Science1 distinct publisher2 min readPublished
A VUB-led study in Science Advances splits attributable humid heat by age band. The gap between children and people in their 60s is part risk and part demography.
The Scientist · Science desk

Compiled by The ScientistSomething wrong?How this is made
Divide the two headline figures by their own percentages and the cohorts behind them fall out. Up to 560 million children is 43 percent of the world's under-10s, which implies an age band of about 1.30 billion [2][15]. The 190 million people in their 60s carrying the same extra month are 30 percent of theirs, implying roughly 630 million [3][16]. The children's band is a little more than twice as large before any heat is counted [16]. So the per-person gap is 43 against 30, about 1.4 to one, and the near-triple headcount gap is that 1.4 multiplied by the demography [17][14]. Both are true statements about unfairness. They are not the same statement, and the one that survives a shrinking birth cohort is the smaller one.
That distinction runs through the projections, where the shares move faster than the heads. Between the 1.5C and 2C snapshots the count of children with an extra month rises by 30 million, while the share rises ten points, from 49 to 59 percent [4][20]. Back out the denominators and the reason shows: about 1.27 billion children in the first, about 1.10 billion in the second, a fall of roughly 13 percent [19]. The published figures are rounded to two digits, so those implied totals are approximate. Anyone quoting 59 percent should say that part of the increase is fewer children rather than more heat.
The cleanest attribution number in the release is the other one. For total heat stress of five months or more, the study sets 120 million children in a world without human influence against 350 million today, which is 230 million children in that bracket who would not otherwise be in it [6][18]. That statistic names both worlds, which the extra-month figure does not. It is also worth noting that 560 million is presented as an upper bound, with no lower bound printed [2].
None of this lands evenly. The largest increases in humid heat are in tropical regions and poorer countries, where populations are younger, and South Asia, Southeast Asia and West Africa hold the most exposed children today [8]. Humid heat is the chosen variable because it stops sweat from doing its job [9], and small bodies heat up faster and sweat less efficiently than adult ones [10]. Rosa Pietroiusti of VUB, the lead author, said climate change is "already causing hundreds of millions of children to suffer from dangerous heat worldwide," and that children in developing countries face disproportionate exposure now and in future despite having contributed least historically [12].
Ranked by verification strength, evidence, and original report placement.
A study published in Science Advances by an international team led by Vrije Universiteit Brussel is the first analysis to quantify how heat exposure directly linked to climate change affects different age groups globally, both now and in the near future.
The researchers combined population data with state-of-the-art climate models and attribution science that identifies the role of climate change in extreme weather events.
Up to 560 million children ages 0-9, or 43% of all children in that age group, experience at least one additional month (30 or more days) of heat stress per year because of human-induced climate change.
That exposure outpaces any other age group and is nearly triple the 190 million people, or 30%, ages 60-69 experiencing the same additional threat.
Climate change has almost tripled the number of children exposed to at least five months (150 days) of total heat stress globally, from 120 million (9% of children) in a world without climate change to 350 million (27%) today.
The 350 million children exposed to at least five months of total heat stress is nearly three times the 120 million adults ages 60-69 (18%) in the same bracket.
Follow any of these and your For You feed starts watching them — no settings page required.
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Peer-reviewed study, press-release-only visibility
The underlying artefact is a peer-reviewed Science Advances paper with a DOI and a named method stack (population data, climate model ensembles, event attribution), which lifts evidence above promotional baseline. But everything in this cluster comes from the authoring institution's release: no uncertainty ranges, no heat-stress threshold definition, no population baseline year, no independent comment, and two assertions (scenario exceedance likelihood, 2026 record heat waves) carry no supporting citation. The one check available from the release's own figures - the implied 0-9 denominators - does not reconcile across scenarios.
No uptake signals supplied
The supplied material contains no adoption evidence of any kind: no citations, no policy or agency uptake, no heat action plans adopting the study's metrics, no dataset or code release, no downstream use. A newly published paper's uptake cannot be inferred from its own announcement, so this dimension is left unmeasured rather than estimated.
Headline ratio overstates the per-person gap
The substance is real and moderately well grounded, but the framing is stretched in three specific ways: the marquee 'nearly triple' comparison is a headcount ratio driven substantially by the 0-9 band being roughly 2.1 times larger than the 60-69 band, whereas the per-person gap implied by the release's own shares is about 1.4 to one; 'up to 560 million' presents an upper bound as the headline number; and the internally inconsistent scenario denominators are presented as precise projections. The gap is positive but modest, not fabrication - the attributable increment of roughly 230 million children into five-month-plus heat stress stands on its own without the ratio rhetoric.
Authoring institution's release with explicit policy asks
The single source is a promotional release from the institution that produced the study, distributed through an aggregator that republishes such releases largely intact. Both quoted voices are the paper's own authors, and their statements move beyond findings into advocacy - drastic fossil fuel emission cuts, Paris-aligned warming limits, a 'huge increase' in climate finance to developing countries - alongside framing that ties the results to school strikes and climate litigation. Those are legitimate positions, but they are interested ones, and the release's choice of the most striking available comparison (headcount 'nearly triple' rather than the 1.4:1 per-person gap) is consistent with that incentive.
Solid artefact, thin corroboration
Confidence is moderate. The peer-reviewed paper and DOI make the core existence and direction of the finding reliable, and the derived arithmetic is checkable from the text. Against that: one publisher, one source, that source being the interested party; no uncertainty bounds; an unresolved denominator inconsistency; and two uncited assertions. Enough to report the finding and the framing caveat, not enough to treat any single headcount as settled.
science
Smarter Models Are Better Followers: LLM Agents Lock Into Majority Views At 1,000-Agent Scale1 distinct publisher
science
Eighteen volcano myths, itemized: a correction list hazard communicators can cite1 distinct publisher
science
Forty-two Jomon genomes put a date on Ice Age isolation and a frequency on cold alleles1 distinct publisher
science
Valley fever tracks what grows next door, and freshly idled fields sit on the wrong side1 distinct publisher
Distinct publishers with included, body-backed reporting in this cluster.
phys.org
1 article · August 26, 2026