GLP-1 medicines, pancreatitis and gallbladder events in US claims data

Why this study matters

Acute pancreatitis is a sudden, painful inflammation of the pancreas that often needs a hospital stay. Biliary events include gallstones and problems with the gallbladder and bile ducts. Both have been raised as possible risks of GLP-1 medicines such as semaglutide (Wegovy, Ozempic). Earlier evidence has been mixed. This study used insurance claims from more than 1.2 million people in each comparison to look at the question in routine care. Canadian readers who take, or are considering, a GLP-1 medicine will want to know what it found.

Who was in the study

The researchers used three US claims databases covering 2014 to 2021: Medicare Fee-for-Service, which is the US public insurance program for people aged 65 and over, and two commercial insurance databases. Everyone had type 2 diabetes and no previous record of acute pancreatitis or biliary disease. Adults aged 18 and over were included in the commercial databases, and adults aged 65 and over in Medicare.

Three separate comparisons were built, each with more than 1.2 million people. The extracted data note that the three cohorts overlap and their numbers must not be added together. The summary we reviewed does not report how long people were followed.

What the researchers did

This was a cohort study using an active-comparator new-user design. New-user means everyone was followed from the day they started a medicine. Active-comparator means people starting a GLP-1 medicine were compared with people starting a different diabetes medicine at the same stage of care, rather than with people taking nothing.

The three comparisons were:

1. GLP-1 receptor agonists versus SGLT2 inhibitors (the main comparison)
2. DPP-4 inhibitors versus SGLT2 inhibitors
3. GLP-1 receptor agonists versus DPP-4 inhibitors

GLP-1 medicines and DPP-4 inhibitors both work through the incretin system, a set of gut hormones that affect insulin. SGLT2 inhibitors work in the kidney and were used as the reference group.

The researchers used propensity score fine stratification weighting, based on 92 measured characteristics, to balance the groups. They then used Cox models to estimate hazard ratios for hospitalization because of acute pancreatitis and for biliary events. A hazard ratio compares how often an event happens in one group with another over time. A value of 1 means no difference.

What they found

For acute pancreatitis, people starting a GLP-1 medicine had a similar rate to people starting an SGLT2 inhibitor (R32-C122). A confidence interval is the range in which the true value is likely to sit, and here it includes 1, so no clear difference was found.

For biliary events, both GLP-1 medicines and DPP-4 inhibitors showed a modest increase compared with SGLT2 inhibitors. The authors describe this as equivalent to fewer than one additional event per 1000 person-years. Person-years add up the time each person was followed.

When GLP-1 medicines were compared directly with DPP-4 inhibitors, there was no difference in either outcome.

Side effects and people who stopped

This study looked only at pancreatitis and biliary events. It does not report other side effects or how many people stopped their medicine.

What this study does not tell you

• This is an observational study. Differences between the groups that were not among the 92 measured characteristics could still affect the results.
• Everyone had type 2 diabetes. The findings are not a direct estimate for people using these medicines at weight-management doses without diabetes, or for any single product.
• Insurance claims record hospital admissions and diagnoses, not milder cases treated outside hospital.
• The summary we reviewed does not report how long people were followed or the actual number of events in each group.
• Our summary is based on the abstract and extracted data. Two attempts to retrieve the full paper were unsuccessful.
• The summary we reviewed does not state who funded the study.

What it means in Canada

For Canadian readers, this study adds to the evidence that acute pancreatitis was not more common after starting a GLP-1 medicine than after starting a comparison diabetes medicine in a very large population. It also confirms a small increase in gallbladder and bile duct events, in line with the randomized-trial meta-analysis covered on a separate page. For how these medicines are used in Canada, see semaglutide in Canada and tirzepatide in Canada. For coverage questions, see public drug coverage by province.

Related reading

Other studies in this library that bear on the same question.

Source