Characterising seizure cycles in pediatric epilepsy

Area of Science:

• Epilepsy research
• Chronobiology
• Pediatric neurology

Background:

• Multiday cyclic patterns in seizure timing are known in adults with epilepsy.
• Longer-term seizure patterns in pediatric epilepsy cohorts remain underexplored.
• This study investigates multiday seizure cycles in pediatric and adult epilepsy patients.

Purpose of the Study:

• To identify and compare multiday seizure cycles between pediatric and adult epilepsy cohorts.
• To validate cycle-based methods for estimating seizure likelihood in pediatric epilepsy.

Main Methods:

• Retrospective analysis of multiday seizure cycles from 325 electronic seizure diary users (71 pediatric).
• K-means clustering grouped cycles; synchronization index quantified cycle strength.
• Rayleigh test identified significant cycles; Wilcoxon rank-sum test compared cohorts; AUC calculated for model validation.

Main Results:

• Analysis of 30,019 seizures revealed significant circadian, weekly, fortnightly, and monthly seizure cycles.
• Prevalence and strength of these cycles did not differ between pediatric and adult cohorts.
• Cycle-based models reliably predicted seizure occurrence, outperforming random models and moving average models.

Conclusions:

• Multiday seizure cycles derived from seizure diaries are a viable method for understanding pediatric epilepsy patterns.
• Identifying individual seizure cycles can decrease seizure unpredictability.
• This approach has potential to inform clinical decision-making for epilepsy management.