Stomach Pain on a GLP-1: What Is Expected and What Is a Warning Sign

Stomach pain is one of the common complaints on Ozempic, Wegovy, Mounjaro and Zepbound, and most of it is mild and arrives with the dose increases. What separates the ordinary kind from the kind that needs a phone call is not how often it happens but when it starts and what it feels like. The labels are specific about that, and they are worth reading in their own words before any percentage.

How common it is on each drug

Every label reports abdominal pain from its own placebo-controlled trials. Read each row against its own placebo figure. The rows are not a league table, for reasons the next paragraph explains.

The biggest figures are bundles. Wegovy’s 20% folds seven terms into one row, among them upper and lower abdominal pain, abdominal tenderness and epigastric discomfort. Zepbound folds five, and Foundayo says only that it includes other related terms.^[1][2][7] Counted as a single term in the pooled STEP trials, abdominal pain was 8.8% on semaglutide against 4.3% on placebo, with upper abdominal pain a separate 7.9% against 4.2%.^[8] In SURMOUNT-1, the single term on tirzepatide was 4.9% to 5.3% against 3.3% on placebo, a gap the trial did not find statistically significant, while constipation and indigestion clearly were.^[9] Three of the labels also come from diabetes trials, where the people and the doses differ.

Only one label shows the pain climbing clearly with the dose. In the Wegovy 7.2 mg trials, grouped abdominal pain was 12% on 7.2 mg, 9% on 2.4 mg and 7% on placebo. Zepbound’s figure barely moves between its lowest and highest dose.^[1][2]

When it starts, and whether it passes

The Wegovy label names abdominal pain, indigestion and bloating among the stomach reactions reported most often while the dose was being increased.^[1] The pooled STEP analysis puts shape on that. First stomach side effects bunched up in the escalation period and leveled off after week 20. Nearly all of them, 98.1%, were rated mild or moderate.^[8] Our guide to how long GLP-1 side effects last covers the escalation pattern for each drug.

How long an episode of stomach pain lasts is a question nobody has answered. The same STEP analysis gives median durations for nausea (8 days), diarrhea (3) and vomiting (2), and a much longer one for constipation (47 days, against 35 on placebo). It gives none for abdominal pain, indigestion or bloating, in the paper or its supplement.^[8]

What the trials can show is how often pain turns up once the dose ladder is finished. In STEP 4, everyone took semaglutide for 20 weeks and was then randomized to keep going or switch to placebo. New abdominal pain over the following 48 weeks was reported by 6.5% of those who stayed on the drug and 3.0% of those switched to placebo.^[8] In SURMOUNT-4, the comparable figures after tirzepatide’s escalation were 3.3% against 1.8%, and that difference was not statistically significant.^[9] So pain that appears for the first time on a settled dose is less typical of the drug, and more worth describing to a prescriber.

Is it constipation, gas, or something else?

The trials cannot tell you. Every label and every trial table lists abdominal pain, constipation, bloating and gas as separate rows, and none of the pain groupings includes constipation or gas. No source reports how often they occur together in the same person.^[1][2][8] What the data do show is that constipation behaves differently from the other stomach effects: it runs longer, although in STEP 4 it became less common over time on continued treatment.^[8] If the cramping comes with days between bowel movements, constipation on a GLP-1 covers what has and has not been shown to help. Burning high in the stomach with burping or heartburn is covered in burping, reflux and gas on a GLP-1.

A common explanation online is that food sits in a slowed stomach. The labels do say these drugs delay stomach emptying. The STEP authors are more careful: they describe the cause of the stomach side effects as uncertain, note that the longer-acting drugs do not appear to delay emptying much, and point to a 20-week study of semaglutide 2.4 mg that found no measurable delay.^[8]

The serious causes, and how strong the evidence is

Pancreatitis

Every current label carries the same picture: persistent, severe stomach pain that sometimes reaches the back, with or without vomiting, and an instruction to stop the drug if pancreatitis is suspected.^[1][2] It is rare in the trials. Confirmed cases ran at 0.2 per 100 patient-years on Wegovy against under 0.1 on placebo, and at 0.14 against 0.15 per 100 years of exposure on Zepbound.^[1][2] Two details are worth knowing. On Saxenda, cases were still being confirmed 74 and 124 days after the last dose, so stopping the drug does not close the window at once.^[6] And the pancreatic enzymes lipase and amylase rise on these drugs without any pancreatitis: lipase rose 28% to 35% on Zepbound, and the label says the rise means nothing on its own without symptoms.^[2] The full risk picture is in what the evidence says about pancreatitis and gallbladder disease. Back pain on its own, without severe stomach pain, is a different question, covered in joint pain, back pain and aches on a GLP-1.

