Post-Stroke Epilepsy Mortality Is Driven by Age and Seizure Type, Not Sex

Mortality among 84 post-stroke epilepsy patients reached 47% in women against 22% in men (p=0.02) — yet sex itself did not drive the difference. A study from University Hospital Innsbruck, published in Epilepsy & Behavior (December 2026, Volume 185, article 111231), found that age, stroke severity at discharge, and pre-stroke disability fully accounted for the higher death rate in women.

The cohort is worth a reader's attention. Luisa Delazer and colleagues included consecutive ischaemic stroke patients treated in Innsbruck between 2014–2017 and 2020–2025 who developed post-stroke epilepsy, with a median follow-up of 40 months. The group comprised 38 women and 46 men, median age 72.4 years. Investigators extracted data on PSE onset, stroke severity, aetiology, SeLECT score, and mortality from electronic medical records and analysed them with regression models.

What stayed the same across age and sex?

Neither first seizure type nor latency to the first unprovoked seizure depended on demographics. The median latency was 266 days, and the distribution of seizure types was similar across age groups and between sexes. For reading-room and monitoring-unit practice, this means the epileptologist cannot infer presentation patterns from a patient's age or sex alone.

What differed in women?

Cardioembolism stood out. It was the most common stroke aetiology in women with PSE, at 47% — roughly half the female cohort. Women also tended to present with more severe strokes. The authors, however, report that sex differences in stroke severity and aetiology did not reach statistical significance in this cohort. The signal is present; the sample size of 84 tempers it.

Which seizure presentations predicted death?

The strongest prognostic finding concerns the initial seizure type. Patients presenting with status epilepticus had 71% mortality, and those presenting with focal to bilateral tonic-clonic seizures had 39% mortality. The adjusted hazard ratio was 2.98 (p=0.02) for these presentations combined. In Cox models, the independent predictors of mortality were:

• Increased age (HR 1.09, p<0.001)
• Greater stroke severity at discharge (p=0.005)
• Poorer pre-stroke function on the modified Rankin Scale (p=0.004)

Once these variables entered the model, the crude excess mortality in women disappeared. Sex was a marker, not a mechanism.

What does this change for clinical practice?

The study answers the practical question directly: mortality risk in post-stroke epilepsy is legible from variables clinicians already record — age, discharge stroke severity, pre-stroke mRS, and the character of the first seizure. An initial status epilepticus or a focal to bilateral tonic-clonic seizure warrants closer surveillance than seizure timing or demographic profile does. The cardioembolic predominance in women also argues for rigorous cardiac evaluation in that subgroup, though the authors themselves flag the statistical weakness of that finding.

Context matters for scale: post-stroke epilepsy affects 2–14% of ischaemic stroke survivors and carries increased mortality and poorer functional recovery. Demographic differences within PSE cohorts have been understudied, and this single-centre Austrian series is a corrective rather than a final word. Larger multicentre cohorts will need to confirm whether the female cardioembolic signal holds.