Epilepsy is a complex neuropsychiatric illness that has been studied for millennia. Hippocrates commented not only about epilepsy but about the common overlap between epilepsy and mood disorders, especially depression. Hippocrates is quoted circa 420 B.C.E. as saying, “Melancholics ordinarily become epileptics, and epileptics melancholics: what determines the preference is the direction the malady takes: if it bears upon the body, epilepsy, if upon the intelligence, melancholy” (
1). This profound statement by Hippocrates is as relevant today as it was more than 2,000 years ago. Psychiatric illness is exceedingly common in epilepsy and continues to be prevalent well beyond the levels seen in other chronic medical conditions.
The 20th century brought the maturation of biological approaches to psychiatric illness as well as the advent of the electroencephalogram (EEG) to diagnose and monitor seizure disorders. Despite the clinical divergence of epilepsy to more neurologically geared physicians, the relationship between the two conditions persisted. Today, epilepsy often is considered a “neuropsychiatric” illness—a medical condition involving both psychiatric and neurologic manifestations (
2). Attention to mental health considerations in epilepsy programs has grown substantially in recent years, ostensibly in recognition of the clinical necessity for a neuropsychiatric approach to treatment (
3). The objective of this review article is to highlight diagnostic and treatment strategies for patients with epilepsy and comorbid depression or bipolar disorder.
Epilepsy Classification and Subtypes
A seizure itself is defined as “a transient occurrence of signs and/or symptoms due to abnormally excessive or synchronous neuronal activity in the brain” (
4). A single seizure may occur from a variety of medical conditions, including dehydration, hypoglycemia, toxic exposures, or trauma. However, a single seizure does not suffice for the diagnosis of epilepsy and typically is not managed with anticonvulsant drug treatment. A diagnosis of epilepsy requires the existence of recurrent seizures, with specific events separated by 24 hours or more and recurring within approximately 18 months. Epilepsy is prevalent in about 1%−1.2% of the adult population, according to epidemiologic studies in the United States and in Europe (
5).
An epilepsy diagnosis becomes applicable in the context of recurrent seizures. However, even in those cases, seizures that recur with childhood febrile illness or that recur in the context of substance abuse also may not be classified as epilepsy. Instead, an intrinsic central nervous system disruption is necessary for epilepsy to be diagnosed. Epilepsy may then be described in one of three ways. The first category is symptomatic epilepsy, where the cause is known. Examples include a brain tumor or stroke, where the seizures emerge from known areas of pathology. The second category is cryptogenic epilepsy, where the cause is suspected on the basis of reasonable assessment of neurologic state, although it may not be proven definitively. For example, a patient might have a history of trauma that might not clearly correlate with evident seizure localization, but epilepsy nonetheless is suspected on the basis of available evidence. The third description is idiopathic epilepsy, where the cause is unknown. Most cases of epilepsy fall into this category. Although genetic or developmental anomalies may be suspected, there is insufficient evidence to suggest a clear relationship between etiology and semiology—observable clinical symptoms or signs that result from seizure activity.
Epilepsy classification has evolved over recent decades. Terms such as “petit mal” and “grand mal” are antiquated, although still part of the vernacular for patients and physicians alike. “Grand mal seizures” refers to classic generalized tonic-clonic seizures, with visible full-body convulsions. “Petit mal” essentially refers to smaller (petit) and shorter events, without full-body convulsions. Usually, these events are consistent with absence seizures.
Classification is dependent primarily on seizure-related characteristics and usually involves combining behavioral observations of seizures as well as ascertaining the initial pathologic changes detectable on an EEG. Many seizure types exist, and although most are descriptive, some seizure types are related specifically to epilepsy syndromes that are identifiable. Many epilepsy syndromes have been named, but to more fully understand the disorder, the first step is to describe the seizure activity. The International League Against Epilepsy recently revised its classification schemes, and most descriptive parameters are intuitive (
6). Ultimately, seizure classification may have important implications for psychiatric comorbidity.
The first step in classifying seizure activity is to determine whether the seizure is focal (initial symptoms starting in one portion of the body or in a limited section on an EEG) or generalized (initial symptoms affecting the entire body at once or widespread across the EEG). Some seizures begin as focal events but then become generalized as the seizure propagates (i.e., secondary generalization).
