Climate change could make humidity-driven heat risks more dangerous, study finds

By Vijay Kumar MalesuReviewed by Susha Cheriyedath, M.Sc.Apr 19 2026

Temperature alone may miss the real danger: this study shows that humidity can sharply raise health risks during both hot and cold weather, with climate change likely to make these compound threats more common.

Study: Humidity may amplify the temperature-related health risks in the context of climate change. Image Credit: Quality Stock Arts / Shutterstock

In a recent study published in the journal Scientific Reports, a group of researchers evaluated how combined temperature and humidity exposures influence health risks and identified high-risk thresholds under climate change scenarios.

Temperature-Humidity Risk Background

What if the real danger during extreme weather is not just heat or cold, but how humidity amplifies it? Thousands of people die every year from extreme temperatures, but temperature alone is not the only risk. Humidity affects how our bodies adapt to temperature changes. As climate change increases both temperature extremes and humidity, it is essential to understand how these factors interact to affect human health, and further research is needed to improve risk assessment processes.

China Ambulance Dispatch Climate Study Design

Researchers analyzed emergency ambulance dispatch data from 13 large cities across China during 2013-2019, using daily data collected from local Centers for Disease Control (CDC) and ambulance services. These data provided an index of acute health outcomes.

Daily mean temperature and relative humidity were obtained from national meteorological databases. Environmental confounding factors were accounted for using average air pollution levels, including particulate matter less than 2.5 micrometers (PM2.5), ozone (O3), and sulfur dioxide (SO2). Climate projections were generated using 12 global climate models participating in the Coupled Model Intercomparison Project Phase 6 (CMIP6), with Shared Socioeconomic Pathways (SSPs) and Representative Concentration Pathways (RCPs).

A time-series quasi-Poisson regression model with a distributed lag non-linear model (DLNM) was applied to assess delayed health effects over seven days. Temperature and humidity were divided into percentile-based categories, generating 400 compound exposure scenarios, and grouped into warm-wet, warm-dry, cold-wet, and cold-dry categories.

Researchers pooled city-specific estimates into national-level risk estimates using traditional meta-analysis methods. Thresholds for high-risk events were identified based on relative risk (RR) patterns and statistical inflection points.

Compound Temperature-Humidity Health Risk Findings

The analysis included approximately 2.46 million ambulance dispatch records, revealing a clear relationship between weather conditions and health risks. It also showed that the association between temperature and ambulance dispatches follows a U-shaped curve, with both extreme heat and extreme cold associated with a higher risk of dispatches.

The analysis of combined temperature and humidity effects showed that all four compound event types were associated with increased risk, and that compound temperature-humidity events posed higher risks than single temperature exposures under the same temperature thresholds. The greatest risk assessed was for the cold-dry event type (RR = 1.102; 95% CI = 1.045 - 1.161), followed by warm-wet (RR = 1.093; 95% CI = 1.068 - 1.118). Warm-dry and cold-wet events also showed elevated risks, with pooled RRs of 1.091 and 1.081, respectively.

Importantly, the findings suggest that temperatures need not be at the extremes of the distribution to pose a high risk when combined with unfavorable humidity. This finding suggests that relying solely on temperature thresholds may underestimate real-world health risks. The study also demonstrated that compound events consistently produced higher risks than single temperature exposures under the same thresholds.

Age-Specific Vulnerability and Climate Projections

Older adults were more likely to call an ambulance during the hot-wet and hot-dry periods than younger adults. Also, older people aged 80+ had the highest risk during warm-wet events, while people aged 60-79 years also had elevated risk. Younger and middle-aged adults were more vulnerable during cold-wet events, although some increased sensitivity was also observed among those aged 80 years and older and those younger than 17 years, showing that the risk patterns varied across age groups.

The spatial analysis revealed a high frequency of compound events in the southeastern and central climate regions, where it is warm and humid. Historically, these regions experienced between 20 and 40 compound events each year. However, with projected future climate scenarios, the frequency and area of compound events will increase significantly, particularly under high-emission scenarios such as SSP585.

The number of ambulance dispatches attributable to compound events is expected to increase over the coming decades, reach its highest levels around the mid-21st century, and then decline later in the century. Heat-related events, particularly warm-wet conditions, are projected to become the dominant driver of health risks. In contrast, cold-related risks are expected to decline over time.

Early Warning and Public Health Implications

The study suggests that humidity significantly amplifies the health risks associated with temperature extremes, making compound temperature-humidity events more dangerous than temperature alone. Both hot and cold conditions become dangerous when combined with unfavorable humidity levels, particularly for vulnerable populations. As climate change continues, these combined risks will occur regularly, shifting the burden toward heat-related risks. It is important to include multiple meteorological factors in health risk assessments to achieve more accurate predictions. These results underscore the importance of improved early warning systems and public health strategies grounded in cumulative environmental exposures.