Supplementary Material for: Association between Ambient Temperature and Risk of Ischaemic Stroke: An Observational Cohort Study

Introduction
The influence of ambient temperature on the risk of stroke has remained unclear, particularly in relation to stroke subtypes. This study aimed to investigate the association between ambient temperature and the risk of ischaemic stroke from various aetiologies.
Methods
We investigated the onset of acute ischaemic stroke in the prospective stroke registry conducted from October 2007–September 2019 in Fukuoka, Japan. Stroke aetiology was classified into cardioembolism (CE), large-artery atherosclerosis (LAA), small-vessel occlusion (SVO), and other aetiologies. Hourly, maximum, and minimum temperature data were obtained from the Japan Meteorological Agency. After controlling for the nonlinear and delayed effects of daily average temperature, we employed a quasi-Poisson generalised additive model with the natural logarithm of daily stroke counts as a function of predictors, including temperature metrics and humidity.
Results
A total of 17,755 patients with acute ischaemic stroke (mean age ± SD: 73.1 ± 12.6 years; 41.4% female) were included in the analysis. Lower temperatures were associated with an increased risk of CE and LAA but not with that of SVO. Conversely, higher temperatures were associated with a decreased risk of CE, whereas no association was found with LAA or SVO. The association was observed within 5 days before stroke onset for LAA but after a longer period for CE. The risk of CE increased with rising variability in daily temperature preceding stroke onset.
Conclusion
We observed distinct associations between ambient temperature and the risk of ischaemic stroke, contingent upon stroke aetiology, temperature levels, exposure duration, and patient characteristics.