Address correspondence to Dr. Marco Mula, Department of Neurology, Amedeo Avogadro University, C.so Mazzini, 18-28100 Novara, Italy. E-mail: email@example.com
Purpose: Different authors suggested the occurrence of a pleomorphic affective syndrome in patients with epilepsy named interictal dysphoric disorder (IDD). We sought to investigate whether IDD occurs only in patients with epilepsy and to validate IDD features against DSM-IV criteria.
Methods: Consecutive patients with a diagnosis of epilepsy (E) or migraine (M) have been assessed using the BDI, MDQ, and the Interictal Dysphoric Disorder Inventory (IDDI), a questionnaire specifically created to evaluate IDD symptoms. Diagnosis of current and lifetime DSM-IV Axis I disorders was established using the MINI Plus version 5.0.0.
Results: A total of 229 patients (E = 117; M = 112) were evaluated. Females were significantly more represented in the migraine group (E = 46.5% vs. M = 73.3% p = 0.009), but there was no difference in age, duration of the disease, or education level. Patients with epilepsy were more likely to screen positively at MDQ (E = 17% vs. M = 5.3% p = 0.006) and to have a diagnosis of bipolar disorder (E = 14.5% vs. M = 4.5% p = 0.013) as compared to migraine patients. There was no between-groups difference in IDD prevalence (E = 17%; M = 18.7%) and IDDI total scores (E = 4.1 ± 2.0 vs. M = 3.8 ± 2.0). Validation of IDD against DSM-IV categories showed current major depression being the foremost diagnostic category correlated with IDD in both epilepsy (OR = 0.32–0.12–0.88, p = 0.028) and migraine (OR = 0.10, 95% CI = 0.02–0.49, p = 0.004) samples. Current anxiety disorder correlated with IDD only in migraine patients (OR = 0.19, 95% CI = 0.05–0.77, p = 0.02).
Conclusion: IDD represents a homogenous construct that can be diagnosed in a relevant proportion of patients but it is not typical only of epilepsy, occurring in other central nervous system disorders such as migraine.
Despite an extensive literature indicating that mood disorders are a frequent psychiatric complication of epilepsy, reports continue to indicate that depression often goes unrecognized or untreated (Schmitz, 2005). The underdiagnosis of affective comorbidity in epilepsy is likely to have many sources: an insufficient sensitivity among epileptologists toward psychiatric symptoms and the atypical clinical presentation of depressive disorders in patients with epilepsy, which often does not allow a straightforward classification according to standardized psychiatric diagnostic systems such as the ICD-10 or the DSM-IV (American Psychiatric Association, 1994).
The issue of phenomenology of depression in epilepsy is very much a matter of debate. Mendez et al. (1986) investigated the clinical semiology of depression in 175 patients with epilepsy, and reported that 22% could be classified as having atypical features. In particular, classic “endogenous”-type depressive symptoms, including feelings of guilt, Gefühl der Gefühllosigkeit or “loss of feelings,” and a circadian pattern of symptom severity were rarely reported. Kanner et al. (2000) showed that 71% of patients with refractory epilepsy and depressive episodes, severe enough to need a psychopharmacological treatment, failed to meet criteria for any DSM-IV Axis I diagnosis. In contrast, a paper by Jones et al. (2005) showed that a current diagnosis of any mood disorder could be allocated in 24% of patients recruited in five tertiary referral centers for epilepsy in the United States, using the Mini International Neuropsychiatric Interview (MINI), clearly suggesting that it is possible to apply standardized criteria of DSM in a not negligible proportion of patients. In general terms, it is reasonable to hypothesize that patients with epilepsy can experience forms of mood disorders identical to those of patients without epilepsy. However, it is equally reasonable to assume that the underlying brain pathology can influence the expression of mood disorder symptoms making less evident some aspects or emphasizing others. Thus, the spectrum of phenomenology of mood disorders in epilepsy is likely to be large and only partly defined.
Blumer et al. (2004), referring to the classic German psychiatric literature, especially Kraepelin (1923), described a pleomorphic affective disorder in epilepsy, named “interictal dysphoric disorder” (IDD) and characterized by eight key symptoms. Among these are labile depressive symptoms (depressive mood, anergia, pain, insomnia), labile affective symptoms (fear, anxiety), and supposedly “specific” symptoms (paroxysmal irritability and euphoric moods). Blumer preferred the term dysphoria to more accurately translate the original definition of Kraepelin's Verstimmungszustand to stress the periodicity of mood changes of the patients and the presence of irritability and outbursts of aggressive behavior as key symptoms.
