AES-NINDS Epilepsy Benchmarks Area IV: Comorbidities of Epilepsy – Where Are We and Where Will We Go?

Abstract

Area IV of the 2021 Epilepsy Research Benchmarks aims to establish research goals to limit or prevent adverse consequences of seizures and their treatment across the life span. In this commentary, we will highlight key advances in Area IV. We will focus here on the comorbidities of epilepsy with special emphasis on neuropsychiatry, the endocrine system, and sudden unexpected death in epilepsy, which were discussed at the Epilepsy Benchmarks Area IV Symposium during the Annual Meeting of the American Epilepsy Society in December 2025.

Introduction

The Epilepsy Research Benchmarks were first established in 2000 as a joint effort by the National Institute of Neurological Disorders and Stroke (NINDS) and the American Epilepsy Society (AES) to better understand current progress in epilepsy research and to define priority areas of research for the future. The Epilepsy Research Benchmarks have regularly been updated in approximately 7-year intervals. Area IV of the 2021 Epilepsy Research Benchmarks aims to establish research goals to limit or prevent adverse consequences of seizures and their treatment across the life span. In this commentary, we will highlight key advances in basic and clinical knowledge pertaining to Area IV, presented at the Epilepsy Benchmarks Symposium during the American Epilepsy Society Annual Meeting held on December 5 to 9, 2025 in Atlanta, GA.

Epilepsy is associated with a high burden of comorbidities, with approximately 50% of adults and 80% of children with epilepsy experiencing comorbid conditions.^1,2 Cognitive disabilities and psychiatric comorbidities (PCs), in particular anxiety and depression, are most reported, as they have the most significant impact on patients’ quality of life. Other common physical conditions associated with epilepsy include cardiovascular, respiratory, digestive, metabolic and endocrine conditions, autoimmune disorders, and cancer. Epilepsy is also associated with an increased risk of sudden unexpected death in epilepsy (SUDEP).

In the following sections, we summarize 3 research topics pertaining to the comorbidities of epilepsy: neuropsychiatry, the endocrine system, and SUDEP.

Psychiatric Comorbidities in Epilepsy

A review of most presentations of PCs in people with epilepsy (PWE) would give the impression that mood disorders are the only 1 that merits the clinician's attention. Yet, compared to the general population, a recent systemic review of the prevalence of PCs in epilepsy demonstrates comparable odds to develop anxiety disorders, suicidality, obsessive compulsive disorder, alcohol and drug abuse disorders, and higher odds for psychotic disorders and attention-deficit hyperactivity disorders (ADHD). ³ In fact, 1 in 3 PWE is expected to experience a PC in their life, while this occurs in half of patients with treatment-resistant epilepsy (TRE).

PCs preceding the onset of epilepsy are common, but often go unrecognized by clinicians. For example, a recent study of 347 patients with newly diagnosed focal epilepsy found that 39% met criteria for a mood and/or anxiety disorder and 23% for suicidality. ⁴ In fact, population-based studies have demonstrated the existence of a bidirectional relationship between epilepsy and PCs (mood, anxiety, ADHD, suicidality, and psychotic disorders), whereby PWE are at an increased risk of developing these PCs, while patients with these primary psychiatric disorders have an increased risk of developing epilepsy.

Why Should Neurologists Care?

These facts have a direct impact on the management of the seizure disorder at several levels:

1. A history of a PC preceding the onset of epilepsy has been associated with an increased risk of TRE. For example, in a study of 780 patients with newly diagnosed epilepsy, a history of depression preceding the epilepsy had an odds ratio of 2.2 against reaching seizure-freedom. ⁵ In a recently published study of patients with newly diagnosed focal epilepsy, having a history of a PC at the time of diagnosis was associated with a 1.78 relative risk of developing TRE at follow-up after 4 years. ⁶ In a separate study of a subset of these patients, a history of suicidality and mood disorders identified at the time of diagnosis of epilepsy were the 2 variables associated with TRE at 4 years of follow-up, but suicidality was the 1 variable predictive of TRE (relative risk 2.89 [confidence interval: 1.65-5.05]). ⁷

