Exploring associations between ambient temperatures and seizure occurrence

Environmental factors may influence seizure occurrence; however, the relationship between short- and medium-term temperature fluctuations and seizure occurrence remains unclear. This exploratory study investigated associations between daily ambient temperatures and seizure occurrence in patients with epilepsy.

A subset of patients from the RADAR-II cohort filled a seizure diary. Temperature information was obtained based on patients' home address postal codes. Four features were derived: diurnal temperature amplitude, day-to-day temperature change, deviation of the daily mean from the 7-day rolling mean, and deviation of the daily minimum from its 30-day rolling mean. Associations with seizure occurrence were evaluated using Generalized Estimating Equations with exchangeable correlation structure, adjusted for season, age, and sex. Robust sandwich variance estimation was applied; Benjamini-Hochberg FDR correction was used to account for multiple testing across four features.

Twelve adults (mean age 38.7 years, SD 13.4, 50% female) documented seizures over 1884 days. Three of four temperature features were significantly associated with seizure occurrence after FDR correction: temperature amplitude (p_FDR​=0.044), day-to-day temperature change (p_FDR=0.044), deviation of the daily minimum from its 30-day rolling mean (p_FDR=0.012). For the latter, the daily minimum lay 0.47 °C below its 30-day baseline on seizure days (β=-0.47 °C, 95%CI [-0.78, -0.16]).

Ambient temperature patterns at multiple timescales and seizure occurrence may be linked. It is likely that these associations are substantially influenced by both regional context and individual characteristics. While limited by sample size and regional focus, this associational and hypothesis-generating study underscores the relevance of environmental factors to seizure dynamics and warrants further investigation in larger populations.