Epilepsy and Psychiatric Comorbidity: Depression, Anxiety, Psychosis, Cognitive Effects, and Evidence-Based Treatment Considerations

Epilepsy is among the most common serious neurological conditions, affecting approximately 50 million people worldwide according to the World Health Organization, with an estimated lifetime prevalence of 7.6 per 1,000 persons. While seizure control remains the primary treatment objective, the field has increasingly recognized that psychiatric comorbidity is not a peripheral concern—it is a central determinant of quality of life, treatment adherence, seizure outcomes, and mortality risk in people with epilepsy (PWE). Large population-based studies consistently demonstrate that psychiatric disorders occur in 30–50% of PWE, a rate roughly two to three times higher than in the general population.

The relationship between epilepsy and psychiatric illness is not simply one of psychological reaction to chronic disease. Landmark epidemiological work, including the study by Hesdorffer and colleagues (2012) using Swedish national registry data, demonstrated that a history of depression increases the risk of developing epilepsy by approximately 1.7-fold—establishing a bidirectional relationship that implicates shared neurobiological substrates rather than a simple cause-and-effect model. This bidirectionality has been replicated for anxiety disorders and psychosis, fundamentally reframing how clinicians should conceptualize the epilepsy–psychiatry interface.

Despite the magnitude of this burden, psychiatric comorbidity in epilepsy remains systematically underdiagnosed and undertreated. Studies using structured diagnostic interviews reveal that 50–60% of depression in PWE goes unrecognized in routine neurological care. This treatment gap carries measurable consequences: untreated depression in epilepsy is associated with poorer seizure control, increased healthcare utilization, higher rates of adverse effects from antiseizure medications (ASMs), elevated suicidality, and a roughly threefold increase in all-cause mortality. This article provides a detailed clinical review of the epidemiology, neurobiology, diagnostic challenges, and treatment evidence for the major psychiatric comorbidities of epilepsy—depression, anxiety, psychosis, and cognitive impairment.

The epidemiology of psychiatric comorbidity in epilepsy has been established through multiple large-scale population studies and meta-analyses, though prevalence estimates vary depending on the population studied (community vs. tertiary care), the assessment methodology (screening instrument vs. structured diagnostic interview), and the epilepsy subtype.

Depression

Depression is the most extensively studied psychiatric comorbidity in epilepsy. A meta-analysis by Fiest et al. (2013), encompassing 51 studies and over 80,000 PWE, reported a pooled lifetime prevalence of major depressive disorder (MDD) of approximately 23.1% (95% CI: 20.6–28.3%) in community samples and up to 30–35% in tertiary epilepsy centers. Point prevalence estimates using validated screening tools such as the Neurological Disorders Depression Inventory for Epilepsy (NDDI-E) typically range from 20–30%. The incidence of new-onset depression in PWE is estimated at 7–22 per 1,000 person-years, compared to approximately 4–8 per 1,000 in the general population. The odds ratio for depression in PWE versus the general population is consistently reported between 2.0 and 3.0.

Anxiety Disorders

Anxiety disorders are at least as common as depression in epilepsy but have received considerably less research attention. The meta-analysis by Scott et al. (2017) reported pooled prevalence estimates of 20.2% for any anxiety disorder in PWE, with generalized anxiety disorder (GAD) being the most common subtype (approximately 10–15%), followed by social anxiety disorder (5–8%) and panic disorder (5–7%). The odds ratio for anxiety disorders in PWE versus the general population is approximately 2.4 (95% CI: 1.6–3.6). Notably, comorbid depression and anxiety co-occur in roughly 40–60% of PWE who have either condition, creating a particularly treatment-resistant phenotype.

Psychosis

Psychotic disorders occur at elevated rates in epilepsy, though with lower absolute prevalence than mood and anxiety disorders. The classic estimate, drawn from Clancy and colleagues and refined by subsequent systematic reviews, places the prevalence of interictal psychosis at approximately 5–7% in epilepsy populations, roughly 6–12 times higher than the approximately 1% general population rate of schizophrenia-spectrum disorders. Postictal psychosis occurs in approximately 2–7.8% of refractory epilepsy patients, while ictal psychosis is rarer and may be underrecognized. Temporal lobe epilepsy (TLE) carries the highest risk, particularly with left-sided or bilateral foci.

ADHD and Neurodevelopmental Conditions

Attention-deficit/hyperactivity disorder affects an estimated 12–40% of children with epilepsy, compared to approximately 5–7% in the general pediatric population. In adults with epilepsy, ADHD prevalence is estimated at 12–20%. Autism spectrum disorder co-occurs with epilepsy at rates of approximately 20–30% in individuals with intellectual disability.