2. A history of PC preceding a temporal lobectomy was associated with a lower seizure-freedom rate after a 2-year postsurgical follow-up period. For example, among 378 patients with TRE (47% of whom had a PC), 88% without a PC met Engel I postsurgical seizure outcome in contrast to 45% of those with a history of an axis I diagnosis and a personality disorder. ⁸ Similar findings were reported in 5 other studies.^9–13

3. A personal and/or a family history of PC are associated with a higher incidence of medical, neurologic, and psychiatric iatrogenic adverse events to antiseizure medications (ASMs) and with an increased risk of postsurgical psychiatric complications.

4. A history of PC is associated with poor ASM compliance.

Implications for clinical practice:

1. A psychiatric evaluation or at least a detailed screening of the common PC affecting the patient and first-degree relatives should be included in the initial epilepsy evaluation of every patient. The psychiatric data should include past and current disorders.

2. The data from the initial psychiatric evaluation should be strongly considered in the selection of the first and subsequent ASMs. For example, levetiracetam is most frequently prescribed as the initial ASM, yet psychiatric iatrogenic adverse events are likely to occur in 1 of every 3 to 4 of these patients. These adverse events are more likely to occur in PWE with personal and/or family psychiatric histories. Topiramate, zonisamide, phenobarbital, primidone, perampanel, and vigabatrin are other ASMs with similar risks of psychiatric iatrogenic effects. Conversely, ASMs with mood stabilizing (eg, carbamazepine, oxcarbazepine, valproic acid), antidepressant (lamotrigine) and/or anxiolytic properties (eg, valproic acid, pregabalin, benzodiazepines) should be considered as first choice in the presence of a personal and/or family history of mood and/or anxiety disorders as long as they are appropriate for the type of epilepsy and seizures.

3. The potential impact of presurgical PC should be factored-in to anticipate potential postsurgical psychiatric complications and should be discussed with patients and families before the surgery.

4. The association of PC with an increased risk of TRE should be factored-in when considering the discontinuation of ASMs after seizure-freedom with ASMs or after surgery. However, this question needs to be investigated in future research as the data is limited.

A Final Question

PCs are often untreated because of limited access to mental health professionals. Accordingly, has the time come for neurologists and/or neurology nurse practitioners to learn how to treat the common and uncomplicated PCs which respond readily to psychotropic drugs and which comprise 60% of symptomatic PWE?

Effect of Seizures and Epilepsy on the Reproductive Endocrine System

There are complex bidirectional interactions between reproductive hormones and epilepsy. While there is ample evidence that hormones can affect neuronal excitability and seizure thresholds, it is also apparent that epileptic activity influences hormones leading to hormonal dysregulation in both men and women. This has important clinical consequences as reproductive dysfunction is unusually common in both men and women with epilepsy.^14–17

The Effects of Epilepsy on Reproductive Endocrine Function

Epileptic activity significantly impacts reproductive function in both women and men mainly by influencing the hypothalamic–pituitary–gonadal (HPG) axis. Clinically, this may manifest as menstrual disorders, polycystic ovaries, hypothalamic amenorrhea, and hypogonadism. The brain controls reproductive endocrine function primarily through hypothalamic regulation of pituitary secretion. There are extensive direct connections from the cerebral hemispheres, especially from temporolimbic structures and amygdala to the hypothalamus. This is especially evident in regions involved in regulation of gonadotropin-releasing hormone (GnRH) production, secondarily affecting the release of the gonadotropins follicle-stimulating hormone (FSH) and luteinizing hormone (LH), from the pituitary.

