Pancreatitis is inflammation of the pancreas, the gland behind the stomach that makes digestive enzymes and insulin. Ozempic, a brand of semaglutide, lists it as a warning. How the drug works is covered in the semaglutide guide.
This page sets out what the label says, how often pancreatitis appeared in the trials, and what the larger studies found. The full list of side effects is in semaglutide side effects.
What the Ozempic label says about pancreatitis
Section 5.2 of the label states that “acute pancreatitis, including fatal and non-fatal hemorrhagic or necrotizing pancreatitis, has been observed” in people treated with GLP-1 receptor agonists, including Ozempic [1].
Hemorrhagic and necrotizing pancreatitis are the severe forms, where the gland bleeds or its tissue dies. The label does not say how many of the observed cases were of each kind.
It tells prescribers to watch for symptoms after starting the drug. If pancreatitis is suspected, the label says to discontinue Ozempic and “initiate appropriate management.”
The symptoms the label lists
The label describes the signs as “persistent or severe abdominal pain (sometimes radiating to the back),” which “may or may not be accompanied by nausea or vomiting” [1].
The Medication Guide, written for patients, puts it more plainly. It says to stop Ozempic and call a healthcare provider “right away if you have severe pain in your stomach area (abdomen) that will not go away, with or without vomiting.”
How often it happened in the trials
In the Ozempic glycemic control trials, pancreatitis was confirmed by an independent committee in 7 people on Ozempic and 3 on comparator drugs [1]. That is 0.3 and 0.2 cases per 100 patient-years.
A patient-year is one person followed for one year. So the trial rate on Ozempic was roughly three cases for every 1,000 people followed for a year, in adults with type 2 diabetes.
The Wegovy label reports the weight-loss trials, which used a higher dose of the same molecule. Pancreatitis was confirmed in 4 people on Wegovy (0.2 per 100 patient-years) and 1 on placebo (under 0.1) [2].
| Label | Trials | On semaglutide | On comparator |
|---|---|---|---|
| Ozempic | Glycemic control trials, type 2 diabetes | 7 people; 0.3 per 100 patient-years | 3 people; 0.2 per 100 patient-years |
| Wegovy | Weight-reduction trials, adults | 4 people; 0.2 per 100 patient-years | 1 person; under 0.1 per 100 patient-years |
Raised pancreatic enzymes
Amylase and lipase are pancreatic enzymes, and doctors measure them when pancreatitis is suspected. Both rose on semaglutide in the trials.
On Ozempic, the mean rise from baseline was 13% for amylase and 22% for lipase [1]. In the Wegovy weight trials, amylase rose 15% to 16% and lipase 39% [2]. Neither label saw these changes on placebo.
A raised enzyme level is not the same as pancreatitis. The labels report the rises as a lab finding, separate from the adjudicated cases above.
What pooled trial data show
A 2024 meta-analysis pooled 21 randomized, placebo-controlled semaglutide trials with 34,721 patients [3]. It found no increased risk of acute pancreatitis (odds ratio 0.7, 95% CI 0.5 to 1.2).
The result held for oral semaglutide, low injected doses and high injected doses when each was analyzed alone. An odds ratio below 1 with an interval crossing 1 means no difference was detected either way.
An earlier meta-analysis covered the whole drug class in 43 trials of at least 52 weeks, all in type 2 diabetes [4]. It found “no clear evidence of risk” for pancreatitis (odds ratio 1.24, 95% CI 0.94 to 1.64).
The same paper found the data on pancreatic cancer too scarce to draw any conclusion. Neither analysis was built to detect a very rare event, because trials rarely run long enough.
What the real-world studies found
A 2023 research letter used US insurance claims for people prescribed a GLP-1 drug for weight loss, without diabetes [5]. It compared 613 semaglutide and 4,144 liraglutide users with 654 users of bupropion-naltrexone, a weight-loss pill that works differently.
Pancreatitis ran at 4.6 cases per 1,000 person-years on semaglutide, 7.9 on liraglutide and 1.0 on bupropion-naltrexone. The adjusted hazard ratio for the GLP-1 drugs was 9.09.
That figure needs its limits beside it. Its 95% confidence interval ran from 1.25 to 66.00, and the comparator group had one pancreatitis case in total. The semaglutide group had two.
A 2025 study took the opposite approach, with over 1.2 million people with type 2 diabetes in each comparison [6]. It drew on Medicare and two commercial claims databases from 2014 to 2021.
People starting a GLP-1 drug had the same risk of hospitalization for pancreatitis as those starting an SGLT2 inhibitor, another diabetes class (hazard ratio 1.01, 95% CI 0.90 to 1.13).
| Study | Design | Who | Finding |
|---|---|---|---|
| Masson 2024 | Meta-analysis of 21 placebo-controlled trials | 34,721 patients on semaglutide or placebo | OR 0.7 (0.5 to 1.2); no increase |
| Nreu 2023 | Meta-analysis of 43 trials of 52 weeks or more | Type 2 diabetes, any GLP-1 drug | OR 1.24 (0.94 to 1.64); no clear risk |
| Sodhi 2023 | US claims cohort, weight loss | 613 semaglutide, 4,144 liraglutide, 654 comparator | HR 9.09 (1.25 to 66.00); 1 comparator case |
| Fang 2025 | US claims cohorts, type 2 diabetes | Over 1.2 million per comparison | HR 1.01 (0.90 to 1.13) vs SGLT2 inhibitors |
The gallstone connection
Gallstones are a common cause of pancreatitis, because a stone can block the duct the pancreas drains through. The 2023 letter counted gallstone pancreatitis inside its pancreatitis outcome [5].
The 2025 cohort found a modest rise in gallbladder and bile duct events on GLP-1 drugs versus SGLT2 inhibitors (hazard ratio 1.15), under one extra event per 1,000 person-years [6]. The gallbladder side is covered in Ozempic and gallbladder problems.
Heavy drinking is another common cause of pancreatitis. What is known about combining the two is in Ozempic and alcohol.
Reading the evidence together
The randomized trials, pooled, have not shown a clear increase in pancreatitis on semaglutide. The label still carries the warning because cases, some of them fatal, have been observed in people on these drugs.
The observational studies disagree, and they measured different people. The study that found a raised risk was small and had very few events; the one that did not was large and limited to diabetes.