Association of aerosols, trace gases and black carbon with mortality in an urban pollution hotspot over central Indo-Gangetic Plain

摘要

The short term effect of multiple air pollutants e.g. aerosols (black carbon, BC; PM2.5 and PM10) and trace gases (NO2, SO2, and O-3) on all-cause mortality was systematically investigated in a typical urban pollution hotspot over central Indo-Gangetic Plain (IGP). To our knowledge, this would be the first report of mortality estimates for exposure to BC aerosols and multiple trace gases over South Asia. Daily all-cause mortality and ambient air quality were analyzed from 2009 to 2016 following a semiparametric quasi-Poisson regression model adjusting mean temperature (Tmean), relative humidity (RH), and long term time trend (Time) as potential confounders. Single pollutant model clearly established the significant impact of BC aerosols (against 10-unit increase in pollutant; 4.95%, 95% CI: 2.16-7.74), NO2 (2.38%, 95% CI: 0.88-3.87%) and PM2.5 exposure (1.06%, 95% CI: 0.45-1.66%) on mortality. The inclusion of co-pollutants in the multi-pollutant model increased the individual mortality risks for BC aerosols (7.3%). Mortality estimates were further stratified considering different effect modifiers viz. sex, age, place of death, and season. Almost in all the cases statistically insignificant differences in effect modification were noted for all the pollutants except PM10. We also explored a distributed lag nonlinear model to estimate the lag effect and all the pollutants showed significant lag up to 3 days while BC showed lag effect up to 5 days. The exposure-response curves for individual air pollutants were mostly linear, while a considerable increase in mortality was noted for an exposure 15 mu gm(-3) for BC aerosols and 60 mu gm(-3); 3 for PM2.5. The effect estimates of air pollutants during haze and no-haze days were also defined. During haze days, mortality rose to 6.11% and 3.06% for each 10-unit increase in BC and NO2 exposure, respectively. Significant effect of BC aerosol exposure on human mortality was established which reaffirms its inclusion as a potential health regulator for epidemiological studies.