Background: Indian cities are among the most polluted globally, yet assessments of short term mortality impacts due to pollution have been limited. Furthermore, studies examining temperature-pollution interactions on mortality are largely absent. Addressing this gap remains important in providing research evidence to better link health outcomes and air quality standards for India.

Methods: Daily all-cause mortality, temperature, humidity and particulate matter less than 10 microns (PM10) data were collected for five cities-Ahmedabad, Bangalore, Hyderabad, Mumbai and Shimla spanning 2005 - 2012. Poisson regression models were developed to study short term impacts of PM10 as well as temperature-pollution interactions on daily all-cause mortality.

Results: We find that mortality associated with a 10 µg/m3 PM10 increase is highest for Shimla (1.36%, 95% confidence interval = -0.38% to 3.1%) and the least for Ahmedabad (0.16%, 95% CI = -0.31% to 0.62%). The corresponding values for Bangalore, Hyderabad and Mumbai are 0.22% (-0.04% - 0.49%), 0.85% (0.06% - 1.63%) and 0.2% (0.1% - 0.3%) respectively. The relative health benefits of reducing pollution are higher for cleaner cities (Shimla) as opposed to dirtier cities (Mumbai). Overall we find that temperature and pollution interactions do not significantly impact mortality for the cities studied.

Conclusions: This is one of the first multi-city studies that assess heterogeneity of air pollution impacts and possible modification due to temperature in Indian cities that are spread across climatic regions and topographies. Our findings highlight the need for pursuing stringent pollution control policies in Indian cities to minimize health impacts.