Abstract
Conventional temperature-based approaches often overlook the intricate nature of thermal stress experienced by individuals. To address this limitation, climatologists have developed thermal indices-composite measures designed to reflect the complex interaction of meteorological factors influencing human perception of temperature. Our study focuses on Brazil, estimating the association between thermal comfort conditions and mortality related to respiratory and circulatory diseases. We examined four distinct thermal indices: the discomfort index (DI), net effective temperature (NET), humidex (H), and heat index (HI). Analyzing a comprehensive dataset of 2,872,084 deaths from 2003 to 2017, we found significant variation in relative risk (RR) based on health outcomes, exposure lag, percentile of exposure, sex/age groups, and specific thermal indices. For example, under high exposure conditions (99th percentile), we observed that the shorter lags (3, 5, 7, and 10) had the most robust effects on all-cause mortality. For example, under lag 3, the pooled national results for the overall population (all ages and sexes) indicate an increased risk of all-cause mortality, with an RR of 1.17 (95% CI: 1.13; 1.122) for DI, 1.15 (95% CI: 1.12; 1.17) for H, 1.15 (95% CI: 1.09; 1.21) for HI, and 1.18 (95% CI: 1.13; 1.22) for NET. At low exposure levels (1st percentile), all four distinct thermal indices were linked to an increase in all-cause mortality across most sex and age subgroups. Specifically, for lag 20, we observed an estimated RR of 1.19 (95% CI: 1.14; 1.23) for DI, 1.12 (95% CI: 1.08; 1.16) for H, 1.17 (95% CI: 1.12; 1.22) for HI, and 1.18 (95% CI: 1.14; 1.23) for NET. These findings have important implications for policymakers, guiding the development of measures to minimize climate change's impact on public health in Brazil.