GnRH pulsatile release frequency seems to play a critical role in determining gonadotropin output. High GnRH pulse frequency promotes LH release, failure of ovarian maturation, aromatase deficiency, and an increase in testosterone.^14,18 A high GnRH pulse frequency is seen in women with left temporal lobe epilepsy^14,18,19 and primary generalized epilepsy. ²⁰ The combination of neuroendocrine changes may increase the risk of polycystic ovary syndrome (PCOS),^18,20 which is found in up to 20% of women with epilepsy compared to 5% to 6% in the general population. ¹⁴ The high frequency of PCOS in women with epilepsy has also been attributed to valproate (VPA) use,^17,21,22 and remains the subject of heated debate. A possible unifying consideration might be that epilepsy is a risk factor for the development of PCOS and that the risk is further increased in women who take VPA.

Reduced GnRH pulse frequency favors FSH secretion and lower LH and estradiol levels. This will lead to hypothalamic amenorrhea with menstrual disorders, infertility and reduced libido.^14,17 Continuous GnRH secretion suppresses both FSH and LH, inhibiting ovulation and estrogen production.

In men, the effects of epileptic activity on the HPG axis have been less studied and the effects are difficult to dissect from the effects of ASMs or psychosocial stress. In men with temporal lobe epilepsy the pulsatile secretion of LH has been found to be disturbed ²³ and symptoms in line with hypogonadism with lower serum testosterone, reduced sperm production, diminished sexual interest, and potency has been described.^14,24

The Effects of Reproductive Hormones on Epilepsy

The male and female peripheral sex steroid hormones, estradiol, progesterone, and testosterone are all derived from cholesterol and are closely linked. ⁸ In general, estrogens have been considered proconvulsant, progesterone, and its metabolites as anticonvulsants and androgens as mainly anticonvulsant, although their effects are more varied probably due to metabolism to, among others, estradiol. It should be remembered that the effects of the different reproductive steroid hormones may vary in relation to dose, duration of exposure, and site of action. ¹⁷ All steroid hormones act both at intracellular steroid hormone receptors with effects evolving over time and at membrane receptors responsible for immediate effects on excitability. ¹⁷

Estrogens

Estrogens have both acute, membrane-related effects and more slowly developing effects mediated through their intracellular receptors. An acute effect was shown already in 1980, ²⁵ and is considered to be related to a potentiation of N-methyl-D-aspartate receptor (NMDA) on its receptor.^25,26 Over time, gamma-aminobutyric acid (GABA)-ergic inhibition may be reduced, at least partly due to reduced GABA release. Estrogens can also increase dendritic spine density and then the number of NMDA type glutamate receptors. ²⁷ These plastic changes may be of major importance for catamenial epilepsy, in which seizure incidence fluctuates with hormone concentrations during the menstrual period.

Progesterone and Its Metabolites

Progesterone exerts its effect on brain excitability mainly through its reduced metabolites, especially allopregnanolone, which is a very potent, positive allosteric modulator of the postsynaptic GABA-A membrane receptor.^13,14 Progesterone metabolites seem to have a unique binding site on the GABA-A receptor, increasing their opening time. ²⁸

Progesterone and its metabolites may also antagonize excitatory mechanisms as it has been found that they may decrease glutamatergic responsiveness after systemic or topical application. ²⁹ Another way to affect seizure susceptibility is by altering subunit composition of the GABA-A receptor to 1 less sensitive to ASMs. Cyclic changes in subunit composition seem to be of relevance in understanding the mechanisms behind catamenial epilepsy.^30,31 Finally, although the effects on nonclassical receptors are the most important, effects on classical intracellular receptors cannot be ruled out and probably participate in the overall effect of progesterones.

Androgens

The effects of androgens are mainly anticonvulsant. The variable actions of testosterone may be partly due to its metabolism. Testosterone can be metabolized to 17β-estradiol, which is generally excitatory, but also to androstanediol and dihydrotestosterone, which exert potent antiepileptic effects.^8,23

As with the other sex steroids, androgens act via both classical and nonclassical mechanisms, the latter being the most important related to epilepsy. Here they act as positive modulators of the GABA-A receptor. ³² Like estradiol, it seems that androgens also have effects on neuronal structure and function, increasing the number of spine synapses. ³³ The relevance of these plastic changes is, however, uncertain.

