Ultraprocessed and non-ultraprocessed meals can trigger different short-term metabolic responses even when they are closely matched for calories and major nutrients, according to a new controlled human study.
The experiment found that an ultraprocessed meal produced a stronger insulin response and different patterns of energy use than a nutritionally similar non-ultraprocessed meal. The findings add experimental evidence to a field in which many links between ultraprocessed foods, or UPFs, and poorer health have come from observational studies.
But the results do not show that a single ultraprocessed meal causes diabetes, obesity or another chronic disease. Instead, the researchers say the experiment offers clues about how food processing itself might influence what happens in the body after eating.
The peer-reviewed study, published in Nature Metabolism on October 5, 2026, involved 57 healthy-weight adults and examined both metabolic and neural responses to nutritionally matched foods. The findings were also reported by Newsweek.
What the researchers tested
The researchers used a randomized crossover design, meaning participants experienced both food conditions rather than being permanently assigned to only one group. That approach helps reduce the influence of differences between individuals when comparing responses.
For the metabolic portion of the experiment, 32 participants completed both meal sessions on separate days. The meals contained about 300 calories and were designed to be closely matched for characteristics including macronutrients, available carbohydrates, fiber, sodium, energy density and glycemic measures.
The distinction was that one meal consisted of ultraprocessed foods and the other of non-ultraprocessed foods.
Researchers then monitored participants for roughly three hours, collecting blood samples and measuring energy expenditure and the fuels their bodies were using through indirect calorimetry.
This design allowed the team to ask a narrower question than many population studies: if two meals look similar on conventional nutritional measures, can their degree of processing still be associated with different immediate physiological responses?
The biggest difference appeared in insulin, not total glucose exposure
The answer was more complicated than a simple difference in blood sugar.
Total glucose exposure over the measurement period, assessed using area under the curve, was not significantly different between the two meals. However, the timing of the glucose response differed. Glucose initially rose faster after the non-ultraprocessed meal, while it remained elevated later after the ultraprocessed meal.
The insulin response showed a clearer distinction. Participants produced a greater overall insulin response after the ultraprocessed meal, with higher insulin measurements during much of the post-meal period.
That distinction matters because glucose and insulin are related but are not interchangeable measurements. Insulin is a hormone that helps regulate circulating glucose and influences how nutrients are stored and used. A stronger acute insulin response is therefore a physiological observation, but it should not automatically be interpreted as evidence that participants developed insulin resistance or another metabolic disorder.
The body also used fuel differently
Processing level was associated with differences in how participants used energy after eating.
The ultraprocessed meal produced a greater post-meal metabolic-rate response but a smaller shift toward carbohydrate oxidation compared with the non-ultraprocessed meal. Carbohydrate oxidation refers to the body's use of carbohydrates as an energy source.
These measurements suggest that two meals with similar conventional nutritional profiles can be handled differently in the hours after consumption.
What the experiment does not establish is whether those short-term differences are beneficial, harmful or important enough to produce long-term health consequences. Answering that would require longer studies designed specifically around chronic outcomes.
Researchers also saw differences in food-related brain activity
The study also investigated whether metabolic responses were related to how the brain responded to food cues.
Fifty-two participants took part in the functional MRI component. The researchers reported differences in brain responses associated with food valuation, including activity in the striatum, a region involved in reward processing.
Differences between the two meal conditions in peak carbohydrate oxidation were also associated with differences in striatal activation when participants viewed food cues.
Those results require careful interpretation. An association between metabolic changes and brain activity does not establish a direct causal chain in which ultraprocessed food changes metabolism and therefore changes behavior. The experiment also did not show that UPFs “rewire” or damage the brain.
Participants' reported willingness to pay for foods did not significantly differ between conditions, further underscoring the difference between measured neural activity and demonstrated behavioral effects.
Why processing might matter even when nutrients look similar
The study raises the possibility that nutritional labels do not capture every property that determines how a food is processed by the body.
Researchers are investigating several possible explanations. Industrial processing can alter a food's physical structure, potentially affecting how quickly or where nutrients become available. Ultraprocessed products can also contain additives and other characteristics that are not reflected simply by comparing calories, carbohydrates, fat or protein.
However, this experiment did not identify one specific processing-related feature as the cause of the observed differences.
The meals could not be identical in every respect while simultaneously belonging to different processing categories. Characteristics such as food structure, ingredients and protein sources therefore remain possible contributors that future studies would need to separate experimentally.
What the study does not prove
The experiment provides stronger evidence about immediate meal responses than an observational association alone, but its scope remains limited.
It tested acute responses rather than months or years of dietary exposure. The participants were healthy-weight adults, so the findings cannot automatically be generalized to every age, body type or metabolic condition. Some analyses also involved substantially fewer participants than the study's overall enrollment.
Most importantly, measurable differences in insulin, energy expenditure, carbohydrate oxidation or brain activity are not themselves evidence that a disease will develop.
The study therefore cannot establish that the measured responses explain the associations between high UPF consumption and conditions such as obesity or type 2 diabetes reported in other research.
Why the experiment still matters
Ultraprocessed foods have been associated with numerous health outcomes in population research, but observational studies can struggle to separate food processing from other differences in diet and lifestyle.
By closely matching two meal conditions and testing the same participants under both, the new experiment narrows the question and provides evidence that processing-related characteristics may influence immediate physiology beyond calories and major macronutrients alone.
The next step is determining which characteristics are responsible and whether the short-term differences persist or matter clinically. Larger and more diverse experiments, longer controlled feeding studies and studies that isolate specific processing-related factors could help establish whether the acute effects observed here connect to meaningful long-term outcomes.
For now, the study provides a mechanistic clue rather than a final explanation: foods that look nutritionally similar on paper may not necessarily produce identical responses inside the body.
0 Comments