Long-term effects of fine particulate PM2.5 estimated from NASA satellite remote sensor data on mortality in an elderly cohort in Hong Kong

Abstract

Background The limited available studies have suggested long-term effects of air pollution are several-fold those of short-term effects. In Asia where air quality is deteriorating fast due to increasing reliance on burning of fossil fuel for transportation and electricity generation, there are no data describing the long-term effects. Aim To assess long-term effects of PM2.5 on mortality in acohort of the older population in Hong Kong Methods This prospective cohort study included 66,820 participants aged 65 or older who were enrolled and interviewed in all the 18 government Elderly Health Centres of Hong Kong in 1998-2001 and were followed up for mortality outcomes until 2011. Their proxy exposures to PM2.5 were estimated from NASA’s satellite surface extinction coefficients at 1x1 km2 resolution at their geo-coded residential address. Cox regression models using time on study as timescale were fitted to assess the hazard ratios (HR) of mortality associated with PM2.5 adjusted for potential confounding factors. Results The proxy exposures of individuals to PM2.5 (?g/m3) ranged from 21.5 to 63.9 with an inter-quartile range of 32.6 to 38.2 and a median of 35.0. The excess risks (i.e. HR-1) in percentage (%) (95% CI) of mortality associated with 10 ?g/m3 increase in PM2.5 for all-natural causes [ICD10:A00-R99; n=17,599] and cardiorespiratory diseases [I00-I99, J00-J47; n=8,532] were 6.3 (2.5, 10.0) and 8.3 (2.8, 13.8), respectively. Concentration-response relationships were observed particularly in all-natural cause mortality with excess risks at the 2nd- 4th quartile relative to the 1st quartile of 1.4 (-4.2, 4.6), 2.9 (-1.5, 7.5) and 5.0 (0.5, 9.7), respectively. Conclusions Estimates of long-term health effects of air pollution in Hong Kong, the most urbanized Chinese city in Asia, are comparable with those from North America and Europe. The results will provide important but much overdue information for proper air quality management and effective risk communication. Acknowledgement We thank Wellcome Trust for fnding of this study.