How droughts affect human health: mortality impacts attributed to events of different time scales and vulnerability drivers 

Epidemiological evidence on the effects of droughts on human health is limited and heterogeneous, and drivers of vulnerability are still uncertain. The IGIA-SETH project aims to address these research gaps by using advanced epidemiological models and unique health and climate datasets. In particular, the present study aims to estimate drought-related mortality risks and identify vulnerability patterns on a global scale, using a robust and common approach and a large multi-location mortality dataset.We analyse mortality data from 832 locations distributed around the world with a wide range of climatic, demographic and socioeconomic characteristics over the period 1969-2019. We use a two-stage time series analytical design with a quasi-Poisson regression and a threshold function to model the association between droughts and mortality. Droughts at short and long- time scales are defined using the Standardized Precipitation Evaporation Index (SPEI) computed at one- and twelve-month accumulation periods. Potential effect modification by climatic, demographic, socioeconomic and environmental factors are also evaluated.Our findings suggest that extreme short-term and long-term drought events are associated with an increased mortality risk at 1% (95% confidence interval: 0.7%-1.2%) and 0.7% (0.01%-1.3%), respectively, at a SPEI=-2 vs. SPEI=0. Countries with higher mean temperatures and lower annual precipitation show a higher vulnerability to short-term droughts, while for long-term droughts, higher vulnerability is mostly found in countries with lower temperature range, lower annual average precipitation, and with a higher Gross Domestic Product per Capita.To our knowledge, this study represents the first comprehensive quasi-global analysis providing robust evidence of increased mortality risk associated with different drought exposures. Different mechanisms interacting at different levels, as well as different distribution of climatic, socioeconomic and demographic vulnerability factors between countries can driver disparities in drought-related mortality risks worldwide.