
Blog · Lifestyle & Wellness
Why Most Cold Weather Deaths Happen on Mild Days
By Vitae Team ·
Most cold-related illness and death occurs in mild winter weather, not severe cold. The headline mortality figures measure something much narrower.
Winter cold is associated with substantially more deaths in England than extreme heat, but the coldest days do not account for the whole burden. NICE says most cold-related illness and death occurs during relatively mild winter weather, not severe cold snaps. The latest ONS figures measure something narrower: deaths associated with the coldest 2.5% of days.
TL;DR
- ONS estimates 138,944 deaths in England between 1988 and 2025 associated with the coldest days, against 40,063 associated with the hottest. The two are converging.
- Those are modelled estimates, not counts, and they cover only the coldest and hottest 2.5% of days.
- Most cold-related harm does not happen on the coldest days. NICE says it occurs at ordinary winter temperatures, below roughly 5°C to 8°C outdoors.
- The effect lingers. Death rates stay raised for up to two weeks after a cold spell, where heat effects fade within a day.
- Housing changes the size of the risk. Death rates rise about 2.8% per degree of outdoor cooling in the coldest tenth of homes, against 0.9% in the warmest.
- 18°C is a recommended target, not a proven biological threshold. WHO strongly supports maintaining adequate warmth while acknowledging uncertainty about the exact temperature at which risk increases.
- Warming homes improves symptoms, though the best trial missed its primary outcome.
01 Most of the harm lands on ordinary days
NICE guideline NG6, which deals with excess winter deaths and cold homes, is direct about this. Its position is that most excess winter deaths and illnesses are not caused by hypothermia or extremes of cold at all. They are respiratory and cardiovascular problems occurring at ordinary winter temperatures, once the mean outdoor temperature falls below around 5°C to 8°C.
Risk keeps climbing as it gets colder, and the per-day risk on a freezing day is higher. There are simply far fewer of those days, and a season is made of the ordinary ones, so most of the harm lands on the mild ones. A cold winter does its damage cumulatively, through months of unremarkable weather, rather than through the week the forecast turns severe.
It does not stop when the cold does, either. NICE notes that raised death rates persist for up to two weeks after a cold spell ends, where the increase associated with hot weather lasts a day or so beyond the heat. Our piece on what heatwaves do to the body covers the other half of that pattern. The strongest single link is between cold and respiratory deaths, though because far more people die of cardiovascular disease overall, cardiovascular events account for the larger number of cases. One caveat applies throughout: NG6 was published in March 2015 and its figures are older still, so it is the framing rather than the numbers that has carried forward.
02 What the mortality figures count
The Office for National Statistics published its latest climate-related mortality estimates on 26 August 2026, covering 1988 to 2025. For England, ONS estimates 138,944 deaths associated with the coldest days against 40,063 associated with the hottest, and for Wales 10,098 against 2,070. The cold figure peaked in 2010, at 17,584 in England, and the heat figure in 2022, at 4,430. ONS notes that the two are starting to converge.
Two features of that series matter as much as the totals. These are modelled estimates rather than counted deaths, produced by fitting daily temperature against mortality while adjusting for air pollution, wind, pressure and humidity, and ONS classes them as statistics in development. The model defines cold and heat as the coldest and hottest 2.5% of days against regional thresholds, which is why a single severe year can account for a large share of the total. These are not estimates of the total mortality burden from cold weather. By restricting its calculation to the coldest 2.5% of days, ONS excludes the wider range of temperatures that NICE identifies as responsible for much of the winter health burden.
ONS is candid about the rest. The model does not control for seasonal diseases such as influenza, and it cannot tell whether anyone was outdoors at the temperatures being measured. The estimates are not age-standardised, the 2025 figures are incomplete, and ONS notes its estimates supersede earlier publications. None of that makes the series wrong, and none of it puts ONS at odds with NICE. The two are measuring different things, one the sharp end of the season and the other a long, shallow effect spread across all of it. A reader who sees only the first will think the risk arrives with the cold snap and leaves with it.
03 Where the cold is
If the risk sits in ordinary winter weather rather than extremes, the question becomes where people are when they meet it, and for most of a British winter that is indoors. NICE cites research finding that death rates rise by about 2.8% for every degree the outdoor temperature falls among people in the coldest tenth of homes, against 0.9% in the warmest tenth. The Marmot review team put excess winter deaths in the coldest quarter of housing at almost three times those in the warmest.
The exposure is unevenly distributed. In 2025 around 2.36 million English households, 9.4%, were in fuel poverty on the official measure, a figure the government still describes as provisional, and 39.4% of homes sat in the lower energy efficiency bands. Age concentrates it further. In the year NICE drew on, 51% of cold-related deaths were among people aged 85 and over and 27% among those aged 75 to 84, the groups most likely to be home all day with the heating off.
A 2024 European analysis offers another possible explanation for regional differences. Its authors proposed that climate-related patterns of respiratory infection and acquired immunity might help explain why some places with mild winters show greater cold-associated mortality than colder regions, with London patterning more like Madrid than Stockholm. The analysis used population-level data and could not establish those mechanisms directly. It does not rule out housing quality or fuel affordability as contributors, and its UK series ran only five years.
04 What warming a home has been shown to change
The familiar advice is to keep the home at 18°C, which comes from the World Health Organization's housing guidelines. WHO grades that recommendation strong on moderate-quality evidence. It is also unusually clear about where the number came from, stating that 18°C was proposed by an earlier working group considering healthy sedentary people, and that evidence cannot establish the temperature below which harm begins. The supporting cardiovascular evidence is blood pressure rather than heart attacks, with a Scottish cohort finding raised risk below 18°C and a steeper rise below 16°C.
