Research explained: prospective cohort study
Kids who ate more ultra-processed food showed stronger brain reward responses and larger gains in waist size
A one-year USC cohort of 46 children of healthy weight linked a larger share of calories from ultra-processed food to stronger food-cue responses in a brain reward area and to gains in waist size.
Children who got more of their calories from ultra-processed food showed stronger responses to pictures of food in the dorsal striatum, a brain area tied to reward and habit, and that stronger response was linked to gains in body mass index and body fat over the next year, according to a cohort study in JAMA Network Open (published October 1, 2026). A higher ultra-processed share at the start was also linked to a larger increase in waist size. The study drew wide news coverage in recent weeks.
What the researchers did
The team at the Keck School of Medicine of the University of Southern California followed 46 children of healthy weight from the BrainChild study. Their mean age was 8.7 years, and 26 of them (57%) were girls. At the start, each child had a brain MRI scan while looking at food images, several 24-hour dietary recalls, and body measurements. The share of calories from ultra-processed food was scored with the NOVA system. Body mass index, body fat percentage and waist size were measured again 1 year later. The analysis adjusted for age, sex, starting BMI and total calories.
Higher ultra-processed intake went with a stronger dorsal striatum response (β = 0.34), but not with a clear change in the ventral striatum. A stronger dorsal striatum response was linked to 1-year gains in BMI (β = 0.32) and body fat percentage (β = 0.44). Ultra-processed intake itself was linked to a rise in waist size (β = 0.38), while its links with BMI and body fat leaned upward at the edge of statistical significance. Taking the brain response into account shrank those links by 31% for BMI and 44% for body fat, which suggests reward responses may be part of the pathway.
How strong is the evidence?
This is a small observational study, so it shows links, not cause and effect. Diet came from self-reported recalls, and one year is a short window. The children were all of healthy weight, so the results may not apply to every child. The authors report no conflicts of interest.