Just a few days of dangerous heat each year may be linked to greater dementia risk, according to a 20-year US study that reveals how neighborhood disadvantage and limited cooling infrastructure could leave some communities more vulnerable.
Study: Outdoor extreme heat-related dementia risk and the role of socioeconomic disparities: evidence from 20 years of data on US older adults. Image credit: Rainer Fuhrmann/Shutterstock.com
Exposure to just a few days of extreme outdoor heat each year may be linked to a higher risk of dementia, according to a study of more than 28,000 US adults using data spanning two decades, published in the journal Alzheimer’s & Dementia. The findings suggest that socially vulnerable communities with limited cooling infrastructure may face greater risks.
Extreme heat raises concerns about long-term dementia risk
Extreme heat poses significant health risks, particularly for older adults, who are more vulnerable to its effects. Previous studies have linked hot weather to poorer cognitive performance and faster cognitive decline, while growing evidence suggests that socioeconomic disadvantage may further increase vulnerability. People with fewer financial resources may have limited access to healthcare and cooling facilities, greater exposure to outdoor heat, and less social support during heat-related illness.
Housing conditions can also influence heat exposure. Older, poorly insulated homes and mobile homes may retain more heat, while rural communities may face additional challenges because of aging housing, limited healthcare access, and lower air-conditioning (AC) use.
Despite these concerns, little is known about whether prolonged exposure to extreme outdoor heat is associated with the development of dementia in the general older population. Two previous studies from Taiwan examined dementia risk among people with a history of heat-related illness, but their findings may not apply to the broader population.
To address this gap, the researchers investigated whether cumulative outdoor heat exposure was associated with new-onset dementia in a nationally representative US cohort and whether socioeconomic circumstances and neighborhood conditions influenced this association.
Twenty years of data reveal heat exposure patterns
The study drew on data obtained from 2000 to 2020 from the Health and Retirement Study (HRS). It included 28,512 community-dwelling, dementia-free adults aged 50 to 89 at baseline, with an average starting age of 59.8 years, with at least one follow-up over an average of 10.5 years. During that time, 3,773 developed dementia (13.2%) and 7,887 died without a dementia classification (27.7%).
Dementia was identified using the validated Langa–Weir algorithm, based on cognitive assessments and proxy reports. To reduce false positives, researchers generally required an initial dementia classification to be followed by another assessment indicating dementia or cognitive impairment. Participants classified with dementia who died before the next survey wave were also counted as cases.
Heat exposure was categorized by the daily heat index (which combines air temperature and humidity to estimate how hot conditions feel), which was estimated for each respondent's census tract from 4-km gridMET data. Days were classified into two National Weather Service categories: Extreme Caution, at 90–102°F (32.2–39.3°C), and Danger+ at 103°F (39.4°C) or higher. Across the study observations, participants experienced an average of 5.8 Danger+ heat days per year.
To obtain the exposure window, heat days were counted over the preceding five years and recalculated at each wave. The researchers used move dates to map changes in residential location.
The researchers also examined whether personal circumstances and neighborhood conditions influenced the association between extreme heat and dementia. At the individual level, these included markers of socioeconomic deprivation, such as low education, non-housing wealth, living alone, and living in a mobile home. They also attempted to capture differences in neighborhood heat-mitigative resources, such as the Social Vulnerability Index (SVI) of the Centers for Disease Control and Prevention (CDC), tract-level estimated prevalence of central air conditioning (AC), and the degree of urbanization.
They used Fine–Gray weighted subdistribution hazard models to estimate dementia risk, with death as a competing event and age as the time scale. They adjusted for demographics, cardiometabolic conditions, smoking, drinking, region, calendar year, and the effects of the moderating factors.
Dementia associations emerge at the highest heat-index levels
The researchers found no statistically significant association between the number of days of Extreme Caution heat-index conditions and dementia, irrespective of exposure quartile. Conversely, Danger+ heat-index conditions were associated with higher dementia incidence in some exposure categories.
Compared with the lowest exposure (0.2 or fewer Danger+ days per year), moderate exposure (0.4–1.8 days/year) carried a 14% higher subdistribution hazard.
At moderate-high exposure (2–7 days/year), there was no significant increase in risk compared to the lowest exposure. High exposure (7.2 or more days/year) carried a 23% higher hazard.
Thus, the dose-response pattern was not monotonic. However, they found clearly separated cumulative curves for low and high exposure to Danger+ heat, particularly among older people. In contrast, the Extreme Caution curves remained close together. However, these were unadjusted curves.
Neither heat category was significantly associated with mortality in the primary analysis. Therefore, the authors suggest that a five-year cumulative measure captures something different from the acute heat spikes that drive heat-related deaths.
Neighborhood disadvantage strengthens heat-associated dementia risk
The link between extreme heat and dementia was particularly pronounced in disadvantaged neighborhoods, suggesting that where people live may influence their vulnerability to prolonged heat exposure.
Among those experiencing the highest levels of Danger+, the predicted dementia hazard was 40% higher in the most socially vulnerable neighborhoods, compared with 11% higher in less vulnerable areas. Both figures were calculated against the lowest heat-exposure group, with a 26% relative difference between the predicted hazard ratios.