Himmeloch (1984), and subsequently Kanner (2003), highlighted the chronic course of this state of moderate neurotic depression with symptom-free intervals, using the term dysthymia-like.
Many clinicians believe in the existence of this dysphoric or dysthymic disorder. Unfortunately, no systematic studies proved that this empirically identified syndrome is specific for epilepsy in general or more specifically for limbic epilepsy. The identification of epilepsy-specific psychiatric disorders would be of relevance because, theoretically, they implicate therapeutic strategies tailored to individual patient needs.
The aim of this study is to characterize the clinical and psychopathological features of the IDD clarifying whether it is typical for patients with epilepsy and to identify similarities and differences with standardized criteria of DSM-IV for unipolar and bipolar depression.
This is a cross-sectional study of consecutive patients with a diagnosis of epilepsy or migraine (with or without aura) recruited in two tertiary referral centers in Europe, namely, the Department of Neurology, Amedeo Avogadro University in Novara, Italy, and the Department of Neurology, Universitätsmedizin-Charité, Berlin, Germany. We chose patients with migraine as a comparison group because migraine is a paroxysmal disorder of the central nervous system (CNS) characterized by a high comorbidity with mood disorders (Radat & Swendsen, 2005). To be enrolled, patients had to fulfill the following criteria: (1) diagnosis of epilepsy or migraine according to the ILAE (Commission on classification and terminology of the International League Against Epilepsy, 1989) or ICHD-II (Headache Classification Committee, 2004) criteria; (2) age more than 18; (3) absence of learning disabilities, a reading level more than sixth grade and/or a MMSE >24; (4) no psychotropic drug therapy or psychotherapy during last year; (5) absence of severe medical diseases or a high risk of suicide; and (6) be willing to provide a written informed consent to undergo the experimental procedures. We excluded patients on psychotropic drug therapy or psychotherapy because our study represents not only a preliminary investigation but also the first study trying to dissect the psychopathological characteristics of the IDD and different treatments could strongly influence the pheno type.
Subjects were assessed with the Beck Depression Inventory (BDI) (Beck et al., 1961), the Mood Disorder Questionnaire (MDQ) (Hirschfeld et al., 2003), and the Interictal Dysphoric Disorder Inventory (IDDI) (Mula & Trimble, in press). Diagnosis of current and lifetime DSM-IV Axis I disorders was established using the MINI Plus version 5.0.0 (Sheehan et al., 1998). The BDI is a well-known self-report instrument designated to assess and detect the severity of current depressive symptoms. Scores up to 9 are considered normal, between 10 and 18 mild depression, 19 and 29 moderate depression, and 30 and 63 severe depression (Groth-Marnat, 1990). The MDQ is a validated self-report instrument that screens for the presence of a lifetime history of bipolar disorder. An individual is scored positive if 7 or more of the 13 symptom items are endorsed and a moderate or serious degree of functional impairment is reported. The IDDI is a 38-item self-report questionnaire specifically created to investigate IDD. The eight key symptoms are evaluated in the first 32 items in terms of presence, frequency, severity, and global impairment. The time interval explored is the last 12 months. The six questions of the Appendix concern the time course of the disorder and relations of symptoms with attacks and therapy. As suggested by Blumer et al. (2004), a definite diagnosis of IDD is defined by the presence of at least three symptoms of at least “moderate” or “severe” severity and causing “moderate” or “severe” distress. Apart from the diagnosis, it is possible to obtain total and separate scores for key symptoms and “severeness” of key symptoms, defined by severity, frequency, and impairment scores. The MINI Plus 5.0.0 was developed from the MINI (Sheehan et al., 1998) as an efficient diagnostic interview, to be used in clinical as well as research settings, and follows DSM-IV and ICD-10 criteria, screening for a number of Axis I diagnoses with brief suicidality and antisocial personality modules. It has been validated in the United States and Europe and is available in several languages. All questionnaires were administered in a standardized way and in the same sequence in all patients.
The comparisons of primary importance in these analyses are between epilepsy and migraine subjects. The two groups were compared for age, gender, age at onset, and duration of the disease, DSM-IV diagnoses of affective disorders, prevalence, and clinical features of IDD, BDI, MDQ, and IDDI scores. The frequency of categorical variables was compared between groups using the chi-square analysis or Fisher's exact test. Continuous demographic and clinical variables were compared using the independent-sample t-test and the Mann–Whitney test. Correlations were tested using the bivariate two-tailed nonparametric correlation procedures. Subsequently, we evaluated correlations between a definite diagnosis of IDD and DSM-IV categories, assessed with the MINI Plus, using a backward stepwise logistic regression model. After the significant likelihood ratio test for the full model, a backward selection method was used to reduce the number of variables. A variable was eliminated if its removal statistic had a probability ≥0.10. Because of the exploratory nature of the study, the probability level was set at p = 0.05. All statistical analyses were two-sided and conducted using the Statistical Package for Social Sciences (Version 12 for Windows, SPSS Inc., Chicago, IL, U.S.A.).
