Health Effects of Air Pollutant Mixtures on Overall Mortality Among the Elderly Population Using Bayesian Kernel Machine Regression (BKMR)

Abstract

BACKGROUND AND AIM: It is well documented that air pollutants are associated with a range of adverse health outcomes. In reality, humans are exposed to multiple pollutants that are likely highly correlated with each other. However, the majority of the existing research has primarily focused on the health effect of a single pollutant, partially because we lack the methodology to evaluate the complicated relationship of mixture exposures. METHODS: We collected the health data of 1,406,185 Medicare enrollees residing in North Carolina, South Carolina, and Georgia through the Centers for Medicare and Medicaid Services. Ambient PM2.5, O3, and NO2 concentrations, derived from a well-validated ensemble machine learning model, were assigned to individuals based on their ZIP codes. We assessed the individual and joint effect of air pollutant mixtures (PM2.5, O3, and NO2) on all-cause mortality by applying a novel mixture modeling approach, Bayesian Kernel machine regression (BKMR). RESULTS:We observed a statistically significant adverse effect of the multi-pollutant mixture (PM2.5, O3, and NO2) on overall mortality. We found significant evidence for the association between PM2.5 and increased mortality on the population level in this study; the positive association with mortality appears stronger at lower percentiles of other pollutants. An interquartile range (IQR) change in PM2.5 concentration was associated with a significant increase in mortality of 1.7 (95% CI: 0.5, 2.9), 1.6 (95% CI: 0.4, 2.7), and 1.4 (95% CI: 0.1, 2.6) SDs when O3 and NO2 were set at the 25th, 50th, and 75th percentiles, respectively. CONCLUSIONS:This finding suggests a strong association between pollutant mixture and all-cause mortality, mainly driven by PM2.5. BKMR analysis did not identify statistically significant interactions among PM2.5, O3, and NO2. However, since the small sub-population might weaken the study power, additional studies (in larger sample size and other regions in the US) are in need to reinforce the current finding. KEYWORDS: Air pollution, Mixtures, Long-term exposure, Mortality, Bayesian kernel machine regression (BKMR)