Climate change is impacting human health. The 2020 report of the Lancet Countdown on health and climate change1Watts N Amann M Arnell N et al.The 2020 report of The Lancet Countdown on health and climate change: responding to converging crises.Lancet. 2021; 397: 129-170Summary Full Text Full Text PDF PubMed Scopus (316) Google Scholar estimates a 53·7% increase in heat-related mortality in people older than 65 years during the past 2 decades. Nowadays, most of the record-breaking temperature extremes are directly attributable to climate change,2Coumou D Robinson A Rahmstorf S Global increase in record-breaking monthly-mean temperatures.Clim Change. 2013; 118: 771-782Crossref Scopus (167) Google Scholar and these events are continuously redefining the range of observed climatological temperatures to which populations are exposed. These previously unobserved temperatures pose an additional threat to human health, as exemplified by the record-breaking heatwave in the summer of 2003, which caused a mortality excess of more than 70 000 premature deaths in Europe.3Robine JM Cheung SLK Le Roy S et al.Death toll exceeded 70,000 in Europe during the summer of 2003.C R Biol. 2008; 331: 171-178Crossref PubMed Scopus (875) Google Scholar In this Comment, we show for the first time the contribution of previously unobserved extreme heat to the trends and seasonality changes of temperature-attributable mortality (TAM) projections in 147 contiguous regions in 16 European countries (Austria [n=9 regions], Belgium [n=11], Croatia [n=2], Czech Republic [n=8], Denmark [n=1], France [n=22], Germany [n=16], Italy [n=21], Luxembourg [n=1], the Netherlands [n=1], Poland [n=16], Portugal [n=5], Slovenia [n=1], Spain [n=16], Switzerland [n=7], and the UK [n=10 regions in England and Wales only]). Several studies4Martínez-Solanas È Quijal-Zamorano M Achebak H et al.Projections of temperature-attributable mortality in Europe: a time series analysis of 147 contiguous regions in 16 countries.Lancet Planet Health. 2021; 5: e446-e454Summary Full Text Full Text PDF PubMed Scopus (3) Google Scholar, 5Gasparrini A Guo Y Sera F et al.Projections of temperature-related excess mortality under climate change scenarios.Lancet Planet Health. 2017; 1: e360-e367Summary Full Text Full Text PDF PubMed Scopus (256) Google Scholar have used epidemiological models to transform climate change simulations into TAM projections, and they generally found a reduction in cold-attributable mortality and an increase in heat-attributable deaths. Following the data and methods of Martínez-Solanas and colleagues,4Martínez-Solanas È Quijal-Zamorano M Achebak H et al.Projections of temperature-attributable mortality in Europe: a time series analysis of 147 contiguous regions in 16 countries.Lancet Planet Health. 2021; 5: e446-e454Summary Full Text Full Text PDF PubMed Scopus (3) Google Scholar we compute the monthly TAM by combining the observed risks of death due to temperatures in the period 1998–2012 with historical and scenario temperatures from climate change model simulations. These mortality projections point to a huge change in the seasonality of TAM. In the historical period (1971–2005), TAM is found to be greatest in winter, with a small secondary maximum in July and August (figure A). This well-established seasonal pattern6Achebak H Devolder D Ingole V Ballester J Reversal of the seasonality of temperature-attributable mortality from respiratory diseases in Spain.Nat Commun. 2020; 112457Crossref PubMed Scopus (8) Google Scholar is projected to change progressively throughout the present century (figure E–G), with a moderate, albeit sustained, decrease of TAM in winter and a much larger increase in summer (figure B–D).7Ballester J Robine JM Herrmann FR Rodó X Long-term projections and acclimatization scenarios of temperature-related mortality in Europe.Nat Commun. 2011; 2: 358Crossref PubMed Scopus (95) Google Scholar The magnitude of these opposing trends is found to directly depend on the greenhouse gas emission scenario, with a complete reversal of the seasonality in Representative Concentration Pathway (RCP) 6.0 and especially in RCP8.5. We note that summer TAM is projected to more than double winter TAM at the end of the century in the highest emission scenario. To better understand the reversal of the seasonality of TAM, we have broken down contributions of monthly TAM by temperature ranges. Because it is customary in the field,5Gasparrini A Guo Y Sera F et al.Projections of temperature-related excess mortality under climate change scenarios.Lancet Planet Health. 2017; 1: e360-e367Summary Full Text Full Text PDF PubMed Scopus (256) Google Scholar, 8Achebak H Devolder D Ballester J Trends in temperature-related age-specific and sex-specific mortality from cardiovascular diseases in Spain: a national time-series analysis.Lancet Planet Health. 