Suicidality

The risk of completed suicide in PWE is approximately 3–5 times that of the general population, and up to 25 times higher in temporal lobe epilepsy with comorbid psychiatric illness. The standardized mortality ratio (SMR) for suicide in epilepsy has been estimated at 3.3 (95% CI: 2.8–3.7) in a large meta-analysis by Bell et al. (2009). This formed part of the backdrop for the controversial 2008 FDA black-box warning regarding suicidality and antiseizure medications, based on a meta-analysis of 199 trials showing an odds ratio of 1.8 for suicidal ideation or behavior with ASMs versus placebo—though the absolute risk increase was small (0.43% vs. 0.24%).

The bidirectional relationship between epilepsy and psychiatric disorders points to shared neurobiological vulnerabilities operating at multiple levels—neurotransmitter systems, neural circuits, neuroinflammatory pathways, and genetic architecture.

Serotonergic and Noradrenergic Dysfunction

Serotonin (5-HT) and norepinephrine (NE) are centrally implicated in both seizure susceptibility and mood regulation. Animal models of epilepsy demonstrate consistent reductions in serotonergic transmission, particularly in the raphe-hippocampal and raphe-amygdalar projections. In the genetic epilepsy-prone rat (GEPR), a well-established model, deficits in 5-HT1A receptor density in the hippocampus and amygdala precede both seizure onset and depressive-like behavior—suggesting that serotonergic dysfunction is not merely a consequence of seizures but a shared predisposing factor. In human PET studies, reduced 5-HT1A binding has been demonstrated in the mesiotemporal structures of patients with TLE and comorbid depression (Hasler et al., 2007).

The locus coeruleus-norepinephrine system plays a dual role: norepinephrine is a potent endogenous anticonvulsant (NE depletion lowers seizure threshold in virtually all animal models), and NE deficiency is a core pathophysiological feature of depression. This convergence provides a mechanistic explanation for why depression and epilepsy can potentiate each other.

GABAergic and Glutamatergic Imbalance

The excitation-inhibition (E/I) imbalance central to seizure generation—specifically, reduced GABAergic inhibition and/or enhanced glutamatergic excitation—also features prominently in the neurobiology of anxiety and depression. CSF and magnetic resonance spectroscopy (MRS) studies in PWE with depression show reduced GABA concentrations in the prefrontal cortex and anterior cingulate cortex. Conversely, glutamate excitotoxicity in the hippocampus, a hallmark of mesial temporal sclerosis, contributes to both epileptogenesis and hippocampal dysfunction underlying depressive cognition and memory impairment.

Hypothalamic-Pituitary-Adrenal (HPA) Axis Dysregulation

Chronic HPA axis hyperactivation is well-documented in both epilepsy and depression. Elevated cortisol levels promote hippocampal neuronal loss, impair neurogenesis in the dentate gyrus, and lower seizure threshold. In PWE, interictal hypercortisolism correlates with both seizure frequency and depressive symptom severity. This creates a pathological feedback loop in which seizures elevate cortisol, cortisol worsens depression and lowers seizure threshold, and depression—through behavioral mechanisms such as sleep disruption and medication nonadherence—further worsens seizure control.

Mesiotemporal and Prefrontal Circuit Dysfunction

The mesiotemporal structures—hippocampus, amygdala, and entorhinal cortex—are the most epileptogenic regions of the brain and are simultaneously critical nodes in the limbic circuitry of mood regulation. Structural MRI studies consistently demonstrate that hippocampal volume loss in TLE correlates with both seizure burden and depression severity. Functional connectivity analyses using resting-state fMRI reveal disrupted connectivity between the amygdala and prefrontal cortex in PWE with depression, paralleling findings in primary MDD. The default mode network shows abnormal hyperconnectivity in epilepsy patients with comorbid depression, suggesting shared circuit-level pathology.

Neuroinflammation

Emerging evidence implicates neuroinflammatory pathways as a shared mechanism. Interleukin-1β (IL-1β), interleukin-6 (IL-6), and tumor necrosis factor-alpha (TNF-α) are elevated in both epileptic brain tissue and in the peripheral blood of patients with depression. IL-1β in particular is proconvulsant (acting via IL-1R1 on hippocampal neurons to enhance glutamate release and reduce GABA currents) and prodepressant (activating the indoleamine 2,3-dioxygenase [IDO] pathway, which shunts tryptophan away from serotonin synthesis and toward neurotoxic kynurenine metabolites). The kynurenine pathway has become a major research focus, with elevated quinolinic acid (an NMDA receptor agonist) found in the CSF of patients with epilepsy and comorbid depression.

