Ultra-processed meal lifts insulin more than a nutrient-matched minimally processed meal in 32 adults

DiFeliceantonio, a Virginia Tech neuroscientist whose team reported the work in Nature Metabolism, framed the experiment as a problem of controls ^[3]. "We know that the majority of ultra-processed foods are high in fat, they're high in sugar, they're low in fiber, and they're low in protein," she said ^[11]. "But if we artificially hold all of those things constant, is there something about the processing that leads to a different outcome?" ^[12]

Her team's answer was a pair of plates that look nothing alike. One held deli turkey, instant mashed potatoes, a chocolate chip cookie, frosted wheat cereal, veggie chips and a small peanut butter and jelly sandwich on white bread ^[6]. The other held a hard-boiled egg, a banana, cranberries, cheese and sea salt ^[7]. "The two meals are matched within 1 percent of carbs, fat, proteins, calories, water, and salt," said Zach Hutelin, the study's first author ^[9]. "If these meals had a nutritional label, they would be practically identical." ^[10] Because each person ate both, their response to one plate is set against their own response to the other, so differences between people drop out of the comparison ^[5].

The thing this doesn't tell you is which difference between the plates produced the effect. Hutelin's matching list names carbohydrate as a whole, while DiFeliceantonio's list of things to hold constant also named sugar and fiber ^[9][11]. The ingredients differ too: cookie and frosted cereal on one side, egg and fruit on the other ^[6][7]. We think the cleaner test would process the same ingredients two ways. As run, the study compares two particular meals that differ in more than their processing.

ScienceAlert's account describes the insulin rise as significantly higher and does not give the figures ^[14]. Hutelin put it more loosely. "What we noticed with the ultra-processed food is that insulin response was much higher and blood sugar stays a little bit higher for a little bit longer," he said ^[15]. He also said, "We were shocked when every single metabolic metric differed." ^[2] Those metrics come from a single 300-calorie meal eaten by healthy-weight adults whose average age was 26 ^[8][4].

The brain-imaging arm is correlational by design. Participants viewed pictures of food in an MRI scanner in a session apart from the meals, and the striatum link was found among those who did both ^[16]. Of the 57 adults recruited, 25 did not take part in the meal test ^[18]. The researchers suggest that how the body burns fuel may be tied to cues about the food, possibly including how processed it is ^[17]. The reported result is an association between two measurements and does not show which way any influence runs ^[16].

For a cause, the team points first to food structure. "These mechanisms might involve altered nutritional availability as a result of altered physical structure in ultra-processed foods," the researchers said in a briefing on the study ^[20]. Commercial food additives are the other candidate, and the group intends to examine them next ^[21].