GLP-1 Pancreatitis and Gallbladder Risk: What the Labels Counted

Pancreatitis and gallbladder disease sit next to each other on every GLP-1 side effect list, and they are not the same kind of problem. Gallstones and gallbladder inflammation show up more often on these drugs, and much of that appears to track the speed of the weight loss, though two of the labels say not all of it does. Pancreatitis is rarer, the trial evidence is thin in exactly the group that worries about it most, and the observational studies disagree with each other. Both produce abdominal pain that needs a clinician the same day, not a search engine. Most stomach pain on these drugs is neither, and what separates the ordinary kind from these two is covered in its own guide.

Pancreatitis: what the approved labels commit to

Every approved GLP-1 and dual-agonist label in this class carries a pancreatitis warning, and the four we read carry it in nearly identical words. Wegovy's section 5.2 names acute pancreatitis, including the hemorrhagic and necrotizing forms, and including fatal cases.^[1] Then it makes a careful choice of verb.

… observed in patients treated with …

— Wegovy prescribing information, §5.2 Acute Pancreatitis

Observed in. That is a statement about who the events happened to, and it is not a statement that the drug produced them. The Zepbound, Mounjaro and Ozempic warnings are built the same way.^[2] All four also name the same forms of the illness, fatal ones included.^[3] A label resting on a causal finding reads differently, and the same document shows you how: a few paragraphs away, the counseling section on thyroid tumors tells prescribers flatly that semaglutide causes them in rodents.^[1] That is the verb regulators use when they have the finding. They did not use it here.

The counted numbers behind the warning are small and they do not point one way. In the pooled Zepbound weight-reduction trials, adjudication-confirmed acute pancreatitis occurred in 0.2% of patients on the drug and 0.2% on placebo — 0.14 versus 0.15 events per 100 years of exposure.^[2] In the tirzepatide diabetes program the same adjudication process found 14 events in 13 treated patients, a rate of 0.23 per 100 years of exposure against 0.11 in comparator arms.^[3] Those are the numbers a regulator saw. They are compatible with a small excess and they are compatible with none.

One finding is easy to misread. Both tirzepatide labels report that treatment raises pancreatic enzymes measurably in ordinary use: amylase up by 20% to 25% and lipase by 28% to 35% in the Zepbound weight-reduction pool,^[2] and by 33% to 38% and 31% to 42% respectively in the Mounjaro trials.^[3] Both labels then state plainly that the clinical significance of such an elevation is unknown when nothing else is wrong. A raised lipase on routine bloodwork is not a diagnosis, and it is a question for the clinician who ordered the test.

The trials left out the people who most want an answer

If you have had pancreatitis before, the pivotal trial record cannot tell you what happens to you, because you would not have been enrolled in it. We pulled the registry eligibility records to check rather than assume. SURMOUNT-1, the 2,539-person tirzepatide obesity trial, lists “History of pancreatitis” flatly among its exclusion criteria.^[5][6] SURMOUNT-2 and SURMOUNT-5 exclude a history of chronic or acute pancreatitis in the same terms.^[7][8] SELECT, the large semaglutide cardiovascular outcomes trial, excludes chronic pancreatitis and any acute episode in the 180 days before screening.^[9]

This is the honest shape of the gap. Randomization gives you the cleanest possible comparison, and it gives it to you about a population that was screened to be low-risk for the exact outcome in question. Nobody hid this. It is standard trial design and it is the reason the answer for a person with prior pancreatitis has to come from somewhere else.

What the observational studies do and do not settle

Somewhere else means large database studies, which include the excluded people but cannot randomize anyone. They have produced a genuinely mixed record, and any page that reports only one of them is selecting.

The last row is the useful one for an individual reader. Whatever the population-level effect turns out to be, the people who went on to have pancreatitis were disproportionately the ones who had already had it, or who had gallstones. That is worth saying out loud to a prescriber before a first injection, and it is a question about your own history rather than about the drug.

Regulators looked hard at this class once before. In 2014 staff from the FDA and the European Medicines Agency published a joint assessment of the pancreatic safety of incretin drugs in the New England Journal of Medicine.^[15] We flag it here because it is the single most-cited item in this debate and because we want to be straight with you about it: it is a four-page Perspective piece with no abstract indexed and no open copy we could retrieve, so we are pointing you at it rather than summarizing a conclusion we did not read. It is also twelve years old, and most of the studies on this page did not exist when it was written.

