Nutrition labels provide calorie and macronutrient information, but processing may influence physiological responses beyond listed nutrients. Researchers at Virginia Tech investigated whether two meals with similar nutrient profiles but differing in processing level would elicit different metabolic and brain responses. The study, published in Nature Metabolism, tested whether food processing independently affects postprandial physiology and whether metabolic responses relate to brain responses to food cues.
The trial enrolled 57 healthy-weight adults aged 18–45. Of these, 32 participants completed both meal-testing sessions and were included in the within-subject metabolic comparisons; 52 participants contributed to the brain imaging analysis. Participants fasted overnight before each meal visit. The order of the two meal conditions was randomized for each participant who completed both sessions.
Each test meal provided about 300 calories. The two meals were matched for total calories, carbohydrates, fat, protein, fiber, and sodium. Examples of items used in the less-processed meal session included banana slices and cheese. Examples in the ultra-processed meal session included a portion of a peanut butter and jelly sandwich and veggie chips. After participants ate the meal, researchers collected baseline and multiple postprandial blood samples over the ensuing three hours.
In a separate visit, participants performed a bidding task: they viewed pictures of foods and placed monetary bids, which served as a behavioral measure of valuation. During that task, researchers recorded brain activity with functional MRI (fMRI) to examine neural responses to pictured ultra-processed and less-processed foods.
The investigators measured blood glucose, insulin, and energy expenditure and substrate oxidation after each meal. The study found that post-meal blood glucose responses were similar between the two meal types. However, the ultra-processed meal produced a larger insulin response, with insulin concentrations remaining higher from approximately 40 to 120 minutes after eating.
Alongside the higher insulin response, participants exhibited shifts in substrate use and energy expenditure: after the ultra-processed meal they burned fewer carbohydrates and more fat, and overall energy expenditure was higher compared with the less-processed meal condition. The authors noted these metabolic changes but did not establish that the observed insulin pattern constitutes harm or signals insulin resistance.
During the fMRI session, participants bid on pictured foods to indicate how much they valued each item. Bid amounts did not differ systematically between ultra-processed and less-processed foods. The researchers nevertheless observed that differences in carbohydrate oxidation after the meals correlated with differences in neural responses to food images. This suggests a possible relationship between how the brain processes food cues and postprandial metabolic phenotypes, although the study did not establish causation.
The primary interpretation is that processing level may influence postprandial physiology independent of matched nutrient content. Key limitations reported in the source article include:
Because of these limitations, the authors call for larger and longer trials to determine whether the observed acute differences persist with repeated exposure and whether they translate into meaningful health effects over time.
Two external experts commented on the findings. Mir Ali, MD, a bariatric surgeon and medical director not involved in the study, stated that the research demonstrated that ultra-processed foods can have notable hormonal and metabolic effects, altering insulin response and carbohydrate metabolism. He emphasized that further work is needed to determine which features of ultra-processed foods are most harmful, as the study did not separate processing methods from extra ingredients.
Qianzhi Jiang, PhD, RD, a pediatric and family dietitian, observed that the study adds evidence that processing may matter beyond label nutrients. Jiang noted that insulin release can respond to protein and amino acids and suggested that differences in ingredient composition or protein sources could have contributed to the higher insulin after the ultra-processed meal. She also highlighted the absence of measurements of incretin hormones such as GIP and GLP-1 and pointed out that the NOVA classification used to define processing was not intended as a direct metric of nutritional quality.
Both experts recommended caution in interpretation: an acute higher insulin response does not necessarily indicate long-term harm, and more research is required to identify mechanisms and health consequences. Practically, clinicians who counsel patients may continue to recommend limiting frequent intake of highly processed foods while recognizing accessibility and affordability challenges.
In a randomized, within-subject comparison of two nutrient-matched meals, the ultra-processed meal elicited a larger and longer postprandial insulin response, altered substrate oxidation toward less carbohydrate burning and more fat burning, and increased energy expenditure. Brain imaging data linked metabolic differences to neural responses to food cues, though participants’ explicit valuations of foods did not differ by processing category. The study supports the hypothesis that processing level can shape acute metabolic responses beyond macronutrient content, but it is limited by small meal size, the specific population studied, and the absence of mechanistic hormone measurements or long-term outcome data. Larger, longer, and mechanistic studies are needed to determine which processing features drive these effects and whether repeated exposure affects metabolic health over time.