WHO also flags the obvious confounding. Cold homes tend to come with damp, mould, poor repair and low income, and separating the temperature from everything travelling with it is difficult. Its insulation evidence was mixed across eleven studies, and it warns that insulating a home can cut ventilation enough to create damp if airflow is not maintained. The 18°C figure is also a general minimum rather than a guarantee of safety. WHO suggests a higher minimum for older people, young children and anyone with a heart or lung condition.
One useful randomised test of whether warming a home changes health was published in the BMJ in 2008. Researchers randomised 409 New Zealand households with a child who had asthma, to receive improved heating either immediately or once the study was over. After exclusions and withdrawals 349 remained, and the number analysed varies by outcome.
The primary outcome, lung function, did not reach significance, at 130.7 ml with an interval running from minus 20.3 to 281.7. The secondary outcomes did. Sleep disturbed by wheeze fell, odds ratio 0.55 with an interval of 0.35 to 0.85, as did night cough and parent-rated poor health. Children had 1.8 fewer days off school and 0.40 fewer asthma-related doctor visits, an interval of 0.11 to 0.62. The living rooms warmed by 1.1°C.
That is the honest shape of the evidence. A modest temperature gain produced consistent improvements in how ill the children were and how much care they needed, in a trial that could not blind anyone and did not demonstrate the physiological change it set out to find. The authors suggested the lung function result may have been underpowered and the meters sensitive to technique. Either way, the useful reading is that warmth changed how the children lived rather than how their lungs tested.
What This Actually Means
Whether it is the weather or the house has no clean answer, and the more useful finding is that the two are hard to separate. Outdoor cold is what the mortality models measure because it is what gets recorded. Indoor cold is where most people spend the winter, and NICE treats cold homes as the part that can be acted on, while respiratory infection, underlying health and circumstances all sit in the same picture.
What a reader can take from it is a change of emphasis. Cold weather risk is not mainly about the week the forecast turns severe, and it is not about hypothermia. It is about months of being somewhat too cold, in rooms that cost too much to heat, with effects that build and then persist for a fortnight after a cold spell. If you have a cardiovascular or respiratory condition, that steady background matters more than any single freezing night, much as blood pressure in hot weather is a seasonal problem rather than a single-day one.
It is also worth holding the 18°C figure loosely. It is a reasonable target with a strong recommendation behind it, and not a cliff edge discovered in the data. Nobody has established the temperature at which harm starts, and the evidence that warming a home helps is stronger for symptoms and sleep than for the outcomes everyone would most like to see moved.
Frequently Asked Questions
Is cold weather responsible for more deaths than heat?
On ONS estimates, deaths associated with the coldest days have outnumbered those associated with the hottest by a wide margin. That gap is narrowing, and both figures are modelled estimates covering only the extremes of the temperature range.
Is hypothermia the main danger?
No. NICE is explicit that most excess winter deaths are caused not by hypothermia or extreme cold but by cardiovascular and respiratory problems at ordinary winter temperatures.
What temperature should I keep my home at?
WHO recommends 18°C for general populations in cold seasons, with a higher minimum for older people, young children and those with heart or lung conditions. It notes the figure is a guide rather than a measured threshold.
Why do deaths keep rising after a cold spell ends?
The effect is delayed rather than immediate. Raised rates have been observed for up to two weeks afterwards, which is one reason the risk is easy to underestimate at the time.
Does insulating a home improve health?
The evidence is mixed by type of insulation, and WHO rates that recommendation conditional. It warns that reduced ventilation can cause damp, so airflow matters alongside warmth.
The Bottom Line
ONS estimates 138,944 deaths in England between 1988 and 2025 associated with the coldest days, against 40,063 associated with the hottest, while restricting that calculation to the most extreme 2.5% of days and not accounting for influenza. NICE, looking at the same season from a different angle, finds that most cold-related illness and death occurs at ordinary winter temperatures rather than extremes, and lingers for a fortnight after a cold spell.
Together, the findings suggest that preventing winter illness cannot be limited to warnings about severe cold. Housing conditions are one important part of the picture, alongside respiratory infections, underlying health and the cumulative effects of lower temperatures. Death rates rise roughly three times faster per degree in the coldest homes than the warmest, and the most exposed are the oldest and the least able to pay to heat a room. The evidence that warming a cold home helps is real but modest, and 18°C is a recommended target rather than a line discovered in the data.
This article is for general information and is not medical advice. If you have symptoms or concerns, speak to your GP.
Sources
- Office for National Statistics. Climate-related mortality, England and Wales: 1988 to 2025. Published 26 August 2026. https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/bulletins/climaterelatedmortalityenglandandwales/1988to2025
- National Institute for Health and Care Excellence. Excess winter deaths and illness and the health risks associated with cold homes (NG6), context. Published 5 March 2015. https://www.nice.org.uk/guidance/ng6/chapter/context
- World Health Organization. Low indoor temperatures and insulation, WHO Housing and Health Guidelines. https://www.ncbi.nlm.nih.gov/sites/books/NBK535294/
- Howden-Chapman P, Pierse N, Nicholls S, et al. Effects of improved home heating on asthma in community dwelling children: randomised controlled trial. BMJ 2008;337:a1411. https://pmc.ncbi.nlm.nih.gov/articles/PMC2658826/
- Exploring the paradoxical nature of cold temperature mortality in Europe. https://pmc.ncbi.nlm.nih.gov/articles/PMC10850168/
- Department for Energy Security and Net Zero. Annual Fuel Poverty Statistics in England, 2026 (2025 data). Published 26 March 2026. https://assets.publishing.service.gov.uk/media/69c3af123ed0546101e0dc3e/Main_Report__2026_Fuel_Poverty_Statistics_Publication_.pdf