Mortality Burden Of Ambient PM2.5 Exposure In Megacity Delhi: Spatially Disaggregated v/s Aggregated Approach To Exposure Assignment

Abstract:

Exposure to fine particulate matter (PM_2.5) presents a major environmental and public health challenge for India. Delhi, India’s capital is the world’s second-most populous urban agglomeration. Delhi is infamous for poor air quality, with annual PM_2.5 concentration exceeding 8-10 times of WHO air quality guideline. The adverse human physiological effects of PM_2.5 exposure are now scientifically well documented across the developed world. However, such evidence is quite limited for the Indian region, partly due to the lack of consistent and long-term ground-based PM_2.5 measurements and partly due to the lack of systematic health outcome data. Here we examine the short-term impacts of PM_2.5 exposure on all-cause mortality in the megacity Delhi.

We use daily average PM_2.5 concentration derived from the satellite data and all-cause non-trauma mortality count data for Delhi over 2013-2016. Satellite data are used in this analysis because ground-based PM_2.5 measurements before 2015 are available only from a few sites and are not adequate for exposure assessment. We use variants of quasi-Poisson generalized additive model adjusted for non-linear confounding of time, temperature and relative humidity to estimate the short term impact of PM_2.5 exposure on daily non-trauma mortality count. The analysis is carried out at two levels: 1) for Delhi as a whole with the state average PM_2.5 concentration and mortality count and 2) Delhi district wise spatial disaggregates with each death event assigned district-specific exposure. A significant association of all-cause non-trauma mortality with short-term exposure to PM_2.5 is observed at both levels. However, the estimated impact (quantified in terms of percentage increase in all-cause mortality per 10 mg m^-3 rise in PM_2.5) of spatially disaggregated exposure at district level (0.51%, 95^% CI = 0.3%-0.72%) is found to be double the estimated impact of aggregated exposure for Delhi as a whole (0.27%, 95% CI=0.18%-0.36%). We note these impacts are much higher than the impact observed for PM₁₀ previously in the PAPA study. This is the first evidence of PM_2.5 impact on mortality burden in Delhi. Our results suggest an immediate health benefit from any reduction in PM_2.5, especially during the peak pollution season from October-January.