Long-term exposure to ultrafine particles and cause-specific mortality in the Rome Longitudinal Study

Abstract

OPS 51: Air pollution and mortality: what’s new? Beatrix Theater, August 26, 2019, 4:30 PM - 5:30 PM Background/Aim: Few studies evaluated the association between long-term exposure to ultrafine particles (UFP) and mortality. We aimed to study the association between UFP and mortality in a large administrative cohort (Rome Longitudinal Study, RoLS) and to evaluate whether the role of UFP is independent of other exposures. Methods: We selected 1,009,434 30+ year olds living in the Rome urban area, and we followed them from 2001 to 2013. We used land use regression models for PNC (particle number concentrations) as proxy for UFP, NO2, PM2.5, PM2.5 absorbance (as proxy for soot), and zinc component of PM2.5 (as proxy for non-tail-pipe emissions). We used the acoustic model SoundPLAN7.4 to assess road traffic noise indicators (Lden, Lnight). All exposures were attributed to the residential address of each subject. We analyzed non-accidental, cardiovascular, and ischemic heart disease (IHD) mortality. We applied Cox proportional hazard models adjusted for age, sex, education, place of birth, marital status, occupation, area-level socioeconomic position. We performed single-pollutant and two-pollutant models. Results: The average population exposure was 16,014 (interquartile range, iqr=2,659) n/cm3 PNC, 47 (iqr=8) µg/m3 NO2, 20 (iqr=2) µg/m3 PM2.5, 2.8 (iqr=0.3) 10–5/m PM2.5-absorbance, 25 (iqr=3.4) ng/m3 PM2.5-Zn, 61 (iqr=11) dB. The correlation between UFP and the other exposures ranged from 0.45 (UFP-noise) to 0.61 (UFP-PM2.5). We observed 166,787 deaths for non-accidental causes, 69,034 for cardiovascular causes, and 25,092 for ischemic heart disease. We estimated an increase in non-accidental (Hazard Ratio [HR]=1.010, 95%CI:1.004-1.016), cardiovascular (HR=1.029, 95%CI:1.020-1.038), and IHD (HR=1.032, 95%CI:1.016-1.047) mortality, per iqr increments in PNC. Results were robust to adjustment for all other exposures for cardiovascular and IHD mortality. For non-accidental mortality, adjustment for NO2 and PM2.5-absorbance substantially decreased the PNC effect estimates. Conclusions: We found an effect of UFP exposure on cardiovascular mortality independent of the concomitant exposure to other air pollutants and traffic noise.