Cutting ultra-processed food by 85% moved depression scores in a 20-person crossover trial

Twenty adults with depression, an average age of 44, spent four weeks eating their usual diet and four weeks eating one that cut ultra-processed food by more than 85 percent. Their depression scores improved by moderate to large margins on three separate measures during the low-processed stretch, researchers at the University of California, San Francisco reported in JAMA Psychiatry on September 23.

Nobody involved calls it proof. Dr. Andrew Krystal, the senior author, said the pilot “successfully established the viability of carrying out this research in a larger-scale trial,” and a phase 2 trial is expected to begin enrolling in 2028.

A crossover with a small cast and a big dietary change

The design put each volunteer through both conditions, so every person served as their own comparison. Half began with the usual diet and half with the low-ultra-processed-food diet, and the groups swapped after four weeks, according to the UCSF release. Seventy percent of the participants were female, and all had depression, elevated body mass index and at least one metabolic disorder such as hypertension, diabetes or high LDL cholesterol.

The research team, led by Krystal and corresponding author Dr. Nyasha Chagwedera, did not hand out meals. Participants reported what they ate, and StudyFinds’ account of the paper says a screening score for processed-food intake fell from about 10 points to 1.4, which the authors read as confirmation that eating habits really changed. That self-reported measure is the weak link on the dietary side.

Three depression measures and one anxiety signal

The primary result is the clinical one: moderate to large improvements across three different depression measures, and a modest improvement in anxiety, the ScienceDaily summary reports. StudyFinds, which had more of the raw numbers, reports depression scores falling from about 21.6 at baseline to roughly 8 on the reduced-processed diet, against about 19 on the regular one, and says 19 of the 20 participants completed the study. Those figures come from one outlet’s reading of the paper and are not repeated in the university’s release.

The participants were not new to the illness. The same account puts the average duration of their depression at nearly 13 years, with close to two-thirds taking psychiatric medication and about a third unmedicated, which makes the size of the improvement in a short window notable but also raises the question of whether it would survive a longer one. Medication status was not the variable under test, and four weeks is long enough to register a change in mood scores but too short to say anything about relapse, weight, or whether people can keep eating this way outside a study.

The trial’s novelty is in its framing. According to the ScienceDaily summary, it is the first study to compare changes in diet within the same individuals who have depression, rather than relying on large surveys that link processed-food consumption to mood across different people. Those population surveys have long reported a correlation, but they cannot say which direction it runs, since low mood can drive people toward convenience food just as easily as convenience food might feed low mood. A within-person swap addresses that order-of-events problem even though it leaves other questions standing.

The gut hypothesis and the limits of a 20-person pilot

Chagwedera offered a mechanism, flagged as a hypothesis. Ultra-processed foods “may damage the protective lining of the gut,” she said, and that damage “can create low-grade, chronic inflammation throughout the body that either directly causes depression or makes people more vulnerable to it.” She also pointed to dopamine, saying ultra-processed food triggers the pleasure chemical in a way that promotes craving, which in her framing could make the diet harder to leave as well as harder on mood. Neither mechanism was measured in this pilot, so both remain proposals for the larger trial to test.

The trial cannot test that chain. Because it compared diets and did not isolate ingredients, StudyFinds notes that the researchers could not say whether less sugar, less fat, fewer calories or fewer additives drove the change. It was also unblinded; people know when they have swapped their groceries, and expectation alone can move a depression rating.

Scale matters for the next step. Roughly 8 percent of American adults had at least one major depressive episode in 2021, according to NIMH survey data, and a dietary approach that worked in a controlled trial would reach a very large group. UCSF’s Diet and Depression study page lists Chagwedera and Krystal as leading a continuing enrollment effort testing whether reducing ultra-processed food is feasible for patients with major depressive disorder who eat a lot of it.

News-Medical’s coverage of the release echoes the preliminary framing and the 2028 phase 2 start, and the work was supported by the National Institutes of Health and UCSF’s Weill Institute for Neurosciences, per that report. A blinded trial with measured meals would show whether a drop from about 21.6 to about 8 survives, and the phase 2 study planned for 2028 is the first that could.

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

More from Morning Overview