Diarrhoea is common on GLP-1 medications, but a large part of what the trials recorded also happened on placebo. In the Wegovy weight-loss trials, 30% of people on the drug reported it — and so did 16% of people on dummy injections. The drug's own contribution is the gap between those two numbers, and it varies by drug, dose and who was in the trial. This page sets out the rate printed in each of nine current US labels, the placebo rate beside it, and what the trial record says about how long episodes last.
The labels were read on 9 October 2026 from DailyMed, using the manufacturers' own label versions. This page reports what the documents record; it does not advise on treatment, and anyone with persistent symptoms should speak to their prescriber.
The rate in every current label
Each label prints diarrhoea in its adverse-reaction table. The "excess" column is glp1ledger's subtraction — the drug rate minus the placebo rate in the same table — which no label prints.
| Label (effective date) | Trial population | Placebo | Drug, by dose | Excess over placebo |
|---|---|---|---|---|
| Zepbound (2026-08-28) | Adults with obesity or overweight | 8% | 19% (5 mg) · 21% (10 mg) · 23% (15 mg) | 11–15 points |
| Wegovy injection (2026-06-18) | Adults with obesity or overweight | 16% | 30% (2.4 mg) | 14 points |
| Wegovy injection, adolescents | Ages 12 and older with obesity | 19% | 22% (2.4 mg) | 3 points |
| Saxenda (2026-02-25) | Adults with obesity or overweight | 9.9% | 20.9% (3 mg) | 11 points |
| Foundayo tablets (2026-07-29) | Adults with obesity or overweight | 11% | 21% (5.5 mg) · 23% (9 mg) · 25% (17.2 mg) | 10–14 points |
| Mounjaro (2026-08-27) | Adults with type 2 diabetes | 9% | 12% (5 mg) · 13% (10 mg) · 17% (15 mg) | 3–8 points |
| Ozempic injection (2026-06-01) | Adults with type 2 diabetes | 1.9% | 8.5% (0.5 mg) · 8.8% (1 mg) | 6.6–6.9 points |
| Semaglutide tablets, Rybelsus/Ozempic (2026-01-30) | Adults with type 2 diabetes | 4% | 9% (7 mg) · 10% (14 mg) | 5–6 points |
| Victoza (2025-10-14) | Adults with type 2 diabetes | 4% | 10% (1.2 mg) · 12% (1.8 mg) | 6–8 points |
| Trulicity (2026-06-16) | Adults with type 2 diabetes | — | 7.0% (1.5 mg) · 11.4% (3 mg) · 10.7% (4.5 mg) | not printed: the table compares doses, with no placebo arm |
Three patterns stand out:
- Weight-loss trials record more diarrhoea than diabetes trials, on both arms. The weight-management doses are higher, and the placebo groups in those trials reported diarrhoea at 8% to 16%, against 1.9% to 9% in the diabetes trials. Comparing a weight-trial rate with a diabetes-trial rate compares two different studies, not two drugs.
- The dose gradient is clearest for tirzepatide and orforglipron. Zepbound, Mounjaro and Foundayo rise with each step; Ozempic's 0.5 mg and 1 mg differ by 0.3 points, and Trulicity's highest dose sits below its middle one.
- In adolescents, the excess nearly disappears. Wegovy's paediatric table shows 22% against 19% on placebo — a 3-point gap on a 133-person drug arm.
Zepbound's figure also counts "frequent bowel movements" under diarrhoea, according to the table's footnote, so its definition is slightly wider than a label that lists only the one term.
How long an episode lasts
No label prints an episode length for its own drug. The fullest figures come from a pooled analysis of the STEP 1, 2 and 3 trials of semaglutide 2.4 mg, the Wegovy dose (Wharton et al., 2022):
- diarrhoea was reported by 29.7% on semaglutide against 15.9% on placebo over 68 weeks;
- the median length of one episode was 3 days, and similar on placebo;
- across all stomach side effects on semaglutide, 98.1% were mild to moderate and 99.5% were non-serious;
- the share of people with diarrhoea fell gradually after the dose-escalation period.
