Skip to content

EHP logoISEE logo

2016 Conference

Abstract Number: O-003 | ID: 3367

Long-Term Exposure to Ambient Particulate Matter and Incident Stroke: Results from the Women’s Health Initiative

Gregory Wellenius*, Brown University, United States, gregory_wellenius@brown.edu; Melissa Eliot, Brown University, United States, melissa_eliot@brown.edu; Adam Szpiro, University of Washington, United States, aszpiro@u.washington.edu; Brent Coull, Harvard University, United States, bcoull@hsph.harvard.edu; Silas Bergen, University of Washington, United States, SBergen@winona.edu; Lesley Tinker, Fred Hutchinson Cancer Research Center, United States, ltinker@whi.org; Charles Eaton, Miriam Hospital of Rhode Island, United States, cbeaton51@gmail.com; Eric Whitsel, University of North Carolina, Chapel Hill, United States, eric_whitsel@med.unc.edu; Sverre Vedal, University of Washington, United States, svedal@u.washington.edu; Joel Kaufman, University of Washington, United States, joelk@uw.edu;
Introduction: Long-term exposure to ambient air pollution has been associated with risk of cerebrovascular events. However, cerebrovascular disease manifests in acute events with heterogeneous etiopathologies, not all of which are necessarily related to air pollution.
Methods: We used Cox models to evaluate the association between long-term residential exposure to fine (PM2.5) and respirable (PM10) particulate matter and subtypes of incident cerebrovascular events among 146,802 post-menopausal women participating in the US Women’s Health Initiative (WHI) and free of cardiovascular disease at enrollment in 1993-1998. We estimated residential annual average PM2.5 and PM10 using national geostatistical models. All analyses were adjusted for age, race, smoking, BMI, high cholesterol, family history of stroke, alcohol consumption, physical activity, markers of individual and neighborhood socioeconomic status, study component, and WHI study region at enrollment.
Results: Over a median follow-up time of 14.0 years, there were 4,089 hospitalized strokes (2,854 for ischemic stroke, 643 for hemorrhagic stroke) and 928 cerebrovascular deaths. Annual average PM2.5 levels at enrollment ranged from 3.0 to 23.7 μg/m3 with a mean ± SD of 14.1 ± 2.7 μg/m3. An interquartile range shift in PM2.5 (3.5 μg/m3) was associated with a hazard ratio (HR) of hospitalized stroke of 1.07 (95% CI: 1.03, 1.12). The data suggest a more pronounced association for hospitalized ischemic stroke (HR: 1.08; 95% CI: 1.03-1.13) than for hospitalized hemorrhagic stroke (HR: 1.03; 0.92-1.14) or cerebrovascular death (HR: 1.05; 95% CI: 0.97, 1.14). PM10 was not associated with any outcome.
Conclusions: In this cohort of US post-menopausal women, long-term residential PM2.5 was associated with rates of hospitalized ischemic stroke. These results suggest that the association between particulate matter and cerebrovascular events may differ by both particulate matter size fraction and stroke etiology.