Even with similar calories and nutrients, ultra-processed meals triggered stronger insulin responses and altered how the body burned fuel, while researchers uncovered intriguing links between metabolism and the brain's response to food.
Study: Metabolic and neural responses to ultraprocessed foods: a randomized, controlled, crossover study. Image credit: Julio Cruces Carvallo/Shutterstock.com
A recent study in the journal Nature Metabolism found that UPFs trigger different metabolic responses compared to non-UPFs of comparable calorie and macronutrient value, and are also associated with differential brain responses to food. The findings suggest that food processing may influence metabolism and brain activity through mechanisms beyond nutritional composition alone.
What is a UPF?
According to the Nova classification, UPFs are defined as “industrially manufactured products that undergo multiple physical and chemical transformations and typically contain added industrial ingredients, specifically additives not commonly used in home cooking.”
Ultra-processed foods (UPFs) are making up an increasing share of consumed calories, accounting for more than half of calories consumed in the United States. However, a higher UPF intake is linked to cardiometabolic disease and cancer risk, though this does not mean the relationship is causal.
Processing may alter nutrient delivery and food preferences
Food processing can change more than a food's nutritional composition. By altering its physical structure, processing can affect how nutrients are released, absorbed, and metabolized. For example, breaking down the cellular structure of chickpeas can lead to greater increases in blood glucose and insulin, while starch extrusion can make carbohydrates easier to digest and alter post-meal metabolic responses.
These changes may also influence how the brain learns to value food. When we eat, sensory cues such as taste and smell become associated with the nutrients delivered to the body. This process, known as flavor-nutrient learning, may involve signals traveling from the gut to the brain after eating, helping shape future food preferences.
Previous research suggests that these signals are processed partly in the striatum, a brain region involved in reward and learning. Evidence from preclinical studies indicates that the brain may respond differently to signals from different macronutrients, potentially influencing food choices. Research involving sugar-sweetened beverages has also linked nutrient intake to changes in brain responses to food cues and subsequent food liking.
Building on these findings, the researchers investigated whether UPFs produce different metabolic responses from nutritionally matched non-UPFs and whether those differences are associated with how the brain responds to food images and assigns value to food. They proposed that processing could alter nutrient availability and gut-brain signaling, potentially helping explain why UPFs are associated with overconsumption.
Researchers compare nutritionally matched meals with different processing levels
The study included 57 participants, with 52 included in the final functional magnetic resonance imaging (fMRI) analyses, and 32 completing both metabolic sessions in a randomized crossover design. Of these, 29 contributed to analyses linking metabolic responses with brain activity.
Most participants were female, with a mean age of 26 years and mean body mass index (BMI) of 22.8 kg/m2. The average proportion of energy derived from UPFs resembled the national average, at 54.4%.
A subgroup of 32 participants was asked to fast overnight, and baseline testing was done. Subsequently, participants received UPF and non-UPF meals yielding ~300 kcal each on separate days in randomized order, to be finished within ten minutes. Notably, these meals were also matched for macronutrients and glycemic index, as well as energy density, available carbohydrates, fiber, sodium, water, and other nutritional characteristics. The participants were further assessed by calorimetry and blood tests over a three-hour post-meal period.
On a separate day, participants completed an fMRI task in which they viewed pictures of 28 familiar foods, comprising 14 UPFs and 14 non-UPFs, before bidding their willingness to pay for each item using a Becker–DeGroot–Marschak (BDM) auction procedure. This was aimed at examining the link between metabolic responses and food cue reactivity by assessing associations between separately measured metabolic and neural responses.
In this study, the food pictures represented generally familiar and commonly consumed foods, increasing the likelihood that their metabolic effects had already been learned.
The participants also self-rated hunger, fullness, and thirst just before and after the fMRI.
Ultraprocessed meals trigger stronger insulin responses despite matched nutrients
Despite containing similar calories and nutrients, the two meals produced noticeably different metabolic responses. Before eating, participants showed no differences in metabolic rate or how their bodies used carbohydrates and fat for energy. They also reported similar levels of liking and wanting immediately after consuming either meal.
