Depression in Epilepsy: Ignoring Clinical Expression of Neuronal Network Dysfunction?


Address correspondence and reprint requests to Dr. F. Gilliam at Washington University Adult Epilepsy Program, Box 8111, 660 South Euclid Avenue, St. Louis, MO 63110, U.S.A. E-mail:


Summary:  Epilepsy is a chronic disorder that adversely affects social, vocational, and psychological functioning. Despite the variety and complexity of the negative clinical associations with epilepsy, depression is remarkable in prevalence and related adverse effects on health status. An estimated 30–50% of persons with refractory epilepsy have major depression, and depression has a stronger correlation than seizure rate with quality of life. Suicide is one of the leading causes of death in epilepsy. Available data indicate that depression may result from underlying brain dysfunction rather than social and vocational disability. Most patients with depression are not screened systematically for the diagnosis, and are subsequently not treated. Although the density of serotonin receptors is greatest in limbic brain regions commonly involved in human epilepsy, such as the mesial temporal and prefrontal areas, no prior randomized controlled trials have evaluated the efficacy of serotonin reuptake inhibitors for depression in epilepsy.

Epilepsy is a chronic disorder that has complex interactions with social, vocational, and psychological functioning. Although multiple epidemiological studies indicate that depression is a common comorbid condition in persons with epilepsy, the relative contributions of social and vocational disability compared to limbic system injury are not known (1). Many common syndromes, such as temporal lobe epilepsy and frontal lobe epilepsy, are associated with injury or dysfunction in brain regions implicated in the etiology of depression based on density of serotonin receptors. Despite the plausibility of the variable efficacy of serotonin-reuptake inhibitors related to limbic system responsiveness, few randomized controlled trials of the treatment of depression in epilepsy are published (2).

The consequences of depression in persons with epilepsy have not been fully examined. Some clinical studies have suggested synergistic negative effects of epilepsy and depression on various aspects of health status, and suicide is one of the leading causes of death in epilepsy. Although depression appears to be highly prevalent in epilepsy and has significant consequences, available data indicate that most patients with depression are not screened or treated. This article reviews the available information regarding the etiologies of depression in epilepsy and current standards of care to offer areas for further research and to alleviate the burden of depression in persons with epilepsy.


The prevalence of depression ranges from 20 to 55% in patients with recurrent seizures and from 3 to 9% in patients with well-controlled seizures, based on available epidemiological studies. In an inpatient sample in a video/EEG seizure monitoring unit, Boylan et al. (3) found that 50% were depressed, 19% had recent suicidal ideation, and only 17% were being treated with an antidepressant medication. Mendez et al. (4) used the Hamilton Depression Rating Scale in 175 consecutive patients in an outpatient epilepsy clinic, finding that 55% met criteria for depression. In a community-based study that used the Hospital Anxiety and Depression Scale, Jacoby et al. (5) observed, in a large survey, that 21% of 168 patients with recurrent seizures were depressed. Approximately 9% of controlled patients had significant symptoms of depression. O'Donoghue et al. (6) used the same scale to demonstrate that, out of a group of 155 patients identified through two large primary care practices in the United Kingdom, 33% with recurrent seizures and 6% of those in remission had probable depression. Edeh and Toone (7) found that 22% of 88 epilepsy patients identified from general practices in the United Kingdom using the Clinical Interview Schedule had probable depression. Although these studies have methodological considerations, including potential selection bias and varying diagnostic techniques, depression is consistently found to be at least five to 10 times more prevalent in association with uncontrolled epilepsy than in the general population.

Suicide has one of the highest standardized mortality rates (SMR) of all causes of death in persons with epilepsy. Robertson reviewed 17 studies pertaining to mortality in epilepsy, summarizing that suicide was nearly 10 times more frequent than in the general population (8). Rafnsson et al. (9) recently reported the results of a population-based incidence cohort study from Iceland in which suicide had the highest SMR (5.8) of all causes of death. A Swedish study of cause specific mortality among 9,000 previously hospitalized patients with epilepsy found an SMR of 3.5 (10). The findings from these studies underscore the importance of depression in epilepsy and the need to effectively screen and treat depressive disorders.


