Timonin, SergeyShartova, NataliaWen, BoWu, YaoAndreev, EvgenyGuo, YumingBallester, Joan2025-05-232025-05-232542-5196RIS:urn:E98830D3C7C89823B55C36A8517E5FD2ORCID:/0000-0001-6651-2023/work/184313197https://hdl.handle.net/1885/733753082Background: Despite a substantial body of evidence on the association between ambient temperature and mortality worldwide, there has not yet been a comprehensive country-wide assessment of the health effects of temperature in Russia. Moreover, there is no consensus on the effect of non-optimal temperatures on age-specific and sex-specific mortality. Our study aimed to provide the first analysis of temperature-related mortality in a large assembly of cities located in different geographical and socioeconomic zones of Russia. Methods: We analysed 19 044 538 non-accidental deaths in the 300 most populated cities in Russia between 2000 and 2019. A two-stage analysis strategy was used. First, a quasi-Poisson time-series model with distributed lag non-linear model was fitted to estimate city-specific associations. Second, these associations were pooled with multivariate multilevel meta-regression, from which we also calculated temperature-attributable mortality. Findings: Relative risks were generally higher for cold than for heat, except for cities in southern European Russia. Cold had a similar effect in both sexes, with a varying age gradient across cities. Although the effect of heat was generally stronger in women than in men, with the relative risk increasing steadily with age in both sexes, men younger than 60 years had a significantly higher risk of dying from heat than women of the same age. With a total of 106 007 (95% empirical CI [eCI]: 88 942–121 318) temperature-attributable deaths, there was a higher mortality attributable fraction for cold (10·74%, 95% eCI 8·80–11·99) than for heat (0·67%, 0·42–0·88). Interpretation: Russia has a high temperature-related mortality burden, with large differences in risk between cities and subpopulations. This information should be taken into account when planning public health interventions. Funding: European Research Council, National Health and Medical Research Council, and Australian Research Council.JB and NS gratefully acknowledge funding from the European Union's Horizon 2020 and Horizon Europe research and innovation programmes under grant number 865564 (ERC Consolidator Grant EARLY-ADAPT; https://www.early-adapt.eu/), support from the grant CEX2023-0001290-S funded by MCIN/AEI/10.13039/501100011033, and from the Generalitat de Catalunya through the CERCA Program. JB acknowledges funding from 101069213 (European Research Council Proof-of-Concept HHS-EWS, https://forecaster.health/) and 101123382 (European Research Council Proof-of-Concept FORECAST-AIR), and from the Spanish Ministry of Science and Innovation under grant agreement number RYC2018-025446-I (programme Ramón y Cajal). ST acknowledges support from the Australian Research Council (Discovery Project DP210100401). BW and YW are supported by China Scholarship Council funds (number 202006010043 and 202006010044). YG was supported by a Career Development Fellowship (APP1163693) and a Leader Fellowship (APP2008813) of the Australian National Health and Medical Research Council.11en© 2025 The Author(s)202510.1016/S2542-5196(25)00084-1105004919514