ATLANTA — Depression shapes the course of epilepsy at multiple stages — raising the risk of developing the disorder by about twofold and, once epilepsy is present, increases the chance of early treatment failure, two new studies showed.

Together, these findings raise the possibility that depression and epilepsy may share underlying vulnerabilities that influence both seizure risk and early treatment response, although the mechanisms remain unclear.

Ali Rafati, MD, a post-doctoral research fellow in neurology at Johns Hopkins University School of Medicine, Baltimore, said this emerging evidence strengthens the view that depression and epilepsy influence one another in multiple ways and highlights the need for neurologists and psychiatrists to work “as a team” to optimize patient outcomes. 

The studies were presented on December 6 at American Epilepsy Society (AES) 79th Annual Meeting 2025.

Need for Routine Depression Screening

Depression is the most common psychiatric comorbidity in individuals with epilepsy, and untreated depression is associated with lower quality of life, poor treatment adherence, higher healthcare utilization, and mortality.

“Yet time and again, we neurologists still don’t screen for depression in patients with epilepsy,” said Rafati.

Research, including a recent meta-analysis by Rafati and his team, shows individuals with epilepsy often experience psychiatric comorbidities. But less is known about whether depression itself predisposes individuals to develop epilepsy.

To explore this question, Rafati and his team conducted a comprehensive literature search for prospective cohort and case-control studies that examined epilepsy incidence in individuals with and without depression. The primary outcome was the pooled hazard ratio (HR) for epilepsy among individuals with a history of depression.

The meta-analysis of two large cohort studies that included almost 11.5 million individuals followed for at least 5 years found a pooled HR of 2.39 (95% CI, 2.35-2.43; P < .001), suggesting a significant increase in epilepsy risk among individuals with prior depression.

Rafati noted this is the first time that a meta-analysis has “pulled large cohort studies” together to provide “top level” evidence linking depression to epilepsy risk.

In contrast, the group’s secondary analysis of four case-control studies assessing depression history in individuals with and without epilepsy produced a pooled odds ratio (OR) of 2.09 (95% CI, 0.92-4.76; P = .08), showing a nonstatistically significant association.Cohort studies provide higher quality evidence and have less selection or recall bias than retrospective case-control studies, Rafati noted.

Shared Neurobiological Pathways?

The underlying mechanisms that increase epilepsy risk among people with depression remain unclear, but they may involve shared neurobiological pathways, chronic stress responses, or inflammatory processes, said Rafati. “Other shared pathways may involve the endocrine system and the hormones that are involved in that, and the hypothalamus-pituitary axis,” he added.

Future work will examine whether early identification and treatment of depression could reduce epilepsy risk. Rafati and colleagues also plan to study whether other psychiatric disorders — including anxiety and psychotic disorders — similarly increase epilepsy risk.

Principal investigator Churl-Su Kwon, MD, faculty member in the Departments of Neurology, Epidemiology, and Neurosurgery at Columbia University, said the mechanisms underlying the association between depression and epilepsy remain poorly defined.

“We need more studies that explore whether risks vary among different epilepsy subtypes, and we need more genetic studies and functional imaging studies. These will play a crucial role in providing insight into the connection between depression and epilepsy and all the other psychiatric conditions that we will be studying in the future,” Kwon told Medscape Medical News.

While Kwon and colleagues’ study focused on the risk of developing epilepsy, another analysis presented at AES examined what happens after diagnosis — and showed that depression may also influence early treatment response.

To evaluate how depression at the time of diagnosis affects early treatment patterns and outcomes, investigators led by Samuel W. Terman, MD, conducted a retrospective observational study of 90,738 adults newly diagnosed with epilepsy identified through two large US claims databases. Patients were stratified by the presence of depression; about 24% met criteria. The median ages in the depression and nondepression groups were 59 and 58 years, respectively.

Women were more common in the depression group (62.4% vs 54.8%), and these patients had a heavier burden of almost all chronic conditions, including psychiatric comorbidities such as anxiety, sleep disorders, psychosis, and bipolar disorder, said Terman, assistant professor in the Division of Epilepsy, Department of Neurology, Michigan Medicine, University of Michigan health, Ann Arbor, Michigan. This pattern may partly reflect increased healthcare contact among people with depression, leading to greater detection of underlying conditions, he noted.

Researchers examined “lines of therapy,” which indicate whether treatment is uninterrupted. “A line of therapy could be disrupted by a drug discontinuation or a switch or augmentation or cessation of a drug,” Terman explained.

Patients with depression remained on their initial therapy for a shorter period (median, 4.8 months vs 5.8 months). They were also less likely to respond to treatment. In an analysis adjusted for age, sex, race/ethnicity, and payer, patients with depression had significantly higher odds of treatment failure (OR, 1.41; 95% CI, 1.34-1.48) compared with those without depression.

“We found patients with epilepsy and depressionhad a 40% higher odds of treatment failure in the first month on average. We need to be mindful of this and maybe keep a closer watch on people with depression because we know they’re at greater risk for bouncing around between medications, either because of intolerance or inefficacy,” said Terman.

Future research may be able to clarify the factors driving shorter time to treatment failure in these patients, he added.

Depression status was identified using claims datasets, which generally lack clinical details — for example, specific reasons for discontinuation — limiting the ability to determine what prompted treatment changes. Another potential limitation is that more than 60% of the study population were Medicare beneficiaries, which may reduce generalizability to other groups such as those with commercial insurance.

Clinically Meaningful Research

Commenting on the studies, Aatif M Husain, professor of neurology, and chief of Epilepsy, Sleep and Clinical Neurophysiology, Duke University Medical Center, Durham, North Carolina, said the findings reinforce a growing body of evidence that the relationship between depression and epilepsy is bidirectional.

Husain said the finding that people with depression are roughly twice as likely to develop epilepsy is clinically meaningful. The link has long been suspected, but this study underscores it in a compelling way, he noted.

He added that large claims-based studies examining treatment patterns in individuals with and without depression are valuable because they capture real-world experience across very large patient populations. At the same time, he cautioned that diagnostic codes have limitations and may not always accurately capture psychiatric conditions.

He said the study’s most meaningful contribution is the clear picture it provides of the substantial comorbidity burden and reduced medication adherence seen in patients with both epilepsy and depression. These individuals, he added, tend to be medically complex and may require closer monitoring because they are more vulnerable to treatment disruption than other patients.

Findings from studies like these, Husain said, support growing calls for routine depression screening in epilepsy clinics — an approach that has been recommended but is still not consistently implemented in practice.

The meta-analysis received no outside funding. Rafati reported no relevant conflicts of interest. The observational study was funded by Xenon Pharmaceuticals Inc. Terman reported he has been a paid consultant for Jazz Pharmaceuticals, Inc. and currently being a paid consultant for Xenon Pharmaceuticals Inc. He is an employee of the University of Michigan with support from the National Institutes of Health. Husain reported being a consultant to Xenon Pharmaceuticals Inc.