Published in Nature Health, the study examined 2,618,307 hospitalization records collected between 2000 and 2019 across 852 locations. The researchers focused on the warm seasons in each country and investigated whether hospital admissions for mental and behavioural disorders were more common during periods of unusually high temperatures. Rather than comparing different groups of people, the analysis compared individuals with themselves across different days, allowing the researchers to examine whether hospitalization risk changed when a person’s location experienced a heatwave.

The primary definition of a heatwave was deliberately strict. It required daily mean temperatures to remain above the location-specific 97.5th percentile for at least four consecutive days. In practical terms, this threshold identified temperatures that were exceptionally high for a particular place, taking local climate into account. A temperature that is extreme in one region may be relatively ordinary in another, so using location-specific percentiles allowed the analysis to measure heat stress against the conditions to which local populations are normally exposed.

Under this definition, the relative risk of hospitalization was 1.033 on the same day as the heatwave condition. This corresponds to an estimated 3.3% increase in hospitalization risk compared with non-heatwave control days, with a 95% confidence interval ranging from 1.007 to 1.059. The confidence interval indicates the range of values compatible with the statistical analysis, and because it remained above 1, the researchers classified the association as statistically meaningful. Although a 3.3% increase may appear modest at the individual level, even a small percentage change can translate into a substantial number of additional admissions when extreme heat affects large populations.

The consequences became more pronounced when the researchers considered delayed effects. From the same day through the following eight days, the cumulative relative risk reached 1.056, corresponding to an estimated 5.6% increase in hospitalization risk. The confidence interval for this cumulative estimate was 1.011 to 1.103. This lagged pattern suggests that the mental-health effects of sustained heat may not end when temperatures begin to decline. Physiological strain, sleep disruption, dehydration, medication-related vulnerabilities and the psychological stress of prolonged heat could continue to influence health after the most intense conditions have passed.

To identify these patterns, the researchers used a time-stratified case-crossover design. This method is commonly used to study short-term environmental exposures, such as air pollution, temperature extremes or infectious disease outbreaks. Each hospitalization day was compared with control days for the same location and the same day of the week within a defined time period. By making these comparisons within individuals or closely matched time windows, the approach automatically controls for characteristics that do not change rapidly, such as genetics, long-term socioeconomic status and chronic baseline health. The time-stratified structure also helps reduce bias caused by seasonal trends and weekly hospitalization patterns.

The analysis showed that the association was stronger among older adults. Aging can increase sensitivity to heat through reduced thermoregulatory capacity, a higher prevalence of chronic illness and the use of medications that affect hydration, sweating, blood pressure or sleep. Older people may also be more vulnerable to social isolation during extreme weather, particularly if they live alone or have limited access to cooling environments. The researchers also observed stronger associations in low-population-density areas, where hospitals and emergency services may be farther away, cooling infrastructure may be less available and public-health monitoring may be more limited.

The findings do not establish that every heatwave directly causes a mental-health hospitalization, and the study does not identify a single mechanism responsible for the observed increase. Mental and behavioural disorders encompass a broad range of conditions, while hospitalization is influenced by healthcare access, clinical decisions, family support and the availability of community services. Nevertheless, the consistency of the association across multiple countries and hundreds of locations strengthens the evidence that prolonged extreme heat can act as an acute environmental stressor with consequences for mental health. The use of a large international dataset also helps demonstrate that the relationship is not restricted to one climate zone or healthcare system.

The study arrives as heatwaves are becoming more frequent, intense and persistent in many parts of the world. Climate change is increasing the likelihood of days that exceed historical temperature extremes, raising concern that hospitals will face overlapping surges in heat-related illness and mental-health demand. The researchers’ results point to the importance of including mental and behavioural health in heatwave preparedness plans. Early warnings, welfare checks for vulnerable residents, accessible cooling spaces, medication guidance and coordination between emergency departments and community mental-health services could help reduce the burden. As extreme heat becomes a recurring feature of modern summers, the study suggests that protecting public health will require treating mental-health preparedness as an essential part of heatwave response.

Subject of Research: The association between sustained extreme heat and hospitalizations for mental and behavioural disorders across Brazil, Canada, Chile and New Zealand.

Article Title: Mental health hospitalizations associated with sustained extreme heat in multiple countries

Article References: Liu, Y., Xu, Z., Huang, W. et al. Mental health hospitalizations associated with sustained extreme heat in multiple countries. Nat. Health (2026). https://doi.org/10.1038/s44360-026-00166-2

Image Credits: AI Generated

DOI: https://doi.org/10.1038/s44360-026-00166-2

Keywords: heatwaves, extreme heat, mental health, behavioural disorders, hospitalizations, climate change, public health, older adults, case-crossover analysis, environmental epidemiology

The investigation, published as Strengthening England’s Heat Resilience: From Warnings to Long-Term Adaptation, combines survey data from adults in England with interviews involving professionals responsible for communicating heat risks. Around seven in ten people surveyed said they had encountered a Heat Health Alert. Yet the most common explanation among those who did nothing was not that the message was inconvenient or impossible to follow. Instead, many respondents said they did not consider themselves personally vulnerable to high temperatures. That response points to a psychological barrier at the centre of the problem: people may recognise that heat can be dangerous in general while excluding themselves from the group they believe is at risk.

The results arrive after a summer that placed England’s heat-health infrastructure under exceptional pressure. On 26 June, a provisional maximum temperature of 37.7°C was recorded at Lingwood, Strumpshaw Hill, breaking a June record that had stood since 1976. England was also placed under only the second Red Heat Health Alert since the current alert system was introduced. At Hastings, the minimum temperature reached 23.2°C, setting a new June record for the warmest night. Such high overnight temperatures are medically important because the body depends on cooler conditions during sleep to dissipate stored heat and recover from daytime thermal stress. When homes remain hot through the night, physiological strain can accumulate, especially among older adults and people with underlying illnesses.

