Mood disorders in patients with medical illness are generally poorly identified and appropriately treated. It has been estimated that only 19% of patients seen in a general medical clinic receive proper treatment for their depressive disorder. Patients with subsyndromal depressive disorders may also experience deteriorations in quality of life because of depressive symptoms (12).

In over two thirds of patients with an idiopathic depressive disorder, the diagnosis is missed. Unfortunately, the fate of PWE is not any better. In a study of adult PWE and depression, 43% with a current MDD and 68% with a minor depression were untreated and in 38% with lifetime histories of MDD, no treatment had ever been instituted (13). A recent survey of neurologists found that 80% do not routinely screen PWE for depression (14). Children with epilepsy and depression also go unrecognized. In a study by Ettinger et al. 26% (n = 44) of the children studied had a significant depression and all went undiagnosed and untreated (15). As will be discussed later, quality of life is also severely affected by the presence of depression and with increased severity, suicide is a potential consequence.

When evaluating a patient with epilepsy and symptoms of depression, we need to keep in mind two major issues. The first is to avoid the frequently quoted misconception that depression from a “good cause” is
understandable and does not require treatment. This is a major fallacy and can lead to needless morbidity and potential mortality. The second results from the difficulty in determining the extent of depression in patients suffering from side effects from medications like antiepileptic drugs (AEDs) that can also be associated with a depressive disorder. Such side effects include weight gain, lethargy, poor concentration, and so on. The presence of anhedonia is an excellent marker for the presence of depression and is often impervious to physical complaints secondary to drugs or underlying illness. It is also a barometer of the intensity of the depression in the medically ill (16). The patient who is noted to display little interest in his or her surroundings or significant others needs to be assessed for depression and treated accordingly.

Using research diagnostic criteria (RDC), Robertson et al. (17) showed that PWE can meet standard measures for what is described as a MDD (Table 13.2).

This observation was confirmed by Mendez et al. (18), who used DSM-III standards to compare epileptic and nonepileptic depressed inpatients. Although the PWE met the criteria for MDD, they presented atypical features with more paranoia and psychotic symptoms that were peri-ictal and resolved with anticonvulsant adjustment. In addition, they tended to show a more chronic dysthymic course between
MDD episodes. At these times, the PWE demonstrated more irritability and emotionality (18). Psychotic features may be part of a MDD, representing a more severe form of the disorder (MDD with psychotic features).

In a recent study by Jones et al., 174 PWE were evaluated from five tertiary epilepsy centers (19). The current rate of DSM-IV MDD as determined by Structured Clinical Interview for DSM-IV (SCID) was 17.2% (12-month prevalence in the general population is approximately 10.3% [20,21]. Clearly, PWE can demonstrate classic symptoms of a depressive disorder that meet the criteria for a DSM-IV diagnosis of a MDD. Psychosis may also occur separately from, but coexist with, affective complaints, and this phenomenon is classified as a schizoaffective disorder.

Depressive symptoms secondary to medical illnesses from causes other than epilepsy not only meet standard criteria for MDD (16), but also respond to medication in a similar fashion as that seen in idiopathic depression. This has also been our experience in patients with interictal depression, with the caveat that there may be unique mood changes surrounding the peri-ictal period, which will be discussed later in the chapter.

Blumer (22) emphasized the pleomorphic pattern of mood complaints in epilepsy consistent with the observations of Kreaplin, Gastaut, and others, and coined the term interictal dysphoric disorder (IDD). The symptoms have an intermittent course and can be categorized into depressive-somatoform and affective symptoms. The depressive-somatoform symptoms include (i) depressive mood, (ii) anergia, (iii) pain, and (iv) insomnia. The affective symptoms include (i) irritability, (ii) euphoric mood, (iii) fear, and (iv) anxiety. Unfortunately, there are no direct comparisons in the literature evaluating depression in PWE using standard diagnostic techniques with those yielding a diagnosis of IDD. It is possible that a spectrum exists with a chronic dysthymic state characterized by the features of the IDD that may intermittently exacerbate and at that time meet the criteria for MDD. This pattern may be similar to “double depression” seen in nonepileptic patients (8). Longitudinal studies are needed to clarify these issues. The uniqueness of the depression seen in PWE and that in the general population has been discussed elsewhere (23).

The semiology of depression in PWE has been further clarified by Kanner et al. (24). In two studies, he noted the presence of atypical symptoms of depression in PWE. In a review of 97 patients with symptoms of depression, 28 met the criteria for a diagnosis of a MDD. The remaining patients had a pleomorphic picture of depressive complaints that were termed a dysthymic-like disorder of epilepsy. The symptoms often waxed and waned with one group (n = 36) complaining of more irritability and poor frustration tolerance and in the remaining 33, anhedonia was the most pronounced depressive feature (25). In another study, 205 patients were evaluated using the MINI-International Neuropsychiatric Interview (MINI) and the SCID. Sixty-seven
met DSM-IV criteria for a mood disorder. Excluding these patients, of the 132 remaining patients, 52 had some depressive symptoms associated with anxiety and irritability (26). Whether this symptom complex is unique from that seen in patients with idiopathic depressive disorders is debated (23).

