Ulcerative colitis patients who happened to be on a GLP-1 weight-loss drug went into symptom remission almost three times as often as matched patients who were not. At twelve weeks, 66.7 percent of the GLP-1 group had quiet disease versus 25.3 percent of the controls, and the gap held after statistical adjustment.
That comes from a retrospective cohort study ↗ published August 21 in Inflammatory Bowel Diseases. Researchers at one tertiary academic health system pulled electronic records from 2022 to 2024, found 150 adults with ulcerative colitis who started liraglutide or semaglutide for metabolic reasons and stayed on the drug at least twelve weeks, and matched each to a colitis patient who did not. Ulcerative colitis is chronic inflammation and ulceration of the colon lining. Remission here meant a partial Mayo score of 2 or less with no rectal bleeding, the standard symptom-based yardstick.
The separation showed up early and widened. At four weeks the remission split was 34.7 versus 15.3 percent, at eight weeks 54.7 versus 18.0 percent, at twelve weeks 66.7 versus 25.3 percent. Mean symptom scores in the GLP-1 group fell from 5.8 at baseline to 2.1, while the controls landed at 4.6. After adjustment, GLP-1 use carried an odds ratio of 5.90 for remission (95 percent CI 3.52 to 9.86), meaning the treated patients had roughly six times the odds of quiet disease. Semaglutide (Ozempic, Wegovy) edged out liraglutide (Victoza, Saxenda), 72.5 versus 60 percent.
The number that makes it interesting
Weight loss did not track with remission. GLP-1 drugs work partly by shrinking appetite and body weight, and excess weight is itself pro-inflammatory, so the obvious story would be that lighter patients ended up with calmer guts. The authors did not find that link. That points instead at a more direct effect. The GLP-1 receptor sits on immune cells and along the gut wall, and a growing preclinical literature argues these drugs dampen inflammatory signaling regardless of what the scale does. A designed cyclic peptide aimed at the immune checkpoint CD28 ↗ took a deliberate route to the same disease earlier this year. The GLP-1 result is the accidental version: drugs already sitting in millions of medicine cabinets appearing to move colitis on the side.
The number that makes it suspect
An odds ratio of 5.90 from a single-center chart review is a very large effect, and large effects in this kind of data are usually where confounding lives. Nobody was randomized. The GLP-1 patients were people whose doctors chose to prescribe a metabolic drug, who then filled it and stayed on it for three months, a group that skews toward better follow-up and steadier overall care. The primary endpoint is a symptom score, and symptoms are the softest thing to move. The harder read, endoscopic inflammation seen on a scope, was available only in a subset, and it moved less: 58 versus 38 percent for remission. A twenty-point gap on the objective measure is real, but it is a long way from the thirty-nine-point gap on symptoms.
So the study cannot show that GLP-1 drugs treat ulcerative colitis. It shows that in one hospital's records, colitis patients on these drugs looked better on paper, mostly by symptoms, in a way that weight loss does not explain. That is a reason to run a randomized trial of a GLP-1 as add-on therapy in colitis, especially for the many patients who also carry metabolic disease. It is not evidence that the trial has already succeeded.
The distinction matters because GLP-1 drugs are being credited with benefits across nearly every organ system right now, and most of those credits come from exactly this design: a records database, a matched comparison, an outcome that favors the drug, and a healthier, more adherent group on one side. Semaglutide and liraglutide are real molecules with real anti-inflammatory biology worth testing here. Whether they earn a colitis indication depends on a trial nobody has run yet.