Summer heat may have caused more than 50,000 deaths in European cities, study estimates

Heat during the summer of 2026 may have contributed to about 50,391 deaths in 854 European cities, according to a new estimate by researchers at Imperial College London and the London School of Hygiene & Tropical Medicine. The team attributes 32,619 of those deaths, about two-thirds, to warming caused by human activity. The figures come from epidemiological models covering June through August. They are not a count of death certificates identifying individual victims.

The analysis covers urban areas in 30 countries: the 27 members of the European Union, the United Kingdom, Norway and Switzerland. The cities studied account for about 40% of those countries' combined population. The estimate therefore should not be described as the total number of heat-related deaths across Europe. Nor can it simply be extended to rural areas, where temperatures, populations and access to services differ.

Imperial College said the analysis of summer 2026 had been accepted for publication in the journal Nature Health. Its authors compared it with historical research on the same cities from 2000 to 2025, published this month. On that consistent basis, 2026 had the highest estimated heat mortality in the series, above approximately 29,480 deaths in 2022 and 28,462 in 2003.

Estimating deaths when heat is rarely recorded as the cause

Heat can worsen cardiovascular, respiratory and kidney disease without appearing on a death certificate. To assess its effect, the scientists combined observed temperatures in each city with historical relationships between heat and mortality, adjusted for local characteristics and age. They then compared the real-world conditions with a hypothetical world without the warming caused by human emissions.

That comparison yields a fraction attributable to climate change; it cannot identify which particular person would have survived in a different climate. The model estimates how many additional deaths would be expected statistically at given temperatures. It measures risk across populations and does not replace a clinical investigation of each death.

Complete observed mortality data for this summer were not yet available for all 854 cities. David García García, a researcher with Spain's MoMo mortality surveillance system, said the common method was useful for comparing regions. But in comments to Science Media Centre Spain, he cautioned that the estimates do not incorporate every death actually recorded throughout the period. He also noted that the choice of temperature threshold can affect the calculation. Official national figures and the study's estimates are not interchangeable.

Large estimated tolls in Italy, France and Spain

Among the cities studied, Italy had an estimated 13,303 heat-related deaths, followed by France with 7,268 and Spain with 6,824. The estimates were 6,340 for Germany and 4,354 for the United Kingdom. Those are figures for the urban areas in the study, not complete national death tolls. Differences reflect the intensity of the heat as well as the size, age structure and vulnerability of the populations analyzed.

In Spain, a mortality working group of the Spanish Society of Epidemiology noted that the sample comprised 44 cities and about 22 million residents. The model attributed 77% of estimated heat deaths there to human-caused warming. Spain's official MoMo system estimated 5,807 deaths attributable to high temperatures over a longer period, from May 15 to September 30. The apparent disparity cannot be resolved by selecting one number and discarding the other: the populations, dates and methods need to be explained.

The most intense spell ran from June 17 to July 1, the researchers said, accounting for about a third of the summer's estimated deaths. Early-season heat may have found some communities less acclimatized, while humidity and unusually warm nights limited relief between hot days. The study also estimated several thousand deaths among people younger than 65; the risk was not confined to older adults.

Looking back, and preparing cities for future heat

The historical study published October 14 estimated more than 430,000 heat-related deaths in the same 854 cities over the summers of 2000 through 2025. The share attributed to human-caused warming rose from 39% at the start of that period to 68% in 2025. That work provides a benchmark for the new 2026 estimate. Each figure belongs to its own year and should not be added to the new result as though it were an additional death toll announced today.

The authors and other specialists point to responses on two timescales. Cities can improve warnings, provide more shade, adapt housing and help people exposed during hot nights. Over the longer term, reducing emissions limits the warming that raises the risk. None of those measures eliminates uncertainty in a model, but they address a hazard documented by a wider body of evidence.

The next test will come when national systems publish complete mortality records for the summer. Those data will allow the projections to be compared with observed deaths, although death certificates will not always record heat's contribution. For now, the estimate indicates the potential scale of the harm while requiring a clear account of what was measured and what remains unknown.