Two meals can look similar on a nutrient comparison and still produce different short-term responses in the body. That is the central finding of a human study published in Nature Metabolism on October 5, 2026, examining ultraprocessed and non-ultraprocessed meals.
The study offers evidence that food processing deserves attention beyond a simple calories-and-macronutrients comparison. It does not prove that every packaged food causes disease, that a single insulin rise is a diagnosis, or that one meal experiment can settle a lifetime of eating choices.
What the researchers compared
The full study included 57 healthy-weight adults. Of these, 32 completed both metabolic meal sessions in the randomised crossover design, and 52 completed the brain-imaging session. A crossover compares the meal conditions within the same people rather than only comparing unrelated groups. The researchers selected closely matched meals.
The matching extended beyond calories, carbohydrate, protein and fat to features including fibre, sodium, water and energy density. That matters because an unequal comparison could simply reveal the consequences of different nutrient amounts. Here, the researchers were deliberately asking whether responses still differed when those familiar differences were reduced.
This was not a comparison of an entire long-term diet with another diet. It measured responses to specific experimental meals. The participants also represented a particular adult study population, not everyone with every metabolic condition or dietary need.
What changed after the meals
After the ultraprocessed meal, the researchers reported a larger insulin response and greater energy expenditure, alongside reduced carbohydrate oxidation compared with the matched alternative. Carbohydrate oxidation refers to using carbohydrate as fuel; it is not the same thing as the amount of carbohydrate written on a label.
In the brain-imaging component, metabolic differences were associated with responses to food cues. These findings add a possible link between what happens after eating and the way food is represented in the brain. They do not show that the participants had become addicted to a food.
Nor does greater energy expenditure in this experiment establish a weight-loss benefit. A short-term physiological measurement cannot be interpreted as a complete account of energy intake, appetite, body weight or health over time. The direction of one measurement is not enough to rank the whole diet.
Why matching nutrients is useful but not the final answer
Closely matching meals strengthens a specific comparison. It helps challenge the idea that calories and the usual nutrient totals describe everything relevant about food. But it does not make the meals identical in every physical feature or identify one ingredient as the cause of all the observed differences.
The authors discuss altered physical structure and nutritional availability as plausible contributors. A proposed mechanism is something to investigate. It should not be presented as a finished explanation for every association between ultraprocessed-food intake and health.
For readers, this is a useful distinction between asking whether something matters and asking exactly how it matters. A study can provide an answer to the first question while leaving several credible explanations for the second.
Put the news beside the wider evidence
This work joins an existing research programme, rather than creating the subject from scratch. A separate 2019 NIH-led inpatient feeding trial compared ultraprocessed and unprocessed dietary patterns and found differences in intake and weight over a short period. That older study asked a different question under different conditions.
The two studies should not be collapsed into one claim. Repeatedly citing different outcomes from different experiments can make a conclusion sound broader than any individual design supports. The October study contributes a more controlled meal-level investigation, not a new demonstration of every long-term outcome discussed in food headlines.
What would make the next study more informative?
We will look for replication, more varied meal comparisons and populations, and studies connecting acute responses with outcomes that matter over time. It would also help to learn which specific processing features explain a response, rather than treating a broad classification as a single ingredient.
Those are research questions, not a requirement to wait for perfect certainty before making ordinary food choices. People still need meals that fit their budget, access, preferences and health needs. A dramatic headline should not turn eating into a purity contest or replace individual clinical nutrition advice.
The measured advance is narrower and more interesting: a controlled human experiment found that closely matched nutrient profiles did not eliminate differences in metabolic and neural responses. Understanding which food characteristics explain those differences is the next part of the story.