Genetic Overlap

Genome-wide association studies (GWAS) have identified shared genetic risk variants between epilepsy and psychiatric disorders. Variants in genes encoding sodium channels (SCN1A, SCN2A), calcium channels (CACNA1A), and synaptic proteins (SV2A, the target of levetiracetam) are implicated in both epilepsy and mood/psychotic disorders. Common variants in the major histocompatibility complex (MHC) region on chromosome 6, long associated with schizophrenia risk, have also been linked to certain epilepsy subtypes. Copy number variants (CNVs) at 15q13.3, 16p13.11, and 22q11.2 are risk factors for both epilepsy and psychosis.

Depression in epilepsy presents with distinctive phenomenological features that can complicate recognition and diagnosis. Blumer's concept of the "interictal dysphoric disorder" (IDD), while not formally codified in DSM-5-TR or ICD-11, describes a pleomorphic affective syndrome characterized by intermittent irritability, depressive moods, anergia, insomnia, anxiety, and pain—often waxing and waning over hours to days rather than following the persistent pattern of a classic major depressive episode. Epidemiological studies using structured interviews suggest that approximately 30–50% of depression in epilepsy does not neatly conform to standard DSM-5-TR criteria for major depressive disorder or persistent depressive disorder (dysthymia), which creates a significant detection gap when using standard diagnostic frameworks.

Diagnostic Tools and Screening

The Neurological Disorders Depression Inventory for Epilepsy (NDDI-E) is the most widely validated epilepsy-specific screening instrument. A 6-item self-report scale, it has demonstrated sensitivity of 81% and specificity of 90% at a cutoff score of >15 for detecting major depression in epilepsy, and it avoids somatic items that overlap with ASM side effects or seizure symptoms (a problem with the PHQ-9 and BDI-II in this population). The International League Against Epilepsy (ILAE) Commission on Neuropsychiatry recommends routine screening for depression at epilepsy diagnosis, at treatment changes, and at annual reviews.

Differential Diagnostic Pitfalls

• ASM-related mood effects: Barbiturates, topiramate, levetiracetam, perampanel, and vigabatrin are associated with significant depressogenic effects. Levetiracetam-induced behavioral adverse effects (irritability, aggression, depression) occur in approximately 10–16% of patients and are a leading cause of drug discontinuation. Conversely, lamotrigine, valproate, and carbamazepine have mood-stabilizing properties and may improve comorbid depression.
• Peri-ictal mood symptoms: Preictal dysphoria can occur hours to days before a seizure, and postictal depression—sometimes severe and including suicidal ideation—occurs in approximately 40–50% of patients with refractory epilepsy. These temporally linked mood changes must be distinguished from interictal depression, as treatment implications differ.
• Iatrogenic factors: Post-surgical depression occurs in approximately 20–30% of patients after temporal lobectomy, including in patients who achieve seizure freedom—a phenomenon sometimes termed the "burden of normality," reflecting adjustment difficulties when seizures cease after years of illness identity.
• Forced normalization/alternative psychosis: Rarely, effective seizure control (particularly through ASM escalation) can be followed by the emergence of depression or psychosis—a phenomenon termed forced normalization, described originally by Landolt (1953). This represents a crucial diagnostic consideration when new psychiatric symptoms emerge coincident with EEG improvement.

Impact on Epilepsy Outcomes

Comorbid depression is not merely an epiphenomenon—it actively worsens epilepsy outcomes. The landmark study by Kanner et al. (2012) demonstrated that pretreatment depression was the strongest predictor of pharmacoresistance to antiseizure medications, stronger than seizure type, epilepsy syndrome, or MRI findings. Depression is also the strongest independent predictor of reduced quality of life in epilepsy, outweighing seizure frequency, seizure severity, and ASM side effects in multivariate analyses.

Anxiety Disorders

Anxiety in epilepsy is phenomenologically heterogeneous and includes anticipatory anxiety about seizures, phobic avoidance of seizure-triggering situations, generalized worry, and panic-like symptoms that can be either ictal (arising from seizure activity, particularly from amygdalar foci) or interictal. The distinction between ictal panic and panic disorder is clinically critical: ictal panic arises from amygdalar or insular seizure discharges, typically lasts 30–120 seconds, may be accompanied by automatisms or impaired awareness, and does not respond to anxiolytics—whereas interictal panic disorder follows the standard phenotype and treatment approach. Ictal fear is estimated to be the aura symptom in approximately 10–15% of temporal lobe seizures.

GAD is the most common interictal anxiety disorder, but its phenomenology in epilepsy is often contaminated by realistic health-related worries (fear of seizures in public, fear of SUDEP, concerns about driving restrictions), making it difficult to apply standard DSM-5-TR criteria that require "excessive" worry. Social anxiety disorder is particularly prevalent, driven by the stigma and unpredictability of seizures. Obsessive-compulsive disorder occurs at approximately 2–3 times the general population rate, particularly in patients with frontal lobe epilepsy.

