Exposure to ambient air pollution and the incidence of congestive heart failure and acute myocardial infarction: a population-based study of 5.1 million Canadian adults living in Ontario
Bai L, Shin S, Burnett RT, Kwong JC, Hystad P, van Donkelaar A, Goldberg MS, Lavigne E, Copes R, Martin RV, Kopp A, Chen H. Environ Int. 2019; 132:105004. Epub 2019 Aug 3. DOI: https://doi.org/10.1016/j.envint.2019.105004.
Long-term exposure to ambient air pollution has been linked to cardiovascular mortality, but the associations with incidence of major cardiovascular diseases are not fully understood, especially at low concentrations. We aimed to investigate the associations between exposure to fine particulate matter (PM2.5), nitrogen dioxide (NO2), ozone (O3), redox-weighted average of NO2 and O3 (Ox) and incidence of congestive heart failure (CHF) and acute myocardial infarction (AMI). Our study population included all long-term residents aged 35-85 years who lived in Ontario, Canada, from 2001 to 2015 (~5.1 million). Incidence of CHF and AMI were ascertained from validated registries. We assigned estimates of annual concentrations of pollutants to the residential postal codes of subjects for each year during follow-up. We estimated hazard ratios (HRs) and 95% CIs for each pollutant separately using Cox proportional hazards models. We examined the shape of concentration-response associations using shape-constrained health impact functions. From 2001 to 2015, there were 422,625 and 197,628 incident cases of CHF and AMI, respectively. In the fully adjusted analyses, the HRs of CHF corresponding to each interquartile range increase in exposure were 1.05 (95% CI: 1.04-1.05) for PM2.5, 1.02 (95% CI: 1.01-1.04) for NO2, 1.03 (95% CI: 1.02-1.03) for O3, and 1.02 (95% CI: 1.02-1.03) for Ox, respectively. Similarly, exposure to PM2.5, O3, and Ox were positively associated with AMI. The concentration-response relationships were different for individual pollutant and outcome combinations (e.g., for PM2.5 the relationship was supralinear with CHF, and linear with AMI).
View full text