Acute unstable depressive syndrome (AUDS) is associated more frequently with epilepsy than major depression

Depression is more frequent in patients with epilepsy compared to the general population [1, 2]. Every second patient admitted to a tertiary epilepsy center may suffer from concurrent depressive disorders [3, 4]. Depression has greater negative impact on quality of life than seizure frequency or the number of prescribed antiepileptic drugs [4–6]. In addition, depression in epilepsy is associated with reduced seizure control [7–10] and dramatically increased suicide risk [11–14] and health care costs [15]. Recent evidence suggests that the relationship between epilepsy and depression is bidirectional [2, 16–18]. Thus, patients with epilepsy are more likely to develop depressive disorders, and patients with a history of depression or suicide attempt have a higher risk of developing epilepsy.

Interictal depressive disorders in patients with epilepsy have unique manifestations that are poorly reflected in the established classification systems ICD-10 and DSM-IV [19–24]. Symptoms such as suicidal ideation, frustration intolerance, irritability and motor agitation can rapidly alternate with symptom-free periods lasting one to several days. Blumer and co-workers have called this type of depression "an interictal dysphoric disorder with a characteristic intermittent and pleomorphic symptomatology" [23]. Kanner has termed it the "dysthymic-like disorder of epilepsy" [3]. Indeed, almost 50% of interictal depressive disorders may have to be classified as atypical depressions according to DSM-IV criteria [23]. It can be concluded that the physician must be cautious not to overlook atypical interictal mood disturbances. However, this sometimes leads to a well-known clinical dilemma in which the patient tends to attribute all mood disturbances to sub-clinical epileptic activity (and consequently demands revision of antiepileptic therapy), whereas the neurologist may not at all be confident with the patient's interpretation. Depression and many types of epilepsy share indeed several pathogenic mechanisms [7, 8]. To determine when interictal mood disturbance in a given patient is due to sub-clinical epileptic activity and when not, is often impossible.

Clinical research on the association of depression and epilepsy has for the sake of simplicity focused on depressive symptoms in defined populations with epilepsy (e.g., patients admitted to a tertiary epilepsy center). This approach has the obvious disadvantage that the study population is highly pre-selected. In order to examine epilepsy and depression from a different angle, we have therefore chosen the opposite approach in the present study. We hypothesized that it is possible to define by clinical means a subgroup of psychiatric patients with a higher than expected prevalence of epilepsy and seizures. We hypothesized further that this patient group presents to the psychiatric emergency unit with rapidly fluctuating mood symptoms that do not meet DSM-IV criteria of a Major Depressive Episode (MDE). For the purpose of the present study we have called this presentation the Acute Unstable Depressive Syndrome (AUDS). In a prospective study we have been able to show that AUDS patients indeed more often have a history of epileptic seizures [25] and abnormal quantitative EEG recordings [26] than age- and sex-matched MDE patients. The main aim of the present paper was to further classify the differences of depressive symptoms at admittance and follow-up of patients with AUDS and MDE. For this purpose we have developed a 5-item diagnostic scale to measure the prevalence and degree of atypical depressive mood symptoms. We have called this scale the Symptomatic Organic Mental disorder Assessment Scale (SOMAS).

In the five AUDS patients fulfilling clinical criteria for epilepsy, two had juvenile generalised epilepsy syndromes, both with generalized tonic-clonic (GTC) seizures and absences. One had a syndrome of posttraumatic localisation-related epilepsy with nocturnal seizures presumed to be secondarily generalized, and two had scattered GTCs with a syndromic classification that remained undetermined [25].

At clinical neurological examination, three AUDS patients and one control patient had signs of CNS pathology. One AUDS patient had ophtalmoplegia and bilaterally inverted plantar reflexes (positive Babinski's sign), whereas another had inverted plantar reflexes as the only sign of pathology. The third patient had bilateral horizontal nystagmus. In the control group, there was a unilaterally inverted plantar reflex in one patient [25].

The AUDS patients had significantly higher temporal delta band amplitude than the MDE patients. Delta asymmetry was larger in AUDS than MDE in all regions. The delta activity dominated in the right hemisphere. Twelve of 16 AUDS patients had temporal delta amplitude above 8.9 μV as compared to 3 of 16 MDE patients. Three MDE - and 11 AUDS patients had absolute temporal delta amplitude side difference above 0.1 μV [26].

There were no statistical differences between AUDS and MDE patients with respect to antiepileptic medication, benzodiazepine intake, drug abuse, alcohol consumption and alcohol withdrawal symptoms.

We have compared two groups of patients with depressive disorders admitted to a psychiatric emergency unit. AUDS patients presented with rapidly fluctuating mood disturbances, motor agitation and relative lack of insight and concern, whereas MDE patients showed rather stable mood levels, motor retardation and better understanding of and greater interest in their symptoms. Differences between AUDS and MDE patients were present until remittance of the affective episodes. This may indicate that AUDS and MDE are clinically distinct syndromes. AUDS patients had significantly more often a history of seizures and diagnose of epilepsy as well as more abnormalities on standard and quantitative EEG. Hence, atypical depressive disorders are not only frequent in patients in tertiary epilepsy centers [35], but seizures, epilepsy and EEG abnormalities are also overrepresented in AUDS patients in psychiatric emergency units.