2019; 3: e297-e306Summary Full Text Full Text PDF PubMed Scopus (39) Google Scholar cold days are defined as the days with temperatures lower than the temperature of minimum mortality and heat days heat days are defined as the days with temperatures higher than the temperature of minimum mortality. These subsets of days are further broken down into extreme cold days (cold days colder than the 2·5th percentile in the observational period), moderate cold days (cold days warmer than the 2·5th percentile), moderate heat days (heat days colder than the 97·5th percentile), and extreme heat days (heat days warmer than the 97·5th percentile). Unlike in Martínez-Solanas et al,4Martínez-Solanas È Quijal-Zamorano M Achebak H et al.Projections of temperature-attributable mortality in Europe: a time series analysis of 147 contiguous regions in 16 countries.Lancet Planet Health. 2021; 5: e446-e454Summary Full Text Full Text PDF PubMed Scopus (3) Google Scholar we here use the observational temperature records of the last climatological 30-year period (1991–2020)9Cornes RC van der Schrier G van den Besselaar EJM Jones PD An ensemble version of the E-OBS temperature and precipitation data sets.J Geophys Res Atmos. 2018; 123: 9391-9409Crossref Scopus (364) Google Scholar, 10Copernicus Climate Change ServiceClimate Indicators, Temperature.https://climate.copernicus.eu/climate-indicators/temperatureDate accessed: August 9, 2021Google Scholar to compute these temperature percentile thresholds. This choice allows us to further break down the temperature range of extreme heat by defining the set of previously unobserved extreme heat days as those with temperatures warmer than the maximum observational value—ie, 100th percentile of the 1991–2020 period. The break down of TAM into these five temperature categories highlights the very large contribution of extreme heat days to the increase in summer TAM. January TAM is projected to progressively decrease from 13·36% in the historical period (1971–2005) to 11·08% in RCP2.6, 9·69% in RCP6.0, and 8·59% in RCP8.5 at the end of the century (2070– 99). However, July TAM is projected to increase from 3·89% in the historical period to 7·80% in RCP2.6, 12·81% in RCP6·0, and 21·59% in RCP8.5, but in this case, differences between end-of-the-century scenarios are mostly projected to arise from the set of previously unobserved extreme heat days. The contribution of these unobserved days to July TAM is projected to be negligible in RCP2.6 (0·61%), small in RCP6.0 (2·54%), and very large in RCP8.5 (9·21%). In the absence of mitigation, these projections entirely driven by changes in temperature emphasise the vital role of adaptation measures specifically targeted towards the negative impacts of previously unobserved extreme heat. Nonetheless, it is not clear whether these currently unobserved temperatures actually exceed the limits of human body physiology, and whether societal mechanisms and behavioural changes can reduce the health burden beyond these eventual body limits. Moreover, it is still unknown whether all societies, and all population groups within them, will have the capacity to implement these adaptation processes to the same degree. All these uncertainties ultimately highlight the importance of mitigation actions as the safer and more direct path to prevent eventual tipping points in human health vulnerability, such as those shown here for unobserved extreme heat days in the business-as-usual scenario. MQ-Z, HA, DP, and JB acknowledge funding from the EU's Horizon 2020 research and innovation programme under grant agreement No 865564 (European Research Council Consolidator Grant EARLY-ADAPT). MQ-Z, EM-S, DP, and JB acknowledge funding from the EU's Horizon 2020 research and innovation programme under grant agreement No 727852 (project Blue-Action). HA acknowledges funding from the Secretariat for Universities and Research of the Ministry of Business and Knowledge of the Government of Catalonia (grant numbers B00391 [FI-2018], B100180 [FI-2019] and B200139 [FI-2020]). JB acknowledges funding from the EU's Horizon 2020 research and innovation programme under grant agreement No 956396 (project EDIPI), and from the Ministry of Science and Innovation under grant agreements RYC2018-025446-I (programme Ramón y Cajal) and EUR2019-103822 (project EURO-ADAPT). J-MR acknowledges funding from the EU Community Action Program for Public Health (grant agreement 2005114). ISGlobal recieves support from the Spanish Ministry of Science and Innovation through the Centro de Excelencia Severo Ochoa 2019-2023 Program (CEX2018-000806-S), and support from the Generalitat de Catalunya through the CERCA Program. All other authors declare no competing interests. We acknowledge the E-OBS dataset from the EU-FP6 project UERRA and the Copernicus Climate Change Service, and the data providers in the ECA&D project. For their roles in producing, coordinating, and making available the ISIMIP input data and impact model output we acknowledge the modelling groups, the ISIMIP sector coordinators, and the ISIMIP cross-sectoral science team for the health sector.