possibly represent a true difference between males and females with PNES or may represent a sampling effect in that only the more maladjusted males with PNES were likely to seek medical treatment and thus present to the EMU (Holmes et al., 2001). Still, if it is true that males with PNES tend to have more maladjustment than females with PNES,
S.D. Gale et al. / Epilepsy Research 115 (2015) 153–157
then it is possible that the expression of this emotional maladjustment may differ based on gender. Furthermore, Kaplan et al. (2013) found an association between alexithymia and conversion disorder in PNES and suggested that the emotion that has not been properly exhibited because of alexithymia may present physically, as in resulting in PNES. There is prior work suggesting higher levels of alexithymia in males compared to females (Levant et al., 2009) though whether this might be related to male differences in semiology when somatization is present is uncertain. Future studies may wish to analyze seizure semiology in the context of psychological symptoms. Limitations of the current study are those associated with crosssectional studies. Thus, though there was an association between seizure semiology and presence or absence of somatic complaints in males, the direction of this association cannot be determined. Other limitations include the fact that the patients in this study may be affected by selection bias in that we could only include those that completed the psychological testing and had their stereotypic events captured while on the EMU. Noe et al. (2012) found that in those patients predicted to demonstrate PNES with vEEG on an EMU, failure to capture events was 39% in probable PNES males but only 12% in probable PNES females. Furthermore, there may be a difference between those patients, particularly males with PNES, which are evaluated on an EMU compared to those that are not. Another limitation is the fact that the finding of differences in seizure type between males and females with PNES based on SOM score was a post hoc analysis and increases the possibility that it was a spurious or false positive finding. Finally, though our sample was larger than most in terms of the number of males with PNES, it was still much smaller compared to our sample of females with PNES. Thus, the findings from this study will need to be replicated. In conclusion, we describe a sample of males with PNES evaluated on an EMU that suggested an increased likelihood of motor seizure semiology when the males with PNES had abnormal scores on the somatic complaints subscale of the PAI. Thus, while we confirmed that males with PNES compared to males with ES have elevated somatization on psychological testing as shown previously, we suggest that the presence of self-reported somatization is associated with a more dramatic seizure presentation which may have implications for diagnosis and treatment. Further, although the proportion of clinically elevated psychopathology was greater in PNES compared to the epilepsy group, there was no significant difference in the rate of psychopathology between PNES gender groups, which is in contrast to prior evidence suggesting greater psychopathology in PNES males compared to PNES females. Future studies may wish to determine if potential gender differences in the interaction between self-reported symptoms of psychopathology and their relationship to the behavioral manifestation of PNES, specifically seizure semiology, may influence diagnostic or treatment decisions.
157
References Asadi-Pooya, A.A., Emami, M., Emami, Y., 2013. Gender differences in manifestations of psychogenic non-epileptic seizures in Iran. J. Neurol. Sci. 332, 66–68. Asmussen, S.B., Kirlin, K.A., Gale, S.D., Chung, S.S., 2009. Differences in self-reported depressive symptoms between patients with epileptic and psychogenic nonepileptic seizures. Seizure 18, 564–566. Beck, A.T., Steer, R.A., Brown, G.K., 1996. BDI-II Manual. The Psychological Corporation, San Antonio. Binder, L.M., Salinsky, M.C., 2007. Psychogenic nonepileptic seizures. Neuropsychol. Rev. 17, 405–412. Busse, M., Whiteside, D., Waters, D., Hellings, J., Ji, P., 2014. Exploring the reliability and component structure of the personality assessment inventory in a neuropsychological sample. Clin. Neuropsychol. 28, 237–251. Dickinson, P., Looper, K.J., 2012. Psychogenic nonepileptic seizures: a current overview. Epilepsia 53, 1679–1689. Dworetzky, B.A., Strahonja-Packard, A., Shanahan, C.W., Paz, J., Schauble, B., Bromfield, E.B., 2005. Characteristics of male veterans with psychogenic nonepileptic seizures. Epilepsia 46, 1418–1422. Gale, S.D., Hill, S.W., 2012. Concurrent administration of the MMPI-2 and PAI in a sample of patients with epileptic or non-epileptic seizures: implications for an inpatient epilepsy monitoring unit. Epilepsy Behav. 25, 181–184. Griffith, N.M., Szaflarski, J.P., Schefft, B.K., Isaradisaikul, D., Meckler, J.M., McNally, K.A., Privitera, M.D., 2007. Relationship between semiology of psychogenic nonepileptic seizures and Minnesota Multiphasic Personality Inventory profile. Epilepsy Behav. 11, 105–111. Hill, S.W., Gale, S.D., 2011a. Neuropsychological characteristics of nonepileptic seizure semiological subgroups. Epilepsy Behav. 22, 255–260. Hill, S.W., Gale, S.D., 2011b. Predicting psychogenic nonepileptic seizures with the Personality Assessment Inventory and seizure variables. Epilepsy Behav. 22, 505–510. Holmes, M.D., Dodrill, C.B., Bachtler, S., Wilensky, A.J., Ojemann, L.M., Miller, J.W., 2001. Evidence that emotional maladjustment is worse in men than in women with psychogenic nonepileptic seizures. Epilepsy Behav. 2, 568–573. Kaplan, M.J., Dwivedi, A.K., Privitera, M.D., Isaacs, K., Hughes, C., Bowman, M., 2013. Comparisons of childhood trauma, alexithymia, and defensive styles in patients with psychogenic non-epileptic seizures vs. epilepsy: implications for the etiology of conversion disorder. J. Psychosom. Res. 75, 142–146. Levant, R.F., Hall, R.J., Williams, C.M., Hasan, N.T., 2009. Gender differences in alexithymia. Psychol. Men Masc. 10, 190–203. Mason, S.L., Doss, R.C., Gates, J.R., 2000. Clinical utility of the Personality Assessment Inventory in the diagnosis of psychogenic non-epileptic seizures. Epilepsia 41, 156. Morey, L., 1991. The Personality Assessment Inventory. Psychological Assessment Resource, Odessa, FL. Morey, L., 2003. Essentials of PAI Assessment. John Wiley & Sons, Hoboken, NJ. Noe, K.H., Grade, M., Stonnington, C.M., Driver-Dunckley, E., Locke, D.E., 2012. Confirming psychogenic nonepileptic seizures with video-EEG: sex matters. Epilepsy Behav. 23, 220–223. Oto, M., Conway, P., McGonigal, A., Russell, A.J., Duncan, R., 2005. Gender differences in psychogenic non-epileptic seizures. Seizure 14, 33–39. Thomas, A.A., Preston, J., Scott, R.C., Bujarski, K.A., 2013. Diagnosis of probable psychogenic nonepileptic seizures in the outpatient clinic: does gender matter? Epilepsy Behav. 29, 295–297. Thompson, A.W., Hantke, N., Phatak, V., Chaytor, N., 2010. The Personality Assessment Inventory as a tool for diagnosing psychogenic nonepileptic seizures. Epilepsia 51, 161–164. van Merode, T., de Krom, M.C., Knottnerus, J.A., 1997. Gender-related differences in non-epileptic attacks: a study of patients’ cases in the literature. Seizure 6, 311–316. Walker, J.A., 2000. Use of neuropsychological testing to differentiate neurologic from non-neurologic disorders. In: Gates, J.R., Rowan, A.J. (Eds.), Non-epileptic Seizures. , second ed. Butterworth-Heinemann, Boston, pp. 115–122.