Community-level risk factors for temperature-related mortality in France.

Background: Vulnerability to nonoptimal temperatures varies from one geographical location to another, but the contextual factors accounting for these spatial differences are still poorly understood. We aimed to identify the community-level characteristics contributing to geographical disparities in heat-related and cold-related mortality risk in France. Methods: We conducted a country-wide analysis using data on all-cause mortality, temperature, and contextual characteristics across 1,967 pseudo-cantons in France between 2004 and 2019. We first estimated the daily temperature-mortality association in each pseudo-canton using a time-series quasi-Poisson regression in combination with distributed lag nonlinear models, and then we fitted univariable and multivariable multivariate meta-regression models to assess the effect modification of the contextual factors on heat-related and cold-related mortality risk. Findings: Over the 16-year study period, metropolitan France recorded 8,807,376 deaths out of an average population of 63·2 million inhabitants, which corresponds to an average annual mortality rate of 8.7 per 1,000 people. The country-level percent change (%CR) in mortality risk at the 1st and 99th daily temperature percentiles versus the minimum mortality temperature was, respectively, 31.2% (95% CI = 29.0, 33.5) and 11.0% (95% CI = 9.4, 15.5). The mortality risk associated with low temperatures was not modified by any of the contextual factors considered in the study, while the mortality risk associated with high temperatures was independently modified by NO Interpretation: This study suggests that urban areas in France are more vulnerable to heat, compared to rural communities, and that this disparity is probably driven by air pollution (NO