Gallbladder disease

The Medication Guides list upper stomach pain, fever, yellowing of the skin or eyes, and clay-colored stools. They say upper stomach; none of them says the right side.^[2][1] Gallstones were reported in 1.6% of adults on Wegovy against 0.7% on placebo, and in 1.1% on Zepbound against 1%, with gallbladder inflammation 0.7% against 0.2%.^[1][2] Lilly’s labels tie these events to weight reduction, but the Wegovy and Saxenda labels say the excess remained even after the amount of weight lost was taken into account.^[2][1][6] Across 76 randomized trials, the risk ratio was 2.29 in weight-loss trials, higher than in diabetes trials, and the overall absolute increase was about 27 extra cases per 10,000 people treated for a year.^[10]

Bowel blockage

Ileus and intestinal obstruction appear on all seven labels, but only in the postmarketing list of voluntary reports, where no rate can be worked out.^[1] The study most often quoted on it, a 2023 insurance-claims analysis, found a hazard ratio of 4.22 for bowel obstruction, with a confidence interval running from 1.02 to 17.40.^[11] The semaglutide group in that study had no obstruction events at all; the signal came from liraglutide, and the result lost statistical significance when the authors excluded people with high cholesterol.^[11] A 2025 pooling of 55 placebo-controlled trials found the drugs probably have little or no effect on obstruction or ileus.^[12] In SELECT, which followed 17,604 people for up to about four and a half years, serious intestinal obstruction was reported 11 times on semaglutide and 8 times on placebo.^[13] That is reassuring, but it is not zero, and the labels still ask you to tell your prescriber about any stomach problem that is severe or will not go away.^[1]

Gastroparesis

Every label says the drugs are not recommended for people who already have severe gastroparesis, and none gives a rate for developing it.^[1] A 2025 cohort of 55,460 people with obesity found gastroparesis diagnosed at 6.5 per 1,000 person-years on semaglutide against 2.1 on bupropion-naltrexone. The diagnoses came from medical records rather than stomach-emptying tests, and the authors could not tell whether they were lasting or temporary.^[14]

One more, on one label only: appendicitis was reported in 0.5% of adults on Wegovy against 0.2% on placebo.^[1]

What helps, and how much of it has been tested

The only tools the labels give prescribers are about the dose. Wegovy’s label suggests delaying the next increase by four weeks if a dose is not tolerated; Zepbound’s suggests a lower maintenance dose.^[1][2] In the STEP trials, 12.5% of people had a dose reduced or paused for stomach side effects, and 74.0% of them finished the trial still on treatment.^[8] A slower schedule may help from the start: in a 104-person pilot trial, a gradual semaglutide titration cut stomach-related withdrawals from 19% to 2%, although it did not report on pain.^[15] That trial and the case for smaller dose steps are covered separately.

The familiar diet advice, smaller meals, stopping when full and going easy on fatty food, is not on any FDA label for these drugs. It comes from the SURMOUNT trial procedure, which told participants with stomach symptoms to eat smaller meals, split three meals into four or more, and stop when full, before trying medicines for the symptoms, a skipped dose or a lower dose.^[9] That sequence was how the trial managed people, not something it tested. An expert consensus recommends similar habits, but its sections cover nausea, vomiting, diarrhea and constipation and never mention abdominal pain.^[16] The advice is reasonable and low-risk. It is also unproven for pain.

No randomized trial has tested any treatment for stomach pain on a GLP-1. Searches of PubMed and ClinicalTrials.gov found drug trials, reviews and case reports, but nothing that tested a remedy for the pain. The trials of remedies that are registered target nausea and vomiting. The same gap exists for diarrhea, where no trial has tested a treatment either.