The next step is to determine whether there is change of consciousness. Every generalized seizure by definition has disrupted consciousness. The change in conscious awareness accompanies many focal seizures as well, especially if seizures emerge from temporal lobes of the brain. Some focal seizures do not have disrupted consciousness—that is, patients are fully alert and understand that they are experiencing a seizure. Seizures may then be described in a variety of ways (e.g., focal with disrupted consciousness, primary generalized) (
7).
Prevalence of Mood Disorder in Epilepsy
The overall prevalence of psychiatric illness associated with epilepsy is much higher than in the general population, ranging from 25% to 50% (
8). Comorbidities range from psychosis to mood disorders and cognitive dysfunction. For adults, comorbidities seem to be more frequent in the context of poorly controlled epilepsy and may approach a prevalence rate of 60% (
9).
Mood disorder may represent the most common, and likely the most worrisome, psychiatric manifestation associated with epilepsy. Depression is frequently associated, although anxiety and bipolar disorder may also co-occur with epilepsy. Mood symptoms may also be present in the context of seizure events. Some people with epilepsy may have depressed mood immediately following a seizure (i.e., in a postictal state). The mood states usually return to baseline within a day or so. In some cases, hypomania or other manic symptoms may occur immediately before or after seizures. Some pediatric patients have been reported to have increased motor activity or mood lability a day or so before seizures occur.
However, pure bipolar disorder comorbid with epilepsy seems uncommon. In a recent study, when symptoms related to seizure events were discounted, the presence of bipolar disorder was only 1.4% of the sample, thus not reflecting a higher prevalence than in the general population without epilepsy (
10). The commonality with bipolar disorder may be theoretical in nature. The propagation of manic episodes as well as seizure episodes seems to depend on kindling phenomena, where neural activity is more likely to recur with each subsequent exacerbation (
11,
12).
Another curious overlap of bipolar symptoms and epilepsy occurs with the description of “temporal lobe personality.” In the 1970s, Waxman and Geschwind, two prominent neuropsychiatrists, observed that persons with temporal lobe seizure foci often had a unique personality style (
13). They were characterized as having irritability, altered sexual appetites, preoccupation with religious or moral issues, and hypergraphia. Loosely applied, these criteria may be consistent with some symptoms of manic episode, particularly the irritability and religiosity. Ultimately, the association of mood disorder with temporal lobe pathology became viewed more consistently.
Yet the intrigue of the overlap between bipolar disorder and epilepsy may be in the use of the same class of medicines to treat both conditions. Some anticonvulsants are excellent mood stabilizers and also excellent adjunctive treatments for depression. A large database exists for medicines such as carbamazepine, divalproex sodium, and lamotrigine for the treatment of bipolar mood disorders (
14). Other anticonvulsants may also have benefit for improvement of aggression or irritability (
15). Even levetiracetam has been reported to improve bipolar mood symptoms (
16). In some cases, bipolar symptoms and epilepsy may be treated simultaneously by the same anticonvulsant (
17). However, regardless of the intriguing overlap between bipolar symptoms and epilepsy and relevancies for treatment, the overlap between epilepsy and depression is most striking.
Overrepresentation of Depression
A recent meta-analysis revealed prevalence of depression in epilepsy to be as high as 23% in some studies. In this same analysis, the odds ratios for depression in epilepsy were consistently ≥2.0 throughout studies reviewed (
18). However, other studies have identified the rate of depression as much higher, even as high as 50% in some samples (
19,
20). Even more concerning is the high suicide rate that may be present as well. Although it is not clear whether depression confers a higher suicide risk with epilepsy, suicide risk and seizure risk do seem to be interrelated (
21). Suicide risk seems to be fivefold more common for people with epilepsy, and suicide attempts may be prominent even prior to manifestations of epilepsy (
22). In addition, epilepsy was found to be overrepresented among hospitalized children and adolescents who attempt suicide (
23).
The cause for such overrepresentation of depression is unclear. One reason posited is that epilepsy involves sudden exacerbations that may be life threatening. Sudden seizure events are very disconcerting and may lead to anxiety or depressive symptoms, among other psychiatric issues. However, an illness such as asthma also involves sudden exacerbations that may be life threatening and is possibly even more frightening, given that consciousness is preserved. Yet psychiatric illness is more than three times as common with epilepsy as it is with asthma (
24,
25). Similarly, epilepsy may involve lifestyle change and lack of independence and in that way possibly contribute to depression. However, even diabetes, an illness involving much lifestyle change, has less associated psychiatric illness than epilepsy (
26). Clearly, the relationship between epilepsy and psychiatric illness, particularly mood disorder, is complex. Many differing views characterize the etiology of depression in the context of epilepsy and have important implications for determining a treatment strategy.
