A randomized crossover study suggests that food processing itself can change short-term metabolic and brain responses, even when meals are closely matched for calories, macronutrients, fiber, sodium and estimated glycemic impact.
What changed
Researchers enrolled 57 healthy-weight adults; 32 completed both metabolic meal sessions and 52 completed brain imaging. On separate days, participants ate roughly 300-calorie ultraprocessed or non-ultraprocessed meals. The ultraprocessed meal produced a larger insulin response and higher energy expenditure, while overall blood-glucose exposure did not differ. Differences in carbohydrate oxidation were also associated with activity in reward-related brain regions when participants viewed food cues.
How strong is the evidence?
This was a human randomized crossover experiment, a strong design for comparing acute responses because each participant served as their own control. The meals were closely nutrient-matched, helping isolate processing-related differences.
What remains uncertain
The metabolic comparison involved only 32 mostly young, healthy-weight adults and measured responses for three hours after a single meal. Insulin and brain-imaging signals are surrogate outcomes; the study did not test weight gain, diabetes or cardiovascular events. Protein sources, additives and food structure also differed, so the specific cause cannot be separated.
Why it matters
The findings offer a plausible mechanism for why nutritionally similar foods may not act identically after eating. They support longer trials that test whether repeated exposure produces meaningful health effects, but they do not prove that one meal causes chronic disease.
Illustrative stock photo by Fallon Michael on Unsplash; it does not depict the study meals, participants or results.
Sources
Original study in Nature Metabolism, published October 5, 2026.
