# Immediate and delayed effects of thermal stress on fever-associated seizures in children: A time-stratified case-crossover study in Japan

## TL;DR

This study found that both cold and heat stress can increase the risk of fever-associated seizures in young children, with effects appearing hours after exposure.

## Contribution

The study reveals a bimodal relationship between thermal stress and pediatric fever-associated seizures using high-resolution bioclimatic data.

## Key findings

- Moderate cold stress increased seizure risk with a cumulative odds ratio of 2.22 after 72–96 hours.
- Moderate heat stress showed a peak risk at 100 hours with a cumulative odds ratio of 2.26.
- High-resolution thermal data revealed non-linear and delayed effects of temperature on seizure risk.

## Abstract

This study aimed to examine the non-linear and delayed effects of thermal stress, measured by the hourly Universal Thermal Climate Index (UTCI), on the risk of pediatric fever-associated seizures (FAS). We conducted a time-stratified case-crossover study in Okayama, Japan (May 2015–March 2023), analyzing 3,201 ambulance-attended FAS cases in children younger than 7 years. Using a distributed lag non-linear model (DLNM) with a 144-h lag, we estimated the association between UTCI and FAS. The analysis revealed a bimodal exposure–response relationship. Moderate Cold Stress (10th percentile, –1.6 °C) was associated with a significant cumulative odds ratio (OR) of 2.22 (95% CI: 1.22–4.06). Risk also increased at the upper range of No Thermal Stress (24.2 °C; cumulative OR 2.74, 95% CI: 1.63–4.63), extending into Moderate Heat Stress (28.7 °C; cumulative OR 2.26, 95% CI: 1.33–3.84). These effects were primarily delayed to 72–96 h for Moderate Cold and reached a peak around 100 h for Moderate Heat. Strong Heat Stress showed immediate but non-significant risk patterns. These findings suggest that infection-mediated pathways likely drive the observed bimodal risk pattern, demonstrate the utility of high-resolution bioclimatic indices, and can inform the development of temperature-specific public health alerts.