Etiology of Depression in Epilepsy
Although the idea that depression stems from intrinsic aspects of epilepsy is increasingly well accepted, in the past, depression was thought to be related only circumstantially. A large literature exists that faults psychosocial stressors or the burden of chronic disease as leading to mood disorder. Another important perspective is consideration of iatrogenic causes of comorbid depression. Although these potential etiologies of depression may be related to epilepsy indirectly, they have important implications for clinical management.
Antiepileptic Medication as a Cause of Depressive Symptoms
Unfortunately, in some cases, anticonvulsant drugs used to treat epilepsy can lead to depressive symptoms. The most likely offender usually is considered to be phenobarbital. Phenobarbital is a very powerful agent acting on postsynaptic GABA-A receptors. The enhancement of GABAergic transmission increases global inhibitory networks and decreases overall neural activity of the brain. At high doses, phenobarbital may induce a neurologic coma, which may be a necessary treatment in
status epilepticus, a condition where seizures are prolonged and do not remit spontaneously with other medications. However, phenobarbital may also cause cognitive dysfunction and sedation similar to other GABAergic drugs, such as the benzodiazepines. In the case of phenobarbital, depressive symptoms have been noted throughout the age span and in many instances resolve when phenobarbital is removed (
27,
28).
It may be intuitive to consider that other medicines that similarly enhance GABAergic transmission could also increase the risk for depression (
29). Although specific mechanisms are not known fully, other strong GABA enhancers likely include clobazam, primidone, tiagabine, pregabalin, gabapentin, and vigabatrin. Yet these medicines do not seem to consistently increase the incidence of depression. Some case reports show that pregabalin and gabapentin may improve anxiety (
30). In fact, most benzodiazepines are anxiolytic and often used as adjunctive treatments for the many persons who have pathologic anxiety along with depression.
Other anticonvulsants may have depression as a prominent side effect. Irritability has been noted in as many as 15% of those patients receiving levetiracetam, although the mechanism is largely unknown. However, in some cases, levetiracetam may also improve mood symptoms. Use of folic acid supplementation may help mitigate irritability in some circumstances (
31).
Still, it is true that any antiepileptic drug may have side effects, which could include depression. Side effects are extremely difficult to predict, given the heterogeneity of both psychiatric illness and epilepsy. More details on neuropsychiatric effects of antiepileptic drugs are shown in
Table 1 (
14,
32,
33).
Lifestyle Stress
Another theory for mood disorder and epilepsy relates to lifestyle stress. The threat of seizures occurring at any moment can be particularly anxiety provoking for persons with epilepsy. Anxiety and stress may be significant not only for the patients themselves but also for caregivers, spouses, or parents of individuals with epilepsy (
34). For individuals with epilepsy, caregivers may unwittingly prevent independent activity and foster dependency, ostensibly for fear of a seizure occurring in a location where no one may be available to assist.
A well-developed literature on locus of control may also be relevant for people with epilepsy. Individuals with an internal locus of control seem to have less depression than those with an external locus of control. This theory holds that if individuals perceive efficacy or control over their actions and, ultimately, their fate, then they are less likely to develop learned helplessness, which can lead to depressive illness.
Seizure events may seem random and out of an individual’s control; however, certain lifestyle changes may help to reduce the potential for seizures to occur. Having a proper diet, maintaining hydration, and avoiding sleep disruption may all have an impact on the seizure threshold. An individual who recognizes the benefit of health-promoting lifestyles not only will have improved seizure control but also may have a more empowering and internalized locus of control.
Finally, stress itself may be related to both the exacerbation of seizures and the development of epilepsy. Stressful events may alter steroid hormones and, in turn, influence neuronal excitability such that the seizure threshold is affected (
35). Stress is reported frequently by patients as relevant to the timing of seizure events. One small study in pediatrics found that 49% of the time, an acute stressor preceded a seizure (
36). A study in Israel found that exposure to combat led to an increase in the development of epilepsy compared with a control group (
37).
Bidirectionality of Epilepsy and Depression
The evidence of psychosocial stressors as potentially causative of seizure disorders is a fascinating development and validates the likely physiologic overlap between mood disorder and epilepsy. The connection between the two conditions involves intrinsic pathophysiologic processes that complicate any discussion of etiology. Epilepsy is known to lead to the development of depression, but, recently, the converse also has been found to be true: that depression may lead to the subsequent development of epilepsy (
38). Adults with depression are more likely than those without depression to develop epilepsy during the course of their illness (
21). Scholars now consider that a “bidirectional relationship exists between depression and epilepsy where one condition may exacerbate and even lead to the development of the other” (
38,
39).
