Air pollution and stroke: A two-year analysis linking short-term particulate matter 2.5 exposure to stroke severity and etiology

Background Air pollution is increasingly implicated in cerebrovascular disease through mechanisms of systemic inflammation, endothelial dysfunction, and neurovascular injury. Although long-term exposure to particulate matter contributes to stroke risk, the impact of short-term fluctuations, particularly during extreme pollution events such as wildfires, on stroke severity and etiology remains poorly defined. Methods We conducted a retrospective study of adults with imaging-confirmed strokes admitted to a single center in the mid-Atlantic United States from 2022–2023. Daily particulate matter 2.5 (PM 2.5 ) and ozone levels from Environmental Protection Agency monitors were assigned based on hospital presentation date and evaluated as both categorical (above vs. below median) and continuous exposures. Outcomes included daily stroke incidence, initial National Institutes of Health Stroke Scale (NIHSS), change in modified Rankin Scale (ΔmRS), length of stay, and stroke etiology. Associations were assessed using multivariable logistic and linear regression models adjusted for age, sex, and race, with PM 2.5 modeled continuously. Lagged exposures of 1–2 days were also examined. Results Among 1,184 patients, presentation on high PM2.5 days was associated with greater stroke severity, including higher initial NIHSS (10.0 vs. 8.25; p=0.004) and greater functional decline by discharge (ΔmRS 2.6 vs. 2.15; p=0.001). In multivariable linear regression adjusting for age, sex, race, smoking status, and stroke etiology, higher PM2.5 levels remained independently associated with increased stroke severity, with each 1 µg/m³ increase corresponding to higher NIHSS (β = 0.047, 95% CI 0.027–0.067; p < 0.001). Hemorrhagic stroke was more frequent on high PM2.5 days (p=0.027), while overall ischemic stroke incidence did not differ. LAA stroke was more common during high PM2.5 exposure (adjusted OR 1.45; 95% CI 1.08–1.86). Lagged analyses demonstrated significant associations at two days for ischemic stroke incidence, NIHSS, and ΔmRS (all p