UPFs spike insulin, even with matched nutrients

Clinical takeaway: The level of processing of food appears to affect insulin response beyond calories and macronutrients. 

Ultraprocessed foods (UPFs) supply more than half of the calories Americans eat, and higher intake has tracked with obesity, metabolic dysfunction and cancer. Because these foods differ from less processed ones in composition as well as processing, the evidence so far has struggled to separate the two. A randomized crossover trial now tests processing apart from nutrient content. 

Insulin response was greater after the ultraprocessed meal, running higher at 40, 60, 90 and 120 minutes before both conditions returned to near baseline after three hours. Blood glucose didn't follow. Total glucose exposure over the three hours was the same, and both meals peaked at 40 minutes. Glucose rose faster after the non-ultraprocessed meal, then trended higher after the ultraprocessed meal at two and three hours. 

Metabolic rate over the three hours was greater after the ultraprocessed meal, which also blunted the post-meal shift from fat to carbohydrate burning: carbohydrate oxidation was lower and fat oxidation higher. 

Participants also viewed pictures of ultraprocessed and non-ultraprocessed foods in an MRI scanner, on a separate day. Those whose peak carbohydrate burning was lower after the ultraprocessed meal than after the non-ultraprocessed meal showed a relatively stronger brain response to the ultraprocessed pictures, among the 29 who completed both parts.

The link appeared in the caudate and ventral striatum, parts of the brain's reward circuitry, and in the left superior temporal gyrus. Across all 52, participants were not willing to pay more for either type of food in an auction task, yet brain activity tracked what they were willing to pay in opposite directions: in the visual cortex and striatum it rose with willingness to pay for non-ultraprocessed foods and fell with willingness to pay for ultraprocessed foods. 

In this randomized crossover design, 32 healthy-weight adults ages 18 to 45 ate one matched 300-calorie ultraprocessed meal and one non-ultraprocessed meal on separate days inside a whole-room calorimeter, with blood drawn repeatedly over the next three hours. Brain imaging covered 52 adults, 29 of whom also completed the metabolic sessions. 

The investigators want to repeat the comparison in other populations, over longer periods, with bigger meals and more food pairings. They also want to test whether specific additives or processing steps account for the different responses, which this trial could not separate from processing as a whole. 

"We know that the majority of ultraprocessed foods are high in fat, they're high in sugar, they're low in fiber, and they're low in protein," says Alexandra DiFeliceantonio, PhD, associate professor at the Fralin Biomedical Research Institute at VTC. "But if we artificially hold all of those things constant, is there something about the processing that leads to a different outcome?" 

Source: Hutelin Z, et al. (2026 Oct 5) Nat Metab. Metabolic and neural responses to ultraprocessed foods: a randomized, controlled, crossover study