Deaths per year due to heat

FOI reference: FOI-2026-3720

You asked

It has been reported in various tabloids that over 40,000 excess deaths have occurred between 1988 and 2025 from reports apparently conducted by yourselves due to excess heat.

Under the Freedom of information please can you state:

1. Total amount of deaths per year due apparently to heat (broken down by year ) 

2. The ages of said people in number 1 

3. The cause of death in those in question 1 mentioned 

4. Any comorbidities in those people or not

5. Whether it was due to the heat or with the heat

We said

Thank you for your request. 

To note, the data provided in this response are modelled estimates of mortality attributable to the hottest temperatures. They are not direct counts of deaths certified as caused by heat, and they do not identify individual people. Estimates are presented with 95% confidence intervals to show statistical uncertainty. 

The ONS holds data on all deaths registered in England and Wales, including the causes of death recorded by the certifying doctor or coroner. In line with international standards for death certification, doctors generally record the medical conditions or injuries that led directly to death, rather than the environmental circumstances contributing to it. As a result, while a small number of deaths may be certified as being due to the effects of heat, most deaths where heat is likely to have contributed are recorded under other causes, such as cardiovascular or respiratory conditions. Therefore, the ONS does not hold a direct count of all deaths caused by heat. 

The published climate-related mortality dataset provides the following estimates of heat-attributable deaths by year for England and Wales. Table 5 of these figures relate to year of death occurrence. The spreadsheet provides results by English region, and Wales. The figures found in Tab '1.Year' in the associated download are the sum of the nine English regions, and Wales, with England calculated as the sum of the nine English regions.  

The publication does not provide age-by-year estimates or the ages of identifiable individuals. Table 8 of the accompanying dataset provides modelled estimates for two broad age groups for the full period 1988 to 2025. The spreadsheet provides results by English region, and Wales. The figures found in Tab '2.Age' are the sum of the nine English regions, and Wales, with England calculated as the sum of the nine English regions. 

These age-specific estimates come from separate statistical models. They should not be summed together or expected to equal the all-age estimate from Table 5. A confidence interval that includes zero indicates substantial uncertainty and that the estimate is not statistically distinguishable from zero at the 95% confidence level. 

Table 12 of the accompanying dataset provides cause-specific modelled estimates for selected underlying causes of death for the full period 1988 to 2025. It does not provide a cause of death for each annual estimate or identify individual deaths. Table 12 provides results by English region, and Wales. The figures found in Tab '3.Underlying cause of death' in the associated download are the sum of the nine English regions, and Wales, with England calculated as the sum of the nine English regions. 

These results cover selected underlying-cause groups and should not be summed to derive an all-cause total. Each cause-specific estimate is produced from a separate model, and the categories do not represent every possible underlying cause of death. 

This analysis did not examine comorbidities or link the modelled estimates to individual-level hospital or clinical records. Therefore, we cannot provide information from this study on whether the people represented by the estimates had other health conditions. Answering that question would require a different study design and appropriate linkage to individual-level health data. 

Neither death "due to heat" nor occurred "with the heat" precisely reflects what the statistics measure. These are modelled population-level estimates, not direct counts from death certificates. The analysis models the relationship between daily all-cause mortality and daily mean temperature, then estimates how many deaths were attributable to exposure to the hottest temperatures. 

The figures therefore do not mean that heat was recorded as the underlying cause of death on an individual's death certificate. They estimate deaths that would not have occurred, at that time, in the absence of exposure to the defined hottest temperatures, subject to the assumptions and uncertainty of the statistical model. 

Further information, including definitions, interpretation guidance, methodological notes and the full tables, is available in our Climate-related mortality, England and Wales: 1988 to 2025 bulletin, accompanying dataset, and quality and methods guide. The statistics are official statistics in development and will continue to be reviewed and improved as part of our work on climate and health. 

You may also be interested in UK Health Security Agency publications on this topic.