A total of 229 subjects were enrolled in the present study, 117 with epilepsy and 112 with migraine. There was no between-groups difference in the two centers (Berlin and Novara) in terms of age, gender, duration of the disease, marital status, education level, employment, distribution of epilepsy/migraine diagnoses, or pattern and frequency of seizures or migraine attacks.
The demographic and clinical characteristics of the sample are presented in Table 1. Females were significantly more represented in the migraine group, but there was no difference in age, duration of the disease, or education level. As shown in Table 1, a greater part of patients (72.6%) had a diagnosis of partial epilepsy, 66.4% of whom had temporal lobe epilepsy while 27.4% had a diagnosis of generalized epilepsy. Surgery cases represented less than 5% of the total sample. The majority of patients (67.5%) were on a monotherapy regime, while 21.4% on dualtherapy, 8.5% in a polytherapy with three or more AEDs, and 2.6% were free of medications. As part of the routine assessment, all patients with epilepsy underwent magnetic resonance imaging (MRI) scans that, in the majority of cases (76.5%), were normal. In the migraine sample, 75% of patients had MRI scans that, in large proportion (84.5%), were normal.
Table 1. Demographic and clinical data of the population investigated
Epilepsy (n = 117)
Migraine (n = 112)
aFisher's two-sided exact test p = 0.009.
bChi-square = 15.422, df = 4, p = 0.004.
cChi-square = 22.413, df = 2, p < 0.001.
Age, years (SD)
Epilepsy syndrome (%)
Migraine diagnosis (%)
Migraine without aura
Migraine with aura
Age at onset of epilepsy/migraine (SD), years
Duration of the disease (SD), years
Frequency of seizures/attacks (%)
Less than 10/years
Previous psychiatric history (%)
Previous psychiatric history (%)
Two or more psychiatric comorbidities
Patients with epilepsy were more likely to refer or present documentation of a previous psychiatric history of mood disorders (E = 29.6% vs. M = 11.9%, p = 0.004) and a more complex psychopathology with several comorbidities (E = 22.6% vs. M = 1.8%, p < 0.001) (Table 1).
IDDI total and partial (labile affective symptoms, labile depressive symptoms, and specific symptoms) scores showed strong or very strong correlations among themselves in both groups, supporting the validity of the IDD as a construct (Table 2). IDDI scores displayed a good correlation with BDI, especially IDDI partial scores for depressive symptoms (E = 0.636; M = 0.664). Conversely, correlations with MDQ were weak (Table 2). However, BDI displayed a low sensitivity (69% in epilepsy and 81.0% in migraine) and low specificity (65.9% in epilepsy and 82.4% in migraine) for IDD diagnosis. Conversely, the specificity of MDQ was good (86.0% in epilepsy and 97.8% in migraine), although sensitivity was unacceptable (27.6% in epilepsy and 19.0% in migraine).
Table 2. Correlations between IDDI total and partial scores, BDI and MDQ scores (Spearman's rho correlations)
The percentage of patients with a definite diagnosis of IDD was similar in the two groups (17% in epilepsy vs. 18.7% in migraine) as were IDDI total and partial scores. In the epilepsy sample, 44.8% of patients with IDD have never had a prior psychiatric diagnosis while 45% of IDD patients had a prior psychiatry diagnosis in the migraine group.
BDI scores where similar in the two groups, while patients with epilepsy were more likely to score positively at the MDQ questionnaire (17% vs. 5.3%, p = 0.006) (Table 3). Axis I diagnoses of affective disorders were more prevalent in epilepsy than in migraine (64.1% vs. 50.0%, p = 0.033), in particular, bipolar disorder (14.5% vs. 4.5%, p = 0.013) (Table 4). A lifetime diagnosis of major depressive episode was the most frequently diagnosed mood disorder in both groups. A current major depressive episode was identified in 24.8% of patients with epilepsy and 14.3% of patients with migraine, being mild to moderate in severity (BDI = 19.0 ± 6.7 in epilepsy; BDI = 18.2 ± 6.7 in migraine).
Table 3. MDQ, BDI, IDDI total and partial scores in the population investigated
Table 4. MINI: DSM-IV Axis I diagnoses of affective disordersa
Epilepsy (n = 117)
Migraine (n = 112)
aA large number of patients had more than one diagnoses. Percentages across the Axis I disorders exceeded 100%, reflecting comorbidity.
bFisher's two-sided exact test, p = 0.033.
cFisher's two-sided exact test, p = 0.013.