The Zepbound label puts the timing the same way in words: "The majority of nausea, vomiting, and/or diarrhea events occurred during dose escalation and decreased over time." The full timeline for every common stomach symptom, including constipation's much longer episodes, is on how long GLP-1 side effects last; nausea, the most common of them, has its own page.
Is diarrhoea why people lose weight?
The STEP analysis tested it. Using mediation analysis, the authors estimated that less than one percentage point of semaglutide's weight loss beyond placebo was explained by nausea, diarrhoea or vomiting, and weight change was similar in people with and without those side effects. The paper also notes that the mechanism behind GLP-1 diarrhoea is not established: changes in how fast the lower gut moves are a proposed explanation, and the authors write that "evidence is lacking".
How many people stop because of it
Labels report discontinuations for stomach side effects as a group, not for diarrhoea alone. Zepbound's: 1.9%, 3.3% and 4.3% of people on 5, 10 and 15 mg stopped because of gastrointestinal reactions, against 0.5% on placebo. In the STEP pool, 4.3% of people on semaglutide stopped for gastrointestinal reasons.
What the labels warn about
Two warnings in the labels concern diarrhoea directly:
- Acute kidney injury from fluid loss. Every label read for this page carries it. Zepbound's says there have been postmarketing reports of acute kidney injury, "in some cases requiring hemodialysis", and that most occurred in people with gastrointestinal reactions leading to dehydration.
- Severe gastrointestinal adverse reactions. A section the Zepbound label lists among its recent major changes, dated February 2026: severe stomach reactions were reported by 1.7%, 2.5% and 3.1% of people on 5, 10 and 15 mg, against 1% on placebo.
The labels' other stomach-related warnings, on pancreatitis and gallbladder disease, are set out on GLP-1 and pancreatitis and GLP-1 and the gallbladder. The overview of every side effect in the class is on GLP-1 side effects.
What this page cannot tell you
- The excess figures are subtractions across one table, not adjusted estimates, and they carry the uncertainty of both arms.
- The rates are any report during the trial, of any severity, counted once per person; they are not the share of people with diarrhoea at a given moment.
- No trial compared two GLP-1 drugs for diarrhoea in a way that any label reports, so the table ranks nothing.
- Compounded and research products were not studied in any of these trials.
Whether symptoms on a particular medicine need a change of plan is a decision for the prescriber.
Sources
- Prescribing information read on DailyMed, 2026-10-09 (manufacturer label versions, effective dates as shown): ZEPBOUND (Eli Lilly, set ID 487cd7e7-434c-4925-99fa-aa80b1cc776b), Table 1 and sections 5.2–5.3; WEGOVY (Novo Nordisk, ee06186f-2aa3-4990-a760-757579d8f77b), Tables 3 and 5; SAXENDA (Novo Nordisk, 3946d389-0926-4f77-a708-0acb8153b143), Table 2; FOUNDAYO (Eli Lilly, 8ac446c5-feba-474f-a103-23facb9b5c62), Table 1; MOUNJARO (Eli Lilly, d2d7da5d-ad07-4228-955f-cf7e355c8cc0), Table 1; OZEMPIC injection (Novo Nordisk, adec4fd2-6858-4c99-91d4-531f5f2a2d79), Table 1; semaglutide tablets, OZEMPIC/RYBELSUS (Novo Nordisk, 27f15fac-7d98-4114-a2ec-92494a91da98), Table 2; VICTOZA (Novo Nordisk, 5a9ef4ea-c76a-4d34-a604-27c5b505f5a4), Table 1; TRULICITY (Eli Lilly, 463050bd-2b1c-40f5-b3c3-0a04bb433309), Table 2. All at dailymed.nlm.nih.gov.
- Wharton S, Calanna S, Davies M, et al., "Gastrointestinal tolerability of once-weekly semaglutide 2.4 mg in adults with overweight or obesity, and the relationship between gastrointestinal adverse events and weight loss", Diabetes, Obesity and Metabolism 2022;24:94–105, PMC9293236.