Deaths related to heat are being substantially undercounted in the United States, according to an investigation by NPR and Boston University. The analysis estimated about 9,000 annual deaths due to heat in the US between 2018 to 2025. The Centers for Disease Control and Prevention, or CDC, recorded an average of 1,700 heat-related deaths a year over the same period.
The difference reflects how official figures are compiled. The CDC relies on mentions of “heat” in death certificates, yet heat is often absent from those records even when it contributed to a death. Heat can increase the risk of heart attack and worsen serious health conditions, while death certificates rarely cite it as the cause of death.
Tennessee illustrates the gap. The NPR-Boston University model estimated that about 125 people died each year from heat exposure in the state between 2018 and 2025. The CDC officially recorded an average of 25.6 heat-related deaths per year in Tennessee during 2018 to 2025. The investigation concluded that Tennessee likely has five times more heat-related deaths than official tallies show.
Quinn Adams, a research scientist at Boston University School of Public Health’s Center for Climate and Health, was the lead partner in the joint NPR-Boston University investigation. Its analysis found that about a third of heat-related deaths in the US are happening in Florida and Texas alone.
Official heat-related death counts are widely understood to be undercounts. Public health researchers have shown that death-certificate coding misses a substantial proportion of heat deaths, especially where heat acts indirectly but decisively.
Other methods produce higher estimates. Excess-mortality analysis compares the total deaths during a heat event with the number statistically expected under normal conditions. Such analyses more consistently capture a higher number of deaths than cause-of-death coding alone.
Another approach is active case investigation. The method involves systematically reviewing deaths during heat events against weather data, the location of exposure and medical history. It was pioneered by Maricopa County and adopted as a national model for counting heat-related deaths. Undercounting is more common where heat-death investigations are not standardised across medical examiners in a jurisdiction and over time.
Research on air conditioning points to a means of reducing the toll. A paper published on 31 August in Nature Health estimated that household air conditioning use prevents about 5,267 heat-related deaths across the contiguous United States every year.
Researchers at Yale analysed nearly 30 million death records from 2010 through 2020 and matched each death to the daily temperature in the county where it occurred. They found that, if Americans had used almost no air conditioning, the annual toll from heat would have been more than double what the country actually recorded.
In counties with the least air conditioning, the hottest days of the year carried 2.2% higher odds of death than the mildest days. In counties at the national median for cooling, the excess risk was 0.8%, an effect small enough to be statistically insignificant.
Heat mortality has also increased over time. An estimated 210,000 more people died of heat-related causes each year from 2012 to 2021 than from 1990 to 1999, on average. The NPR-Boston University estimates add to evidence that death-certificate counts alone leave public health agencies with an incomplete measure of the risk.
Comments (0)