Long-term exposure to PM2.5 and all-cause mortality in India using a Difference in Difference approach
Year: 2023
Type: conference-abstract
Abstract: BACKGROUND AND AIM: Global Burden of Disease estimated 1.7 million deaths were attributed to air pollution in India in 2019. In contrast, a retrospective cohort study ‘Million Death Study’ reported no excess deaths from all non-accidental causes due to ambient air pollution. These studies used exposure-response functions derived from countries with low air pollution exposure levels. To strengthen the evidence on health effects of air pollution, we studied the association between annual exposure to particulate matter measuring 2.5 µg/m3 (PM2.5) and all-cause mortality in India using a modified difference-in-difference approach. METHOD: We used district level annual all-cause mortality and annual population weighted district-level PM2.5 exposure from a high-resolution spatiotemporal model for 2009- 2020. Difference-in-difference, a causal modeling approach that controls for unmeasured confounders by design was employed to analyze effect of PM2.5 and mortality. We performed sensitivity analyses using state-wide death registration levels to account for measurement error. We estimated exposure-response function from our main model. RESULTS: In 655 Indian districts, the population weighted PM2.5 median was 38.9µg/m3 (19.7 (5th percentile) and 71.8 (95th percentile)) over a decade. Over the study period, death registration levels ranged from 68.7 to 83.7% with less than 10 states having 100% registration through out. For every 10 µg/m3 increase in PM2.5, we estimated that mortality rate increased by 5.9% (95% CI 3.8,8.1) which were robust to sensitivity analyses. The exposure-response function was slightly non-linear, with steeper slope up to ~ 95 µg/m3 and levelling afterward. CONCLUSIONS: Using a causal modelling approach allowed us to harness the full extent of registered deaths in one of the world’s most populated and polluted country, we provide new evidence of increased risk of all-cause mortality from long-term exposure to PM2.5 in India. We expect these results to have important implications for current environmental health policy in India.
Source: ISEE Conference Abstracts
Language: English
Subfield: Health, Toxicology and Mutagenesis
Field: Environmental Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4390965304