A global tally puts 2023's heat deaths near 180,000, over half tied to warming

A study analysing mortality across 2,013 locations in 67 countries and territories estimates that the 2023 global heatwave was associated with 178,486 excess deaths, and that 54.29% of them would not have happened without human-induced climate change [s1]. In Europe alone, a separate model put the 2023 heat toll at 47,690 deaths — a burden that would have been 80% higher without decades of societal adaptation [s2]. Between them, the two papers do something official statistics rarely manage in real time: they attach a firm number, and a cause, to a year of heat.

The global count

Published in The Innovation in 2025, the global analysis used a three-stage model — fitting historical heatwave-mortality associations, pooling them across locations with meta-regression, then projecting grid-specific mortality for 2023 [s1]. It estimated 178,486 excess deaths (95% empirical confidence interval 159,892–204,147) linked to the 2023 heatwave, equal to 0.73% of all deaths globally that year, or 23 deaths per million people [s1].

The burden was wildly uneven. The highest death rates fell in Southern Europe, at 120 per million (95% eCI 116–126), Eastern Europe at 107 per million (100–114), and Western Europe at 66 per million (62–70) — several times the global average of 23 [s1]. That concentration in Europe, rather than in the hottest places on Earth, is a recurring lesson of heat epidemiology: mortality tracks not raw temperature but the gap between the temperature and what a population is acclimatised and built for.

The attribution step

What separates this from a body count is the counterfactual. The authors ran the same mortality model under a "factual" 2023 climate and a "counterfactual" world without human-induced warming, and compared the two [s1]. The difference — 54.29% of the heatwave-related deaths, with a confidence interval of 45.71% to 61.36% — is the share attributable to climate change [s1]. In plain terms, more than half of the 2023 heat deaths are estimated to be the direct product of warming already in the system, not of weather that would have happened anyway.

Europe, and the adaptation the numbers hide

The European figure comes from a different team using a different method, and it adds the variable the global paper does not isolate: what societies have already done to protect themselves. Applying epidemiological models to temperature and mortality records in 823 regions across 35 countries, the Nature Medicine study estimated 47,690 heat-related deaths in Europe in 2023 (95% CI 28,853–66,525) — the second-highest annual burden of its 2015–2023 study period, behind only 2022 [s2]. 2023 was the warmest year on record globally and the second warmest in Europe [s2].

The striking result is what adaptation prevented. The authors estimated the 2023 European toll would have been 80.0% higher without the protective adaptation accumulated since 2000 — and 100.7% higher among people aged 80 and over, the group most exposed [s2]. Air conditioning, heat-health warning systems, cooling centres, changed behaviour and better clinical response are not usually visible in a death count, precisely because they show up as deaths that did not occur. This paper makes that invisible saving legible, and it is roughly the difference between the toll Europe recorded and one nearly twice as large.

What the two numbers can and cannot bear

Both are model estimates with wide intervals, and the two are not directly additive — different methods, different reference periods, overlapping geography [s1][s2]. Excess-death and attribution modelling infers deaths statistically rather than counting them on certificates, which is why the confidence intervals are broad and why the point estimates should be read as central guesses, not exact tallies [s1]. Heat deaths are also systematically under-recorded on death certificates, attributed to the cardiac or respiratory event a heatwave triggers rather than to heat itself — one reason modelling, for all its uncertainty, tends to reveal a larger toll than direct counts.

What to watch

Whether the adaptation dividend the European study measured keeps pace as heat intensifies is the open question; the same paper frames it as evidence that current strategies are working but insufficient for hotter summers ahead [s2]. And whether attribution moves from retrospective research into something faster — numbers produced within weeks of an event rather than a year or two later — determines whether findings like these can inform response while a heatwave is still killing people, rather than only memorialising it afterward.

Sources

1. [s1] Global excess deaths associated with heatwaves in 2023 and the contribution of human-induced climate change. The Innovation (Cell Press), 2025. https://doi.org/10.1016/j.xinn.2025.101110
2. [s2] Heat-related mortality in Europe during 2023 and the role of adaptation in protecting health. Nature Medicine, August 12, 2024. https://doi.org/10.1038/s41591-024-03186-1