The psychosocial ramifications of epilepsy have been recognized for many decades, and the long-term disability is clearly characterized even for patients in extended remission. However, few studies have systematically evaluated the potential relationship between epilepsy-related disability and depression. Hermann et al. (11) examined the relevance of the theory of learned helplessness and also attributional style to depression in epilepsy. Using the Beck Depression Inventory (BDI), Center for Epidemiological Studies-Depression scale (CES-D), and the Optimism/Pessimism Scale, they demonstrated that attributional style was significantly associated with increased self- reported depression and remained significant when the effects of several confounding variables were controlled (age, age at onset, laterality of temporal lobe epilepsy, sex, and method variance). They concluded that attributional style was a component of the genesis of depression in some patients with epilepsy, as it is in the general population. These results emphasize the complexity of the etiologies of depression in epilepsy, and also the potential exacerbating factors of social isolation and vocational disability in patients with recurrent seizures.

Adverse psychosocial conditions, e.g., stress, social stigma, employment limitations, are an intuitively compelling explanation for the high prevalence of depression in epilepsy, but emerging evidence suggests that brain dysfunction in regions related to mood regulation may be more important. As Engel et al. (12) stated nearly two decades ago, “although arguments favoring the occurrence of epilepsy-induced disorders of behavior must not be presented in a way that adds to the stigmata associated with epilepsy, it is not in the best interest of epileptic patients to deny this possible relationship and overlook an opportunity to prevent or treat a major cause of disability.”

More than 10 years ago, Bromfield et al. (13) evaluated FDG-PET abnormalities and Beck Depression Inventory scores in nine brain regions of 23 patients and found significant correlation of higher depression scores with hypometabolism in the inferior frontal regions. It is noteworthy that similar observations have been described in patients with depression who had no history of neurological disorder (14). Victoroff et al. (15) studied 53 epilepsy patients with FDG-PET and structured psychiatric interviews, and found significant correlation with left temporal lateralization and severity of hypometabolism with a history of major depressive disorders. However, severity of hypometabolism in the right temporal lobes also correlated with history of depression.

Quiske et al. (16) examined the association of MRI-defined mesial temporal sclerosis (MTS) with Beck Depression scores in 60 patients with temporal lobe epilepsy. Mean depression scores were significantly higher in patients with MTS, independent of the lateralization of MTS. The investigators described depression as “a good indicator of MTS, but not vice versa.”

Schmitz et al. (17) compared clinical and psychosocial variables in 25 patients with epilepsy and depression to 50 otherwise healthy epilepsy patients. Although the mean age was higher in patients with depression and epilepsy, no differences were found between any psychosocial variables. Based on their findings, the authors questioned the “simple hypothesis” that social and vocational problems could predict depression in epilepsy.

We performed a pilot study of the association of severity of depression symptoms, clinical variables, and FDG-PET. Sixty-two consecutive patients from our outpatient clinics at Washington University who had previously had a PET scan as part of a presurgical evaluation were enrolled. After approval from our Institutional Review Board, each patient completed a demographic and seizure assessment, the BDI, and the Adverse Events Profile (AEP) (18).

During the PET imaging, patients were awake and resting with eyes closed and ears unplugged. PET scans were performed on a high-resolution head-dedicated camera (ECAT HR plus Siemens or ECAT-931). Imaging was initiated 30 min after a bolus of 250 MBq 18-FDG and ended 20 min later. Images were obtained as 63 slices of 0.25 cm thickness. All PET scans were evaluated qualitatively, and images were classified by brain region and severity of hypometabolism by a nuclear medicine physician who was not aware of the results of mood or epilepsy assessments.

The mean age, complex partial and generalized tonic-clonic seizure rate, number of current antiepileptic drugs, and AEP scores were similar between the patients who had an abnormal PET (n = 55) compared to the group with a normal PET (n = 7), as shown in Table 1. However, the mean BDI scores were higher in the abnormal PET group (Mann-Whitney U; p < 0.01), as shown in the box plot in Fig. 2. Fifty-one (82%) of the PET scans showed abnormalities limited to the temporal lobes (maximal hypometabolism; 23 left versus 28 right). Figure 1 shows a typical example of temporal lobe hypometabolism. These findings were consistent with the previous published studies that supported the association of symptoms of depression with functional neuroimaging abnormalities (13, 15).

Table 1. Comparison of clinical and mood assessments in 62 patients with refractory localization-related epilepsy divided by FDG-PET findings
 Abnormal FDG-PET
(n = 55; mean (SD)
Normal FDG-PET
(n = 7; mean (SD)
  1. GTC, generalized tonic-clonic; AEDs, antiepileptic drugs.

  2. aMann-Whitney U, p < 0.01; no other significant between-group differences.

