Do ultra-processed foods make people eat more, even when the nutrients are matched?

In one carefully controlled hospital trial, yes. Twenty weight-stable adults were admitted to the NIH Clinical Center and randomly given an ultra-processed diet for two weeks and an unprocessed diet for another two, in either order. The meals were designed to be matched for presented calories, energy density, macronutrients, sugar, sodium and fibre, and the volunteers were told to eat as much or as little as they wanted.

On the ultra-processed menu they ate about 508 kilocalories a day more. The extra came from carbohydrate and fat, not protein. They gained 0.9 kilograms on that diet and lost 0.9 kilograms on the unprocessed one, and weight changes tracked how much they ate.

Ultra-processed food is a classification, not a nutrient: the Nova system groups foods by how industrially they are made and says nothing about their nutritional content. That is why the trial mattered. Epidemiological data had associated ultra-processed food with obesity and non-communicable disease, but before this trial there had never been a randomised controlled trial of increasing or reducing such foods.

The result does not say why. The authors write that the study was not designed to identify the cause of the differences in intake. It also had limits of size and setting: 20 people, 28 days that the authors call relatively modest, and an inpatient ward that makes it difficult to generalise to free-living conditions.