BACKGROUND:Little is known about the role of wildfire smoke PM2.5 in cancer burden, despite ambient PM2.5 being classified as a human carcinogen. METHODS:Daily wildfire smoke PM2.5 estimates were aggregated into cumulative (2006-2020) and recent (2018-2020) averages, and linked to 2022 tract-level cancer history prevalence estimates for all tracts in the contiguous US. Quasi-Poisson generalized additive models with random effects and spatial splines estimated prevalence ratios and 95% CIs for each exposure window, modeling smoke PM2.5 continuously (per IQR) and across tertiles. Models adjusted for tract-level sociodemographic factors, health behaviors, health status, population density, climate, and Census region. Effect modification was assessed using interaction terms and stratified analyses. RESULTS:Smoke PM2.5 exposure was not associated with cancer history prevalence in fully adjusted models across either exposure window. However, we observed effect modification by Census region. Higher cumulative and 3-year smoke PM2.5 exposure was associated with higher cancer history prevalence in the Midwest (IQR-difference in 3-year exposure PR: 1.020; 95% CI: 1.013-1.026) and Northeast (PR: 1.040; 95% CI: 1.032-1.049). CONCLUSION:Wildfire smoke PM2.5 was not associated with cancer history prevalence at the tract level in this cross-sectional ecological study, although there may be potential effect modification by region. Further investigation using individual-level data on smoke PM2.5 exposure and cancer outcomes is needed to confirm these findings and better understand the implications of wildfires for cancer burden.
原文链接
分享
引用
