Long-term exposure to low levels of air pollution and mortality adjusting for road traffic noise: A Danish Nurse Cohort study

BACKGROUND: The association between air pollution and mortality is well established, yet some uncertainties remain: there are few studies that account for road traffic noise exposure or that consider in detail the shape of the exposure-response function for cause-specific mortality outcomes, especially at low-levels of exposure.

OBJECTIVES: We examined the association between long-term exposure to particulate matter [(PM) with a diameter of <2.5 µm (PM2.5), <10 µm (PM10)], and nitrogen dioxide (NO2) and total and cause-specific mortality, accounting for road traffic noise.

METHODS: We used data on 24,541 females (age > 44 years) from the Danish Nurse Cohort, who were recruited in 1993 or 1999, and linked to the Danish Causes of Death Register for follow-up on date of death and its cause, until the end of 2013. Annual mean concentrations of PM2.5, PM10, and NO2 at the participants' residences since 1990 were estimated using the Danish DEHM/UBM/AirGIS dispersion model, and annual mean road traffic noise levels (Lden) were estimated using the Nord2000 model. We examined associations between the three-year running mean of PM2.5, PM10, and NO2 with total and cause-specific mortality by using time-varying Cox Regression models, adjusting for individual characteristics and residential road traffic noise.

RESULTS: During the study period, 3,708 nurses died: 843 from cardiovascular disease (CVD), 310 from respiratory disease (RD), and 64 from diabetes. In the fully adjusted models, including road traffic noise, we detected associations of three-year running mean of PM2.5 with total (hazard ratio; 95% confidence interval: 1.06; 1.01-1.11), CVD (1.14; 1.03-1.26), and diabetes mortality (1.41; 1.05-1.90), per interquartile range of 4.39 μg/m3. In a subset of the cohort exposed to PM2.5 < 20 µg/m3, we found even stronger association with total (1.19; 1.11-1.27), CVD (1.27; 1.01-1.46), RD (1.27; 1.00-1.60), and diabetes mortality (1.44; 0.83-2.48). We found similar associations with PM10 and none with NO2. All associations were robust to adjustment for road traffic noise.

DISCUSSION: Long-term exposure to low-levels of PM2.5 and PM10 is associated with total mortality, and mortality from CVD, RD, and diabetes. Associations were even stronger at the PM2.5 levels below EU limit values and were independent of road traffic noise.