Skip to content
This GLP
← Research
Evidence

Irritable bowel symptoms fell on the drug famous for causing them

Diarrhea, constipation, pain and bloating were all coded less often in treated IBS patients. Coded is doing a lot of work in that sentence.

Ruth Alvarez6 min read
Coded in 90 days — on a GLP-1 vs noton onenot on oneabdominal pain31.635.8constipation19.822diarrhea8.910.6bloating8.310.9Gaps of 2 to 4 points, in about 6,665 people per arm.

Nausea, diarrhea, constipation and bloating are the defining side effects of this drug class, described from the mechanism side in nausea and appetite on one circuit. So a study finding fewer of those symptoms recorded in people who already have irritable bowel syndrome is worth reading carefully [1].

The direction is consistent across every outcome measured. In the 90-day cohort, chronic diarrhea was coded in 8.9% of the treated group against 10.6% of controls, constipation in 19.8% against 22.0%, abdominal pain in 31.6% against 35.8%, and bloating in 8.3% against 10.9%. Similar patterns appeared in the diarrhea-predominant and constipation-predominant subtypes.

The authors provide a clue that points the same way. They note that differences in outcome recurrence — how many separate episodes people had — were smaller than differences in incidence. If the drug were genuinely settling irritable bowels, you would expect ongoing symptom burden to fall at least as much as first occurrences. A gap concentrated in first coding is more consistent with a change in how symptoms are attributed than in how often they happen.

The absolute differences are also modest: two to four percentage points across the outcomes. With roughly 6,665 people in each arm, small differences reach significance easily, and the P values say nothing about whether a patient would notice — the distinction drawn in the FLOW numbers needed to treat applies here too.

A real effect is nonetheless plausible. Slower gastric emptying could genuinely help diarrhea-predominant IBS, and weight loss reduces the systemic inflammation that some IBS phenotypes involve. What this design cannot do is separate that from a coding artifact, and a prospective study with symptom diaries rather than billing codes would settle it — the same gap running through the Crohn's disease cohort and the gastroparesis evidence.

Frequently asked

Do these drugs improve irritable bowel syndrome?
Coded rates of IBS symptoms were lower among treated patients, but every outcome is a billing code rather than a measured symptom, and the pattern is also consistent with symptoms being attributed to the drug instead of the IBS.
How large were the differences?
Two to four percentage points across outcomes — for example abdominal pain in 31.6% against 35.8%. With about 6,665 people per arm, small gaps reach statistical significance readily.
Is there a plausible mechanism?
Yes. Slowed gastric emptying could help diarrhea-predominant IBS, and weight loss reduces inflammation relevant to some phenotypes. This study design cannot distinguish that from a coding effect.

Sources

  1. [1] Khan SA, Marasini A, Shrestha A, Bharadwaj HR, et al. (2026). Impact of GLP-1 Analogue Therapy on Gastrointestinal Outcomes in Patients with Irritable Bowel Syndrome: A Real-World TriNetX Analysis Digestive Diseases and Sciences. PMID 42570075

Where to get it

Best GLP-1 injections

Every injectable seller we can verify, with the price each one publishes and an honest read of what the trials measured.

Compare providers →

More in Evidence