Accurate recognition is especially important because the depression seen in PWE has also been associated with a significantly higher suicide rate than expected in the general population. This was noted by Barraclough (27), who reviewed 11 studies showing the overall rate in PWE to be five times higher than in the general population. It was 25 times expected in those patients with complex partial seizures (CPS) of temporal lobe origin. Mendez et al. investigated this occurrence in two studies. The first study looked at causative factors in completed suicide in PWE and found more psychotic symptoms in these patients (n = 4) (28). In a follow-up investigation, they reviewed the cases of 22 PWE who had attempted suicide. They concluded that there was a relationship between interictal pathology such as borderline personality disorder and psychosis and none related directly to the peri-ictal phenomenon (29). Of note, 54% of these patients were taking phenobarbital or primidone, a medication metabolized to phenobarbital.

Jones et al. (19) also reviewed the rates and risk factors for suicide and suicidal ideation and attempts in PWE. The authors found that the lifetime average rate was 12% in PWE as compared with 1.1% to 1.2% in the general population. The elevated risk appears to be present for children and adolescents as well. Although the existence of a mood disorder, in particular, in a MDD is important, the presence of psychosocial issues, personality, stress, physical health, and access to firearms are also important factors. In a multicenter study involving 139 patients evaluated by the use of the MINI, the rate of current suicidal ideation was 12.2% (n = 17) and the lifetime prevalence of suicide attempts was 20.8% (n = 29) (19).

One of the reasons for the lack of detection of depression in medical illnesses in general, and in PWE in particular, appears to be the time pressure that seems to be an omnipresent problem for all physicians in their medical clinics. The use of psychometric tools that are self-administered can help resolve this problem. These devices do not diagnose the presence of a MDD but identify symptoms and quantify severity that can be later categorized using DSM-IV criteria. Concerns have been raised about the valid use of these screening instruments in medical illnesses. These concerns stem from the fact that the most commonly used psychometric instruments like the Beck Depression Inventory (BDI) and the Hamilton Depression Rating Scale (HAMD) have almost a third of their criteria associated with somatic symptoms. These physical symptoms may be the result of the disease itself or of the medications used to treat the medical illness. Newer psychometric instruments have been devised to circumvent this problem by the removal of somatic questions coupled with standardization to a gold standard of depression identification, that is, the SCID and MINI. The six-item neurological disorders depression inventory for epilepsy (NDDI-E) was designed to identify major depression in PWE by minimizing the potential for being confounded by adverse AED side effects or cognitive problems associated with epilepsy (30). Psychometric testing in a sample of 205 PWE demonstrated that the association of major depression with the BDI and the CES-D, but not the NDDI-E, was significantly affected by medication toxicity, suggesting that the NDDI-E is a more accurate screening instrument for depression in epilepsy.

Despite this fact, Jones et al. have examined the validity of the BDI and the CES-D in PWE and compared it with MINI. Both psychometric tools had good correlation (31) and can apparently also be used with validity in this patient population.
In addition, AED medication side effects can be evaluated by another self-administered instrument, the Adverse Events Profile (32). Finally, the Mood Disorders Questionnaire (MDQ) has been used in PWE (refer to the study by Ettinger et al.) as a screen for the presence of symptoms consistent with a bipolar affective disorder (BAD) (33).

The approach to depression or manic symptoms in PWE must be evaluated in the context of their temporal relation to the ictal event. In 1828, Burrows wrote that “the epileptic attack may be preceded by a furious paroxysm, or merely by elevated ideas, by great depression … or the reverse” (3). In the 1860s, Falret crystallized these concepts by noting that psychiatric symptoms could be associated with the ictal event or the interval between attacks (3). Patients and their families will frequently note mood changes that herald the onset of an epileptic event. Blanchet and Frommer (34) evaluated this phenomenon in a study involving 27 patients over the course of 56 days. Thirteen patients experienced at least one seizure each during this period. Mood ratings declined 3 days before the event and, most significantly, 24 hours before it. A return to baseline took place over the next 3 days, but most rapidly 1 day postictally. Interestingly, one patient noted the reverse, that is, experienced elation instead of depression. More negative life events were also described prodromally (34).