Age35.9 (11.7)31.2 (8.6) 
Monthly complex partial 8.7 (14.6)10.2 (14.3)
 and GTC seizures
Number of AEDs 2.2 (0.85)1.7 (1.0)
Adverse Events39.8 (10.8)33.5 (10.7)
 Profile score
Beck Depression9.7 (8.5)1.8 (2.6)
 Inventory scorea
Figure 2.

Box plots of median and quartile Beck Depression Inventory scores from patients presented in Table 1 (Mann-Whitney U; p < 0.01).

Figure 1.

FDG-PET image of right temporal (white arrow) and right basal ganglia hypometabolism in a patient with refractory right temporal lobe epilepsy and recurrent depression.


Identification of depression in persons with epilepsy

The epidemiology of the treatment of depression in epilepsy is not well characterized, but the few available studies indicate that depression usually goes untreated (19,20). Ojemann et al. (21) investigated an adult clinic population of 2,000 at a tertiary epilepsy center caring for complicated cases, finding that <8% were on antidepressant medications. If an estimated 30% of patients were depressed, then only 27% of the depressed group was receiving treatment.

To assess the approach to the diagnosis of depression in persons with epilepsy, we surveyed a representative sample of neurologists in the U.S. to determine their practice patterns and their perspective on relevant factors that might affect their screening for depression in epilepsy patients. After obtaining approval from the Washington University Institutional Review Board, we generated a random code to select 135 neurologists from the published membership of the American Academy of Neurology. Through a brief e-mail questionnaire that disclosed our intentions, each neurologist was asked to answer the following two questions:

  • (a) Do you routinely screen epilepsy patients for depression in your outpatient clinics?
  • (b) If a randomized controlled trial demonstrated that the treatment of depression improved compliance and health-related quality of life in epilepsy patients, would you systematically screen for depression in your outpatient clinic?

E-mail tracking confirmed that 113 (84%) neurologists received the questionnaire. Seventy-five (66%) from this group responded. Eight neurologists reported that they did not see epilepsy outpatients. The results from the group of neurologists that see epilepsy patients (75 − 8 = 67) are summarized in Table 2.

Table 2. Questions asked of 135 randomly selected neurologists from the American Academy of Neurology Directory and the answers given by 67/135 respondents who see epilepsy patients
 (a) Do you routinely screen epilepsy patients for depression
    in your outpatient clinics?
 (b) If a randomized controlled trial demonstrated that the treatment
    of depression improved compliance and health-related quality
    of life in epilepsy patients, would you systematically screen
    for depression in your outpatient clinic?
Answer toAnswer toTotal
question (a)question (b)neurologists (%)
NoYes53/67 (79%)
 (95% CI; 67%–91%)
NoNo2/67 (3%)
YesYes4/67 (6%)
YesNo1/67 (1%)
Did not answer questions 7/67 (10%)

Fifty-three (79% of responders who see epilepsy patients) reported that they did not screen for depression in their clinic, but that they would if an appropriate trial demonstrated that treatment improved compliance and health outcomes such as quality of life. Only 4% of neurologists responded that they would not routinely screen for depression if they were presented with positive results from a trial such as the one that we are proposing. Although this survey does not provide indisputable evidence that the proposed trial will improve the detection of depression by neurologists treating refractory epilepsy, it indicates that neurologists are willing to systematically screen for depression and will seriously consider data from clinical trials supporting the need to address the negative impact of comorbid depression on patients with epilepsy.

Treatment of depression in persons with epilepsy

Reasons for inadequate identification and treatment include lack of appreciation of the impact of depression on quality of life (22–24), fear of seizure exacerbation by antidepressants (25,26), and belief that antidepressants are ineffective in epilepsy patients (20). The relative importance of depression in epilepsy compared to other variables, such as seizure frequency and severity, has only recently been examined (3,27).

In 1959, seizures were first reported in patients taking antidepressants, soon after the release of imipramine in the United Kingdom (28). Although numerous cases of isolated seizures after initiation of tricyclic and SSRI antidepressants have been reported (19,25,26), the only prospective controlled trial of antidepressants in epilepsy patients found no difference in seizure rates between placebo and tricyclic agents (29). This study included only a total of 42 patients in three treatment arms; the small sample size had inadequate power to detect any less than a very large effect. In a retrospective 2-month study of the effect of doxepin on seizure frequency in 19 patients with epilepsy, Ojemann et al. (21) found that 15 patients experienced a reduction in seizures, and only two patients experienced increased seizures. Kanner et al. (30) also reported a prospective study of sertaline, finding that only 6% had any significant increase in seizure rate or severity. Pisani et al. (26) concluded in a recent comprehensive review that “the use of antidepressants in patients with epilepsy still raises uncertainties because of the widespread conviction that this class of drugs facilitates seizures.” They also added that “the question of whether or not antidepressants at therapeutic doses cause seizures still merits great attention.”