Psychosis of Epilepsy

Psychotic phenomena in epilepsy are classified temporally relative to seizures:

• Ictal psychosis: Psychotic symptoms arising directly from seizure activity, typically during non-convulsive status epilepticus or complex partial status. Characterized by paranoid ideation, hallucinations (often visual), and disorganized behavior concurrent with EEG seizure activity. Treatment is seizure termination, not antipsychotics.
• Postictal psychosis (PIP): Occurs after a cluster of seizures, typically after a lucid interval of 24–72 hours. Presents with paranoid delusions, auditory and visual hallucinations, grandiosity, and sometimes affective symptoms. Duration is typically 2–14 days, and it is usually self-limiting. PIP affects approximately 2–7.8% of patients with refractory epilepsy and is a risk factor for developing chronic interictal psychosis (approximately 14–25% of patients with PIP eventually develop interictal psychosis). Risk factors include bilateral seizure foci, temporal lobe epilepsy, and intellectual disability.
• Interictal psychosis: A chronic psychotic disorder occurring independently of seizure timing. Clinically resembles schizophrenia but often features better preservation of affect and personality (the so-called "schizophrenia-like psychosis of epilepsy"), more visual hallucinations, more religious/mystical content, and less negative symptoms. The latency from epilepsy onset to psychosis onset is typically 10–15 years, a phenomenon described as the Slater latency.
• Forced normalization / alternative psychosis: Emergence of psychotic (or depressive) symptoms when seizures are suppressed, associated with EEG normalization. Clinically challenging because it may require tolerating some seizure activity or adjusting ASMs.

Cognitive and Behavioral Comorbidity

Cognitive impairment is pervasive in epilepsy, affecting an estimated 70–80% of patients with refractory epilepsy and 20–50% of patients with well-controlled seizures. The most commonly affected domains are episodic memory (particularly in TLE, reflecting hippocampal dysfunction), attention and processing speed, and executive function (particularly in frontal lobe epilepsy). Accelerated cognitive decline occurs in a subset of patients and may reflect the cumulative effects of seizures, interictal epileptiform activity, ASM cognitive toxicity (particularly topiramate, phenobarbital, and benzodiazepines), and comorbid depression. Depression itself independently worsens cognitive performance in epilepsy, creating a bidirectional relationship between mood and cognition.

Personality changes have been historically described in chronic TLE (Geschwind syndrome: hypergraphia, hyperreligiosity, altered sexuality), though the specificity and prevalence of these features remain debated. Aggression and irritability are common, affecting approximately 20–30% of PWE, and may be interictal traits, peri-ictal phenomena, or ASM adverse effects (notably levetiracetam and perampanel).

The treatment of psychiatric comorbidity in epilepsy has been historically hampered by the pervasive—and largely unfounded—fear that antidepressants lower seizure threshold. This concern, rooted primarily in early case reports with tricyclic antidepressants (TCAs) at supratherapeutic doses and the established proconvulsant effect of bupropion at high doses, has led to widespread therapeutic nihilism. However, contemporary evidence strongly supports the safety and efficacy of most modern antidepressants in epilepsy.

Selective Serotonin Reuptake Inhibitors (SSRIs)

SSRIs are the first-line treatment for depression and anxiety in epilepsy, supported by the largest evidence base. The pivotal study by Kanner et al. (2000) demonstrated that sertraline was effective and safe in PWE with depression, with response rates of approximately 54–60% and no increase in seizure frequency. A large retrospective cohort study by Alper et al. (2007), analyzing FDA clinical trial databases, found that patients randomized to SSRIs actually had lower seizure rates than those receiving placebo—with a standardized incidence ratio of 0.48 for seizures with SSRIs versus placebo. This anticonvulsant effect is consistent with the known seizure-protective role of serotonin in animal models.

Specific SSRIs with the most evidence in epilepsy include:

• Sertraline: Best studied, minimal ASM interactions via CYP enzymes, anticonvulsant properties demonstrated in animal models.
• Citalopram/Escitalopram: Favorable pharmacokinetic profile with minimal drug-drug interactions, widely used in clinical practice. Citalopram's QTc prolongation risk requires monitoring, particularly with concurrent carbamazepine (which may reduce citalopram levels via CYP3A4 induction).
• Fluoxetine and Paroxetine: Effective but carry higher risk of CYP2D6 inhibition, which can increase levels of phenytoin, carbamazepine, and other ASMs metabolized through this pathway.

