In the last years, the popularity of glucagon-like peptide 1 receptor agonists (GLP-1 RA), a drug class traditionally employed to treat type 2 diabetes, has grown exponentially, mostly around their use for weight loss, after the authorisation of some GLP-1RA as weight loss agents, and fuelled by media and celebrities attention. GLP-1 RA use has been linked to potentially severe gastrointestinal adverse effects such as pancreatitis, bowel obstruction or biliary disease, but evidence is based on studies that either are not adequately designed to capture those events or to infer causality,1–3 or in observational studies with a debatable choice of comparator groups.4,5 We evaluated the association between GLP-1 RA use and increased risk for acute pancreatitis and biliary disease using population-based real-world data from a Spanish region with 5 million inhabitants.

We included 11,508 new users of SGLT-2i and 3021 new users of GLP-1 RA in the acute pancreatitis analysis, and 10,664 and 2760 new users in the biliary disease analysis, respectively (Table 1). Mean follow-up was 482.9 days for GLP-1 RA users and 397.1 days for SGLT-2i users in per protocol analyses, and 991.8 and 969.7 in intention-to-treat analyses, respectively. Standardized mean differences in covariates between groups after PS weighting showed adequate balance after adjustment for main analyses (see Table S3 in online Supplementary Material). In adjusted, per protocol, main analyses, no risk differences were found for acute pancreatitis (hazard ratio [HR]: 0.56; 95% confidence interval [95CI%]: 0.17-1.91) nor for biliary disease (HR: 1.12; 95%IC: 0.79-1.58) (Table 2). Intention-to-treat analyses and analyses in patients with previous outcomes (including 119 and 19 patients in the SGLT-2i and GLP1 RA groups, respectively, in the pancreatitis analysis, and 963 and 280 patients, respectively, in the biliary disease analysis) also yielded non-significant results (Table 2).

In this propensity-weighted, new user, active comparator study, we did not observe increased risk of acute pancreatitis nor biliary complications in patients with diabetes and obesity initiating GLP-1 RA when compared to SGLT-2i initiators. We evaluated patients with and without previous acute pancreatitis and biliary disease, and we provided per protocol and intention-to-treat estimates, consistently finding no association between GLP1 RA and increased risk. Our results do not confirm the higher risk observed in other observational studies. However, these studies do not control for important confounders such as BMI, and comparator groups differ significantly at baseline from GLP-1 RA patients.4,5 A systematic review including 76 randomized clinical trials found increased risk of gallbladder and biliary disease associated with GLP-1 RA; however, no significant increase was observed when including only active-comparator trials.8 A potential effect of both comparator and GLP-1 RA groups in these studies (and ours) cannot therefore be ruled out.

Our study is subject to some limitations: first, real-world, daily practice observational data may be affected by many potential biases; second, despite inverse probability of treatment weighting adjustment, residual confounding is possible; third, small number of events in the acute pancreatitis analysis result in potentially imprecise estimates, a cautious interpretation is warranted; fourth, we did not include private care data, even if it is marginal in our setting; fifth, we cannot discard a potential effect of both drugs on the outcomes; sixth, even if it is highly unlikely, there is a potential use of GLP1 RA in non-diabetic patients; and finally, extrapolation to other settings should be approached with prudence. Use of GLP-1 RA is growing exponentially. Despite we did not observe an increase in gastrointestinal events associated with this drug class when compared to SGLT-2i in patients with diabetes type 2 and obesity, we cannot fully discard a risk increase. Adherence to approved indications and close monitoring of potential adverse events and of patients at a potentially higher risk is warranted to ensure a safe use of GLP-1 RA.

Data are available upon reasonable request. The datasets presented in this article are not readily available due to legal restrictions on sharing the data set as regulated by the Valencia regional government by means of legal resolution by the Valencia Health Agency which forbids the dissemination of data to third parties (accessible at: https://www.san.gva.es/es/web/investigacio/introduccion). Upon request, authors can allow access to the databases to verify the accuracy of the analysis or the reproducibility of the study. Requests to access the datasets should be directed to Management Office of the Data Commission in the Valencia Health Agency.

Pilar Pinilla Domínguez.

The corresponding author, on behalf of the other authors guarantee the accuracy, transparency and honesty of the data and information contained in the study, that no relevant information has been omitted and that all discrepancies between authors have been adequately resolved and described.

G. Sanfélix-Gimeno had full access to all the study data and assumes responsibility for the integrity of the data and the precision of the analysis. Concept and design: C. Robles Cabanillas, I. Hurtado Navarro, G. Sanfélix-Gimeno and A. García-Sempere. Data acquisition, analysis, and interpretation: all authors. Writing of the text: C. Robles Cabanillas and A. García-Sempere. Critical review of the text and important intellectual contribution: all authors. Statistical analysis: C. Robles Cabanillas and I. Hurtado Navarro. Supervision: G. Sanfélix-Gimeno, I. Hurtado Navarro and A. García-Sempere.

This research was co-funded by RD21/0016/0006 of the Carlos III Health Institute (ISCIII), Ministry of Science, Innovation and Universities, and by the Recovery, Transformation and Resilience Plan through NextGenerationEU European funds.

All authors are employees of FISABIO, a research organization of in Spain affiliated with the Department of Health of the Generalitat Valenciana. FISABIO signed a collaboration agreement (2021-25) with RTI International to conduct research on management patterns and incidence of cardiovascular and renal outcomes in patients with diabetes and kidney disease, unrelated with the present work.