Conclusions

There are close bidirectional interactions between reproductive hormones and epileptic activity in the brain explaining the high frequency of reproductive endocrine dysfunction in patients with epilepsy. A major problem is that much of our current understanding in this field is based on studies conducted in the 1990s. Therefore, there is a need for further and more updated information from both experimental and clinical studies in this field. Recently however, very promising results have been obtained using a new experimental model, the intrahippocampal kainic acid model, in rats and mice. ³⁴ Important clinical questions still must be answered like treatment of catamenial epilepsy, treatment of sexual problems, and issues related to menopause. During recent years additional questions have arisen related to in vitro fertilization and transgender issues, which makes it even more necessary to have updated information and newer studies.

SUDEP: What Have We Learnt and What Comes Next?

This National Institutes of Health (NIH) benchmark update signals an inflection point in SUDEP research. Indeed, work performed within and outside the NINDS Center without Walls ³⁵ and fueled by deeply committed advocacy groups has driven significant advances in understanding the clinical neurophysiology, genetics, and cellular mechanisms of postictal cardiorespiratory failure, bringing us closer to preventing this tragic outcome. New insights emerge from investigators sharing data from monitoring studies of high-risk patients at the epilepsy monitoring unit (EMU) bedside with neurobiology laboratories analyzing genetically engineered molecular defects in mouse brain, heart, and human organoids. Initially faced with a blank list of explanations for the rare and sudden death in bed of a person with long-standing epilepsy, the last decade has witnessed a solid framework for hypothesis-driven studies, lighting the way for therapeutic trials that target reduction of SUDEP risk.

However, we have now reached a pivotal realization: most patients at high epidemiological risk survive. Predicting exactly who is at risk, and when, is still difficult. While certain genes, most notably those linked to developmental epilepsies and cardiac arrhythmias, markedly increase premature mortality risk, not all variants within each gene are equally harmful. The risk landscape is complex, and no single factor clearly determines SUDEP risk for everyone. Without additional functional biomarkers to validate variant pathogenicity, individual prediction remains challenging. Despite this diversity, there are encouraging signs of shared downstream mechanisms, and strategies to improve recovery after seizures that may help reduce SUDEP incidence for many patients.

Refinements in SUDEP Diagnosis

An ILAE Task Force on SUDEP is refining the working definition of SUDEP, essential for forensic diagnosis and research. Narrowing the criteria delimiting definite SUDEP facilitates isolation of a shared SUDEP mortality pathway differentiated from other causes of death. This strict definition focuses attention on the canonical pattern of terminal events shared by all monitored EMU cases described in the MORTEMUS study, ³⁶ namely apneas, arrhythmias, and postictal bradycardia that begin minutes following a generalized tonic–clonic seizure. During this aftermath, respiratory arrest invariably precedes cardiac arrest, and ventricular fibrillation is absent, distinguishing this pattern from sudden cardiac death.

Isolation of Forebrain–Brainstem Pathways

Stimulation of the central amygdala depresses respiration and hyperactivates vagal tone, providing a clear link between limbic seizures, brainstem dysfunction, and SUDEP risk. ³⁷ Mouse models of lethal tonic audiogenic seizures activate brainstem nuclei without evoking electroencephalographic seizures in forebrain, further implicating cardiorespiratory brainstem pacemaking centers in the SUDEP pathway. Recent imaging advances ³⁸ that define ascending brainstem reticulo–thalamic pathways promise to facilitate neuromodulation of innate arousal mechanisms, where serotonin, adenosine, and other prime signaling targets reside.

