A probiotic for psoriatic arthritis, tested against a dummy powder for 12 weeks. More people got better on the dummy

Psoriatic arthritis is the joint disease that comes with psoriasis: swollen fingers and toes, sore tendons, stiff backs and, often, the skin plaques too. Rheumatologists have grown interested in the gut, because about one in five people with the condition has quiet, symptom-free bowel inflammation and many have an unusual mix of gut bacteria. Many patients have drawn their own conclusion and take a probiotic. At the Medical University of Graz in Austria, a team noticed that a lot of their patients were using the same over-the-counter powder of Lactobacillus and Bifidobacterium strains, and an open-label pilot they ran in 2020 suggested it helped. So they tested it properly.

The trial, published online in Rheumatology on 5 October 2026, was double-blind: neither patients nor doctors knew who was getting what. Between November 2020 and September 2022, 66 adults with moderately active disease, all on a stable treatment for at least 12 weeks, were randomised to the probiotic or a placebo powder that looked, tasted and smelled the same. Each took a dose stirred into a glass of water once a day on an empty stomach for 12 weeks. Five people turned out to have been enrolled by mistake because of an error calculating their disease score, and four dropped out, which left 56 in the main analysis: 30 on the probiotic and 26 on placebo.

The goal was to bring disease activity down to low or remission on a composite score called PASDAS, without needing a change of medicine. That happened for 13 of 30 people on the probiotic (43%) and 17 of 26 on placebo (65%). The gap could be chance, but it ran the wrong way for the probiotic, and it did so in every check the authors made: counting the five wrongly enrolled people, filling in missing data statistically, or looking only at probiotic takers whose stool showed the bacteria had arrived (19 of the 23 tested). Swollen joint counts, blood inflammation, gut-leakiness markers such as zonulin, immune cells and the stool microbiome showed no difference between the groups either. Side effects were about as common on both: 19% on the probiotic and 21% on placebo.

Why did so many placebo takers improve? The authors point out that people with moderate rather than severe disease cross the line into 'low activity' more easily, and that the parts of the score that moved most were pain and patients' own rating of how they felt, which shift for many reasons besides inflammation. That also helps explain why the pilot, where everyone knew they were on the probiotic, saw 60% respond. The trial was small, at one centre, all participants were White, it tested a single product for three months, and the groups weren't perfectly matched at the start (the probiotic group was older, had had the disease longer and used more biologic drugs). The study was paid for by an Austrian public research programme; one co-author has had research grants from probiotic makers. What it shows is narrow but useful: this common kind of probiotic, taken for three months, did not calm psoriatic arthritis better than a dummy.