Two Meals, Nearly Identical Labels—and Different Insulin Responses

Here’s a puzzle. Build two meals with the same calories, nearly the same carbs, fat and protein, even closely matched water and salt. Feed one to a person on one day and the other on another day. Shouldn’t the body respond the same way?

In a new trial, some responses differed—even though the nutrition profiles were almost identical.

A shopper reading the information panel on a package in a store
Photo by Rachel Claire / Pexels

What the researchers actually did

The study, published in Nature Metabolism by researchers at Virginia Tech’s Fralin Biomedical Research Institute, used a randomized crossover design. In its metabolic testing, the same participants ate both meals on separate days, allowing researchers to compare each person with themselves.

  • The full study included 57 healthy-weight adults ages 18 to 45. Thirty-two participated in both metabolic test sessions; 52 contributed usable brain-imaging data. These were different parts of the research, not simply a case of 25 people failing to finish both meals.
  • Each meal supplied about 300 calories and was closely matched for calories, macronutrients and several other factors. Across the full set of measures, the reported differences were no more than 1.6%.
  • The ultra-processed meal included a bite of peanut butter and jelly sandwich, deli turkey, veggie chips, a cookie, cereal and instant mashed potatoes.
  • The minimally processed meal included sliced banana, dried cranberries, cheese and egg.
  • Researchers collected blood samples during a three-hour period and monitored energy use in a metabolic chamber.

What they found

The ultra-processed meal produced a significantly greater overall insulin response: insulin levels ran higher from 40 to 120 minutes after eating. Energy expenditure and patterns of fuel use also differed.

The blood-sugar story is subtler, and worth telling precisely. Added up over the full three hours, total blood-glucose exposure was not significantly different between the meals (P = 0.92). But the shape of the curve differed. After the minimally processed meal, glucose rose faster early on and was higher at 20 minutes. After the ultra-processed meal, it came down more slowly — trending higher at 120 and 180 minutes and not returning to baseline the same way. Same total, different shape.

So the interesting finding isn’t simply “more sugar in the bloodstream.” It’s that similar nutrient totals can be accompanied by different insulin and metabolic responses — in this case, a markedly greater insulin response to essentially the same glucose.

One possible explanation is physical, not just chemical. Industrial processing can alter a food’s structure and how nutrients become available during digestion. But the study did not directly isolate or measure the effects of food structure; the meals also differed in ingredients, protein sources and additives. That explanation remains a promising hypothesis, not a settled cause.

There was a brain finding, too. In a separate imaging session, the researchers observed differences in how brain regions involved in food valuation responded to pictures of food. Some of those brain responses were associated with each participant’s metabolic differences. The study cannot show that one meal changed what people would choose to eat later, and participants’ reported liking and food valuations did not differ significantly between conditions.

What the study does not show

The honest limits matter, because the headline may travel further than the caveats:

  1. It tested one pair of meals, once each. A three-hour laboratory response does not establish long-term weight gain, diabetes risk or the effects of years of eating different foods.
  2. The participants were young and healthy. We don’t know whether the same results would appear in older adults or people with diabetes.
  3. The meals differed in more than processing level. Deli turkey versus cheese and egg, plus different additives and physical structures, could have contributed. Researchers couldn’t identify which characteristic drove the effects.
  4. The metabolic sample was modest. Thirty-two people completed the metabolic testing; the other enrolled participants were not all intended to undergo that demanding component. This is a limit on precision and generalizability, not evidence of unexplained dropouts.
  5. Some outcomes did not differ. Total measured glucose exposure and subjective liking were not significantly different, which is important context for the findings that did change.
Photo by Kampus Production via Pexels

How to use this

  1. Read beyond the label. Nutrition Facts are useful, but the ingredient list, fiber content and overall food pattern can add context that calories and macronutrient totals don’t capture.
  2. Build more meals from recognizable ingredients. Banana, cheese and egg aren’t exotic. Neither are beans, vegetables, oats or other everyday foods you can combine into satisfying meals.
  3. Don’t turn “processed” into a blanket warning. Plain yogurt, frozen vegetables and canned beans can be practical, nutritious choices. This trial tested a specific set of foods; it did not prove that every processed product causes the same response.
  4. Rebalance, rather than demand perfection. Ultra-processed foods supply more than half of the average American’s calories, so shifting the proportion of your diet toward less ultra-processed foods is a reasonable food-quality goal — but this experiment alone can’t tell us exactly how much difference a particular swap will make to long-term health.

Same nutrients on paper, different responses in the laboratory. The nutrition label remains useful—but it isn’t the whole story.

We break down new health research like this most weeks — honestly, caveats included. Join the Project Life Wellness newsletter and we’ll send it straight to you.

Cultivate a Life of Balance!

Sources

  1. Hutelin Z, et al. (Oct. 5, 2026). “Metabolic and neural responses to ultraprocessed foods: a randomized, controlled, crossover study.” Nature Metabolism 8:2152–2164. https://doi.org/10.1038/s42255-026-01619-4
  2. Virginia Tech (Oct. 5, 2026). “Food processing influences metabolism and brain activity.” https://news.vt.edu/articles/2026/10/research_fralinbiomed_upfhutelin.html

This article summarizes published research for general information and is not medical advice.