As the monsoon begins its advance this week, pushing damp air over a subcontinent still hot from a long summer, India is entering its season of humid heat, the very conditions a new study has now linked to lasting harm in the womb.
During the final stretch of pregnancy, a spell of hot, humid weather is associated with a fall in a child’s eventual height almost four times larger than the same heat gauged by temperature alone, according to the research, published in the journal Science Advances, which tracked nearly 2,00,000 children, almost nine in 10 of them Indian. Its central warning is blunt: climate projections that follow the thermometer but ignore the moisture riding with it are undercounting the harm to children by millions of cases.
The distinction is not academic, because humid heat and dry heat fall on two different Indias. Dry, scorching heat concentrates over the western desert and the arid interior. Humid heat, the kind that leaves the body unable to cool itself, settles over the eastern Gangetic plain, through Bihar, eastern Uttar Pradesh and Bengal, among the most densely populated places on earth and home to some of India’s highest birth rates.
At stake is whether India’s heat plans, which are built largely around temperature, are aimed at the right threat for the youngest and least visible victims of a warming climate: children still in the womb.
What the thermometer misses
The study, led by researchers at the University of California, Santa Barbara, linked household survey data on nearly 2,00,000 children in India, Bangladesh and Nepal to daily records of temperature and humidity. For India, the data for children came from the National Family Health Survey of 2015–16. To separate the two kinds of heat, the team counted days above two thresholds tuned to be equally common: a dry-bulb temperature above 35 degrees Celsius, and a wet-bulb globe temperature, which folds in humidity, above 29 degrees.
What is WBGT?
Wet-bulb globe temperature folds humidity, sunlight and wind into a single heat-stress measure, how hot conditions actually feel to the body, not just the air temperature. When the air is humid, sweat cannot evaporate, and the body cannot cool down, which is why a humid 30 °C is far more punishing than a dry 30 degrees Celsius. WBGT readings run lower than the temperature on a forecast: a value near 30 degrees Celsius already marks dangerous heat.
The gap between them was stark. A standard increase in humid-heat days during the third trimester was associated with a 5.1% fall in a child’s height-for-age score, the standard yardstick of healthy growth. The same rise, measured by temperature alone, was associated with a 1.3% fall, a little more than a quarter as much. Children whose height-for-age falls far enough below the global benchmark are classed as stunted, a condition that shadows learning, earnings and health for life.
Two windows of danger
Timing mattered as much as type. The damage clustered at two moments: the earliest weeks of pregnancy, when the foetus is most fragile, and the final trimester, when the strain on the mother peaks. The middle of pregnancy looked comparatively protected.
The survey data carried a second clue to the mechanism. Spells of extreme humid heat were followed by fewer births six to 12 months later, a pattern consistent with lost or disrupted pregnancies. Spells of extreme temperature, by contrast, were followed by a rise in births within three months, the signature of labour brought on early.
What the numbers can’t say
The findings come with real limits, and they cut in more than one direction. These are statistical associations drawn from a single snapshot of children, not a group followed through time, and stunting has many causes, including nutrition, sanitation and income, that no model fully strips out. The study’s design compares children born in the same community but in different years, which holds steady the things about a place that do not change, but cannot account for every shock. The humid-heat metric is calculated from a heat-index formula, not read off an instrument in a field.
Millions in the balance
The reason the humidity question matters now is what it does to the future. By 2050, the study projects that humid heat could leave between 3 million and 3.7 million additional children across South Asia stunted: the lower figure under a moderate-emissions path, the higher under the high-emission scenario climate scientists label SSP5-8.5. A model that tracks temperature alone projects a fraction of that, 0.3 million to 0.4 million, across the same two scenarios. The difference, between 2.7 million and 3.3 million children, is the harm that simply disappears when humidity is left out of the sum.
Built for the wrong heat
India does not enter that calculation as a bystander. More than a third of the country’s children under five, 35.5%, were already stunted at the last count, according to the National Family Health Survey of 2019–21. The states carrying the humid-heat burden, Bihar, eastern Uttar Pradesh and West Bengal, are among those where that rate is highest, and adaptation money is thinnest. India’s heat action plans, modelled on the one Ahmedabad pioneered in 2013 after a deadly 2010 heatwave, are organised around temperature and around cities. Pregnant women are rarely named as a group at risk, and humidity rarely sets the alarm.
Districts across the east and the Gangetic plain are shaded darkest, indicating the most humid-heat days a year; the Himalayan north and the Deccan plateau are palest.
The damage this study describes is the hardest kind to see. It happens before a child is born, to someone not yet counted in any survey, on days that never trouble a record book, days that are not the hottest, only the most humid. The thermometer, it turns out, has been telling India only half the story.
Data Note
1,98,710 children across India, Bangladesh and Nepal (India 1,74,668; NFHS-4, 2015–16), via the Demographic and Health Surveys / IPUMS-DHS. The child-level data is access-restricted and was not used by DIU; this story relies on the published results. CHIRTS daily temperature and ERA5 humidity (1983–2016); 2050 projections from CHC-CMIP6. Effect sizes are per one-standard-deviation rise in third-trimester exceedance days (the 4 comparison is 5.1% vs 1.3%). The 2050 range spans the SSP2-4.5 and SSP5-8.5 scenarios. The comparison throughout is between humid heat and heat measured by temperature alone (humidity ignored), not between humid heat and “dry heat.” All figures are modelled estimates and associations, not confirmed counts; the 2050 numbers are a scenario projection, not a forecast.
– Ends