Extreme cold outranks heat as the deadlier extreme

Extreme temperature events are forecast to become more frequent and intense, and heat dominates the public conversation about climate and health. Mortality is only one part of the burden, though: hospital admissions, emergency department visits, length of stay and cost all fall on health systems, and those indicators are far less well documented than deaths. Comparing cold and heat side by side within one analytical framework is the only way to see which imposes the larger burden and on which indicator. That comparison shapes where preparedness effort is worth putting.

The authors searched for observational quantitative studies published between 2000 and 2024, building search terms using a MeSH approach and applying PECOS criteria for inclusion. Urgent care utilisation was defined broadly, covering hospitalisation, emergency department visits, length of stay and medical costs, alongside all-cause mortality. From 3411 screened papers, 58 studies met the criteria and were carried into a meta-analysis to produce pooled relative risks for extreme cold, extreme heat and heatwaves across those outcomes.

The pooled relative risk for all-cause mortality was 1.478 for extreme cold, 1.280 for heatwaves and 1.145 for extreme heat, making cold the most hazardous of the three on this measure. Urgent care effects were considerably smaller: heatwaves were associated with a pooled hospitalisation risk of 1.060 and an emergency department visit risk of 1.047. Only limited studies examined length of stay, though those available indicated that both extreme cold and extreme heat increased the risk of stays longer than 24 hours. Economic losses were reported as potentially amounting to €25.5 billion.

The finding that cold carries the higher mortality risk aligns with a long-running strand of temperature-health research, and it complicates the assumption that heat is the dominant threat, even as heat events grow. What this synthesis adds is the direct comparison across urgent care indicators, where the evidence base turns out to be thin: length of stay and cost were covered by few studies. Definitions of what counts as an extreme event differ across the included work, which limits how precisely the pooled numbers can be read. Whether cold's relative burden shifts as climates warm is not answered here.

The reasonable takeaway is that cold snaps deserve as much attention as heatwaves when thinking about weather and health risk, and that mortality is the outcome most sensitive to extreme temperature. Because these are pooled population associations, they describe how death and hospital use rise across a region during an event, not personal odds. The smaller effect sizes for hospitalisation and emergency visits suggest the mortality signal is not simply a reflection of more people turning up at hospital. Nothing in this synthesis identifies which protective measures work.