Depression in the postictal period was evaluated by Kanner et al. in 100 patients (24). Forty-eight of these patients had a mean of 5 postictal depressive complaints out of a possible 12 depressive symptoms. In 13 of these patients, postictal suicidal ideation was noted. These symptoms persisted for 24 hours in two thirds of these patients. Interictal depressive symptoms were seen in 37 patients, and 7 of the 100 patients had postictal psychotic symptoms (24). It was postulated by the authors that postictal depressive symptoms might account for the atypical presentation of depression in this population. Postictal depressive symptoms may also be especially difficult to treat.

Several reports examined the phenomena of manic and depressive symptoms presenting as an ictal event. Williams (35) evaluated 2,000 PWE “living normal lives.” Of 100 patients who experienced ictal emotions, 65 noted ictal fear and 21 experienced a depression of varying severity (35). In eight patients, ictal depression persisted in the postictal period, lasting up to 3 days. Five patients developed suicidal ideation, with one completing suicide. Taylor and Lochery (36) evaluated simple partial seizures (auras) presenting as psychiatric symptoms in 88 patients and confirmed the frequency of fear, but not the complaints of depression. Ictal laughter (gelastic) and crying (dacrystic) also have been documented, with the former being more frequent than the latter (>100 cases reported in the literature) (37) and both phenomena characteristically presenting in association with CPS. These events usually take place independent of the patients’ surroundings and are nonprovoked and inappropriate (37).

Peri-ictal mania is unusual, but isolated case reports exist. Williams (35) found nine patients who experienced “pleasure” peri-ictally, three of whom noted “elation.” Barczak (38) reported three cases of manic symptoms postictally, all following a flurry of seizure activity. This was followed by four other letter-to-the-editor case reports with similar descriptions (39,40,41). All instances followed increased seizure activity, with a “normal” period and then, hours later, the appearance of manic-type symptoms. A similar pattern is seen in patients experiencing postictal psychotic events (42). All of these case reports found seizure activity to be localized to the right temporal lobe. A detailed discussion of postictal symptoms of depression can be found in Chapter 19.

Many reviews have noted an increased incidence of affective disorders in PWE. When
reviewing this literature, it is imperative to be aware of several factors. Of utmost importance is the sample population being evaluated, that is, is it representative of the entire population of PWE? Second, what types of seizure disorders were identified in the patients of the group being studied? For example, are they all patients with a temporal lobe focus, and is there a comparison or control group? Finally, what types of measures of depression are being used? Are they self-report or interview techniques, and can these measures be compared? Also, are there descriptions of the severity, symptom pattern, and longevity of the depression?

It is important to note that the assessment techniques described in the following sections, except for the Bear-Fedio Inventory and the NDDI-E, have been designed for the general population. As stated earlier, the affective disorder complaints of PWE may have some unique features especially in those patients with milder forms of the illness. It does appear that at least the BDI and the CES-D can be used with validity in PWE but other (22) contemporary measures of depression might present an erroneous picture of disease prevalence.

The actual incidence and prevalence of interictal depression in epilepsy remain uncertain, despite the numerous research studies addressing this issue. Comparing the results of one study to another is complicated by the diversity in methodologies and sample populations across studies. The ambiguity inherent in the diagnosis of epileptic subtypes is superimposed on the variability and ambiguity of depression measurement scales. Although the relevant literature is difficult to consolidate, the preponderance of evidence seems to favor a few basic conclusions. First, the incidence of depression in epilepsy is higher than in a matched population of healthy controls (range 11% to approximately 62%). Second, within a cohort of individuals with both epilepsy and depression, no specific subtype of epilepsy shows a higher incidence of depression than any other subtype, but a left-sided seizure focus may be important. Whether the incidence of depression in epilepsy is clearly elevated above those with other chronic illnesses, both neurological and non-neurological, has yet to be fully established as seen in the following text.

The diversity in methodologies in this body of work makes coherent synthesis of the data difficult. Assessment scales for depression range from self-reporting questionnaires (18,43,44,45,130), to objective measures of current but not past mood and anxiety states (44), to the use of the Clinical Interview Schedule (CIS) and/or DSM/RDC criteria, which incorporate both historical data and current mood states (17,44,47). Some investigators use “cutoff” scores to group depressed and nondepressed subjects. Others use the results of assessment scales although they represent continued variables and compare mean scores, often with considerable variance that might skew the data. The lack of uniformity in the assessment of depression across studies is complicated further by the wide spectrum of epilepsy subtypes. Whereas many investigators use single electroencephalographic (EEG) measurements to detect seizure focus, others rely on 24-hour EEG telemetry (45). Perhaps the biggest challenge to interpretation of the literature relates to the vast differences in sample populations. Most studies cull subjects from subspecialty neurology and epilepsy clinics, which in, and of, itself represent a skewed population. The variety of sample populations ranges from the community (48), to inpatient neurology units, female only (47), to psychiatric inpatient units (18), to medically intractable PWE scheduled for surgery (3,49). Meaningful comparison of results across studies is not trivial, given these vastly diverse sample populations.