Access to cooling infrastructure also appeared to matter. In neighborhoods with lower estimated central air-conditioning (AC) prevalence, the predicted dementia hazard was 36% higher among those with the greatest heat exposure, compared with 13% higher in areas where central AC was more common. This represented a 20% relative difference between the predicted hazard ratios.
These differences emerged only at the highest levels of dangerous heat. At lower exposure levels, neighborhood vulnerability and AC prevalence did not significantly alter the association with dementia.
Interestingly, individual circumstances, including education, household wealth, living alone, and mobile home residence, did not significantly modify the association. Neither did whether participants lived in urban or less urban areas.
Longer heat exposure windows show more consistent associations
The association between extreme heat and dementia remained largely consistent when researchers tested their findings using different methods, particularly among those with the highest exposure to Danger+ heat (a heat index of 103°F or higher). However, the association at moderate exposure levels was less consistent.
After accounting for residential moves and transitions into nursing homes, the dementia hazard remained 19% higher in the highest Danger+ exposure group, while the association at moderate exposure was no longer statistically significant.
The researchers also examined whether unusually hot conditions relative to each neighborhood's historical climate were associated with dementia. Using this alternative measure, those in the highest exposure group had a 15% higher dementia hazard. The smaller association suggested that the absolute intensity of heat may be more important than how unusual the temperature is for a particular area.
The findings changed little after accounting for air pollution. The length of heat exposure also appeared to matter. When researchers extended the exposure window from five to seven years, the association remained similar, with a 21% higher dementia hazard in the highest exposure group. However, shortening the window to three years reduced the estimate to 12%, which was not statistically significant. This suggests that the association may become more apparent over longer periods, although the study could not establish a cumulative causal effect.
To determine whether the findings were driven by the hotter southern United States, researchers repeated the analysis among 17,298 participants living outside the South. The association persisted, with dementia hazards 26% and 30% higher in the moderate and highest Danger+ exposure groups, respectively, compared with the lowest exposure group.
Finally, the researchers broadened their definition of dementia to include 370 additional participants whose classifications had not remained consistent across assessments. Even with these cases included, the highest Danger+ exposure group had a 19% higher dementia hazard, reinforcing the consistency of the main finding.
Severe heat may trigger harmful changes in aging brains
According to the authors, these findings suggest a possible physiological threshold beyond which severe heat exposure may become more harmful to the aging brain. However, the study did not directly measure the biological processes involved or establish a precise temperature threshold.
Preclinical experimental evidence in rats also shows a threshold for blood-brain barrier impairment at higher temperatures (38°C for four hours) compared to 35°C.
The authors offer plausible biological mechanisms, including reduced cerebral blood flow, neuroinflammation, oxidative stress, chronic cortisol elevation, and secondary effects such as poor sleep and dehydration. However, these were not demonstrated in the current study and remain hypotheses only.
The findings suggest that neighborhood infrastructure and cooling capacity may be associated with heat-related dementia risk beyond individual-level characteristics. However, the researchers measured estimated neighborhood AC prevalence rather than individual household AC ownership or use, and did not establish that air conditioning directly protects against dementia.
These findings do not agree completely with prior work, where heat-related cognitive effects were concentrated among people with lower education or income.
Survey-based dementia assessments limit certainty about heat associations
The findings should be interpreted with some caution. As an observational study, the research cannot establish whether extreme heat directly contributes to dementia.
The researchers estimated outdoor heat exposure at the neighborhood level, meaning they could not account for differences in how much heat individuals experienced or the temperatures inside their homes. Estimates of central AC prevalence were also based on 2016 data and applied across the entire study period. If access to AC increased over time, this approach may have overestimated cooling availability in earlier years, potentially weakening the observed association.
Another limitation was that dementia was identified through survey-based cognitive assessments rather than clinical diagnoses. The researchers also lacked information on dementia subtypes, making it impossible to determine whether the associations differed between Alzheimer's disease, vascular dementia, and other forms of the condition.
Other unmeasured factors, including environmental pollutants, diet, and genetic susceptibility, may have influenced the findings. Although the researchers could not completely rule out the possibility that people experiencing early cognitive decline had moved to hotter areas, additional analyses provided little evidence that such moves explained the results.
Finally, the study could not assess lifetime heat exposure, leaving questions about the longer-term effects of extreme heat on dementia risk.
Study highlights need for community-focused extreme heat adaptation
The authors conclude that outdoor heat-index conditions of 103°F or more, occurring even a few days each year over several years, may be associated with increased dementia risk. They suggest that dementia prevention strategies should consider local extreme heat exposure and prioritize approaches that may reduce cumulative thermal stress, particularly in disadvantaged communities.
In high-heat areas, the authors recommend moving beyond generic heat action plans designed for whole populations, toward locally appropriate strategies developed with residents. These could combine structural improvements, including reflective roofs and better-insulated housing, with affordable measures such as staying hydrated and limiting outdoor activity during extreme heat. They also recommend prioritizing resources for communities facing socioeconomic and infrastructure constraints.
Overweight dementia patients likely to live longer than normal weight peers