Serotonin-Norepinephrine Reuptake Inhibitors (SNRIs)

Venlafaxine and duloxetine are second-line options with efficacy data primarily from clinical experience and small open-label studies rather than controlled trials. Both appear safe at standard therapeutic doses and may be particularly useful when depression is comorbid with neuropathic pain or significant fatigue. Animal data support anticonvulsant properties of SNRIs at therapeutic concentrations.

Medications to Use with Caution or Avoid

• Bupropion: The most proconvulsant antidepressant, with dose-dependent seizure risk estimated at 0.4% at doses ≤450 mg/day (comparable to other antidepressants) but approximately 2.1% at doses >450 mg/day. The immediate-release formulation carries higher risk than extended-release. Bupropion is generally avoided in epilepsy, though some experts consider extended-release formulations at low doses in selected cases.
• Tricyclic antidepressants (TCAs): Clomipramine and amitriptyline carry the highest seizure risk among TCAs. However, at therapeutic doses, the absolute seizure risk with most TCAs is low (approximately 0.1–0.4%), and some TCAs (notably amitriptyline at low doses) may actually have anticonvulsant properties. The primary concern is overdose risk, which carries substantial proconvulsant and cardiotoxic danger.
• Maprotiline: Tetracyclic antidepressant with the highest proconvulsant risk among all antidepressants; contraindicated in epilepsy.

Anxiolytic Treatment

Benzodiazepines are effective anxiolytics and are also anticonvulsants, making them theoretically attractive for anxiety in epilepsy. However, chronic use carries risks of tolerance, dependence, cognitive impairment, and withdrawal seizures (which can be life-threatening). They are best reserved for acute or short-term use. Pregabalin, while effective for GAD and also an ASM, has limited evidence specifically in epilepsy-comorbid anxiety. Buspirone, a 5-HT1A partial agonist, has theoretical appeal given serotonergic dysfunction in epilepsy but lacks epilepsy-specific efficacy data.

ASM Selection to Optimize Psychiatric Outcomes

The choice of ASM itself has significant implications for psychiatric comorbidity. The ILAE and American Academy of Neurology guidelines increasingly emphasize considering psychiatric comorbidity when selecting ASMs:

• Favorable for mood/anxiety: Lamotrigine (FDA-approved for bipolar maintenance, NNT for mood stabilization ≈ 7), valproate, carbamazepine, and oxcarbazepine have mood-stabilizing properties.
• Negative psychiatric profiles: Levetiracetam (behavioral adverse effects in 10–16%), topiramate (depression, cognitive slowing, psychosis risk), phenobarbital (depression, cognitive impairment), vigabatrin (depression, psychosis), perampanel (irritability, aggression in 5–12%), and zonisamide (depression, psychosis).

The management of psychosis in epilepsy requires balancing antipsychotic efficacy against the proconvulsant potential of many antipsychotic drugs. All antipsychotics lower seizure threshold to some degree, but the magnitude varies substantially across agents.

Seizure Risk by Antipsychotic Agent

• High seizure risk (avoid in epilepsy): Clozapine is the most proconvulsant antipsychotic, with dose-dependent seizure rates of approximately 1% at doses <300 mg/day, 2.7% at 300–600 mg/day, and 4.4% at doses >600 mg/day. Chlorpromazine also carries elevated risk, approximately 1.2% at standard doses.
• Moderate risk (use with caution): Olanzapine and quetiapine have intermediate proconvulsant potential. Seizure rates in clinical trials were approximately 0.9% for olanzapine and 0.8% for quetiapine.
• Lower risk (preferred in epilepsy): Haloperidol (seizure rate approximately 0.1–0.3%), risperidone (approximately 0.3%), and aripiprazole (approximately 0.1%) are considered the safest options. Risperidone and aripiprazole are generally regarded as first-line antipsychotics in epilepsy.

Treatment by Psychosis Subtype

Postictal psychosis (PIP): PIP is typically self-limiting (duration 2–14 days), and the primary intervention is prevention of seizure clusters and ensuring patient safety during the psychotic episode. Short-term benzodiazepines (e.g., lorazepam) are often sufficient and offer the advantage of also treating the underlying seizure cluster. If antipsychotic medication is needed, low-dose risperidone or haloperidol for a limited duration (1–4 weeks) is standard practice. Long-term antipsychotic prophylaxis is reserved for patients with recurrent PIP episodes.

Interictal psychosis: Chronic interictal psychosis requires ongoing antipsychotic treatment, following principles similar to the management of schizophrenia but with careful ASM optimization and selection of lower-risk antipsychotics. Response rates to antipsychotics in interictal psychosis are generally favorable—estimated at 60–80%—and often better than in primary schizophrenia, possibly reflecting the generally preserved premorbid functioning and less prominent negative symptoms in epilepsy-related psychosis.