MINI positive (%)
Major Depressive Episode – Current (%)
Anxiety Disorders – Current (%)
Major Depressive Episode – Lifetime (%)
Dysthymia – Lifetime (%)
Manic/Hypomanic Episode – Lifetime (%)
Social Phobia – Lifetime (%)
Generalized Anxiety Disorder – Lifetime (%)
Agoraphobia – Lifetime (%)
Panic Disorder – Lifetime (%)
There was no statistically significant difference in terms of DSM-IV categories comparing patients with a definite diagnosis of IDD in the two groups (Fig. 1). A backward stepwise logistic regression model applied to the whole population, including DSM-IV diagnoses assessed with the MINI Plus, showed a significant correlation between IDD and a diagnosis of major depressive episode lifetime (OR = 0.29, 95% CI = 0.13–0.65, p = 0.002) and current (OR = 0.30, 95% CI = 0.13–0.68, p = 0.004). Analyzing the two populations separately, IDD diagnosis correlated with major depressive episode lifetime (OR = 0.27, 95% CI = 0.10–0.76, p = 0.014) and current (OR = 0.32–0.12–0.88, p = 0.028) in the epilepsy group and with major depressive episode current (OR = 0.10, 95% CI = 0.02–0.49, p = 0.004) and anxiety disorder current (OR = 0.19, 95% CI = 0.05–0.77, p = 0.02) in the migraine group.
This is the first study investigating the prevalence and psychopathological features of the so-called IDD of epilepsy evaluating its psychopathological characteristics against standardized diagnostic criteria in psychiatry such as the DSM. Our data suggest that IDD is a robust construct and can be diagnosed in a significant proportion of patients (about 17%), being one of the most frequent mood disorders in epilepsy. Moreover, IDD seems to be largely unrecognized, considering that 44.8% of patients with epilepsy have never had any psychiatric diagnosis. Nevertheless, it seems that IDD is not typical only of patients with seizures but can be seen also in other CNS disorders such as migraine.
Our results about the prevalence of affective disorders in epilepsy are in keeping with previously published studies using standardized DSM-based structured interviews (Jones et al., 2005), reflecting the rates of Axis I affective disorders in a sample of outpatients with epilepsy followed at University-based clinics that are usually higher than those reported in nonselected populations of people with epilepsy (Tellez-Zenteno et al., 2007). In general terms, patients with epilepsy are more likely to have a previous diagnosis of mood disorders and a more complex psychiatric comorbidity when compared to patients with migraine (Table 1).
The higher rates of positive MDQ scores in epilepsy than in migraine were previously suggested by a large U.S. survey, showing similar percentages (Ettinger et al., 2005). In our study, we replicated their findings, further suggesting that bipolar disorder actually represents a significant comorbidity in epilepsy. It used to be confidently stated that bipolar disorder was rare in patients with seizure disorders (Wolf, 1982). Such statements were made prior to the use of standardized diagnostic manuals, such as the DSM, and were also based on clinical impressions rather than being assessed by the use of rating scales. Our data, along with those of other authors (Ettinger et al., 2005), raise some doubts about the previous suggestions that bipolar disorder is rare, at least in selected populations of patients with epilepsy, and raise doubts as to our knowledge of the association between these two disorders. It is, therefore, evident that more studies are warranted to clarify the clinical significance and the potential impact of this comorbidity in patients with epilepsy.
The concept of IDD, as a psychopathological construct, is an important point of our study. The stability of its clinical and psychopathological features in both epilepsy and migraine is suggested not only by the similar prevalence rates of definitive IDD diagnosis but also by the comparable total and partial IDDI scores (Table 1) and the similar spectrum of DSM-IV diagnoses for IDD patients in both groups (Fig. 1). Although these data support the existence of IDD, at the same time they reinforce the idea that IDD may arise autonomously by the epilepsy. This concept was partly suggested by Blumer himself, stating that IDD can occasionally occur in the absence of clinical seizures, in patients with brain lesions (with or without an abnormal EEG) (Blumer et al., 1988). However, we observed no associations between IDD diagnosis and MRI features in both samples. Some authors speculated that pathophysiological mechanisms are partly shared by these two neurological disorders, mainly accounting for the efficacy of anticonvulsant drugs in the prophylaxis of headache attacks (Post & Silberstein, 1994), but serial EEG recordings have not been performed in our migraine sample making impossible, therefore, to know whether those who developed IDD had underlying epileptic abnormalities.
