Incident cardiovascular disease and long-term exposure to source-specific air pollutants in a Swedish cohort.

ABSTRACT

BACKGROUND: Air pollution is associated with cardiovascular morbidity and mortality, but its role in the development of congestive heart failure (CHF) and the role of different pollution sources in cardiovascular disease remain uncertain. METHODS: Participants were enrolled in the Malmö Diet and Cancer cohort in 1991-1996 with information on lifestyle and clinical indicators of cardiovascular disease. The cohort participants were followed through registers until 2016. Annual total and local source-specific concentrations of particulate matter less than 10 µm and 2.5 µm (PM10 and PM2.5), black carbon (BC), and nitrogen oxides (NOx) from traffic, residential heating, and industry were assigned to each participant's address throughout the study period. Cox proportional hazards models adjusted for possible confounders was used to estimate associations between air pollution 1-5 years prior to outcomes of incident CHF, fatal myocardial infarction (MI), major adverse coronary events (MACE), and ischemic stroke. RESULTS: Air pollution exposure levels (mean annual exposures to PM2.5 of 11 µg/m3 and NOx of 26 µg/m3) within the cohort were moderate in terms of environmental standards. After adjusting for confounders, we observed statistically significant associations between NOx and CHF (hazard ratio [HR] 1.11, 95% confidence interval [CI] 1.01-1.22) and NOx and fatal MI (HR 1.10, 95%CI 1.01-1.20) per interquartile range (IQR) of 9.6 µg/m3. In fully adjusted models, the estimates were similar, but the precision worse. In stratified analyses, the associations were stronger in males, ever-smokers, older participants, and those with baseline carotid artery plaques. Locally emitted and traffic-related air pollutants generally showed positive associations with CHF and fatal MI. There were no associations between air pollution and MACE or stroke. DISCUSSION/CONCLUSION: In an area with low to moderate air pollution exposure, we observed significant associations of long-term residential NOx with increased risk of incident CHF and fatal MI, but not with coronary events and stroke.