The first differences emerged in blood glucose responses. Although overall glucose exposure was similar, levels rose more rapidly after the non-UPF meal and were significantly higher at 20 minutes. Both meals produced peak glucose levels around 40 minutes, but glucose tended to remain elevated longer after UPF consumption. By the end of the three-hour monitoring period, glucose remained above baseline after the UPF meal, while levels following the non-UPF meal had returned closer to baseline. However, differences between meals at 120 and 180 minutes were not statistically significant.
The insulin response told a different story. Despite similar overall glucose exposure, the UPF meal triggered significantly greater insulin release. Levels were comparable during the first 20 minutes but were higher after UPF consumption at 40, 60, 90, and 120 minutes, before returning close to baseline in both groups by three hours.
Differences also emerged in how the body used energy. Although both meals initially increased metabolic rate to a peak at around 40 minutes, the increase was more sustained following the UPF meal, resulting in greater overall post-meal energy expenditure.
More notably, the meals shifted the body's fuel use in different directions. After eating, the body would normally increase its use of carbohydrates for energy. This shift was more pronounced following the non-UPF meal, whereas UPF consumption was associated with relatively greater fat oxidation and a weaker increase in carbohydrate oxidation. This difference was also reflected in the respiratory exchange ratio (RER), a measure of the body's relative use of carbohydrates and fat as fuel.
The researchers suggested that this reduced reliance on carbohydrates after UPF consumption might help explain the greater insulin response. The pattern also resembled aspects of altered fuel metabolism seen in insulin resistance and type 2 diabetes. However, insulin sensitivity was not directly measured, and the findings do not establish that eating UPFs causes either condition.
One possible explanation lies in the physical structure of processed foods. Many solid UPFs are designed to break down readily during chewing, potentially changing how nutrients are released and digested. Such changes could influence metabolic responses even when meals have similar nutritional compositions, although the study did not directly establish which processing-related characteristics were responsible.
Interestingly, participants took approximately 1.5 minutes longer to eat the UPF meal than the non-UPF meal, contrary to earlier research linking UPFs with faster eating. The researchers considered this small difference unlikely to account for the metabolic findings.
Together, the results suggest that food processing may influence how the body handles nutrients beyond differences in calories and macronutrients alone. Whether these short-term changes contribute to longer-term metabolic dysfunction, however, remains to be determined.
Brain responses to food images reflect metabolic differences
The metabolic differences between the two meals were also linked to how participants' brains responded to food images, suggesting a possible connection between the body's use of nutrients and the brain's processing of food cues.
Brain scans revealed that participants who showed greater differences in carbohydrate oxidation after the two meals also tended to respond differently to pictures of UPFs and non-UPFs. These associations emerged in the left superior temporal gyrus, right caudate, and left ventral striatum, regions involved in sensory processing, learning, and reward-related functions.
Specifically, participants whose carbohydrate oxidation increased more after the non-UPF meal relative to the UPF meal showed relatively weaker brain responses to non-UPF than UPF food images in these regions. A similar pattern emerged for the respiratory exchange ratio (RER), which reflects the body's relative use of carbohydrates and fat for energy, including associations in the left superior temporal gyrus and two sites within the right caudate.
Other measured metabolic differences were not significantly associated with brain responses, suggesting that the observed relationships were particularly linked to carbohydrate metabolism.
The findings build on earlier research involving sugar-sweetened beverages by suggesting that links between nutrient metabolism and brain responses to food cues may also extend to whole foods. However, the direction of the associations differed from some previous studies, potentially because the researchers measured brain activity while participants viewed food pictures rather than during flavor delivery.
Although hunger and thirst increased slightly and fullness declined during the scanning session, the study did not establish that the metabolic changes caused the differences in brain activity. Instead, the findings point to a possible relationship between how the body processes nutrients and how the brain responds to foods with different levels of processing, a connection that will require further investigation.
Subjective food value
Although the two types of food were associated with different brain responses, participants did not appear to value one more highly than the other.
