Depressed patients who cut ultra-processed food reported notable symptom improvement

Twenty adults with depression spent four weeks eating almost no ultra-processed food and four weeks eating their usual diet, and on three separate depression measures they scored better in the low-processed weeks. Psychiatrist Andrew Krystal’s group at the University of California, San Francisco, with Nyasha Chagwedera as corresponding author, reported the result in JAMA Psychiatry on September 23, 2026.

The participants cut their ultra-processed intake by more than 85 percent. The study was built to test whether such a trial could be run at all, and its authors describe it that way: a pilot, with a larger trial pencilled in for later. A pilot of this kind cannot say whether eating differently helps people with depression in general, and nothing in it amounts to advice for patients about their care or treatment.

A pilot crossover design with 20 participants

According to UCSF’s own account of the work, the trial enrolled 20 participants with an average age of 44, 70 percent of them women. All had depression, an elevated body mass index and at least one metabolic problem such as high blood pressure, diabetes or raised LDL cholesterol. The News-Medical write-up adds that 14 of the 20 also had anxiety.

In a crossover trial, each person serves as their own comparison. One group followed their regular diet for four weeks while the other moved to a low ultra-processed diet, and then the groups swapped, for eight weeks in total. The paper is titled “Reducing Ultraprocessed Foods From the Diets of Patients With Depression: A Pilot Crossover Trial”. The publisher’s page blocked automated access when this article was checked, so the design details here come from UCSF and secondary coverage of the paper.

Three scales behind the “notable symptom improvement”

UCSF says participants showed “moderate to large improvements across three measures of depression,” with a more modest improvement in anxiety scores. Newsweek’s account names the three tools: the Montgomery-Asberg Depression Rating Scale, which is rated by a clinician, and the Patient Health Questionnaire-9 and the Quick Inventory of Depressive Symptomatology, which participants fill in themselves. So the improvement is partly what patients reported and partly what an interviewer scored.

The release does not give scale-level scores, and the figures circulating in coverage differ from outlet to outlet, so none are repeated here as the paper’s own. What can be said is the direction and the size category the authors chose.

A clinical summary of the trial describes it as open label, meaning participants knew which diet they were on, and notes that depression scores also fell during the weeks on the usual diet. Diet cannot be blinded the way a pill can, so some of the gain may reflect expectation, the attention of the study team, or the natural course of depression. A crossover design reduces some of those problems because each person is compared with themselves, but it does not remove them, and with 20 people a single unusual responder can move a group average.

A proposed gut mechanism and a trial still to come

Krystal’s group proposes a biological route. As UCSF puts it, the researchers think ultra-processed food damages the gut’s protective lining, producing “low-grade, chronic inflammation throughout the body that either directly causes depression or makes people more vulnerable to it.” That is a hypothesis the authors offer to explain the result, and a 20-person pilot is not built to prove it.

The release also ties the idea to craving, quoting a UCSF researcher who says ultra-processed food triggers dopamine, which promotes craving, and that food companies work with scientists to find the amounts of sugar, salt and fat that make people want to eat more. That craving link is a second proposed pathway, and like the gut account it was not the thing the pilot set out to measure. Newsweek quotes Krystal calling the study preliminary and quoting another physician, Uma Naidoo, describing it as a first step.

Funding came from the National Institutes of Health, including the National Heart, Lung, and Blood Institute and the National Institute of Mental Health, along with UCSF Weill Institute for Neurosciences grants. The authors plan a phase 2 trial, with enrollment expected to begin in 2028, according to the ScienceDaily repost of the release.

The participants were adults who already had depression and metabolic risk factors and who changed their eating for four weeks, so the result does not speak to people without depression, to other age groups, or to a diet kept for months. The coverage read for this article does not say whether participants stayed on antidepressants or therapy during the trial, and the trial did not test diet as a replacement for either. Whether a larger, longer, better-controlled trial will show the same size of effect is what the phase 2 study is meant to settle.

This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.

More from Morning Overview