Ultraprocessed meals: what a similar glucose response cannot tell us

Text written with the help of AI.

Similar nutrient totals and blood-glucose responses do not necessarily mean two meals produce the same metabolic response. Glucose measures what is circulating in the blood; it does not show how much insulin the body released or which fuels it used after eating. The harder question is whether a difference in those measurements tells us anything useful about long-term health.

The reported meal comparison goes beyond glucose

A reported 2026 randomised crossover study compared responses to two roughly 300 kcal meals in 32 healthy-weight adults. The meals were closely matched for calories, macronutrients and fibre. Researchers measured blood responses and estimated fuel use for three hours after each meal. Hutelin and colleagues, *Nature Metabolism*, 2026.

Overall glucose exposure reportedly did not differ significantly, although glucose was higher 20 minutes after the less-processed meal. The ultraprocessed meal produced a greater insulin response and a smaller shift towards carbohydrate use. If confirmed against the paper, the result would show why a similar glucose response cannot, by itself, establish that two meals were handled in the same way.

Insulin helps regulate blood glucose after eating. A glucose reading therefore reflects a regulated concentration, not the amount of insulin needed to produce it. The CDC’s explanation of insulin provides useful background, but a larger insulin response to one meal is not a diagnosis of insulin resistance.

Matching nutrients does not isolate processing

The meal comparison cannot identify which food property produced the reported difference. Protein sources and additives differed, and food structure was not directly characterised. It would be premature to attribute the response to a particular additive or manufacturing technique.

Food structure is one plausible consideration, though evidence from another experiment cannot explain this result. In a small crossover study, meals made with disrupted chickpea cells produced higher glucose and insulin responses than nutrient-matched meals retaining intact cell clusters. That finding shows that the physical arrangement of food can affect digestion; it does not establish what caused the difference between the meals in the 2026 comparison. Cai and colleagues, *Nature Metabolism*, 2025.

A fuel-use difference is an observation, not a diagnosis

The reported smaller shift towards carbohydrate use also needs its experimental context. Estimates of carbohydrate and fat oxidation describe fuel use during the measurement period. They do not show that a person has persistently impaired metabolic flexibility, will gain body fat or faces a higher disease risk.

Glucose availability and uptake can affect that estimate. In a study comparing adults with and without type 2 diabetes, an apparent difference in metabolic flexibility disappeared after researchers accounted for the rate of glucose uptake. The study illustrates why a fuel-use measure cannot stand alone as a health score; it does not resolve the cause of the response in the meal comparison. Galgani and colleagues, *Diabetes*, 2008.

The meal study also enrolled adults aged 18 to 45, with an average age reported at about 26. Its three-hour observations cannot establish what repeated consumption does in adults over 40, let alone whether it affects longevity.

Repeated diets answer a different question

Longer feeding trials offer evidence about outcomes closer to everyday eating, although they cannot explain the acute insulin result. In a 2019 inpatient crossover trial, 20 adults ate ultraprocessed and unprocessed diets for two weeks each, with portions left to their choice. Average energy intake was reported to be about 508 kcal per day higher during the ultraprocessed phase. That is a group result under trial conditions, not a prediction for an individual meal. Hall and colleagues, *Cell Metabolism*, 2019.

A 2025 trial compared minimally processed and ultraprocessed diets that both followed the UK Eatwell Guide. Both groups lost weight over eight weeks; estimated average weight loss was greater on the minimally processed diet. The diets also differed in energy density, so the trial does not isolate an effect of additives or connect weight change to the insulin response described above. Dicken and colleagues, *Nature Medicine*, 2025.

For someone tracking glucose, even a meal ranking based only on that measure warrants care. A study of duplicate meals in adults without diabetes reported substantial variation in individual glucose-monitor responses. Repeated readings may clarify a glucose pattern, but they still do not reveal insulin release or fuel use. Hengist and colleagues, *The American Journal of Clinical Nutrition*, 2025.

The useful conclusion is limited: similar glucose responses may conceal other short-term metabolic differences, while those differences cannot establish personal disease risk. The priority before treating this particular meal comparison as a reason to change eating habits is to verify its design and findings in the primary paper.

Sources