Our pre-study observations indicated that SOMAS measures a different concept than depression. This was supported by the present study. Whereas no differences between groups were observed with MADRS (suggesting that the level of mood decrease was the same), all SOMAS items were significantly different between AUDS and MDE patients at almost any given time point. Symptom fluctuation (Item A) was the only item that was no longer different between groups at 3 months follow-up. Motor agitation and lack of insight (Item B-D) gradually diminished during the observation period, but remained rather pronounced until the end. Agitation, defined as unusual motor restlessness accompanied by emotional tension, is frequently encountered in depressive disorders [36, 37]. The present data indicate that in AUDS patients, agitation is a core symptom. The presence of agitation has clinical implications for the choice between antidepressants and antiepileptic mood stabilizers [36–39]. Antiepileptic mood stabilizers have a positive effect on agitated depression with or without organic brain disorders [36, 40], but antidepressants are sometimes detrimental in these conditions [38, 39]. Interestingly, whereas symptom fluctuation and motor agitation decreased in AUDS patients with time, lack of understanding (Item D) increased during the observation period. At 3 months follow-up AUDS patients often stated that they had given up searching for an explanation and did not want to put further concerns into it. This is perhaps in analogy to patients with epilepsy who more often than others believe that their life is controlled by external factors (fate, luck, chance) rather than internal factors (efforts, skills) [41, 42]. It can be concluded that atypical depressive symptoms cannot be classified by DSM-IV and that SOMAS provides different information than MADRAS. The advantage of SOMAS is that it can be used to evaluate and classify clinical symptoms suggestive of AUDS.

The interictal dysphoric disorder has a characteristic chronic intermittent and pleomorphic symptomatology [23]. Also the AUDS clinical syndrome features pleomorphic psychiatric symptoms, but it differs from interictal dysphoric disorder in that it has a more acute onset and a more varied symptomatology including behaviorioral problems such as aggression. Therefore, patients with AUDS require acute psychiatric care. However, episodic interictal psychiatric conditions resembling interictal dysphoric disorder with slightly different characteristics are described and termed by different authors in the same manner as the AUDS [3, 23, 43]. Core characteristics of the AUDS clinical picture has similarities to some of the symptoms in these conditions, for instance, according to Himmelhoch such patients have "brief depressive dips with impulsive suicide attempts" [43].

Clinical observations made by neurologists, psychiatrists and epileptologists of epilepsy-specific psychiatric syndromes have led to an expert consensus proposing a new system of classification of these disorders [21]. These proposals are based on data primarily generated in tertiary epilepsy centers. The present study suggests that possibly, epilepsy-specific psychiatric syndromes might also be identified in psychiatric emergency units. Obviously, much larger patient populations must be studied before such a claim can be verified. However, our data indicate that assessment of AUDS patients is a possible approach to this issue in the future.

This study has a number of limitations and thus, the present results must be interpreted with caution. Patients were recruited in daily routine clinical practice and the ward staff was not blinded to the outcome of our assessment. We included patients with substance abuse, alcohol withdrawal symptoms, antiepileptic drug treatment and benzodiazepine intake. These factors probably affected the expression of symptoms as well as EEG recordings. However, in order to obtain a more naturalistic study, we chose not to exclude these patients [44]. The small number of subjects led to relatively weak statistical power. However, nearly all SOMAS-scores reached statistical significance, which indicates a strong correlation between epileptic activity, depression and the different SOMAS items. Further, one might argue that our study was affected by a pre-selection bias. Yet as stated earlier, previous studies on the prevalence of depressive symptoms in epilepsy populations have been affected by a similar selection bias.

The study also has some significant advantages. It was a prospective study in a naturalistic patient population from a defined catchments area. All patients admitted within a three-year period were evaluated for inclusion. We have used a robust validated instrument with documented sensitivity to symptom fluctuation [27]. All patients had EEG recordings shortly after admittance. We have extensively screened for alcohol and illicit drug use and medication serum levels. Moreover, we screened for a broad range of confounding factors. Cognition, life events and life style changes were assessed, but there were no differences between the groups.

AUDS patients present with rapidly fluctuating mood symptoms, motor agitation when depressed and relative lack of insight and concern. These patients have more often a history of seizures, diagnose of epilepsy and EEG abnormalities than MDE patients. Whereas atypical depressive disorders are frequently encountered in patients in tertiary epilepsy centers [35], it seems that seizures, epilepsy and EEG abnormalities are also overrepresented in AUDS patients in psychiatric emergency units. Obviously, further studies are needed to corroborate these results. In a sentinel paper on epilepsy and depression, Kanner complains that "neurologists and psychiatrists [have] stopped talking to each other" [45]. It is time that both specialties join forces again. We suggest that the study of AUDS patients may offer a new approach to better understanding epilepsy and its association with depressive disorders.