Short-term effects of PM2.5, PM10 and PM2.5–10 on daily mortality in the Netherlands

INTRODUCTION: Information on the relationship between levels of particulate matter (PM) smaller than 2.5 μm and mortality rates in Europe is relatively sparse because of limited availability of PM2.5 measurement data. Even less information is available on the health effects attributable to PM2.5-10, especially for North-West Europe. OBJECTIVES: To investigate the relationship between various PM size fractions and daily mortality rates. METHODS: Daily concentrations of PM from the Dutch National Ambient Air Quality Monitoring Network as well as all cause and cause-specific mortality rates in The Netherlands were obtained for the period 2008-2009. Poisson regression analysis using generalized additive models was used, with adjustment for potential confounding including long-term and seasonal trends, influenza incidence, meteorological variables, day of the week, and holidays. Different measures of PM (PM2.5, PM10 and PM2.5-10) were analysed. RESULTS: PM10 and PM2.5 levels were statistically significantly (p<0.05) associated with all cause and cause-specific deaths. For example, a 10 μg/m(3) increase in previous day PM was associated with 0.8% (95% CI 0.3-1.2) excess risk in all cause mortality for PM2.5 and a 0.6% (CI 0.2-1.0) excess risk for PM10. No appreciable associations were observed for PM2.5-10. Effects of PM10, and PM2.5 were insensitive to adjustment for PM2.5-10, and vice-versa. PM10 and PM2.5 were too highly correlated to disentangle their independent effects. CONCLUSIONS: PM10 and PM2.5 both were significantly associated with all cause and cause-specific mortality. We were unable to demonstrate significant effects for PM2.5-10, possibly due to the lower temporal variability and the higher exposure misclassification in PM2.5-10 compared to PM10 or PM2.5. The lack of effects of PM2.5-10 in our study should therefore not be interpreted as an indication that PM2.5-10 can be considered harmless.