Association of winter exposure with ischemic stroke risk in nonvalvular atrial fibrillation within a CHADS₂-based...

Ischemic stroke, including transient ischemic attack, exhibits seasonal variation, and the association between winter exposure and ischemic stroke risk in patients with nonvalvular atrial fibrillation (NVAF) has not been fully characterized at the population level. We investigated the association of winter exposure with ischemic stroke risk in Japanese patients with NVAF within a CHADS₂-based risk framework.

We analyzed nationwide Japanese insurance claims data (2014-2022) to identify patients ≥ 60 years with newly diagnosed NVAF and subsequently conducted Poisson regression analyses. The outcome was hospitalization for ischemic stroke. Among 389,314 patients (median age 80 years), 24,576 (6.3%) had ischemic stroke. The incidence rate peaked in winter (December-February), reaching 35.0 per 1,000 person-years, the highest among all seasons. In a Poisson regression model adjusted for CHADS₂ components, with winter treated as a binary exposure, winter remained independently associated with ischemic stroke risk (adjusted incidence rate ratio 1.06, 95% CI 1.03-1.09). Adding winter exposure to the CHADS₂ framework improved model fit but resulted in only negligible improvement in discrimination (AUC 0.690 vs. 0.691). In interaction analyses with CHADS₂ components, the association between winter exposure and ischemic stroke risk was stronger in patients without prior ischemic stroke than in those with prior ischemic stroke (p = 0.033).

In patients with NVAF, winter exposure was associated with a modest increase in ischemic stroke risk at the population level. The stronger association observed among patients without prior ischemic stroke may support seasonal risk awareness and population-level risk communication in the primary prevention setting.