A0886
Title: Comparing spatiotemporal temperature models for heat-related mortality risk assessment in Lazio, Italy
Authors: Emiliano Ceccarelli - Sapienza University of Rome (Italy) [presenting]
Giovanna Jona Lasinio - Sapienza University of Rome (Italy)
Giada Minelli - Statistical Service - Italian National Health Institute (Italy)
Marta Blangiardo - Imperial College London (United Kingdom)
Jorge Castillo-Mateo - University of Zaragoza (Spain)
Sandra Gudziunaite - Imperial College London (United Kingdom)
Abstract: The aim is to investigate how different spatiotemporal temperature models affect the estimation of heat-related mortality in Lazio, Italy (2008-2022). Three methods are compared to reconstruct daily maximum temperature at the municipality level: Two Bayesian station-based approaches, a quantile autoregressive model with spatial interpolation, a Gaussian model via INLA-SPDE, and a satellite-based method using ERA5. Station-based models show higher and more spatially variable temperatures than satellite-based ones, especially in warmer provinces. Using individual mortality data for cardiovascular and respiratory causes, temperature-mortality associations are estimated through Bayesian conditional Poisson models in a case-crossover design. Exposure is defined as the mean maximum temperature over the previous three days. Additional models include heatwave definitions combining different thresholds and durations. All models show a marked increase in relative risk at high temperatures, but the temperature of minimum risk varies notably across methods. Station-based models estimate higher minimum-risk temperatures compared to ERA5. Stratified analyses reveal higher RR increases in females and the elderly (80+). Heatwave effects depend on definitions, but all methods capture the prolonged heat exposure effect. Overall, results confirm the importance of temperature model choice in epidemiology and provide insights for early warning systems and climate-health adaptation strategies.