Temporal Lobe Pathophysiology
The temporal lobes are a common location for seizure foci, particularly the amygdala and hippocampus in the medial regions of either temporal lobe. Seizure foci in the temporal lobe typically lead to focal seizures with disrupted consciousness. There may be some laterality of function, such that verbal memory is subtended by the left hippocampus and nonverbal memory by the right. Thus, memory impairment is common with seizures localized in these areas. Additionally, the seizures commonly include an aura of several seconds prior to disruption of consciousness. If the seizure focus is in the amygdala, the aura may involve a sensation or feeling of fear or foreboding. The sensation may be phenomenologically very similar to the initial stages of a panic attack.
Today, a growing body of evidence suggests that the temporal lobes of the brain are closely associated with mood and anxiety pathophysiology (
40). Neuroimaging studies consistently have implicated the temporal lobe, especially the amygdala, in a circuit involving the prefrontal cortex and caudate nucleus (
41–
44). The amygdala seems to be larger and the hippocampus smaller, particularly if depression has been present for a long period (
45–
47). In addition, if both epilepsy and depression are present, the hippocampi are structurally smaller than found in either condition by itself. The two conditions together may result in more severe anatomic loss (
48). Although depression seems to be present with many different seizure foci, it may be most prominent, especially in medically refractory epilepsy, when the seizure foci are in either temporal lobe (
49).
Diagnosis of Mood Disorder
Much of psychiatric illness is determined by duration of symptoms and the length of time that the symptoms occur. Categorical criteria for depressive disorder require time spans of symptoms that are rated consistently and then may effectively differentiate diagnostic entities, such as manic episode, dysthymia, and major depressive episode. Also present in psychiatric diagnostic criteria are exclusions for medical illness or attribution of symptoms to other conditions.
These types of categorical criteria present unique challenges in terms of epilepsy. Although seizure episodes may have a readily identifiable onset and endpoint, the pathophysiology inherent in epilepsy is less clear in terms of timing. Typical seizure episodes may have an aura or prodrome that lasts three to five seconds, then 90 seconds of focal or generalized seizure activity, and then a recovery phase involving headache or fatigue that may last a few hours after the event.
However, the actual chemical and physiologic changes associated with seizure events may occur well beyond those time points. The biological precursors to seizures and postictal recovery periods may involve hours or even days beyond the actual events. Patients who have seizures frequently may have lengthy periods of dysfunction during recovery phases that are difficult to discern from underlying psychiatric conditions. Thus, it may be very difficult to distinguish a chronic condition, such as depression, as independent of epilepsy, also a chronic illness.
In an attempt to isolate these conditions, researchers use the term interictal dysphoric disorder (IDD) to characterize depression that is present beyond acute seizure events (
50,
51). The definitions of IDD are somewhat arbitrary in terms of timing, especially given the variability of seizure events, but the criteria are not wholly inconsistent with the approach leading to categorical criteria for major depressive disorder.
Although the differentiation of diagnostic terms such as IDD versus depressed mood secondary to epilepsy may be impossible, attempting to characterize the distinction may be useful for treatment planning. Patients with IDD may benefit from primarily depression-guided treatment, whereas those with the secondary condition may benefit more from primarily epilepsy-focused medication adjustments. Nevertheless, identification of depression in the context of epilepsy can be done readily with several straightforward steps, regardless of distinctions about the endpoint diagnostic terminology. These steps are outlined in
Box 1.
Treatment Strategies
Anticonvulsants as Mood Stabilizers and Antidepressants
It is important to note that for patients with epilepsy, treatment of depression versus treatment of epilepsy may be a moot point. In many cases, anticonvulsants have dual roles and may simultaneously improve seizure disorders as well as mood symptoms. Anticonvulsants as a class are considered to be mood stabilizers and represent powerful tools for clinicians seeking to optimize complex treatment regimens.
As mood stabilizers, anticonvulsants may be especially effective to treat neuronal excitability leading to impulsivity and disruptive behavior. Mood lability and thought disorder may also be reasonable targets for these medicines. For younger patients, sodium valproate has been effective for extreme irritability and for explosive mood outbursts (
52). For people with epilepsy and intellectual and developmental disabilities, the impulse control afforded by anticonvulsants may offer a multiplicity of benefits (
53).