Forced normalization: This paradoxical phenomenon may require reduction or adjustment of ASMs, even at the cost of some return of epileptiform activity. Antipsychotic treatment alone may be insufficient if the underlying mechanism is excessive EEG suppression.

Drug Interactions

Pharmacokinetic interactions between ASMs and antipsychotics are clinically significant. Enzyme-inducing ASMs (carbamazepine, phenytoin, phenobarbital) reduce plasma levels of most antipsychotics via CYP3A4 and CYP1A2 induction, sometimes by 40–50%, necessitating dose adjustments. Valproate inhibits glucuronidation and can increase levels of lamotrigine and some antipsychotics. Conversely, some antipsychotics may alter ASM levels—for example, chlorpromazine can increase phenytoin levels through CYP2C19 inhibition.

Psychotherapeutic interventions are increasingly recognized as important components of comprehensive psychiatric care in epilepsy, both as monotherapy for mild-to-moderate symptoms and as adjuncts to pharmacotherapy.

Cognitive Behavioral Therapy (CBT)

CBT has the strongest evidence base among psychotherapeutic modalities for depression and anxiety in epilepsy. A randomized controlled trial by Thompson et al. (2010)—one of the few adequately powered psychotherapy trials in epilepsy—demonstrated that a manualized CBT program adapted for epilepsy produced significant reductions in depression (BDI-II scores) and anxiety, with a moderate-to-large effect size (Cohen's d ≈ 0.63 for depression), compared to treatment as usual. The Managing Epilepsy Well (MEW) network has developed and tested several evidence-based self-management programs incorporating CBT principles, including Project UPLIFT (an internet- and phone-based mindfulness and CBT program) that showed significant reductions in depression scores with an NNT of approximately 4–5 for clinically meaningful improvement.

Mindfulness-Based Interventions

Mindfulness-based stress reduction (MBSR) and mindfulness-based cognitive therapy (MBCT) have shown preliminary efficacy in epilepsy, with small RCTs demonstrating reductions in perceived stress, anxiety, and depressive symptoms, along with modest improvements in seizure frequency and quality of life. Effect sizes have been small-to-moderate (Cohen's d ≈ 0.3–0.5). The theoretical rationale is strong: stress is a major seizure precipitant reported by 60–70% of PWE, and mindfulness-based approaches directly target stress reactivity and HPA axis dysregulation.

Acceptance and Commitment Therapy (ACT)

ACT, which focuses on psychological flexibility and valued action rather than symptom reduction, has shown promise in epilepsy populations, particularly for improving quality of life and reducing the functional impact of seizure-related anxiety. A small RCT by Lundgren et al. (2008) showed improvements in seizure frequency and quality of life with ACT compared to supportive therapy, but larger replication studies are needed.

Neurostimulation Approaches

Vagus nerve stimulation (VNS), FDA-approved for refractory epilepsy, has demonstrated ancillary mood benefits. Post-hoc analyses of VNS epilepsy trials and the pivotal D-21 study showed significant improvement in depression scores independent of seizure reduction, with response rates of approximately 30–40% for mood improvement. VNS has since received FDA approval for treatment-resistant depression (independent of epilepsy), and its mechanism—involving ascending vagal afferents to the nucleus tractus solitarius, locus coeruleus, and raphe nuclei—directly modulates the monoamine systems implicated in both seizure control and mood regulation.

Responsive neurostimulation (RNS) and deep brain stimulation (DBS) of the anterior nucleus of the thalamus are primarily seizure-targeted, but their effects on psychiatric comorbidity are under active investigation. Preliminary data suggest that thalamic DBS may have anxiolytic effects, though psychiatric adverse effects (including depression) have also been reported.

Cognitive impairment is a major contributor to disability in epilepsy and interacts bidirectionally with psychiatric comorbidity. The prevalence, pattern, and severity of cognitive deficits vary by epilepsy type, duration, treatment, and comorbid psychiatric status.

Patterns of Cognitive Impairment

In temporal lobe epilepsy, the hallmark deficit is material-specific memory impairment: left TLE typically produces verbal memory deficits (affecting word lists, stories, verbal paired associates), while right TLE affects visuospatial memory (affecting design recall, face recognition, spatial navigation). However, meta-analyses show that this lateralization is imperfect, with considerable overlap. Beyond memory, generalized attentional deficits and slowed processing speed are common across epilepsy types and often have more functional impact than memory deficits per se.

Frontal lobe epilepsy produces executive dysfunction—impairments in planning, cognitive flexibility, working memory, and response inhibition—that can mimic ADHD or the cognitive profile of depression. Generalized epilepsies, particularly juvenile myoclonic epilepsy, show frontal-predominant cognitive profiles with executive dysfunction and attentional deficits even when seizures are well-controlled.

