Three trial programmes, three answers
Section 5.5 warns that acute pancreatitis, including haemorrhagic or necrotising forms in which the pancreas bleeds or its tissue dies, some fatal and some not, has been seen with GLP-1 receptor agonists and with Zepbound® [1]. Section 6.1 then reports only cases that an independent committee confirmed against set criteria. Because people stayed on treatment for different lengths of time, some results are given per 100 patient-years, in effect, cases counted against every 100 years that people spent on treatment in total [1].
| Trial pool (source) | Zepbound® or tirzepatide | Comparison |
|---|---|---|
| Weight reduction, SURMOUNT-1 and -2 (US label) [1] | 0.2% of people; 0.14 per 100 patient-years | Placebo 0.2%; 0.15 |
| Placebo-controlled weight trials (Canadian monograph) [3] | 6 events in 6 people (0.2%); 0.15 | Placebo: 3 events in 3 people (0.2%); 0.18 |
| Sleep apnea, SURMOUNT-OSA-1 and -2 [1][3] | 0.84 per 100 patient-years | Placebo 0 |
| Tirzepatide trials in type 2 diabetes [1][3] | 14 events in 13 people; 0.23 | Comparators: 3 events in 3 people; 0.11 |
How much weight the sleep apnea figure can bear
The 0.84 figure is the one people notice, and it deserves neither alarm nor dismissal. The two sleep apnea trials enrolled 467 adults with obesity, split evenly between Zepbound® and placebo, for 52 weeks [1]. The label gives the rate without a case count, and with so little treatment time a single case moves an exposure-adjusted rate a long way. The label also says reactions in those trials were otherwise similar to the weight trials [1].
The bottom row comes from tirzepatide studies in type 2 diabetes, the setting in which the molecule is sold as Mounjaro®, where tirzepatide was compared with other treatments rather than placebo alone [1][3]. In the much larger weight programme, the rate on Zepbound® matched or sat below placebo [1][3]. Read together, the rows show a rare event without a consistent excess, which is the reason the warning stays in place rather than a reason to ignore it.
When a blood test shows raised enzymes
Blood levels of amylase and lipase, two digestive enzymes from the pancreas, rise on Zepbound® even in people who feel well. In SURMOUNT-1 and -2, average amylase rose 20% to 25% from baseline and lipase 28% to 35%, against 2.1% and 5.8% on placebo [1]. Canada’s Table 3 counts the bigger jumps: readings more than triple the upper limit of normal at any visit occurred for amylase in 1.4% on Zepbound® and 0.8% on placebo, and for lipase in 1.5% and 1.0% [3]. The EU information reports average rises of 25% for amylase and 39% for lipase in the sleep apnoea trials [4].
Without other signs or symptoms, the US label calls the clinical significance of these rises unknown, and the Canadian and EU documents add that enzyme elevations on their own are not predictive of pancreatitis [1][3][4]. When pancreatitis is suspected, the NIH describes doctors combining your history, an examination of the abdomen, blood tests for amylase and lipase and imaging such as ultrasound or CT to confirm it and look for causes like gallstones [13].
Research beyond the label
Pancreatic adverse events did not rise significantly with tirzepatide in a 2025 meta-analysis that pooled 13 randomised trials and 26,894 adults with obesity and no diabetes [5]. A post hoc analysis of SURMOUNT-1 to -5, the sleep apnea trials and SUMMIT, by authors largely employed by Lilly, found no clinically meaningful difference from placebo in pancreatitis among adults aged 65 and over [6]. Trials see few cases, so both results leave room for a small risk.
Observational records add different questions. A TriNetX study of adults with type 2 diabetes found that GLP-1 medicines as a group, tirzepatide included, did not raise the risk of acute pancreatitis and were linked with fewer complications in those who developed it [7]. Another TriNetX study followed people with diabetes or obesity who had already had pancreatitis; recurrence was lower among GLP-1 users, and 6.2% among tirzepatide users [8]. Both were retrospective, so differences between the people treated may explain part of the result.
One published case describes a 59-year-old man with type 2 diabetes who developed acute pancreatitis two days after switching from semaglutide to tirzepatide; he recovered after treatment and tirzepatide was stopped [9]. A single case cannot show how often this happens. In FDA spontaneous reports from 2022 to early 2025, pancreatitis was among the tirzepatide signals more often linked with serious outcomes; such reports cannot prove cause or give a frequency [10].
Possible routes, from established to unknown
- What the product information establishes: the warning rests on cases seen with tirzepatide and its drug class, and no Zepbound® document proposes a mechanism [1][3][4]. Zepbound® separately raises gallbladder inflammation, 0.7% against 0.2% on placebo in the weight trials [1], and in the NIH’s account gallstones are the leading cause of acute pancreatitis [11].
- Proposed but not measured: whether any Zepbound® pancreatitis cases came through gallstones has not been reported. The authors of the switching case suggested that moving between GLP-1 medicines may add risk, an idea one patient cannot test [9].
- General medical knowledge: pancreatitis happens when the pancreas’s own digestive enzymes damage it. Most people with acute pancreatitis recover within several days of treatment, though a severe form can mean a long hospital stay [12].
Pancreatitis without Zepbound®
The NIH lists gallstones, heavy alcohol use, genetic disorders of the pancreas and some medicines as the most common causes, with infections, injury to the abdomen and a bile-duct procedure called ERCP among others [11]. People are also more likely to get pancreatitis if they are men, African American, have a personal or family history of pancreatitis or gallstones, have diabetes, high triglycerides or obesity, drink heavily or smoke [12].
Several of those factors are common in people prescribed Zepbound®: in the US weight pool, a quarter had type 2 diabetes and more than a third had abnormal blood fats [1]. That is one reason trials compare against placebo, and one reason a single case in a person on the medicine cannot be blamed on it without tests.
Before starting, while on treatment and afterwards
- Before the first injection, mention any pancreas trouble you have now or have had before; the Medication Guide and Canada’s patient information both ask for this [2][3].
- Canada and the EU note that tirzepatide was not studied in people with past pancreatitis and should be used with caution in them [3][4].
- If pain is severe and persists, the Medication Guide says to stop Zepbound® and call your provider right away [2]. Canada’s patient information asks you to speak to your healthcare professional immediately about severe stomach or back pain that does not go away [3].
- If pancreatitis is confirmed, Canadian and EU guidance is not to restart tirzepatide; the US label says to stop it when pancreatitis is suspected [1][3][4].
- After an episode, the NIH describes a healthy, low-fat eating plan, plenty of fluids, limited caffeine and no alcohol at all, even when the pancreatitis was mild; check any change with your care team [14].
Zepbound® and Mounjaro® are registered trademarks of Eli Lilly and Company. PepLabo is not affiliated with or endorsed by Eli Lilly and Company.
When to get help
The signs and contact information below come from this article’s sourced guidance.
Emergency
- Intense belly pain that keeps going, whether or not you feel sick or vomit, at times reaching round to your back
Contact: Stop Zepbound® and call your healthcare provider right away
Medication Guide [2]
- Worsening or severe belly pain or tenderness together with breathlessness, a racing heart, yellow skin or eyes, or fever or chills
Contact: Seek care right away; emergency services if you are very unwell
NIH [11]