Before their brain scans, participants rated pictures of UPFs and non-UPFs based on factors including liking, familiarity, how often they ate them, perceived healthiness, and expected fullness. They also estimated calorie and energy content and later assessed prices.
The ratings were largely similar across both food categories, although UPFs were considered less healthy and non-UPFs were reportedly consumed more frequently. When participants were asked how much they would pay for each food, up to a maximum of $5, there was no significant overall difference between UPFs and non-UPFs, even after accounting for differences in their initial ratings.
However, brain activity revealed a more complex picture. In the fusiform and lingual gyri, regions involved in visual processing, activity was positively associated with willingness to pay for non-UPFs but negatively associated with willingness to pay for UPFs. Similar opposing patterns emerged in the right putamen and caudate, brain regions involved in reward-related processing and food valuation.
These findings suggest that the brain may represent the subjective value of UPFs and non-UPFs differently, even when people express similar willingness to pay for them. However, the results do not mean that UPFs were more rewarding or desirable, and the researchers could not establish which underlying signals accounted for the differences.
The findings are consistent with the possibility that processing-related changes in nutrient availability influence how the brain evaluates food. Such mechanisms could potentially contribute to the links between UPF consumption, overeating, and metabolic health reported in earlier research. However, this study did not measure actual overconsumption or long-term health outcomes, leaving these proposed connections unconfirmed.
Ingredient differences complicate conclusions about food processing
Despite the differences observed between the two meals, the researchers cautioned that food processing itself could not be isolated as the sole explanation.
Although the meals were closely matched for calories and nutritional composition, they contained different protein sources, ingredients, and additives. These differences may have contributed to the metabolic responses. The researchers also did not directly measure the physical and chemical structure of the foods beyond their total fiber content.
The study's relatively small sample also limits how widely the findings can be applied. Participants were healthy-weight adults, and only 29 contributed to the combined metabolic and brain-imaging analysis. Whether similar responses occur in people with obesity or existing metabolic disorders remains uncertain.
Another consideration is that participants consumed relatively small meals of approximately 300 calories, which may not reflect responses to larger meals or typical eating patterns.
Crucially, the research captured only short-term responses to single meals; therefore, it cannot establish whether the observed changes in insulin release, fuel metabolism, or brain activity eventually contribute to insulin resistance, increased food intake, weight gain, or chronic disease.
Further studies will be needed to separate the effects of food processing from differences in ingredients and to determine whether these immediate physiological responses have lasting health consequences.
UPFs affect metabolism, but long-term consequences remain uncertain
The findings suggest that foods with similar calories and nutritional compositions can affect the body differently depending on how they are processed.
Compared with the non-UPF meal, the UPF meal triggered a greater insulin response, increased overall post-meal metabolic rate, and reduced the shift toward using carbohydrates for energy. These metabolic differences were also associated with distinct brain responses to pictures of UPFs and non-UPFs, suggesting a possible connection between nutrient metabolism and food cue processing.
Yet these changes did not translate into a measurable difference in willingness to pay. Instead, brain imaging revealed opposing relationships between neural activity and subjective food value for the two food categories, raising questions about how processing might influence the brain's representation of food.
The researchers propose that changes in the physical structure of foods during processing could alter nutrient availability, potentially influencing metabolic responses and food-related learning. This offers one possible explanation for previously reported associations between UPF consumption, overeating, and poor metabolic health.
However, the study does not establish that these short-term differences cause overeating, weight gain, or metabolic disease. Larger and longer-term studies are needed to identify which aspects of processing drive these responses and whether they have meaningful consequences for health.
Journal reference:
-
Hutelin, Z., Ahrens, M., Baugh, M. E., et al. (2026). Metabolic and neural responses to ultraprocessed foods: a randomized, controlled, crossover study. Nature Metabolism. DOI: https://doi.org/10.1038/s42255-026-01619-4. https://www.nature.com/articles/s42255-026-01619-4
Scientists discover a novel strategy to unmask cancer cells in high-risk neuroblastoma