Mechanisms of Cognitive Impairment

• Seizure-related factors: Status epilepticus, prolonged seizure duration, and high seizure frequency are associated with cumulative cognitive decline. Interictal epileptiform discharges (IEDs), even without clinical seizures, produce "transient cognitive impairment" measurable on EEG-triggered cognitive testing.
• Structural pathology: Hippocampal sclerosis, the most common pathological substrate of TLE, directly underlies memory deficits. Progressive hippocampal volume loss has been demonstrated over 3–4 years in longitudinal MRI studies of refractory TLE.
• ASM effects: Topiramate has the most adverse cognitive profile among current ASMs, significantly impairing verbal fluency, working memory, and processing speed, with measurable deficits in approximately 20–40% of patients. Phenobarbital impairs processing speed and attention. Levetiracetam and lamotrigine have the most favorable cognitive profiles.
• Depression-cognition interaction: Comorbid depression independently reduces cognitive performance on neuropsychological testing by approximately 0.3–0.5 standard deviations across multiple domains, complicating the interpretation of cognitive assessments and the attribution of deficits.

Assessment and Management

Baseline and serial neuropsychological assessment is recommended for all patients with refractory epilepsy, before and after epilepsy surgery, and when cognitive complaints are present. Cognitive rehabilitation strategies have limited evidence but show modest benefits in compensatory memory training (e.g., errorless learning, use of external memory aids). Treating comorbid depression produces measurable cognitive improvement, as does switching from cognitively impairing ASMs (e.g., topiramate) to more favorable agents (e.g., lamotrigine, levetiracetam). Aerobic exercise has emerged as a promising intervention, with small trials showing improvements in both mood and cognition in PWE.

Identifying patients at highest risk for psychiatric comorbidity—and those most likely to respond to treatment—is critical for clinical decision-making. Research has identified several consistent predictive factors.

Risk Factors for Psychiatric Comorbidity in Epilepsy

• Seizure-related factors: Temporal lobe epilepsy (particularly mesial TLE with hippocampal sclerosis) carries the highest risk for depression, anxiety, and psychosis. Refractory epilepsy confers approximately double the psychiatric comorbidity risk compared to well-controlled epilepsy. Higher seizure frequency, bilateral seizure foci, and history of status epilepticus are independent risk factors.
• Pre-epilepsy psychiatric history: A personal or family history of psychiatric disorder before epilepsy onset is the strongest predictor of post-onset psychiatric comorbidity, supporting the shared vulnerability model.
• Psychosocial factors: Unemployment, social isolation, stigma perception, and low self-efficacy are robust predictors of depression and anxiety in epilepsy. The perceived impact of epilepsy on daily life is a stronger predictor of depression than objective seizure frequency.
• ASM factors: Polytherapy (≥3 ASMs) is associated with higher rates of depression and cognitive impairment. Use of ASMs with negative psychotropic profiles (levetiracetam, topiramate, phenobarbital) increases risk.
• Neurobiological markers: Hippocampal volume loss, reduced 5-HT1A binding on PET, and elevated inflammatory markers have been associated with psychiatric comorbidity in research settings, though none are currently used clinically for risk stratification.

Predictors of Treatment Response

Shorter duration of untreated psychiatric illness predicts better response to both pharmacological and psychotherapeutic treatment, underscoring the importance of early screening and intervention. Patients with depression clearly related to ASM initiation (iatrogenic depression) typically improve rapidly with medication switching. Comorbid personality disorder, substance use disorder, and treatment-resistant epilepsy predict poorer psychiatric treatment response. Achievement of seizure freedom—whether through ASMs, surgery, or neurostimulation—is associated with significant psychiatric improvement in approximately 40–60% of patients, though the paradoxical worsening phenomenon (de novo depression after surgery) must be anticipated.

Post-Surgical Psychiatric Outcomes

Anterior temporal lobectomy for refractory TLE produces seizure freedom (Engel Class I) in approximately 60–70% of patients at 1 year. Among those achieving seizure freedom, depression and anxiety often improve significantly, with some studies showing a 50% reduction in depression prevalence at 2-year follow-up. However, approximately 10–20% develop new-onset depression or anxiety postoperatively, even after successful seizure control. Risk factors for post-surgical psychiatric decline include pre-surgical psychiatric history, right-sided resection (paradoxically), poor post-surgical psychosocial adjustment, and loss of "seizure identity." The Spencer and colleagues (2003) study and subsequent longitudinal analyses have emphasized that psychiatric assessment must be integral to the pre-surgical evaluation and long-term follow-up of epilepsy surgery candidates.

Despite significant advances in understanding the epilepsy-psychiatry interface, substantial gaps remain in evidence, clinical practice, and research methodology.