The Expanding SUDEP Genome

Cardiac long QT (LQT) genes that populate sinoatrial conduction pathways are coexpressed in brain and include known genes for epilepsy. Mutations in these genes cause arrhythmias in heart and brain that reproduce the canonical MORTEMUS sequence, providing the first biomarker contributing monogenic risk for SUDEP. ³⁹ Epilepsy is often discounted or escapes detection in syncope patients, and the list of genes in this category will grow as they are also identified as a cause of prolonged apneas. Functional links between LQT syndromes and sleep apneas are also becoming recognized, implicating a circadian mechanism. Sleep apnea elevates nocturnal LQT arrhythmias, suggesting a hypoxic threshold for these genes may be exceeded. ⁴⁰ Many variants for the most common developmental epileptic encephalopathies (DEE) syndromes are linked to high SUDEP mortality and map onto respiratory circuits.^41,42 Genes isolated for sudden death in the young and sudden cardiac death await analysis for occult epilepsy using transgenic models.

Genetic Risk Revisited

While variants identified in SUDEP cases contribute to lethality, they do not identify the age and penetrance profile which varies considerably. Longevity curves for many SUDEP genes expressed in transgenic mice reveal that most animals survive, and those that die do so at widely varying ages. ⁴³ The incomplete penetrance of the SUDEP phenotype in isogenic inbred mouse strains undoubtedly persists in human, and must inform clinical genetic counseling, even in the case of a validated SUDEP variant.

The Finality of Brainstem Spreading Depolarization

The variable onset of SUDEP genes raises a major dilemma in conceptualizing SUDEP mechanisms and therapy. If a mutation cripples brainstem autoresuscitation pathways, how do patients survive many seizures? Why is the first seizure not also the last? This concern gave rise to the second hit hypothesis, that a wave of spreading depolarization triggered in epileptic brain silences cardiorespiratory function. ⁴⁴ Spontaneous spreading depolarization (SD) waves are detected with electrodes in cortex and medulla of mouse DEE models, including Scn1a, Cacna1a, Scn8a, Kcnq2, Kcn1a, and in vivo imaging reveals that a forebrain seizure is not lethal unless the SD wave reaches the lower brainstem. ⁴⁵ Since SD thresholds are considerably higher than seizures, this lethal transition can be targeted by medicine or other genes. ⁴⁶

In summary, further advances in SUDEP gene detection, functional validation of candidate variants in patients using circadian autonomic and oximetry readouts, and regular outreach informing caregivers, families, genetic counselors, and forensic pathologists of recent advances will allow the field to better navigate the uncertainty of SUDEP risk until the arrival of preventative therapies.

Overall Conclusions

Psychiatric and other comorbidities of epilepsy remain a major concern. On the bright side, even though 60% of persons with symptomatic epilepsy have PCs, many of those are common, uncomplicated and treatable with common psychotropic drugs. A broader understanding of these comorbidities and treatment options remains an important goal.

Despite progress being made, a major gap remaining in our understanding of the effects of epilepsy on the endocrine system and vice versa is the difficulty in separating direct interactions between hormones and epilepsy from psychosocial factors and effects of ASMs. Other clinical challenges that need to be addressed include improving treatments for catamenial epilepsy and chronic reproductive dysfunction. It is also imperative to take advantage of the existing preclinical models presenting similar reproductive endocrine phenotypes to define the natural history of epilepsy comorbidities, establish mechanisms, and develop new treatments.

In view of a lack of preventative therapies for SUDEP and the need to address the uncertainty of SUDEP risk it will be essential to advance SUDEP gene detection. Studies should include the functional validation of candidate variants in patients. It is also critical to promote outreach activities to inform PWE, their families and caregivers as well as other stakeholders such as genetic counselors and forensic pathologists of recent advances in the field.

In summary, addressing the comorbidities of epilepsy remains a highly active research area with an urgent need to limit or prevent the adverse consequences of seizures and epilepsy.