Long-term exposure to ambient particulate matter components and mortality: results from six European administrative cohorts within the ELAPSE project

Abstract

BACKGROUND AND AIM: Evidence of the association between long-term exposure to ambient particulate matter components and mortality is inconsistent. We evaluated this association in six large administrative cohorts in the Effects of Low-Level Air Pollution: A Study in Europe (ELAPSE) project. METHODS: We analysed data from country-wide administrative cohorts in Norway, Denmark, the Netherlands, Belgium, Switzerland and in Rome (Italy). Annual 2010 mean concentrations of copper (Cu), iron (Fe), sulfur (S) and zinc (Zn) in PM2.5 were estimated using 100x100m Europe-wide land use regression models assigned to the participants’ residential addresses. We applied cohort-specific Cox proportional hazard models controlling for area- and individual-level covariates to evaluate associations with natural, cardiovascular, non-malignant respiratory and lung cancer mortality. Two pollutant models adjusting for PM2.5 or NO2 were also applied. We pooled cohort specific estimates using a random effects meta-analysis. RESULTS:We included 27 million participants contributing more than 240 million person-years. Cu, Fe and S were associated with natural mortality [Hazard Ratios (HRs) (95% CI): 1.037 (1.014, 1.060) per 5 ng/m³ Cu; 1.069 (1.031, 1.108) per 100 ng/m³ Fe; 1.036 (1.016, 1.057) per 200 ng/m³ S] even after controlling for PM2.5 [e.g. 1.048 (1.022, 1.074) for Fe]. Adjustment for NO2 decreased HRs to unity [e.g. 0.994 (0.954, 1.036) for Fe]. Associations with cardiovascular mortality were weaker and non-significant. We found positive significant associations of Cu, Fe and S with non-malignant respiratory [HRs: 1.047, 1.099 and 1.043, respectively] and lung cancer mortality [HRs: 1.091, 1.160 and 1.102, respectively] in single pollutant models, though only the effect of Fe persisted after control for PM2.5. CONCLUSIONS:Our results indicate associations of long-term exposure to PM2.5 components with natural and cause-specific mortality outcomes related to non-tailpipe emissions (Cu and Fe) and long-range transport (S). However, associations did not persist after NO2 adjustment. KEYWORDS: Air pollution, Particle components, Long-term exposure, Environmental epidemiology, Survival, Mortality