According to the 1981 National Institute of Mental Health (NIMH) Epidemiological Catchment Area study, using DSM-III criteria the rate of depression in
the general population was determined to be 4.9% for MDD, with a 3.3% incidence of dysthymia (lifetime prevalence). The National Comorbidity Study was completed in 1996 and showed a greater lifetime frequency of 17% for MDD (20). The rate of depression in PWE reported in the literature ranges from a low of 11% with a current depression to approximately 62% with a lifetime-to-date depressive disorder (9,50). As previously discussed, Jones et al. found a current rate of MDD in PWE of 17.2%. (19). In a community sample of PWE compared to those with asthma and a no disease group, the rate of severe depressive symptoms as measured by the CES-D was 26.5%, 20.2%, and 5.2%, respectively (51). To clarify these findings further, we have looked only at controlled studies that compare epilepsy to well-defined comparison populations (Table 13.3).

At least five controlled studies found that depression in PWE is significantly increased as compared with healthy controls (46,51,62,63,130). In contrast, Fiordelli et al. (54), in a large, well-designed study, reported no increased risk of depression in PWE as compared with a population of matched controls undergoing minor outpatient surgery. One might validly argue that the control population studied by Fiordelli et al. was not adequately screened for major illnesses and so does not represent a healthy control.

When PWE are compared with patients who suffer from other neurological diseases, including traumatic brain injury, neuromuscular diseases, and multiple sclerosis, several controlled studies suggest that there is no difference in the rates of depression (64,65,66,67) between these groups. For the most part, both groups seem to be at increased risk of depression as compared with healthy controls. We know of only three controlled studies suggesting that PWE have a higher rate of depression than patients with nonepileptic neurological diseases (9,43,55).

Numerous studies compared depression in PWE with depression in individuals with chronic illnesses that are not of neurological origin. In support of psychosocial burden as a major risk factor for depression, most studies found equal rates of depression in these two groups (Table 13.3). There are few reports showing an increased incidence of depression in PWE as compared with other chronic illnesses (18,51,64). Mendez et al. (18) compared 175 outpatients with epilepsy with 70 nonepileptic controls attending a vocational center for disabled people. They reported depression in 55% of the epilepsy group and 30% of the control group. However, the study was based on a self-report questionnaire, which was returned by only 35% of the epilepsy group and 38% of the controls. Dodrill and Batzel (64) reviewed the literature and separated out neurological and non-neurological illness control groups for comparison. Interestingly, they found increased rates of depression in PWE as compared with those having non-neurological disorders, but equal rates of depression in epilepsy and neurological disorders (64). As noted previously, Ettinger et al. (51) also noted increased rates in PWE compared to those with asthma and a no-disease group. In summary, depression occurs more commonly in the epilepsy population as compared with healthy populations, but whether there is something integral to the biology of epilepsy and other neurological diseases acting through a common physiological pathway that predisposes to depression or whether psychosocial factors play the dominant role common to other chronic illnesses continues to elude investigation.

During the past 15 years, multiple studies have evaluated aspects of health-related quality of life in epilepsy. Most investigations have found that achieving complete seizure cessation offers the greatest gains in quality of life, but also that relative differences in complex, partial, or generalized seizures short of total seizure freedom are not associated with significant differences in subjective health status (73,74,75). These data suggest that other epilepsy-related factors may be critically important targets for interventions intended to improve quality of life.

Multiple subsequent studies have confirmed the importance of mood for optimal quality of life in epilepsy (76). Coefficients for the correlations of severity of depression symptoms with the Quality of Life in Epilepsy Inventory are consistently 0.6 to 0.8 (75,77). Anxiety may have as great an influence as depression on quality of life, but has received less attention (77). Although reduced mood might be anticipated in the setting of poor quality of life in epilepsy, there is minimal evidence to support that the direction of causality is from poor quality of life toward depression. In fact, most studies have not found an association of social and vocational status with mood in epilepsy (78). Furthermore, multiple studies including neuroimaging have found association of markers of brain injury or dysmetabolism with depression in epilepsy, suggesting a significant role of limbic system dysfunction (refer to following discussion on neurobiological risk factors). For additional discussion of mood and quality of life in epilepsy, refer Chapter 29.

