Does a toddler's sweet tooth start in breast milk? Hormones in the milk at six months lined up with tastes a year later

Breast milk carries more than nutrients. It also holds hormones, among them insulin, leptin (a fullness signal made by fat tissue), the stress hormone cortisol, and a newer one called LEAP2, which works against ghrelin, the body's hunger signal. A team at a children's hospital in La Plata, Argentina, working with colleagues in Sweden and France, asked whether these hormones in the first months of life have anything to do with how much a child later likes sweet and fatty food. Their paper appeared in the Journal of Neuroendocrinology on 1 October 2026.

At the six-month check-up, mothers gave a milk sample and their babies a blood sample. At 18 months the toddlers came back after a four-hour fast for a taste test: four bottles, each offered for 45 seconds, water, then two of a lightly sweet lactose drink at the strength found in breast milk, then water again. The same was done with a plain oil-in-water emulsion. The bottles were weighed to see how much each child drank, and a researcher who didn't know which drink was which scored the child's face and body language on a five-point scale from strong rejection to strong acceptance. Parents also recalled everything the child ate the day before, and foods were sorted by whether they are 'hyperpalatable', combining fat, sugar or salt in ways that make them hard to stop eating.

The links pointed in several directions at once. More insulin in the milk went with toddlers drinking less of the sweet drink and liking it less. More leptin in the milk went with more interest in the sweet drink, yet also with a smaller share of hyperpalatable foods in the day's diet, the strongest link in the study. Cortisol in the milk went with drinking more of both the sweet and fatty drinks. The baby's own hormones often told a different story from the milk's: higher cortisol in the baby's blood went with eating more hyperpalatable food, while higher LEAP2 went with eating less of it. The team tested whether the milk hormones worked by raising the baby's own levels and found no sign of that. Weight at 18 months mostly stood apart: the share of hyperpalatable food and the fat scores weren't linked to it, and the one link found ran the unexpected way, with toddlers who liked the sweet drink more being slightly lighter for their length.

The size of the study is the first thing to keep in mind. Of 725 mother and baby pairs who volunteered, 82 were enrolled and 51 came back at 18 months, but many toddlers refused the bottle or didn't finish the test, leaving 28 in the analysis. Many hormone and outcome pairs were tested, each hormone was measured once, and the day's diet came from a single recall, which the authors list among their limits along with the observational design. The 'hyperpalatable' list at this age was mostly ordinary toddler food: milk, bread, yogurt, crackers, cheese and sweet biscuits. So this is an early map of possible pathways, not a reason for any mother to worry about what is in her milk. What it does suggest is that the appetite system is already being shaped before a child can ask for a biscuit, and that breast milk may be one of the signals involved.