key: cord-0701233-0kcdrf3b authors: Trias-Llimos, S.; Blanes, A.; Franco, M.; Bilal, U.; Riffe, T. title: Declines in life expectancy following the COVID-19 pandemic in provinces of Spain date: 2021-04-17 journal: nan DOI: 10.1101/2021.04.15.21255545 sha: 158c6deaff0a78bb87020859687c999dc6580860 doc_id: 701233 cord_uid: 0kcdrf3b The COVID-19 pandemic has impacted population health on a global scale. Most of the studies on mortality impacts are at national level, while broad evidence exists on heterogeneous COVID-19 incidence across regions and within countries. Using Spanish data for 2020, we estimate life expectancy changes in 2020 compared with the 2017-19 period in 50 Spanish provinces. We visualize longer-term trends (1990-2020), and compare the robustness of our province-specific results with cumulative COVID-19 incidence using regional data from the Spanish ENECOVID seroprevalence study. In 2020 there was a 1.2 and 1.1 year drop in life expectancy for men and women in Spain, but this impact was heterogeneous across regions. For men these losses were highest in the province of Segovia (-3.5 years decline), while for women the highest drop was observed in Salamanca (-2.8 years decline). Life expectancy actually increased in Santa Cruz de Tenerife (+1.1 and +0.6 years for men and women, respectively). Declines in life expectancy in 2020 were also highly correlated with the cumulative seroprevalence through November 2020 ({rho}=0.80 and 0.77 in men and women, respectively). Monitoring regional life expectancy dynamics provide valuable and granular information on the heterogeneous impacts of the pandemic on health at the population level. Similar exercises in other European countries may reveal insightful geographic patterns in mortality impacts in COVID-19 pandemic years. In this study, we aimed to estimate life expectancy changes in 2020 compared with the 2017-19 period in 50 Spanish provinces with a complete calendar year of data for 2020 at a lower regional level, NUTS-3 level (provincias), accounting for all the waves contained in the first calendar year of the pandemic. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 17, 2021. ; https://doi.org/10.1101/2021.04.15.21255545 doi: medRxiv preprint We used mortality data from the "Estimación del número de defunciones semanales (EDeS) durante el brote del covid-19" from the Instituto Nacional de Estadística (INE), aggregated by 5-year age groups up to age 90+, gender, and province of occurrence (NUTS-3) for both 2019 and 2020. As the underlying electronic reporting system only captures 93% of deaths, the INE has corrected for this undercoverage using provincespecific correction weights (8) . We used populations as of July 1st, 2019 and 2020, as denominators to calculate mortality rates. To compare our results with longer term trends, we also obtained province-level life expectancy estimates for the 1990-2019 period. We applied conventional life table techniques to estimate expectancy at birth by gender and province for both 2019 and 2020 (4) . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 17, 2021. ; https://doi.org/10.1101/2021.04.15.21255545 doi: medRxiv preprint This study reported regional life expectancies for 2020 in Spain, the EU country that experienced the highest declines in life expectancy (6) . Life expectancy changes between the period 2017-19 and 2020 were found to be heterogeneous across Spanish provinces with drops up to 3.5 years among men (Segovia) and 2.8 years among women (Salamanca). This study used provisional data for 2020 on deaths reported by province of occurrence. The coverage of the original data was 93%, but the INE corrected for the province-specific undercoverage by applying weights. The data and the INE correction proves to be consistent and unbiased for 2019. That is, our life expectancy estimates using coverage-corrected estimates for 2019 are almost identical to those estimated by INE using final 2019 data (Tables S1 and S2). We should also acknowledge that, due to data limitations, the first age group used to estimate life tables was 0-4. In this case we assumed that the mean years of life lived by those dying in the interval (a0) was 0.5. This assumption implies a trivial effect on life expectancy estimates in low-mortality populations. Finally, we found a strong correlation with SARS-CoV-2 seroprevalence as of November 2020, lending validity to our findings. This study reflects for the first time the extraordinary impacts that the COVID-19 has had on most of the Spanish provinces using conventional anual life expectancy estimates. The observed drops over 1 year of life expectancy between 2017-19 and 2020 in several regions as well as the national drops have not been experienced since the Civil War, in the late 1930s (7). In the first weeks of 2021, the third COVID-19 wave has also produced an observable excess mortality at the national level, but this impact differs from the one we reported here for 2020. For example, the provinces from the Valencian Community were highly affected, while its impact on 2020 was one of the smallest. This indicates that life expectancy for 2021 may not fully recover to the levels of previous years. Figure S1 . . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted April 17, 2021. ; https://doi.org/10.1101/2021.04.15.21255545 doi: medRxiv preprint Quantifying impacts of the COVID-19 pandemic through life expectancy losses. medRxiv Years of life lost to COVID-19 in 81 countries COVID-19: a need for real-time monitoring of weekly excess deaths. The Lancet Monitoring life expectancy levels during the COVID-19 pandemic: Example of the unequal impact of the first wave on Spanish regions Life expectancy: frequently used, but hardly understood Life expectancy decreased in 2020 across the EU Estimación del número de defunciones semanales (EDeS) durante el brote de covid-19 Prevalence of SARS-CoV-2 in Spain (ENE-COVID): a nationwide, population-based seroepidemiological study. The Lancet Estudio ENE-COVID: Cuarta ronda. Estudio Nacional de sero-epidemiología de la infección por Sars-COV-2 en España