Official statistics estimate that cold weather was associated with 2,200 deaths in England during the winter of 2025 to 2026. The estimate covers five regionally defined cold-weather episodes and has a 95% confidence interval of 1,627 to 2,759 deaths.
The figure is a statistical estimate of deaths associated with cold, not a count of death certificates listing cold as the cause. UKHSA analyses daily death registrations alongside temperature data and models how mortality changes as temperatures fall, including effects that can occur in the days after exposure.
One January episode accounted for most of the estimate
The largest of the five episodes lasted nine days in early January 2026. It was associated with an estimated 1,563 deaths, with a 95% confidence interval of 1,180 to 1,923. The other four episodes lasted two or three days each and had substantially lower estimates.
Across the winter, the report estimates 1,565 deaths were directly associated with cold, independently of influenza circulation. A further 635 were linked to cold weather contributing to increases in influenza infection. These components together make up the total estimate of 2,200.
The figures describe modelled excess mortality associated with cold episodes. They should not be read as identifying particular people or establishing that cold was the sole cause of an individual death. The analysis accounts for delayed effects because health impacts do not necessarily occur on the days when temperatures fall.
Regional patterns differed
Statistically significant cold-associated mortality was identified in every English region except the North East. The North West had the highest estimated rate, at 52 deaths per million people, with a 95% confidence interval of 18 to 85. It also had the highest estimated number of deaths over the winter, at 393.
The absence of a statistically significant estimate for the North East does not mean that cold weather presents no health risk there. UKHSA notes that the region experienced colder temperatures than many other areas, while its model found no statistically significant cold-associated mortality during the monitoring period. The agency says regional differences may reflect factors including housing, heating, population health and adaptation, but further research is needed to understand them.
The report found the strongest relationship between lower temperatures and mortality among people aged 85 and over. Its findings also describe cold-associated deaths in care homes, hospitals and people’s own homes. The statistics are intended to inform public-health monitoring and winter preparedness, rather than provide individual-level conclusions.
New method changes the year-on-year picture
This edition is the first to be published as official statistics rather than official statistics in development. It also introduces a revised method: cold episodes are now identified separately for English regions instead of using an England-wide temperature average. An episode is defined as at least two consecutive days when a region’s average daily mean temperature is below 2°C.
That change affects comparisons with the previous winter. Under the new regional method, UKHSA estimates 3,351 cold-associated deaths in 2024 to 2025; the earlier method produced an estimate of 2,788 for that winter. The new 2025 to 2026 estimate of 2,200 should therefore not be presented as a like-for-like year-on-year fall against the earlier published figure. The difference in the 2024 to 2025 figures reflects methodology, not evidence of a change in deaths.
For 2025 to 2026, regional episode definitions yielded five episodes lasting 18 days in total. The approach is intended to reflect differences in where cold conditions occur and which populations are exposed. UKHSA’s estimates provide one part of the evidence used in winter health surveillance, alongside monitoring of seasonal infections and other pressures.
