The UCL ultra-processed food trial

Read it as support for eating mostly minimally processed food, not as proof that processing itself is what does the harm. The difference was small, the trial was short, and nine in ten participants were women.

What the trial tested

It was a randomised crossover trial: every participant ate both diets, in random order.^[1] Fifty-five adults took part, nine in ten of them women, with an average body-mass index (BMI) of 33, which is in the obesity range. Each spent eight weeks on each diet, with a four-week break in between, and 50 provided data for the main analysis.

Both diets were built to the UK's Eatwell Guide, the government's healthy-eating model. One was made of ultra-processed foods, the other of minimally processed foods. Participants could eat as much or as little as they liked.

What it found

• Weight. The minimally processed diet produced roughly twice the weight loss: 2.06% of body weight against 1.05%, a difference of about 1 kg over eight weeks.
• Body composition. On the minimally processed diet the weight lost was largely fat, with lean muscle preserved — as measured by a bioimpedance scale (see below).
• Cravings and intake. People reported better control of cravings on the minimally processed diet. Their food diaries showed a spontaneous deficit of about 330 kcal a day, though only about two-thirds of them returned diaries.

Why it doesn't settle the question

Both diets produced weight loss. The authors say that ran counter to what they had expected from the observational studies, though the trial's registration specified only that the two diets would be compared, not which way the result would go.

Three formal critiques appeared together in the February 2026 issue of Nature Medicine. Between them they argue that the result can be explained without processing itself doing the work:

• The diets were nutritionally mismatched. The critics argue the diets differed in a way that could explain the result without invoking processing.^[2] On the trial's own figures, corrected in December 2025, the ultra-processed diet was about a quarter more energy-dense: 1.60 against 1.25 kcal per gram.^[3]
• The effect is small and the analysis questionable. A difference of about 1 kg is of limited public-health importance and does not justify processing-based dietary advice, and the handling of missing data and of carry-over between the two diet periods is questionable.^[4]
• The fat loss may be partly water. Body fat was measured with a bioimpedance scale and never with a DXA (dual-energy X-ray absorptiometry) scan, and the critics argue the body-composition signal may be an artifact of shifts in water and stored carbohydrate.^[5] The trial's own data show total body water fell about half a kilogram more on the minimally processed diet.^[6]

The critics have interests too: a co-author of the first critique declares speaking fees from Ferrero, PepsiCo, General Mills, Mondelez and the World Sugar Research Organization.

The authors replied that energy density is an intrinsic part of ultra-processing and should not be "matched away".^[7] That is a fair point about what ultra-processed food is in practice, but it means the trial cannot say whether processing matters beyond energy density.

Who ran it

Several trial authors have ties to the makers of obesity drugs: one is an Eli Lilly employee, and another has Novo Nordisk and Eli Lilly links. A co-author wrote a popular book against ultra-processed food.

Kevin Hall, who led the foundational 2019 US government metabolic-ward trial — in which people ate about 500 kcal a day more on an ultra-processed diet^[8] — is a co-author of this one. The two trials point the same way, but they are not fully independent lines of evidence.

What to do

1. Shift toward minimally processed food anyway. The trial points the same way as the other feeding trials and the large observational studies.
2. Don't expect large weight loss from the switch alone. On guideline-built diets, the gap was about 1 kg in eight weeks.
3. Watch energy density and eating speed. Those are the explanations this trial could not rule out, and they are where the other feeding trials point. The wider picture is on Ultra-processed food.

Further reading

• Dicken SJ et al. Ultraprocessed or minimally processed diets following healthy dietary guidelines on weight and cardiometabolic health: a randomized, crossover trial (with three Matters Arising critiques and reply). Nat Med 2025.^[9]
• Hall KD et al. Ultra-Processed Diets Cause Excess Calorie Intake and Weight Gain: An Inpatient Randomized Controlled Trial of Ad Libitum Food Intake. Cell Metab 2019.^[10]