Gallbladder disease and pancreatitis appear one after the other in the warnings section of both branded labels, which makes them look like a matched pair. They are not. One has a consistent, dose-related signal with a measurable size. The other has been looked for repeatedly in randomized data and mostly has not been found. Telling them apart is what turns a scary label into a usable one.
The gallbladder signal is real, and it has a size
The largest synthesis pooled 76 randomized trials covering 103,371 patients. Assignment to a GLP-1 receptor agonist was associated with a relative risk of 1.37 (95% CI 1.23 to 1.52) for gallbladder or biliary disease, made up of cholelithiasis at 1.27, cholecystitis at 1.36 and biliary disease at 1.55.[1]
Two subgroup contrasts inside that analysis matter more than the headline. In the thirteen trials run for weight loss the relative risk was 2.29 (95% CI 1.64 to 3.18), against 1.27 in the diabetes trials. And the risk concentrated at higher doses, 1.56, while lower doses came in at 0.99.[1] Weight-loss dosing is the high end of that range, which is the relevant end for anyone reading a weight-loss price board.
A separate meta-analysis of 55 placebo-controlled trials in 106,395 participants put the same finding in the units that answer the question a reader is actually asking. Cholelithiasis rose at a risk ratio of 1.46 (95% CI 1.09 to 1.97), which the authors translate as two additional cases per 1,000 people.[2]
What that looks like on the label
The prescribing information puts the same numbers in percentages. In the adult weight-reduction trials of injected semaglutide, gallstones were reported by 1.6% of treated patients against 0.7% on placebo, and cholecystitis by 0.6% against 0.2%. The oral form ran higher, at 2.5% against 1%.
Tirzepatide reads differently. In the pooled weight-reduction trials, gallstones were reported by 1.1% of treated patients and 1% of placebo patients — effectively level. Cholecystitis was 0.7% against 0.2%, and gallbladder removal 0.2% against none. A trial-level meta-analysis of nine tirzepatide studies in 9,871 participants found the composite gallbladder or biliary outcome elevated at a risk ratio of 1.97 (95% CI 1.14 to 3.42) while none of the individual components reached significance on their own.[3]
A 2025 analysis restricted to 13 trials of semaglutide and tirzepatide in 26,894 people without diabetes found the same split. Semaglutide raised gallbladder disorders by a factor of about 2.6 (95% CI 1.40 to 4.82); tirzepatide showed no significant biliary excess.[4] That is one of the few places where the two molecules appear to differ on a safety outcome rather than an efficacy one.
Rapid weight loss is part of the mechanism, but not all of it
Losing weight quickly raises gallstone risk on its own, through bile that becomes more saturated with cholesterol and a gallbladder that empties less often. That is the obvious confounder, and the obvious objection to blaming the drug.
The semaglutide labeling addresses it directly and states that the incidence of acute gallbladder disease was greater on the drug than on placebo even after accounting for the degree of weight loss. The tirzepatide labeling, by contrast, notes that its gallbladder events were associated with weight reduction. Both statements sit in the current prescribing information, and they do not say the same thing.
Pancreatitis is the risk the trials keep not finding
This is the one people fear most and the one with the weakest evidence behind it. Across the adult weight-reduction program for injected semaglutide, acute pancreatitis was confirmed by adjudication in four treated patients, a rate of 0.2 cases per 100 patient-years, against one placebo patient at under 0.1 cases per 100 patient-years.
The tirzepatide weight-reduction pool is even flatter: 0.2% of treated patients had adjudicated acute pancreatitis against 0.2% on placebo, which works out to 0.14 cases per 100 years of exposure against 0.15. A meta-analysis of nine tirzepatide trials found a risk ratio of 1.46 with a confidence interval running from 0.59 to 3.61 — an interval that comfortably includes no effect.[3] The 55-trial synthesis reached the same conclusion for the class, finding little or no effect on pancreatitis.[2]
An umbrella review of 60 meta-analyses, covering 1,751 trials and more than 3.5 million participants, graded the gastrointestinal signals as the most credible ones in the class and classed biliary events as exploratory rather than established.[5] Pancreatitis did not emerge as a credible signal at all.
Two caveats keep this from being an all-clear. Trials exclude people with a history of pancreatitis, so the population most at risk is the one least studied. And the labeling records enzyme changes without clinical illness: mean lipase rose 39% from baseline across the weight-reduction trials, a shift not seen in the placebo group. In a separate semaglutide trial in adults with liver disease, lipase above three times the upper limit of normal occurred in 4.7% of treated patients against 1.3% on placebo. The labeling states plainly that the significance of those elevations is unknown without other signs of pancreatitis.
The symptom that is worth acting on
Both labels describe the same presentation: persistent or severe abdominal pain, sometimes radiating to the back, with or without nausea and vomiting. Both instruct discontinuing the drug if pancreatitis is suspected. Gallbladder attacks present similarly, often after a fatty meal, sometimes with fever or yellowing of the skin.
The practical distinction is between pain that escalates and pain that fades, because escalating pain is the one that does not belong to ordinary nausea — and ordinary nausea has a shape that most people recognize by the second month.
What these figures do not cover
Every rate above comes from the branded products at labeled doses. Sellers here mostly supply compounded versions, which ship without that prescribing information, and a compounded vial carries no adjudication committee behind its event counts. Whether an intake asks about prior gallbladder or pancreatic disease at all is recorded in each provider review, and the criteria behind those are in the methodology. Commercial relationships are listed in the disclosures.
None of this is medical advice, and no number on this page can tell an individual person what their own risk is.