key: cord-0840621-kl3mfkoh authors: Woolf, Steven H.; Masters, Ryan K.; Aron, Laudan Y. title: Changes in Life Expectancy Between 2019 and 2020 in the US and 21 Peer Countries date: 2022-04-13 journal: JAMA Netw Open DOI: 10.1001/jamanetworkopen.2022.7067 sha: 027668f68bf41f5d8539e0c66921df3b1664d1f0 doc_id: 840621 cord_uid: kl3mfkoh IMPORTANCE: Prior studies reported that US life expectancy decreased considerably in 2020 because of the COVID-19 pandemic, with estimates suggesting that the decreases were much larger among Hispanic and non-Hispanic Black populations than non-Hispanic White populations. Studies based on provisional data suggested that other high-income countries did not experience the large decrease in life expectancy observed in the US; this study sought to confirm these findings according to official death counts and to broaden the pool of comparison countries. OBJECTIVE: To calculate changes in US life expectancy between 2019 and 2020 by sex, race, and ethnicity and to compare those outcomes with changes in other high-income countries. DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study involved a simulation of life tables based on national death and population counts for the US and 21 other high-income countries in 2019 and 2020, by sex, including an analysis of US outcomes by race and ethnicity. Data were analyzed in January 2022. EXPOSURES: Official death counts from the US and 21 peer countries. MAIN OUTCOMES AND MEASURES: Life expectancy at birth and credible range (CR) based on 10% uncertainty. RESULTS: Between 2019 and 2020, US life expectancy decreased by a mean of 1.87 years (CR, 1.70-2.03 years), with much larger decreases occurring in the Hispanic (3.70 years; CR, 3.53-3.87 years) and non-Hispanic Black (3.22 years; CR, 3.03-3.40 years) populations than in the non-Hispanic White population (1.38 years; CR, 1.21-1.54 years). The mean decrease in life expectancy among peer countries was 0.58 years (CR, 0.42-0.73 year) across all 21 countries. No peer country experienced decreases as large as those seen in the US. CONCLUSIONS AND RELEVANCE: Official death counts confirm that US life expectancy decreased between 2019 and 2020 on a scale not seen in 21 peer countries, substantially widening the preexisting gap in life expectancy between the US and peer countries. The decrease in US life expectancy was experienced disproportionately by Hispanic and non-Hispanic Black populations, consistent with a larger history of racial and ethnic health inequities resulting from policies of exclusion and systemic racism. Policies to address the systemic causes of the US health disadvantage relative to peer countries and persistent racial and ethnic inequities are essential. Death counts in 2018, 2019, and 2020 by sex and by age in each peer country were obtained from the January 17, 2022 release of the Human Mortality Database (HMD) -Short-term Mortality Fluctuations original input data in standardized format files. The age-specificity of death counts varied across peer countries. For example, 2020 death counts in England/Wales were provided across seven age groups of varying intervals [0-1), [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] , , , [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] , [75-85), 85+, while 2020 death counts in Norway were provided across 21 five-year age intervals [0-5), [5] [6] [7] [8] [9] [10] , ..., [95-100), 100+. See the HMD metadata for additional details about each peer country's reports of 2020 deaths. Population counts in 2018, 2019, and 2020 by sex and age were obtained from each country's central statistical agency. Death counts for Canada, the Netherlands, New Zealand, and South Korea were available for five age groups that match the five age groups for the year-specific mortality rates published by the HMD-STMF (i.e., [0-15), [15-65), [65-75), , 85+). Therefore, we did not combine original death counts with population estimates for these countries, but instead used the mortality rates provided by the HMD-STMF data for five age groups: To estimate 2019 and 2020 life tables for each peer country's male and female population, we first estimated period age-specific death rates (m x ) for each country's male and female population in years 2018, 2019, and 2020. We merged estimates of age-specific death counts in the HMD-STMF original input data in standardized format files with the age-specific population counts provided by each country's central statistical agency. Next, we estimated the 2019:2018 and 2020:2018 age-specific mortality rate ratios (RR 2019,STMF and RR 2020,STMF ) for each country's total, male, and female populations using the estimated 2018, 2019, and 2020 m x . Third, to estimate and for the 22 age groups used to generate 2019 and 2020 life tables, we multiplied the age-specific RR 2019,STMF and RR 2020,STMF with the age-specific 2018 m x reported in each country's HMD 2018 life where is the width of the age interval. Age-specific death rates for US male and female populations in 2018, 2019, and 2020 were calculated by merging estimates of age-specific counts of death (d x ) 1 where is the width of the age interval. To generate distributions of 2020 life expectancies for each country, we added 10% random uncertainty to and used Python version 3.9.1 to simulate 50,000 life tables. 5 We report the median (P 50 ) estimate for each life expectancy as well as the fifth (P 5 ) and ninety-fifth (P 95 ) percentiles as credible ranges for 2020 life expectancies. This was done for the total, male, and female populations of each country, and for the United States, also by race-ethnicity. We Vintage 2020 Estimates, July 1st resident population by state, county, age, sex, single-race, and Hispanic origin United States life tables Python Language Reference, version 3.9 Provisional life expectancy estimates for 2020. Vital Statistics Rapid Release 15 Mortality in the United States Effects of covid-19 pandemic on life expectancy and premature mortality in 2020: time series analysis in 37 countries Mortality in the United States