The psychopathological definition of IDD against commonly used DSM-IV categories, assessed with the MINI, is another relevant issue. In our study, we observed major depressive episode (current and lifetime) to be correlated with IDD definitive diagnosis. However, it would be quite incorrect to state that IDD is similar to a major depressive disorder, being an extreme simplification. In fact, patients with IDD, diagnosed with the MINI as having neither a major depression nor a dysthymic disorder, were 48.2% in the epilepsy group and 35% in the migraine sample. Some features, especially the cooccurrence of mood instability and irritability, belong to the bipolar spectrum rather than to unipolar depression (Moller & Curtis, 2004; Benazzi, 2007). This is further confirmed by the higher specificity for IDD diagnosis of MDQ (86.0% in epilepsy and 97.8% in migraine), when compared to BDI (65.9% in epilepsy and 82.4% in migraine), and having the MDQ a very high specificity also for bipolar symptoms (Hirschfeld et al., 2003). In our opinion, from a clinical point of view, IDD patients have several features in common with a specific subset of cyclothymic subjects where depressive periods and labile–angry–irritable moods dominate the clinical picture, representing the more unstable form of bipolar II disorder (Akiskal & Pinto, 1999). Cyclothymia was first introduced into DSM-III and DSM-IV as a form of attenuated chronic mood disorder, and its diagnosis is not commonly made in clinical practice, because it is almost always seen when a patient presents with a major depressive episode, warranting sometimes the definition of bipolar II disorder (Akiskal, 2001). We fail to demonstrate an association with cyclothymic disorder because it is not listed among diagnoses performed with the MINI; nevertheless, Blumer reported that patients with IDD benefit from a combined therapy of anticonvulsants and antidepressants (Blumer et al., 2004), a combination that is extensively used in bipolar depression. Furthermore, bipolar disorder is characterized by a high comorbidity with anxiety disorders configuring, in some selective cases, a unique affective syndrome (Freeman et al., 2002). This concept fits well with the observed high prevalence of comorbid anxiety disorders in IDD patients (IDD + E = 51%, IDD + M = 72%). Thus, these evidence taken together further suggest that IDD may be closer to bipolar rather than unipolar mood disorders.
It has to be acknowledged that IDD diagnosis showed slight differences in the two groups (epilepsy and migraine), for what concerns the spectrum of DSM categories, namely, current anxiety disorders being significantly correlated to IDD only in the migraine sample. Although IDD patients with epilepsy did not statistically differ from those with migraine in terms of absolute DSM categories (Fig. 1), it is evident that anxiety disorders are slightly more common in the migraine group, suggesting that this may be more evident in larger samples. Moreover, it has to be taken into account that the chronic effects of the antiepileptic drug therapy, in the epilepsy population, may have masked some features.
Our results should be considered keeping in mind the following limitations. First, the small sample size reduces the statistical power of our study. However, the accurate psychiatric assessment, based on widely standardized clinical instruments, makes our results reliable. Second, our findings may not be representative for epilepsy patients in general because our population represents a highly selected sample coming from tertiary referral centers. Moreover, our sample includes different epilepsy syndromes, and further studies are warranted focusing on a specific syndrome such as temporal lobe epilepsy. Third, we decided not to correct for gender all the comparisons between epilepsy and migraine samples, considering the exploratory nature of our study that represents the first investigation of the clinical and psychopathological characteristics of IDD. In fact, it is unknown whether this syndrome is more frequent in men or women and the correction may have underestimated some aspects or, conversely, overestimated others. Fourth, it should be acknowledged that AEDs are psychoactive medications, thus limiting the “purity” of the exclusion of patients on psychoactive medications.
Finally, several clinical issues remain to be addressed in further investigations. It still needs to be clarified whether IDD is an organic affective syndrome that occurs in patients with brain disturbances or whether it can be diagnosed also in subjects with chronic medical conditions not affecting the CNS. It should be elucidated whether IDD patients are subjects with cyclothymia or an underlying cyclothymic temperament. To date, there are no studies investigating the characteristics of affective temperaments in an epilepsy population despite the high relevance of this variable for prognosis and treatment of mood disorders.
In conclusion, mood and anxiety disturbances represent a relevant comorbidity in patients with epilepsy, with bipolar disorder being more prevalent as compared to patients with migraine. IDD represents a homogenous construct with specific clinical features and affecting a relevant proportion of patients with epilepsy. However, it seems not to be specifically associated to the epilepsy, being diagnosed also in subjects with migraine. Major depression represents the foremost feature, but clinical impressions suggest that bipolar spectrum symptoms may be of relevance in IDD phenomenology, deserving further investigations on this issue.
Disclosure of conflict of interest: The authors report no conflicts of interest. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.