Projections of Extreme Temperature–Related Deaths in the US

Abstract

ImportanceExtreme heat in the US is increasing due to climate change, while extreme cold is projected to decline. Understanding how extreme temperature along with demographic changes will affect population health is important for devising policies to mitigate the health outcome of climate change.ObjectiveTo assess the burden of extreme temperature–related deaths in the contiguous US currently (2008-2019) and estimate the burden in the mid–21st century (2036-2065).Design, Setting, and ParticipantsThis cross-sectional study used historical (1979-2000) daily mean temperatures to calculate monthly extreme heat (>97.5th percentile value) and extreme cold days (<2.5th percentile value) for all contiguous US counties for 2008 to 2019 (current period). Temperature projections from 20 climate models and county population projections were used to estimate extreme temperature–related deaths for 2036 to 2065 (mid–21st century period). Data were analyzed from November 2023 to July 2024.ExposureCurrent monthly frequency of extreme heat days and projected mid–21st century frequency using 2 greenhouse gas emissions scenarios: Shared Socioeconomic Pathway (SSP)2-4.5, representing socioeconomic development with a lower emissions increase, and SSP5-8.5, representing higher emissions increase.Main Outcomes and MeasuresMean annual estimated number of extreme temperature–related excess deaths. Poisson regression model with county, month, and year fixed effects was used to estimate the association between extreme temperature and monthly all-cause mortality for older adults (aged ≥65 years) and younger adults (aged 18-64 years).ResultsAcross the contiguous US, extreme temperature days were associated with 8248.6 (95% CI, 4242.6-12 254.6) deaths annually in the current period and with 19 348.7 (95% CI, 11 388.7-27 308.6) projected deaths in the SSP2-4.5 scenario and 26 574.0 (95% CI, 15 408.0-37 740.1) in the SSP5-8.5 scenario. The mortality data included 30 924 133 decedents, of whom 15 573 699 were males (50.4%), with 6.3% of Hispanic ethnicity, 11.5% of non-Hispanic Black race, and 79.3% of non-Hispanic White race. Non-Hispanic Black adults (278.2%; 95% CI, 158.9%-397.5%) and Hispanic adults (537.5%; 95% CI, 261.6%-813.4%) were projected to have greater increases in extreme temperature–related deaths from the current period to the mid–21st century period compared with non-Hispanic White adults (70.8%; 95% CI, −5.8% to 147.3%).Conclusions and RelevanceThis cross-sectional study found that extreme temperature–related deaths in the contiguous US were projected to increase substantially by mid–21st century, with certain populations, such as non-Hispanic Black and Hispanic adults, projected to disproportionately experience this increase. The results point to the need to mitigate the adverse outcome of extreme temperatures for population health.