Heat kills: several hundred deaths were recorded in Paris and other major European capitals during the heatwave from mid-June to early July 2025. The number of deaths is thought to have tripled compared to normal levels due to climate change, according to estimates by British researchers.
These figures are frightening. However, they mask the great progress that has been made in limiting our vulnerability to the increasing frequency of heat waves. Heat kills, but less and less so, thanks to our individual and collective adaptation efforts. We should celebrate this, not ignore it.
As mortality observation data from health agencies is not yet available, the above calculation is based on models and methods known to specialists. Most of these are now sufficiently developed to provide reliable results on the progress of adaptation.
Let’s begin our examination by taking the 2003 heatwave in France as a reference point. That summer, the country experienced a veritable bloodbath: nearly 15,000 excess deaths.
In response, the public authorities decided on a series of preventive measures to protect the population from extreme heat: setting up a heatwave warning system, public information campaigns on ways to protect oneself, training for healthcare staff, opening air-conditioned spaces in retirement homes and hospital wards, etc.
A few years later, a team of researchers from Inserm, Météo France, and Santé publique France wondered whether these measures had been effective. Using a model linking mortality and temperatures observed in France over a 25-year period, they estimated that the heatwave of summer 2006 had resulted in a reduction of more than half in the number of excess deaths.
This progress cannot, of course, be attributed solely to public action. The 2003 heatwave raised widespread awareness of the harmful effects of heat, leading to changes in individual behavior and the purchase of cooling devices such as fans and air conditioners, as well as more innovative equipment that appeared on the market later, such as air coolers and reversible heat pumps.
Please note that refrigerators with ice dispensers are not included in this range! In high temperatures, you should avoid drinking ice-cold water to cool down. It slows down sweating, which is the body’s fundamental mechanism for combating heat.
The summer of 2022 was the second hottest season ever recorded in France. However, the number of excess deaths was five times lower than in 2003, but it is not clear exactly how much of this decrease is simply due to slightly less severe heatwave conditions.
The 2003/2022 comparison is just as difficult at the European level. Estimates of excess mortality linked to heat on both dates are available at this level, but they are based on different methods that make their results difficult to compare: 74,483 deaths for the 2003 heatwave in Europe compared to 61,672 deaths during the 2022 heatwave.
In fact, the first figure measures excess deaths compared to a reference period, while the second is derived from the application of an epidemiological method. This method, which is more sophisticated but also more rigorous, consists of estimating the relative mortality risk for a city, region, or country based on temperature, and plotting an “exposure-response curve,” as specialists call it.
For Europe as a whole, the risk is lowest around 17°C to 19°C, then rises sharply above that. Knowing the daily temperatures reached on heatwave days, we can then deduce the number of deaths associated with the heat.
Summarized in this way, the work seems easy. However, it requires a myriad of data and is based on numerous calculations and assumptions.
This method is used to estimate the excess mortality linked to the temperatures that we will experience by the middle or end of this century, based, of course, on different global warming scenarios. For example, it is expected to increase tenfold in Europe by 2100 in the event of a 4°C rise in temperature.
This figure is frightening, but it does not take into account the future adaptation of humans and societies to global warming. One way to measure this, in terms of the past, is to look at how the exposure-response curve to temperature shifts over time. If adaptation occurs, we should see mortality rising less sharply with heat than before.
This is what was observed in Paris when comparing the relative risk of heat-related mortality between the period 1996-2002 and the period 2004-2010. At the highest temperatures, those in the upper quarter of the distribution, the risk decreased by 15%.
This figure may not seem very impressive, but it should be noted that it only takes into account mortality on the day when extreme temperatures are measured. However, heat-related deaths can occur with a delayed effect lasting several days or even weeks.
Taking this effect into account further reduces the risk between the two periods: from 15% to 50%. This 50% reduction is greater than that observed in other European capitals such as Athens and Rome. In other words, Paris is not lagging behind in adapting to heat waves.
In general, regardless of the method used, the trend toward a decrease in the population’s susceptibility to heat is evident in many other cities and countries in the developed world. Adaptation and the resulting decline in mortality are the norm.
This is good news, but this decline in excess mortality remains relative. If progress in adaptation is slower than warming, it is still possible that the absolute number of deaths will increase. In other words, is mortality declining faster or slower than warming is increasing?
Faster, if we look at the trends observed in ten European countries between 1985 and 2012. As the authors write:
The reduction in heat-attributable mortality has occurred despite the gradual shift toward warmer temperatures observed in recent decades.
Will the same be true tomorrow? We mentioned above a tenfold increase in heat-related excess mortality in Europe by 2100. This comes from an article published in Nature Medicine, which estimated that it would rise from 9 to 84 heat-related deaths per 100,000 inhabitants.
But beware: this figure assumes no adaptation whatsoever. To take this into account in their results, the authors of the article postulate that progress between now and 2100 will lead to a maximum 50% reduction in mortality.
However, given the progress made in the past, as examined above, over a shorter period, a greater reduction does not seem out of reach.
This is especially true if the use of air conditioning continues to grow. The rate of air conditioning installation in Europe is currently only 19%, compared to over 90% in the United States. Its widespread use in the US over the past half-century has led to a sharp decline in heat-related mortality.
Behind the question of whether future progress in adaptation will reduce heat-related mortality by more than 50% in Europe by 2100, there is another question at stake: will the excess mortality associated with global warming lead to an overall positive or negative outcome?
Indeed, while climate change leads to hotter summers, it also leads to milder winters—and therefore fewer deaths. Cold-related mortality was estimated in 2020 at 82 deaths per 100,000 inhabitants. With a temperature rise of 4°C, it should, according to the authors of the article in Nature Medicine, reach 39 deaths per 100,000 people by the end of the century.
If we compare this figure to the 84 deaths per 100,000 inhabitants linked to heat, cited above, it is easy to calculate that a 55% improvement in heat adaptation would be enough to equalize cold- and heat-related mortality. Global warming would then become neutral for Europe, if we consider only human mortality linked to extreme temperatures.
But everyone knows that global warming also causes fires, floods, and deadly storms, as well as the destruction of ecosystems.
From this very simplistic perspective, global warming would even be beneficial, as long as the progress made in adapting to it exceeded this 55% threshold. If we consider France alone, this threshold is only slightly higher, at 56%.
The lower winter mortality rate will no doubt surprise readers, who are more accustomed to being informed and alarmed during the summer months about the negative health effects of global warming. The unsettling idea that rising temperatures in Europe could ultimately be beneficial is also disturbing. Could it not reduce Europeans’ motivation and incentives to reduce their greenhouse gas emissions?
It is perhaps this fear that leads the authors of the Nature Medicine article to conclude that:
Net mortality will increase substantially if we consider the most extreme warming scenarios, and this trend can only be reversed by considering implausible levels of adaptation.
It should also be noted that these prospects concern Europe. In low-latitude countries, temperature-related excess mortality is appalling. Their populations are much more exposed to climate change than ours; they are also more vulnerable, with adaptation capacities limited by poverty.
By 2100, net temperature-related mortality is estimated at more than 200 deaths per 100,000 inhabitants in sub-Saharan Africa and nearly 600 deaths per 100,000 inhabitants in Pakistan.
Let us conclude that we must not relax our efforts to adapt to heat in Europe, even if they result in a net benefit. Individual and collective actions to adapt to heat save lives. Let us continue them. And let us not relax our efforts to reduce emissions, which will save lives elsewhere, particularly in poor countries in low latitudes.
Faced with the acceleration of climate disruption and its already visible impacts – deadly heat waves, floods, giant fires, the weakening of agricultu...