The efficacy of antidepressants on mood outcome in patients with epilepsy also remains controversial (2,20). Despite the high prevalence of depression in patients with epilepsy and the important influence of depression on health outcomes in chronic medical illness, only one controlled trial of the efficacy of antidepressant medications in epilepsy has been reported. In 1985, Robertson and Trimble described the results of a randomized, double-blind comparison of amitriptyline, nomifesine, and placebo in 42 patients with depression and epilepsy (29). Although the mean scores of the Hamilton Depression Rating Scale and the Beck Depression Inventory scores improved by 50% after treatment, similar improvement in the placebo group resulted in no significant difference in outcome between any groups at 6 weeks. A second 6-week treatment phase without placebo control compared higher doses of each drug (150 mg), finding that the nomifesine group had significantly better Hamilton Depression Rating Scale but not Beck Depression Inventory scores. The authors concluded “our results suggest that, in patients with depression and epilepsy, immediate prescription with antidepressants may not be indicated (29).” However, the small sample size (14 patients in each treatment group) limits the interpretation of the results. Blumer (31) subsequently reported the results of an open-label, uncontrolled study of the addition of a seratonin reuptake inhibitor in 22 patients who had not responded to a tricyclic antidepressant. The sample from which these patients were chosen was not clearly defined in the article. The outcome was stratified as responders, partial responders, and nonresponders. After 20 months of follow-up, 15 (68%) patients were rated as excellent or good responders.

Few prior studies have addressed specifically the relationship of depression and noncompliance in epilepsy (32). However, Ojemann et al. (21) found that seizure rates improved in the 2 months following psychotropic medication treatment in 59 epilepsy patients, compared to the previous 2 months. The authors speculated that the improvement might be due to better compliance. Helgeson et al. (33) reported that compliance and mood improved after a brief psychoeducational program, but could not confirm a causal relationship. Evidence from studies of other chronic illnesses suggests that mood is an important factor in compliance (34–39). A recent study of multiple sclerosis found that patients who responded to antidepressant therapy had better adherence to the prescribed beta-interferon regimen than those with continued depression (40).

Although some studies indicate that quality of life is improved after treatment of major depression (41), this response has not been established in patients with other confounders, such as the adverse effects of chronic medical illness. The complex effects of epilepsy on mood, cognition, social function, employment, and autonomy make establishing the effect of the treatment of depression critical to our understanding of optimum health outcomes.


Depression has emerged, through recent patient-oriented outcome studies, as a formidable problem for a large minority of patients with epilepsy (19,20). Significantly increased rates of suicide underscore the importance of definitively addressing the problem of major depressive disorders in epilepsy, including epidemiological, etiological, and health services considerations. Although complete seizure control is the primary treatment consideration for most patients with epilepsy, depression and the adverse effects of antiepileptic medication appear to be important considerations for the optimal outcome for many patients (18,27,42). Certain personality characteristics, such as attributional style and learned helplessness, may be risk factors for depression in epilepsy, but multiple studies have found functional neuroimaging abnormalities to have significant association with depressive symptoms.

Available evidence indicates that most patients with epilepsy are not screened for depression, and a small proportion of affected patients are subsequently treated. Reasons for this include time constraints on busy clinicians that limit the opportunity for diagnosing depression. The hesitance to screen and treat may also be partially based on the belief that antidepressant medications lower the seizure threshold, despite the sparse evidence supporting this assumption. Further research is needed to clarify the relative risks of treating depression versus continued mood dysfunction, as well as to increase knowledge about optimal interventions. Many unanswered questions regarding risk factors and etiologies of depression in epilepsy, chronicity, contributions to long-term disability and poor health status, most effective treatments, and predictors of refractoriness will need to be addressed before optimal management strategies can be implemented throughout the medical care system to improve the lives of persons with epilepsy and depression.


Acknowledgment:  Supported by National Institutes of Health grants NS01794 and NS40808, and a grant from the Epilepsy Foundation of America.