Gallbladder disease: a different question, a clearer answer

The gallbladder signal is stronger, better counted, and better explained. In the adult weight-reduction trials behind the Wegovy label, gallstones were reported by 1.6% of patients on the injection against 0.7% on placebo, and gallbladder inflammation by 0.6% against 0.2%.^[1] The pooled Zepbound weight-reduction trials found gallstones in 1.1% against 1%, inflammation in 0.7% against 0.2%, and gallbladder removal in 0.2% against none.^[2] Ozempic's postmarketing section lists cholecystitis and cholecystectomy under hepatobiliary reports.^[4]

Pooling helps. A systematic review of 76 randomized trials, with 103,371 patients between them, put the relative risk on GLP-1 treatment at 1.37 for disease of the gallbladder or bile ducts, with risk rising at higher doses and longer durations.^[17] The number that matters for anyone taking these drugs to lose weight is the split inside it: the relative risk was 2.29 in the 13 trials run for weight loss and 1.27 in the 63 trials run for diabetes or other conditions, and the difference between those two was itself statistically significant.^[17] The trials with more weight loss in them had more gallbladder disease in them.

A tirzepatide-only review of nine trials lands in the same place from the other direction. It found no significant association with pancreatitis (relative risk 1.46, confidence interval 0.59 to 3.61) and a significant one with the composite of gallbladder or biliary disease (1.97, 1.14 to 3.42).^[18] Two concerns, two different verdicts, one drug.

Why the mechanism matters to you

Losing weight quickly is an established way to form gallstones, independent of any medication. Bile chemistry shifts, the gallbladder empties less often when there is less fat in the diet to prompt it, and stones form. That is why gallstones are a known complication of very-low-calorie dieting and of bariatric surgery, neither of which involves a GLP-1 at all.

So the likeliest reading is that much of this is a rapid-weight-loss risk that these drugs deliver by being effective, though not all of it: the Wegovy label says the excess gallbladder events remained even after accounting for how much weight people lost.^[1] That framing should change what you do with it, but it should not make you dismiss it. A gallstone attack is the same illness whichever route brought it on.

Two details from the label are worth having, because they answer the two questions people actually ask. The first is whether this is simply what rapid weight loss does — losing weight quickly causes gallstones on its own, so a drug that causes rapid loss would produce more of them without touching the gallbladder directly. The label addresses it: acute gallbladder disease was still more common on the drug than on placebo after the degree of weight loss was accounted for.^[1]

The second is age. In the adolescent trial the label reports cholelithiasis in 3.8% of treated patients aged 12 and over against 0% on placebo, and cholecystitis in 0.8% against 0% — more than double the adult rate, and the label says directly that these events were more common in treated adolescents than in treated adults.^[1] Anyone weighing this for a teenager is weighing a different risk profile than the adult trials describe.

The symptoms that mean today, not Monday

Both label sets translate the warning into patient-facing language, and the wording barely changes across products. For pancreatitis, Wegovy's counseling section instructs prescribers to describe severe abdominal pain that sometimes travels through to the back. Nausea and vomiting may accompany it, and may not.^[1] Mounjaro and Zepbound give prescribers the same script.^[2]

• Severe pain in the stomach area that will not go away, sometimes spreading through to the back. Vomiting may or may not be present. This is the pattern every label describes for pancreatitis. Stop the drug and seek care the same day.
• Pain in the upper stomach with fever, yellowing of the eyes or skin, or clay-colored stools. The Medication Guides list these as gallbladder signs; the combination points at the gallbladder or the bile duct and is an emergency-department problem, not a message-your-clinic problem.
• Repeated attacks of pain in the upper right abdomen, often after a heavy meal and often in the evening or at night. The NIH describes attacks that sometimes last several hours, and says more usually follow the first.^[19] Less urgent than the two above, still not something to sit on: it is worth a same-week appointment and an ultrasound.

Ordinary GLP-1 nausea and cramping are common, usually at their peak in the few weeks following each dose step, and they generally settle. We cover that arc separately in how long GLP-1 side effects last. The distinction this page cares about is different in kind, not degree: routine queasiness is uncomfortable, and the pain described above is a reason to be seen.

One thing all four labels are unanimous about, and we will not soften it: where pancreatitis is suspected, the instruction to the prescriber is to stop the drug and manage the episode.^[1] Zepbound, Mounjaro and Ozempic carry that instruction too.^[2] It is a decision for a clinician to make with you. Nothing on this page tells you to stop or continue anything.

Questions worth asking before the first injection

1. Have I ever had pancreatitis, and does my prescriber know? It was an exclusion criterion in the tirzepatide obesity trials, and it was the single strongest predictor of pancreatitis in the one study that looked for predictors.^[6][14]
2. Do I have gallstones, or have I had them? A history of gallstone disease nearly tripled the odds in that same analysis.^[14]
3. Am I taking this through a seller who will still be reachable at 11pm? A telehealth program with a message queue is not the same thing as an on-call clinician, and abdominal pain does not wait for business hours.
4. If the product is compounded, who is accountable for it? A compounded preparation has not been FDA-reviewed, so the label facts on this page describe the approved products rather than the vial in your fridge. Our page on whether compounded tirzepatide is safe sets out what that changes and what it does not.

Related reading on this register: tirzepatide side effects in compounded form, what is known about tirzepatide over the long term, and hair loss and lean mass on GLP-1 medication. Every seller we track carries a dated price on the live price tracker.