Environment/Climate Change
Estimate Joint Effects of Time-varying Air Pollutant Mixtures on Mortality in Canada Using a Novel Methodologic Approach Juwel Rana* Juwel Rana Alexander P. Keil Hong Chen Tarik Benmarhnia Scott Weichenthal Jay S. Kaufman
Background and Objective: Studies have established the association between total PM2.5 mass and mortality in Canada. However, no research has specifically investigated the joint effects of PM2.5 constituents on mortality in the Canadian context. This study examines the joint effects of time-varying PM2.5 constituent mixtures on nonaccidental all-cause and cardiovascular disease (CVD)-related mortality in Canada. Methods and Materials: We developed a novel approach: time-varying quantile g-computation (tvcQGcomp), which estimates the joint effects of increasing all exposures in a quantile across all time points. To evaluate tvcQGcomp, we conducted simulations involving nonaccidental deaths; time-varying black carbon (BC), sulfate (SO₄²⁻), ammonium (NH₄⁺), nitrate (NO₃⁻), and organic matter (OM), and both time-fixed and time-varying confounders. We then constructed a population-based cohort involving approximately 3 million adults (ages 25 to 89) from the Canadian Census Health and Environment Cohort (CanCHEC, comprising 3,070,547 observations followed from 2007 to 2019. We estimated annual mean concentrations of BC, SO₄²⁻, NH₄⁺, NO₃⁻, and OM using satellite data and a global atmospheric chemistry transport model, assigned by annual postal code. Results: Simulation results show that tvcQGcomp yields unbiased joint effect estimates for time-varying mixtures, with ~95% confidence interval coverage and acceptable type I error rates (5–7%). The correlation among these five PM2.5 constituents ranged from 0.20 to 0.91. We will apply tvcQGcomp to estimate 13-year nonaccidental all-cause and CVD-related mortality risks associated with the time-varying PM2.5 constituents. Conclusion: Our novel approach can elucidate dose-response relationships and effect estimates to better inform composition-specific targeted public health policies across Canada.