The evidence base for some medications that function as anticonvulsants and mood stabilizers is quite extensive. Classically, divalproex sodium and carbamazepine have long been used as primary agents for bipolar mood disorders and adjuncts for depressive disorders (
54,
55). Some of them may also be effective for mood stability, although the evidence base is not well developed.
Although many anticonvulsants have negative clinical trial data when used for primary bipolar disorder, some of them may offer symptomatic benefit, such as oxcarbazepine and lamotrigine. Lamotrigine may also have benefit for depressive symptoms associated with bipolar disorder (
56).
Similar to the reported concerns with antidepressants, anticonvulsants also were reconsidered in terms of suicidality. The Food and Drug Administration added precautions to drug information labels for anticonvulsants as a class after post hoc review of clinical trial data suggested that suicidality was increased. However, many of the clinical trials reviewed did not establish baseline psychiatric status and also did not differentiate anticonvulsant trials geared toward psychiatric illness versus epilepsy (
57,
58). As such, the consensus conclusion in the field was that careful consideration of risks and balances is the best practice (
59). Overall, for persons with epilepsy, anticonvulsants continue to be versatile tools for the treatment of a variety of psychiatric illnesses, including mood disorder.
Antidepressant Usage
Historically, clinicians have been hesitant to use antidepressants for fear of adversely altering the seizure threshold. However, only a few antidepressants have relative contraindications. Tricyclic antidepressants in the context of toxicity or purposeful overdose may cause seizures. High-dose clomipramine and immediate-release bupropion may also have a dose-dependent increased risk of seizures. However, most antidepressants seem to be well tolerated and have not yielded de novo seizures in clinical trials.
Although the literature is not very well developed, selective serotonin reuptake inhibitors have been used with some success in the epilepsy population. No double-blind, randomized, placebo-controlled studies have been done, but some uncontrolled and observational studies have been reported that reasonably add to the evidence base.
In one study, sertraline was used with 100 patients with epilepsy and depression. Seizures worsened for six patients but resolved after antiepileptic drug dosages were adjusted. Patients also subjectively reported significant improvement in depressive symptoms (
60).
Citalopram was used in two smaller studies without significantly exacerbating seizures and also yielded apparent mood improvement in a significant portion of the sample (
61,
62). Reboxetine and mirtazapine were also found to be well tolerated and effective for patients with temporal lobe epilepsy and major depression (
63). Ultimately, most modern antidepressants seem to be safe for persons with epilepsy when used in therapeutic dosages (
64).
Psychotherapy
Individual psychotherapy is an important strategy to complement medical treatment of depression comorbid with epilepsy. A person with epilepsy often has difficulties with vocational opportunities and may have limited mobility because of restrictions on driving. Additionally, stress related to stigma associated with epilepsy is still a reality for many. Individual therapy focusing on cognitive-behavioral techniques has been done with moderate success (
65). Group therapy and recreational support also have been attempted with success for adolescents, although much of the progress may be geared to improving social skills (
66).
Support for caregivers is also important, especially for parents of children with epilepsy. Parents often have significant stress that is difficult to manage (
34). Caring for a child with epilepsy often involves fear of medical destabilization and uncertainty regarding how much independence to allow. Spouses or significant others in the social network may benefit from support, particularly in understanding the disease process and limitations that may be present when mood disorder comorbidity is present.
Quality of Life
The necessity for treatment of depression is intuitive, to prevent negative outcomes related to depression. However, it has recently become more apparent that treatment of depression may be critically important for quality of life. Depression seems to have more to do with quality of life than epilepsy-related characteristics, including frequency of seizures. Even after surgical procedures have been done to treat medically refractory epilepsy, addressing potential depression is a critical component of the treatment outcome (
67). Especially if seizure control is incomplete, addressing depression is critical for health-related quality of life (
68).
Conclusions
Mood disorders, and particularly depression, are commonly associated with epilepsy. Fortunately, both medical and nonmedical treatments may be very effective. Although the evidence base for treatment is not well developed, small samples and observational studies are promising and help guide clinical practice. Seizure disorders that involve the temporal lobe and that do not involve generalized seizure episodes should alert clinicians to the possibility of comorbid depression. As the evidence base grows, more details regarding vulnerable subtypes and optimal usage of anticonvulsants and antidepressants will become available. Until that time, identification of presence of depression will, by itself, go a long way toward improvement of quality of life for persons with epilepsy and comorbid mood disorder.