Wildland Fire-Related Smoke PM2.5 and Cardiovascular Disease ED Visits in the Western United States

BackgroundThe impact of short-term exposure to fine particulate matter (PM2.5) due to wildland fire smoke on the risk of cardiovascular disease (CVD) remains unclear. We investigated the association between short-term exposure to wildfire smoke PM2.5 and Emergency Department (ED) visits for acute CVD in the Western United States from 2007 to 2018. MethodsED visits for primary or secondary diagnoses of atrial fibrillation (AF), acute myocardial infarction (AMI), heart failure (HF), stroke, and total CVD were obtained from hospital associations or state health departments in California, Arizona, Nevada, Oregon, and Utah. ED visits included those that were subsequently hospitalized. Daily smoke, non-smoke, and total PM2.5 were estimated using a satellite-driven multi-stage model with a high resolution of 1 km. The data were aggregated to the zip code level and a case-crossover study design was employed. Temperature, relative humidity, and day of the year were included as covariates. ResultsWe analyzed 49,759,958 ED visits for primary or secondary CVD diagnoses, which included 6,808,839 (13.7%) AFs, 1,222,053 (2.5%) AMIs, 7,194,474 (14.5%) HFs, and 808,396 (1.6%) strokes. Over the study period from 2007-01-01 to 2018-12-31, the mean smoke PM2.5 was 1.27 (Q1: 0, Q3: 1.29) {micro}g/m3. A 10 {micro}g/m3 increase in smoke PM2.5 was associated with a minuscule decreased risk for AF (OR 0.994, 95% CI 0.991-0.997), HF (OR 0.995, 95% CI 0.992-0.998), and CVD (OR 0.9997, 95% CI 0.996-0.998), but not for AMI and stroke. Adjusting for non-smoke PM2.5 did not alter these associations. A 10 {micro}g/m3 increase in total PM2.5 was linked to a small increased risk for all outcomes except stroke (OR for CVD 1.006, 95% CI 1.006-1.007). Associations were similar across sex and age groups. ConclusionWe identified an unexpected slight lower risk of CVD ED visits associated with short-term wildfire smoke PM2.5 exposure. Whether these findings are due to methodological issues, behavioral changes, or other factors requires further investigation.