Evidence Gaps

There are no large-scale, randomized, placebo-controlled trials of antidepressants specifically in epilepsy populations. The evidence base relies primarily on open-label studies, retrospective analyses, and extrapolation from primary psychiatric disorder trials. This gap is partly attributable to historical concerns about antidepressant-seizure interactions (now largely dispelled) and partly to the difficulty of conducting psychiatric trials in neurological populations. The need for epilepsy-specific antidepressant RCTs has been emphasized by the ILAE Commission on Neuropsychiatry and the American Epilepsy Society.

Similarly, there is a near-complete absence of controlled psychotherapy trials in epilepsy with sample sizes sufficient for definitive conclusions. The few existing RCTs (Thompson et al., 2010; Cillessen et al., 2020) are small and vary in intervention design, outcome measures, and follow-up duration.

Emerging Research Areas

• Neuroinflammation-targeted therapies: Given the evidence for shared inflammatory mechanisms, anti-inflammatory interventions (including trials of IL-1β antagonists, TNF-α inhibitors, and minocycline) are being investigated for dual seizure and mood benefits.
• Kynurenine pathway modulation: Agents targeting the kynurenine pathway (e.g., IDO inhibitors, kynurenine aminotransferase II inhibitors) are in preclinical development for their potential to simultaneously reduce seizure susceptibility and depressive symptoms.
• Biomarker development: Efforts to identify blood-based or neuroimaging biomarkers that predict psychiatric comorbidity risk in epilepsy—enabling preemptive intervention—are ongoing. Candidates include hippocampal connectivity measures, 5-HT1A PET binding, and inflammatory cytokine panels.
• Precision psychiatry in epilepsy: Pharmacogenomic approaches to ASM selection that incorporate psychiatric risk profiles are being explored. For example, HLA typing to prevent carbamazepine hypersensitivity is established; extending pharmacogenomics to predict psychiatric adverse effects of ASMs is a logical next step.
• Digital health and remote monitoring: Smartphone-based ecological momentary assessment (EMA) for real-time mood and seizure tracking, and digital CBT platforms adapted for epilepsy, represent scalable solutions for the treatment access gap.

Practice Gaps

The implementation gap between evidence and practice remains wide. Surveys of neurologists consistently reveal low rates of psychiatric screening, limited comfort with prescribing psychotropic medications, and inadequate referral networks for psychological therapy. Models of integrated care—embedding psychiatrists or psychologists within epilepsy clinics—have shown improved detection and treatment rates but are not widely implemented. The development and dissemination of epilepsy-specific psychiatric care pathways, workforce training in neuropsychiatry, and insurance coverage for integrated mental health services in epilepsy remain pressing priorities.

The following evidence-based recommendations synthesize the current literature for clinicians managing psychiatric comorbidity in epilepsy:

• Screen routinely: Use the NDDI-E or GAD-7 at every epilepsy visit. Screen for suicidality at all encounters using direct questioning or validated instruments.
• Consider the bidirectional model: Psychiatric symptoms may precede, coincide with, or follow epilepsy onset. A new psychiatric symptom is not automatically "reactive"—it may reflect shared neurobiology or ASM effects.
• First-line pharmacotherapy for depression/anxiety: SSRIs (particularly sertraline, citalopram, or escitalopram) are safe, effective, and may have modest anticonvulsant properties. Start low, titrate slowly, and monitor for pharmacokinetic interactions with ASMs.
• Choose ASMs wisely: Favor lamotrigine, valproate, or carbamazepine when psychiatric comorbidity is present. Avoid or use with caution levetiracetam, topiramate, phenobarbital, and perampanel in patients with active psychiatric illness.
• Antipsychotic selection: Prefer risperidone, aripiprazole, or haloperidol for psychosis in epilepsy. Avoid clozapine unless all alternatives have failed and close monitoring is feasible.
• Psychotherapy: Offer adapted CBT as monotherapy for mild-to-moderate depression/anxiety or as adjunct to pharmacotherapy. Consider digital CBT platforms when access to epilepsy-specialized therapists is limited.
• Monitor cognition: Obtain baseline neuropsychological assessment in refractory epilepsy and before epilepsy surgery. Consider ASM-related cognitive effects when patients report cognitive complaints.
• Peri-surgical psychiatric care: Include psychiatric assessment in pre-surgical evaluation. Plan for post-surgical psychiatric monitoring, recognizing that both improvement and de novo onset of psychiatric symptoms can occur after successful surgery.
• Collaborate: The complexity of epilepsy-psychiatry interactions demands multidisciplinary care involving neurologists, psychiatrists, neuropsychologists, and psychotherapists. Integrated care models produce the best outcomes.