Supplementary Material for: Seasonal variations in stroke occurrence

Background Understanding seasonal variations in stroke can help stakeholders identify underlying causes in seasonal trends, and tailor resources appropriately to times of highest needs. We sought to evaluate seasonal occurrence of stroke and its subtypes. Methods We conducted a retrospective cohort study using administrative data from January 1st, 2003, to December 31st, 2017, in Ontario, Canada’s most populous province. We evaluated seasonal variations in stroke occurrence by subtype, via age/sex standardized rates and adjusted rate ratios using Poisson regressions. In those with stroke, we evaluated 30-day case fatality risks by season, adjusted for age, sex, stroke type and comorbid conditions, and then used Cox proportional hazard models to estimate the effect of season on the fatality. The administrative data used in this study was from the Canadian Institute for Health Information’s Discharge Abstract Database (DAD), the National Ambulatory Care Reporting System (NACRS) database, the Ontario Registered Persons Database, and from the 2006 and 2011 Canada Census and linked administrative databases. Results During our study period, we observed 394,145 stroke or TIA events, with a decrease in monthly hospitalization/emergency department visits per 100,000 people between January 2003 and December 2017 from 24.22 to 17.43. Compared to the summer, overall stroke occurrence was similar in the spring but slightly lower in the fall (adjusted relative risk (aRR) 0.97, 95% confidence interval (CI) 0.96 to 0.98) and winter (aRR 0.94, 95% CI 0.94 to 0.95). There were minor variations by stroke subtype. Winter was associated with the highest risk of stroke case fatality compared to the summer (12.4% vs. 11.4%, adjusted hazard ratio 1.10, 95% CI 1.07 to 1.13). Conclusions We found seasonal variations in stroke occurrence and case fatality, although the absolute differences were small. Further work is needed to better understand how environmental or meteorological factors might affect stroke risk.