The evidence favors no unique association between depression and epilepsy subgroups with respect to seizure type, frequency, seizure duration, and age at onset (17,43,57,79,130). Much enthusiasm has focused on TLE as a possible risk factor for depression, because it topographically subsumes the limbic system that is intimately involved in affect and mood regulation. Bear and Fedio (46) distilled their epilepsy population into those with TLE and found an increased risk of depression in TLE when compared with nonepileptic neurological diseases. However, there is no overwhelming evidence to show a difference between TLE and other types of epilepsy (Table 13.3). It may be that the increased rate of psychiatric disorders seen in patients with TLE simply represents the increased frequency of TLE in contrast to other seizure types (81). In addition, TLE with secondary generalization is often miscategorized and therefore may be a significant confounding variable (9).

Laterality of the seizure focus has been evaluated as a risk factor for depression in epilepsy. Using 24-hour EEG monitoring on a female subject with TLE and agitated depression, Hurwitz et al. (82) described left-sided EEG discharges with depressed affect and right-sided discharges with laughter and seductive behavior. It has been postulated that the dominant hemisphere is responsible for positive emotional states, with the nondominant hemisphere displaying the opposite effect. Seizure activity in one hemisphere might “release” the contralateral hemisphere (81). Another theory postulates that nondominant hemispheric activity may result in denial and neglect of negative emotions (81). Several controlled studies comparing seizure foci and rates of depression found increased rates of depression with left-sided foci, independent of seizure type, whereas other well-done studies found no correlation (Table 13.3). The most likely explanation argues for complex factors.

Some recent work suggests that clarifying the phenomenology of depression and epilepsy may have less to do with the risk factors just mentioned and more to do with finding a possible link between epilepsy and frontal lobe dysfunction and hypometabolism. Positron emission tomography (PET) in particular, and single photon emission computed tomography (SPECT) in psychiatric patients with depression, showed frontal hypometabolism, especially in the left prefrontal cortex (20).

Similarly, PWE have been studied using these modalities. Both PET and SPECT have been utilized for localization of a seizure focus before anterior temporal lobectomy (ATL). Bromfield et al. (60) evaluated 23 patients with CPS and at least mild depressive features (BDI >11) and those without symptoms as compared with normal controls. Results of this study showed that patients with a left temporal lobe focus were associated with more depressive features and with bilateral inferior frontal hypometabolism. Of note are similar findings in Parkinson’s patients with depression, which perhaps explains the uniformity in rates of depression between the two patient populations (20).

Victoroff et al. (49) looked at 53 medically intractable PWE scheduled for surgery and used standardized measures to assess for lifetime history of depression and current mood states (DSM-III-R, Structured Clinical Interview for Diagnosis-P, HAMD). They then used EEG telemetry and fluorine-18 (¹⁸F) PET scans to assess seizure laterality and frontal lobe hypometabolism. They found that left ictal onset was associated with greater frequency of depression: 79% versus 50% (nonsignificant). No correlation was found between current mood state and hypometabolism, but, interestingly, a history of depression was significantly correlated with left frontal lobe hypometabolism.

Along the same lines, Hermann et al. (45) used the Wisconsin Card Sorting Test as a cognitive measure of frontal lobe function. They found that, although there
was no correlation with mood and laterality, a left-sided seizure focus was significantly correlated with the degree of frontal lobe dysfunction and dysphoria. In contrast, a right-sided focus was nonsignificantly inversely related to frontal lobe dysfunction and dysphoria. These two studies, along with that of Bromfield et al. (60), suggest that the etiological underpinnings of depression in epilepsy may be related to associated frontal lobe hypometabolism, analogous to that seen in patients with idiopathic depression. Certain types of seizures may induce frontal lobe dysfunction and hypometabolism directly, or they may ignite a kindling phenomenon that leads to subsequent depression when other risk factors for depression are present. Therefore, the interplay between factors is complex.

Although hippocampal volume loss has been reported in persons with depression (83), few studies have evaluated this association in epilepsy. Quiske et al. found higher BDI scores in patients with TLE with mesial temporal sclerosis as compared with those with normal magnetic resonance imaging (MRI) (84). Another investigation identified an association of higher depression scores with larger left hippocampal volume in persons with right hippocampal sclerosis (85). Temporal lobe abnormalities on fluorodeoxyglucose-positron emission tomography (FDG-PET) imaging have also been reported to be associated with higher depression scores as compared with PWE with normal FDG-PET (14).

Another possible etiological factor concerns the observation of “paradoxical” or “forced” normalization. This phenomenon was first described in 1953 by Landolt, who was astonished to find that behavioral disturbance seemed to be associated with improvements on the patient’s electroencephalograph. The behavioral changes were originally noted to be in the form of psychotic symptoms and hence were referred to as “alternative psychosis” (86). Therefore, as seizures and EEG findings returned to an improved state, behavior deteriorated. It should be noted that several of these patients were taking ethosuximide, which may have been responsible for the behavioral disturbances (87). Wolf (86) reported four cases presenting as affective disorders: two depressive and two manic. The pathogenesis of this phenomenon has focused on the interplay of dopamine (DA), glutamate, and γ-aminobutyric acid (GABA). Although more frequently associated with ethosuximide, other AEDs have been implicated, often in patients with temporal lobe localization–related epilepsies (87). Forced normalization has been reviewed by Krishnamoorthy et al. (87) and Trimble et al. (88) and more recently by Janz (89) and Schmitz (90). This concept is an issue of hot debate. Clinically, affective symptoms that are directly related to the cessation of seizure activity are not seen frequently (also debated).

