Examining Fine Particulate Matter and Cause-Specific Morbidity During the 2017 North San Francisco Bay Wildfires

Abstract Background Recent increases in wildfire frequency and severity necessitate better understanding of health effects of wildfire smoke to protect affected populations. Objectives We examined relationships between fine particulate matter (PM2.5) and morbidity during wildfires in California, and whether those relationships differed during the fire compared to a similar non-fire period. Methods For nine San Francisco Bay Area counties, daily county-level diagnosis-specific counts of emergency department visits (EDVs) and hospitalizations were linked with county-level estimates of daily mean PM2.5 during the October 2017 Northern California wildfires and similar October days in 2015, 2016, and 2017. Associations were estimated using Poisson regression. Results The median difference between county PM2.5 during the fire versus the non-fire period was 23.4 μg/m3, with days exceeding 80 μg/m3 in some counties. Over the entire study period, PM2.5 was most consistently linked to EDVs for respiratory disease (RREDV(lag0) per 23.4 μg/m3 increase: 1.25, 95% CI: 1.21, 1.30), asthma, chronic lower respiratory disease (CLRD; RREDV(lag0): 1.18, 95% CI: 1.10, 1.27), and acute myocardial infarction (RREDV(lag0): 1.14, 95% CI: 1.03, 1.25). Increases in acute upper respiratory infections and decreases in mental/behavioral EDVs were observed but were sensitive to model specification, specifically the inclusion of time-related covariates. Comparing fire and non-fire period EDV associations, we observed indications that PM2.5 during the fire was more strongly associated with asthma (RRlag0: 1.46, 95% CI: 1.38, 1.55) compared to non-fire period PM2.5 (RRlag0: 0.77, 95% CI: 0.55, 1.08), with the opposite observed for dysrhythmia, with the asthma difference being particularly robust to model choice. For hospitalizations, the most robust PM2.5 relationships were positive associations with respiratory, CLRD, and diabetes, and inverse associations with pneumonia. Respiratory and CLRD effect estimates were generally similar or smaller than for EDVs. Conclusions Elevated short-term PM2.5 levels from wildfire smoke appears to impact respiratory and other health domains.