Ultra-processed foods and weight loss: two healthy diets, different results

The two diets both produced weight loss

A 2025 Nature Medicine crossover trial randomized 55 adults to two eight-week diets in different orders. Both followed the UK Eatwell Guide; one emphasized minimally processed food and the other ultra-processed food. Participants could eat freely. Average weight change was −2.06% on the minimally processed diet and −1.05% on the ultra-processed diet: a difference of about one percentage point.

This was not a comparison of identical consumed calories. The diets also differed in features such as energy density. The trial therefore does not isolate an effect of processing independent of everything else. Food was provided, and the findings do not automatically predict what happens in a different shopping or household setting.

What the October 2026 experiment adds

A separate study published on 5 October 2026 examined responses to nutritionally matched ultra-processed and non-ultra-processed meals in a randomized crossover design involving 57 healthy-weight adults overall. The researchers reported differences in post-meal metabolism and brain responses, including a greater insulin response after the ultra-processed meal.

That was an acute meal experiment, not a long-term weight-loss trial. An insulin response after a meal is not itself a diagnosis, and it does not tell us how much fat someone will gain. The two studies address different questions rather than serving as interchangeable proof.

Read the whole comparison

The 2025 study is particularly useful because both diets met the same national guidance, yet the resulting eating patterns differed. But “met guidance” is not the same as “matched in every property.” A broad food category can change several features of a menu at once.

The later experiment asks a narrower physiological question. Its matching helps investigate responses to those particular meals. It does not make a single meal representative of every product in the same category.

Turn a headline into better questions

Here is an editorial shopping exercise rather than a forbidden-food list. Compare two realistic meals you could actually prepare or buy. Note the portion you would eat, the protein and fibre listed, the energy per serving, the price and the preparation time. Then ask which comparison changes if the serving size changes.

For example, a packet may list values per 30 g while your usual portion is 60 g. Doubling the portion doubles the listed energy and nutrient amounts. That is label arithmetic, not a claim about processing or a reason to attach moral value to the food.

A sustainable food choice also exists within a budget and routine. Treating the category name as the only decision rule can hide the practical comparison you are trying to make.

Keep the finding attached to the experiment

The defensible takeaway is that these diet patterns and test meals produced different measured outcomes under their study conditions. The tempting leap is that every packaged item causes the same effect. The evidence described here does not establish that.

Use our macro calculator to inspect an allocation of energy and nutrients, while remembering that arithmetic does not evaluate the complete quality, convenience or suitability of a meal.