ATL for intractable TLE is highly effective in a select sample of patients, with up to 80% of patients being seizure-free and with less than 5% morbidity and mortality (91). However, a significant number of patients may develop postoperative psychosis, depression, and anxiety (92). Jensen and Larsen (93) noted that depressive symptoms usually occur acutely, with six of their patients (n = 72) committing suicide in the first postoperative month. Bruton (94) reviewed 249 patients who underwent ATL and noted that only one was depressed preoperatively (0.4%). After surgery, depression was found in 24 (10%) patients, with 6 committing suicide. This contrasts with several other studies that noted high frequencies of depression in patients being evaluated for ATL (95,96). For example, Altshuler et al. (1999) reviewed the results of 49 surgical patients (59). Of the total group, 77% had a prior history of depression and of these, 46% had no further episodes of depression after surgery. Only 10% (n = 5) developed a depression de novo and four out of five developed a depression within 1 year. The conclusion of the authors was that a prior depression is not a contraindication to ATL. Kohler et al. (2001) noted that patients with preoperative fear auras, after ATL, had a higher incidence of depression and anxiety disorders than controls, especially if seizure-free (97). Blumer et al.
(1998) evaluated the surgical outcome in 44 patients who underwent ATL, and 57% (n = 25) had IDD before surgery (98). After an ATL, of the 44, 39% (n = 17) had either a de novo worsening (six psychotic, six dysphoric and two depressive disorders) or an exacerbation of a preexisting dysphoria (n = 3). Most occurred in the first 2 months after surgery. All responded to psychotropic medication. Glosser et al. (2000) reviewed 39 patients who underwent an ATL (99). Of this group, 65% had an Axis 1 diagnosis before and after surgery. There was a trend association between right anterior temporal lobectomy (RT ATL) and more psychopathology both before and after ATL. Of note was the observation that 31% had developed new psychiatric problems and 15% had a resolution. Psychopathology was lower at 6 months post-ATL than before surgery. Quigg et al. (100) used the Minnesota Multiphasic Personality Inventory (MMPI) (n = 90) and concluded that patients undergoing an RT ATL had a higher susceptibility to postsurgical depression (100). Finally, Carran et al. (101) reviewed their results of patients undergoing an ATL over a 5-year period (101). There were 16 patients who developed manic episodes within the first year of ATL. These patients were compared with those who developed a depression after ATL and those without a psychiatric illness. Factors of importance were bitemporal EEG abnormalities and RT ATL. Manic episodes were transient. Psychosocial issues are important as well in people undergoing ATL but will be discussed in the next section.

Neurochemical risk factors for depression in PWE are also important. By an extrapolation from the literature on depression in patients with an idiopathic depression, some clues into the relationship of depression and epilepsy might ensue. In the 1960s, American and European researchers introduced the concept of biogenic amine depletion as a cause of depression (20). The major neurotransmitters that have been implicated are norepinephrine (NE), serotonin (5-hydroxytryptamine [5-HT]), and finally DA. Other important neurotransmitters in epilepsy and possibly depression are GABA and glutamate. Changes in the neuroendocrine system and neuropeptides may also play a role. Epilepsy may precipitate unique biologic changes modulating these endogenous factors, resulting in the features of an endogenous depression.

Depletion of NE and 5-HT with consequent up-regulation of the postsynaptic receptor is of interest in PWE. Elevation of both of these neurotransmitters has been noted to increase the seizure threshold (102,103). 5-HT depletion has been postulated to be involved in the pathophysiology of epilepsy, but the exact mechanism is uncertain (103).

Both mood disorders and epilepsy are associated with episodic discontrol type behavior. In addition, 5-HT (emanating from the dorsal raphe nucleus) and NE (produced from the locus ceruleus) have widespread projections to pivotal areas in the cortex (104). Depletion of these neurotransmitters appears to be associated with depression. For example, the examination of brain tissue from patients who had committed suicide showed depleted stores of 5-HT as well as an increase in the suppressor autoreceptor 5-HT_1A in the dorsal raphe nucleus (105). In addition, arousal and the modulation of stress appear to be modulated by NE. The platelets of patients who have committed suicide show elevated levels of the suppressor autoreceptor α₂ adrenergic receptor, which suggests a depletion of NE in the central nervous system (CNS) (105). Deficient CNS NE arborization has also been seen in the brain tissue of patients with MD, evaluated postmortem (106).

