Two meals with the same nutrition label produced the same blood sugar, but the ultra-processed one pushed insulin higher for most of the next two hours. Ultra-processed foods supply most of the calories Americans eat, and an umbrella review covering almost 10 million people found convincing evidence that eating more of them tracks with a higher risk of type 2 diabetes. The strongest experiment so far came in 2019, when 20 adults living on a research ward could eat as much as they wanted of ultra-processed or unprocessed diets matched for calories, sugar, fat, fiber and sodium on offer. They ate about 508 more calories a day on the ultra-processed diet and gained 0.9 kg in two weeks. That trial could not separate the extra eating from the food itself. In the new study, 32 healthy-weight adults aged 18 to 45 ate two 300-calorie meals on separate mornings in random order: whole foods such as egg, banana and cheese, or ultra-processed foods such as a peanut butter and jelly sandwich, a cookie and veggie chips. The meals were matched within 1.6% on calories, carbohydrate, fat, protein, fiber, sodium, water and glycemic index, and each was eaten in full in a sealed room that measured energy use, with blood drawn seven times over three hours. 1. Total blood sugar over three hours was the same after both meals, although it rose faster after the whole-food meal and peaked at 40 minutes either way. 2. Insulin was higher after the ultra-processed meal at 40, 60, 90 and 120 minutes, and its total over three hours was clearly larger. 3. After the ultra-processed meal, the body used more energy and more fat but less carbohydrate, and in brain scans this shift tracked with how reward areas responded to food pictures, even though liking, hunger and fullness did not differ. Taken together, processing changed how the body handled a meal with an identical label, but the trial does not show that the difference matters. Body weight follows total energy, and across 32 controlled feeding studies, diets that lowered insulin by cutting carbohydrate did not produce more fat loss at equal calories. Here the ultra-processed meal even raised energy use slightly. The proven cost of ultra-processed food is that people eat about 500 more calories a day when it fills the plate, so the practical lever is how much of it you eat. This was one 300-calorie meal eaten once by young, lean, healthy adults, so it cannot show long-term effects. Protein came from different foods in each meal, the ultra-processed meal had more additives, and food structure was not measured, so the trial cannot say which part of processing mattered. The researchers knew which meal was served. References Hutelin Z et al., Nature Metabolism, 2026 Lane MM et al., BMJ, 2024 Hall KD et al., Cell Metabolism, 2019 Hall KD and Guo J, Gastroenterology, 2017