Many diseases are affected by changes in weather. There have been limited studies, however, which have examined the relationship between heat and cold and cause-specific mortality in low and middle-income countries. In this study, we aimed to estimate the effects of heat and cold days on total and cause-specific mortality in the Vadu Health and Demographic Surveillance System (HDSS) area in western India. We used a quasi-Poisson regression model allowing for over-dispersion to examine the association of total and cause-specific mortality with extreme high (98th percentile, >39 °C) and low temperature (2nd percentile, <25 °C) over the period January 2003 to December 2012. Delays of 0 and 0-4 days were considered and relative risks (RR) with 95% confidence intervals (CI) were calculated. Heat was significantly associated with daily deaths by non-infectious diseases (RR = 1.57; CI: 1.18-2.10). There was an increase in the risk of total mortality in the age group 12-59 years on lag 0 day (RR = 1.43; CI: 1.02-1.99). A high increase in total mortality was observed among men at lag 0 day (RR = 1.38; CI: 1.05-1.83). We did not find any short-term association between total and cause-specific mortality and cold days. Deaths from neither infectious nor external causes were associated with heat or cold. Our results showed a strong and rather immediate relationship between high temperatures and non-infectious disease mortality in a rural population located in western India, during 2003-2012. This study may be used to develop targeted interventions such as Heat Early Warning Systems in the area to reduce mortality from extreme temperatures.
Keywords: India; cause-specific mortality; cold; heat; mortality; temperature.