Deficits of 5-HT and NE have been found in animal models and in humans with epilepsy. In the genetically epilepsy-prone rat (GEPR), depressive symptoms and seizure predisposition coexist. In the GEPR, depletion of 5-HT and noradrenergic arborization in the locus ceruleus are coupled with an increase in 5-HT and NE autoreceptor inhibition (105) and a subsequent decrease in both neurotransmitters. The 5-HT_1A autoreceptor appears to have opposing effects depending on its location
with a resultant inhibitory effect in the raphe nucleus and a stimulatory role in the hippocampus (103). Two strains of the animals, type 3 and 9, show noradrenergic and serotonergic pre- and postsynaptic transmission defects. An increase of either NE or 5HT decreases seizure activity with a depletion having the opposite effect (24). In other epilepsy-prone animal species, the administration of selective serotonin reuptake inhibitors (SSRIs) or monoamine oxidase inhibitors (MAOIs) can be anticonvulsant (24).

In addition, by stimulation of the 5-HT_1A receptor on hippocampal neurons, 5-HT may have anticonvulsant activity. Several studies have also shown an in vivo decrease of 5-HT_1A binding in PWE when compared with controls and this depletion was highly correlated with the apparent focus of epileptogenesis (107,108,109). Finally, 5-HT appears to be necessary for the antiseizure activity of carbamazepine, and monoamine increases are also seen with the use of AEDs like lamotrigine (LTG), zonisamide, and valproate (VPA). Further evidence for the bidirectional relationship of epilepsy and depression (which will be discussed later) comes from studies involving reserpine that causes monoamine storage vesicle inactivation and apparent depressive symptom production and the exacerbation of seizures in GEPR rats. Common brain dysfunction clearly seems to be apparent in epilepsy and depression.

Clearly, the monoamine hypothesis of depression does not provide the entire explanation for depression onset. Complex intracellular functions are determined by a vast array of intracellular signaling pathways. 5-HT and NE help regulate cyclic adenosine monophosphate (cAMP). The activation of G-proteins catalyzes a pathway that results in cAMP production, subsequent protein kinase production, and modulation of transcription activity. Other neurotropic factors may be of importance including brain-derived neurotropic factor (BDNF). Hippocampal atrophy may result from prolonged depression and may be reversed by antidepressants, perhaps by BDNF activation. Finally, stress with the production of glucocorticoids and resultant decrease of BDNF may be important. Glucocorticoids appear to damage the hippocampus; especially the CA3 region and corticotropin-releasing factor may cause depression. In addition, steroid antagonists-like mifepristone are being evaluated as antidepressants especially in patients with psychotic depression where steroid elevation is most prominent (104,110,111,112) These observations have led to the development of a neurotropic hypothesis of depression, especially for stress-related mood changes. This model notes the association of atrophy and subsequent death of neurons, especially the CA3 pyramidal cells with a down-regulated neurogenic response in the hippocampus (113). PET data noted in the preceding text underscores the relationship between depression and epilepsy (84,112). The hypothalamic–pituitary axis (HPA) may be significantly affected by epilepsy. Interictal neurohormonal changes in epilepsy include alterations of prolactin and gonadotropins (114). Some of these effects may be the result of AEDs themselves, either working directly on the brain or by secondary effects on sex hormone binding globulin (114). Depression can likewise cause a variety of neurohormonal aberrations, especially when associated with psychotic features. In unipolar depression, increases of corticotropin-releasing factor, adrenocorticotropic hormone (ACTH), and abnormal thyroid dysfunction, especially hypothyroidism, occur along with unclear changes in the hypothalamic–pituitary–gonadal system (20). It seems possible that the relationship between epilepsy and depression could hinge on these neurotransmitter changes. The observation of increased epileptogenesis with increasing dosages of antidepressants is an interesting one, considering the role of serotonin and NE in the generation of seizures. Except for the tottering mouse model, in all animal studies, NE attenuates seizure occurrence. Serotonin also appears to be responsible for increasing the seizure threshold (103). The expectation, therefore, would be a decrease in seizures with the use of antidepressants. In fact, this has been the
case in the studies of Ojemann (described later in the chapter) and Favale et al. (115), who studied 17 PWE who had fluoxetine 20 mg per day added to their AED regimen. Six patients had a complete remission of their seizure activity, and the remaining patients were noted to have a 30% reduction of seizure frequency (115). In a confirmatory animal study, monoamine depletion through reserpine had no effect on seizure incidence associated with increasing dosages of desipramine in rats (103). Serotonin depletion blocks the anticonvulsant properties of fluoxetine; conversely, the augmentation of fluoxetine with 5-hydroxytryptophan increases the seizure threshold (103).