Another spell of extreme heat in August prompted widespread amber alerts across England, again bringing attention to the country’s exposure to increasingly intense heat events. Heat is not simply an issue of discomfort. The human body maintains its core temperature through mechanisms including sweating and increased blood flow to the skin. In hot conditions, evaporation of sweat becomes the principal route for losing heat, but that process becomes less effective when humidity is high or when people cannot access water, shade or ventilation. Older people may have a reduced ability to sweat or regulate circulation, while cardiovascular, respiratory, renal and neurological conditions can make thermal stress more dangerous. Some medications can also affect hydration, sweating or blood pressure, meaning that heat risk depends on health, housing and social circumstances as well as the temperature shown on a weather forecast.

The UNU-INWEH analysis found that only 18% of people in England reported a high perception of heat risk. Approximately 30% said they knew little or nothing about the actions that could protect them. A similar proportion had encountered no heat alerts or heat-protection information at all, indicating that two different failures can occur: some people never receive a warning, while others receive one without understanding its relevance or knowing how to respond. The report also identifies uncertainty about the alerts themselves. Some respondents did not recognise what they were looking at, while others found the message difficult to interpret or were unclear about the practical steps expected of them. A warning may therefore fail before it becomes a decision, particularly when the language is abstract or does not connect the risk to everyday experiences such as sleeping in an overheated bedroom, travelling on public transport or caring for a family member.

“People aren’t ignoring these alerts because they don’t care,” said Dr Mehri Khosravi, a research fellow in behavioural change and risk communication at UNU-INWEH and lead author of the policy brief. “They read them and think, this isn’t about me.” Khosravi said heat continues to be perceived as good weather or as a threat limited to the very old and very ill. That perception can produce what risk researchers describe as a gap between hazard recognition and personal susceptibility. If people do not identify themselves as vulnerable, they are less likely to change routines, check on neighbours, drink more fluids, seek cooler environments or alter strenuous activity. The challenge is therefore not only to distribute technically accurate forecasts, but also to make the causal pathway from heat exposure to health effects visible and personally meaningful.

Interviews with professionals involved in England’s heat-risk system suggest that its communication architecture has been designed primarily for coordination among institutions. Among the practitioners interviewed, 76% identified informing the public as a communication objective, while substantially less emphasis was placed on motivating people to act or enabling them to do so. During a heatwave, alerts help national agencies, local authorities, health services and emergency responders coordinate their activities. That institutional function is essential, but it does not automatically translate into household-level protection. A message can move efficiently through government and health networks while failing to reach a person living alone, working outdoors, caring for an older relative or occupying a poorly ventilated home.

Digital communication has increased the speed and scale of heat warnings, but the report argues that speed should not be confused with equal access. Online platforms can distribute information rapidly, yet their reach depends on connectivity, digital confidence, language, trust and the habits of different age groups. Social media was identified as the least trusted source of heat information, particularly among older adults. The channels most heavily used by authorities may consequently be least effective for some of the people most exposed to harm. The researchers point to libraries, faith groups, voluntary organisations, social-care providers and other trusted intermediaries as ways to carry heat information into communities. These organisations can do more than repeat an alert: they can explain what it means, identify who may need help and connect general advice to the physical realities of local homes and neighbourhoods.

The report’s central argument is that heat resilience cannot be created by warnings alone. Housing conditions, deprivation, existing health problems and environmental exposure determine how much heat people experience and how effectively they can escape it. A warning cannot cool a top-floor flat that stores heat, provide shade in a treeless neighbourhood or repair a home with inadequate ventilation. The authors call for heat resilience to be incorporated into housing standards, planning policy and retrofit programmes, alongside stronger enforcement and implementation of Part O of England’s Building Regulations, which addresses overheating in residential buildings. They also recommend investment in local-authority and community capacity, because protective action often depends on organisations that know residents personally and can respond before a heatwave becomes a medical emergency.

“A warning tells you a heatwave is coming. It doesn’t tell you how to get through the night in a flat that traps heat,” said Professor Kaveh Madani, director of UNU-INWEH and a co-author of the policy brief. The researchers argue that future evaluation should measure whether alerts prompt protective behaviour rather than simply counting how many people saw them. That shift would treat communication as part of a technical intervention whose success can be tested through outcomes: whether people understand their personal risk, know what to do, can access a cooler place and receive support when they need it. England’s warning system may be effective at detecting and announcing dangerous heat, but the new findings suggest that the next stage of heat adaptation must move beyond notification. As extreme temperatures become a recurring public-health hazard, the decisive question will be whether warnings reach people in forms they trust—and whether the places where they live allow them to act.

Subject of Research: England’s heat-health warning system, public responses to heat alerts, risk communication and long-term heat adaptation.

Article Title: Strengthening England’s Heat Resilience: From Warnings to Long-Term Adaptation

News Publication Date: 14 August 2026

Web References: https://unu.edu/publication/strengthening-englands-heat-resilience-warnings-long-term-adaptation; https://unu.edu/inweh

References: Khosravi, M., Assan, A. and Madani, K. (2026). Strengthening England’s Heat Resilience: From Warnings to Long-Term Adaptation. United Nations University Institute for Water, Environment and Health. DOI: 10.53328/INP26PMK001. Support paper: Assan, A. A., Khosravi, F. and Osei, G. (2026). “The heat is on: Understanding public responses to heat-health alerts in England.” Energy Research & Social Science, 135, 104685.

Keywords: Heat waves, climatology, climate change, climate change adaptation, climate change effects, environmental health, human health, public policy, risk management, risk communication, behaviour modification, disaster management.