The role of DA in epilepsy may be similar to the role of neurotransmitters noted earlier, but it is far from clear. DA antagonists, like the phenothiazines, lower the seizure threshold, whereas DA agonists frequently increase the threshold (103,116).

GABA and glutamate levels may be of particular importance in the development of unipolar depression. GABA is one of the most important neurotransmitters in the brain. It acts at two types of GABA receptors, A and B. One of the major modes of seizure control consists of pharmacological augmentation of GABA at the A receptor site. Many antimanic agents increase GABA, and low GABA levels have been found in patients with depression (117). GABA may interact with NE and serotonin and thereby exert an impact on depression (24). GABA levels may be elevated by both SSRIs and by electroconvulsive therapy (ECT) (118). However, several drugs, such as phenobarbital and vigabatrin, which are GABA agonists, are associated with affective symptoms. Therefore, the actual pathogenic mechanisms of GABAergic effects are yet to be identified. A complex interplay may be operant.

The role of excitatory neurotransmitters and the glutamate receptor in the genesis of epilepsy has been the focus of much research (119). This receptor site is an extremely complex macromolecule with an N-methyl-D-aspartate-aspartate (NMDA) site, which also may be important in depression. NMDA antagonists have antiepileptogenic (i.e., they block “kindling” of epileptic activity in animals) and antiepileptic properties. Although there is sparse data on the interplay of glutamate in the pathogenesis of depression, NMDA antagonists also possess antidepressant qualities and can be mildly euphorogenic in nondepressed patients (120). In addition, adaptive changes are seen in the NMDA receptor site with chronic administration of most antidepressants and may predict antidepressant response (121). Therefore, modulation of the NMDA receptor site may be yet another area where the presence of epilepsy may create an environment that predisposes to neurochemical changes that result in the phenomenon of depression (120).

Finally, folic acid deficiency is yet another potential risk factor for the development of depression in epilepsy. Folate is involved in one-carbon transfer reactions important in the synthesis of biological macromolecules, including neurotransmitters (4). Folate depletion has been observed in some PWE taking AEDs such as VPA and carbamazepine with a resultant elevation of a metabolite from the demethylation of methionine to homocysteine. The role of folic acid depletion in psychiatric disorders had been noted in several studies, with the gradual onset of depression, psychosis, and, finally, dementia (122,123). One review quoted the incidence of folate deficiency in 21% (124) of psychiatric inpatients and another a rate of 30% as compared with a rate of 2.5% in a control group (125). PWE are especially prone to folate deficiency. Phenobarbital and phenytoin, in particular, deplete red blood cell, serum, and cerebrospinal fluid folate levels. Rösche et al. noted a negative correlation between folate levels and scores on the Self-Rating Depression Scale in 46 patients with chronic epilepsy (126). Whether associated depression with folate reduction is purely an epiphenomenon is unclear. Charney (124) and Froscher et al. (127) noted improvement in patient mood with folate replacement, as did Reynolds et al. (122) in an epilepsy group. Papakostas et al. noted an association
between low folate and depressive relapse in a group with MDD (128). Double-blind placebo-controlled studies with folic acid replacement, however, have not confirmed a direct association (48). This finding is in contrast to similar investigations with methylfolate, which showed clinical improvement (4). A systematic review and meta-analysis of randomized trials by Taylor et al. concluded that there is a potential role of folate as a supplement to other treatments for depression. Therefore, an association between folate deficiency and psychiatric dysfunction in PWE is clear, but a causal link is uncertain (129).

The impact of seizure frequency on patients’ quality of life is an important factor that mediates much of the psychopathology seen in PWE (116,130). The psychopathogenic roles of psychosocial factors in epilepsy can be appreciated in patients undergoing surgery for epilepsy. For example, levels of “learned helplessness,” that is, the feeling that one is powerless to control external events and depression before surgery have been shown to be associated with higher levels of dysfunction postsurgically (131). The relationship between external and internal loci of control appears to be important. External locus of control refers to a feeling that the individual is powerless to influence outside events. In contrast, having an internal locus of control is associated with a feeling of being able to positively influence the world around you. Depression prior and subsequent to ATL—was evaluated by Hermann and Wyler (132). They found a correlation of locus of control preceding but not following surgery, with complete seizure cessation being the most pivotal factor in outcome. This finding is somewhat in conflict to our clinical observations. Patients who have never adjusted to having epilepsy, especially those who are overprotected and become socially agoraphobic, are often presented with a frightful period of postsurgical adjustment and present with an external locus of control. These patients and their families require close psychiatric follow-up and are more prone to depression, divorce, and interpersonal dysfunction. Unclear biological factors are also important and require further investigation.